Economics Assignment

Economic Development

T w e l f T h e d i T i o n

Michael P. Todaro New York University

Stephen C. Smith The George Washington University

Boston Columbus Indianapolis New York San Francisco Upper Saddle River Amsterdam Cape Town Dubai London Madrid Milan Munich Paris Montréal Toronto

Delhi Mexico City São Paulo Sydney Hong Kong Seoul Singapore Taipei Tokyo

For Donna Renée

and

For Renee, Martin, and Helena

ISBN 10: 0-13-340678-4 ISBN 13: 978-0-13-340678-8

Vice President, Editor in Chief: Donna Battista Executive Editor: David Alexander Program Manager: Lindsey Sloan Director of Marketing: Maggie Moylan Leen Marketing Manager: Lori DeShazo Team Lead, Project Management: Jeff Holcomb Senior Production Project Manager: Liz Napolitano Supplements Editor: Andra Skaalrud Manager, Rights & Permissions: Michael Joyce Rights & Permissions Coordinator: Samantha Graham Senior Procurement Specialist: Carol Melville

Cover Designer: Jonathan Boylan Cover Art: motif: Elenasz/Shutterstock; top left: Ranplett/Vetta/Getty Images; top right: Charles Harker/Shutterstock; bottom right: Dmitryh Kalinovsky/Shutterstock; bottom left: Steve Estvanik/Shutterstock. Project Coordination, Text Design, and Electronic Page Makeup: Cenveo®

Publisher Services Printer/Binder: Edwards Brothers Malloy Cover Printer: Lehigh-Phoenix Color/ Hagerstown Typeface: 10/12 Palatino

Credits and acknowledgments borrowed from other sources and reproduced, with permis- sion, in this textbook appear on the appropriate page within text.

Copyright © 2015, 2012, 2009 by Michael P. Todaro and Stephen C. Smith. All rights reserved. Manufactured in the United States of America. This publication is protected by Copyright, and permission should be obtained from the publisher prior to any prohibited reproduction, stor- age in a retrieval system, or transmission in any form or by any means, electronic, mechanical, photocopying, recording, or likewise. To obtain permission(s) to use material from this work, please submit a written request to Pearson Education, Inc., Permissions Department, One Lake Street, Upper Saddle River, New Jersey 07458, or you may fax your request to 201-236-3290.

Many of the designations by manufacturers and sellers to distinguish their products are claimed as trademarks. Where those designations appear in this book, and the publisher was aware of a trademark claim, the designations have been printed in initial caps or all caps.

Library of Congress Cataloging-in-Publication Data

Todaro, Michael P. Economic development / Michael P. Todaro, New York University, Stephen C. Smith, The George Washington University. -- Twelfth Edition. pages cm Includes bibliographical references and index. ISBN 978-0-13-340678-8 -- ISBN 0-13-340678-4 1. Economic development. 2. Developing countries--Economic policy. I. Smith, Stephen C., Date- II. Title. HD82.T552 2014 338.9009172’4--dc23 2014011530 10 9 8 7 6 5 4 3 2 1

www.pearsonhighered.com

Abel/Bernanke/ Croushore

Macroeconomics*

Bade/Parkin Foundations of Economics*

Berck/Helfand The Economics of the Environment

Bierman/Fernandez Game Theory with Economic Applications

Blanchard Macroeconomics*

Blau/Ferber/Winkler The Economics of Women, Men, and Work

Boardman/Greenberg/ Vining/Weimer

Cost-Benefit Analysis

Boyer Principles of Transportation Economics

Branson Macroeconomic Theory and Policy

Bruce Public Finance and the American Economy

Carlton/Perloff Modern Industrial Organization

Case/Fair/Oster Principles of Economics*

Chapman Environmental Economics: Theory, Application, and Policy

Cooter/Ulen Law & Economics

Daniels/VanHoose International Monetary & Financial Economics

Downs An Economic Theory of Democracy

Ehrenberg/Smith Modern Labor Economics

Farnham Economics for Managers

Folland/Goodman/ Stano

The Economics of Health and Health Care

Fort Sports Economics

Froyen Macroeconomics

Fusfeld The Age of the Economist

Gerber International Economics*

González-Rivera Forecasting for Economics and Business

Gordon Macroeconomics*

Greene Econometric Analysis

Gregory Essentials of Economics

Gregory/Stuart Russian and Soviet Economic Performance and Structure

Hartwick/Olewiler The Economics of Natural Resource Use

Heilbroner/Milberg The Making of the Economic Society

Heyne/Boettke/ Prychitko

The Economic Way of Thinking

Holt Markets, Games, and Strategic Behavior

Hubbard/O’Brien Economics*

Money, Banking, and the Financial System*

Hubbard/O’Brien/ Rafferty

Macroeconomics*

Hughes/Cain American Economic History

Husted/Melvin International Economics

Jehle/Reny Advanced Microeconomic Theory

Johnson-Lans A Health Economics Primer

Keat/Young/Erfle Managerial Economics

Klein Mathematical Methods for Economics

Krugman/Obstfeld/Melitz International Economics: Theory & Policy*

Laidler The Demand for Money

Leeds/von Allmen The Economics of Sports

Leeds/von Allmen/ Schiming

Economics*

Lynn Economic Development: Theory and Practice for a Divided World

Miller Economics Today*

Understanding Modern Economics

Miller/Benjamin The Economics of Macro Issues

Miller/Benjamin/North The Economics of Public Issues

Mills/Hamilton Urban Economics

Mishkin The Economics of Money, Bank- ing, and Financial Markets*

The Economics of Money, Banking, and Financial Mar- kets, Business School Edition*

Macroeconomics: Policy and Practice*

Murray Econometrics: A Modern Introduction

O’Sullivan/Sheffrin/ Perez

Economics: Principles, Applica- tions and Tools*

Parkin Economics*

Perloff Microeconomics*

Microeconomics: Theory and Applications with Calculus*

Perloff/Brander Managerial Economics and Strategy*

Phelps Health Economics

Pindyck/Rubinfeld Microeconomics*

Riddell/Shackelford/ Stamos/Schneider Economics: A Tool for Critically Understanding Society

Roberts The Choice: A Fable of Free Trade and Protection

Rohlf Introduction to Economic Reasoning

Roland Development Economics

Scherer Industry Structure, Strategy, and Public Policy

Schiller The Economics of Poverty and Discrimination

Sherman Market Regulation

Stock/Watson Introduction to Econometrics

Studenmund Using Econometrics: A Practi- cal Guide

Tietenberg/Lewis Environmental and Natural Resource Economics

Environmental Economics and Policy

Todaro/Smith Economic Development

Waldman/Jensen Industrial Organization: Theory and Practice

Walters/Walters/Appel/ Callahan/Centanni/ Maex/O’Neill

Econversations: Today’s Students Discuss Today’s Issues

Weil Economic Growth

Williamson Macroeconomics

The Pearson Series in Economics

*denotes MyEconLab titles. Visit www.myeconlab.com to learn more.

This page intentionally left blank

v

Part One Principles and Concepts 1

1 Introducing Economic Development: A Global Perspective 2

2 Comparative Economic Development 40

3 Classic Theories of Economic Growth and Development 118

4 Contemporary Models of Development and Underdevelopment 164

Part Two Problems and Policies: Domestic 215

5 Poverty, Inequality, and Development 216

6 Population Growth and Economic Development: Causes, Consequences, and Controversies 284

7 Urbanization and Rural-Urban Migration: Theory and Policy 330

8 Human Capital: Education and Health in Economic Development 382

9 Agricultural Transformation and Rural Development 437

10 The Environment and Development 490

11 Development Policymaking and the Roles of Market, State, and Civil Society 541

Part Three Problems and Policies: International and Macro 599

12 International Trade Theory and Development Strategy 600

13 Balance of Payments, Debt, Financial Crises, and Stabilization Policies 678

14 Foreign Finance, Investment, Aid, and Conflict: Controversies and Opportunities 731

15 Finance and Fiscal Policy for Development 781

Brief Contents

This page intentionally left blank

vii

Case Studies and Boxes xvii Preface xix

Part One Principles and Concepts 1

1 Introducing Economic Development: A Global Perspective 2 Prologue: An Extraordinary Moment 2

1.1 How the Other Half Live 4

1.2 Economics and Development Studies 9 The Nature of Development Economics 9 Why Study Development Economics? Some Critical Questions 11 The Important Role of Values in Development Economics 14 Economies as Social Systems: The Need to Go Beyond Simple Economics 15

1.3 What Do We Mean by Development? 16 Traditional Economic Measures 16 The New Economic View of Development 17 Amartya Sen’s “Capability” Approach 18 Development and Happiness 21 Three Core Values of Development 22 The Central Role of Women 24 The Three Objectives of Development 24

1.4 The Future of the Millennium Development Goals 24

1.5 Conclusions 28

■ Case Study 1: Progress in the Struggle for More Meaningful Development: Brazil 30

2 Comparative Economic Development 40 2.1 Defining the Developing World 42

2.2 Basic Indicators of Development: Real Income, Health, and Education 45 Purchasing Power Parity 45 Indicators of Health and Education 49

2.3 Holistic Measures of Living Levels and Capabilities 51 The New Human Development Index 51

2.4 Characteristics of the Developing World: Diversity within Commonality 55 Lower Levels of Living and Productivity 57 Lower Levels of Human Capital 59 Higher Levels of Inequality and Absolute Poverty 60 Higher Population Growth Rates 63

Contents

viii Contents

Greater Social Fractionalization 64 Larger Rural Populations but Rapid Rural-to-Urban Migration 65 Lower Levels of Industrialization and Manufactured Exports 66 Adverse Geography 67 Underdeveloped Markets 69 Lingering Colonial Impacts and Unequal International Relations 70

2.5 How Low-Income Countries Today Differ from Developed Countries in Their Earlier Stages 73 Physical and Human Resource Endowments 74 Relative Levels of Per Capita Income and GDP 75 Climatic Differences 75 Population Size, Distribution, and Growth 75 The Historical Role of International Migration 76 The Growth Stimulus of International Trade 78 Basic Scientific and Technological Research and Development Capabilities 79 Efficacy of Domestic Institutions 79

2.6 Are Living Standards of Developing and Developed Nations Converging? 80

2.7 Long-Run Causes of Comparative Development 85

2.8 Concluding Observations 93

■ Case Study 2: Comparative Economic Development: Pakistan and Bangladesh 96

Appendix 2.1 The Traditional Human Development Index (HDI) 112

3 Classic Theories of Economic Growth and Development 118 3.1 Classic Theories of Economic Development: Four Approaches 119

3.2 Development as Growth and the Linear-Stages Theories 119 Rostow’s Stages of Growth 120 The Harrod-Domar Growth Model 121 Obstacles and Constraints 123 Necessary versus Sufficient Conditions: Some Criticisms of the Stages Model 123

3.3 Structural-Change Models 124 The Lewis Theory of Economic Development 124 Structural Change and Patterns of Development 129 Conclusions and Implications 130

3.4 The International-Dependence Revolution 131 The Neocolonial Dependence Model 131 The False-Paradigm Model 133 The Dualistic-Development Thesis 133 Conclusions and Implications 134

3.5 The Neoclassical Counterrevolution: Market Fundamentalism 135 Challenging the Statist Model: Free Markets, Public Choice, and Market-Friendly Approaches 135 Traditional Neoclassical Growth Theory 137 Conclusions and Implications 139

3.6 Classic Theories of Development: Reconciling the Differences 140

■ Case Study 3: Schools of Thought in Context: South Korea and Argentina 142

Appendix 3.1 Components of Economic Growth 149

Appendix 3.2 The Solow Neoclassical Growth Model 155

Appendix 3.3 Endogenous Growth Theory 159

ixContents

4 Contemporary Models of Development and Underdevelopment 164 4.1 Underdevelopment as a Coordination Failure 165

4.2 Multiple Equilibria: A Diagrammatic Approach 168

4.3 Starting Economic Development: The Big Push 174 The Big Push: A Graphical Model 176 Other Cases in Which a Big Push May Be Necessary 181 Why the Problem Cannot Be Solved by a Super-Entrepreneur 182

4.4 Further Problems of Multiple Equilibria 183 Inefficient Advantages of Incumbency 183 Behavior and Norms 184 Linkages 185 Inequality, Multiple Equilibria, and Growth 186

4.5 Michael Kremer’s O-Ring Theory of Economic Development 187 The O-Ring Model 187 Implications of the O-Ring Theory 190

4.6 Economic Development as Self-Discovery 192

4.7 The Hausmann-Rodrik-Velasco Growth Diagnostics Framework 193

4.8 Conclusions 197

■ Case Study 4: Understanding a Development Miracle: China 200

Part Two Problems and Policies: Domestic 215

5 Poverty, Inequality, and Development 216 5.1 Measuring Inequality 218

Size Distributions 218 Lorenz Curves 220 Gini Coefficients and Aggregate Measures of Inequality 222 Functional Distributions 224 The Ahluwalia-Chenery Welfare Index (ACWI) 225

5.2 Measuring Absolute Poverty 226 Income Poverty 226

5.3 Poverty, Inequality, and Social Welfare 230 What’s So Bad about Extreme Inequality? 230 Dualistic Development and Shifting Lorenz Curves: Some Stylized Typologies 232 Kuznets’s Inverted-U Hypothesis 235 Growth and Inequality 239

5.4 Absolute Poverty: Extent and Magnitude 240 The Multidimensional Poverty Index (MPI) 242 Growth and Poverty 248

5.5 Economic Characteristics of High-Poverty Groups 250 Rural Poverty 250 Women and Poverty 251 Ethnic Minorities, Indigenous Populations, and Poverty 255

x Contents

5.6 Policy Options on Income Inequality and Poverty: Some Basic Considerations 256 Areas of Intervention 256 Altering the Functional Distribution of Income through Relative Factor Prices 257 Modifying the Size Distribution through Increasing Assets of the Poor 258 Progressive Income and Wealth Taxes 260 Direct Transfer Payments and the Public Provision of Goods and Services 260

5.7 Summary and Conclusions: The Need for a Package of Policies 262

■ Case Study 5: Institutions, Inequality, and Incomes: Ghana and Côte d’Ivoire 264

Appendix 5.1 Appropriate Technology and Employment Generation: The Price Incentive Model 272

Appendix 5.2 The Ahluwalia-Chenery Welfare Index 275

6 Population Growth and Economic Development: Causes, Consequences, and Controversies 284 6.1 The Basic Issue: Population Growth and the Quality of Life 284

6.2 Population Growth: Past, Present, and Future 285 World Population Growth throughout History 285 Structure of the World’s Population 287 The Hidden Momentum of Population Growth 291

6.3 The Demographic Transition 293

6.4 The Causes of High Fertility in Developing Countries: The Malthusian and Household Models 296 The Malthusian Population Trap 296 Criticisms of the Malthusian Model 301 The Microeconomic Household Theory of Fertility 303 The Demand for Children in Developing Countries 305 Implications for Development and Fertility 306

6.5 The Consequences of High Fertility: Some Conflicting Perspectives 307 It’s Not a Real Problem 308 It’s a Deliberately Contrived False Issue 309 It’s a Desirable Phenomenon 309 It Is a Real Problem 311 Goals and Objectives: Toward a Consensus 314

6.6 Some Policy Approaches 315 What Developing Countries Can Do 316 What the Developed Countries Can Do 318 How Developed Countries Can Help Developing Countries with Their Population Programs 319

■ Case Study 6: Population, Poverty, and Development: China and India 321

7 Urbanization and Rural-Urban Migration: Theory and Policy 330 7.1 Urbanization: Trends and Living Conditions 331

7.2 The Role of Cities 339 Industrial Districts 339 Efficient Urban Scale 343

7.3 The Urban Giantism Problem 344 First-City Bias 345 Causes of Urban Giantism 346

xiContents

7.4 The Urban Informal Sector 348 Policies for the Urban Informal Sector 350 Women in the Informal Sector 354

7.5 Migration and Development 355

7.6 Toward an Economic Theory of Rural-Urban Migration 357 A Verbal Description of the Todaro Model 358 A Diagrammatic Presentation 360 Five Policy Implications 362

7.7 Conclusion: A Comprehensive Urbanization, Migration, and Employment Strategy 365

■ Case Study 7: Rural-Urban Migration and Urbanization in Developing Countries: India and Botswana 369

Appendix 7.1 A Mathematical Formulation of the Todaro Migration Model 375

8 Human Capital: Education and Health in Economic Development 382 8.1 The Central Roles of Education and Health 382

Education and Health as Joint Investments for Development 384 Improving Health and Education: Why Increasing Income Is Not Sufficient 385

8.2 Investing in Education and Health: The Human Capital Approach 388

8.3 Child Labor 391

8.4 The Gender Gap: Discrimination in Education and Health 396 Education and Gender 396 Health and Gender 398 Consequences of Gender Bias in Health and Education 399

8.5 Educational Systems and Development 401 The Political Economy of Educational Supply and Demand: The Relationship between

Employment Opportunities and Educational Demands 401 Social versus Private Benefits and Costs 403 Distribution of Education 404

8.6 Health Measurement and Disease Burden 406 HIV/AIDS 412 Malaria 415 Parasitic Worms and Other “Neglected Tropical Diseases” 418

8.7 Health, Productivity, and Policy 420 Productivity 420 Health Systems Policy 422

■ Case Study 8: Pathways Out of Poverty: Progresa/Oportunidades in Mexico 425

9 Agricultural Transformation and Rural Development 437 9.1 The Imperative of Agricultural Progress and Rural Development 437

9.2 Agricultural Growth: Past Progress and Current Challenges 440 Trends in Agricultural Productivity 440 Market Failures and the Need for Government Policy 446

9.3 The Structure of Agrarian Systems in the Developing World 448 Three Systems of Agriculture 448 Traditional and Peasant Agriculture in Latin America, Asia, and Africa 449

xii Contents

Agrarian Patterns in Latin America: Progress and Remaining Poverty Challenges 451 Transforming Economies: Problems of Fragmentation and Subdivision of Peasant Land in Asia 453 Subsistence Agriculture and Extensive Cultivation in Africa 456

9.4 The Important Role of Women 458

9.5 The Microeconomics of Farmer Behavior and Agricultural Development 462 The Transition from Traditional Subsistence to Specialized Commercial Farming 462 Subsistence Farming: Risk Aversion, Uncertainty, and Survival 462 The Economics of Sharecropping and Interlocking Factor Markets 466 The Transition to Mixed or Diversified Farming 468 From Divergence to Specialization: Modern Commercial Farming 469

9.6 Core Requirements of a Strategy of Agricultural and Rural Development 471 Improving Small-Scale Agriculture 472 Institutional and Pricing Policies: Providing the Necessary Economic Incentives 473 Conditions for Rural Development 474

■ Case Study 9: The Need to Improve Agricultural Extension for Women Farmers: Kenya 477

10 The Environment and Development 490 10.1 Environment and Development: The Basic Issues 490

Economics and the Environment 490 Sustainable Development and Environmental Accounting 492 Environment Relationships to Population, Poverty, and Economic Growth 493 Environment and Rural and Urban Development 496 The Global Environment and Economy 496 Natural Resource–Based Livelihoods as a Pathway Out of Poverty: Promise and Limitations 498 The Scope of Domestic-Origin Environmental Degradation 499 Rural Development and the Environment: A Tale of Two Villages 500 Environmental Deterioration in Villages 501

10.2 Global Warming and Climate Change: Scope, Mitigation, and Adaptation 502 Scope of the Problem 502 Mitigation 505 Adaptation 506

10.3 Economic Models of Environmental Issues 508 Privately Owned Resources 508 Common Property Resources 513 Public Goods and Bads: Regional Environmental Degradation and the Free-Rider Problem 515 Limitations of the Public-Good Framework 517

10.4 Urban Development and the Environment 518 Environmental Problems of Urban Slums 518 Industrialization and Urban Air Pollution 519 Problems of Congestion, Clean Water, and Sanitation 522

10.5 The Local and Global Costs of Rain Forest Destruction 523

10.6 Policy Options in Developing and Developed Countries 526 What Developing Countries Can Do 526 How Developed Countries Can Help Developing Countries 528 What Developed Countries Can Do for the Global Environment 529

■ Case Study 10: A World of Contrasts on One Island: Haiti and the Dominican Republic 532

xiiiContents

11 Development Policymaking and the Roles of Market, State, and Civil Society 541 11.1 A Question of Balance 541

11.2 Development Planning: Concepts and Rationale 542 The Planning Mystique 542 The Nature of Development Planning 543 Planning in Mixed Developing Economies 543 The Rationale for Development Planning 544

11.3 The Development Planning Process: Some Basic Models 546 Three Stages of Planning 546 Aggregate Growth Models: Projecting Macro Variables 547 Multisector Models and Sectoral Projections 549 Project Appraisal and Social Cost-Benefit Analysis 550

11.4 Government Failure and Preferences for Markets over Planning 554 Problems of Plan Implementation and Plan Failure 554 The 1980s Policy Shift toward Free Markets 556 Government Failure 557

11.5 The Market Economy 558 Sociocultural Preconditions and Economic Requirements 558

11.6 The Washington Consensus on the Role of the State in Development and Its Subsequent Evolution 560

Toward a New Consensus 561

11.7 Development Political Economy: Theories of Policy Formulation and Reform 562 Understanding Voting Patterns on Policy Reform 564 Institutions and Path Dependency 566 Democracy versus Autocracy: Which Facilitates Faster Growth? 567

11.8 Development Roles of NGOs and the Broader Citizen Sector 569

11.9 Trends in Governance and Reform 576 Tackling the Problem of Corruption 576 Decentralization 578 Development Participation 580

■ Case Study 11: The Role of Development NGOs: BRAC and the Grameen Bank 583

Part Three Problems and Policies: International and Macro 599

12 International Trade Theory and Development Strategy 600 12.1 Economic Globalization: An Introduction 600

12.2 International Trade: Some Key Issues 603 Five Basic Questions about Trade and Development 606 Importance of Exports to Different Developing Nations 608 Demand Elasticities and Export Earnings Instability 610 The Terms of Trade and the Prebisch-Singer Hypothesis 610

12.3 The Traditional Theory of International Trade 612 Comparative Advantage 613

xiv Contents

Relative Factor Endowments and International Specialization: The Neoclassical Model 614 Trade Theory and Development: The Traditional Arguments 619

12.4 The Critique of Traditional Free-Trade Theory in the Context of Developing-Country Experience 619

Fixed Resources, Full Employment, and the International Immobility of Capital and Skilled Labor 620 Fixed, Freely Available Technology and Consumer Sovereignty 623 Internal Factor Mobility, Perfect Competition, and Uncertainty: Increasing Returns, Imperfect Competition,

and Issues in Specialization 624 The Absence of National Governments in Trading Relations 626 Balanced Trade and International Price Adjustments 627 Trade Gains Accruing to Nationals 627 Some Conclusions on Trade Theory and Economic Development Strategy 628

12.5 Traditional Trade Strategies and Policy Mechanisms for Development: Export Promotion versus Import Substitution 630

Export Promotion: Looking Outward and Seeing Trade Barriers 632 Expanding Exports of Manufactured Goods 635 Import Substitution: Looking Inward but Still Paying Outward 637 Tariffs, Infant Industries, and the Theory of Protection 637 The IS Industrialization Strategy and Results 639 Foreign-Exchange Rates, Exchange Controls, and the Devaluation Decision 644 Trade Optimists and Trade Pessimists: Summarizing the Traditional Debate 648

12.6 The Industrialization Strategy Approach to Export Policy 651 Export-Oriented Industrialization Strategy 651 The New Firm-level International Trade Research and the Developing Countries 655

12.7 South-South Trade and Economic Integration 655 Economic Integration: Theory and Practice 655 Regional Trading Blocs, the Globalization of Trade, and Prospects for South-South Cooperation 657

12.8 Trade Policies of Developed Countries: The Need for Reform and Resistance to New Protectionist Pressures 659

■ Case Study 12: A Pioneer in Development Success through Trade: Taiwan 663

13 Balance of Payments, Debt, Financial Crises, and Stabilization Policies 678 13.1 International Finance and Investment: Key Issues for Developing Countries 678

13.2 The Balance of Payments Account 679 General Considerations 679 A Hypothetical Illustration: Deficits and Debts 681

13.3 The Issue of Payments Deficits 685 Some Initial Policy Issues 685 Trends in the Balance of Payments 689

13.4 Accumulation of Debt and Emergence of the Debt Crisis in the 1980s 691 Background and Analysis 691 Origins of the 1980s Debt Crisis 693

13.5 Attempts at Alleviation: Macroeconomic Instability, Classic IMF Stabilization Policies, and Their Critics 695

The IMF Stabilization Program 695 Tactics for Debt Relief 697

xvContents

13.6 The Global Financial Crisis and the Developing Countries 706 Causes of the Crisis and Challenges to Lasting Recovery 707 Economic Impacts on Developing Countries 710 Differing Impacts and Continuing Challenges across Developing Regions 714 Prospects for Recovery and Stability 717 Opportunities as Well as Dangers? 718

■ Case Study 13: Trade, Capital Flows, and Development Strategy: South Korea 720

14 Foreign Finance, Investment, Aid, and Conflict: Controversies and Opportunities 731 14.1 The International Flow of Financial Resources 731

14.2 Private Foreign Direct Investment and the Multinational Corporation 732 Private Foreign Investment: Some Pros and Cons for Development 736 Private Portfolio Investment: Benefits and Risks 743

14.3 The Role and Growth of Remittances 744

14.4 Foreign Aid: The Development Assistance Debate 747 Conceptual and Measurement Problems 747 Amounts and Allocations: Public Aid 748 Why Donors Give Aid 750 Why Recipient Countries Accept Aid 754 The Role of Nongovernmental Organizations in Aid 755 The Effects of Aid 756

14.5 Conflict and Development 757 The Scope of Violent Conflict and Conflict Risks 757 The Consequences of Armed Conflict 758 The Causes of Armed Conflict and Risk Factors for Conflict 761 The Resolution and Prevention of Armed Conflict 763

■ Case Study 14: Costa Rica, Guatemala, and Honduras: Contrasts and Prospects for Convergence 767

15 Finance and Fiscal Policy for Development 781 15.1 The Role of the Financial System in Economic Development 782

Differences between Developed and Developing-Country Financial Systems 784

15.2 The Role of Central Banks and Alternative Arrangements 787 Functions of a Full-Fledged Central Bank 787 The Role of Development Banking 791

15.3 Informal Finance and the Rise of Microfinance 792 Traditional Informal Finance 792 Microfinance Institutions: How They Work 793 MFIs: Three Current Policy Debates 796 Potential Limitations of Microfinance as a Development Strategy 798

15.4 Formal Financial Systems and Reforms 799 Financial Liberalization, Real Interest Rates, Savings, and Investment 799 Financial Policy and the Role of the State 801 Debate on the Role of Stock Markets 803

xvi Contents

15.5 Fiscal Policy for Development 805 Macrostability and Resource Mobilization 805 Taxation: Direct and Indirect 805

15.6 State-Owned Enterprises and Privatization 810 Improving the Performance of SOEs 811 Privatization: Theory and Experience 812

15.7 Public Administration: The Scarcest Resource 815

■ Case Study 15: African Success Story at Risk: Botswana 817 Glossary 826 Name Index 839 Subject Index 851

xvii

Case Studies 1 Progress in the Struggle for More Meaningful Development: Brazil 30 2 Comparative Economic Development: Pakistan and Bangladesh 96 3 Schools of Thought in Context: South Korea and Argentina 142 4 Understanding a Development Miracle: China 200 5 Institutions, Inequality, and Incomes: Ghana and Côte d’Ivoire 264 6 Population, Poverty, and Development: China and India 321 7 Rural-Urban Migration and Urbanization in Developing Countries: India and Botswana 369 8 Pathways Out of Poverty: Progresa/Oportunidades in Mexico 425 9 The Need to Improve Agricultural Extension for Women Farmers: Kenya 477 10 A World of Contrasts on One Island: Haiti and the Dominican Republic 532 11 The Role of Development NGOs: BRAC and the Grameen Bank 583 12 A Pioneer in Development Success through Trade: Taiwan 663 13 Trade, Capital Flows, and Development Strategy: South Korea 720 14 Costa Rica, Guatemala, and Honduras: Contrasts and Prospects for Convergence 767 15 African Success Story at Risk: Botswana 817

Boxes 1.1 The Experience of Poverty: Voices of the Poor 8 2.1 Computing the New HDI: Ghana 54 2.2 What Is New in the New Human Development Index 56 2.3 FINDINGS The Persistent Effects of Colonial Forced Labor on Poverty and Development 71 2.4 FINDINGS Instruments to Test Theories of Comparative Development: Inequality 91 2.5 FINDINGS Legacy of Colonial Land Tenure and Governance Systems 92 4.1 Synchronizing Expectations: Resetting “Latin American Time” 171 4.2 FINDINGS Village Coordination and Monitoring for Better Health Outcomes 172 4.3 FINDINGS Three Country Case Study Applications of Growth Diagnostics 196 5.1 The Latin America Effect 238 5.2 Problems of Gender Relations in Developing Countries: Voices of the Poor 254 6.1 FINDINGS The 2012 Revised United Nations Population Projections 291 6.2 FINDINGS Social Norms and the Changing Patterns of Fertility in Bangladesh 300 6.3 FINDINGS Contraceptives Need and Use in Developing Countries, 2003 to 2012 317 7.1 FINDINGS The Emergence of Industrial Districts or Clusters in China 341 8.1 Health and Education: Voices of the Poor 384 8.2 Linkages between Investments in Health and Education 385 8.3 FINDINGS Mothers’ Health Knowledge Is Crucial for Raising Child Health 386 8.4 FINDINGS School Impact of a Low-Cost Health Intervention 387 8.5 FINDINGS Cash or Condition? Evidence from Malawi 395

Case Studies and

Boxes

xviii Case Studies and Boxes

8.6 FINDINGS Impacts of Tutor and Computer-Assisted Learning Programs 407 8.7 Health Challenges Faced by Developing Countries 410 8.8 AIDS: Crisis and Response in Uganda 416 9.1 Development Policy Issues: Famine in the Horn of Africa 444 9.2 FINDINGS Learning about Farming: The Diffusion of Pineapple Growing in Ghana 470 10.1 FINDINGS Autonomous Adaptation to Climate Change by Farmers in Africa 508 10.2 One of the World’s Poorest Countries Tries to Prepare for Climate Change: Niger 509 10.3 FINDINGS Elinor Ostrom’s Design Principles Derived from Studies

of Long-Enduring Institutions for Governing Sustainable Resources 516 11.1 The Washington Consensus and East Asia 560 11.2 The New Consensus 562 11.3 FINDINGS Reducing Teacher Absenteeism in an NGO School 576 12.1 FINDINGS Four Centuries of Evidence on the Prebisch-Singer Hypothesis 612 13.1 The History and Role of the International Monetary Fund 682 13.2 The History and Role of the World Bank 686 13.3 Mexico: Crisis, Debt Reduction, and the Struggle for Renewed Growth 700 13.4 “Odious Debt” and Its Prevention 704 14.1 Seven Key Disputed Issues about the Role and Impact of Multinational

Corporations in Developing Countries 740 15.1 FINDINGS The Financial Lives of the Poor 794 15.2 FINDINGS Combining Microfinance with Training 798 15.3 Privatization—What, When, and to Whom? Chile and Poland 814

xix

Preface

Economic Development, Twelfth Edition, presents the latest thinking in eco- nomic development with the clear and comprehensive approach that has been so well received in both the developed and developing worlds.

The pace and scope of economic development continues its rapid, uneven, and sometimes unexpected evolution. This text explains the unprecedented progress that has been made in many parts of the developing world but fully confronts the enormous problems and challenges that remain to be addressed in the years ahead. The text shows the wide diversity across the developing world and the differing positions in the global economy that are held by developing countries. The principles of development economics are key to understanding how we got to where we are, how great progress has been made in recent years, and why many development problems remain so difficult to solve. The princi- ples of development economics are also key to the design of successful economic development policy and programs as we look ahead.

The field of economic development is versatile and has much to contribute re- garding these differing scenarios. Thus, the text also underlines common features that are exhibited by a majority of developing nations, using the insights of the study of economic development. The few countries that have essentially com- pleted the transformation to become developed economies, such as South Korea, are also examined as potential models for other developing countries to follow.

Both theory and empirical analysis in development economics have made major strides, and the Twelfth Edition brings these ideas and findings to stu- dents. Legitimate controversies are actively debated in development econom- ics, and so the text presents contending theories and interpretations of evi- dence, with three goals. The first goal is to ensure that students understand real conditions and institutions across the developing world. The second is to help students develop analytic skills while broadening their perspectives of the wide scope of the field. The third is to provide students with the resources to draw independent conclusions as they confront development problems, their sometimes ambiguous evidence, and real-life development policy choices— ultimately, to play an informed role in the struggle for economic development and ending extreme poverty.

New to This Edition

• Global crisis. This edition includes a major update and expansion of the new section on the impacts and potential longer-term implications of the recent global financial crisis on economic development, examining

xx Preface

conditions that caused the crisis, its aftermath, and possible broader im- plications and large differences across developing nations and regions.

• Prologue in Chapter 1. Chapter 1 is launched with a new introductory sec- tion that describes for students how much has changed over the past two decades in a majority of countries in the developing world and in greater autonomy and nascent leadership of some developing countries in inter- national economic and political relationships. The chapter compares con- ditions today to those prevailing in 1992—a pivotal period in a number of ways, which is also close to the time when many students were born.

• Violent conflict. The Eleventh Edition provided an entirely new major sec- tion on the causes and consequences of violent conflict, postconflict re- covery and development, and prevention of conflict through an improved understanding of its major causes; the Twelfth Edition more fully devel- ops and extends this section, incorporating recent developments.

• Findings Boxes. The Eleventh Edition also introduced a new textbook fea- ture of Findings boxes, reporting on empirical research results in the field that are wide-ranging in both methods and topics. New Findings boxes ad- dress such topics as long-lasting impacts of colonial institutions (Peru); how coordination and monitoring by villagers leads to better health outcomes (Uganda); how social norms facilitated or constrained changing patterns of fertility (Bangladesh); and comparative impacts of conditional versus un- conditional cash transfers to the poor (Malawi). Other boxes examine global findings such as unmet contraceptives demand across countries. The num- ber of Findings boxes has been approximately doubled for the Twelfth Edi- tion. The Findings boxes also illustrate empirical methods for students—in an intuitive introductory manner—such as the use of instruments; random- ized control trials; regression discontinuity; and fixed effects; as well as the painstaking design, implementation, and robust analysis of survey data; growth diagnostics; and systematically applied qualitative research. The Findings boxes in this edition are listed on pages xvii-xviii.

• Policy Boxes. Other boxes address policy issues. New policy boxes examine such topics as the efforts of Niger—one of the world’s poorest countries— to adapt to the climate change already impacting the country and to build resilience against unknown future climate change; and what we learned from the 2011–2012 famine in the Horn of Africa. Other new policy boxes address global findings, such as the extent of contraception use and the extent of still-unmet demand for contraceptives in developing countries; and the UN’s new unexpectedly increased population projections through this century. Policy boxes in this edition are listed on pages xvii-xviii.

• New, full-length, three-way comparative case study of Costa Rica, Guate- mala, and Honduras. The full-length, end-of-chapter comparative case studies have long been one of the most popular features of the text. For this edition, an entirely new three-way comparative case study of Costa Rica, Guatemala, and Honduras is introduced at the end of Chapter 14, which addresses topics of conflict, foreign investment, remittances, and foreign aid; the study also addresses the themes of very long-term com- parative development addressed in some of the existing and updated case

xxiPreface

studies, such as those comparing Ghana and Côte d’Ivoire; Pakistan and Bangladesh; and Haiti and the Dominican Republic. Each of the compara- tive cases also has a special theme, such as human development, poverty, environment, and structural transformation.

• New topics. Other new topics briefly introduced in this edition include short sections on the new firm-level international trade research and the develop- ing countries; the emergence of “Sustainable Development Goals” as suc- cessors to the MDGs; corporate social responsibility; and food price trends.

• New measures. Measurement is an ever-present issue in the field of eco- nomic development. The United Nations Development Program released its Multidimensional Poverty Index in August 2010 and its New Human Development Index in November 2010. The text examines the index for- mulas, explains how they differ from earlier indexes, reports on findings, and reviews issues surrounding the active debate on these measures. Each has been updated since its initial release, as covered in the Twelfth Edition. Note: From surveys we know many instructors are still using the tradi- tional Human Development Index (HDI), which is reasonable, since it per- meates a majority of the literature on the subject. So, we have maintained a very substantial and detailed section on the traditional HDI, which now appears in a new Appendix 2.1 in Chapter 2; it includes a number of coun- try applications and extensions, as in previous editions. You can teach ei- ther or both of the indexes, without losing the thread in later chapters.

• Updated statistics. Change continues to be very rapid in the developing world. Throughout the text, data and statistics have been updated to re- flect the most recent available information at the time of revision, typically 2011 or 2012, and sometimes 2013.

• Additional updates. Other updates include a further expansion of the sec- tion on microfinance, including new designs, potential benefits, successes to date, and some limitations; further expanded coverage of China; and expanded coverage and analysis of the growing environmental problems facing developing countries.

Audience and Suggested Ways to Use the Text

• Flexibility. This book is designed for use in courses in economics and other social sciences that focus on the economies of Africa, Asia, and Latin America, as well as developing Europe and the Middle East. It is written for students who have had some basic training in economics and for those with little formal economics background. Essential concepts of economics that are relevant to understanding development problems are highlighted in boldface and explained at appropriate points throughout the text, with glossary terms defined in the margins as well as collected together at the end of the book in a detailed Glossary. Thus, the book should be of special value in undergraduate development courses that attract students from a variety of disciplines. Yet the material is sufficiently broad in scope and rigorous in coverage to satisfy any undergraduate and some graduate economics requirements in the field of development. This text has been

xxii Preface

widely used both in courses taking relatively qualitative and more quan- titative approaches to the study of economic development and emphasiz- ing a variety of themes, including human development.

• The text features a 15-chapter structure, convenient for use in a comprehen- sive course and corresponding well to a 15-week semester but with enough breadth to easily form the basis for a two-semester sequence. However, the chapters are now subdivided, making it easier to use the text in targeted ways. To give one example, some instructors have paired the sections on conflict (14.5) and on informal and micro finance (15.4) with Chapter 5 on poverty.

• Courses with a qualitative focus. For qualitatively oriented courses, with an institutional focus and using fewer economic models, one or more chap- ters or subsections may be omitted, while placing primary emphasis on Chapters 1, 2, 5, 6, 8, and 9, plus parts of Chapters 7 and 10, and other se- lected sections, according to topics covered. The text is structured so that the limited number of graphical models found in those chapters may be omitted without losing the thread, while the intuition behind the models is explained in detail.

• Courses with a more analytic and methods focus. These courses would focus more on the growth and development theories in Chapter 3 (including Ap- pendices such as 3.3 on endogenous growth) and Chapter 4, and highlight and develop some of the core models of the text, including poverty and in- equality measurement and analysis in Chapter 5, microeconomics of fertil- ity and relationships between population growth and economic growth in Chapter 6, migration models in Chapter 7, human capital theory, including the child labor model and empirics in Chapter 8, sharecropping models in Chapter 9, environmental economics models in Chapter 10, tools such as net present benefit analysis and multisector models along with political economy analysis in Chapter 11, and trade models in Chapter 12. Regard- ing methods, these courses could also expand on material introduced in some of the Findings boxes and subsections into more detailed treatments of methods topics such as use of instrumental variables, randomization, regression discontinuity, and growth empirics, including origins of com- parative development and analysis of convergence (which is examined in Chapter 2). Endnotes and sources suggest possible directions to take. The text emphasizes in-depth institutional background reading accompanying the models that help students to appreciate their importance.

• Courses emphasizing human development and poverty alleviation. The Twelfth Edition can be used for a course with a human development focus. This would typically include the sections on Amartya Sen’s capabil- ity approach and Millennium Development Goals in Chapter 1, the new section on conflict in Chapter 14, the discussion of microfinance institu- tions in Chapter 15, and a close and in-depth examination of Chapters 2 and 5. Sections on population policy in Chapter 6; diseases of poverty and problems of illiteracy, low schooling, and child labor in Chapter 8; prob- lems facing people in traditional agriculture in Chapter 9; relationships between poverty and environmental degradation in Chapter 10; and roles of nongovernmental organizations (NGOs) in Chapter 11 would be likely highlights of this course.

xxiiiPreface

• Courses emphasizing macro and international topics. International and macro aspects of economic development could emphasize sections 2.6 and 2.7 on convergence, and long-run growth and sources of comparative de- velopment; Chapter 3 on theories of growth (including the three detailed appendixes to that chapter); Chapter 4 on growth and multiple-equilibrium models; and Chapters 12 through 15 on international trade, international fi- nance, debt and financial crises, direct foreign investment, aid, central bank- ing, and domestic finance. The book also covers other aspects of the interna- tional context for development, including the new section on financial crisis (13.6), implications of the rapid pace of globalization and the rise of China (Chapter 12 and such case studies as Brazil in Chapter 1 and China in Chap- ter 4), the continuing struggle for more progress in sub-Saharan Africa, and controversies over debt relief and foreign aid (Chapter 14).

• Broad two-semester course using supplemental readings. Many of the chap- ters contain enough material for several class sessions, when their topics are covered in an in-depth manner, making the text also suitable for a yearlong course or high-credit option. The endnotes and sources offer many starting points for such extensions.

Guiding Approaches and Organization

The text’s guiding approaches are the following:

1. It teaches economic development within the context of a major set of prob- lems, such as poverty, inequality, population growth, the impact of very rapid urbanization and expansion of megacities, persistent public health challenges, environmental decay, and regions experiencing rural stagna- tion, along with the twin challenges of government failure and market failure. Formal models and concepts are used to elucidate real-world de- velopment problems rather than being presented in isolation from these problems.

2. It adopts a problem- and policy-oriented approach, because a central objec- tive of the development economics course is to foster a student’s ability to understand contemporary economic problems of developing countries and to reach independent and informed judgments and policy conclusions about their possible resolution.

3. It simultaneously uses the best available data from Africa, Asia, Latin Amer- ica, and developing Europe and the Middle East, as well as appropriate theoretical tools to illuminate common developing-country problems. These problems differ in incidence, scope, magnitude, and emphasis when we deal with such diverse countries as India, Pakistan, Bangladesh, China, the Philippines, Kenya, Botswana, Nigeria, Ghana, Côte d’Ivoire, Argentina, Brazil, Chile, Mexico, Haiti, and the Dominican Republic. Still, a majority face some similar development problems: persistent poverty and large in- come and asset inequalities, population pressures, low levels of education and health, inadequacies of financial markets, and recurrent challenges in international trade and instability, to name a few.

xxiv Preface

4. It focuses on a wide range of developing countries, not only as independent nation-states, but also in their growing relationships to one another, as well as in their interactions with rich nations in a globalizing economy.

5. Relatedly, the text views development in both domestic and international contexts, stressing the increasing interdependence of the world economy in ar- eas such as food, energy, natural resources, technology, information, and financial flows.

6. It recognizes the necessity of treating the problems of development from an institutional and structural as well as a market perspective, with appropriate modifications of received general economic principles, theories, and poli- cies. It thus attempts to combine relevant theory with realistic institutional analyses. Enormous strides have been made in the study of these aspects of economic development in recent years, which are reflected in this edition.

7. It considers the economic, social, and institutional problems of underdevel- opment as closely interrelated and requiring coordinated approaches to their solution at the local, national, and international levels.

8. The book is organized into three parts. Part One focuses on the nature and meaning of development and underdevelopment and its various manifes- tations in developing nations. After examining the historical growth ex- perience of the developed countries and the long-run experience of the developing countries, we review four classic and contemporary theories of economic development, while introducing basic theories of economic growth. Part Two focuses on major domestic development problems and policies, and Part Three focuses on development problems and policies in international, macro, and financial spheres. Topics of analysis include eco- nomic growth, poverty and income distribution, population, migration, urbani zation, technology, agricultural and rural development, education, health, the environment, international trade and finance, debt, financial crises, domestic financial markets, direct foreign investment, foreign aid, violent conflict, and the roles of market, state, and nongovernmental or- ganizations in economic development. All three parts of the book raise fundamental questions, including what kind of development is most de- sirable and how developing nations can best achieve their economic and social objectives.

9. As part of the text’s commitment to its comprehensive approach, it covers some topics that are not found in other texts on economic development, in- cluding growth diagnostics, industrialization strategy, innovative policies for poverty reduction, the capability approach to well-being, the central role of women, child labor, the crucial role of health, new thinking on the role of cities, the economic character and comparative advantage of nongovern- mental organizations in economic development, emerging issues in environ- ment and development, financial crises, violent conflict, and microfinance.

10. A unique feature of this book is the in-depth case studies and compara- tive case studies appearing at the end of each chapter. Each chapter ’s case study reflects and illustrates specific issues analyzed in that chapter. In- chapter boxes provide shorter case examples.

xxvPreface

Comments on the text are always welcome; these can be sent directly to Stephen Smith at [email protected].

Supplementary Materials

The Twelfth Edition comes with PowerPoint slides for each chapter, which have been expanded and fully updated for this edition.

The text is further supplemented with an Instructor’s Manual by Chris Marme of Augustana College. It has been thoroughly revised and updated to reflect changes to the Twelfth Edition. Both the PowerPoint slides and the Instructor’s Manual can also be downloaded from the Instructor’s Resource Center at www.pearsonhighered.com/irc.

Acknowledgments

Our gratitude to the many individuals who have helped shape this new edition cannot adequately be conveyed in a few sentences. However, we must record our immense indebtedness to the hundreds of former students and contemporary colleagues who took the time and trouble during the past several years to write or speak to us about the ways in which this text could be further improved. We are likewise indebted to a great number of friends (far too many to mention individually) in both the developing world and the developed world who have directly and indirectly helped shape our ideas about development economics and how an economic development text should be structured. The authors would like to thank colleagues and students in both developing and developed countries for their probing and challenging questions.

We are also very appreciative of the advice, criticisms, and suggestions of the many reviewers, both in the United States and abroad, who provided detailed and insightful comments for the Eighth, Ninth, Tenth, Eleventh, and Twelfth Editions:

U.S. Reviewers

Mohammed Akacem, METROPOLITAN STATE UNIVERSITY OF DENVER William A. Amponsah, GEORGIA SOUTHERN UNIVERSITY Erol Balkan, HAMILTON COLLEGE Karna Basu, HUNTER COLLEGE, CITY UNIVERSITY OF NEW YORK Valerie R. Bencivenga, UNIVERSITY OF TEXAS, AUSTIN

xxvi Preface

Sylvain H. Boko, WAKE FOREST UNIVERSITY Michaël Bonnal, UNIVERSITY OF TENNESSEE AT CHATTANOOGA Milica Z. Bookman, ST. JOSEPH’S UNIVERSITY Jim Cobbe, FLORIDA STATE UNIVERSITY Michael Coon, HOOD COLLEGE Lisa Daniels, WASHINGTON COLLEGE Fernando De Paolis, MONTEREY INSTITUTE Luc D’Haese, UNIVERSITY OF GHENT Quentin Duroy, DENISON UNIVERSITY Can Erbil, BRANDEIS UNIVERSITY Yilma Gebremariam, SOUTHERN CONNECTICUT STATE UNIVERSITY Abbas P. Grammy, CALIFORNIA STATE UNIVERSITY, BAKERSFIELD Caren Grown, AMERICAN UNIVERSITY Kwabena Gyimah-Brempong, UNIVERSITY OF SOUTH FLORIDA Bradley Hansen, MARY WASHINGTON COLLEGE John R. Hanson II, TEXAS A&M UNIVERSITY Seid Hassan, MURRAY STATE UNIVERSITY Jeffrey James, TILBURG UNIVERSITY Barbara John, UNIVERSITY OF DAYTON Pareena G. Lawrence, UNIVERSITY OF MINNESOTA, MORRIS Tung Liu, BALL STATE UNIVERSITY John McPeak, SYRACUSE UNIVERSITY Michael A. McPherson, UNIVERSITY OF NORTH TEXAS Daniel L. Millimet, SOUTHERN METHODIST UNIVERSITY Camille Soltau Nelson, TEXAS A&M UNIVERSITY Thomas Osang, SOUTHERN METHODIST UNIVERSITY Elliott Parker, UNIVERSITY OF NEVADA, RENO Julia Paxton, OHIO UNIVERSITY Meenakshi Rishi, SEATTLE UNIVERSITY James Robinson, UNIVERSITY OF CALIFORNIA, BERKELEY Monthien Satimanon, MICHIGAN STATE AND THAMMASAT UNIVERSITY Andreas Savvides, OKLAHOMA STATE UNIVERSITY Rodrigo R. Soares, UNIVERSITY OF MARYLAND Michael Twomey, UNIVERSITY OF MICHIGAN, DEARBORN Wally Tyner, PURDUE UNIVERSITY Nora Underwood, UNIVERSITY OF CALIFORNIA, DAVIS Jogindar Uppal, STATE UNIVERSITY OF NEW YORK Evert Van Der Heide, CALVIN COLLEGE Adel Varghese, ST. LOUIS UNIVERSITY Sharmila Vishwasrao, FLORIDA ATLANTIC UNIVERSITY Bill Watkins, CALIFORNIA LUTHERAN UNIVERSITY Janice E. Weaver, DRAKE UNIVERSITY Jonathan B. Wight, UNIVERSITY OF RICHMOND Lester A. Zeager, EAST CAROLINA UNIVERSITY

U.K. Reviewers

Arild Angelsen, AGRICULTURAL UNIVERSITY OF NORWAY David Barlow, NEWCASTLE UNIVERSITY

xxviiPreface

Sonia Bhalotra, UNIVERSITY OF BRISTOL Bernard Carolan, UNIVERSITY OF STAFFORDSHIRE Matthew Cole, UNIVERSITY OF BIRMINGHAM Alex Cunliffe, UNIVERSITY OF PLYMOUTH Chris Dent, UNIVERSITY OF HULL Sanjit Dhami, UNIVERSITY OF NEWCASTLE Subrata Ghatak, KINGSTON UNIVERSITY Gregg Huff, UNIVERSITY OF GLASGOW Diana Hunt, SUSSEX UNIVERSITY Michael King, TRINITY COLLEGE DUBLIN Dorothy Manning, UNIVERSITY OF NORTHUMBRIA Mahmood Meeskoub, UNIVERSITY OF LEEDS Paul Mosley, UNIVERSITY OF SHEFFIELD Bibhas Saha, UNIVERSITY OF EAST ANGLIA Colin Simmons, UNIVERSITY OF SALFORD Pritam Singh, OXFORD BROOKES UNIVERSITY Shinder Thandi, UNIVERSITY OF COVENTRY Paul Vandenberg, UNIVERSITY OF BRISTOL

Their input has strengthened the book in many ways and has been much ap- preciated.

Our thanks also go to the staff at Pearson in both the United States and the United Kingdom, particularly David Alexander, Lindsey Sloan, Liz Napoli- tano, and Kate Brewin.

Finally, to his lovely wife, Donna Renée, Michael Todaro wishes to express great thanks for typing the entire First Edition manuscript and for providing the spiritual and intellectual inspiration to persevere under difficult circum- stances. He reaffirms here his eternal devotion to her for always being there to help him maintain a proper perspective on life and living and, through her own creative and artistic talents, to inspire him to think in original and some- times unconventional ways about the global problems of human development.

Stephen Smith would like to thank his wonderful wife, Renee, and his chil- dren, Martin and Helena, for putting up with the many working Saturdays that went into the revision of this text.

Michael P. Todaro

Stephen C. Smith

This page intentionally left blank

Pa r t O n e Principles and Concepts

Prologue: An Extraordinary Moment

Two pictures of the developing world compete in the media for the public’s attention. The first is misery in places like rural Africa or unsanitary and overcrowded urban slums in South Asia. The second is extraordinary dyna- mism in places like coastal China. Both pictures convey important parts of the great development drama. Living conditions are improving significantly in most, though not all, parts of the globe—if sometimes slowly and unevenly. The cumulative effect is that economic development has been giving rise to unprecedented global transformations.

Consider the world of 1992, a time when the divide between the rich developed nations and the low-income developing nations was apparently widening. Rich countries were growing faster than poor countries; and the dominance of high-income industrialized nations in the global order was clear-cut. The United States had just won the Cold War, with the Soviet Union disintegrating in the last days of 1991. The end of the Cold War also saw the European Union in the ascendency, full of confidence with its high- profile Europe ‘92 Single Market project. The real estate and stock market bubble in Japan was just beginning to deflate, with almost no one predicting

2

Introducing Economic Development: A Global Perspective

Development can be seen . . . as a process of expanding the real freedoms that people enjoy. —Amartya Sen, Nobel laureate in economics

Our vision and our responsibility are to end extreme poverty in all its forms in the context of sustainable development and to have in place the building blocks of sustained prosperity for all.

—Report of the High-Level Panel of Eminent Persons on the Post-2015 Development Agenda, 2013

Under necessaries, therefore, I comprehend, not only those things which nature, but those things which the established rules of decency, have rendered necessary to the lowest rank of people.

—Adam Smith, The Wealth of Nations

We are at an auspicious moment in history when successes of past decades and an increasingly favorable economic outlook combine to give developing countries a chance—for the first time ever—to end extreme poverty within a generation.…to create a world for our children which is defined not by stark inequities but by soaring opportunities. A sustainable world where all households have access to clean energy. A world where everyone has enough to eat. A world where no one dies from preventable diseases. A world free of poverty.

—Jim Yong Kim, World Bank President, 2013

1

3CHAPTER 1 Introducing Economic Development: A Global Perspective

the protracted stagnation that would follow Japan’s long period of high eco- nomic growth.

Yet in 1992, many developing nations, including Brazil, Russia, India, China, and South Africa (now sometimes grouped by the media as the “BRICS”), found themselves in precarious conditions if not full-scale crisis. Brazil—like most of Latin America—was still struggling to emerge from the 1980s’ debt crisis. Russia was descending into depression after the collapse of its Soviet economy. India was trying to rebound from its worst economic crisis since independence. China had launched its period of very rapid growth, but the 1989 massacre in Tiananmen Square was a fresh memory and future prospects for reform and growth in China were uncertain. Meanwhile, the end of apartheid was still being negotiated in South Africa, while the continent as a whole was entering its second consecutive lost decade of slow economic growth, and pessimism prevailed. Despite pressing development needs, there were widespread concerns that with the end of the Cold War, the rich world would lose interest in development assistance. And at the 1992 Earth Summit, while the world was taking its first tentative steps to acknowledge and try to restrain climate change due to global warming, almost no one imag- ined that 20 years later China and India would be among the top three green- house gas emitters.

But since 1992, we have moved from a sharp dualism between a rich Center and a backward Global South periphery to more dynamic and com- plex relationships. Asia has been growing at an average rate almost triple that of high-income Western countries, and growth has returned to Africa, herald- ing the promise of an era of global convergence.1 The scale of transformation is immense.

Health has improved strongly, with dramatic declines in child mortality; and the goal of universal primary education is coming into sight. Poverty has fallen. While about two-fifths of the global population lived in extreme poverty in 1990, the fraction has fallen to about one-fifth today. The number of people living in extreme poverty in China (on less than $1.25 per day) fell from about 743 million in 1992 to 157 million in 2009. India has seen substan- tial, if less dramatic, reductions in poverty; social programs in Brazil such as Bolsa Familia have helped substantially reduce the country’s once intractible poverty problems. The enormous growth of innovations such as mobile phones and of availability of credit for small enterprises have led to benefits and fueled a new optimism.

At the same time, the future of economic development and poverty reduc- tion is far from assured—many people who have come out of poverty remain vulnerable, the natural environment is deteriorating, and national economic growth remains uncertain. Economic development is a process, not of years, but of many decades. After the 2011 media celebration of the “BRICS” economic growth, there were reminders that the process remains uneven and uncer- tain. In Brazil economic growth fell from a spike of close to 7.5% in 2010 to under 1% in 2012. Growth in India, topping 10% for the first time in 2010, fell to barely a third that level in 2012. Growth in China fell from over 10% in 2010 to below 8% in 2012 with projections of a permanently slower pace of perhaps 7%. In 2012 growth in South Africa was little more than 3%. Growth per person was slower as populations continued to grow. When financial markets were

4 PART one Principles and Concepts

unsettled during the summer of 2013, many investors started withdrawing money from these and other developing countries.

Meanwhile, many in the development community were dismayed by a 2013 report showing the number of people living in poverty in Africa had yet to decline, and the average income of those remaining poor had still not risen above its long-term level of just 70 cents per day. And climate change talks, also launched in 1992, proceeded at a snail’s pace, even as greenhouse gas emissions reached record levels and the impacts of climate change had become all too visible in low-income countries, threatening to reverse progress in South Asia as well as Africa.

But while optimism that other countries could soon match China’s histori- cally high growth rates dimmed, nonetheless the potential for dramatic catch- up remained as bright as ever. The media pessimism that prevailed in the summer of 2013 was no more warranted than the blind optimism of just two years earlier. Realism is needed—both about the daunting challenges and the exciting opportunities. Gains for the developing world in recent years have been genuine and substantial—in some cases transformative—with many developing countries steadily closing the gap with the developed world, par- ticularly in health and education, and very often in income. Prospects remain strong in coming years, particularly for middle-income countries; yet the high volatility of growth is just one hint at the remaining broader development challenges, as we will examine throughout this text.

This book will explain what lies behind the headline numbers and the sweep of development patterns, presenting the necessary analytic tools and the most recent and reliable data—on challenges ranging from poverty to international finance. To begin, even today many of the world’s poorest people have benefited little, if at all, from the new global prosperity.

1.1 How the Other Half Live

As people throughout the world awake each morning to face a new day, they do so under very different circumstances. Some live in comfortable homes with many rooms. They have more than enough to eat, are well clothed and healthy, and have a reasonable degree of financial security. Others—and these constitute a majority of the earth’s more than 7 billion people—are much less fortunate. They may have inadequate food and shelter, especially if they are among the poorest third. Their health is often poor, they may not know how to read or write, they may be unemployed, and their prospects for a better life are uncertain at best. About two-fifths of the world’s population lives on less than $2 per day, part of a condition of absolute poverty. An examination of these global differences in living standards is revealing.

If, for example, we looked first at a family of four in North America, we would probably find an annual income of over $50,000. They would live in a comfortable suburban house with a small yard or garden, and two cars. The dwelling would have many comfortable features, including a separate bedroom for each of the two children. It would be filled with numerous consumer goods, electronics, and electrical appliances, many of which were manufactured outside North America in countries as far away as South Korea and China. Examples might

Absolute poverty A situa- tion of being unable to meet the minimum levels of income, food, clothing, health care, shelter, and other essentials.

5CHAPTER 1 Introducing Economic Development: A Global Perspective

include computer hard disks made in Malaysia, DVD players manufactured in Thailand, garments assembled in Bangladesh, and mountain bikes made in China. There would always be three meals a day and plenty of processed snack foods, and many of the food products would also be imported from overseas: coffee from Brazil, Kenya, or Colombia; canned fish and fruit from Peru and Australia; and bananas and other tropical fruits from Central America. Both chil- dren would be healthy and attending school. They could expect to complete their secondary education and probably go to a university, choose from a variety of careers to which they might be attracted, and live to an average age of 78 years.

This family, which is typical of families in many rich nations, appears to have a reasonably good life. The parents have the opportunity and the neces- sary education or training to find regular employment; to shelter, clothe, feed, and educate their children; and to save some money for later life. Against these “economic” benefits, there are always “noneconomic” costs. The competitive pressures to “succeed” financially are very strong, and during inflationary or recessionary times, the mental strain and physical pressure of trying to provide for a family at levels that the community regards as desirable can take its toll on the health of both parents. Their ability to relax, to enjoy the simple plea- sures of a country stroll, to breathe clean air and drink pure water, and to see a crimson sunset is constantly at risk with the onslaught of economic progress and environmental decay. But on the whole, theirs is an economic status and lifestyle toward which many millions of less fortunate people throughout the world seem to be aspiring.

Now let us examine a typical “extended” family in a poor rural area of South Asia. The household is likely to consist of eight or more people, including par- ents, several children, two grandparents, and some aunts and uncles. They have a combined real per capita annual income, in money and in “kind” (meaning that they consume a share of the food they grow), of $300. Together they live in a poorly constructed one- or two-room house as tenant farmers on a large agricultural estate owned by an absentee landlord who lives in the nearby city. The father, mother, uncle, and older children must work all day on the land. The adults cannot read or write; the younger children attend school irregularly and cannot expect to proceed beyond a basic primary education. All too often, when they do get to school, the teacher is absent. They often eat only two (and some- times just one) meals per day; the food rarely changes, and the meals are rarely sufficient to alleviate the children’s persistent hunger pains. The house has no electricity, sanitation, or fresh water supply. Sickness occurs often, but quali- fied doctors and medical practitioners are far away in the cities, attending to the needs of wealthier families. The work is hard, the sun is hot, and aspirations for a better life are continually being snuffed out. For families such as theirs, the only relief from the daily struggle for physical survival lies in the spiritual tradi- tions of the people.

Shifting to another part of the world, suppose we were to visit a large city situated along the coast of South America. We would immediately be struck by the sharp contrasts in living conditions from one section of this sprawling metropolis to another. There would be a modern stretch of tall buildings and wide, tree-lined boulevards along the edge of a gleaming white beach; just a few hundred meters back and up the side of a steep hill, squalid shanties would be pressed together in precarious balance.

6 PART onE Principles and Concepts

If we were to examine two representative families—one a wealthy and well- connected family and the other of peasant background or born in the slums— we would no doubt also be struck by the wide disparities in their individual living conditions. The wealthy family lives in a multiroom complex on the top floor of a modern building overlooking the sea, while the peasant family is cramped tightly into a small makeshift dwelling in a shantytown, or favela (squatters’ slum), on the hill behind that seafront building.

For illustrative purposes, let us assume that it is a typical Saturday evening at an hour when the families should be preparing for dinner. In the penthouse apartment of the wealthy family, a servant is setting the table with expensive imported china, high-quality silverware, and fine linen. Russian caviar, French hors d’œuvres, and Italian wine will constitute the first of several courses. The family’s eldest son is home from his university in North America, and the other two children are on vacation from their boarding schools in France and Switzerland. The father is a prominent surgeon trained in the United States. His clientele consists of wealthy local and foreign dignitaries and business- people. In addition to his practice, he owns a considerable amount of land in the countryside. Annual vacations abroad, imported luxury automobiles, and the finest food and clothing are commonplace amenities for this fortunate family in the penthouse apartment.

And what about the poor family living in the dirt-floored shack on the side of the hill? They too can view the sea, but somehow it seems neither scenic nor relaxing. The stench of open sewers makes such enjoyment rather remote. There is no dinner table being set; in fact, there is usually too little to eat. Most of the four children spend their time out on the streets begging for money, shining shoes, or occasionally even trying to steal purses from unsuspecting people who stroll along the boulevard. The father migrated to the city from the rural hinterland, and the rest of the family recently followed. He has had part-time jobs over the years, but nothing permanent. Government assistance has recently helped this family keep the children in school longer. But lessons learned on the streets, where violent drug gangs hold sway, seem to be making a deeper impression.

One could easily be disturbed by the sharp contrast between these two ways of life. However, had we looked at almost any other major city in Latin America, Asia, and Africa, we would have seen much the same contrast (although the extent of inequality might have been less pronounced).

Now imagine that you are in a remote rural area in the eastern part of Africa, where many small clusters of tiny huts dot a dry and barren land. Each cluster contains a group of extended families, all participating in and sharing the work. There is little money income here because most food, cloth- ing, shelter, and worldly goods are made and consumed by the people them- selves—theirs is a subsistence economy. There are few passable roads, few schools, and no hospitals, electric wires, or water supplies. In many respects, it is as stark and difficult an existence as that of the people in that Latin Ameri- can favela across the ocean. Yet perhaps it is not as psychologically troubling because there is no luxurious penthouse by the sea to emphasize the relative deprivation of the very poor. With the exception of population growth and problems of the increasingly fragile environment, life here seems to be almost eternal and unchanging—but not for much longer.

Subsistence economy An economy in which production is mainly for personal con- sumption and the standard of living yields little more than basic necessities of life—food, shelter, and clothing.

7CHAPTER 1 Introducing Economic Development: A Global Perspective

A new road is being built that will pass near this village. No doubt it will bring with it the means for prolonging life through improved medical care. But it will also bring more information about the world outside, along with the gadgets of modern civilization. The possibilities of a “better” life will be pro- moted, and the opportunities for such a life will become feasible. Aspirations will be raised, but so will frustrations as people understand the depth of some of their deprivations more clearly. In short, the development process has been set in motion.

Before long, exportable fruits and vegetables will probably be grown in this region. They may even end up on the dinner table of the rich South American family in the seaside penthouse. Meanwhile, radios made in Southeast Asia and playing music recorded in northern Europe have become prized possessions in this African village. In villages not far away, mobile phone use has been intro- duced and is growing rapidly. Throughout the world, remote subsistence villages such as this one are being linked up with modern civilization in an increasing number of ways. The process, well under way, will become even more intensi- fied in the coming years.

Finally, imagine you are in booming East Asia; to illustrate, a couple born in obscure zhuangs (rural areas) in populous central Sichuan Province grew up in the 1960s, going to school for six years and becoming rice farmers like their parents. The rice grew well, but memories of famine were still sharp in their commune, where life was also hard during the Cultural Revolution. Their one daughter, let’s call her Xiaoling, went to school for ten years. Much of the rice they and their commune grew went to the state at a price that never seemed high enough. After 1980, farmers were given rights to keep and sell more of their rice. Seeing the opportunity, they grew enough to meet government quotas and sold more of it. Many also raised vegetables to sell in a booming city 100 kilometers up the river and other towns. Living standards improved, though then their incomes stagnated for many years. But they heard about peasants moving first to cities in the south and recently to closer cities—making more money becoming factory workers. When their daughter was 17, farmers from the village where the mother grew up were evicted from their land because it was close to lakes created by an immense dam project. Some were resettled, but others went to Shenzhen, Guangzhou, or Chongqing. Xiaoling talked with her family, saying she too wanted to move there for a while to earn more money. She found a city that had already grown to several million people, quickly find- ing a factory job. She lived in a dormitory, and conditions were often harsh, but she could send some money home and save toward a better life. She watched the city grow at double digits, becoming one of the developing world’s new mega- cities, adding territories and people to reach over 15 million people. After a few years, she opened a humble business, selling cosmetics and costume jewelry to the thousands of women from the countryside arriving every day. She had five proposals of marriage, with parents of single men near where she grew up offering gifts, even an enormous house. She knows many people still live in deep poverty and finds inequality in the city startling. For now she plans to stay, where she sees opportunities for her growing business and a life she never imag- ined having in her village.

Listening to the poor explain what poverty is like in their own words is more vivid than reading descriptions of it. Listen to some of the voices of the

Development The process of improving the quality of all human lives and capabili- ties by raising people’s levels of living, self-esteem, and freedom.

8 PART onE Principles and Concepts

poor about the experience of poverty in Box 1.1.2 From these, together with the voices of the poor recorded in Box 5.1 and Box 8.1, it is clear that what people living in poverty need and want extend beyond increased income to health, education, and—especially for women—empowerment. These correspond to enhanced capabilities and to the achievement of the Millennium Development Goals (and its emerging successor, the Sustainable Development Goals), intro- duced later in this chapter.

This first fleeting glimpse at life in different parts of our planet is suffi- cient to raise various questions. Why does affluence coexist with dire poverty, not only on different continents, but also within the same country or even the same city? Can traditional, low-productivity, subsistence societies be trans- formed into modern, high-productivity, high-income nations? To what extent are the development aspirations of poor nations helped or hindered by the economic activities of rich nations? By what process and under what condi- tions do rural subsistence farmers in the remote regions of Nigeria, Brazil, or the Philippines evolve into successful commercial farmers? What are the implications of the surprisingly long stagnation in rich countries following the financial crisis for further progress on development and poverty reduction? These and many other questions concerning international and national differ- ences in standards of living, in areas including health and nutrition, educa- tion, employment, environmental sustainability, population growth, and life expectancies, might be posed on the basis of even this very superficial look at life around the world.

BOX 1.1 the experience of Poverty: Voices of the Poor

When one is poor, she has no say in public, she feels inferior. She has no food, so there is famine in her house; no clothing, and no progress in her family.

—a poor woman from Uganda

For a poor person, everything is terrible—illness, humiliation, shame. We are cripples; we are afraid of everything; we depend on everyone. No one needs us. We are like garbage that everyone wants to get rid of.

—a blind woman from tiraspol, Moldova

Life in the area is so precarious that the youth and every able person have to migrate to the towns or join the army at the war front in order to escape the hazards of hunger escalating over here.

—Participant in a discussion group

in rural ethiopia

When food was in abundance, relatives used to share it. These days of hunger, however, not even relatives would help you by giving you some food.

—Young man in nichimishi, Zambia

We have to line up for hours before it is our turn to draw water.

—Participant in a discussion group from

Mbwadzulu Village (Mangochi), Malawi

[Poverty is] . . . low salaries and lack of jobs. And it’s also not having medicine, food, and clothes.

—Participant in a discussion group in Brazil

Don’t ask me what poverty is because you have met it outside my house. Look at the house and count the number of holes. Look at the utensils and the clothes I am wearing. Look at everything and write what you see. What you see is poverty.

—Poor man in Kenya

9CHAPTER 1 Introducing Economic Development: A Global Perspective

This book is designed to help students obtain a better understanding of the major problems and prospects for broad-based economic development, paying special attention to the plight of the half or more of the world’s popu- lation for whom low levels of living are a fact of life. However, as we shall soon discover, the process in developing countries cannot be analyzed real- istically without also considering the role of economically developed nations in directly or indirectly promoting or retarding that development. Perhaps even more important to students in the developed nations is that as our earth shrinks with the spread of modern transport and communications, the futures of all peoples on this small planet are becoming increasingly interdependent. What happens to the health and economic welfare of poor rural families and many others in the developing regions of Asia, Africa, the Middle East, or Latin America will in one way or another, directly or indirectly, affect the health and economic welfare of families in Europe and North America, and vice versa. The steady loss of tropical forests contributes to global warming; new diseases spread much more rapidly thanks to increased human mobility; economic interdependence steadily grows. It is within this context of a common future for all humankind in the rapidly shrinking world of the twenty-first century that we now commence our study of economic development.

1.2 Economics and Development Studies

The study of economic development is one of the newest, most exciting, and most challenging branches of the broader disciplines of economics and politi- cal economy. Although one could claim that Adam Smith was the first “devel- opment economist” and that his Wealth of Nations, published in 1776, was the first treatise on economic development, the systematic study of the problems and processes of economic development in Africa, Asia, and Latin America has emerged only over the past five decades or so. Although development eco- nomics often draws on relevant principles and concepts from other branches of economics in either a standard or modified form, for the most part it is a field of study that is rapidly evolving its own distinctive analytical and meth- odological identity.3

The Nature of Development Economics

Traditional economics is concerned primarily with the efficient, least-cost allocation of scarce productive resources and with the optimal growth of these resources over time so as to produce an ever-expanding range of goods and services. Traditional neoclassical economics deals with an advanced capitalist world of perfect markets; consumer sovereignty; automatic price adjust- ments; decisions made on the basis of marginal, private-profit, and utility calculations; and equilibrium outcomes in all product and resource markets. It assumes economic “rationality” and a purely materialistic, individualistic, self-interested orientation toward economic decision making.

Political economy goes beyond traditional economics to study, among other things, the social and institutional processes through which certain groups of economic and political elites influence the allocation of scarce productive

Developing countries Countries of Asia, Africa, the Middle East, Latin America, eastern Europe, and the former Soviet Union that are presently characterized by low levels of living and other development deficits. Used in the develop- ment literature as a synonym for less developed countries.

Traditional economics An approach to economics that emphasizes utility, profit maximization, market effi- ciency, and determination of equilibrium.

Political economy The attempt to merge economic analysis with practical politics— to view economic activity in its political context.

10 PART onE Principles and Concepts

resources now and in the future, either for their own benefit exclusively or for that of the larger population as well. Political economy is therefore concerned with the relationship between politics and economics, with a special emphasis on the role of power in economic decision making.

Development economics has an even greater scope. In addition to being concerned with the efficient allocation of existing scarce (or idle) productive resources and with their sustained growth over time, it must also deal with the economic, social, political, and institutional mechanisms, both public and private, necessary to bring about rapid (at least by historical standards) and large-scale improvements in levels of living for the peoples of Africa, Asia, Latin America, and the formerly socialist transition economies. In comparison with the more developed countries (MDCs), in most less developed countries, commodity and resource markets are typically highly imperfect, consumers and producers have limited information, major structural changes are taking place in both the society and the economy, the potential for multiple equilibria rather than a single equilibrium is more common, and disequilibrium situations often prevail (prices do not equate supply and demand). In many cases, economic calculations are heavily influenced by political and social priorities such as unifying the nation, replacing foreign advisers with local decision makers, resolving tribal or ethnic conflicts, or preserving religious and cultural tradi- tions. At the individual level, family, clan, religious, or tribal considerations may take precedence over private, self-interested utility or profit-maximizing calculations.

Thus, development economics, to a greater extent than traditional neoclas- sical economics or even political economy, must be concerned with the eco- nomic, cultural, and political requirements for effecting rapid structural and institutional transformations of entire societies in a manner that will most effi- ciently bring the fruits of economic progress to the broadest segments of their populations. It must focus on the mechanisms that keep families, regions, and even entire nations in poverty traps, in which past poverty causes future poverty, and on the most effective strategies for breaking out of these traps. Consequently, a larger government role and some degree of coordinated eco- nomic decision making directed toward transforming the economy are usually viewed as essential components of development economics. Yet this must somehow be achieved despite the fact that both governments and markets typically function less well in the developing world. In recent years, activi- ties of nongovernmental organizations, both national and international, have grown rapidly and are also receiving increasing attention (see Chapter 11)

Because of the heterogeneity of the developing world and the complexity of the development process, development economics must be eclectic, attempt- ing to combine relevant concepts and theories from traditional economic analy- sis with new models and broader multidisciplinary approaches derived from studying the historical and contemporary development experience of Africa, Asia, and Latin America. Development economics is a field on the crest of a breaking wave, with new theories and new data constantly emerging. These theories and statistics sometimes confirm and sometimes challenge traditional ways of viewing the world. The ultimate purpose of development economics, however, remains unchanged: to help us understand developing economies in order to help improve the material lives of the majority of the global population.

Development economics The study of how economies are transformed from stagna- tion to growth and from low- income to high-income status, and overcome problems of absolute poverty.

More developed countries (MDCs) The now economi- cally advanced capitalist coun- tries of western Europe, North America, Australia, New Zealand, and Japan.

Less developed countries A synonym for developing countries.

11CHAPTER 1 Introducing Economic Development: A Global Perspective

Why Study Development Economics? Some Critical Questions

An introductory course in development economics should help students gain a better understanding of a number of critical questions about the economies of developing nations. The following is a sample list of 30 such questions, followed by the chapters (in parentheses) in which they are discussed. They illustrate the kinds of issues faced by almost every developing nation and, indeed, every development economist.

1. What is the real meaning of development? Do the Millennium Development Goals fit with these meanings? (Chapter 1)

2. What can be learned from the historical record of economic progress in the now developed world? Are the initial conditions similar or different for contemporary developing countries from what the developed coun- tries faced on the eve of their industrialization or in their earlier phases? (Chapter 2)

3. What are economic institutions, and how do they shape problems of under- development and prospects for successful development? (Chapter 2)

4. How can the extremes between rich and poor be so very great? Figure 1.1 illustrates this disparity. (Chapters 2, 3, 4, and 5)

5. What are the sources of national and international economic growth? Who benefits from such growth and why? (Chapters 3 and 5)

6. Why do some countries make rapid progress toward development while many others remain poor? (Chapters 2, 3, and 4)

7. Which are the most influential theories of development, and are they com- patible? Is underdevelopment an internally (domestically) or externally (internationally) induced phenomenon? (Chapters 2, 3, and 4)

8. What constraints most hold back accelerated growth, depending on local conditions? (Chapter 4)

9. How can improvements in the role and status of women have an espe- cially beneficial impact on development prospects? (Chapters 5, 6, 7, 8, 9, and 10)

10. What are the causes of extreme poverty, and what policies have been most effective for improving the lives of the poorest of the poor? (Chapters 5, 6, 7, 8, 9, 10, and 11)

11. With world population superseding 7 billion people, on its way to a pro- jected 9 billion before mid-century, is rapid population growth threatening the economic progress of developing nations? Does having large families make economic sense in an environment of widespread poverty and financial insecurity? (Chapter 6)

12. Why is there so much unemployment and underemployment in the developing world, especially in the cities, and why do people continue to migrate to the cities from rural areas even when their chances of finding a conventional job are slim? (Chapter 7)

12 PART onE Principles and Concepts

13. Under what conditions can cities act as engines of economic transformation? (Chapter 7)

14. Wealthier societies are also healthier ones because they have more resources for improving nutrition and health care. But does better health also help spur successful development? (Chapter 8)

15. What is the impact of poor public health on the prospects for develop- ment, and what is needed to address these problems? (Chapter 8)

16. Do educational systems in developing countries really promote economic development, or are they simply a mechanism to enable certain select groups or classes of people to maintain positions of wealth, power, and influence? (Chapter 8)

FIGUre 1.1 World Income Distribution

1 High-income OECD

2 Eastern and central Europe and CIS

3 Latin America and the Caribbean

4 East Asia and the Pacific

5 South Asia

6 Sub-Saharan Africa

Per capita income

6125_01_FG001

Poorest

Richest

Poorest

Richest

Regional percentage of the population for each 20% of income

World income distributed by percentiles of the population, 2000

(a)

(b)

1 2 3 4 5 6

Part (a) shows world income distribution by percentile. The huge share controlled by the top percentiles gives the graph its “champagne glass shape.” Part (b) shows the regional shares of global income. For example, a large majority of people in the top 20% of the global income distribution live in the rich countries. Most of those in the bottom 60% live in sub-Saharan Africa and Asia. OECD is the Organization for Economic Coopera- tion and Development. CIS is the Commonwealth of Independent States.

Source: From Human Development Report, 2005, p. 37. Reprinted with permission from the United Nations Development Programme.

13CHAPTER 1 Introducing Economic Development: A Global Perspective

17. As more than half the people in developing countries still reside in rural areas, how can agricultural and rural development best be promoted? Are higher agricultural prices sufficient to stimulate food production, or are rural insti- tutional changes and infrastructure (land redistribution, local government reform, roads, transport, education, credit, etc.) also needed? (Chapter 9)

18. What do we mean by “environmentally sustainable development”? Are there serious economic costs for pursuing sustainable development as opposed to simple output growth, and who bears the major responsibility for global environmental damage—the developed North or the developing South? (Chapter 10)

19. Are free markets and economic privatization the answer to development problems, or do governments in developing countries still have major roles to play in their economies? (Chapter 11)

20. Why do so many developing countries select such poor development policies, and what can be done to improve these choices? (Chapter 11)

21. Is expanded international trade always desirable from the point of view of the development of poor nations? Who gains from trade, and how are the advantages distributed among nations? (Chapter 12)

22. When and under what conditions, if any, should governments in develop- ing countries adopt a policy of foreign-exchange control, raise tariffs, or set quotas on the importation of certain “nonessential” goods in order to promote their own industrialization or to ameliorate chronic balance of payments problems? (Chapter 12)

23. What has been the impact of International Monetary Fund “stabilization programs” and World Bank “structural adjustment” lending on the balance of payments and growth prospects of heavily indebted less developed countries? (Chapters 12 and 13)

24. What is meant by globalization, and how is it affecting the developing countries? (Chapters 12, 13, and 14)

25. Should exports of primary products such as agricultural commodities and iron ore be promoted, or should all developing countries attempt to indus- trialize by developing their own manufacturing industries as rapidly as possible? (Chapter 13)

26. How did so many developing nations get into such serious foreign-debt problems, and what are the implications of debt problems for economic development? How do financial crises affect development? (Chapter 13)

27. What is the impact of foreign economic aid from rich countries? Should developing countries continue to seek such aid, and if so, under what con- ditions and for what purposes? Should developed countries continue to offer such aid, and if so, under what conditions and for what purposes? (Chapter 14)

28. Should multinational corporations be encouraged to invest in the economies of poor nations, and if so, under what conditions? How have the emergence

Globalization The increas- ing integration of national economies into expanding international markets.

14 PART onE Principles and Concepts

of the “global factory” and the globalization of trade and finance influenced international economic relations? (Chapter 14)

29. What is the role of financial and fiscal policy in promoting development? (Chapter 15)

30. What is microfinance, and what are its potential and limitations for reduc- ing poverty and spurring grassroots development? (Chapter 15)

The following chapters analyze and explore these and many related questions. The answers are often more complex than one might think. Remember that the ultimate purpose of any course in economics, including development eco- nomics, is to help students think systematically about economic problems and issues, and formulate judgments and conclusions on the basis of relevant ana- lytical principles and reliable statistical information. Because the problems of development are in many cases unique in the modern world and not often easily understood through the use of traditional economic theories, we may often need unconventional approaches to what may appear to be conven- tional economic problems. Traditional economic principles play a useful role in enabling us to improve our understanding of development problems, but they should not blind us to the realities of local conditions in less developed countries.

The Important Role of Values in Development Economics

Economics is a social science. It is concerned with human beings and the social systems by which they organize their activities to satisfy basic material needs (e.g., food, shelter, clothing) and nonmaterial wants (e.g., education, knowledge, spiritual fulfillment). It is necessary to recognize from the outset that ethical or normative value premises about what is or is not desirable are central features of the economic discipline in general and of development economics in particular. The very concepts of economic development and modernization represent implicit as well as explicit value premises about desirable goals for achieving what Mahatma Gandhi once called the “real- ization of the human potential.” Concepts or goals such as economic and social equality, the elimination of poverty, universal education, rising levels of living, national independence, modernization of institutions, rule of law and due process, access to opportunity, political and economic participation, grassroots democracy, self-reliance, and personal fulfillment all derive from subjective value judgments about what is good and desirable and what is not. So too, for that matter, do other values—for example, the sanctity of private property, however acquired, and the right of individuals to accumulate unlimited personal wealth; the preservation of traditional hierarchical social institutions and rigid, inegalitarian class structures; the male head of house- hold as the final authority; and the supposed “natural right” of some to lead while others follow.

When we deal in Part Two with such major issues of development as poverty, inequality, population growth, rural stagnation, and environmental decay, the mere identification of these topics as problems conveys the value

15CHAPTER 1 Introducing Economic Development: A Global Perspective

judgment that their improvement or elimination is desirable and therefore good. That there is widespread agreement among many different groups of people—politicians, academics, and ordinary citizens—that these are desir- able goals does not alter the fact that they arise not only out of a reaction to an objective empirical or positive analysis of what is but also ultimately from a subjective or normative value judgment about what should be.

It follows that value premises, however carefully disguised, are an inherent component of both economic analysis and economic policy. Economics cannot be value-free in the same sense as, say, physics or chemistry. Thus, the validity of economic analysis and the correctness of economic prescriptions should always be evaluated in light of the underlying assumptions or value premises. Once these subjective values have been agreed on by a nation or, more specifically, by those who are responsible for national decision making, specific development goals (e.g., greater income equality) and corresponding public policies (e.g., tax- ing higher incomes at higher rates) based on “objective” theoretical and quan- titative analyses can be pursued. However, where serious value conflicts and disagreements exist among decision makers, the possibility of a consensus about desirable goals or appropriate policies is considerably diminished. In either case, it is essential, especially in the field of development economics, that one’s value premises always be made clear.4

Economies as Social Systems: The Need to Go Beyond Simple Economics

Economics and economic systems, especially in the developing world, must be viewed in a broader perspective than that postulated by traditional eco- nomics. They must be analyzed within the context of the overall social system of a country and, indeed, within an international, global context as well. By “social system,” we mean the interdependent relationships between economic and noneconomic factors. The latter include attitudes toward life, work, and authority; public and private bureaucratic, legal, and administrative struc- tures; patterns of kinship and religion; cultural traditions; systems of land tenure; the authority and integrity of government agencies; the degree of popu- lar participation in development decisions and activities; and the flexibility or rigidity of economic and social classes. Clearly, these factors vary widely from one region of the world to another and from one culture and social setting to another. At the international level, we must also consider the organization and rules of conduct of the global economy—how they were formulated, who controls them, and who benefits most from them. This is especially true today with the spread of market economies and the rapid globalization of trade, finance, corporate boundaries, technology, intellectual property, and labor migration.

Resolving problems to achieve development is a complicated task. Increas- ing national production, raising levels of living, and promoting widespread employment opportunities are all as much a function of the local history, expectations, values, incentives, attitudes and beliefs, and institutional and power structures of both the domestic and the global society as they are the direct outcomes of the manipulation of strategic economic variables such as

Social system The orga- nizational and institutional structure of a society, including its values, attitudes, power structure, and traditions.

16 PART onE Principles and Concepts

savings, investment, product and factor prices, and foreign-exchange rates. As the Indonesian intellectual Soedjatmoko, former rector of the United Nations University in Tokyo, so aptly put it:

Looking back over these years, it is now clear that, in their preoccupation with growth and its stages and with the provision of capital and skills, development theorists have paid insufficient attention to institutional and structural problems and to the power of historical, cultural, and religious forces in the development process.5

Just as some social scientists occasionally make the mistake of confusing their theories with universal truths, they also sometimes mistakenly dismiss these noneconomic variables as “nonquantifiable” and therefore of dubious impor- tance. Yet these variables often play a critical role in the success or failure of the development effort.

As you will see in Parts Two and Three, many of the failures of develop- ment policies have occurred precisely because these noneconomic variables (e.g., the role of traditional property rights in allocating resources and distribut- ing income or the influence of religion on attitudes toward modernization and family planning) were excluded from the analysis. Although the main focus of this text is on development economics and its usefulness in understanding problems of economic and social progress in poor nations, we will try always to be mindful of the crucial roles that values, attitudes, and institutions, both domestic and international, play in the overall development process.

1.3 What Do We Mean by Development?

Because the term development may mean different things to different people, it is important that we have some working definition or core perspective on its meaning. Without such a perspective and some agreed measurement criteria, we would be unable to determine which country was actually developing and which was not. This will be our task for the remainder of the chapter and for our first country case study, Brazil, at the end of the chapter.

Traditional Economic Measures

In strictly economic terms, development has traditionally meant achieving sus- tained rates of growth of income per capita to enable a nation to expand its output at a rate faster than the growth rate of its population. Levels and rates of growth of “real” per capita gross national income (GnI) (monetary growth of GNI per capita minus the rate of inflation) are then used to measure the overall economic well-being of a population—how much of real goods and services is available to the average citizen for consumption and investment.

Economic development in the past has also been typically seen in terms of the planned alteration of the structure of production and employment so that agri- culture’s share of both declines and that of the manufacturing and service indus- tries increases. Development strategies have therefore usually focused on rapid industrialization, often at the expense of agriculture and rural development.

With few exceptions, such as in development policy circles in the 1970s, devel- opment was until recently nearly always seen as an economic phenomenon in

Values Principles, standards, or qualities that a society or groups within it considers worthwhile or desirable.

Attitudes The states of mind or feelings of an individual, group, or society regarding issues such as material gain, hard work, saving for the future, and sharing wealth.

Institutions Norms, rules of conduct, and generally accepted ways of doing things. Economic institu- tions are humanly devised constraints that shape human interactions, including both informal and formal “rules of the game” of economic life in the widely used framework of Douglass North.

Income per capita Total gross national income of a country divided by its total population.

Gross national income (GnI) The total domestic and foreign output claimed by residents of a country. It comprises gross domestic product (GDP) plus factor incomes accruing to residents from abroad, less the income earned in the domestic economy accruing to persons abroad.

17CHAPTER 1 Introducing Economic Development: A Global Perspective

which rapid gains in overall and per capita GNI growth would either “trickle down” to the masses in the form of jobs and other economic opportunities or create the necessary conditions for the wider distribution of the economic and social benefits of growth. Problems of poverty, discrimination, unemployment, and income distribution were of secondary importance to “getting the growth job done.” Indeed, the emphasis is often on increased output, measured by gross domestic product (GDP).

The New Economic View of Development

The experience of the first decades of post–World War II and postcolonial development in the 1950s, 1960s, and early 1970s, when many developing nations did reach their economic growth targets but the levels of living of the masses of people remained for the most part unchanged, signaled that something was very wrong with this narrow definition of development. An increasing number of economists and policymakers clamored for more direct attacks on widespread absolute poverty, increasingly inequitable income dis- tributions, and rising unemployment. In short, during the 1970s, economic development came to be redefined in terms of the reduction or elimination of poverty, inequality, and unemployment within the context of a growing econ- omy. “Redistribution from growth” became a common slogan. Dudley Seers posed the basic question about the meaning of development succinctly when he asserted:

The questions to ask about a country’s development are therefore: What has been happening to poverty? What has been happening to unemployment? What has been happening to inequality? If all three of these have declined from high levels, then beyond doubt this has been a period of development for the country con- cerned. If one or two of these central problems have been growing worse, espe- cially if all three have, it would be strange to call the result “development” even if per capita income doubled.6

This assertion was neither idle speculation nor the description of a hypo- thetical situation. A number of developing countries experienced relatively high rates of growth of per capita income during the 1960s and 1970s but showed little or no improvement or even an actual decline in employment, equality, and the real incomes of the bottom 40% of their populations. By the earlier growth definition, these countries were developing; by the newer poverty, equality, and employment criteria, they were not. The situation in the 1980s and 1990s worsened further as GNI growth rates turned negative for many developing countries, and governments, facing mounting foreign-debt problems, were forced to cut back on their already limited social and economic programs.

But the phenomenon of development or the existence of a chronic state of underdevelopment is not merely a question of economics or even one of quantitative measurement of incomes, employment, and inequality. As Denis Goulet forcefully portrayed it:

Underdevelopment is shocking: the squalor, disease, unnecessary deaths, and hopelessness of it all!…The most empathetic observer can speak objectively about underdevelopment only after undergoing, personally or vicariously, the “shock of

Gross domestic product (GDP) The total final output of goods and services produced by the country’s economy, within the country’s territory, by residents and nonresidents, regardless of its allocation between domestic and foreign claims.

18 PART onE Principles and Concepts

underdevelopment.” This unique culture shock comes to one as he is initiated to the emotions which prevail in the “culture of poverty.” The reverse shock is felt by those living in destitution when a new self-understanding reveals to them that their life is neither human nor inevitable.…The prevalent emotion of underdevel- opment is a sense of personal and societal impotence in the face of disease and death, of confusion and ignorance as one gropes to understand change, of ser- vility toward men whose decisions govern the course of events, of hopelessness before hunger and natural catastrophe. Chronic poverty is a cruel kind of hell, and one cannot understand how cruel that hell is merely by gazing upon poverty as an object.7

Development must therefore be conceived of as a multidimensional pro- cess involving major changes in social structures, popular attitudes, and national institutions, as well as the acceleration of economic growth, the reduc- tion of inequality, and the eradication of poverty. Development, in its essence, must represent the whole gamut of change by which an entire social system, tuned to the diverse basic needs and evolving aspirations of individuals and social groups within that system, moves away from a condition of life widely perceived as unsatisfactory toward a situation or condition of life regarded as materially and spiritually better. No one has identified the human goals of economic development as well as Amartya Sen, perhaps the leading thinker on the meaning of development.

Amartya Sen’s “Capability” Approach

The view that income and wealth are not ends in themselves but instruments for other purposes goes back at least as far as Aristotle. Amartya Sen, the 1998 Nobel laureate in economics, argues that the “capability to function” is what really matters for status as a poor or nonpoor person. As Sen puts it, “the expansion of commodity productions...are valued, ultimately, not for their own sake, but as means to human welfare and freedom.”8

In effect, Sen argues that poverty cannot be properly measured by income or even by utility as conventionally understood; what matters fundamentally is not the things a person has—or the feelings these provide—but what a person is, or can be, and does, or can do. What matters for well-being is not just the character- istics of commodities consumed, as in the utility approach, but what use the con- sumer can and does make of commodities. For example, a book is of little value to an illiterate person (except perhaps as cooking fuel or as a status symbol). Or as Sen noted, a person with a parasitic disease will be less able to extract nourish- ment from a given quantity of food than someone without parasites.

To make any sense of the concept of human well-being in general, and poverty in particular, we need to think beyond the availability of commodities and consider their use: to address what Sen calls functionings, that is, what a person does (or can do) with the commodities of given characteristics that they come to possess or control. Freedom of choice, or control of one’s own life, is itself a central aspect of most understandings of well-being. A function- ing is a valued “being or doing,” and in Sen’s view, functionings that people have reason to value can range from being healthy, being well-nourished, and well-clothed, to being mobile, having self-esteem, and “taking part in the life of the community.”9

Functionings What people do or can do with the com- modities of given characteris- tics that they come to possess or control.

19CHAPTER 1 Introducing Economic Development: A Global Perspective

Sen identifies five sources of disparity between (measured) real incomes and actual advantages:10 first, personal heterogeneities, such as those connected with disability, illness, age, or gender; second, environmental diversities, such as heating and clothing requirements in the cold or infectious diseases in the tropics, or the impact of pollution; third, variations in social climate, such as the prevalence of crime and violence, and “social capital”; fourth, distribution within the family—economic statistics measure incomes received in a family because it is the basic unit of shared consumption, but family resources may be distributed unevenly, as when girls get less medical attention or education than boys do; fifth, differences in relational perspectives, meaning that some goods are essential because of local customs and conventions. For example, nec- essaries for being able, in Adam Smith’s phrase, “to appear in public without shame,” include higher quality clothing (such as leather shoes) in high-income countries than in low-income countries.

In a richer society, the ability to partake in community life would be extremely difficult without certain commodities, such as a telephone, a televi- sion, or an automobile; it is difficult to function socially in Singapore or South Korea without an e-mail address. And minimal housing standards to avoid social disgrace also rise strongly with the average wealth of the society.

Thus, looking at real income levels or even the levels of consumption of spe- cific commodities cannot suffice as a measure of well-being. One may have a lot of commodities, but these are of little value if they are not what consumers desire (as in the former Soviet Union). One may have income, but certain com- modities essential for well-being, such as nutritious foods, may be unavailable. Even when providing an equal number of calories, the available staple foods in one country (cassava, bread, rice, cornmeal, potatoes, etc.) will differ in nutri- tional content from staple foods in other countries. Moreover, even some sub- varieties of, for example, rice, are much more nutritious than others. Finally, even when comparing absolutely identical commodities, one has to frame their consumption in a personal and social context. Sen provides an excellent exam- ple of bread, the most basic of commodities. It has product “characteristics” such as taste and nutrition such as protein; and it helps to meet conventions of social exchange in the sense of breaking bread together. But many of these benefits depend on the person and her circumstances, such as her activity level, metabolism, weight, whether she is pregnant or lactating, nutrition knowledge, whether she is infected with parasites, and her access to medical services. Sen goes on to note that functioning depends also on (1) “social conventions in force in the society in which the person lives, (2) the position of the person in the family and in the society, (3) the presence or absence of festivities such as marriages, seasonal festivals and other occasions such as funerals, (4) the physical distance from the homes of friends and relatives...”11

In part because such factors, even on so basic a matter as nutrition, can vary so widely among individuals, measuring individual well-being by levels of consumption of goods and services obtained confuses the role of com- modities by regarding them as ends in themselves rather than as means to an end. In the case of nutrition, the end is health and what one can do with good health, as well as personal enjoyment and social functioning. Indeed, the capacity to maintain valued social relationships and to network leads to what James Foster and Christopher Handy have termed external capabilities,

20 PART onE Principles and Concepts

which are “abilities to function that are conferred by direct connection or relationship with another person.” But measuring well-being using the concept of utility, in any of its standard definitions, does not offer enough of an improve- ment over measuring consumption to capture the meaning of development.12

As Sen stresses, a person’s own valuation of what kind of life would be worthwhile is not necessarily the same as what gives pleasure to that person. If we identify utility with happiness in a particular way, then very poor people can have very high utility. Sometimes even malnourished people either have a disposition that keeps them feeling rather blissful or have learned to appreci- ate greatly any small comforts they can find in life, such as a breeze on a very hot day, and to avoid disappointment by striving only for what seems attain- able. (Indeed, it is only too human to tell yourself that you do not want the things you cannot have.) If there is really nothing to be done about a person’s deprivation, this attitude of subjective bliss would have undoubted advan- tages in a spiritual sense, but it does not change the objective reality of depri- vation. In particular, such an attitude would not prevent the contented but homeless poor person from greatly valuing an opportunity to become freed of parasites or provided with basic shelter. The functioning of a person is an achievement. Sen provides the example of bicycling “[B]icycling has to be dis- tinguished from possessing a bike. It has to be distinguished also from the happiness generated by [bicycling]... A functioning is thus different both from (1) having goods (and the corresponding characteristics), to which it is poste- rior, and (2) having utility (in the form of happiness resulting from that func- tioning), to which it is, in an important way, prior.”13

To clarify this point, in his acclaimed 2009 book, The Idea of Justice, Sen suggests that subjective well-being is a kind of psychological state of being—a functioning— that could be pursued alongside other functionings such as health and dignity. In the next section, we return to the meaning of happiness as a development outcome, in a sense that can be distinguished from conventional utility.

Sen then defines capabilities as “the freedom that a person has in terms of the choice of functionings, given his personal features (conversion of char- acteristics into functionings) and his command over commodities.”14 Sen’s perspective helps explain why development economists have placed so much emphasis on health and education, and more recently on social inclu- sion and empowerment, and have referred to countries with high levels of income but poor health and education standards as cases of “growth without development.”14a Real income is essential, but to convert the characteristics of commodities into functionings, in most important cases, surely requires health and education as well as income. The role of health and education ranges from something so basic as the nutritional advantages and greater personal energy that are possible when one lives free of parasites to the expanded ability to appreciate the richness of human life that comes with a broad and deep edu- cation. People living in poverty are often deprived—at times deliberately— of capabilities to make substantive choices and to take valuable actions, and often the behavior of the poor can be understood in that light.

For Sen, human “well-being” means being well, in the basic sense of being healthy, well nourished, well clothed, literate, and long-lived, and more broadly, being able to take part in the life of the community, being mobile, and having freedom of choice in what one can become and can do.

Capabilities The freedoms that people have, given their personal features and their command over commodities.

21CHAPTER 1 Introducing Economic Development: A Global Perspective

Development and Happiness

Clearly, happiness is part of human well-being, and greater happiness may in itself expand an individual’s capability to function. As Amartya Sen has argued, a person may well regard happiness as an important functioning for her well-being.15 In recent years, economists have explored the empirical rela- tionship across countries and over time between subjectively reported satis- faction and happiness and factors such as income. One of the findings is that the average level of happiness or satisfaction increases with a country’s average income. For example, roughly four times the percentage of people report that they are not happy or satisfied in Tanzania, Bangladesh, India, and Azerbaijan as in the United States and Sweden. But the relationship is seen only up to an average income of roughly $10,000 to $20,000 per capita, as shown in Figure 1.2.16 Once incomes grow to this point, most citizens have usually escaped extreme poverty. At these levels, despite substantial variations across countries, if inequality is not extreme, a majority of citizens are usually relatively well nourished, healthy, and educated. The “happiness science” findings call into question the centrality of economic growth as an objective for high-income countries. But they also reaf- firm the importance of economic development in the developing world, whether the objective is solely happiness or, more inclusively and persuasively, expanded human capabilities.

0

Russia

Romania Bulgaria

Belarus Latvia

Georgia

Pakistan

Macedonia Turkey Lithuania

Estonia Slovakia

Poland

South Africa Hungary

South Korea Greece Slovenia

Argentina

Czech RepublicChile

Israel Spain Japan

Italy

France Germany

Britain Belgium Australia

Austria Finland

Singapore Sweden

Denmark Ireland Netherlands

Canada Switzerland

Norway USA

Portugal

Albania

Ukraine

Zimbabwe

5,000 10,000 15,000 20,000 25,000 30,000 35,000

Income per head (U.S. $ per year)

100

90

80

70

60

50

40

30

A ve

ra g

e o

f p

e rc

e n

t “

h a

p p

y” a

n d

p e

rc e

n t

“s a

ti sf

ie d

”

•• ••

•• • •••

•

• •

• • •• ••

• • •

•• • •

•• • •

• • •••

• ••• •

• • •• • • • • •

• ••

• • • •• •

• •• • •••••

• • •

•

Moldova

Iran Peru

Azerbaijan

Morocco Egypt

China Philippines Croatia

Brazil Venezuela

El Salvador Colombia

Nigeria

Indonesia

Vietnam

Mexico New Zealand

Uruguay

Dominican Republic

India Jordan

Uganda

Algeria

Tanzania Bangladesh

Source: From Happiness: Lessons from a New Science by Richard Layard, copyright © 2005 by Richard Layard. Used by permission of The Penguin Press, a division of Penguin Group (USA) Inc. and United Agents Ltd. (www.unitedagents.co.uk) on behalf of the author.

FIGUre 1.2 Income and Happiness: Comparing Countries

22 PART onE Principles and Concepts

Not surprisingly, studies show that financial security is only one factor affect- ing happiness. Richard Layard identifies seven factors that surveys show affect average national happiness: family relationships, financial situation, work, com- munity and friends, health, personal freedom, and personal values. In particular, aside from not being poor, the evidence says people are happier when they are not unemployed, not divorced or separated, and have high trust of others in society, as well as enjoy high government quality with democratic freedoms and have reli- gious faith. The importance of these factors may shed light on why the percentage of people reporting that they are not happy or satisfied varies so widely among developing countries with similar incomes. For example, the fraction of people who are not happy and satisfied on average is 4½ times as great in Zimbabwe as in Indonesia, despite somewhat higher incomes in Zimbabwe, and over 3 times as great in Turkey as in Colombia, despite somewhat higher incomes in Turkey at the time of the study. Many opinion leaders in developing nations hope that their societies can gain the benefits of development without losing traditional strengths such as moral values and trust in others—sometimes called social capital.

The government of Bhutan’s attempt to make “gross national happiness” rather than gross national income its measure of development progress has attracted considerable attention.17 Informed by Sen’s work, its indicators extend beyond traditional notions of happiness to include capabilities such as health, education, and freedom. Happiness is not the only dimension of subjective well- being of importance. As the 2010 Stiglitz-Sen-Fitoussi (“Sarkozy”) Commission on the Measurement of Economic Performance and Social Progress put it:

Subjective well-being encompasses different aspects (cognitive evaluations of one’s life, happiness, satisfaction, positive emotions such as joy and pride, and negative emotions such as pain and worry): each of them should be measured sep- arately to derive a more comprehensive appreciation of people’s lives.18

Although, following Sen, what people say makes them happy and satisfied as just one among valued functionings is at best only a rough guide to what people value in life, this work adds new perspectives to the multidimensional meaning of development.

Three Core Values of Development

Is it possible, then, to define or broadly conceptualize what we mean when we talk about development as the sustained elevation of an entire society and social system toward a “better” or “more humane” life? What constitutes the good life is a question as old as philosophy, one that must be periodically reevaluated and answered afresh in the changing environment of world society. The appropriate answer for developing nations today is not necessarily the same as it would have been in previous decades. But at least three basic components or core values serve as a conceptual basis and practical guideline for understanding the inner meaning of development. These core values—sustenance, self-esteem, and freedom—represent common goals sought by all individuals and societies.19 They relate to fundamental human needs that find their expression in almost all societies and cultures at all times. Let us therefore examine each in turn.

Sustenance: the ability to Meet Basic needs All people have certain basic needs without which life would be impossible. These life-sustaining basic

Sustenance The basic goods and services, such as food, clothing, and shelter, that are necessary to sustain an average human being at the bare mini- mum level of living.

Self-esteem The feeling of worthiness that a society enjoys when its social, political, and economic systems and institu- tions promote human values such as respect, dignity, integ- rity, and self-determination.

Freedom A situation in which a society has at its disposal a variety of alterna- tives from which to satisfy its wants and individuals enjoy real choices according to their preferences.

23CHAPTER 1 Introducing Economic Development: A Global Perspective

human needs include food, shelter, health, and protection.20 When any of these is absent or in critically short supply, a condition of “absolute underdevelopment” exists. A basic function of all economic activity, therefore, is to provide as many people as possible with the means of overcoming the helplessness and misery arising from a lack of food, shelter, health, and protection. To this extent, we may claim that economic development is a necessary condition for the improvement in the quality of life that is development. Without sustained and continuous eco- nomic progress at the individual as well as the societal level, the realization of the human potential would not be possible. One clearly has to “have enough in order to be more.”21 Rising per capita incomes, the elimination of absolute poverty, greater employment opportunities, and lessening income inequalities therefore constitute the necessary but not the sufficient conditions for development.22

Self-esteem: to Be a Person A second universal component of the good life is self-esteem—a sense of worth and self-respect, of not being used as a tool by others for their own ends. All peoples and societies seek some basic form of self-esteem, although they may call it authenticity, identity, dignity, respect, honor, or recognition. The nature and form of this self-esteem may vary from society to society and from culture to culture. However, with the proliferation of the “modernizing values” of developed nations, many societies in develop- ing countries that have had a profound sense of their own worth suffer from serious cultural confusion when they come in contact with economically and technologically advanced societies. This is because national prosperity has become an almost universal measure of worth. Due to the significance attached to material values in developed nations, worthiness and esteem are nowadays increasingly conferred only on countries that possess economic wealth and technological power—those that have “developed.”

As Denis Goulet put it, “Development is legitimized as a goal because it is an important, perhaps even an indispensable, way of gaining esteem.”23

Freedom from Servitude: to Be able to Choose A third and final universal value that we suggest should constitute the meaning of development is the concept of human freedom. Freedom here is to be understood in the sense of emancipation from alienating material conditions of life and from social servitude to nature, other people, misery, oppressive institutions, and dog- matic beliefs, especially that poverty is predestination. Freedom involves an expanded range of choices for societies and their members together with a minimization of external constraints in the pursuit of some social goal we call development. Amartya Sen writes of “development as freedom.” W. Arthur Lewis stressed the relationship between economic growth and freedom from servitude when he concluded that “the advantage of economic growth is not that wealth increases happiness, but that it increases the range of human choice.”24 Wealth can enable people to gain greater control over nature and the physical environment (e.g., through the production of food, clothing, and shelter) than they would have if they remained poor. It also gives them the freedom to choose greater leisure, to have more goods and services, or to deny the importance of these material wants and choose to live a life of spiritual contemplation. The concept of human freedom also encompasses various components of political freedom, including personal security, the rule of law, freedom of expression, political participation, and equality of opportunity.25

24 PART onE Principles and Concepts

Although attempts to rank countries with freedom indexes have proved highly controversial,26 studies do reveal that some countries that have achieved high economic growth rates or high incomes, such as China, Malaysia, Saudi Arabia, and Singapore, have not achieved as much on human freedom criteria.

The Central Role of Women

In light of the information presented so far, it should come as no surprise that development scholars generally view women as playing the central role in the development drama. Globally, women tend to be poorer than men. They are also more deprived in health and education and in freedoms in all its forms. More- over, women have primary responsibility for child rearing, and the resources that they are able to bring to this task will determine whether the cycle of trans- mission of poverty from generation to generation will be broken. Children need better health and education, and studies from around the developing world confirm that mothers tend to spend a significantly higher fraction of income under their control for the benefit of their children than fathers do. Women also transmit values to the next generation. To make the biggest impact on develop- ment, then, a society must empower and invest in its women. We will return to this topic in more depth in Chapters 5 through 9 and 15.

The Three Objectives of Development

We may conclude that development is both a physical reality and a state of mind in which society has, through some combination of social, economic, and institutional processes, secured the means for obtaining a better life. Whatever the specific components of this better life, development in all societies must have at least the following three objectives:

1. To increase the availability and widen the distribution of basic life-sustaining goods such as food, shelter, health, and protection

2. To raise levels of living, including, in addition to higher incomes, the provi- sion of more jobs, better education, and greater attention to cultural and human values, all of which will serve not only to enhance material well- being but also to generate greater individual and national self-esteem

3. To expand the range of economic and social choices available to individuals and nations by freeing them from servitude and dependence, not only in rela- tion to other people and nation-states, but also to the forces of ignorance and human misery

1.4 the Future of the Millennium Development Goals

In September 2000, the 189 member countries of the United Nations at that time adopted eight Millennium Development Goals (MDGs), committing themselves to making substantial progress toward the eradication of poverty

Millennium Development Goals (MDGs) A set of eight goals adopted by the United Nations in 2000: to eradicate extreme poverty and hunger; achieve universal primary education; promote gender equality and empower women; reduce child mortality; improve maternal health; combat HIV/AIDS, malaria, and other diseases; ensure environmental sustainability; and develop a global partner- ship for development. The goals are assigned specific tar- gets to be achieved by 2015.

25CHAPTER 1 Introducing Economic Development: A Global Perspective

and achieving other human development goals by 2015. The MDGs are the strongest statement yet of the international commitment to ending global poverty. They acknowledge the multidimensional nature of development and poverty alle- viation; an end to poverty requires more than just increasing incomes of the poor. The MDGs have provided a unified focus in the development community unlike anything that preceded them.27

The eight goals are ambitious: to eradicate extreme poverty and hunger; achieve universal primary education; promote gender equality and empower women; reduce child mortality; improve maternal health; combat HIV/AIDS, malaria, and other diseases; ensure environmental sustainability; and develop a global partnership for development. The goals are then assigned specific targets deemed achievable by 2015 based on the pace of past international development achievements. The goals and targets are found in Table 1.1.

Appropriately, the first MDG addresses the problem of extreme poverty and hunger. The two targets for this goal are more modest: to reduce by half

taBLe 1.1 Millennium Development Goals and targets for 2015

Source: From “Millennium Development Goals” (accessed via www.undp.org). Reprinted with permission from the United Nations Development Programme.

Goals targets

1. Eradicate extreme poverty and hunger • Reduce by half the proportion of people living on less than $1 a day • Reduce by half the proportion of people who suffer from hunger

2. Achieve universal primary education • Ensure that all boys and girls complete a full course of primary schooling 3. Promote gender equality and empower

women • Eliminate gender disparity in primary and secondary education, preferably by 2005,

and at all levels by 2015 4. Reduce child mortality • Reduce by two-thirds the mortality rate among children under 5 5. Improve maternal health • Reduce by three-quarters the maternal mortality ratio 6. Combat HIV/AIDS, malaria, and other

diseases • Halt and begin to reverse the spread of HIV/AIDS • Halt and begin to reverse the incidence of malaria and other major diseases

7. Ensure environmental sustainability • Integrate the principles of sustainable development into country policies and programs; reverse the loss of environmental resources

• Reduce by half the proportion of people without sustainable access to safe drinking water

• Achieve significant improvement in the lives of at least 100 million slum dwellers by 2020

8. Develop a global partnership for development

• Develop further an open, rule-based, predictable, nondiscriminatory trading and financial system; includes a commitment to good governance, development, and poverty reduction—both nationally and internationally

• Address the special needs of the least developed countries; includes tariff and quota free access for least developed countries' exports; enhanced program of debt relief for heavily indebted poor countries (HIPCs) and cancellation of official bilateral debt; and more generous official development assistance (ODA) for countries committed to poverty reduction

• Address the special needs of landlocked countries and small-island developing states • Deal comprehensively with the debt problems of developing countries through

national and international measures in order to make debt sustainable in the long term • In cooperation with developing countries, develop and implement strategies for

decent and productive work for youth • In cooperation with pharmaceutical companies, provide access to affordable essential

drugs in developing countries • In cooperation with the private sector, make available the benefits of new technologies,

especially information and communications

26 PART onE Principles and Concepts

the proportion of people living on less than $1 a day and to reduce by half the proportion of people who suffer from hunger. “Halving poverty” has come to serve as a touchstone for the MDGs as a whole. To achieve this target requires that progress be made on the other goals as well.

As reported by the United Nations Development Programme (UNDP), the halving of global poverty was achieved by 2012, but if current trends con- tinue, not all of the other targets will be achieved; and great regional dispar- ity is obscured when global averages are reported, as East Asia has done far better than sub-Saharan Africa.28 Shockingly, in a world of plenty, the target of cutting the proportion of people who are chronically hungry in half by 2015 is very unlikely to be achieved. Some conditions even worsened after a food price spike in 2008 and thereafter as a result of the global economic cri- sis. And the UNDP highlights that if global trends continue through 2015, the reduction in under-5 mortality will reach roughly one-quarter, far below the target reduction of two-thirds. This means that the target will be missed by 4.4 million avoidable deaths in 2015. Universal primary enrollment will not be achieved unless faster progress can be made in sub-Saharan Africa. Projecting current trends, there will still be 47 million children out of school in 2015. And the UNDP reports that the gap between the current trends and the target of halving poverty represents an additional 380 million people still living on less than $1 a day in 2015.

The goal of ensuring environmental sustainability is essential for secur- ing an escape from poverty. This is immediately seen by looking at two of the targets: reduce by half the proportion of people without access to safe drinking water and achieve significant improvement in the lives of at least 100 million slum dwellers. But more generally, without protecting the environment of the poor, there is little chance that their escape from poverty can be permanent. Finally, the governments and citizens of the rich coun- tries need to play their part in pursuit of the goal of “global partnership for development.”

The MDGs were developed in consultation with the developing coun- tries, to ensure that they addressed their most pressing problems. In addi- tion, key international agencies, including the United Nations, the World Bank, the International Monetary Fund (IMF), the Organization for Economic Cooperation and Development (OECD), and the World Trade Organization (WTO), all helped develop the Millennium Declaration and so have a col- lective policy commitment to attacking poverty directly. The MDGs assign specific responsibilities to rich countries, including increased aid, removal of trade and investment barriers, and eliminating unsustainable debts of the poorest nations.29

However, the MDGs have also come in for some criticism.30 For example, some observers believe that the MDG targets were not ambitious enough, going little beyond projecting past rates of improvement 15 years into the future. Moreover, the goals were not prioritized; for example, reducing hunger may leverage the achievement of many of the other health and education targets. At the same time, although the interrelatedness of development objec- tives was implicit in the MDGs’ formulation, goals are presented and treated in reports as stand-alone objectives; in reality, the goals are not substitutes for

27CHAPTER 1 Introducing Economic Development: A Global Perspective

each other but complements, such as the close relationship between health and education. Further, the setting of 2015 as an end date for the targets could discourage rather than encourage further development assistance if it were not met. Moreover, when the MDGs measure poverty as the fraction of the population below the $1-a-day line, this is arbitrary and fails to account for the intensity of poverty—that a given amount of extra income to a family with a per capita income of, say, 70 cents a day makes a bigger impact on poverty than to a family earning 90 cents per day (see Chapter 5). Other critics have complained that $1 a day is too low a poverty line and about the lack of goals on reducing rich-country agricultural subsidies, improving legal and human rights of the poor, slowing global warming (which is projected to harm Africa and South Asia the most), expanding gender equality, and leveraging the con- tribution of the private sector. While the reasonableness of some of these criti- cisms may be questioned, it should be acknowledged that the MDGs do have some inherent limitations.

With the imminent expiration of the MDGs, the UN coordinated global efforts to launch its successor, Sustainable Development Goals (SDGs), with the May 2013 agenda-setting report of the High-Level Panel of Eminent Persons on Development Agenda.31 This highly diverse panel of political leaders from every part of the world agreed upon a bold approach that is expected to sub- stantially influence the eventual shape of the post-2015 agenda, the SDGs. The panel repeatedly stressed that it is “a universal agenda” for all countries, developed as well as developing and without exceptions, “to be driven by five big, transformative shifts.” These universal shifts are:

1. Leave no one behind—to move “from reducing to ending extreme poverty, in all its forms;” in particular, to “design goals that focus on reaching excluded groups.”

2. Put sustainable development at the core, “to integrate the social, economic, and environmental dimensions of sustainability.”

3. Transform economies for jobs and inclusive growth, while moving to sus- tainable patterns of work and life.

4. Build peace and effective, open, and accountable institutions for all, which “encourage the rule of law, property rights, freedom of speech and the media, open political choice, access to justice, and accountable govern- ment and public institutions.”

5. Forge a new global partnership so that each priority should involve gov- ernments and also others, including people living in poverty, civil society and indigenous and local communities, multilateral institutions, business, academia, and philanthropy.

The High-Level Panel also agreed on well-recognized and illustrative uni- versal goals and national targets for the SDGs, including an outright end by 2030 of poverty, hunger, child marriage, and preventable under-5 deaths, and specific targets on stunting, social protection coverage, and maternal mortality. The debate will be lively throughout 2014 and 2015.

Sector A subset (part) of an economy, with four usages in economic development: technology (modern and traditional sectors); activity (industry or product sectors); trade (export sector); and sphere (private and public sectors)

28 PART onE Principles and Concepts

1.5 Conclusions

Development economics is a distinct yet very important extension of both traditional economics and political economy. While necessarily also concerned with efficient resource allocation and the steady growth of aggregate output over time, development economics focuses primarily on the economic, social, and institutional mechanisms needed to bring about rapid and large-scale improvements in standards of living for the masses of poor people in develop- ing nations. Consequently, development economics must be concerned with the formulation of appropriate public policies designed to effect major eco- nomic, institutional, and social transformations of entire societies in a very short time.

As a social science, economics is concerned with people and how best to provide them with the material means to help them realize their full human potential. But what constitutes the good life is a perennial question, and hence economics necessarily involves values and value judgments. Our very concern with promoting development represents an implicit value judgment about good (development) and evil (underdevelopment). But development may mean dif- ferent things to different people. Therefore, the nature and character of develop- ment and the meaning we attach to it must be carefully spelled out. We did this in section 1.3 and will continue to explore these definitions throughout the text.

The central economic problems of all societies include traditional ques- tions such as what, where, how, how much, and for whom goods and services should be produced. But they should also include the fundamental question at the national level about who actually makes or influences economic decisions and for whose principal benefit these decisions are made. Finally, at the inter- national level, it is necessary to consider the question of which nations and which powerful groups within nations exert the most influence with regard to the control, transmission, and use of technology, information, and finance. Moreover, for whom do they exercise this power?

Any realistic analysis of development problems necessitates the supple- mentation of strictly economic variables such as incomes, prices, and savings rates with equally relevant noneconomic institutional factors, including the nature of land tenure arrangements; the influence of social and class strati- fications; the structure of credit, education, and health systems; the organi- zation and motivation of government bureaucracies; the machinery of public administrations; the nature of popular attitudes toward work, leisure, and self-improvement; and the values, roles, and attitudes of political and economic elites. Economic development strategies that seek to raise agricultural output, create employment, and eradicate poverty have often failed in the past because economists and other policy advisers neglected to view the economy as an interdependent social system in which economic and noneconomic forces are continually interacting in ways that are at times self-reinforcing and at other times contradictory. As you will discover, underdevelopment reflects many individual market failures, but these failures often add up to more than the sum of their parts, combining to keep a country in a poverty trap. Govern- ment can play a key role in moving the economy to a better equilibrium, and in many countries, notably in East Asia, government has done so; but all too often government itself is part and parcel of the bad equilibrium.

29CHAPTER 1 Introducing Economic Development: A Global Perspective

Achieving the Millennium Development Goals will be an important mile- stone on the long journey to sustainable and just development. Although prog- ress has been substantial, many of the interim targets remain unachieved—nor do the MDGs include all of the critical objectives of development. The emerging Sustainable Development Goals, planned as the MDGs’ successor after 2015, will be even more ambitious, including the full eradication of extreme poverty.

Despite the great diversity of developing nations—some large, others small; some resource-rich, others resource-barren; some subsistence economies, others modern manufactured-good exporters; some private-sector oriented, others to a large degree run by the government—most share common problems that define their underdevelopment. We will discuss these diverse structures and common characteristics of developing countries in Chapter 2.

The oil price shocks of the 1970s, the foreign-debt crisis of the 1980s, and the twenty-first-century concerns with economic globalization, economic imbalances and financial crises, global warming, and international terrorism have underlined the growing interdependence of all nations and peoples in the international social system. What happens to life in Caracas, Karachi, Cairo, and Kolkata will in one way or another have important implications for life in New York, London, and Tokyo. It was once said that “when the United States sneezes, the world catches pneumonia.” A more fitting expression for the twenty-first century would perhaps be that “the world is like the human body: If one part aches, the rest will feel it; if many parts hurt, the whole will suffer.”

Developing nations constitute these “many parts” of the global organism. The nature and character of their future development should therefore be a major concern of all nations irrespective of political, ideological, or economic orientation. There can no longer be two futures, one for the few rich and the other for the very many poor. In the words of a poet, “There will be only one future—or none at all.”

3030

Case Study 1

Progress in the Struggle for More Meaningful Development: Brazil

there are two faces of development in Brazil. World-competitive industry coexists with stag- nant, protected sectors. Modern agriculture coex- ists with low-productivity traditional practices. But Brazil is in the midst of a spurt of economic development that might herald a lasting transfor- mation for a country often considered synonymous with inequality and unmet potential. Economic growth has returned, health and education have improved markedly, the country’s democratization has proved durable, and inequality—among the highest in the world—has at long last started to fall. But there is still a long way to go to achieve genuine development in Brazil. Growth remains vulner- able to world commodity prices, and social progress remains tenuous as revealed by the widespread pro- test demonstrations that erupted in 2013.

Many Brazilians have been frustrated with the uneven pace of development and are known for telling self-deprecating jokes such as “Brazil is the country of the future—and always will be.” Brazil has even been cited as an example of a country that has experienced “growth without development.” But despite huge inequities, Brazil has made eco- nomic and social progress. Extremely high economic inequality and social divisions do pose a serious threat to further progress in Brazil. But there are growing reasons to hope that Brazil may overcome its legacy of inequality so that the country may yet join the ranks of the developed countries.

Brazil is of special interest in part because its growth performance from the 1960s through the early 1980s was the best in Latin America, with at least some parallels with East Asian export policy and performance, although Brazil had a larger role for state-owned enterprises, much lower education

and other social expenditures, and much higher inflation.

Brazil’s performance is followed widely in the devel- oping world, as it is the largest and most populous country in Latin America; with close to 200 million people, it is the world’s fifth-largest country in both area and population. Brazil is consolidating its role as the lead country in the Latin America and Carib- bean region; it is a key member of the G20 leading economies; and one of a group of developing coun- tries pushing for fairer international trade rules. It is one of four influential countries referred to by the media and financial analysts as the “BRICs” (Brazil, Russia, India, and China, often expanded to include South Africa).

Although over two decades of military rule ended in Brazil in 1985, an ongoing debt crisis, years of stagnant incomes, and extremely high inflation followed. It took drastic policies to reduce inflation, and incomes continued to stagnate in the aftermath. The 1980s and the 1990s have been described as “lost decades” for development. So the recent signs of palpable progress, especially since about 2004, have been welcomed with relief and growing enthusiasm among many Brazilians. Although the country remains politically divided between the center-left and the center-right, a strik- ing convergence has been achieved on policies agreed to be necessary for equitable and sustained growth, ranging from active poverty reduction pro- grams to relatively orthodox monetary policies. The economy has been growing more rapidly (if inconsistently), in part due to commodity exports to China, from soybeans to iron ore. One persistent worry is whether the economy could continue to grow rapidly if commodity prices, which have been

31

31

much higher in recent years, revert to their very long-term trends for decline (see Chapter 12), or the slowing growth in China curtails demand for Brazil’s products—both genuine worries by 2013. High crime remains a problem, especially in the favelas (slums).

But despite the nation’s early and now resumed growth, other indicators of development in Brazil lagged, eventually undermining growth prospects. Benefiting from much higher incomes than Central American countries and spared the destructiveness of civil war, Brazil, it would seem, should have been in a much better position to fight extreme poverty and improve economic equity and social indicators. Instead, despite recent improvement, the country has continued to see a higher percentage of its pop- ulation in poverty than would be expected for an upper-middle-income country. Brazil remains one of the countries with the highest levels of inequality in the world. So how should Brazil’s development performance be evaluated and future priorities chosen?

Income and Growth Growth is generally necessary, though not suffi- cient, for achieving development. In 2011, Brazil’s per capita income was close to $11,000, still well under a quarter of that of the United States but more than nine times greater than that of Haiti (World Bank data).

Growth has been erratic, with substantial swings over time. Data for growth of gross domestic prod- uct (GDP) per capita are sometimes presented for the periods 1965–1990, when for Brazil it was 1.4%, and for 1990–2000, when it was 1.5%. This appears to suggest a remarkable stability. But the former figures combine the booming years from 1967 to 1980 and Brazil’s “lost decade of development” of the 1980s. Nevertheless, performance through this period was still better than most other countries of Latin America. In 2000–2011, average annual per capita growth rose to about 2.8% (World Bank data). But wide swings continue, with a spike close to 7% in 2010 slowing to a near standstill by the end of 2012.

Brazil has had an export policy stressing incentives for manufacturing exports, as well as protections for domestic industry, with numerous parallels with Taiwan and South Korea in their earlier formative

stages (see Chapter 12). Its percentage share of manu- factured exports in total exports grew dramatically, reaching 57% in 1980, although it dropped dramati- cally during the lost decade of the 1980s. Although the share of manufactured exports increased again, reaching a new peak of 58% in 2000, it has fallen steadily since, to 45% in 2008, and by 2011, this figure had fallen to just 34% (World Bank data). Although part of this decline reflected an increase in commod- ity prices, it is still a striking reversal that probably increases the vulnerability of the Brazilian economy (see Chapter 12). Brazil must decide whether to respond to its good fortune of a period of high commodity prices as a spur to action or an excuse for complacency.

Brazil’s prolonged status as a highly indebted country (see Chapter 13) was a substantial drag on growth performance, as were continued prob- lems with infrastructure. Recently, however, the Industrial, Technological and Foreign Trade Policy (PITCE) program has been actively working to upgrade the quality and competitiveness of Brazilian industry.

High and growing taxes may have also slowed formal-sector employment growth. The overall tax burden increased from about 25% of gross national income to nearly 40% in the decade from 1993 to 2004, before leveling off (it was about 38% in 2012). Payroll taxes are high and as many as half of Brazil’s labor force now works in the informal sector, where taxes may be avoided (and labor rights and regula- tion circumvented).

However, Ricardo Hausmann, Dani Rodrik, and Andrés Velasco argue that Brazil does not lack for productive investment ideas, nor is concern about government behavior the factor holding back investment. Using their decision tree framework to identify the most binding constraints on economic growth (see Chapter 4), Hausmann, Rodrik, and Velasco argue that Brazil has high returns to invest- ment and is most constrained by a lack of savings to finance its productive opportunities at reasonable interest rates. In raising domestic savings, Haus- mann has emphasized the importance of “creating a financially viable state that does not over-borrow, over-tax or under-invest.”

Technology transfer is critical to more rapid growth, competing internationally, and beginning to catch up with advanced countries. Brazil has

32

made notable progress. The country is viewed as being at the cutting edge of agricultural research and extension in commercially successful export crops such as citrus and soybeans. After a disastrous attempt to protect the computer industry in the 1980s was abandoned, Brazil has begun to see the expan- sion of a software industry, as also seen in India. But Brazil has not absorbed technology to the degree that East Asian countries have.

Social Indicators Brazil’s human development statistics compare unfavorably with many other middle-income coun- tries such as Costa Rica and quite a few low-income countries, let alone with the advanced industrial- ized countries. As of 2007, Brazil ranked just 85th on the United Nations Development Programme’s 2013 Human Development Index (explained in Chapter 2), eight positions lower than would be predicted by its income, and below such countries as Peru, Mauritius, and Azerbaijan.

In Brazil, life expectancy at birth in 2011 was 73 years, compared with 81 in South Korea. Brazil’s under-5 mortality rate is 16 per 1,000 live births, an impressive improvement from its 2000 rate of 36 per 1,000 but still high compared with 10 in similar- income Costa Rica and just 5 in Korea (World Bank data). But about 7% of all children under the age of 5 still suffer from malnutrition in Brazil (World Bank data).

Brazil suffers from a high incidence of child labor for its income level, as a World Bank study and reports by the International Labor Office have underlined. As many as 7 million children still work in Brazil, despite the country’s having officially made the eradication of child labor a priority. (For an analysis of the problems of child labor and appro- priate child labor policies, see Chapter 8.) In the education sphere, Brazil’s officially reported adult literacy rate has now risen to 90% (independent observers have concluded that Brazil’s effective liter- acy is under 50%), while that of similar-income Costa Rica is 96%. Helping explain this difference, in Costa Rica, six years of school attendance are mandatory, and 99% attendance is reported.

The UNDP concluded that

the unequal distribution of social spending is no doubt a major factor in maintaining inequality and thus poverty.…The bulk of the benefits go to the

middle classes and the rich. Close to a third of the poorest fifth of the population does not attend pri- mary school. But the sharpest differences show up in secondary and tertiary education. More than 90% of the poorest four-fifths of the population do not at- tend secondary school, and practically none make it to universities. Only primary schools end up being relatively targeted to the poor, not because the gov- ernment succeeds in targeting resources, but because richer households send their children to private schools. Public expenditures on secondary and ter- tiary education are very badly targeted to the poor. For scholarships—chiefly to graduate students— four-fifths of the money goes to the richest fifth of the population.

In fact, with public universities offering free tuition to mostly high-income undergrads as well as grad students, the distortion is even greater. Moreover, corruption and waste limit the effectiveness of gov- ernment expenditures. And the quality of primary schools in poor areas remains low.

So while the persistence of poverty in Brazil is undoubtedly due in part to mediocre growth rela- tive to East Asia or to Brazil’s potential, the most important explanation is the highly concentrated distribution of income, worsened by inequitable social spending.

Development depends on a healthy, skilled, and secure workforce. Ultimately, a slower improve- ment in health, education, and community devel- opment can feed back to a slower rate of growth, a process that has plagued Brazilian development. A hopeful sign is the role played now in Brazil by a free press, strengthened basic rights, and a very active but peaceful political competition. These elements can be a precursor of expanded capabilities in Amartya Sen’s analysis.

Poverty Perhaps the most important social indicator is the extent of extreme poverty among a country’s people. Poverty has been high in Brazil for an upper- middle-income country. There has been progress; a World Bank study found that Brazil’s average per capita income grew by 220% in the high-growth years from 1960 to 1980, with a 34% decline in the share of the poor in the population. On the other hand, similarly sized Indonesia grew 108% from 1971 to 1987, with a 42% decline in poverty inci- dence. And some of the ground gained on poverty

was subsequently lost in Brazil in the 1980s and 1990s. According to World Bank estimates, in 2009, some 10.8% of the population of Brazil lived on less than $2 per day. And according to the most recent data, 6.1% actually lived in extreme poverty, with incomes below $1.25 per day (World Bank, 2013 World Development Indicators), worse than some low-income countries such as Sri Lanka. But this may actually be an underestimate. According to a Brazilian government research institute cited by the United Nations Development Programme, an even more shocking 15% of Brazilians have incomes of less than $1 a day. However, poverty is now fall- ing, and the recent Bolsa Familia (family stipend) government program has received high marks for addressing poverty through its “conditional cash transfers” of resources to poor families, provided that they keep children vaccinated and in school; it is similar to the Mexican Progresa/Oportunidades program that is the subject of the end-of-chapter case study for Chapter 8. It must also be mentioned that physical security remains a pressing problem in Brazil, with violent gangs having extensive sway. This problem can have the greatest negative impact on people living in poverty.

Inequality For decades, Brazil’s inequality in income (as well as in land and other assets) has ranked among the worst in the world. High inequality not only pro- duces social strains but can also ultimately retard growth, as examined in detail in Chapter 5. The degree of income inequality in Brazil is reflected in the low share of income going to the bottom 60% and the high share to the top 10% of the population, as seen in the following income distribution data for Brazil (2009 survey data, from the 2013 World Devel- opment Indicators):

As these figures show, the top 10% of income earn- ers receive about 43% of national income, while the

bottom 40% receive just 10%. The UNDP concludes that high inequality is the reason for the high level of extreme poverty and the very slow rate of poverty reduction. Inequality in assets is also high. In recent years, inequality in Brazil has moderated some- what, although it remains among the highest in the world. In addition to Bolsa Familia and other social program innovations, Brazilian analysts generally conclude that a recent increase in (and enforcement of) the minimum wage also has reduced inequality; this has had, wide impact as many local government workers receive the minimum wage.

Land Reform Land is very unequally distributed in Brazil, and there is both an efficiency and a social equity case for land reform (a subject discussed in Chapter 9). But land reform has been repeatedly blocked in Brazil by the political power of large plantation owners (fazenderos). In response, impoverished farmers in the “landless movement,” or MST, have increasingly seized land, often arable but unused land within large plantations. Thousands of families have taken part. Farmers have also settled in fragile rain forest areas, finding them- selves unable to acquire land in areas that are more agriculturally suitable and less ecologically sensi- tive. In response, the government has initiated a land reform program, but the results to date have been modest in relation to the scope of the problem.

Sustainability of Development As described in Chapter 10, growth that relies on running down the natural environment is con- trasted with sustainable development, which pre- serves the ecology on which future income and people’s health vitally depend. But Brazilians across the political spectrum appear determined not to acknowledge destruction of forests as a genuine or pressing problem. Deforestation of the Brazilian Amazon rain forest displays conflicts between short- and long-term development goals and the conse- quences of huge inequality and state intervention on behalf of the rich. Despite their destructiveness, economic activities in the Amazon often benefited in the past from ill-conceived subsidies, now curtailed. Grandiose showcase development projects and schemes, such as subsidized ore mining, charcoal- consuming industries, and cattle ranching, were car- ried out on a large scale.

Fraction of Population Share received (%)

Lowest 10% 0.8 Lowest 20% 2.9 Second 20% 7.1 Third 20% 12.4 Fourth 20% 19.0 Highest 20% 58.6 Highest 10% 42.9

33

34

The encouragement of rain forest settlement seemed to be a politically inexpensive alternative to land reform. In the end, the best lands became concentrated in the hands of large, powerful farm- ers. Rights of indigenous peoples were flagrantly violated, with some terrible atrocities committed by settlers. Ecological campaigners and activists among rubber-tappers whose livelihoods were threatened were attacked and sometimes murdered. In the meantime, much of these fragile lands appears to have become irreversibly degraded. Many of the subsidies have now been withdrawn, and at least some protections and “extractive reserves” have been put in place, but rain forest destruction is hard to reverse. Forest management in other tropical rain forests has led to a rapid growth in ecotourism and very high, profitable, and sustainable fruit yields. Products that can be harvested without serious eco- logical disruption include fibers, latex, resins, gums, medicines, and game. However, it is clear that this cannot protect land on the vast scale at risk. Because the rest of the world benefits from Brazil’s rain forests through prevention of global warming, eco- logical cleansing, and the irreplaceable biodiversity needed for future antibiotics and other medicines and goods, the international community should be prepared to pay something to ensure its continua- tion, such as paying forest dwellers to preserve and protect natural resources. Financial support for land reform outside sensitive areas is one clear direction.

Problems of Social Inclusion Few discussions about poverty in Brazil pay much attention to race. But about half of the population of Brazil is of African or mulatto heritage. As a result, it is sometimes noted that Brazil is the world’s largest black nation after Nigeria. And most of the poor in Brazil are black or mulatto. Although racial discrimination is a crime in Brazil, no one has ever been sent to jail for it. According to one estimate, the average black worker receives only 41% of the salary of the average white worker. Most of the millions of Brazilians living in the worst favelas, or shantytown slums, are black. The endemic extreme poverty of the Northeast, which has lagged develop- ment standards of the Southeast for decades, afflicts indigenous and mulatto populations. Although the Northeast has only about 30% of Brazil’s population, an estimated 62% of the country’s extreme poor live

in the region. Black representation in government is shockingly rare, even in the states where nonwhites make up a majority of the population. Univer- sity places are overwhelmingly claimed by whites. Some progress has been made, but Brazil may need a stronger movement comparable to the U.S. civil rights struggle of the 1960s. But in the absence of overt Jim Crow laws, it is sometimes hard to iden- tify the appropriate target. Some form of meaningful affirmative action may be the only way to begin to overcome the problem.

Conclusion It might be most accurate to say that Brazil has experienced some economic growth without as much social development as could be expected from its income level, rather than the more blan- keting “growth without development,” which applies better to countries such as Pakistan, Gabon, and Equatorial Guinea. But continuing racial dis- parities, unjust treatment of indigenous peoples, lack of access of the poor to fertile land, extremely high inequality and surprisingly high poverty for its income level, and the danger that growth will prove ecologically unsustainable all mean that Brazil will have to continue its recent efforts to make social inclusion and human development, as well as environmental sustainability, top priori- ties if it is to resume and maintain rapid economic growth, let alone achieve true multidimensional development.

Part of the explanation for high rates of income poverty and poor social indicators in Brazil is the relatively slower growth that has prevailed since the early 1980s, excepting its recent spurt. But a major explanation is that government social spending on health, education, pensions, unemployment ben- efits, and other transfers are going to the well-off, fre- quently to those in the top 20% of income distribution. Government policy has often had the effect of wors- ening inequality rather than softening it. The Bolsa Familia program is an important recent exception that has made a substantial impact in Brazil. Bolsa Familia transfers income to poor families on the condition that their children stay in school, thus providing current consumption as well as the potential of future higher earnings for families trapped in chronic poverty.

In November 2002, the left-leaning labor leader Luiz Inacio Lula da Silva, known universally as

Lula, was elected president of Brazil on a platform promising greater equity. This generated a lot of excitement in the country, with renewed hopes for greater social inclusion; his first term saw some renewal of growth and a greater public policy focus on poverty, with some improvements in the favelas and better rural nutrition, for example, but the rate of progress on social inclusion was disappointingly slow for many Brazilians. Lula was reelected in 2006, and the general view is that the following four years went well, with steadier growth and signifi- cant improvements in many social indicators; and Lula’s Worker’s Party successor, Dilma Rousset— who was imprisoned and tortured during military rule—won the 2010 presidential election to become the first woman to lead Brazil. But substantial unrest emerged in the country in 2013, and many questions remain. Many Brazilians find the contrast between slum neighborhoods and gleaming new sports facilities that can be seen from them a grating reminder of dubious development priorities. Can steady progress be made on the racial divide, physi- cal security, environmental decay, poverty, inequality, high borrowing costs, needed diversification of exports, and high and inefficient government spending? If so, the outlook for Brazil is bright.

Sources Anderson, Anthony B. “Smokestacks in the rainfor-

est: Industrial development and deforestation in the Amazon basin.” World Development 18 (1990): 1191–1205.

Anderson, Anthony B., ed. Alternatives to Deforestation. New York: Columbia University Press, 1990.

Assunção, Juliano, “Land reform and landholdings in Brazil,” UNU-WIDER Research Paper No. 2006/137, November 2006.

Baer, Werner. The Brazilian Economy: Growth and Development. Boulder, Colo.: Rienner, 2008.

Bank Information Center. Funding Ecological and Social Destruction: The World Bank and the IMF. Washington, D.C.: Bank Information Center, 1990.

Bauman, Renato, and Helson C. Braga. “Export financing in the LDCs: The role of subsidies for export performance in Brazil.” World Develop- ment 16 (1988): 821–833.

Binswanger, Hans P. “Brazilian policies that encourage deforestation in the Amazon.” World Development 19 (1991): 821–829.

Dinsmoor, James. Brazil: Responses to the Debt Crisis. Washington, D.C.: Inter-American Development Bank, 1990.

Downing, Theodore E., Susanna B. Hecht, and Henry A. Pearson, eds. Development or Destruction? The Conversion of Tropical Forest to Pasture in Latin America. Boulder, Colo: Westview Press, 1992.

The Economist, Special Report on Brazil (November 14, 2009): http://www.economist.com/specialreports /displayStory.cfm?story_id=E1_TQRNJQRV.

Erber, Fabio Stefano. “The development of the elec- tronics complex and government policies in Brazil.” World Development 13 (1985): 293–310.

Fields, Gary. Poverty Inequality and Development. New York: Cambridge University Press, 1980.

Hausmann, Ricardo, Dani Rodrik, and Andrés Velasco. “Growth diagnostics.” In One Economics, Many Recipes: Globalization, Institutions, and Economic Growth, by Dani Rodrik. Princeton, N.J.: Princeton University Press, 2007.

Hausmann Ricardo. “In search of the chains that hold Brazil back.” October 31, 2008. http://papers. ssrn.com/sol3/papers.cfm?abstract_id=1338262.

INCRA (Brazilian agency for land reform), http:// www.incra.gov.br.

Sercovich, Francisco Colman. “Brazil.” World Development 12 (1984): 575–600.

Siddiqi, Faraaz, and Harry Anthony Patrinos. Child Labor: Issues, Causes, and Interventions. World Bank, n.d. http://www.worldbank.org.

United Nations Development Programme. Human Poverty Report, 2000–2009 (annual). New York: United Nations, 2000–2009.

World Bank. Eradicating Child Labor in Brazil. Washington, D.C.: World Bank, 2001.

———. Global Monitoring Report, 2007. Washington, D.C.: World Bank, 2007. (See in particular Table A.1, p. 226.)

———. World Development Indicators, 2010. Wash- ington, D.C.: World Bank, 2010.

Yusuf, Shahid. Globalization and the Challenge for Developing Countries. Washington, D.C.: World Bank, 2001. ■

Notes: The Instituto Brasileiro de Geografia e Estatística (IBGE) provides data on Brazil that supplements that found in the World Development Indicators and other international sources. See www.ibge.gov.br/english. This case study benefits greatly from annual exchanges on evolving policies and conditions with Brazilian civil servants.

35

36 PART onE Principles and Concepts

Concepts for review

Absolute Poverty Attitudes Capabilities Developing countries Development Development economics Freedom Functionings Globalization

Gross domestic product Gross national income (GNI) Income per capita Institutions Less developed countries Millennium Development Goals

(MDGs) More developed countries (MDCs) Political economy

Sector Self-esteem Social system Subsistence economy Sustenance Traditional economics Values

All boldfaced terms that appear in the text are listed in Concepts for Review at the end of each chapter. A glossary at the back of the book provides quick-reference definitions for these and other, more general economic concepts.

Questions for Discussion

1. Why is economics central to an understanding of the problems of development?

2. Is the concept of the developing world a useful one? Why or why not?

3. What do you hope to gain from this course on development economics?

4. Briefly describe the various definitions of the term development encountered in the text. What are the strengths and weaknesses of each approach? Do you think that there are other dimensions of devel- opment not mentioned in the text? If so, describe them. If not, explain why you believe that the text description of development is adequate.

5. Why is an understanding of development cru- cial to policy formulation in developing nations? Do you think it is possible for a nation to agree on a rough definition of development and orient its strategies accordingly?

6. Why is a strictly economic definition of develop- ment inadequate? What do you understand economic

development to mean? Can you give hypothetical or real examples of situations in which a country may be developing economically but may still be underdeveloped?

7. How does the concept of “capabilities to func- tion” help us gain insight into development goals and achievements? Is money enough? Why or why not?

8. What forces may be at work in giving the Millen- nium Development Goals such a high profile in international economic relations?

9. What critical issues are raised from the examina- tion of development problems and prospects facing Brazil?

10. It has been said that ending extreme poverty and achieving genuine development are possible but not inevitable and that this gives the study of economic development its moral and intellectual urgency. What is meant by this? Comment and evaluate.

notes

1. United Nations Development Programme, “The Rise of the South: Human Progress in a Diverse World,” Human Development Report 2013. New York:

United Nations Development Programme, 2003. Poverty figures are drawn from the World Bank.

37CHAPTER 1 Introducing Economic Development: A Global Perspective

2. “Voices of the Poor” boxed quotations throughout the text are for the most part drawn from the World Bank “Voices of the Poor” Web site, http://www .worldbank.org/poverty/voices/overview.htm. The Voices project was undertaken as background for the World Development Report, Attacking Poverty. The results were published for the World Bank by Oxford University Press in a three-volume series titled Can Anyone Hear Us? Crying Out for Change, and From Many Lands, edited by Deepa Narayan.

3. See Paul Krugman, “Toward a counter-counter- revolution in development theory,” Proceedings of the World Bank Annual Conference on Development Economics, 1992 (Washington, D.C.: World Bank, 1993), p. 15. See also Syed Nawab Haider Naqvi, “The significance of development economics,” World Development 24 (1996): 975–987.

4. For a classic argument on the role of values in de- velopment economics, see Gunnar Myrdal, The Challenge of World Poverty (New York: Pantheon, 1970), ch. 1. A more general critique of the idea that economics can be “value-free” is to be found in Robert Heilbroner’s “Economics as a ‘value-free’ science,” Social Research 40 (1973): 129–143, and his Behind the Veil of Economics (New York: Norton, 1988). See also Barbara Ingham, “The meaning of development: Interactions between ‘new’ and ‘old’ ideas,” World Development 21 (1993): 1816–1818; Paul P. Streeten, Strategies for Human Development (Copenhagen: Handelshøjskolens Forlag, 1994), pt. 1; Selo Soemardjan and Kenneth W. Thompson, eds., Culture, Development, and Democracy (New York: United Nations University Press, 1994); and Mozaffar Qizilbash, “Ethical development,” World Develop- ment 24 (1996): 1209–1221.

5. Soedjatmoko and Anne Elizabeth Murase, The Primacy of Freedom in Development (Lanham, Md.: University Press of America, 1985), p. 11.

6. Dudley Seers, “The meaning of development,” paper presented at the Eleventh World Confer- ence of the Society for International Develop- ment, New Delhi (1969), p. 3. See also Richard Brinkman, “Economic growth versus economic development: Toward a conceptual clarification,” Journal of Economic Issues 29 (1995): 1171–1188; and P. Jegadish Gandhi, “The concept of development: Its dialectics and dynamics,” Indian Journal of Applied Economics 5 (1996): 283–311.

7. Denis Goulet, The Cruel Choice: A New Concept in the Theory of Development (New York: Atheneum, 1971), p. 23. Reprinted with permission from Ana Maria Goulet.

8. Amartya Sen, “Development Thinking at the Be- ginning of the 21st Century.” In Economic and Social Development in the XXI Century. Emmerij, Luis (Ed.) Inter-American Development Bank and Johns Hopkins University Press, Washington, D.C. [Also available as LSE working paper, Copyright Amartya Sen, at http://eprints.lse.ac.uk/6711/.] See also Sen, Commodities and Capabilities (Amsterdam: Elsevier, 1985). We thank Sabina Alkire and James Foster for their helpful suggestions on updating this section for the Twelfth Edition to reflect Professor Sen’s latest thinking on his capability approach, including ideas reflected in his recent book, The Idea of Justice.

9. Amartya Sen, Commodities and Capabilities, p. 12.

10. Sen, Commodities and Capabilities, pp. 25–26; and Development as Freedom, pp. 70–71.

11. Sen, Commodities and Capabilities, pp. 25–26. From Commodities and Capabilities by Amartya Sen. Copyright © 1999 by Amartya Sen. Reprinted with permission.

12. Ibid., p. 21. Sen points out that even if we iden- tify utility with “desire fulfillment,” we still suf- fer from the twin defects of “physical-condition neglect” and “valuation neglect.” He notes that “valuing is not the same thing as desiring.” Ig- noring a person’s objectively deprived physical condition just because the person considers this subjectively unimportant yields an obviously de- fective measure of well-being. The paper by Fos- ter and Handy is “External Capabilities,” in Argu- ments for a Better World: Essays in Honor of Amartya Sen, eds. Kaushik Basu and Ravi Kanbur (Oxford: Oxford University Press, 2008).

13. Ibid., pp. 10–11. From Commodities and Capabilities by Amartya Sen. Copyright © 1999 by Amartya Sen. Reprinted with permission.

14. Amartya Sen, Commodities and Capabilities, p. 13.

14a. See, for example, William Easterly, “The politi- cal economy of growth without development: A case study of Pakistan,” in In Search of Prosperity: Analytic Narratives on Economic Growth, ed. Dani Rodrik (Princeton, N.J.: Princeton University Press, 2003).

38 PART onE Principles and Concepts

15. Sen, Commodities and Capabilities, p. 52.

16. See Richard Layard, Happiness: Lessons from a New Science (New York: Penguin, 2005), esp. pp. 32–35 and 62–70. The data on happiness and satisfaction are based on an average of the two responses. For more on the underlying data and analysis, see http://cep.lse.ac.uk/layard/annex.pdf. For a critique of some aspects of this research, see Martin Wolf, “Why progressive taxation is not the route to happiness,” Financial Times, June 6, 2007, p. 12. For an excellent review of the liter- ature through 2010 that puts the data and their interpretation in useful perspective, see Carol Graham, Happiness around the World: The Paradox of Happy Peasants and Miserable Millionaires (New York: Oxford University Press, 2010).

17. For the revised happiness index formula be- ing considered in Bhutan, see http://www .grossnationalhappiness.com/gnhIndex/introduction GNH.aspx. The formula is closely related to the Alkire-Foster Multidimensional Poverty Index, introduced in Chapter 5. For earlier background see Andrew C. Revkin, “A new measure of well- being from a happy little kingdom,” New York Times, October 4, 2005, http://www.nytimes. com/2005/10/04/science/04happ.html.

18. Commission on the Measurement of Economic Performance and Social Progress, p. 16, http:// www.stiglitz-sen-fitoussi.fr/documents/rapport_ anglais.pdf accessed November 12, 2010.

19. See Goulet, Cruel Choice, pp. 87–94.

20. For a description of the “basic needs” approach, see Pradip K. Ghosh, ed., Third World Development: A Basic Needs Approach (Westport, Conn.: Green- wood Press, 1984).

21. Goulet, Cruel Choice, p. 124.

22. For an early attempt to specify and quantify the concept of basic needs, see International Labor Organization, Employment, Growth, and Basic Needs (Geneva: International Labor Organization, 1976). A similar view with a focus on the notion of entitlements and capabilities can be found in Amartya Sen, “Development: Which way now?” Economic Journal 93 (1983): 754–757. See also United Nations Development Programme, Human Development Report, 1994 (New York: Oxford University Press, 1994).

23. Goulet, Cruel Choice, p. 90. For an even more provocative discussion of the meaning of indi- vidual self-esteem and respect in the context of Latin American development, see Paulo Freire, Pedagogy of the Oppressed (New York: Continuum, 1990).

24. W. Arthur Lewis, “Is economic growth desirable?” in The Theory of Economic Growth (London: Allen & Unwin, 1963), p. 420. For an outstanding and thoughtful analysis of the importance of freedom in development by a leading Developing World intellectual, see Soedjatmoko, Primacy of Freedom. See also Sen, Development as Freedom.

25. For a “political freedom index,” see United Nations Development Programme, Human Development Re- port, 1992 (New York: Oxford University Press, 1992), pp. 20, 26–33. The Heritage Foundation and the Wall Street Journal produce an annual “Index of Economic Freedom.” For 2014 rankings of 165 countries from “free” to “repressed,” see http://www.heritage .org/index/.

26. For a commentary from the UNDP on why its free- dom index was discontinued, see United Nations Development Programme, Human Development Report, 2000, pp. 90–93, esp. box 5.2, at http://hdr. undp.org/docs/statistics/understanding/re- sources/HDR2000_5_2_freedom_indices.pdf.

27. United Nations Development Programme, Human Development Report, 2003—Millennium Develop- ment Goals: A Compact among Nations to End Human Poverty (New York: Oxford University Press, 2003), also available at http://hdr.undp.org/reports/ global/2003.

28. The United Nations issues annual reports on progress and challenges toward achieving the MDGs. The 2006 and 2009 reports, on which this section also draws, can be accessed at http:// mdgs.un.org. The World Bank also publishes the Global Monitoring Report on the MDGs. The 2010 monitoring report found that the global economic crisis slowed progress on poverty reduction, hunger, child and maternal health, access to clean water, and disease control, and is expected to have impacts beyond 2015. See Global Monitoring Report 2010: The MDGs after the Crisis, January 1, 2010, at http://web.worldbank. org. See Report of the Secretary-General, Keep- ing the Promise: A Forward-Looking Review to

39CHAPTER 1 Introducing Economic Development: A Global Perspective

Promote an Agreed Action Agenda to Achieve the Millennium Development Goals by 2015, February 12, 2010 http://www.un.org/ga/search/view_ doc.asp?symbol=A/64/665.

29. Despite some disappointments of slow rates of achievement of several targets in some regions, the September 2010 UN summit to review progress on the MDGs underscored its role as a global rallying point and measure of development success.

30. See Jan Vandemoortele, “Can the MDGs foster a new partnership for pro-poor policies?” in NGOs and the Millennium Development Goals: Citizen Ac- tion to Reduce Poverty, eds. Jennifer Brinkerhoff, Stephen C. Smith, and Hildy Teegen (New York:

Palgrave Macmillan, 2007), and Sabina Alkire with James Foster, “The MDGs: Multidimension- ality and Interconnection,” at www.ophi.org.uk/ wp-content/uploads/OPHI-RP-8a.pdf.

31. High-Level Panel of Eminent Persons on the Post- 2015 Development Agenda, A New Global Partnership: Eradicate Poverty and Transform Economies Through Sustainable Development: The Report of the High-Level Panel of Eminent Persons on the Post-2015 Development Agenda, May 30, 2013, http://www.post2015hlp .org/featured/high-level-panel-releases-recommen- dations-for-worlds-next-development-agenda.

The developing world has made substantial economic development progress in recent years. But the most striking feature of the global economy remains its extreme contrasts. Output per worker in the United States is about 10 times higher than it is in India and more than 50 times higher than in the Democratic Republic of Congo (DRC).1 In 2011, real income per capita was $48,820 in the United States, $3,640 in India, and $340 in the DRC.2

If the world were a single country, its income would be distributed more unequally than every nation except Namibia.3 There are also enormous gaps in measures of welfare. Life expectancy is 79 in the United States, 65 in India, and just 48 in the DRC. The percent of children who are underweight is less than 3% in the United States but 43% in India and 24% in the DRC. Whereas almost all women are literate in the United States, just 51% are in India and 57% in the DRC.4 How did such wide disparities come about? In today’s world, with so much knowledge and with the movement of people, informa- tion, and goods and services so rapid and comparatively inexpensive, how have such large gaps managed to persist and even widen? Why have some developing countries made so much progress in closing these gaps while others have made so little?

In this chapter, we introduce the study of comparative economic develop- ment. We begin by defining the developing world and describing how devel- opment is measured so as to allow for quantitative comparisons across countries. Average income is one, but only one, of the factors defining a country’s level of

40

Comparative Economic Development

Among countries colonized by European powers during the past 500 years, those that were relatively rich in 1500 are now relatively poor….The reversal reflects changes in the institutions resulting from European colonialism.

—Daron Acemoglu, Simon Johnson, and James A. Robinson, 2002

Emerging powers in the developing world are already sources of innovative social and economic policies and are major trade, investment, and increasingly development cooperation partners for other developing countries.

—Helen Clark, Administrator, United Nations Development Programme, 2012

2

41CHAPTER 2 Comparative Economic Development

economic development. This is to be expected, given the discussion of the mean- ing of development in Chapter 1.

We then consider 10 important features that developing countries tend to have in common, on average, in comparison with the developed world. In each case, we also discover that behind these averages are very substantial differences in all of these dimensions among developing countries that are important to appreciate and take into account in development policy. These areas are the following:

1. Lower levels of living and productivity

2. Lower levels of human capital

3. Higher levels of inequality and absolute poverty

4. Higher population growth rates

5. Greater social fractionalization

6. Larger rural populations but rapid rural-to-urban migration

7. Lower levels of industrialization

8. Adverse geography

9. Underdeveloped financial and other markets

10. Lingering colonial impacts such as poor institutions and often external dependence.

The mix and severity of these challenges largely set the development con- straints and policy priorities of a developing nation.

After reviewing these commonalities and differences among developing countries, we further consider key differences between conditions in today’s developing countries and those in now developed countries at an early stage of their development, and we examine the controversy over whether developing and developed countries are now converging in their levels of development.

We then draw on recent scholarship on comparative economic development to further clarify how such an unequal world came about and remained so per- sistently unequal, and we shed some light on the positive factors behind recent rapid progress in a significant portion of the developing world. It becomes quite clear that colonialism played a major role in shaping institutions that set the “rules of the economic game,” which can limit or facilitate opportunities for economic development. We examine other factors in comparative develop- ment, such as nations’ levels of inequality. We will come to appreciate why so many developing countries have such difficulties in achieving economic devel- opment but also will begin to see some of the outlines of what can be done to overcome obstacles and encourage faster progress even among today’s least developed countries.

The chapter concludes with a comparative case study of Bangladesh and Pakistan.

42 PART ONE Principles and Concepts

2.1 Defining the Developing World

The most common way to define the developing world is by per capita income. Several international agencies, including the Organization for Eco- nomic Cooperation and Development (OECD) and the United Nations, offer classifications of countries by their economic status, but the best-known system is that of the International Bank for Reconstruction and Develop- ment (IBRD), more commonly known as the World Bank. (The World Bank is examined in detail in Box 13.2). In the World Bank’s classification system, 213 economies with a population of at least 30,000 are ranked by their levels of gross national income (GNI) per capita. These economies are then classified as low-income countries (LICs), lower-middle-income countries (LMCs), upper- middle-income countries (UMCs), high-income OECD countries, and other high-income countries. (Often, LMCs and UMCs are informally grouped as the middle-income countries.)

With a number of important exceptions, the developing countries are those with low-, lower-middle, or upper-middle incomes. These countries are grouped by their geographic region in Table 2.1, making them easier to identify on the map in Figure 2.1. The most common cutoff points for these categories are those used by the World Bank: Low-income countries are defined as having a per capita gross national income in 2011 of $1,025 or less; lower-middle-income countries have incomes between $1,026 and $4,035; upper-middle-income coun- tries have incomes between $4,036 and $12,475; and high-income countries have incomes of $12,476 or more. Comparisons of incomes for several countries are shown graphically in Figure 2.2.

Note that a number of the countries grouped as “other high-income econo- mies” in Table 2.1 are sometimes classified as developing countries, such as when this is the official position of their governments. Moreover, high-income countries that have one or two highly developed export sectors but in which significant parts of the population remain relatively uneducated or in poor health, or social development is viewed as low for the country’s income level, may be viewed as still developing. Examples may include oil exporters such as Saudi Arabia and the United Arab Emirates. Upper-income economies also include some tourism-dependent islands with lingering development prob- lems, which now face daunting climate change adaptation challenges. Even a few of the high-income OECD member countries, notably Portugal and Greece, have been viewed as developing countries at least until recently—a perception that grew again with the ongoing economic crises (e.g., in October 2013 S&P Dow Jones reclassified Greece from “developed market” to “emerg- ing market.”). Nevertheless, the characterization of the developing world as sub-Saharan Africa, North Africa and the Middle East, Asia (except for Japan and, more recently South Korea and perhaps two or three other high-income economies), Latin America and the Caribbean, and the “transition” coun- tries of eastern Europe and Central Asia including the former Soviet Union, remains a useful generalization. In contrast, the developed world constituting the core of the high-income OECD is largely comprised of the countries of western Europe, North America, Japan, Australia, and New Zealand.

Sometimes a special distinction is made among upper-middle-income or newly high-income economies, designating some that have achieved relatively

World Bank An organiza- tion known as an “interna- tional financial institution” that provides development funds to developing countries in the form of interest-bearing loans, grants, and technical assistance.

Low-income countries (LICs) In the World Bank classification, countries with a GNI per capita of less than $1,025 in 2011.

Middle-income countries In the World Bank classification, countries with a GNI per capita between $1,025 and $12,475 in 2011.

43CHAPTER 2 Comparative Economic Development

TABLE 2.1 Classification of Economies by Region and Income, 2013

Country Code Class Country Code Class Country Code Class

East Asia and the Pacific Latin America and the Caribbean Sub-Saharan Africa     American Samoa‡ ASM UMC Antigua and Barbuda ATG UMC Angola* AGO UMC Cambodia* KHM LIC Argentina ARG UMC Benin* BEN LIC China CHN UMC Belize‡ BLZ LMC Botswana† BWA UMC Fiji‡ FJI LMC Bolivia† BOL LMC Burkina Faso*† BFA LIC Indonesia IDN LMC Brazil BRA UMC Burundi*† BDI LIC Kiribati*‡ KIR LMC Chile CHL UMC Cameroon CMR LMC (North) Korea, Dem. Rep. PRK LIC Colombia COL UMC Cape Verde‡ CPV LMC Lao PDR*† LAO LMC Costa Rica CRI UMC Central African Rep.*† CAF LIC Malaysia MYS UMC Cuba‡ CUB UMC Chad*† TCD LIC Marshall Islands‡ MHL LMC Dominica‡ DMA UMC Comoros*‡ COM LIC Micronesia, Fed. Sts.‡ FSM LMC Dominican Republic‡ DOM UMC Congo, Dem. Rep.* COD LIC Mongolia† MNG LMC Ecuador ECU UMC Congo, Rep. COG LMC Myanmar MMR LIC El Salvador SLV LMC Côte d’Ivoire CIV LMC Palau‡ PLW UMC Grenada‡ GRD UMC Eritrea* ERI LIC Papua New Guinea‡ PNG LMC Guatemala GTM LMC Ethiopia*† ETH LIC Philippines PHL LMC Guyana‡ GUY LMC Gabon GAB UMC Samoa*‡ WSM LMC Haiti*‡ HTI LIC Gambia, The* GMB LIC Solomon Islands*‡ SLB LMC Honduras HND LMC Ghana GHA LIC Thailand THA UMC Jamaica‡ JAM UMC Guinea* GIN LIC Timor-Leste*‡ TLS LMC Mexico MEX UMC Guinea-Bissau*‡ GNB LIC Tonga‡ TON LMC Nicaragua NIC LMC Kenya KEN LIC Tuvalu TUV LMC Panama PAN UMC Lesotho*† LSO LMC Vanuatu*‡ VUT LMC Paraguay† PRY LMC Liberia* LBR LIC Vietnam VNM LMC Peru PER UMC Madagascar* MDG LIC Europe and Central Asia St. Kitts and Nevis‡ KNA UMC Malawi*† MWI LIC Albania ALB LMC St. Lucia‡ LCA UMC Mali*† MLI LIC Armenia† ARM LMC St. Vincent and the     Mauritania* MRT LIC Azerbaijan† AZE LMC Grenadines‡ VCT UMC Mauritius‡ MUS UMC Belarus BLR UMC Suriname‡ SUR UMC Mayotte MYT UMC Bosnia and Herzegovina BIH UMC Uruguay URY UMC Mozambique* MOZ LIC Bulgaria BGR UMC Venezuela, RB VEN UMC Namibia NAM UMC Georgia GEO LMC Middle East and North Africa   Niger*† NER LIC Kazakhstan† KAZ UMC Algeria DZA UMC Nigeria NGA LMC Kosovo KSV LMC Djibouti* DJI LMC Rwanda*† RWA LIC Kyrgyz Republic† KGZ LIC Egypt, Arab Rep. EGY LMC Sao Tome and Principe*‡ STP LMC Latvia LVA UMC Iran, Islamic Rep. IRN UMC Senegal* SEN LMC Lithuania LTU UMC Iraq IRQ LMC Seychelles‡ SYC UMC Macedonia, FYR† MKD UMC Jordan JOR LMC Sierra Leone* SLE LIC Moldova† MDA LMC Lebanon LBN UMC Somalia* SOM LIC Montenegro MNE UMC Libya LBY UMC South Africa ZAF UMC Romania ROU UMC Morocco MAR LMC South Sudan SSD LIC Russian Federation RUS UMC Syrian Arab Rep. SYR LMC Sudan* SDN LMC Serbia SRB UMC Tunisia TUN LMC Swaziland† SWZ LMC Tajikistan† TJK LIC West Bank and Gaza WBG LMC Tanzania* TZA LIC Turkey TUR UMC Yemen, Rep.* YEM LMC Togo* TGO LIC Turkmenistan† TKM UMC South Asia   Uganda*† UGA LIC Ukraine UKR LMC Afghanistan*† AFG LIC Zambia*† ZMB LMC Uzbekistan† UZB LMC Bangladesh* BGD LIC Zimbabwe† ZWE LIC

Bhutan*† BTN LMC       India IND LMC       Maldives*‡ MDV UMC       Nepal*† NPL LIC       Pakistan PAK LMC      

      Sri Lanka LKA LMC

(Continued)

44 PART ONE Principles and Concepts

advanced manufacturing sectors as newly industrializing countries (NICs). Yet another way to classify the nations of the developing world is through their degree of international indebtedness; the World Bank has classified countries as severely indebted, moderately indebted, and less indebted. The United Nations Development Programme (UNDP) classifies countries according to their level of human development, including health and education attainments as low, medium, high, and very high. We consider the traditional and new UNDP Human Development Indexes in detail later in the chapter.

Another widely used classification is that of the least developed countries, a UN designation that as of 2012 included 49 countries, 34 of them in Africa, 9 in Asia, 5 among Pacific Islands, plus Haiti. For inclusion, a country has to meet each of three criteria: low income, low human capital, and high economic vulnerability. Other special UN classifications include landlocked developing countries (of which there are 30, with 15 of them in Africa) and small island developing states (of which there are 38).5

Finally, the term emerging markets was introduced at the International Finance Corporation to suggest progress (avoiding the then-standard phrase Third World that investors seemed to associate with stagnation). While the term is appealing, we do not use it in this text for three reasons. First, emerging market is widely used in the financial press to suggest the presence of active

TABLE 2.1 (Continued)

Country Code Class Country Code Class Country Code Class

High-Income OECD Countries Spain ESP   Guam‡ GUM   Australia AUS   Sweden SWE   Hong Kong, China HKG   Austria AUT   Switzerland CHE   Isle of Man IMN   Belgium BEL   United Kingdom GBR   Israel ISR   Canada CAN   United States USA   Kuwait KWT   Czech Rep. CZE   Other High-Income Economies    Liechtenstein LIE   Denmark DNK   Andorra AND   Macao, China MAC   Finland FIN   Antigua and Barbuda‡ ATG   Malta MLT   France FRA   Aruba‡ ABW   Monaco MCO   Germany DEU   Bahamas, The‡ BHS   Netherlands Antilles‡ ANT   Greece GRC   Bahrain‡ BHR   New Caledonia‡ NCL   Hungary HUN   Barbados‡ BRB   Northern Mariana Islands‡ MNP   Iceland ISL   Bermuda BMU   Oman OMN   Ireland IRL   Brunei Darussalam BRN   Poland POL   Italy ITA   Cayman Islands CYM   Puerto Rico‡ PRI   Japan JPN   Channel Islands CHI   Qatar QAT   Korea, Rep. (South) KOR   Croatia HRV   San Marino SMR   Luxembourg LUX   Cyprus CYP   Saudi Arabia SAU   Netherlands NLD   Estonia EST   Singapore‡ SGP   New Zealand NZL   Equatorial Guinea* GNQ   Slovenia SVN   Norway NOR   Faeroe Islands FRO   Taiwan, China TWN   Portugal PRT   French Polynesia‡ PYF   Trinidad and Tobago‡ TTO   Slovak Republic SVK   Greenland GRL   United Arab Emirates ARE  

* least developed countries † landlocked developing countries ‡ small island developing states

Source: Data from World Bank, World Development Indicators, 2013 (Washington, D.C.: World Bank, 2013) and WDI online; United Nations; and http://www.iso.org.

Newly industrializing countries (NICs) Countries at a relatively advanced level of economic develop- ment with a substantial and dynamic industrial sector and with close links to the inter- national trade, finance, and investment system.

Least developed countries A UN designation of countries with low income, low human capital, and high economic vulnerability.

Human capital Productive investments in people, such as skills, values, and health resulting from expenditures on education, on-the-job training programs, and medical care.

45CHAPTER 2 Comparative Economic Development

stock and bond markets; although financial deepening is important, it is only one aspect of economic development. Second, referring to nations as markets may lead to an underemphasis on some non-market priorities in development. Third, usage varies, and there is no established or generally accepted designa- tion of which markets should be labeled as emerging and which as yet to emerge (the latter now sometimes dubbed frontier markets in the financial press).

The simple division of the world into developed and developing countries is sometimes useful for analytical purposes. Many development models apply across a wide range of developing country income levels. However, the wide income range of the latter serves as an early warning for us not to overgeneral- ize. Indeed, the economic differences between low-income countries in sub- Saharan Africa and South Asia and upper-middle-income countries in East Asia and Latin America can be even more profound than those between high- income OECD and upper-middle-income developing countries.

2.2 Basic Indicators of Development: Real Income, Health, and Education

In this section, we examine basic indicators of three facets of development: real income per capita adjusted for purchasing power; health as measured by life expectancy, undernourishment, and child mortality; and educational attainments as measured by literacy and schooling.

Purchasing Power Parity

In accordance with the World Bank’s income-based country classification scheme, gross national income (GNI) per capita, the most common measure of the overall level of economic activity, is often used as a summary index of the relative economic well-being of people in different nations. It is calculated as the total domestic and foreign value added claimed by a country’s residents without making deductions for depreciation (or wearing out) of the domestic capital stock. Gross domestic product (GDP) measures the total value for final use of output produced by an economy, by both residents and nonresidents. Thus, GNI comprises GDP plus the difference between the income residents receive from abroad for factor services (labor and capital) less payments made to nonresidents who contribute to the domestic economy. Where there is a large nonresident population playing a major role in the domestic economy (such as foreign corporations), these differences can be significant (see Chapter 12). In 2011, the total national income of all the nations of the world was valued at more than U.S. $66 trillion, of which about $47 trillion originated in the eco- nomically developed high-income regions and about $19 trillion was generated in the less developed nations, despite their representing about five-sixths of the world’s population. In 2011, Norway had 240 times the per capita income of Ethiopia and 63 times that of India.

Per capita GNI comparisons between developed and less developed coun- tries like those shown in Figure 2.2 are, however, exaggerated by the use of official foreign-exchange rates to convert national currency figures into U.S. dollars. This conversion does not measure the relative domestic purchasing

Gross national income (GNI) The total domestic and foreign output claimed by residents of a country, consist- ing of gross domestic product (GDP) plus factor incomes earned by foreign residents, minus income earned in the domestic economy by non- residents.

Value added The portion of a product’s final value that is added at each stage of pro- duction.

Depreciation (of the capital stock) The wearing out of equipment, buildings, infra- structure, and other forms of capital, reflected in write-offs to the value of the capital stock.

Capital stock The total amount of physical goods existing at a particular time that have been produced for use in the production of other goods and services.

Gross domestic product (GDP) The total final output of goods and services produced by the country’s economy within the country’s territory by residents and nonresidents, regardless of its allocation between domestic and foreign claims.

46 PART ONE Principles and Concepts

FIGURE 2.1 Nations of the World, Classified by GNI Per Capita

Kosov

Western SaharaPuerto

Rico (US)

British Virgin Islands (UK)

The GambiaSt. Lucia

São Tomé and Príncipe

Monaco

Luxembourg

Liechtenstein

Kiribati

Grenada

Dominica

Cape Verde

Andorra

St. Vincent and the Grenadines

St. Kitts and Nevis

Barbados

The Bahamas

Antigua and Barbuda

Martinique (Fr)

Uruguay

U n i t e d S t a t e s

United Kingdom

Tun

Trinidad and Tobago

Togo

Switzerla

Suriname

Spain

Sierra Leone

Senegal

R.B. de Venezuela

Portugal

Peru

Paraguay

Panama

Norway

Nigeria

Nig

Nicaragua

The Netherlands

Morocco

Mexico

Mauritania

Mali

Liberia

Jamaica

Ita

Ireland

Iceland

Faeroe Islands (Den)

Honduras

Haiti

Guyana

Guinea-Bissau Guinea

Guatemala

Ghana

Germ

G

France

Equatoria

El Salvador

Ecuador

Denm

Cuba

Côte d'Ivoire

Costa Rica

Colombia

Chile

C a n a d a

Cam

Burkina Faso

B r a z i l

Bolivia

Benin

Belize

Belgium

Argentina

Algeria Dominican Republic

Curaçao (Neth)

Isle of Man (UK)

Greenland (Den)

Gibraltar (UK)

French Polynesia (Fr)

French Guiana (Fr)

Channel Islands (UK)

Cayman Islands (UK)

Turks and Caicos Islands (UK)

Bermuda (UK)

Aruba (Neth)

Guadeloupe (Fr)

Middle East & North Africa $7,097St. Martin (Fr)US Virgin

Islands (US) Sint Maarten (Neth)

Latin America & Caribbean $8,645

Brazil $10,720

Upper-middle-income countries ($4,036–$12,475)

Lower-middle-income countries ($1,026–$4,035)

Lower-income-countries ($1,025 or less)

High-income countries ($12,476 or more)

NO DATA

GNI per capita, World Bank Atlas method, 2011

Income

Source: Data from Atlas of Global Development, 4th ed., pp. 16-17: World Bank and Collins. 2013. ATLAS OF GLOBAL DEVELOPMENT: A VISUAL GUIDE TO THE WORLD’S GREATEST CHALLENGES, FOURTH EDITION. Washington, DC and Glasgow: World Bank and Collins. doi: 10.1596/978-0-8213-9757-2. License: Creative Commons Attribution CC BY 3.0

47CHAPTER 2 Comparative Economic Development

o

West Bank and Gaza

Vanuatu

Tonga

Timor-Leste

Solomon Islands

Singapore

Seychelles

San Marino

Samoa

Tuvalu

Qatar

Eritrea

Somalia

Kyrgyz Republic

Dem. People's Rep. of Korea

Madagascar

Palau

Nauru

Mauritius

Malta

Maldives

Lebanon

KuwaitIsrael

Fiji

Federated States of Micronesia

Marshall Islands

Cyprus

Comoros

Brunei Darussalam

Bahrain

Zimbabwe

Zambia

Vietnam

Uzbekistan

United Arab Emirates

Ukraine

Uganda

TurkmenistanTurkey

isia

Thailand

Tanzania

Tajikistan

Syrian Arab Rep.

nd

Sweden

Swaziland

Sudan

South Sudan

Sri Lanka

South Africa

Slovenia Slovak Republic

Serbia

Saudi Arabia

Rwanda

R u s s i a n F e d e r a t i o n

Romania

Rep. of Yemen

Rep. of Korea

Poland

Philippines

Papua New Guinea

Pakistan

Oman

y

er

Nepal

Namibia

Myanmar

Mozambique

Mongolia Moldova

Malaysia

Malawi

Lithuania

Libya

Lesotho

Latvia

Lao P.D.R.

Kenya

Kazakhstan

Jordan

Japan

aly

Islamic Republic of Iran

Iraq

Indonesia

India

Hungary

Greece

many

Georgia

abon

FYR Macedonia

Finland

Ethiopia

Estonia

al Guinea

Djibouti

ark

Dem. Rep. of Congo

Czech Republic

Croatia

Rep. of congo

C h i n a

Chad

Central African

Republicmeroon

Cambodia

Burundi

Bulgaria

Botswana

Bosnia and Herzegovina

Bhutan

Belarus

Bangladesh

Azerbaijan

Austria

A u s t r a l i a

New Zealand

Armenia

Arab Rep. of Egypt

Angola

Albania

Montenegro

Afghanistan

Réunion (Fr) New

Caledonia (Fr)

N. Mariana Islands (US)

Mayotte (Fr)

Guam (US)

American Samoa (US)

Sub-Saharan Africa $1,265

Europe & Central Asia $23,562

South Asia $1,299

East Asia & Pacific $7,857

China $4,930

India $1,410

Russian Federation $10,400

48 PART ONE Principles and Concepts

power of different currencies. In an attempt to rectify this problem, researchers have tried to compare relative GNIs and GDPs by using purchasing power parity (PPP) instead of exchange rates as conversion factors. PPP is calculated using a common set of international prices for all goods and services. In a sim- ple version, purchasing power parity is defined as the number of units of a for- eign country’s currency required to purchase the identical quantity of goods and services in the local developing country market as $1 would buy in the United States. In practice, adjustments are made for differing relative prices across countries so that living standards may be measured more accurately.6

Generally, prices of nontraded services are much lower in developing coun- tries because wages are so much lower. Clearly, if domestic prices are lower, PPP measures of GNI per capita will be higher than estimates using foreign- exchange rates as the conversion factor. For example, China’s 2011 GNI per capita was only 10% of that of the United States using the exchange-rate con- version but rises to 17% when estimated by the PPP method of conversion. Income gaps between developed and developing nations thus tend to be less when PPP is used.

Table 2.2 provides a comparison of exchange rate and PPP GNI per capita for 30 countries, 10 each from Africa, Asia, and Latin America, plus Canada, the United Kingdom and the United States. In the first column of Table 2.2, incomes are measured at market or official exchange rates and suggest that income of a person in the United States is 242 times that of a person in the DRC. But this is unbelievable, as many services cost much less in the DRC than in the United States. The PPP rates give a better sense of the amount of goods and services that could be bought evaluated at U.S. prices and suggest that real U.S. incomes are closer to 135 times that of the DRC—still a level of inequality that stretches the imagination. Overall, the average real (PPP) income per capita in

Purchasing power parity (PPP) Calculation of GNI using a common set of inter- national prices for all goods and services, to provide more accurate comparisons of living standards.

FIGURE 2.2 Income Per Capita in Selected Countries, 2011

0 10,000 20,000 30,000 40,000 50,000 60,000

C o

u n

tr y

Annual gross national income per capita (2012 U.S. $)

Canada United States

United Kingdom Brazil

Mexico China

Dominican Republic Indonesia

Egypt, Arab Rep. Ghana

India Pakistan

Côte d’Ivoire Kenya

Bangladesh Haiti

Ethiopia Congo, Dem. Rep.

Source: Data from World Bank, World Development Indicators, 2013 (Washington, D.C.: World Bank, 2013), tab. 1.1.

49CHAPTER 2 Comparative Economic Development

high-income countries is more than 28 times that in low-income countries and more than 5 times higher than in middle-income countries.

Indicators of Health and Education

Besides average incomes, it is necessary to evaluate a nation’s average health and educational attainments, which reflect core capabilities. Table 2.3 shows some basic indicators of income, health (the under-5 mortality rate for 1990 and 2011, plus the rate of malnutrition and life expectancy), and education

TABLE 2.2 A Comparison of Per Capita GNI in Selected Developing Countries, the United Kingdom, and the United States, Using Official Exchange-Rate and Purchasing Power Parity Conversions, 2011

Source: Data from World Bank, World Development Indicators, 2013 (Washington, D.C.: World Bank, 2013), tab. 1.1.

  GNI Per Capita (U.S. $)

Country Exchange Rate Purchasing Power Parity

Bangladesh 770 1,910 Bolivia 2,020 4,890 Botswana 7,070 15,550 Brazil 10,700 11,410 Cambodia 800 2,180 Canada 46,730 41,390 Chile 12,270 19,820 China 4,940 8,390 Colombia 6,090 9,600 Congo, Dem. Rep. 200 360 Costa Rica 7,660 11,910 Côte d’Ivoire 1,140 1,780 Dominican Republic 5,190 9,350 Egypt, Arab Rep. 2,760 6,440 Ghana 1,420 1,830 Guatemala 2,870 4,760 Haiti 700 1,190 India 1,450 3,680 Indonesia 2,930 4,480 Kenya 810 1,690 Korea, Rep. 20,870 29,860 Mexico 8,970 15,930 Niger 330 600 Nigeria 1,260 2,270 Pakistan 1,120 2,880 Peru 5,120 9,390 Philippines 2,200 4,120 Senegal 1,070 1,940 Thailand 4,620 8,710 Uganda 470 1,230 United Kingdom 37,840 35,950 United States 48,550 48,820 Vietnam 1,270 3,250       Low income 554 1,310 Middle income 3,923 6,802 High income 36,390 36,472

50 PART ONE Principles and Concepts

TABLE 2.3 Commonality and Diversity: Some Basic Indicators

Source: World Bank, World Development Indicators 2013, and World Bank WDI online, accessed 1 August 2013.

  Prevalence of Malnutrition

Primary Completion Rate Under-5 Mortality

Rate Total

per 1,000 Live Births Life Expectancy   

  Underweight Total     % of Children Under

Age 5 % of Relevant Age

Group  

  2005-11 1991 2011 1990 2011

Bangladesh 41.3 46 .. 139 46 69   Bolivia 4.5 71 95 120 51 67   Botswana 11.2 89 97 53 26 53   Brazil 2.2 92 .. 58 16 73   Cambodia 29 38 90 117 43 63   Central African Republic 28 28 43 169 164 48   Chile 0.5 .. 95 19 9 79   China 3.4 109 .. 49 15 73   Colombia 3.4 73 112 34 18 74   Congo, Dem. Rep. 28.2 49 61 181 168 48   Costa Rica 1.1 80 99 17 10 79   Côte d’Ivoire 29.4 43 59 151 115 55   Cuba 1.3 94 99 13 6 79   Dominican Republic 3.4 63 92 58 25 73   Egypt, Arab Rep. 6.8 .. 98 86 21 73   Ethiopia 29.2 23 58 198 77 59   Ghana 14.3 65 94 121 78 64   Guatemala 13 .. 86 78 30 71   India 43.5 63 97 114 61 65   Indonesia 18.6 89 108 82 32 69   Mexico 3.4 88 104 49 16 77   Mozambique 18.3 27 56 226 103 50   Niger 39.9 18 46 314 125 55   Nigeria 26.7 .. 74 214 124 52   Pakistan 30.9 .. 67 122 72 65   Peru 4.5 .. 97 75 18 74   Philippines 20.7 89 92 57 25 69   Senegal 19.2 41 63 136 65 59   Uganda 16.4 .. 55 178 90 54   Vietnam 20.2 .. 104 50 22 75  

Low income 22.6 46 67 164 95 59   Middle income 16 83 94 82 46 69   High income 1.7 97 101 12 6 79                   East Asia & Pacific 5.5 84 .. .. 21 72   Latin America & Caribbean

3.1 84 102 53 19 74  

Middle East & North Africa

6.3 77 91 70 32 72  

South Asia 33.2 63 88 119 62 66   Sub-Saharan Africa 21.4 52 69 178 109 55

Note: Some of the specific countries listed in Table 2.3 differ from those listed in Table 2.2 due to differing availability of the most recent comparable data by topic; for example, primary completion rate was not available for Haiti; and income was not available for Cuba.

51CHAPTER 2 Comparative Economic Development

(the primary completion rate for 1991 and 2011). (Each country’s region and income grouping can be found in Table 2.1). Life expectancy is the average number of years newborn children would live if subjected to the mortality risks prevailing for their cohort at the time of their birth. Undernourishment means consuming too little food to maintain normal levels of activity; it is what is often called the problem of hunger. High fertility can be both a cause and a conse- quence of underdevelopment, so the birth rate is reported as another basic indi- cator. Literacy is the fraction of adult males and females reported or estimated to have basic abilities to read and write; functional literacy is generally lower than the reported numbers.

Table 2.3 shows these data for the low-, lower-middle-, upper-middle-, and high-income country groups. The table also shows averages from five devel- oping regions (East Asia and the Pacific, Latin America and the Caribbean, the Middle East and North Africa, South Asia, and sub-Saharan Africa) and from 30 illustrative countries balanced across developing regions similar to those in Table 2.2 (with a few substitutions due to data availability).

Note that in addition to big differences across these income groupings, the low-income countries are themselves a very diverse group with greatly differ- ing development challenges.

For example, even Bangladesh has a real income that is now more than five times greater than the DRC; and India’s income is more than 10 times greater. Under-5 malnutrition (underweight) is higher in Bangladesh, at 41.3%, than DRC (a still very high 28.2%). The under-5 mortality rate in Bangladesh is 46, while that of the DRC is nearly quadruple that number at 168. Life expectancy in Congo is just 48, compared with 69 in Bangladesh. But while India and Bangla- desh clearly do better overall than countries like the DRC, most low- and lower- middle-income countries still face enormous development challenges as seen by comparing these statistics even to Botswana, Peru, or Thailand

2.3 Holistic Measures of Living Levels and Capabilities

The New Human Development Index

The most widely used measure of the comparative status of socioeconomic development is presented by the United Nations Development Programme (UNDP) in its annual series of Human Development Reports. The centerpiece of these reports, which were initiated in 1990, is the construction and refinement of its informative Human Development Index (HDI). This section examines the New HDI, initiated in 2010 (the well-known traditional HDI—the UNDP centerpiece from 1990–2009—is examined in detail in Appendix 2.1). Box 2.2 summarizes “What Is New in the New HDI.”

The New HDI, like its predecessor, ranks each country on a scale of 0 (low- est human development) to 1 (highest human development) based on three goals or end products of development: a long and healthy life as measured by life expectancy at birth; knowledge as measured by a combination of average schooling attained by adults and expected years of schooling for school-age children; and a decent standard of living as measured by real per capita gross

Human Development Index (HDI) An index measuring national socioeconomic devel- opment, based on combining measures of education, health, and adjusted real income per capita.

52 PART ONE Principles and Concepts

domestic product adjusted for the differing purchasing power parity of each country’s currency to reflect cost of living and for the assumption of dimin- ishing marginal utility of income.

There are two steps in calculating the New HDI: first, creating the three “dimension indices”; and second, aggregating the resulting indices to produce the overall New Human Development Index (NHDI).

After defining the relevant minimum and maximum values (or lower and upper “goalposts”), each dimension index is calculated as a ratio that basically is given by the percent of the distance above the minimum to the maximum levels that a country has attained.

Dimension index = Actual Value - Minimum Value

Maximum Value - Minimum Value (2.1)

The health (or “long and healthy life”) dimension of the New HDI is calcu- lated with a life expectancy at birth index, which takes a minimum value of 20 years and a maximum value of 83.57 years (the observed maximum value for any country). For example, for the case of Ghana this is:

Life expectancy index = 164.6 - 202>183.6 - 202 = 0.701 (2.2)

The education (“knowledge”) component of the HDI is calculated with a combination of the average years of schooling for adults aged 25 and older and expected years of schooling for a school-age child now entering school. As explained by the UNDP, these indicators are normalized using a minimum value of 0, and maximum values are set to the actual observed maximum value of mean years of schooling from the countries in the time series, 1980– 2012, which is 13.3 years estimated for the United States in 2010. For Ghana, the average years of schooling among adults is 7 years, so the mean years of schooling subindex is calculated as:

17.0 - 02>113.3 - 02 = 0.527 (2.3)

We can think of this as saying that Ghana is about 53% of the way to the global standard of average education.

In considering expected future education, the highest value (cap, or “goalpost”) is given as 18 years (which we may think of as approximately corresponding to a master ’s degree).

For Ghana, the expected number of years of schooling for a child entering school now is estimated at 11.4 years. The expected years of schooling sub- index is then calculated as:

111.4 - 02>118.0 - 02 = 0.634 (2.4)

The education index is then calculated as a version of the geometric mean of the two subindexes.7

The standard of living (income) component is calculated using purchasing- power-adjusted per-capita gross national income (GNI). For Ghana, the income index then is (where ln stands for natural log):

Income index = 3ln11,6842 - ln110024>3ln187,4782 - ln110024 = 0.417 (2.5)

Diminishing marginal utility The concept that the subjective value of additional consump- tion lessens as total consump- tion becomes higher.

53CHAPTER 2 Comparative Economic Development

Using these three measures of development and applying the formula to data for all 187 countries for which data is available, the HDI currently ranks countries into four groups: low human development (0.0 to 0.535), medium human development (0.536 to 0.711), high human development (0.712 to 0.799), and very high human development (0.80 to 1.0).

The component indexes of the NHDI are computed by taking the differ- ence between the country’s actual achievement and the minimum goalpost value, and then dividing the result by the difference between the overall maxi- mum goalpost and minimum goalpost values. But in calculating the overall index, in place of the arithmetic mean, a geometric mean of the three indexes is used (a geometric mean is also used to build up the overall education index from its two components).

Let’s look at why this change is important and how the calculations are done.

Computing the NHDI The use of a geometric mean in computing the New HDI is very important. When using an arithmetic mean (adding up the com- ponent indexes and dividing by 3) in the HDI, the effect is to assume perfect substitutability across income, health, and education. For example, a higher value of the education index could compensate, one for one, for a lower value of the health index. In contrast, use of a geometric mean ensures that poor performance in any dimension directly affects the overall index. Thus, allow- ing for imperfect substitutability is a beneficial change; but there is active debate about whether using the geometric mean is the most appropriate way to accomplish this.8

Thus, as the UNDP notes, the new calculation “captures how well rounded a country’s performance is across the three dimensions.” Moreover, the UNDP argues “that it is hard to compare these different dimensions of well-being and that we should not let changes in any of them go unnoticed.”

So in the New HDI, instead of adding up the health, education, and income indexes and dividing by 3, the New HDI is calculated with the geometric mean:

NHDI = H1>3E1>3I1>3 (2.6)

where H stands for the health index, E stands for the education index, and I stands for the income index. This is equivalent to taking the cube root of the product of these three indexes. The calculations of the NHDI are illustrated for Ghana in Box 2.1.

Table 2.4 shows the 2013 values of the New HDI for a set of 31 countries. South Korea has achieved the status of a fully developed country, ranking below Canada but above the United Kingdom. Countries such as the United Arab Emirates, Turkey, Guatemala, Gabon, Côte d’Ivoire, Pakistan, Papua New Guinea, and South Africa perform more poorly on the New HDI than would be predicted from their income level, while the reverse is true of South Korea, Chile, Bangladesh, Cuba, Madagascar, and Ghana. Countries such as Russia, Mexico, India, and Niger perform on the New HDI just about as pre- dicted by their income levels.

Income predicts rather weakly how countries will perform on education and health, or on the NHDI in particular. For example, Cuba and Egypt have nearly the same real income per person, but Cuba ranks 59th on the New HDI (44 points

54 PART ONE Principles and Concepts

above where predicted by its income level) and Egypt ranks 112th (6 below where predicted by income). Mexico and Gabon have a very similar income, but Mexico is 4 places above what would be predicted by its income and Gabon is 40 points below. Bangladesh and Pakistan have an identical New HDI ranking, but Paki- stan has a much higher income, and Bangladesh is 9 places higher than expected while Pakistan is 9 places below; see the case study at the end of this chapter for a detailed examination of diverging development in these two countries.

The UNDP now also offers the Inequality-Adjusted Human Development Index (IHDI)—which imposes a penalty on the HDI that increases as inequal- ity across people becomes greater—and the Gender Inequality Index (GII), as well as an important innovation, the Multidimensional Poverty Index (MPI), which is examined in detail in Chapter 5.

Clearly, the Human Development Index, in its Traditional as well as New forms, has made a major contribution to improving our understanding of what constitutes development, which countries are succeeding (as reflected by rises in their NHDI over time), and how different groups and regions within countries are faring. By combining social and economic data, the NHDI allows nations to take a broader measure of their development performance, both relatively and absolutely.

Although there are some valid criticisms, the fact remains that the New HDI and its Traditional version considered in Appendix 2.1, when used in

BOX 2.1 Computing the New HDI: Ghana

Source: UNDP, Human Development Report, 2013, Technical Notes (online):, http://hdr.undp.org/en/media/HDR%202013%20technical%20notes%20EN.pdf.

Indicator Value

Life expectancy at birth (years) 64.6 Mean years of schooling 7.0 Expected years of schooling 11.4 GNI per capita (PPP $) 1,684 Indexes  

Note: Values are rounded.

Life expectancy index = 64.6 - 20 83.6 - 20

= 0.701

Mean years of schooling index = 7.0 - 0

13.3 - 0 = 0.527

Expected years of schooling index = 11.4 - 0 18.0 - 0

= 0.634

Education index = 20.527 * 0.634 - 0

0.971 - 0 = 0.596

Income index = ln11,684) - ln1100)

ln187,478) - ln1100) = 0.417

Human Development Index

= 23 0.701 * 0.558 * 0.417 = 0.596 UN income estimate will differ somewhat from World Bank estimate.

Example: Ghana

55CHAPTER 2 Comparative Economic Development

conjunction with other economic measures of development, greatly increase our understanding of which countries are experiencing development and which are not. And by modifying a country’s overall NHDI to reflect income distribution, gender, regional, and ethnic differentials, as presented in recent Human Development Reports, we are now able to identify not only whether a country is developing but also whether various significant groups within that country are participating in that development.9

2.4 Characteristics of the Developing World: Diversity within Commonality

As noted earlier, there are important historical and economic commonalities among developing countries that have led to their economic development

TABLE 2.4 2013 New Human Development Index and its Components for Selected Countries

Country NHDI Rank

Life Expectancy

at Birth

Mean Yrs Schooling (of Adults)

Expected Years

Schooling (of children)

GNI Per Capita

New HDI value

GNI Per Capita

Rank Minus HDI Rank

United States 3 78.7 13.3 16.8 43,480 0.937 6 Canada 11 81.1 12.3 15.1 35,369 0.911 5 South Korea 12 80.7 11.6 17.2 28,231 0.909 15 United Kingdom 26 80.3 9.4 16.4 32,538 0.875 5 Chile 40 79.3 9.7 14.7 14,987 0.819 13 United Arab Emirates 41 76.7 8.9 12 42,716 0.818 −31 Russian Federation 55 69.1 11.7 14.3 14,461 0.788 0 Cuba 59 79.3 10.2 16.2 5,539 0.78 44 Mexico 61 77.1 8.5 13.7 12,947 0.775 4 Costa Rica 62 79.4 8.4 13.7 10,863 0.773 12 Brazil 85 73.8 7.2 14.2 10,152 0.73 −8 Turkey 90 74.2 6.5 12.9 13,710 0.722 −32 Sri Lanka 92 75.1 9.3 12.7 5,170 0.715 18 China 101 73.7 7.5 11.7 7,945 0.699 −11 Gabon 106 63.1 7.5 13 12,521 0.683 −40 Egypt 112 73.5 6.4 12.1 5,401 0.662 −6 Botswana 119 53 8.9 11.8 13,102 0.634 −55 South Africa 121 53.4 6.7 10.6 9,594 0.629 −42 Guatemala 133 71.4 4.1 10.7 4,235 0.581 −14 Ghana 135 64.6 7 11.4 1,684 0.558 22 Equatorial Guinea 136 51.4 5.4 7.9 21,715 0.554 −97 India 136 65.8 4.4 10.7 3,285 0.554 −3 Kenya 145 57.7 7 11.1 1,541 0.519 15 Bangladesh 146 69.2 4.8 8.1 1,785 0.515 9 Pakistan 146 65.7 4.9 7.3 2,566 0.515 −9 Madagascar 151 66.9 5.2 10.4 828 0.483 28 Papua New Guinea 156 63.1 3.9 5.8 2,386 0.466 −15 Côte d’Ivoire 168 56 4.2 6.5 1,593 0.432 −9 Burkina Faso 183 55.9 1.3 6.9 1,202 0.343 −18 Chad 184 49.9 1.5 7.4 1,258 0.34 −20 Niger 186 55.1 1.4 4.9 701 0.304 −4

Source: 2013 Human Development Report 2013, Table 1, pages 144-147 (New York: United Nations Development Programme, 2013)

56 PART ONE Principles and Concepts

BOX 2.2 What Is New in the New Human Development Index

In November 2010, the UNDP introduced its New Human Development Index (NHDI), which has eight notable changes, each with strengths but also a few potential drawbacks.

1. Gross national income (GNI) per capita replaces gross domestic product (GDP) per capita. This should be an unambiguous improvement: GNI reflects what citizens can do with income they receive, whereas that is not true of value added in goods and services produced in a country that go to someone outside it, and income earned abroad still benefits some of the nation’s citizens. As trade and remittance flows have been expanding rapidly, and as aid has been better targeted to very low-income countries, this distinction has become increasingly important.

2. The education index has been completely revamped. Two new components have been added: the average actual educational attain- ment of the whole population and the expected attainment of today’s children. Each of these changes to the index has implica- tions. Use of actual attainment—average years of schooling—as an indicator is unambigu- ously an improvement. Estimates are regularly updated, and the statistic is easily compared quantitatively across countries. And even though it is at best a very rough guide to what is actually learned—on average, a year of school- ing in Mali provides students with much less than a year of schooling in Norway—this is the best measure we have at present because more detailed data on quality that are credible and comparable are simply not available.

3. Expected educational attainment, the other new component, is somewhat more ambigu- ous: It is not an achievement but a UN fore- cast. History shows that much can go wrong to derail development plans. Nevertheless,

there have also been many development upside surprises, such as rapid improvements in educational attainment in some coun- tries; there is a risk that low expectations will prove discouraging. Note that life expectancy, which remains the indicator for health, is also a projection based on prevailing conditions.

4. The two previous components of the edu- cation index, literacy and enrollment, have been correspondingly dropped. In contrast to expected attainment, literacy is clearly an achievement, and even enrollment is at least a modest achievement. However, literacy has always been badly and too infrequently mea- sured and is inevitably defined more modestly in a less developed country. And enrollment is no guarantee that a grade will be completed or for that matter that anything is learned or that students (or teachers) even attend.

5. The upper goalposts (maximum values) in each dimension have been increased to the observed maximum rather than given a pre- defined cutoff. In some ways, this returns the index to its original design, which was criticized for inadequately recognizing small gains by countries starting at very low levels.

6. The lower goalpost for income has been reduced. This is based on updated estimates for the historic low for recorded income for any country.10

7. Another minor difference is that rather than using the common logarithm (log) to reflect diminishing marginal benefit of income, the NHDI now uses the natural log (ln), as used in the fifth equation in Box 2.1. This reflects a more usual construction of indexes.

8. Possibly the most consequential change is that the NHDI is computed with a geometric mean rather than a simple arithmetic mean, as examined previously.

57CHAPTER 2 Comparative Economic Development

problems being studied within a common analytical framework in develop- ment economics. These widely shared problems are examined here in detail on an issue-by-issue basis. At the same time, however, it is important to bear in mind that there is a great deal of diversity throughout the develop- ing world, even within these areas of broad commonality. The wide range of income, health, education, and HDI indicators already reviewed is sometimes called a “ladder of development.”11 Different development problems call for different specific policy responses and general development strategies. This section examines the 10 major areas of “diversity within commonality” in the developing world.

Lower Levels of Living and Productivity

As we noted at the outset of the chapter, there is a vast gulf in productivity between advanced economies such as the United States and developing nations, including India and the DRC, but also a wide range among these and other developing countries. And as we have seen, all countries with averages below what is defined as high income are considered developing in most taxonomies (and some in the high-income range as defined by the World Bank are still considered developing). The lower average levels but wide ranges of income in developing areas are seen in Table 2.3. Even when adjusted for purchasing power parity and despite extraordinary recent growth in China and India, the low- and middle-income developing nations, with more than five-sixths (84%) of the world’s people, received only about 46% of the world’s income in 2011, as seen in Figure 2.3a. Though resulting from a number of deeper causes, the wide disparity in income largely corresponds to the large gaps in output per worker between developing and developed countries as seen in Figure 2.3b.12

At very low income levels, in fact, a vicious circle may set in, whereby low income leads to low investment in education and health as well as plant and equipment and infrastructure, which in turn leads to low productivity and economic stagnation. This is known as a poverty trap or what Nobel laureate Gunnar Myrdal called “circular and cumulative causation.”13 However, it is important to stress that there are ways to escape from low income, as you will see throughout this book. Further, the low-income countries are themselves a very diverse group with greatly differing development challenges.14

Some star performers among now high-income economies such as South Korea and Taiwan were once among the poorest in the world. Some middle- income countries are also relatively stagnant, but others are growing rapidly— China most spectacularly, as reviewed in the case study at the end of Chapter 4. Indeed, income growth rates have varied greatly in different developing regions and countries, with rapid growth in East Asia, slow or even no growth in sub-Saharan Africa, and intermediate levels of growth in other regions. Problems of igniting and then sustaining economic growth are examined in depth in Chapters 3 and 4.

One common misperception is that low incomes result from a country’s being too small to be self-sufficient or too large to overcome economic inertia. However, there is no necessary correlation between country size in population or area and economic development (in part because each has different advantages and disadvantages that can offset each other).15

58 PART ONE Principles and Concepts

Source: Figure 2.3a, Data from World Bank, World Development Indicators 2013 (Washington, D. C.: World Bank, 2013), p.24. Figure 2.3b, United Nations, Millenium Development Goals Report 2012, p.9.

FIGURE 2.3 (a) Shares of Global Income, 2008. (b) Developing regions lag far behind the developed world in productivity measured as output per worker.

East Asia and Pacific 18%

(a)

Europe and Central Asia

7%

Latin America and Caribbean

8%

Middle East and North Africa

3% South Asia

7%Sub-Saharan Africa

2%

High-Income countries

54%

1991 2001 2011*

* Figures for 2011 are preliminary estimates.

Developing regions

Developed regions

Western Asia

Latin America and the Caribbean

Northern Africa

Eastern Asia

Caucasus and Central Asia

South-Eastern Asia

Southern Asia

Oceania

Sub-Saharan Africa

13

64

40

23

21

14

14

10

9

6

6

8

57

35

21

18

6

7

7

5

5

5

6

48

30

20

17

3

10

6

4

5

5

Output per worker, 1991, 2001, and 2011 (Thousands of constant 2005 PPP-adjusted international dollars)

0 10 20 30 40 50 60 70

(b)

59CHAPTER 2 Comparative Economic Development

The 12 most populous countries include representatives of all four cate- gories: low-, lower-middle-, upper-middle-, and high-income countries (see Table 2.5). The 12 least populous on the list include primarily lower-middle- and upper-middle-income countries, although the 12th least populous country, São Tomé and Príncipe, has a per capita income of just $1,030. And four very small but high-income European countries that are UN members (Andorra, Monaco, Liechtenstein, and San Marino) would appear on the list if compa- rable World Bank income data were available.

Lower Levels of Human Capital

Human capital—health, education, and skills—is vital to economic growth and human development. We have already noted the great disparities in human capital around the world while discussing the Human Development Index. Compared with developed countries, much of the developing world has lagged in its average levels of nutrition, health (as measured, for example, by life expectancy or undernourishment), and education (measured by liter- acy), as seen in Table 2.3. The under-5 mortality is 17 times higher in low- income countries than in high-income countries, although great progress has been made since 1990, as shown graphically in Figure 2.4.

Table 2.6 shows primary school enrollment rates (percentage of students of primary age enrolled in school) and the primary school pupil-to-teacher ratio for the four country income groups and for six major developing regions. Enrollments have strongly improved in recent years, but student attendance and completion, along with attainment of basic skills such as functional lit- eracy, remain problems. Indeed, teacher truancy remains a serious problem in South Asia and sub-Saharan Africa.16

Moreover, there are strong synergies (complementarities) between prog- ress in health and education (examined in greater depth in Chapter 8). For

TABLE 2.5 The 12 Most and Least Populated Countries and Their Per Capita Income, 2008

Source: The World Bank, World Development Indicators, 2010 (Washington, D.C.: World Bank, 2010), tabs 1.1 and 1.6.

Most Populous Population (millions)

GNI Per Capita (U.S. $) Least Populousa

Population (thousands)

GNI Per Capita (U.S. $)

1. China 1,325 2,940 1. Palau 20 8,630 2. India 1,140 1,040 2. St. Kitts and Nevis 49 10,870 3. United States 304 47,930 3. Marshall Islands 60 3,270 4. Indonesia 227 1,880 4. Dominica 73 4,750 5. Brazil 192 7,300 5. Antigua and Barbuda 87 13,200 6. Pakistan 166 950 6. Seychelles 87 10,220 7. Bangladesh 160 520 7. Kiribati 97 2,040 8. Nigeria 151 1,170 8. Tonga 104 2,690 9. Russian Federation 142 9,660 9. Grenada 104 5,880

10. Japan 128 38,130 10. St. Vincent and the Grenadines

109 5,050

11. Mexico 106 9,990 11. Micronesia 110 2,460 12. Philippines 90 1,890 12. São Tomé and Príncipe 160 1,030

aCriteria for inclusion in the least-populous rankings: United Nations member as of mid-2010, with 2008 comparable population and GNI per capita data in tab. 1.6 in the source.

60 PART ONE Principles and Concepts

example, under-5 mortality rates improve as mothers’ education levels rise, as seen in the country examples in Figure 2.5.

The well-performing developing countries are much closer to the devel- oped world in health and education standards than they are to the lowest- income countries.17 Although health conditions in East Asia are relatively good, sub-Saharan Africa continues to be plagued by problems of malnourish- ment, malaria, tuberculosis, AIDS, and parasitic infections. Despite progress, South Asia continues to have high levels of illiteracy, low schooling attainment, and undernourishment. Still, in fields such as primary school completion, low- income countries are also making great progress; for example, enrollments in India are up from 68% in the early 1990s to a reported 94% by 2008.

Higher Levels of Inequality and Absolute Poverty

Globally, the poorest 20% of people receive just 1.5% of world income. The lowest 20% now roughly corresponds to the approximately 1.2 billion people

TABLE 2.6 Primary School Enrollment and Pupil-Teacher Ratios, 2010

Source: Data from World Bank, World Development Indicators, 2010 (Washington, D.C.: World Bank, 2010), tabs 2.11 and 2.12.

Region or Group Net Primary School

Enrollment (%) Primary Pupil- Teacher Ratio

Income Group     Low 80 45 Lower Middle 87 23a Upper Middle 94 22 High 95 15 Region     East Asia and Pacific 93a 19 Latin America and the Caribbean 94 25 Middle East and North Africa 91 24 South Asia 86 40a Sub-Saharan Africa 73 49 Europe and Central Asia 92 16

aData for 2009.

FIGURE 2.4 Under-5 Mortality Rates, 1990 and 2012

Source: Data drawn from World Bank, World Development Indicators, accessed 22 Sept. 2013 Reprinted with permission.

200

160

120

80

40

0 D

ea th

s p

er 1

,0 00

li ve

b ir

th s

Low income

Lower middle income

Upper middle income

High income

1990

2012

61CHAPTER 2 Comparative Economic Development

living in extreme poverty on less than $1.25 per day at purchasing power parity.18

Bringing the incomes of those living on less than $1.25 per day up to this minimal poverty line would require less than 2% of the incomes of the world’s wealthiest 10%.19 Thus, the scale of global inequality is also immense.

But the enormous gap in per capita incomes between rich and poor nations is not the only manifestation of the huge global economic disparities. To appreciate the breadth and depth of deprivation in developing countries, it is also necessary to look at the gap between rich and poor within individual developing countries. Very high levels of inequality—extremes in the relative incomes of higher- and lower-income citizens—are found in many middle-income countries, partly because Latin American countries historically tend to be both middle-income and highly unequal. Several African countries, including Sierra Leone, Lesotho, and South Africa, also have among the highest levels of inequality in the world.20 Inequality is particularly high in many resource-rich developing countries, notably in the Mid- dle East and sub-Saharan Africa. Indeed, in many of these cases, inequality is sub- stantially higher than in most developed countries (where inequality has in many cases been rising). But inequality varies greatly among developing countries, with generally much lower inequality in Asia. Consequently, we cannot confine our attention to averages; we must look within nations at how income is distributed to ask who benefits from economic development and why.

Corresponding to their low average income levels, a large majority of the extreme poor live in the low-income developing countries of sub-Saharan Africa and South Asia. Extreme poverty is due in part to low human capital but also to social and political exclusion and other deprivations. Great progress has already been made in reducing the fraction of the developing world’s population living on less than $1.25 per day and raising the incomes of those still below that level, but much remains to be done, as we examine in detail in Chapter 5.

FIGURE 2.5 Correlation between Under-5 Mortality and Mother’s Education

Source: International Bank for Reconstruction and Development/World Bank, World Development Indicators, 2007 (Washington, D.C.: World Bank, 2007), p. 119. Reprinted with permission.

200

150

100

50

0

D e

a th

s p

e r

1 ,0

0 0

li ve

b ir

th s

Tanzania (2004–05)

Bolivia (2003)

Egypt (2005)

Philippines (2003)

Bangladesh (2004)

No education

Mother’s Education

Some primary

Complete primary/some secondary

Complete secondary/some higher

62 PART ONE Principles and Concepts

Development economists use the concept of absolute poverty to represent a specific minimum level of income needed to satisfy the basic physical needs of food, clothing, and shelter in order to ensure continued survival. A problem, however, arises when one recognizes that these minimum subsistence levels will vary from country to country and region to region, reflecting different physiological as well as social and economic requirements. Economists have therefore tended to make conservative estimates of world poverty in order to avoid unintended exaggeration of the problem.

The incidence of extreme poverty varies widely around the developing world. The World Bank estimates that the share of the population living on less than $1.25 per day is 9.1% in East Asia and the Pacific, 8.6% in Latin America and the Caribbean, 1.5% in the Middle East and North Africa, 31.7% in South Asia, and 41.1% in sub-Saharan Africa.21 The share of the world population living below this level had fallen encouragingly to an estimated 21% by 2010, though there are concerns that the pace of poverty reduction may have slowed recently.22 But as Figure 2.6 shows, the number living on less than $1.25 per day fell from about 1.9 billion in 1981 to about 1.2 billion by 2008, despite a 59% increase in the developing world’s population.

Extreme poverty represents great human misery, and so redressing it is a top priority of international development. Development economists have also increasingly focused on ways in which poverty and inequality can lead to slower growth. That is, not only do poverty and inequality result from distorted growth, but they can also cause it. This relationship, along with

Absolute poverty The situation of being unable or only barely able to meet the subsistence essentials of food, clothing, shelter, and basic health care.

FIGURE 2.6 Number of People Living in Poverty by Region, 1981–2008

Source: World Bank, “World Bank sees progress against extreme poverty, but flags vulnerability,” April 2012, http://web.worldbank.org/WBSITE/EXTERNAL/EXTDEC/ EXTRESEARCH/EXTPROGRAMS/EXTPOVRES/EXTPOVCALNET/0,,contentMDK: 22716987~pagePK:64168435~theSitePK:5280443~isCURL:Y,00.html.

0

2,000

1,500

1,000

500

P o

p u

la ti

o n

li vi

n g

u n

d er

$ 1.

25 a

d ay

in 2

00 5

P P

P (

m il

li o

n s)

2008200520021999199619931990198719841981

Sub-Saharan Africa

South Asia

Rest of the World

Number of poor by region, 1981-2008

East Asia and the Pacific

63CHAPTER 2 Comparative Economic Development

measurements of inequality and poverty and strategies to address these problems, is examined in depth in Chapter 5; because of their central importance in development, poverty reduction strategies are examined throughout the text.

Higher Population Growth Rates

Global population has skyrocketed since the beginning of the industrial era, from just under 1 billion in 1800 to 1.65 billion in 1900 and to over 6 billion by 2000. World population topped 7 billion by 2012. Rapid population growth began in Europe and other now developed countries. But in recent decades, most population growth has been centered in the developing world. Com- pared with the developed countries, which often have birth rates near or even below replacement (zero population growth) levels, the low-income develop- ing countries have very high birth rates. More than five-sixths of all the people in the world now live in developing countries; and some 97% of net popula- tion growth (births minus deaths) in 2012 took place in developing regions.

But population dynamics varies widely among developing countries. Popula- tions of some developing countries, particularly in Africa, continue to grow rap- idly. From 1990 to 2008, population in the low-income countries grew at 2.2% per year, compared to 1.3% in the middle-income countries (the high-income coun- tries grew at 0.7% per year, reflecting both births and immigration).23

Middle-income developing countries show greater variance, with some having achieved lower birth rates closer to those prevailing in rich countries. The birth rate is about three times as high in the low-income countries as in the high-income countries. In sub-Saharan Africa, the annual birth rate is 39 per 1,000—four times the rate in high-income countries. Intermediate but still relatively high birth rates are found in South Asia (24), the Middle East and North Africa (24), and Latin America and the Caribbean (19). East Asia and the Pacific have a moderate birth rate of 14 per 1,000, partly the result of birth control policies in China. The very wide range of crude birth rates around the world is illustrated in Table 2.7. As of 2010, the average rate of population growth was about 1.4% in the developing countries.

A major implication of high birth rates is that the active labor force has to support proportionally almost twice as many children as it does in richer countries. By contrast, the proportion of people over the age of 65 is much

Crude birth rate The number of children born alive each year per 1,000 population.

TABLE 2.7 Crude Birth Rates Around the World, 2012

Source: Population Reference Bureau, Population Data Sheet, 2012.

45+ Chad, Dem. Rep. of Congo, Mali, Niger, Uganda, Zambia 40–44 Afghanistan, Angola, Benin, Burkina Faso, Liberia, Malawi, Mozambique, Nigeria, Somalia, South Sudan, Tanzania 35–39 Central African Republic, Côte d’Ivoire, Eritrea, Iraq, Jordan, Kenya, Madagascar, Senegal, Sierra Leone, Yemen 30–34 Ethiopia, Ghana, Papua New Guinea, Sudan, Timor-Leste, Vanatu, Zimbabwe 25–29 Algeria, Bolivia, Cambodia, Egypt, Guatemala, Haiti, Honduras, Kyrgyzstan, Pakistan, Philippines, Samoa, Tonga 20–24 Dominican Republic, El Salvador, India, Libya, Mexico, Peru, Saudi Arabia, South Africa, Venezuela 15–19 Argentina, Brazil, Colombia, Costa Rica, Indonesia, Jamaica, Sri Lanka, Turkey, Vietnam 10–14 Australia, Canada, China, France, Russia, United Kingdom, United States <10 Austria, Croatia, Germany, Hungary, Italy, Japan, South Korea, Serbia, Portugal, Taiwan

64 PART ONE Principles and Concepts

greater in the developed nations. Both older people and children are often referred to as an economic dependency burden in the sense that they must be supported financially by the country’s labor force (usually defined as citizens between the ages of 15 and 64). In low-income countries, there are 66 children under 15 for each 100 working-age (15–65) adults, while in middle-income countries, there are 41 and in high-income countries just 26. In contrast, low- income countries have just 6 people over 65 per 100 working-age adults, com- pared with 10 in middle-income countries and 23 in high-income countries. Thus, the total dependency ratio is 72 per 100 in low-income countries and 49 per 100 in high-income countries.24 But in rich countries, older citizens are supported by their lifetime savings and by public and private pensions. In contrast, in developing countries, public support for children is very limited. So dependency has a further magnified impact in developing countries.

We may conclude, therefore, that not only are developing countries char- acterized by higher rates of population growth, but they must also contend with greater dependency burdens than rich nations, though with a wide gulf between low- and middle-income developing countries. The circumstances and conditions under which population growth becomes a deterrent to economic development is a critical issue and is examined in Chapter 6.

Greater Social Fractionalization

Low-income countries often have ethnic, linguistic, and other forms of social divisions, sometimes known as fractionalization. This is sometimes asso- ciated with civil strife and even violent conflict, which can lead developing societies to divert considerable energies to working for political accommoda- tions if not national consolidation. It is one of a variety of governance chal- lenges many developing nations face. There is some evidence that many of the factors associated with poor economic growth performance in sub-Saharan Africa, such as low schooling, political instability, underdeveloped financial systems, and insufficient infrastructure, can be statistically explained by high ethnic fragmentation.25

The greater the ethnic, linguistic, and religious diversity of a country, the more likely it is that there will be internal strife and political instability. Some of the most successful development experiences—South Korea, Taiwan, Sin- gapore, and Hong Kong—have occurred in culturally homogeneous societies.

But today, more than 40% of the world’s nations have more than five sig- nificant ethnic populations. In most cases, one or more of these groups face serious problems of discrimination, social exclusion, or other systematic dis- advantages. Over half of the world’s developing countries have experienced some form of interethnic conflict. Ethnic and religious conflicts leading to widespread death and destruction have taken place in countries as diverse as Afghanistan, Rwanda, Mozambique, Guatemala, Mexico, Sri Lanka, Iraq, India, Kyrgyzstan, Azerbaijan, Somalia, Ethiopia, Liberia, Sierra Leone, Angola, Myanmar, Sudan, the former Yugoslavia, Indonesia, and the DRC.

Conflict can derail what had otherwise been relatively positive develop- ment progress, as in Côte d’Ivoire since 2002 (see Chapter 14 and the case study for Chapter 5). There is, however, a heartening trend since the late 1990s toward more successful resolution of conflicts and fewer new conflicts.

Dependency burden The proportion of the total popu- lation aged 0 to 15 and 65+, which is considered economi- cally unproductive and there- fore not counted in the labor force.

Fractionalization Significant ethnic, linguistic, and other social divisions within a country.

65CHAPTER 2 Comparative Economic Development

If development is about improving human lives and providing a widening range of choice to all peoples, racial, ethnic, caste, or religious discrimination is pernicious. For example, throughout Latin America, indigenous popula- tions have significantly lagged behind other groups on almost every measure of economic and social progress. Whether in Bolivia, Brazil, Peru, Mexico, Guatemala, or Venezuela, indigenous groups have benefited little from overall economic growth. Being indigenous makes it much more likely that an indi- vidual will be less educated, in poorer health, and in a lower socioeconomic stratum than other citizens.26 This is particularly true for indigenous women. Moreover, descendants of African slaves brought forcefully to the western hemisphere continue to suffer discrimination in countries such as Brazil.

Ethnic and religious diversity need not necessarily lead to inequality, tur- moil, or instability, and unqualified statements about their impact cannot be made. There have been numerous instances of successful economic and social integration of minority or indigenous ethnic populations in countries as diverse as Malaysia and Mauritius. And in the United States, diversity is often cited as a source of creativity and innovation. The broader point is that the ethnic and religious composition of a developing nation and whether or not that diversity leads to conflict or cooperation can be important determinants of the success or failure of development efforts.27

Larger Rural Populations but Rapid Rural-to-Urban Migration

One of the hallmarks of economic development is a shift from agriculture to manufacturing and services. In developing countries, a much higher share of the population lives in rural areas, and correspondingly fewer in urban areas, as seen in Table 2.8. Although modernizing in many regions, rural areas are poorer and tend to suffer from missing markets, limited information, and social stratification. A massive population shift is also under way as hundreds of mil- lions of people are moving from rural to urban areas, fueling rapid urbaniza- tion, with its own attendant problems. The world as a whole has just crossed the 50% threshold: For the first time in history, more people live in cities than

TABLE 2.8 The Urban Population in Developed Countries and Developing Regions

Source: Population Reference Bureau, 2009 World Data Sheet.

Region Population (millions, 2009) Urban Share (%)

World 6,810 50 More developed countries 1,232 75 Less developed countries 5,578 44 Sub-Saharan Africa 836 35 Northern Africa 205 50 Latin America and the

Caribbean 580 77

Western Asia 231 64 South-central Asia 1,726 31 Southeast Asia 597 43 East Asia 1,564 51 Eastern Europe 295 69

66 PART ONE Principles and Concepts

in rural areas. But sub-Saharan Africa and most of Asia remain predominantly rural. Migration and agriculture issues are examined in Chapters 7 and 9.

Lower Levels of Industrialization and Manufactured Exports

One of the most widely used terminologies for the original Group of Seven (G7) countries28 and other advanced economies such as smaller European countries and Australia is the “industrial countries.” Industrialization is associated with high productivity and incomes and has been a hallmark of modernization and national economic power. It is no accident that most devel- oping-country governments have made industrialization a high national pri- ority, with a number of prominent success stories in Asia.

Table 2.9 shows the relationship between employment and share of GDP in agriculture, industry, and services in selected developing and developed coun- tries, in the 2004 to 2008 period. Generally, developing countries have a far higher share of employment in agriculture than developed countries. More- over, in developed countries, agriculture represents a very small share of both employment and output—about 1% to 2% in Canada, the United States and United Kingdom—although productivity is not below the average for these economies as a whole. This is in sharp contrast to a majority of developing nations, which have relatively low productivity in agriculture in comparison

TABLE 2.9 Share of the Population Employed in the Agricultural, Industrial, and Service Sectors in Selected Countries, 2004–2008 (%)

Source: World Bank, World Development Indicators, 2010 (Washington, D.C.: World Bank, 2010), tabs. 2.3 and 4.2.

  Agriculture Industry Services

  Males Females

Share of GDP (2008) Males Females

Share of GDP (2008) Males Females

Share of GDP (2008)

Africa Egypt 28 43 13 26 6 38 46 51 49 Ethiopia 12 6 44 27 17 13 61 77 42 Madagascar 82 83 25 5 2 17 13 16 57 Mauritius 10 8 4 36 26 29 54 66 67 South Africa 11 7 3 35 14 34 54 80 63 Asia Bangladesh 42 68 19 15 13 29 43 19 52 Indonesia 41 41 14 21 15 48 38 44 37 Malaysia 18 10 10 32 23 48 51 67 42 Pakistan 36 72 20 23 13 27 41 15 53 Philippines 44 24 15 18 11 32 39 65 53 South Korea 7 8 3 33 16 37 60 74 60 Thailand 43 40 12 22 19 44 35 41 44 Vietnam 56 60 22 21 14 40 23 26 38 Latin America Colombia 27 6 9 22 16 36 51 78 55 Costa Rica 18 5 7 28 13 29 54 82 64 Mexico 19 4 4 31 18 37 50 77 59 Nicaragua 42 8 19 20 18 30 38 73 51 Developed Countries United Kingdom 2 1 1 32 9 24 66 90 76 United States 2 1 1 30 9 22 68 90 77

Note: Ethiopia agricultural employment reflects limited coverage.

67CHAPTER 2 Comparative Economic Development

to other sectors of their own economies—particularly industry. Madagascar is a dramatic example: while about 82% of both men and women worked in agri- culture, it represented only a quarter of total output. In Indonesia, 41% of both men and women worked in agriculture, but it represented just 14% of output. The proportion of women who work in the agricultural sector varies greatly across the developing world. Generally, in Latin America a significantly higher proportion of men work in agriculture than women; but in numerous countries in Africa and Asia, a larger proportion of women work in agriculture.

Table 2.10 reveals the structural transformation of employment that has been occurring in developing countries. Where available, the table shows employment shares in both 1990–1992 and 2008–2011 periods. There have been substantial declines over this two-decade period in the share in employ- ment in agriculture in most developing countries for which comparable data is available. For example, in Indonesia the proportion of men who work in agriculture fell from 54% to 37%; and the proportion of women who work in agriculture fell from 57% to 35%. Partial exceptions include Pakistan and Honduras, for which the share of women’s agricultural employment rose by approximately as much as that of men fell.

At the same time, the share of employment in industry in many devel- oped countries is smaller now than in some developing countries, particularly among women, as developed countries continue their secular trend to switch to from industry to service sector employment. However, many developed- country industrial jobs require high skills and pay high wages.

Relatively few countries managed a substantial gain of the fraction in manufacturing in this period; Indonesia, Turkey, and Mexico showed modest gains, particularly for men. (Other evidence indicates that a large fraction of global manufacturing jobs were gained in one country—China—during this period; but comparable data for China were unavailable for comparison.) The share of industrial employment in Africa remains low for both men and women in most countries.

Along with lower industrialization, developing nations tended to have a higher dependence on primary exports. Most developing countries have diver- sified away from agricultural and mineral exports to some degree. The middle- income countries are rapidly catching up with the developed world in the share of manufactured goods in their exports, even if these goods are typically less advanced in their skill and technology content. However, the low-income coun- tries, particularly those in Africa, remain highly dependent on a relatively small number of agricultural and mineral exports. Africa will need to continue its efforts to diversify its exports. We examine this topic in Chapter 12.

Adverse Geography

Many analysts argue that geography must play some role in problems of agri- culture, public health, and comparative development more generally. Land- locked economies, common in Africa, often have lower incomes than coastal economies.29 As can be observed on the map on the inside cover, develop- ing countries are primarily tropical or subtropical, and this has meant that they suffer more from tropical pests and parasites, endemic diseases such as malaria, water resource constraints, and extremes of heat. A great concern

68 P

A R

T O

N E

P rin

cip les an

d C

o n

cep ts

TABLE 2.10 Share of the Population Employed in the Agricultural, Industrial, and Service Sectors in Selected Countries, 1990–92 and 2008–2011 (%)

Source: World Bank, World Development Indicators, 2013 (Washington, D.C.: World Bank, 2013), tab. 2.3.

  Agriculture Industry Services  

  Males Females Males Females Males Females  

  % of Male Employment

% of Female Employment

% of Male Employment

% of Female Employment

% of Male Employment

% of Female Employment

  1990–92 2008–11 1990–92 2008–11 1990–92 2008–11 1990–92 2008–11 1990–92 2008–11 1990–-92 2008–-11 Region

Cameroon .. 49 .. 58 .. 13 .. 12 .. 38 .. 30 Africa Egypt, Arab Rep. 35 28 52 46 25 27 10 6 41 44 37 49 Africa Liberia .. 50 .. 48 .. 14 .. 5 .. 37 .. 47 Africa Mauritius 15 9 13 7 36 32 48 21 48 59 39 73 Africa Namibia 45 23 52 8 21 24 8 9 34 53 40 83 Africa Indonesia 54 37 57 35 15 24 13 15 31 40 31 50 Asia Malaysia 23 16 20 9 31 31 32 21 46 53 48 71 Asia Pakistan 45 37 69 75 20 22 15 12 35 41 16 13 Asia Philippines 53 41 32 23 17 18 14 10 29 41 55 68 Asia Thailand 60 41 62 37 18 23 13 18 22 37 25 45 Asia Turkey 33 18 72 39 26 31 11 15 41 51 17 45 Asia Chile 24 14 6 5 32 31 15 10 45 55 79 85 Latin America Costa Rica 32 20 5 4 27 25 25 11 41 55 69 84 Latin America Dominican

Republic 26 19 3 2 23 21 21 7 52 47 76 60 Latin America

Honduras 53 50 6 12 18 19 25 21 29 31 69 67 Latin America Mexico 34 19 11 4 25 30 19 18 41 51 70 78 Latin America Canada 6 3 2 1 31 32 11 10 64 65 87 89 Developed Japan 6 4 7 4 40 33 27 15 54 62 65 80 Developed United Kingdom 3 2 1 1 41 29 16 8 55 69 82 91 Developed United States 4 2 1 1 34 25 14 7 62 72 85 92 Developed

Note: Country selection reflects that only a limited number of countries are covered or have data over time. Data represent most recent in timeframe if average for the period is not available.

69CHAPTER 2 Comparative Economic Development

going forward is that global warming is projected to have its greatest negative impact on Africa and South Asia (see Chapter 10).30

The extreme case of favorable physical resource endowment is the oil- rich Persian Gulf states. At the other extreme are countries like Chad, Yemen, Haiti, and Bangladesh, where endowments of raw materials and minerals and even fertile land are relatively minimal. However, as the case of the DRC shows vividly, high mineral wealth is no guarantee of development success. Conflict over the profits from these industries has often led to a focus on the distribution of wealth rather than its creation and to social strife, undemocratic governance, high inequal- ity, and even armed conflict, in what is called the “curse of natural resources.”

Clearly, geography is not destiny; high-income Singapore lies almost directly on the equator, and parts of southern India have exhibited enormous economic dynamism in recent years. Prior to colonization, some tropical and subtropical regions had higher incomes per capita than Europe. However, the presence of common and often adverse geographic features in comparison to temperate zone countries means it is beneficial to study tropical and subtropical develop- ing countries together for some purposes. Redoubled efforts are now under way to extend the benefits of the green revolution and tropical disease control to sub- Saharan Africa. In section 2.7 of this chapter, we add further perspectives on the possible indirect roles of geography in comparative development.

Underdeveloped Markets

Imperfect markets and incomplete information are far more prevalent in developing countries, with the result that domestic markets, notably but not only financial markets, have worked less efficiently, as examined in Chapters 4, 11, and 15. In many developing countries, legal and institutional founda- tions for markets are extremely weak.

Some aspects of market underdevelopment are that they often lack (1) a legal system that enforces contracts and validates property rights; (2) a sta- ble and trustworthy currency; (3) an infrastructure of roads and utilities that results in low transport and communication costs so as to facilitate interre- gional trade; (4) a well-developed and efficiently regulated system of banking and insurance, with broad access and with formal credit markets that select projects and allocate loanable funds on the basis of relative economic profit- ability and enforce rules of repayment; (5) substantial market information for consumers and producers about prices, quantities, and qualities of products and resources as well as the creditworthiness of potential borrowers; and (6) social norms that facilitate successful long-term business relationships. These six factors, along with the existence of economies of scale in major sec- tors of the economy, thin markets for many products due to limited demand and few sellers, widespread externalities (costs or benefits that accrue to com- panies or individuals not doing the producing or consuming) in production and consumption, and poorly regulated common property resources (e.g., fisheries, grazing lands, water holes) mean that markets are often highly imperfect. Moreover, information is limited and costly to obtain, thereby often causing goods, finances, and resources to be misallocated. And we have come to understand that small externalities can interact in ways that add up to very large distortions in an economy and present the real possibility of an

Infrastructure Facilities that enable economic activity and markets, such as transporta- tion, communication and distribution networks, utili- ties, water, sewer, and energy supply systems.

Resource endowment A nation’s supply of usable factors of production, including mineral deposits, raw materials, and labor.

70 PART ONE Principles and Concepts

underdevelopment trap (see Chapter 4). The extent to which these imperfect markets and incomplete information systems justify a more active role for gov- ernment (which is also subject to similar problems of incomplete and imperfect information) is an issue that we will be dealing with in later chapters. But their existence remains a common characteristic of many developing nations and an important contributing factor to their state of underdevelopment.31

Lingering Colonial Impacts and Unequal International Relations

Colonial Legacy Most developing countries were once colonies of Europe or otherwise dominated by European or other foreign powers, and institutions created during the colonial period often had pernicious effects on development that in many cases have persisted to the present day. Despite important varia- tions that proved consequential, colonial era institutions often favored extrac- tors of wealth rather than creators of wealth, harming development then and now. Both domestically and internationally, developing countries have more often lacked institutions and formal organizations of the type that have bene- fited the developed world: Domestically, on average, property rights have been less secure, constraints on elites have been weak, and a smaller segment of soci- ety has been able to gain access to and take advantage of economic opportuni- ties.32 Problems with governance and public administration (see Chapter 11), as well as poorly performing markets, often stem from poor institutions.

Decolonization was one of the most important historical and geopolitical events of the post–World War II era. More than 80 former European colonies have joined the United Nations. But several decades after independence, the effects of the colonial era linger for many developing nations, particularly the least developed ones.

Colonial history matters not only or even primarily because of stolen resources but also because the colonial powers determined whether the legal and other institutions in their colonies would encourage investments by (and in) the broad population or would instead facilitate exploitation of human and other resources for the benefit of the colonizing elite and create or reinforce extreme inequality. Development-facilitating or development-inhibiting institutions tend to have a very long life span. For example, when the conquered colonial lands were wealthier, there was more to steal. In these cases, colonial powers favored extractive (or “kleptocratic”) institutions at the expense of ones that encouraged productive effort. When settlers came in large numbers to live permanently, incomes ultimately were relatively high, but the indigenous populations were largely annihilated by disease or conflict, and descendants of those who sur- vived were exploited and blocked from advancement. A growing body of evi- dence demonstrates that practices such as forced labor had pernicious effects on human development even centuries after they were discontinued (see Box 2.3).

In a related point of great importance, European colonization often created or reinforced differing degrees of inequality, often correlated with ethnicity, which have also proved remarkably stable over the centuries. In some respects, postcolonial elites in many developing countries largely took over the exploit- ative role formerly played by the colonial powers. High inequality sometimes emerged as a result of slavery in regions where comparative advantage in crops

Property rights The acknowledged right to use and benefit from a tangible (e.g., land) or intangible (e.g., intellectual) entity that may include owning, using, deriv- ing income from, selling, and disposing.

Imperfect market A market in which the theoretical assumptions of perfect com- petition are violated by the existence of, for example, a small number of buyers and sellers, barriers to entry, and incomplete information.

Incomplete information The absence of information that producers and consumers need to make efficient deci- sions resulting in underper- forming markets.

71CHAPTER 2 Comparative Economic Development

BOX 2.3 FINDINGS The Persistent Effects of Colonial Forced Labor on Poverty and Development

In a 2010 study, Melissa Dell used historical district-level data to examine the long-run impacts of the mita forced labor system in Peru and Bolivia, which “required over 200 indigenous communities to send one-seventh of their adult male population to work in the Potosi silver and Huancavelica mercury mines between 1573 and 1812.” Forced labor can severely harm subjected communities. But Dell finds even today—two centuries later—districts covered by the mita system have lower household consumption and higher probability of stunting in children.

Can development economists conclude with con- fidence that a colonial system ending two centuries ago is the cause of worse performance in the districts it affected? In principle, such correlations could be due to observed or unobserved factors other than the mita. For example, households in mita districts may have been less well off to begin with. To address this question, Dell employed an important tool used by development economists to establish causal effects, known as regression discontinuity design (RDD).

RDD has many uses, including evaluation of devel- opment programs. In evaluating a program, if each individual is associated with an “assignment vari- able,” z, and a “treatment” is assigned to individuals with a value of z less than or equal to a cutoff level z0, then the impact of the treatment on an outcome vari- able, y, can be identified by comparing observations of those who started just below the threshold z0 with those who started just above it. For this group, any difference in the outcome variable between people on each side of the discontinuity would be caused by the treatment. The assignment z can represent many types of threshold variables, including income, birth date, test scores, or a geographic boundary. And it turns out that a very wide range of impacts can be considered as a treatment—whatever impacts only people who are on one side of a threshold, provided that all relevant influences other than treatment vary smoothly across the threshold. Economists have learned that RDD estimates have statistically reliable properties that in

some circumstances can make these studies virtually as informative as a randomized trial.

One basic assumption of RDD is that individuals just below and just above the cutoff are otherwise similar and have the same potential outcomes in the absence of the treatment. This assumption means that individuals cannot “sort themselves” to be just under the cutoff (or over the cutoff, if that is the incentive). For example, people cannot pretend to be poorer in order to get into a poverty program. Otherwise, the estimated effect can be compounded with the char- acteristics of those people who respond by sorting themselves (e.g., people with higher cognitive skills).

Dell’s RDD strategy was to use longitude-latitude, or simply distance to mines, as the assignment vari- able to predict the mita coverage. The effect of the mita system on social or economic outcomes can be estimated by comparing districts with and without the mita system among those close to the mita cov- erage boundary. These districts were considered likely to be similar in all respects except for the mita; and indeed, Dell found that prior to the mita system, fac- tors such as tax rates, steepness of terrain, and ethnic distribution were similar across the boundaries that she studies. Using this strategy, Dell concluded that the “mita effect” lowers household consumption by approximately 25% and that it increases child stunting “by around 6 percentage points.” These are really striking findings: Two centuries have passed since the mita boundary line carried any legal mean- ing whatsoever.

Dell then asked, “Why would the mita affect eco- nomic prosperity nearly 200 years after its aboli- tion?” While “there exist many potential channels,” Dell proposed, “the mita’s influence has persisted through its impacts on land tenure and public good provision.” Outside the mita district boundaries, the Spanish hacienda system emerged—it was a feudal sys- tem, not a market in which labor was free. While the measured impact of the mita likely would have been even worse in comparison with “secure, enfranchised

72 PART ONE Principles and Concepts

such as sugarcane could be profitably produced on slave plantations. It also emerged where a large, settled indigenous population could be coerced into labor. This history had long-term consequences, particularly in Latin Amer- ica.33 Where inequality was extreme, the result was less movement toward democratic institutions, less investment in public goods, and less widespread investment in human capital (education, skills, and health). These are among the ways in which extreme inequality is harmful to development and so is also an important long-term determinant of comparative development. We return to these themes later in this chapter.

The European colonial powers also had a dramatic and long-lasting impact on the economies and political and institutional structures of their African and Asian colonies by their introduction of three powerful and tradition-shattering ideas: private property, personal taxation, and the requirement that taxes be paid in money rather than in kind. These innovations were introduced in ways that facilitated elite rule rather than broad-based opportunity. The worst impact of colonization was probably felt in Africa, especially if one also con- siders the earlier slave trade. Whereas in former colonies such as India local people played a role in colonial governance, in Africa most governance was administered by expatriates. Other well-documented impacts included lasting damage to social trust.34

In Latin America, a longer history of political independence plus a more shared colonial heritage (Spanish and Portuguese) has meant that in spite of geographic and demographic diversity, the countries possess relatively simi- lar economic, social, and cultural institutions and face similar problems, albeit with particular hardships for indigenous peoples and descendants of slaves. Latin American countries have long been middle-income but rarely have advanced to high-income status—reflecting a situation now known as the “middle-income trap.” In Asia, different colonial heritages and the diverse cul- tural traditions of the people have combined to create different institutional and social patterns in countries such as India (British), the Philippines (Span- ish and American), Vietnam (French), Indonesia (Dutch), Korea (Japanese), and China (not formally colonized but dominated by a variety of foreign powers).35

To a widely varying degree, newly independent nations continued to experi- ence foreign domination by former colonial powers and the United States, and in a number of countries by the Soviet Union, particularly during the Cold War

smallholders,” Dell contrasted the two actual histori- cal experiences in this region. Some exploitive condi- tions are worse than land inequality. Dell pointed out that the land tenure system in non-mita districts was more stable compared to mita districts, where there was no system of enforceable peasant titling even after the mita ended. For example, Dell cites a judicial pro- cedure used in mita districts to seize land from peas- ants by falsely claiming their land was abandoned.

Large landowners also had a profit incentive and the political influence to get more roads built in their dis- tricts. Dell argued that in this region of Peru, “large landowners—while they did not aim to promote eco- nomic prosperity for the masses—did shield individu- als from exploitation by a highly extractive state and did ensure public goods.”

Source: Melissa Dell. “The Persistent Effects of Peru’s Mining Mita.” Econometrica 78(2010): 1863–1903.

73CHAPTER 2 Comparative Economic Development

period. The diversity of colonial experiences is one of the important factors that help explain the wide spectrum of development outcomes in today’s world.

External Dependence Relatedly, developing countries have also been less well organized and influential in international relations, with sometimes adverse consequences for development. For example, agreements within the World Trade Organization (WTO) and its predecessors concerning matters such as agricultural subsidies in rich countries that harm developing-country farm- ers and one-sided regulation of intellectual property rights have often been rel- atively unfavorable to the developing world. The “Doha Development Round” of trade negotiations that began in 2001 was supposed to rectify some of these imbalances, but talks have been essentially stalled since 2008 (see Chapter 12). During debt crises in the 1980s and 1990s, the interests of international banks often prevailed over those of desperately indebted nations (discussed in Chapter 13). More generally, developing nations have weaker bargaining posi- tions than developed nations in international economic relations. Developing nations often also voice great concern over various forms of cultural depen- dence, from news and entertainment to business practices, lifestyles, and social values. The potential importance of these concerns should not be underesti- mated, either in their direct effects on development in its broader meanings or indirect impacts on the speed or character of national development.

Developing nations are also dependent on the developed world for envi- ronmental preservation, on which hopes for sustainable development depend. Of greatest concern, global warming is projected to harm developing regions more than developed ones; yet both accumulated and current greenhouse gas emissions still largely originate in the high-income countries, despite the role of developing-country deforestation and growing emissions from lower- middle-income countries such as China and India. Thus the developing world endures what may be called environmental dependence, in which it must rely on the developed world to cease aggravating the problem and to develop solu- tions, including mitigation at home and assistance in developing countries. This topic is explored further in Chapter 10.

2.5 How Low-Income Countries Today Differ from Developed Countries in Their Earlier Stages

The position of developing countries today is in many important ways sig- nificantly different from that of the currently developed countries when they embarked on their era of modern economic growth. We can identify eight signifi- cant differences in initial conditions that require a special analysis of the growth prospects and requirements of modern economic development:

1. Physical and human resource endowments

2. Per capita incomes and levels of GDP in relation to the rest of the world

3. Climate

4. Population size, distribution, and growth

74 PART ONE Principles and Concepts

5. Historical role of international migration

6. International trade benefits

7. Basic scientific and technological research and development capabilities

8. Efficacy of domestic institutions

We will discuss each of these conditions with a view toward formulating require- ments and priorities for generating and sustaining economic growth in develop- ing countries.

Physical and Human Resource Endowments

Contemporary developing countries are often less well endowed with natu- ral resources than the currently developed nations were at the time when the latter nations began their modern growth. Some developing nations are blessed with abundant supplies of petroleum, minerals, and raw materials for which world demand is growing; most less developed countries, however— especially in Asia, where more than half of the world’s population resides— are poorly endowed with natural resources. Moreover, in parts of Africa, where natural resources are more plentiful, and geologists anticipate that there is far more yet to be discovered, heavy investments of capital are needed to exploit them, which until very recently has been strongly inhibited by domestic conflict and perhaps Western attitudes. A new wave of invest- ments from China and other “nontraditional investors” has begun to change the picture, though critics are raising concerns about the process and foreign appropriation of gains.

The difference in skilled human resource endowments is even more pro- nounced. The ability of a country to exploit its natural resources and to initiate and sustain long-term economic growth is dependent on, among other things, the ingenuity and the managerial and technical skills of its people and its access to critical market and product information at minimal cost.36 Paul Romer argues that today’s developing nations “are poor because their citizens do not have access to the ideas that are used in industrial nations to generate economic value.”37 For Romer, the technology gap between rich and poor nations can be divided into two components, a physical object gap, involving factories, roads, and modern machinery, and an idea gap, including knowledge about marketing, distribution, inventory control, transactions processing, and worker motivation. This idea gap, and what Thomas Homer-Dixon calls the ingenuity gap (the ability to apply innovative ideas to solve practical social and technical problems), between rich and poor nations lies at the core of the development divide. There were no comparative human resource gaps for the now developed countries on the eve of their industrialization.

Relative Levels of Per Capita Income and GDP

The people living in low-income countries have, on average, a lower level of real per capita income than their developed-country counterparts had in the nineteenth century. First of all, nearly 40% of the population of developing

75CHAPTER 2 Comparative Economic Development

countries is attempting to subsist at bare minimum levels. Obviously, the aver- age standard of living in, say, early-nineteenth-century England was nothing to envy or boast about, but it was not as economically debilitating or precari- ous as it is today for a large fraction of people in the 40 or so least developed countries, the people now often referred to as the “bottom billion.”

Second, at the beginning of their modern growth era, today’s developed nations were economically in advance of the rest of the world. They could therefore take advantage of their relatively strong financial position to widen the income gaps between themselves and less fortunate countries in a long period of income divergence. By contrast, today’s developing countries began their growth process at the low end of the international per capita income scale.

Climatic Differences

Almost all developing countries are situated in tropical or subtropical climatic zones. It has been observed that the economically most successful countries are located in the temperate zone. Although social inequality and institutional factors are widely believed to be of greater importance, the dichotomy is more than coincidence. Colonialists apparently created unhelpful “extractive” insti- tutions where they found it uncomfortable to settle. But also, the extremes of heat and humidity in most poor countries contribute to deteriorating soil qual- ity and the rapid depreciation of many natural goods. They also contribute to the low productivity of certain crops, the weakened regenerative growth of forests, and the poor health of animals. Extremes of heat and humidity not only cause discomfort to workers but can also weaken their health, reduce their desire to engage in strenuous physical work, and generally lower their levels of productivity and efficiency. As you will see in Chapter 8, malaria and other serious parasitic diseases are often concentrated in tropical areas. There is evidence that tropical geography does pose significant problems for eco- nomic development and that special attention in development assistance must be given to these problems, such as a concerted international effort to develop a malaria vaccine.38

Population Size, Distribution, and Growth

In Chapter 6, we will examine in detail some of the development problems and issues associated with rapid population growth. At this point, it is sufficient to note that population size, density, and growth constitute another important difference between less developed and developed countries. Before and dur- ing their early growth years, Western nations experienced a very slow rise in population growth. As industrialization proceeded, population growth rates increased primarily as a result of falling death rates but also because of slowly rising birth rates. However, at no time did European and North American countries have natural population growth rates in excess of 2% per annum, and they generally averaged much less.

By contrast, the populations of many developing countries have been increasing at annual rates in excess of 2.5% in recent decades, and some are still rising that fast today. Moreover, the concentration of these large and

76 PART ONE Principles and Concepts

growing populations in a few areas means that many developing countries have considerably higher person-to-land ratios than the European countries did in their early growth years. Finally, in terms of comparative absolute size, no country that embarked on a long-term period of successful economic growth approached the present-day population size of India, Egypt, Pakistan, Indonesia, Nigeria, or Brazil. Nor were their rates of natural increase anything like that of present-day Kenya, the Philippines, Bangladesh, Malawi, or Gua- temala. In fact, many observers doubt whether the Industrial Revolution and the high long-term growth rates of contemporary developed countries could have been achieved or proceeded so fast and with so few setbacks and distur- bances, especially for the very poor, had their populations been expanding so rapidly.

The Historical Role of International Migration

In the nineteenth and early twentieth centuries, a major outlet for rural popu- lations was international migration, which was both widespread and large- scale. More than 60 million people migrated to the Americas between 1850 and 1914, a time when world population averaged less than a quarter of its current levels. In countries such as Italy, Germany, and Ireland, periods of famine or pressure on the land often combined with limited economic oppor- tunities in urban industry to push unskilled rural workers toward the labor- scarce nations of North America and Australia. In Brinley Thomas’s famous description, the “three outstanding contributions of European labor to the American economy—1,187,000 Irish and 919,000 Germans between 1847 and 1855, 418,000 Scandinavians and 1,045,000 Germans between 1880 and 1885, and 1,754,000 Italians between 1898 and 1907—had the character of evacua- tions.”39

Whereas the main thrust of international emigration up to World War I was both distant and permanent, the period since World War II witnessed a resurgence of international migration within Europe itself, which is essentially over short distances and to a large degree temporary. However, the economic forces giving rise to this migration are basically the same; that is, during the 1960s, surplus rural workers from southern Italy, Greece, and Turkey flocked into areas of labor shortages, most notably western Germany and Switzer- land. Similar trends have been observed following the expansion of the Euro- pean Union. The fact that this later migration from regions of surplus labor in southern and southeastern Europe was initially of both a permanent and a nonpermanent nature provided a valuable dual benefit to the relatively poor areas from which these unskilled workers migrated. The home governments were relieved of the costs of providing for people who in all probability would remain unemployed, and because a large percentage of the workers’ earnings were sent home, these governments received a valuable and not insignificant source of foreign exchange.40

Historically, at least in the case of Africa, migrant labor both within and between countries was rather common and did provide some relief for locally depressed areas. Until recently, considerable benefits accrued and numerous potential problems were avoided by the fact that thousands of unskilled laborers in Burkina Faso were able to find temporary work in neighboring Côte d’Ivoire.

77CHAPTER 2 Comparative Economic Development

The same was true for Egyptians, Pakistanis, and Indians in Kuwait and Saudi Arabia; Tunisians, Moroccans, and Algerians in southern Europe; Colombians in Venezuela; and Haitians in the Dominican Republic. However, there is far less scope for reducing the pressures of growing populations in developing countries today through massive international emigration, largely due to the very restrictive nature of immigration laws in modern developed countries.

Despite these restrictions, well over 50 million people from the developing world have managed to migrate to the developed world since 1960. The pace of migration from developing to developed countries—particularly to the United States, Canada, and Australia—has picked up since the mid-1980s to between 2 and 3 million people per year. And the numbers of undocumented or illegal migrants have increased dramatically since 1980. Some people in recipient industrialized nations feel that these migrants are taking jobs away from poor, unskilled citizen workers. Moreover, illegal migrants and their families are often believed to be taking unfair advantage of free local health, educational, and social services, causing upward pressure on local taxes to support these services—despite emerging evidence that legalizing immigra- tion actually provides a net positive effect on reducing deficits as well as to overall economic activity.41 As a result, major debates are now under way in both the United States and Europe regarding the treatment of illegal migrants. Many citizens want severe restrictions on the number of immigrants that are permitted to enter or reside in developed countries.42 The anti-immigration law passed in Arizona in 2010 reinforced the deterrent effect of the Mexico- U.S. border fence and also led many legal immigrants to feel vulnerable; a vociferous political debate surrounded proposed immigration reform leg- islation in the United States in 2013. In Europe, anti-immigrant parties have scored major gains, as in the Netherlands and Sweden in 2010.

The irony of international migration today, however, is not merely that this traditional outlet for surplus people has effectively been closed off but that many of the people who migrate from poor to richer lands are the very ones that developing countries can least afford to lose: the highly educated and skilled. Since the great majority of these migrants move on a permanent basis, this perverse brain drain not only represents a loss of valuable human resources but could also prove to be a serious constraint on the future eco- nomic progress of developing nations. For example, between 1960 and 1990, more than a million high-level professional and technical workers from the developing countries migrated to the United States, Canada, and the United Kingdom. By the late 1980s, Africa had lost nearly one-third of its skilled work- ers, with up to 60,000 middle- and high-level managers migrating to Europe and North America between 1985 and 1990. Sudan, for example, lost 17% of its doctors and dentists, 20% of its university teachers, 30% of its engineers, and 45% of its surveyors. The Philippines lost 12% of its professional workers to the United States, and 60% of Ghanaian doctors came to practice abroad.43

India has been concerned that it may be unable to meet its burgeoning require- ments for information technology workers in its growing high-tech enclaves if emigration to the United States, Canada, and the United Kingdom contin- ues at its current pace.44 Globally, remittances from illegal and legal migrants have been topping $100 million annually in this century and approached $200 billion in 2006.45 Migration, when it is permitted, reduces poverty for migrants

Brain drain The emigration of highly educated and skilled professionals and technicians from the developing countries to the developed world.

78 PART ONE Principles and Concepts

and their families, and most of the poverty-reducing benefits of migration for those remaining in the origin countries come through remittances.46 This is an extremely important resource (see Chapter 14).

Paradoxically, a potential benefit is that the mere possibility of skilled emi- gration may encourage many more workers to acquire information technology or other skills than are ultimately able to leave, leading to a net increase in labor force skills. At least in theory, the result could actually be a “brain gain.”47

The fundamental point remains, however, that the possibility of international migration of unskilled workers on a scale proportional to that of the nineteenth and early twentieth centuries no longer exists to provide an equivalent safety valve for the unskilled contemporary populations of Africa, Asia, and Latin America.

The Growth Stimulus of International Trade

International free trade has been called the “engine of growth” that pro- pelled the development of today’s economically advanced nations during the nineteenth and early twentieth centuries. Rapidly expanding export markets provided an additional stimulus to growing local demands that led to the establishment of large-scale manufacturing industries. Together with a rela- tively stable political structure and flexible social institutions, these increased export earnings enabled the developing countries of the nineteenth cen- tury to borrow funds in the international capital market at very low interest rates. This capital accumulation in turn stimulated further production, made increased imports possible, and led to a more diversified industrial structure. In the nineteenth century, European and North American countries were able to participate in this dynamic growth of international exchange largely on the basis of relatively free trade, free capital movements, and the unfettered inter- national migration of unskilled surplus labor.

In the twentieth century, the situation for many developing countries was very different. With the exception of a few very successful Asian countries, the non-oil-exporting (and even some oil-exporting) developing countries faced formidable difficulties in trying to generate rapid economic growth on the basis of world trade. For much of the past century, many developing coun- tries experienced a deteriorating trade position. Their exports expanded, but usually not as fast as the exports of developed nations. Their terms of trade (the price they receive for their exports relative to the price they have to pay for imports) declined over several decades. Export volume therefore had to grow faster just to earn the same amount of foreign currency as in previous years. Moreover, it is unclear whether the commodity price boom of the early twenty-first century, which reversed only a portion of the long-term price declines, and fueled by the spectacular growth in China, can be maintained. Commodity prices are also subject to large, potentially destabilizing price fluctuations (see Chapter 13).

Where developing countries are successful at becoming lower-cost pro- ducers of competitive products with the developed countries (e.g., textiles, clothing, shoes, some light manufactures), the latter have often resorted to various forms of tariff and nontariff barriers to trade, including “voluntary” import quotas, excessive sanitary requirements, intellectual property claims,

Free trade Trade in which goods can be imported and exported without any barriers in the forms of tariffs, quotas, or other restrictions.

Terms of trade The ratio of a country’s average export price to its average import price.

79CHAPTER 2 Comparative Economic Development

antidumping “investigations,” and special licensing arrangements. But in recent years, an increasing number of developing countries, particularly China and others in East and Southeast Asia, have benefited from expanded manufactures exports to developed countries. We will discuss the economics of international trade and finance in the development context in detail in Part Three.

Basic Scientific and Technological Research and Development Capabilities

Basic scientific research and technological development have played a crucial role in the modern economic growth experience of contemporary developed countries. Their high rates of growth have been sustained by the interplay between mass applications of many new technological innovations based on a rapid advancement in the stock of scientific knowledge and further addi- tions to that stock of knowledge made possible by growing surplus wealth. And even today, the process of scientific and technological advance in all its stages, from basic research to product development, is heavily concentrated in the rich nations, despite the emergence of China and India as destinations for research and development (R&D) activities of multinational corporations. Moreover, research funds are spent on solving the economic and technological problems of concern to rich countries in accordance with their own economic priorities and resource endowments.48

In the important area of scientific and technological research, low-income developing nations in particular are in an extremely disadvantageous position vis-à-vis the developed nations. In contrast, when the latter countries were embarking on their early growth process, they were scientifically and techno- logically greatly in advance of the rest of the world. They could consequently focus on staying ahead by designing and developing new technology at a pace dictated by their long-term economic growth requirements.

Efficacy of Domestic Institutions

Another difference between most developing countries and most developed countries at the time of their early stages of economic development lies in the efficacy of domestic economic, political, and social institutions. By the time of their early industrialization, many developed countries, notably the United Kingdom, the United States, and Canada, had economic rules in place that provided relatively broad access to opportunity for individuals with entrepre- neurial drive. Earlier in the chapter, we noted that high inequality and poor institutions facilitating extraction rather than providing incentives for produc- tivity were often established by colonial powers. Today such extraction may be carried out by powerful local interests as well as foreign ones. But it is very difficult to change institutions rapidly. As Douglass North stresses, even if the formal rules “may be changed overnight, the informal rules usually change only ever so gradually.”49 We will return to the question of economic institu- tions later in the chapter.

The developed countries also typically enjoyed relatively stronger political stability and more flexible social institutions with broader access to mobility. States typically emerged more organically over a longer period of time in the

Research and development (R&D) Scientific investi- gation with a view toward improving the existing quality of human life, products, prof- its, factors of production, or knowledge.

80 PART ONE Principles and Concepts

developed regions, and consolidation as nation-states generally occurred before the industrial era. In contrast, particularly in Africa, national boundaries were more arbitrarily dictated by colonial powers. The “failed state,” and states in danger of becoming so, is a phenomenon of the postcolonial period, with roots in imperial and colonial practices. Although many developing nations have roots in ancient civilizations, a long hiatus often existed between autonomous regimes.

2.6 Are Living Standards of Developing and Developed Nations Converging?

At the dawn of the industrial era, average real living standards in the rich- est countries were no more than three times as great as those of the poorest. Today, the ratio approaches 100 to 1. So as noted by Lant Pritchett, there is no doubt that today’s developed countries have enjoyed far higher rates of eco- nomic growth averaged over two centuries than today’s developing countries, a process known as divergence. Theories of economic growth are discussed in Chapter 3. But in comparing development performance among developing nations and between developed and developing countries, it is appropriate to consider whether, with strenuous economic development efforts being made throughout the developing world, living standards of developing and devel- oped nations are exhibiting convergence.

If the growth experience of developing and developed countries were similar, there are two important reasons to expect that developing coun- tries would be “catching up” by growing faster on average than developed countries. The first reason is due to technology transfer. Today’s developing countries do not have to “reinvent the wheel”; for example, they do not have to use vacuum tubes before they can use semiconductors. Even if royalties must be paid, it is cheaper to replicate technology than to undertake original R&D, partly because one does not have to pay for mistakes and dead ends along the way. This should enable developing countries to “leapfrog” over some of the earlier stages of technological development, moving immedi- ately to high-productivity techniques of production. As a result, they should be able to grow much faster than today’s developed countries are growing now or were able to grow in the past, when they had to invent the tech- nology as they went along and proceed step by step through the historical stages of innovation. (This is known as an “advantage of backwardness,” a term coined by economic historian Alexander Gerschenkron.) In fact, if we confine our attention to cases of successful development, the later a country begins its modern economic growth, the shorter the time needed to double output per worker. For example, Britain doubled its output per person in the first 60 years of its industrial development, and the United States did so in 45 years. South Korea once doubled per capita output in less than 12 years, and China has done so in less than 9 years.

The second reason to expect convergence if conditions are similar is based on factor accumulation. Today’s developed countries have high levels of phys- ical and human capital; in a production function analysis, this would explain their high levels of output per person. But in traditional neoclassical analysis,

Divergence A tendency for per capita income (or output) to grow faster in higher-income countries than in lower-income countries so that the income gap widens across countries over time (as was seen in the two centuries after industrialization began).

Convergence The tendency for per capita income (or output) to grow faster in lower-income countries than in higher-income countries so that lower-income countries are “catching up” over time. When countries are hypoth- esized to converge not in all cases but other things being equal (particularly savings rates, labor force growth, and production technologies), then the term conditional convergence is used.

81CHAPTER 2 Comparative Economic Development

the marginal product of capital and the profitability of investments would be lower in developed countries where capital intensity is higher, provided that the law of diminishing returns applied. That is, the impact of additional capital on output would be expected to be smaller in a developed country that already had a lot of capital in relation to the size of its workforce than in a developing country where capital was scarce. As a result, we would expect higher investment rates in developing countries, either through domestic sources or through attracting foreign investment (see Chapter 14). With higher investment rates, capital would grow more quickly in developing countries until approximately equal levels of capital and (other things being equal) out- put per worker were achieved.50

Given one or both of these conditions, technology transfer and more rapid capital accumulation, incomes would tend toward convergence in the long run as the faster-growing developing countries would be catching up with the slower-growing developed countries. Even if incomes did not eventually turn out to be identical, they would at least tend to converge conditional on (i.e., after also taking account of any systematic differences in) key variables such as population growth rates and savings rates (this argument is formalized in the neoclassical growth model examined in Chapter 3 and Appendix 3.2). Given the huge differences in capital and technology across countries, if growth con- ditions were similar, we should see tendencies for convergence in the data.

Whether there is now convergence in the world economy depends on two levels of how the question is framed: whether across average country incomes or across individuals (considering the world as if it were one country); and whether focusing on relative gaps or absolute gaps.

Relative Country Convergence The most widely used approach is simply to examine whether poorer countries are growing faster than richer countries. As long as this is happening, poor countries would be on a path to eventually “catch up” to the income levels of rich countries. In the meantime, the relative gap in incomes would be shrinking, as the income of richer countries would become a smaller multiple of income of poorer countries (or looked at from the other perspective, incomes of poor countries would become an increas- ingly large fraction of income of rich countries). This can be seen on a country- by-country basis. Although China’s average income was just 3% that of the United States in 1980, it was estimated to have reached 14% of U.S. income by 2007. But in the same period, the income of the DRC fell from about 5% of U.S. levels to just 1%. But globally, evidence for relative convergence is weak, even for the most recent decades.

Figure 2.7a illustrates the typical findings of this literature. On the x-axis, income data are plotted from the initial year, in this case 1980; while on the y-axis, the average growth rate of real per capita income is plotted, in this case, over the subsequent 27 years to 2007. If there were unconditional convergence, there would be a tendency for the points plotted to show a clear negative rela- tionship, with the initially lower-income countries growing faster. But as seen in Figure 2.7a there is no apparent tendency toward convergence across coun- tries. In fact, even in this recent period, about 60% of countries grew more slowly than the United States. Looking just at the developing countries, as in Figure 2.7b, it is clear that divergence is occurring: middle-income countries are

82 PART ONE Principles and Concepts

growing faster than low-income countries, so there is a growing gap among developing countries. Many nations, especially among the 49 least developed countries, remain in relative stagnation. Poor developing countries have not been catching up as a group.51

In Figure 2.7c, growth of high-income OECD countries is examined sepa- rately for 1950–2007. The picture here is one of convergence, and we need to interpret it carefully. One explanation is that all of these countries have simi- lar features, including a relatively early start at modern economic growth. This makes the countries more able to borrow technology from each other, as well as trade with and invest in each other ’s economies. We might conjec- ture that if developing countries closely followed the institutions and policies of these OECD economies, they might converge as well. However, as noted earlier, there are many institutional and other differences between low- and

FIGURE 2.7 Relative Country Convergence: World, Developing Countries, and OECD

Source: Data from Center for International Comparisons, University of Pennsylvania, accessed at http://pwt.econ.upenn. edu/php_site/pwt63/pwt63_form.php.

15.00

12.00

9.00

3.00 U.S.

= 1.97

Number below U.S. = 93 Number above U.S. = 63

5,000 15,000 25,0000 35,000

6.00

0.00

–3.00

–6.00

(a) Per capita growth 1980–2007 for 157 countries (b) Per capita growth 1980–2007 for 86 developing countries

Real income per capita, 1980 Real income per capita, 1980

Real income per capita, 1950

A ve

ra g

e r

e a

l p e

r ca

p it

a g

ro w

th (

% )

4.50

4.00

3.50

3.00

2.50

1.50

U.S. = 2.11

2,000 6,000 10,000 14,0000

1.00

0.50

0.00

(c) Per capita growth 1950–2007 for 22 OECD countries

A ve

ra g

e r

e a

l p e

r ca

p it

a g

ro w

th (

% )

8.00

6.00

4.00

0.00

U.S. = 1.97

2,000 6,000 10,000 14,0000 18,000

–2.00

–4.00

–6.00

A ve

ra g

e r

e a

l p e

r ca

p it

a g

ro w

th (

% ) Number below U.S. = 55

Number above U.S. = 31

KHM

BGD

LAO

BTN

ETH

MLI NPL

SOM

BDI

TCD

MOZ BFA

VNM

IND

MWI

RWA

LKA

GIN

SLE

ZAR

NER

BEN

PAK

TZAAFG

CAF

MDG

HTI

MRT

LSOUGA

AGO

SDN

TGO

KEN SEN

IDN EGY

GHA

CMR

LBR

HNDZMB

ZWE

THA

BOL

PHLCOGNGA

PNG SLV

MAR

PER

CIV NIC

COL

PRY

ECU

DOM

GTM

SYR

TUN

JOR

MYS

JAM

LBN

KOR

TUR

DZA MEX

PAN

TWN

CHL

ZAF CRI

BRA

URY ARG

PRT

TTO

VEN

HKG

GRC

SGP

ESP

ISR

AUT

BEL

CAN DNK

FRA

ISL

IRL

ITA

LUX

NLD

NOR

PRT

SWE

CHE

TUR

GBR

USAAUS

FIN

JPN

NZL

ESP

83CHAPTER 2 Comparative Economic Development

high-income economies today, some of which may be very difficult to change; we explore these further in the next section. Moreover, a poor country cannot force a rich country to lower its trade barriers. In any case, one must draw conclusions from the results with great caution because of selection bias. That is, among today’s rich countries, some were relatively rich in the past and some were relatively poor; in order for them all to be rich countries today, the poor countries had to have grown faster than the rich ones, simply as a mat- ter of logic. Confining attention just to the rich countries thus commits the statistical error of selection bias.52 Nevertheless, the strong evidence for con- vergence among the OECD countries, together with the failure at least until very recently to find compelling evidence for longer-term convergence for the world as a whole, particularly divergence for the least developed countries, is likely one reflection of the difference in growth conditions between now developed and developing countries.

Absolute Country Convergence With the recent rapid growth in China, and the acceleration of growth in South Asia as well, these regions are currently on a path of relative country convergence. For example, in the 1990–2003 period, while income grew 24% in high-income OECD countries, it grew 56% in South Asia and 196% in China. But due to their relatively low starting income levels, despite higher growth, income gains were still smaller in absolute amount than in the OECD, as illustrated in Figure 2.8. That is, even when the average income of a developing country is becoming a larger fraction of developed

Source: From Human Development Report, 2005, p. 37. Reprinted with permission from the United Nations Development Programme.

25,000

20,000

1990

Income growth, 1990–2003

2003

15,000

10,000

56%

24%

196% 5,000

0

P e

r ca

p it

a G

D P,

2 0

0 3

( P

P P

)

South Asia

China High-income OECD countries

FIGURE 2.8 Growth Convergence versus Absolute Income Convergence

84 PART ONE Principles and Concepts

country average incomes, the difference in incomes can still continue to widen for some time before they finally begin to shrink. A process of absolute coun- try convergence is a stronger standard than (and appears only with a lag after) a process of relative country convergence.53

Population-Weighted Relative Country Convergence The high growth rate in China and India is particularly important, because more than one- third of the world’s people live in these two countries. This approach frames the question so as to weight the importance of a country’s per capita income growth rate proportionately to the size of its population. A typical study of this type is depicted in Figure 2.9a–d. Instead of points representing the data for each country, bubble sizes are used to depict the relative size of countries’

FIGURE 2.9 Country Size, Initial Income Level, and Economic Growth

0.1

6 7 8 9 105 11 – 0.1

0.0

P e

r C

a p

it a

G ro

w th

R a

te , 1

9 5

0 –1

9 6

3 0.1

6 7 8 9 105 11 – 0.1

0.0

P e

r C

a p

it a

G ro

w th

R a

te , 1

9 7

6 –1

9 8

9

0.1

6 7 8 9 105 11 – 0.1

0.0

P e

r C

a p

it a

G ro

w th

R a

te , 1

9 6

3 –1

9 7

6 0.1

6 7 8 9 105 11 – 0.1

0.0

P e

r C

a p

it a

G ro

w th

R a

te , 1

9 8

9 –2

0 0

3

(c) Country size, initial income, and economic growth, 1976–1989, bubble size proportional to population in 1976

ln (income per capita, 1976)

(a) Country size, initial income, and economic growth, 1950-1963, bubble size proportional to population in 1950

ln (income per capita, 1950)

(b) Country size, initial income, and economic growth, 1963–1976, bubble size proportional to population in 1963

ln (income per capita, 1963)

(d) Country size, initial income, and economic growth, 1989–2003, bubble size proportional to population in 1989

ln (income per capita, 1989)

China

India

Japan

USA

Fed. USSR

China

India

Japan

USA

Fed. USSR

China

India

Japan

USA

Fed. USSR

China

India

Japan

USA

Fed. USSR

Source: Steven Brakmana and Charles van Marrewijk, “It’s a big world after all: On the economic impact of location and distance,” Cambridge Journal of Regions, Economy and Society 1 (2008): 411–437. Reprinted by permission of Oxford University Press.

85CHAPTER 2 Comparative Economic Development

populations. To get a sense of how the acceleration of growth in China and India, along with a few other countries, have changed the picture, the data are broken up into four time periods. Figures 2.9a and 2.9b reflect that there was relative per capita divergence from 1950 through 1976, but Figure 2.9d reflects relative per capita convergence since 1989 (and less unambiguously but plau- sibly from 1977 to 1989 as well—see Figure 2.9c). If current trends continue (a “big if” given widespread predictions for a slowing of their growth rates), then China, India, and Brazil will account for nearly 40% of global output by 2050, compared with about 10% in 1950.54 Although it is true that conditions have remained stagnant or even deteriorated in many of the least developed coun- tries, because of their smaller population sizes with the population-weighted approach, this divergence effect is more than compensated for by growth in countries with large populations. Note that all such trends may be subject to change. For example, the population growth rates of the 49 least developed countries and other low-income countries are much higher than those of the middle- and upper-income countries; so their population-weights are increasing over time. African countries may see a furtherance of their recent trend to faster growth magnifying the new trend to global convergence; or they and other developing regions may see a growth slowdown, with commodity prices fall- ing again and continuing governance problems; and the global economy could return to a period of divergence. These trends will be watched closely.

World-as-One-Country Convergence An alternative approach to the study of convergence is to think of the world as if it were one country. In the first such study, Branko Milanovic stitched together household data sets from around the world and concluded that global inequality rose significantly in the period 1988 to 1993.55 Studies of this kind are difficult to carry out. The most important difference from population-weighted country convergence is that a world-as-one-country convergence study can take into account changes in inequality within countries as well as between them. In particular, the widening gulf between incomes in rural and urban China had a major effect on the finding of global divergence using this method. But most researchers and policymakers frame development as a process that occurs on the national level, something rather different from global inequality; and country convergence studies remain the standard.

Sectoral Convergence Despite evidence that economies are not converging unconditionally, there can be cross-national convergence of economic sectors, which in turn may signal the potential for future convergence. In particular, Dani Rodrik found evidence that there has been convergence in manufactur- ing, with implications that the failure to find overall convergence across coun- tries is due to the small share and slow growth of manufacturing employment in low-income countries.56

2.7 Long-Run Causes of Comparative Development

What explains the extreme variations in development achievement to date among developing and developed countries? The next two chapters examine theories of economic growth and development processes and policy challenges;

86 PART ONE Principles and Concepts

here we present a schematic framework for appreciating the major long-run causes of comparative development57 that have been argued in some of the most influential research literature of this century.58 (Bear in mind that research on this important subject is still at a relatively early stage; scholars have legiti- mate disagreements about emphasis and substance, and new findings are being reported regularly.)

First, in the very long run, few economists doubt that physical geography, including climate, has had an important impact on economic history. Geography was once truly exogenous, even if human activity can now alter it, for better or worse. But the economic role played by geography, such as tropical climate, today is less clear. Some research suggests that when other factors, notably inequality and institutions, are taken into account, physical geography adds little to our understanding of current development levels. However, some evidence is mixed. For example, there is some evidence of an independent impact of malaria and indications that in some circumstances, landlocked status may be an impedi- ment to economic growth; indeed, a direct link is argued by some economists,59

so this possible effect is reflected in Arrow 1 connecting geography to income and human development on the left side of Figure 2.10. Recently, the debate on compar- ative economic development has been widened further with some evidence that an intermediate degree of genetic diversity (heterozygosity) of human popula- tions is most conducive to long-run economic development.60

Economic institutions, which play an important role in comparative development, are defined by Nobel laureate Douglass North as the “rules of the game” of economic life. As such, institutions provide the underpinning of a market economy by establishing the rules of property rights and contract enforcement; improving coordination;61 restricting coercive, fraudulent, and anticompetitive behavior; providing access to opportunities for a broad popu- lation; constraining the power of elites; and managing conflict more generally. Moreover, institutions include social insurance (which also serves to legiti- mize market competition) and the provision of predictable macroeconomic stability.62 Countries with higher incomes can afford better institutions, so it is challenging to identify the impact of institutions on income. But recently, development economists have made influential contributions toward achiev- ing this research goal.

As noted earlier, most developing countries were once colonies. Geogra- phy affected the types of colonies established (Arrow 2), with one of the now best known geographic features being settler mortality rates, whose impact63

was examined in work by Daron Acemoglu, Simon Johnson, and James Robin- son. In this argument, when potential settlers faced higher mortality rates (or perhaps other high costs), they more often ruled at arm’s length and avoided large, long-term settlement. Their interest could be summarized as “steal fast and get out” or “get locals to steal for you.” Unfavorable institutions were therefore established, favoring extraction over production incentives. But where mortality was low, populations were not dense, and exploitation of resources required substantial efforts by colonists, institutions broadly encour- aging investments, notably constraints on executives and protection from expropriation, were established (sometimes as a result of agitation from set- tlers who had the bargaining power to demand better treatment). These effects are reflected by Arrow 3. Acemoglu and colleagues present evidence that after

Economic Institutions “Humanly devised” con- straints that shape interac- tions (or “rules of the game”) in an economy, including formal rules embodied in constitutions, laws, contracts, and market regulations, plus informal rules reflected in norms of behavior and con- duct, values, customs, and generally accepted ways of doing things.

87CHAPTER 2 Comparative Economic Development

accounting for institutional differences, geographic variables (e.g., closeness to the equator) have little influence on incomes today.64 Their statistical esti- mates imply large effects of institutions on per capita income.

The influence of geography on precolonial institutions is captured by Arrow 4. Precolonial institutions also mattered to the extent that they had influence on the type of colonial regime established. This possible effect is reflected by Arrow 5.

Precolonial comparative advantage and evolving labor abundances in the Americas and their relation to the institutions established have been examined in the pioneering work of Stanley Engerman and Kenneth Sokoloff.65 When

FIGURE 2.10 Schematic Representation of Leading Theories of Comparative Development

Physical geography (including climate)

Precolonial institutions

Type of colonial regime

Inequality

Well-functioning markets

Income and human development

Human capital

Public goods quality

Effective civil society

Postcolonial institutional

quality

Precolonial labor abundances, production structure, and potential comparative advantage

Evolution and timing of European

development

Local features affecting types of colonies

established (e.g., potential settler costs such as mortality)

21 2220

11

1914

13

1

2

3

4 6

8

9

5 7

12

1615 17 18

10

88 PART ONE Principles and Concepts

climate was suitable for a production structure featuring plantation agriculture (particularly sugarcane in the early history), slavery and other types of mass exploitation of indigenous labor were introduced. In other areas, when indig- enous peoples survived contact in sufficient numbers and mineral wealth was available, vast land grants that included claims to labor were established (by Spain). Although resulting from different comparative advantage (sugarcane and minerals), economic and political inequality were high and remained high in all of these economies (even among freemen), which had long-lasting nega- tive effects on development. These links are reflected by Arrow 6 and Arrow 7. Early inequities were perpetuated with limits on the nonelite population’s access to land, education, finance, property protection, and voting rights, as well as labor markets. This inhibited opportunities to take advantage of industrialization when they emerged in the nineteenth century, a period when broad participation in commercial activity had high social returns.

The contrast with North American potential production structure is strik- ing. Its comparative (emerging) advantage in grain lacked at the time the scale economies of tropical agriculture and of mineral extraction seen elsewhere in the Americas. Scarce labor with abundant land inhibited the concentration of power (despite efforts of colonizers to do so). The need to attract more settlers and encourage them to engage the colonial economy led to the evolution of more egalitarian institutions in the North American colonies. North Ameri- cans enjoyed greater egalitarianism in access to all of the factors so restricted elsewhere. This environment facilitated broad-based innovation, entrepre- neurship, and investment and gave the United States and Canada a decisive advantage despite their starting out as much poorer societies, which they used to economically surpass societies whose populations were mostly illiterate, disenfranchised, and lacking collateral.66 (We will examine further aspects of Engerman and Sokoloff’s analysis shortly.)

When local populations were larger and denser and social organization was more advanced, it was easier for colonists to take over existing social structures to gain tribute. In such cases, resulting institutional arrangements would tend to favor mechanisms of extraction of existing wealth over the creation of new wealth, often leading to declines in the relative fortunes of these regions. This is pointed up by Acemoglu, Johnson, and Robinson, whose influential research on this historical “reversal of fortune"67 is also reflected by Arrow 5. These authors stress that if geography were fundamental to devel- opment prospects, the most prosperous areas prior to colonization should continue to be relatively prosperous today. But the most prosperous formerly colonized areas today tend to have been least prosperous in the past. Past population density and past urbanization, which are positively correlated with past income, are negatively correlated with current income, these authors show.68 There is evidence that Europeans set up more extractive institutions (ones designed to extract more surplus from colonized populations) in pros- perous areas and that these institutions have often persisted to the contempo- rary period.69

Geography undoubtedly influenced early economic history in Europe.70

This is reflected by Arrow 8, leading to evolution and timing of European devel- opment. Early development in Europe gave it advantages over most other regions—advantages that were used to colonize much of the world. But the

89CHAPTER 2 Comparative Economic Development

types of colonial regimes implemented varied considerably, depending on conditions prevailing at the time of colonization both in the different parts of the world colonized and within the colonizer ’s home country. The timing of European development influenced the type of colonial regime established, reflected by Arrow 9. For example, it has been argued that for various rea- sons, earlier colonization generally involved more plunder and less active production than later colonization, although both occurred at the expense of the indigenous populations.71

Precolonial comparative advantage may also have interacted with the timing of European development in influencing institutions in that settlers in later- colonized temperate zones arrived with more knowledge and more advanced technology. In particular, Europeans brought better agricultural techniques to the later-settled areas such as North America. As noted by David Fielding and Sebastian Torres, by the eighteenth century, population growth in Europe and technical change had produced a large supply of people with temperate-zone agricultural skills in products such as wheat and dairy. They were able to gain higher incomes using these skills in temperate colonies and former colonies (the so-called neo-Europes).72 Thus, precolonial (potential) comparative advan- tage again mattered. This link is reflected in the flow through Arrow 6 and Arrow 7. The possible role played by specific skills also points up the impor- tance of human capital investments for development, reflected by Arrow 14.

Thus, the types of colonial regimes established, while always designed for the benefit of the colonizers, were influenced by local and European supply and demand factors. The type of regime had enormous influence on postco- lonial institutional quality, reflected by Arrow 10. For example, the depraved rule of Belgium’s King Leopold II over the Congo (today’s Democratic Repub- lic of Congo) was arguably an ultimate cause of the oppressive Mobutu reign after independence. Of course, not all influences of colonialism were neces- sarily bad. Along with enslavement, subjugation, exploitation, loss of cultural heritage, and repression, colonists also brought modern scientific methods in fields such as medicine and agriculture. Note that this can be no apologia for colonialism, because these advances could have been gained without the soci- eties’ becoming colonized, as in Japan. Still, there is some evidence that coun- tries and territories that spent a longer time as colonies (at least in the case of islands) have higher incomes than those that experienced shorter colonial periods, with this effect greater for entities colonized later (perhaps because earlier colonial activity had more pernicious effects than later ones). Even so, there are strong caveats to this finding.73

Besides creating specific institutions, European colonization created or reinforced differing degrees of inequality (often correlated with ethnicity), ulti- mately leading to diminished prospects for growth and development, nota- bly in Latin America and the Caribbean. This is reflected by Arrow 11. High inequality often emerged as a result of slavery in regions where crops could be “efficiently” produced on slave plantations. They also emerged where a large, settled indigenous population could be coerced into labor. Such histo- ries had long-term consequences, particularly in Latin America. As Engerman and Sokoloff have argued, the degree of inequality itself can shape the evolu- tion of institutions as well as specific policies. Where inequality was extreme, there was less investment in human capital (Arrow 13) and other public goods

90 PART ONE Principles and Concepts

(Arrow 16) and, as reflected by the bidirectional Arrow 12, a tendency of less movement toward democratic institutions (which could also have facilitated movement to other constructive institutions).74

Thus, extreme inequality is likely to be a long-term factor in explaining comparative development. This is raised in the striking historical contrast between the states of North America and the states of Central and South America. There was greater egalitarianism in North America, though the inhuman treatment of Native Americans and of slaves in the southern colonies (later the United States) reflects the fact that this is not because the English settlers were inherently “nicer masters” than the Spanish. Still, much of the North American experience contrasts strongly with the extreme inequality of Central and South America and the Caribbean.75 Engerman and Sokoloff argued that high inequality in Latin America led to low human capital investments, again in contrast to North America;76 this mechanism is again reflected by Arrow 13. Elites in Latin America then loosened their control only when their returns to increased immigration, and thus to creating more attractive conditions for immigrants, were high. Besides creating specific institutions, then, European colonization created or reinforced different degrees of inequality, often cor- related with ethnicity. This history had long-term consequences, particularly in Latin America. In the direction from inequality to postcolonial institutional quality, Arrow 12 reflects what has been termed the social conflict theory of institutions. Box 2.4 presents findings that inequality does negatively affect per capita income much in the way predicted by Engerman and Sokoloff.

Cultural factors may also matter in influencing the degree of emphasis on education, postcolonial institutional quality, and the effectiveness of civil soci- ety, though the precise roles of culture are not clearly established in relation to the economic factors surveyed in this section and so are not included in the diagram. In addition, institutional quality affects the amount and quality of investments in education and health, via the mediating impact of inequality. In countries with higher levels of education, institutions tend to be more dem- ocratic, with more constraints on elites. The causality between education and institutions could run in either direction, or both could be caused jointly by still other factors. Some scholars argue that some countries with bad institu- tions run by dictators have implemented good policies, including educational investments, and subsequently, after reaping the benefits in terms of growth, those countries have changed their institutions.77 They argue that human cap- ital is at least as fundamental a source of long-run development as institu- tions. In the diagram, this would suggest adding an arrow from human capital back to postcolonial institutional quality; this is intuitively plausible, although additional evidence for this link will be needed for it to become more fully established.78 Clearly, however, in some cases extractive colonial institutions left a legacy that resulted in poor health and education decades after indepen- dence; an example from India is examined in Box 2.5.

For the relatively small number of developing countries never colonized, such as Thailand, type of colonial regime can be reinterpreted in the diagram as institutional quality at an early stage of development (or as cultural influ- ences not shown)—but note that the evidence for causality patterns is not as convincing in these cases. However, the diversity of development experiences of never-colonized countries caution us not to place complete emphasis on the choices of colonizers; preexisting social capital may matter at least as much.79

91CHAPTER 2 Comparative Economic Development

BOX 2.4 FINDINGS Instruments to Test Theories of Comparative Development: Inequality

In a 2007 study, William Easterly used cross-country data to examine the Engerman and Sokoloff hypothesis. His research confirmed that “agricultural endowments predict inequality and inequality pre- dicts development.” Specifically, Easterly finds that inequality negatively affects per capita income; it also negatively affects institutional quality and schooling, which are “mechanisms by which higher inequality lowers per capita income.” That the negative relation- ship between income and inequality is present in the data is clear—but how do development economists take the step to prediction and assignment of causality when measurement error and many confounding fac- tors are present, such as the possible link that under- development itself is a cause of inequality?

Sometimes development economists run field experiments—but obviously, we cannot randomly assign countries various levels of inequality to see what happens! In the many cases when field experiments are impossible, development economists frequently try to understand causality by searching for an instrumen- tal variable (or “instrument”); in fact, many researchers in development economics invest a lot of their time in this search. This is a topic covered in classes in econo- metrics. But the basic idea is that to identify the effect of a potential causal variable c (such as inequality) on a development outcome variable d (such as income or educational attainment), the hunt is on for an elusive instrumental variable e that affects d only through e’s effect on c. So an instrument has no independent

effect on the outcome variable of interest. You saw ear- lier that Acemoglu, Johnson, and Robinson used settler mortality as an instrument for early institutions. East- erly uses “the abundance of land suitable for growing wheat relative to that suitable for growing sugarcane” as an instrument for inequality. Using this strategy, Easterly concludes that high inequality of the Enger- man and Sokoloff variety is independently “a large and statistically significant barrier to prosperity, good quality institutions, and high schooling.” Schooling and institutional quality are precisely the mechanisms proposed by Engerman and Sokoloff by which higher inequality leads to lower incomes. Like a leprechaun, a good instrumental variable is hard to get hold of but when caught can give the researcher’s equivalent of a pot of gold. Though active debate on inequality and development continues, the interplay between the careful institutional analysis and economic history scholarship of Engerman and Sokoloff and the study of causality with larger data sets as used by Easterly gives a window into how the field of development econom- ics continues to make progress.

Sources: William Easterly, “Inequality does cause underde- velopment,” Journal of Development Economics 84 (2007): 755–776; Joshua D. Angrist and Jorn-Steffen Pischke, Mostly Harmless Econometrics: An Empiricist’s Companion (Princeton, N.J.: Princeton University Press, 2008). For an important critique of the use and interpretation of instru- mental variables (and also of randomization) in devel- opment economics research see Angus Deaton, “Instru- ments, randomization, and learning about development,” Journal of Economic Literature, 48, no. 2 (2010): 424–455.

Never-colonized countries also show a dramatic range in performance; Ethio- pia and Afghanistan remain very poor, Thailand is in the lower-middle range, Turkey is in the upper-middle range, and Japan is among the very wealthiest countries; China, starting among the poorest countries 30 years ago, is now rapidly ascending the income tables. The quality of institutions (and inequal- ity) undoubtedly mattered in noncolonized societies; it is just harder to con- clude that institutions led to income rather than only vice versa.

Clearly, human capital has a direct impact on income and on human devel- opment more broadly, as reflected by Arrow 14. The depth and breadth of education in the population will help determine the effectiveness of govern- ment as a force for development, reflected by Arrow 15. This is due not only

92 PART ONE Principles and Concepts

to a better-qualified civil service but also to the understanding of citizens of poor government performance and the knowledge of how to work for a bet- ter outcome and capacity to organize.80 Of course, education could also inde- pendently affect the organization and functioning of markets per se (arrow omitted), but the literature to date has primarily viewed the productive impact of human capital on market outcomes as a direct one, reflected by Arrow 14. These impacts are explored further in Chapter 8.

The type and quality of global integration (particularly trade) have been stressed as a boon to long-run growth and development in many World Bank reports. Trade may be beneficial in that it provides various kinds of access to technology.81 And some economists argue that greater openness to trade ben- eficially affects the subsequent evolution of institutions. On the other hand, critics argue that the wrong kind of integration or the failure to complement integration with appropriate policies could be harmful to development. In fact, evidence suggests that once institutions are accounted for, trade itself explains very little, so for simplicity, integration is left out of the diagram.82

Postcolonial institutional quality has a strong impact on the effectiveness of the private, public, and citizen (or civil society) sectors. Democratic gov- ernance, rule of law, and constraints on elites will encourage more and better

BOX 2.5 FINDINGS Legacy of Colonial Land Tenure and Governance Systems

Substantial evidence on the importance of institu-tions is provided in a study of the impact of land revenue institutions established by the British Raj in India conducted by Abhijit Banerjee and Lakshmi Iyer. Because areas where land revenue collection was taken over by the British between 1820 and 1856 (but not before or after) were much more likely to have a non-landlord system, the authors used being con- quered in this period as an instrument for having a non-landlord system. They also used other statistical tests that showed the results were robust. They found that historical differences in property rights institu- tions led to sustained differences in economic out- comes, in that the regions in which property rights to land were given to landlords have had significantly lower agricultural investments and productivity in the postindependence period than regions in which property rights were given to cultivators. The authors concluded that the divergence occurred because his- torical differences in institutions led to different pol- icy choices. Tellingly, the regions in which landlords

received the proprietary rights also had significantly lower investments in health and education in the postcolonial period.

In subsequent research, Lakshmi Iyer compared economic outcomes across areas in India that expe- rienced direct versus indirect British colonial rule, controlling for the apparent colonial preference to annex higher-quality lands using another instrumen- tal variable strategy. She found evidence that colonial governance quality had persistent effects on postcolo- nial outcomes; areas under direct rule received signifi- cantly less access to schools, health centers, and roads in the postcolonial period, with higher levels of pov- erty and infant mortality.

Sources: Abhijit Banerjee and Lakshmi Iyer, “History, institu- tions, and economic performance: The legacy of colonial land tenure systems in India,” American Economic Review 95 (2005): 1190–1213; and Lakshmi Iyer, “Direct versus indi- rect colonial rule in India: Long-term consequences,” Review of Economics and Statistics 92 (2010): 693–713. Preparation of this box also benefited from a manuscript, Lakshmi Iyer, “The long-run consequences of colonial institutions,” draft, Harvard Business School, 2013.

93CHAPTER 2 Comparative Economic Development

quality public goods, reflected by Arrow 17. Better property rights protections and contract enforcement for ordinary citizens and broad access to economic opportunities will spur private investments, reflected by Arrow 18. And insti- tutions will affect the ability of civil society to organize and act effectively as a force independent of state and market, reflected by Arrow 19. Clearly, the activities of the three sectors will each have an influence on productivity and incomes, and on human development more generally, as reflected by Arrows 20, 21, and 22, respectively.83 These factors are explored further in Chapter 11.

It is not yet entirely clear which economic institutions are most important in facilitating development or the degree to which strength in one institution can compensate for weakness in another.84 Clearly, there are multiple paths to economic development (see, e.g., the case study of China at the end of Chap- ter 4). But a key finding of recent research is that forces that protect narrow elites in ways that limit access of the broader population to opportunities for advancement are major obstacles to successful economic development. If institutions are highly resistant to attempts at reform, this helps clarify why development is so challenging.

Nevertheless, in most countries with poor institutions, there is still much that can be done to improve human welfare and to encourage the development of better institutions. Indeed, economic institutions do change over time, even though political institutions such as voting rules sometimes change without altering the real distribution of power or without leading to genuine reform of economic institutions. Although the evidence of the impact of democracy on growth in the short to medium term is not strong (see Chapter 11), in the long run demo- cratic governance and genuine development do go hand in hand, and the steady spread of more genuinely democratic institutions in the developing world is a very encouraging sign.85 As Dani Rodrik has noted, “Participatory and decen- tralized political systems are the most effective ones we have for processing and aggregating local knowledge. We can think of democracy as a meta-institution for building other good institutions.”86 In addition, development strategies that lead to greater human capital, improve access to new technologies, produce better-quality public goods, improve market functioning, address deep-rooted problems of poverty, improve access to finance, prevent environmental degrada- tion, and foster a vibrant civil society all promote development.

2.8 Concluding Observations

History matters. We have learned that conditions prevailing in a developing nation when European colonialism began had a large impact on the subse- quent history of inequality and institutional development in the nation in ways that either facilitated or thwarted participation in modern economic growth after the Industrial Revolution arrived in the late eighteenth century. And poor institutions have generally proved very resistant to efforts at reform. But the new perspectives do not imply that development is impossible! Instead, they serve to clarify the nature of the great challenges facing many developing nations. The phenomenon of underdevelopment is best viewed in both a national and an international context. Problems of poverty, inequality, low productivity, population growth, unemployment, primary-product export

94 PART ONE Principles and Concepts

dependence, and international vulnerability have both domestic and global origins and potential solutions.

It should be remembered that most developing nations have succeeded in raising incomes significantly. And most developing countries have had nota- ble successes in lowering infant mortality, improving educational access, and narrowing gender disparities.87 By pursuing appropriate economic and social policies both at home and abroad and with effective assistance from devel- oped nations, even the least developed countries do indeed have the means to realize their development aspirations. Parts Two and Three will discuss the ways in which these hopes and objectives can be attained.

But concomitant and complementary human capital, technological, social, and institutional changes must take place if long-term economic growth is to be realized. Such transformations must occur not only within individual developing countries but also in the international economy. In other words, unless there is some major structural, attitudinal, and institutional reform in the world economy, one that accommodates the rising aspirations and rewards the outstanding performances of individual developing nations, particularly the least developed countries, internal economic and social transformation within the developing world may be insufficient.88

There may be some “advantages of backwardness” in development, such as the ability to use existing, proven technologies rather than having to rein- vent the wheel and even leapfrogging over older technology standards that developed countries have become locked into. One can also learn valuable les- sons from economic policies that have been tried in various countries around the world. These advantages are especially helpful if an economy can success- fully manage to get sustained modern economic growth under way, as, for example, in Taiwan, South Korea, China, and a few other cases. However, for most very poor countries, backwardness comes with severe disadvantages, many of which have been compounded by legacies of colonialism, slavery, and Cold War dictatorships. In either case, countries will generally have to do more than simply emulate policies followed by today’s developed countries while they were in their early stages of development.

Despite the obvious diversity of these countries, and growing gaps between middle- and low-income countries, most developing nations share a set of common and well-defined goals. These include a reduction in pov- erty, inequality, and unemployment; the provision of basic education, health, housing, and food to every citizen; the broadening of economic and social opportunities; and the forging of a cohesive nation-state. Related to these eco- nomic, social, and political goals are the common problems shared in varying degrees by most developing countries: chronic absolute poverty, high levels of unemployment and underemployment, wide disparities in the distribution of income, low levels of agricultural productivity, sizable imbalances between urban and rural levels of living and economic opportunities, discontent on the part of the segments of the population not benefiting from economic growth, serious and worsening environmental decay, antiquated and inappropri- ate educational and health systems, and substantial dependence on foreign technologies, institutions, and value systems. It is therefore also possible and useful to talk about the similarities of critical development problems and to analyze these problems in a broad developing world perspective.

95CHAPTER 2 Comparative Economic Development

Economic and social development will often be impossible without cor- responding changes in the social, political, legal, and economic institutions of a nation, such as land tenure systems, forms of governance, educational struc- tures, labor market relationships, property rights, contract law, civic freedoms, the distribution and control of physical and financial assets, laws of taxation and inheritance, and provision of credit. But fundamentally, every develop- ing country confronts its own constraints on feasible policy options and other special circumstances, and each will have to find its own path to effective eco- nomic and social institutions. Examples offered by developed countries’ ear- lier experiences and current institutions, as well as those of other countries in the developing world, provide important insights for policy formulation. Economic institutions of Europe and North America are in most cases closer to efficient than those of many developing countries, although all countries have room for further institutional innovations. But developing countries can- not assume without additional investigation that patterning their policies and institutions on those of developed countries will always provide the fastest route to successful economic development; transitional institutions are likely to be the most effective route to rapid economic growth for at least some developing countries (see the case study of China at the end of Chapter 4).

In sum, this chapter has pointed up some important similarities across most developing countries, in contrast to contemporary and historical char- acteristics of developed countries. It has also shown that developing nations are very heterogeneous, differing in many critical respects. Looming large in explaining the root causes in the levels of incomes and human development are the higher inequality, weaker institutions, and lower levels of education and health. But even starting with these weaknesses, there is much that devel- oping countries can undertake through appropriate policy strategies and at least incremental improvements in institutions to speed economic and social progress.

Indeed, the experience of the past 50 years shows that while development is not inevitable and poverty traps are quite real, it is possible to escape from poverty and initiate sustainable development. Before examining specific poli- cies for doing so, in the next chapters we will set the context further by exam- ining important theories and models of development and underdevelopment. In Chapter 3, we examine classic theories that remain influential and useful in many respects, and in Chapter 4, we consider models of coordination failures and other constraints and conceptual strategies for escaping from them.

96

96

In 1971, Bangladesh declared independence from Pakistan. Previously, Bangladesh had been known as East Pakistan, and what is now Pakistan was called West Pakistan. Though more than 1,000 miles apart, both were part of a single country, with eco- nomic and political power concentrated in West Pakistan. Because they were once the same country, Pakistan and Bangladesh make for an interesting exercise in comparative development, in that the two shared a common national policy in the early years, even if they did not benefit from it equally. Pakistan and Bangladesh had a similar popula- tion in 2012: an estimated 180 million in Pakistan and 153 million in Bangladesh (Population Refer- ence Bureau). They are located in the South Asian region, are both overwhelmingly Islamic, and were both once part of the colonial British Raj of India. Bangladesh was for a long time the global symbol of suffering, from the Bengal famine of 1943 to the 1971 Concert for Bangladesh featuring George Harri- son, Eric Clapton, and Bob Dylan to the horrors of the 1974 postindependence famine.

But analysts such as William Easterly have declared Pakistan a leading example of “growth without devel- opment,” with low social indicators for its income and growth. Meanwhile, Bangladesh, though still very poor and afflicted with many of the social problems found in Pakistan, has been transforming itself from a symbol of famine to a symbol of hope.

When Bangladesh gained its independence, it was viewed as lagging insurmountably behind Pakistan. Indeed, its poor social and economic development in comparison with West Pakistan was a major impetus behind the independence move- ment, which complained that Bangladesh was being drained of tax revenues to benefit West Pakistan.

The war for independence itself and the economic destruction deliberately visited on Bangladesh’s industry left an even wider gap, while abuses left serious psychological scars, and a terrible famine fol- lowed. One U.S. statesman undiplomatically dubbed Bangladesh the “international basket case.” Others somewhat more tactfully called it the “test case for development"—meaning that if development could happen in Bangladesh, it could happen anywhere. Four decades later, Bangladesh is confounding the skeptics; it actually looks like it may pass this test. Although Pakistan still has 44% higher income than Bangladesh according to UNDP estimates, the two countries nonetheless received an identical New HDI ranking for 2013, with Bangladesh 9 places higher on NHDI than predicted for its income level, while Pakistan is 9 places below what would be pre- dicted by income alone.

Not that Bangladesh has dramatically outper- formed Pakistan. Bangladesh continues to have serious development problems. It is rather that Bangladesh has made relatively better progress than Pakistan, particularly on social development indi- cators, despite its handicaps at independence and expectations that it would continue to fare badly. Bangladesh started at a much lower level of social development and still has lower income. But in achieving more progress on social development, Bangladesh now also has the conditions for accel- erating economic progress in the coming years, par- ticularly if continuing problems of governance can be overcome.

Growth PPP-adjusted income estimates vary, but all show average income remains higher in Pakistan than in

Case Study 2

Comparative Economic Development: Pakistan and Bangladesh

97

Bangladesh ($2,880 in Pakistan in 2011 and $1,910 in Bangladesh according to World Bank estimates). In Pakistan, per capita income grew at about 2.2% per year in the half-century from 1950 to 2000. As a result, per capita income tripled. But the growth rate declined decade by decade, even as it rose in other countries, including India. The decline in the growth rate may be a result of the poor per- formance on social indicators. From 2000 to 2011, GDP growth in Pakistan averaged 4.9% (World Bank); with population growth of 1.8%, per capita GDP growth was about 3.1%. It remains to be seen whether Pakistan’s moderately increased growth rate will be sustainable. Indications are that Paki- stan has experienced much less inclusive (pro-poor) growth in comparison with Bangladesh.

In Bangladesh, GDP growth averaged 6% from 2000 to 2011 (World Bank). With a 1.3% population growth in this period, per capita GDP growth was about 4.7%, substantially outpacing Pakistan in this period. Farm yields are up dramatically. When the international textiles quota system of the Multifiber Arrangement ended in 2005, Bangladesh garment factory jobs—a major source of job creation—were at ongoing risk. The speed and astuteness of the market response has been a major test of the resil- iency of the Bangladeshi economy. So far, the out- come is better than many predicted; and the impact of the global crisis on employment in the sector is comparatively modest. But recent factory deaths resulting from disastrous negligence of owners put future growth of this sector in jeopardy—if only because of the resulting global public relations disaster.

Poverty The World Bank 2013 WDI reports (albeit based on only 2005 data) that 23% of the population lives below the $1.25 per-day poverty line in Pakistan, compared with 51% in Bangladesh. But poverty progress has been impressive in the onetime “bas- ket case” of Bangladesh, and incomes of the poor- est people are rising. Many factors have contributed to the relatively rapid decrease in extreme poverty in the country, including the early and quickly dis- seminating green revolution, the impressive role of indigenous nongovernmental organizations (NGOs) fighting poverty in rural areas, opportuni- ties for women’s employment in export industries,

and remittances from relatives working abroad. Bangladesh remains a significantly poorer coun- try, with 80% of Bangladeshis living on less than $2 per day, while the figure is a still very high 61% for Pakistan. But the two countries received much more similar scores on the UNDP’s 2010 multidi- mensional poverty index (discussed in Chapter 5). Pakistan was only slightly less poor, ranking No. 70 with a score of 0.275, while Bangladesh ranked No. 73 with a score of 0.291, when aspects of pov- erty broader than income are considered.

Education and Literacy According to UNESCO, in Pakistan in 2011, the female literacy rate was just 40% (the male rate was 69%) for those 15 and older. In some regions of the country, particularly Baluchistan and the North- west Frontier, it is far lower. Although female lit- eracy is not high in Bangladesh either, it is clearly better than Pakistan by both absolute and relative (gender parity) standards—the UNESCO esti- mate for Bangladesh in 2011 was 53% literacy for all women over age 15 (the male rate was 62%). Thirty times as many public education dollars are spent per pupil for university education as for pri- mary school education. Primary school expendi- tures are extremely unequal, with the lion’s share of funds going to schools that more often train the few students who will eventually go on to univer- sities. Many teachers are hired for political reasons rather than professional competence, and “teacher truancy” is a serious problem. Easterly and other analysts such as Ishrat Husain believe that Paki- stan’s poor performance on education and literacy may result from the incentives of the elite to keep the poor from gaining too much education.

Looking to the future, Bangladesh has the clear edge in school enrollments; for example, in 2011 Bangladesh had a 52% enrollment in secondary school, compared with just 35% in Pakistan (2013 World Bank WDI, Table 2.11). Despite school qual- ity problems in both countries, this differential will translate to higher literacy rates and general knowl- edge in Bangladesh in a few years. In Bangladesh just 30 years ago, attending school was an almost unimaginable luxury for most of the poor. Whereas only half of students completed primary school in 1990, more than two-thirds do today. And recent estimates showed that Bangladesh actually has a

98

female-to-male primary and secondary enrollment ratio of 1.07 to 1, while in Pakistan it is just 0.83. Thus, as we look ahead, then, we can also expect much greater parity in male and female literacy levels in Bangladesh. The nonformal education programs of NGOs such as BRAC provide a major contribution to this progress (see the case study in Chapter 11). But both countries are now making real progress.

Health Life expectancy in Bangladesh is now 69 years, compared with only 65 in Pakistan (2012 Popula- tion Reference Bureau); but in 1970 life expectancy was 54 in Pakistan and only 44 in Bangladesh. Since 1990, the prevalence of child malnutrition in Bangladesh has fallen from two-thirds to less than half. Nutrition in Bangladesh has benefited from a successful green revolution. But child malnutrition remains lower in Pakistan, at about 38%.

Under-5 mortality in Bangladesh has fallen dra- matically. On the eve of independence in 1970, the under-5 mortality rate in Bangladesh was 239 per 1,000 live births; the rate in Pakistan was 180 per 1,000. In 1990, the rate in Bangladesh had fallen to 139, and in Pakistan to 122. By 2011, both countries contin- ued to make strong progress, but again their positions were reversed, with the Bangladesh under-5 mortality rate falling to 46 per 1,000, but that in Pakistan only to 72 per 1,000 (2013 WDI, Table 1.2). Thus, both coun- tries have made progress on health, but the edge is strongly with Bangladesh.

Population Bangladesh has made much greater progress than Pakistan in reducing fertility. Shortly after indepen- dence in 1971, both countries had an extremely high level of over 6 births per woman. In Bangladesh, fertility fell to 2.2 by 2011. But for Pakistan, fertility has fallen only to 3.3 (2013 WDI data), with much of Pakistan’s decline very recent. These changes reflect both cause and effect. Fertility tends to fall as social and economic progress increases. Women perceive better economic opportunities and less need to rely on having several children for security. But with lower fertility, more can be invested in each child in health and education, by families, by govern- ments, and by NGOs. Thus the productivity of the next generation is higher. A virtuous cycle can take

hold as the country passes through its demographic transition (see Chapter 6). Looked at differently, given the negative relationship between population growth and income per capita growth (see Chapter 6), continuing high fertility augurs relatively poorly for Pakistan as we look ahead (though fertility is falling in Pakistan as well). Rather than simply converg- ing, Bangladesh is actually on a trend to pull ahead of Pakistan as they follow divergent paths, with greater human capital investment in Bangladesh. The early and strong emphasis on an effective fam- ily planning strategy was an important factor in the progress of Bangladesh.

Understanding the Divergence What explains the unexpectedly poor performance of Pakistan in social development and recent growth even in relation to Bangladesh, and what might be done to improve it? The most commonly cited examples of countries exhibiting “growth without development” are the Middle Eastern oil-exporting economies of the Persian Gulf states. Elites contest control of natural resources, an enclave economy develops with relatively few strong links to other sectors of the economy, and social spending is crowded out by national defense expenditures— both to ward off external attack, as exemplified by Iraq’s brief conquest of Kuwait in 1990, and at least implicitly also to control the domestic population. In contrast, Pakistan has minimal oil reserves, has to import about four-fifths of its crude oil requirements, and may have to begin importing natural gas.

It is important to note that it is not true that there has been no social progress at all in Pakistan. Rather, the concern is that less progress has been made than in many other countries, even in many that grew much more slowly or experienced negative growth. Why has there been such slow progress?

Geography To the degree that geography constrains develop- ment success, Bangladesh would seem to be at a considerable disadvantage. Tropical countries such as Bangladesh have done more poorly around the world, other things being equal. Pakistan, though facing some geographic disadvantages, would seem to hold the edge here. Moreover, aside from a few city-states and islands, Bangladesh is the

most densely populated country in the world. For perspective, the Netherlands is famous for its crowding and has 495 people per square kilome- ter. But Bangladesh is more than double as densely populated, with 1,174 people per square kilometer (World Bank WDI). Bangladesh has more than half the population of the United States, squeezed into an area less than the size of Wisconsin. (A partial countervailing factor is the greater ease of connect- ing people and economic activity, facilitating the benefits of the division of labor, for example.)

William Easterly and Ross Levine propose that countries with a multitude of social divisions, eth- nic groups, and languages tend to have lower social development and growth rates, although the result is largely muted if the regime is democratic. There is no iron rule here; Mauritius is very diverse but has experienced successful development; India is diverse but has done better than either Pakistan or Bangladesh. Bangladesh is quite homogeneous; as much as 98% of the population is considered ethnic Bangla (Bengali) and speaks the Bangla language. Pakistan has a very high level of ethnic and language diversity. Even its name derives from a compound of Punjab, Afghanistan, Kashmir, and Baluchistan. The official language is Urdu, but it is spoken as a first language by only 7% of the population (the largest language group is Punjabi, at 48%). The failure to provide a fair allocation of revenues and services and resolve other issues for one of the largest ethnic groups, the Bangla, led to the division of Bangladesh from Pakistan in the first place. Easterly concludes that part of the cause of Pakistan’s “fractionalism lies in ethnolinguistic fractionalization” and argues that “Pakistan is the poster child for the hypothesis that a society polarized by class, gender, and ethnic group does poorly at providing public services.”

Gender Equity According to the Social Watch Report, 2013, Ban- gladesh received a gender equity index ranking of 0.55, much higher than the Pakistan score of just 0.29. In Pakistan, as of 2008, only 60% as many women as men were literate—a figure that is little higher in the 15–24 age group. This is a key age group to consider because it represents those old enough to have had a full chance to gain literacy in school yet not be weighted down by past prac- tices, which tend to perpetuate illiteracy in older

groups. In Bangladesh, a significantly higher ratio of female to male literacy of 83% was found in 2008. As already seen, today in Bangladesh, more girls than boys are enrolled in primary education, while in Pakistan, the enrollment level of girls is less than three-quarters that of boys. But both countries have a male-to-female ratio of 1.05, an indicator of gen- der inequality (higher mortality of girls).

The availability of opportunities for work out- side the home, notably in garment factories, has probably increased the autonomy of women. Improved safety is the most urgent priority. Condi- tions are harsh in other ways by Western standards, and many workers are paid below the official mini- mum wage; unions are often suppressed. At the same time, incomes are still far higher than alterna- tives such as domestic work, and the factory jobs have offered a way out for hundreds of thousands of formerly impoverished Bangladeshi women. Ongoing risks facing women factory workers were brought into public view with a factory fire that killed 112 people in November 2012, and a build- ing collapse in April 2013 that killed 1,127 people— the most deadly garment factory disaster in history. More than half of those killed were women; some of their children also died in the buildings. The factory owners may be punished for knowingly subjecting garment workers to risky factory conditions; sus- tained government, union, and civil society action will be needed to help ensure that safety can be instituted before others die needlessly. Fortunately, rather than simply treating this as a public relations disaster and shifting contracts to other countries, in 2013 a group of major European retailers set up an “Accord,” and a grouping of North American retailers set up an “Initiative,” to set standards and monitor workplaces producing their contract gar- ment orders. Of the two programs, the European Accord was viewed by many civil society and union observers as being more legally binding than the North American Initiative—and hence more effective (U.S. retailers claim this is because they could face lawsuit risks). In any case, Bangladeshi workers would benefit from enhanced cooperation and coordination between these two alliances.

Meanwhile, conditions do not seem to be much if any better in Pakistan; for example, in less publicized incidents, more than 300 garment workers died in fac- tory fires in Pakistan in September 2012.

99

100

Aid Pakistan has received a great deal of aid. Since inde- pendence in 1947, it has been one of the top aid- earning countries. In the aftermath of the terrorist attacks on the United States of September 11, 2001, Pakistan assumed great importance as a strategic ally of the United States in the struggle against ter- rorism. Sanctions were lifted, and various forms of aid were greatly increased. Although this should be an opportunity for Pakistan to spur development, and growth has accelerated since 2003 apparently in part as a result, history suggests caution. The country was a major Cold War ally of the United States, but the poor seemed to derive little benefit from that association. Bangladesh has also ben- efited considerably from aid. Effectiveness in the use of aid may be important, particularly the active involvement of effective NGOs in Bangladesh. The major indigenous NGOs and similar groups in Ban- gladesh generally placed a central emphasis on empowerment of women, and the impacts are gen- erally viewed as having been very strong.

Governance and the Role of the Military The military has always played a prominent role in Pakistan, and from 1999 to 2008, the nation was governed by a military ruler, General Pervez Mush- arraf. Pakistan’s long-standing rivalry with India and territorial dispute with it over Kashmir since 1947 have diverted resources as well as government attention from social priorities while reinforcing the influence of the military.

The conflicts in northwest Pakistan and neigh- boring Afghanistan also emphasize a military role. On the other hand, in a heartening sign that democ- racy is taking firmer root, the May 2013 elections were widely considered fair and represented the first time that Pakistan has seen a civilian transfer of power after successful completion of a full term in office of a democratically elected government.

Although the military was very active in Bangla- deshi politics for nearly two decades after indepen- dence in 1971, the military’s relative withdrawal from politics and government after 1990 probably has been a factor in the country’s subsequent prog- ress. Military involvement as the backer of a care- taker government in Bangladesh in 2007 and 2008 was widely viewed as relatively benign, and the country returned to elected civilian rule in 2009,

but political polarization and violence escalated dangerously in late 2013 and early 2014. Neither country has been particularly transparent or free from corruption. In fact, in its 2012 Corruption Per- ceptions Index, Transparency International gave an essentially equally poor score (out of a possible 100) to the two countries, with 27 for Pakistan and only 26 for Bangladesh.

Civil Society Given the weak government and the private sector, one must look to the third sector, variously referred to as the “nongovernmental, nonprofit, or citizen sector.” Here the difference is dramatic. Bangla- desh has one of the most vibrant NGO sectors in the world, the most highly developed in Asia. This will be explored in more detail in the end-of-chapter case studies in Chapter 11, where different approaches of NGOs to poverty action in Bangladesh will be dis- cussed in the cases of BRAC and of the Grameen Bank. If a larger NGO sector could be developed in Pakistan, perhaps led by the many educated Paki- stanis living in the United Kingdom, the United States, and Canada, it might play a similar catalyzing role.

Ishrat Husain proposes that Pakistan has expe- rienced an “elitist growth model,” which he identi- fies as combining a powerful leader or succession of leaders operating without checks and balances, a bureaucratic class that unquestioningly implements the wishes of the leader, and a passive and subser- vient population. He argues that “failure of gover- nance and the consistent domination of political power and state apparatus by a narrowly based elite seeking to advance private and family interests to the exclusion of the majority of the population lies at the root of the problem.” Husain shows that Paki- stan has exhibited these characteristics since inde- pendence and points out that “this combination of strong autocratic leaders, a pliant bureaucracy, and a subservient population made it possible for the benefits of growth to be unequally distributed and concentrated.” He concludes that “the ruling elites found it convenient to perpetuate low literacy rates. The lower the proportion of literate people, the lower the probability that the ruling elite could be replaced.” One reason is that while girls’ education is a boon for development as a whole, it is not neces- sarily in the economic and political interests of some of the elites now in powerful positions, especially at

the local or regional level. The dominance of large landowners over tenants in the social, political, and economic spheres is all too apparent in rural Pakistan. With education, as some landlords and business operators well know, workers, especially women, may finally demand that laws that are in place to protect them be enforced. It is sometimes in the owners’ interest to see that this does not happen.

Concluding Remarks The differences in social development in Bangla- desh and Pakistan are not as overwhelming as would be found in a comparison with Sri Lanka, which has had favorable human development sta- tistics for its low-income level despite enduring civil conflict, or even as dramatic as found between low- income states in India, such as the relatively high human development state of Kerala and the low- development state of Bihar. But Pakistan’s growth has been higher than many countries that have made much greater social improvements and has done much better with available aid. The alterna- tive interpretation of Pakistan’s experience is that economic growth is after all possible even without high investment in health and education. But the long-term trends are for slower growth in Pakistan and higher growth in Bangladesh, making this inter- pretation simply untenable. As Easterly conjectured:

It may be that a certain degree of development and growth was attainable with a skilled managerial elite and unskilled workers, but over time this strategy ran into diminishing returns, as human capital did not grow at the same rate as the other factors. This is consistent with the slowdown in growth from the mid-1980s to the present…. Agricultural growth may have also been possible with the landlord elite taking advantage of the immense potential of the irrigation network and the green revolution, using only unskilled agricultural laborers. But agricultural growth may also have run into diminishing returns, as irrigated land and human capital did not grow at the same rate as other factors of production.

The current development levels of these two countries are not dramatically different. But this itself is the dramatic finding, given the wide dispar- ity when the countries separated in 1971.

Sources Alderman, Harold, and Marito Garcia. “Food security and

health security: Explaining the levels of nutrition in Pakistan.” World Bank Working Paper PRE 865. Wash- ington, D.C.: World Bank, 1992.

Easterly, William. “The political economy of growth without development: A case study of Pakistan.” In In Search of Prosperity: Analytic Narratives on Economic Growth, ed. Dani Rodrik. Princeton, N.J.: Princeton University Press, 2003.

Easterly, William, and Ross Levine. “Africa’s growth trag- edy: Policies and ethnic divisions.” Quarterly Journal of Economics 112 (1997): 1203–1250.

Heston, Alan, Robert Summers, and Bettina Aten. Penn World Table, version 6.3, August 2009. http://pwt. econ.upenn.edu/php_site/pwt63/pwt63_form.php.

Husain, Ishrat. Pakistan: The Economy of an Elitist State. New York: Oxford University Press, 1999.

Hussain, Neelam. “Women and literacy development in Pakistan.” Working paper.

Instituto del Tercer Mundo. Social Watch Report, 2004. Montevideo, Uruguay: Instituto del Tercer Mundo, 2004.

Population Reference Bureau, World Population Datasheet, 2012, and earlier years.

Razzaque, Muhammad Abdur. “Vision 2021: Bangladesh charts a path toward food security.” IFPRI 2012 Global Food Policy Report, pp 80–82.

Sen, Amartya. Development as Freedom. New York: Knopf, 1999.

Smith, Stephen C. “The miracle of Bangladesh: From bas- ket case to case in point.” World Ark, May/June 2009, pp. 13–21.

Summers, Lawrence H. “Investing in all the people.” World Bank Working Paper PRE 905. Washington D.C.: World Bank, 1992.

UNICEF. State of the World’s Children, 2010. New York: UNICEF, 2010.

United Nations. Human Development Report, various years. New York: Oxford University Press.

World Bank. World Development Indicators, various years. Washington, D.C.: World Bank.

——— World Development Report, various years. New York: Oxford University Press. ■

101

102 PART ONE Principles and Concepts

Concepts for Review

Absolute poverty Brain drain Capital stock Convergence Crude birth rate Dependency burden Depreciation (of the capital stock) Diminishing marginal utility Divergence Economic institutions Fractionalization

Free trade Gross domestic product (GDP) Gross national income (GNI) Human capital Human Development Index

(HDI) Imperfect market Incomplete information Infrastructure Least developed countries Low-income countries (LICs)

Middle-income countries Newly industrializing countries

(NICs) Property rights Purchasing power parity (PPP) Research and development

(R&D) Resource endowment Terms of trade Value added World Bank

Questions for Discussion

1. For all of their diversity, many less developed countries are linked by a range of common prob- lems. What are these problems? Which do you think are the most important? Why?

2. Explain the distinction between low levels of liv- ing and low per capita incomes. Can low levels of living exist simultaneously with high levels of per capita income? Explain and give some examples.

3. Can you think of other common (not necessarily universal but widespread) characteristics of less developed countries not mentioned in the text? See if you can list four or five and briefly justify them.

4. Do you think that there is a strong relationship among health, labor productivity, and income lev- els? Explain your answer.

5. What is meant by the statement that many devel- oping nations are subject to “dominance, depen- dence, and vulnerability” in their relations with rich nations? Can you give some examples?

6. Explain the many ways in which developing countries may differ in their economic, social, and political structures.

7. What are some additional strengths and weak- nesses of the Human Development Index as a comparative measure of human welfare? If you were designing the HDI, what might you do dif- ferently, and why?

8. “Social and institutional innovations are as important for economic growth as technologi- cal and scientific inventions and innovations.” What is meant by this statement? Explain your answer.

9. Why do many economists expect income conver- gence between developed and developing coun- tries, and what factors would you look to for an explanation of why this has occurred for only a limited number of countries and in such a limited degree so far?

10. What are good economic institutions, why do so many developing countries lack them, and what can developing countries do to get them? Justify your answer.

11. Which measure shows more equality among countries around the world—GNI calculated at exchange rates or GNI calculated at purchasing power parity? Explain.

12. “South Asia has a lower income per capita than sub-Saharan Africa.” Comment on the validity of this statement.

13. What is the meaning of a “colonial legacy”? Discuss any disadvantages and possible advantages.

14. State five characteristics of the developing world. Discuss diversity within the developing world on these characteristics in relation to the developed world.

103CHAPTER 2 Comparative Economic Development

15. Discuss the differences between the traditional HDI (examined in Appendix 2.1) in comparison to the New NHDI formulation. In what ways do you think either one is a better measure of human development? In your answer, consider the

significance of computing with a geometric mean, instead of an arithmetic mean.

16. What were the central findings of Melissa Dell’s research on the mita system, and what is their sig- nificance for the study of economic development?

Notes

1. Alan Heston, Robert Summers, and Bettina Aten, Penn World Table, version 6.3, Center for Interna- tional Comparisons of Production, Income and Prices, University of Pennsylvania, August 2009, http://pwt.econ.upenn.edu/php_site/pwt63/ pwt63_form.php. Data for 2007.

2. World Bank, World Development Indicators, 2013. (Washington, D.C.: World Bank, 2010), tab. 1.1. Data for 2011. These real measures reflect purchas- ing power parity (explained later in the chapter).

3. United Nations Development Programme (UNDP), Human Development Report, 2005 (New York: Oxford University Press, 2005), p. 38. The global Gini coef- ficient is reported at 0.67 (for details on this mea- sure of inequality, see Chapter 5).

4. World Bank, World Development Indicators, 2010, various tables. Some of these contrasts are sum- marized in Table 2.3 of this text.

5. For more information on country classification systems and other key comparative data, go to the World Bank Web site at http://www.worldbank .org/data, the OECD Web site at http:// www.oecd.org/oecd, and the United Nations Development Programme Web site at http:// www.undp.org. See http://www.unohrlls.org/ en/home/ and http://www.unohrlls.org/en/ ldc/related/59/. Some least developed countries such as Equatorial Guinea are on an “identified for graduation” list; but Equatorial Guinea does not meet “graduation criteria” on human assets or economic vulnerability.

6. Adjustments are made because otherwise the resulting PPP measure would essentially assume that the relative prices prevailing in the United States (i.e., the numeraire currency) also prevailed elsewhere (which would mean that the resulting total incomes would not be “base-country invari- ant”; that is, they would differ if, for example, the

conversions were made to the UK pound ster- ling). Accounting for relative price differences recognizes the substitutions people make toward lower-priced goods in their market basket and thus gives a more accurate comparison of liv- ing standards. For details on calculations of PPP incomes, see the 2011 International Comparison Program site at http://siteresources.worldbank .org/ICPEXT/ Resources/ICP_2011.html, the UN Statistics Division at http://unstats.un.org/unsd/ methods/icp/ipc7_htm.htm, and the Penn World Table site at http://pwt.econ.upenn.edu/about- pwt2.html. These unadjusted figures do provide a useful indicator of the ability of a nation to buy goods and services in dollars abroad, but they are misleading regarding the ability to buy domestically.

There are also other limitations of GNI (and PPP) calculations as measures of economic per- formance and welfare. For example, GNI does not take account of the depletion or degradation of natural resources; it assigns positive values to expenditures resulting from repair and cleanup costs following natural disasters (e.g., earthquakes, hurricanes, floods), to polluting activities, and to the costs of environmental cleanups (see Chapter 10). It frequently ignores nonmonetary transactions, household unpaid labor, and subsistence consump- tion (see Chapter 9). Products consumed by people living in poverty and prices they pay for them dif- fer from the nonpoor. Finally, GNI figures take no account of income distribution (Chapter 5) or capa- bilities other than income.

7. This is accomplished in a special way. First, Equation 2.1 is applied to each of the two subcomponents, just as in Equations 2.3 and 2.4. Then, as explained by the UNDP, “a geometric mean of the resulting indexes is created and and, finally, Equation 2.1 is reapplied to the geometric mean of the indexes using 0 as the minimum and the highest geometric mean of the

104 PART ONE Principles and Concepts

resulting indexes for the time period under consid- eration as the maximum. This is equivalent to apply- ing Equation 2.1 directly to the geometric mean of the two subcomponents.” For further details, see http://hdr.undp.org/en/media/HDR%202013%20 technical%20notes%20EN.pdf.

8. There is still substitutability across the three com- ponents in the NHDI, but not perfect substitut- ability as in the HDI. Regarding the calculation in the last equation of Box 2.1, recall that a geometric mean for the case of three variables is equivalent to the cube root of the product (by the properties of exponents). You can see how a geometric mean is used to build up the overall education index from its two components in the fourth equation in Box 2.1. For an interesting critique of the use of a geometric mean rather than a different functional form that also allows for imperfect substitutability, see Martin Ravallion, “Troubling tradeoffs in the Human Development Index,” World Bank Policy Research Working Paper No. 5484, 2010.

9. The new UNDP measures can be found at http:// hdr.undp.org.

10. It is possible that low income is supplemented by tapping into savings (broadly defined), which would reflect the unsustainable nature of such a low income.

11. See, for example, Jeffrey D. Sachs, The End of Pov- erty: Economic Possibilities for Our Time (New York: Penguin, 2005).

12. World Bank World Development Indicators, 2013.

13. Gunnar Myrdal, Asian Drama (New York: Pan- theon, 1968), app. 2.

14. Recent debates on the incidence of national pov- erty traps and their causes is examined in Chap- ter 4. Economic growth is of course another area of wide variations in the developing world, with historically unprecedented growth in East Asia alongside chronic stagnation—at least until recently—in most of sub-Saharan Africa. Economic growth is a major subject of the next chapter.

15. For a discussion of the relative benefits and costs of country size, see Alberto Alesina and Enrico Spolaore, “On the number and size of nations,” Quarterly Journal of Economics 112 (1997): 1027–1056.

16. For an interesting look at this problem in the con- text of India, see Kaushik Basu, “Teacher truancy

in India: The role of culture, norms and economic incentives,” January 2006, http://ssrn.com/ abstract=956057. We return to this topic in Chap- ters 8 and 11.

17. See Table 2.3, columns 3, 4, and 5. See also World Bank, World Development Indicators, 2007, tabs. 2.14 through 2.20, and World Health Organization, World Health Report, 2006, http://www.who.int/ whr/2006/en/index.html.

18. This widely used benchmark is an updated value for the dollar-a-day level. The standard of $1.25 is increasingly used for reasons described in Chap- ter 5. The 1.2 billion figure, corresponding to 21% of global population, is based on the 2013 World Bank poverty numbers update of April 2013, downloaded July 10, 2013 from: http://www.world bank.org/content/dam/Worldbank/document/ State_of_the_poor_paper_April17.pdf.

19. UNDP, Human Development Report, 2006, p. 269.

20. World Bank, World Development Indicators, 2007, tab. 2.7.

21. World Bank, Global Monitoring Report, 2007 (Washington, D.C.: World Bank, 2007), tab. A.1. For recent evidence, please see Shaohua Chen and Martin Ravallion, “The developing world is poorer than we thought, but no less successful against poverty.” Policy Research Working Paper 4703, World Bank, August 2008.

22. Ibid. and pp. 40–41; World Bank, World Develop- ment Report, 2000/2001 (New York: Oxford Uni- versity Press, 2000); World Development Indicators, 2007, tab. 2.1; Population Reference Bureau, 2006 World Data Sheet, http://www.prb.org/ pdf06/06WorldDataSheet.pdf. Some economists argue that these figures understate poverty inci- dence, but the trend has been clearly favorable, at least until 2006. Since then, the World Bank reports, significantly higher food prices and other consequences of the global economic cri- sis have slowed the pace of poverty reduction substantially.

23. Population Reference Bureau, World Population Trends 2012; World Bank, World Development Indi- cators, 2010, tab. 2.1; Population Reference Bureau, 2009 World Data Sheet, http://www.prb.org/ pdf10/10WorldDataSheet.pdf. For the popula- tion projections, see United Nations, Population Division, “World Population Prospects: The 2008

105CHAPTER 2 Comparative Economic Development

Revision,” June 2009, http://www.un.org/esa/ population/publications/popnews/Newsltr_87.pdf.

24. Population Reference Bureau, 2010 World Data Sheet. 25. See William Easterly and Ross Levine, “Africa’s

growth tragedy: Policies and ethnic divisions,” Quarterly Journal of Economics 112 (1997): 1203–1250, and Alberto Alesina et al., “Fractionalization,” Journal of Economic Growth 8 (2003): 155–194.

26. For a discussion of these issues and the most care- ful attempt at generating the needed data, see Gillette Hall and Harry Anthony Patrinos, eds., Indigenous Peoples, Poverty and Human Develop- ment in Latin America: 1994–2004 (New York: Pal- grave Macmillan, 2006); Haeduck Lee, The Ethnic Dimension of Poverty and Income Distribution in Latin America (Washington, D.C.: World Bank, 1993); and Paul Collier, “The political economy of ethnicity,” Annual World Bank Conference on Development Eco- nomics, 1998 (Washington, D.C.: World Bank, 1999).

27. For a review of the complex statistical issues in sorting out the possible impact of ethnic, reli- gious, and linguistic fractionalization, see Alesina et al., “Fractionalization.” An earlier paper draw- ing somewhat different conclusions using less comprehensive measures is Easterly and Levine, “Africa’s growth tragedy.”

28. The United States, United Kingdom, Japan, Ger- many, France, Italy, and Canada formed the origi- nal Group of Seven (G7) industrial countries, considered the world’s leading economies, to meet annually to deliberate global economic pol- icy; the group was later expanded to include Rus- sia as the Group of Eight (G8).

29. See David Landes, The Wealth and Poverty of Nations: Why Some Are So Rich and Some So Poor (New York: Norton, 1998); Jared Diamond, Guns, Germs, and Steel: The Fates of Human Societies (New York: Norton, 1997); John Luke Gallup, Jeffrey D. Sachs, and Andrew D. Mellinger, “Geography and economic development,” Annual World Bank Con- ference on Development Economics, 1998 (Washington, D.C.: World Bank, 1999), pp. 127–178; and Paul Collier, The Bottom Billion (Oxford: Oxford Univer- sity Press, 2007), who emphasizes the combination of being landlocked with “bad neighbors.”

30. See, for example, Intergovernmental Panel on Climate Change, “Fourth assessment report: Climate change 2007,” http://www.mnp.nl/ipcc/pages_edia/

AR4-chapters.html. The IPCC was established by the World Meteorological Organization (WMO) and the United Nations Environment Program (UNEP) to “assess available scientific, technical, and socioeconomic information relevant for the understanding of climate change, its potential impacts, and options for adaptation and mitiga- tion.” The group won the Nobel Peace Prize in 2007. For more details, see Chapter 10.

31. For a detailed analysis of the importance of information acquisition, its absence in many developing countries, and the consequent role of governments in promoting knowledge and information in the context of limited markets, see World Bank, World Development Report, 1998/99: Knowledge for Development (New York: Oxford University Press, 1998), pp. 1–15.

32. These three factors are identified as critically important in the research by Daron Acemoglu and James A. Robinson; see their Economic Origins of Dictatorship and Democracy (New York: Cambridge University Press, 2005). See also note 58. As Dani Rodrik noted, a caveat is that the institutions gen- erally viewed as favorable are correlated with each other; it is unclear which of these institutions matter most or how specific in form these institu- tions have to be to fulfill their main functions.

33. See Kenneth L. Sokoloff and Stanley L. Engerman, “Factor endowments, institutions, and differential paths of growth among New World economies: A view from economic historians of the United States,” in How Latin America Fell Behind: Essays on the Economic Histories of Brazil and Mexico ed. Stephen Haber, (Stanford, Calif.: Stanford Uni- versity Press, 1997); see also additional works by these authors cited in note 58.

34. See Nathan Nunn and Leonard Wantchekon, “The slave trade and the origins of mistrust in Africa,” American Economic Review 101, No. 7 (December 2011): 3221–3252.

35. Having avoided formal colonization is no guar- antee of development success; Afghanistan and Ethiopia are frequently cited examples. However, it should also be noted that although it was not successfully colonized, Afghanistan was sub- jected to extensive indirect control with British and Russian invasions from the early nineteenth to the early twentieth century (and later by Soviet

106 PART ONE Principles and Concepts

armies with ongoing consequences), and Ethiopia was subject to invasions and intrigue by Italy and Britain. (Liberia, the other frequently cited exam- ple, was also subject to major influence from the developed world.)

36. For an interesting and provocative analysis of the critical role of “ideas” and “ingenuity” in long- term economic growth, see Paul M. Romer, “Idea gaps and object gaps in economic development,” Journal of Monetary Economics 32 (1993): 543–573, and Thomas Homer-Dixon, “The ingenuity gap: Can poor countries adapt to resource scarcity?” Popula- tion and Development Review 21 (1995): 587–612.

37. Romer, “Idea gaps,” 543.

38. See, for example Gallup, Sachs, and Mellinger, “Geography and economic development,” pp. 127–178; Desmond McCarthy, Holger Wolf, and Yi Wu, “The growth costs of malaria,” NBER Work- ing Paper No. W7541, February 2000; and John Luke Gallup and Jeffrey D. Sachs, “The economic burden of malaria,” Harvard University CID Working Paper No. 52, July 2000. See also p. 85.

39. Brinley Thomas, Migration and Economic Growth (London: Cambridge University Press, 1954), p. viii.

40. For an interesting contemporaneous description of the process and implications of international migration from the Mediterranean area to west- ern Europe, see W. R. Böhnung, “Some thoughts on emigration from the Mediterranean basin,” International Labour Review 14 (1975): 251–277.

41. Congressional Budget Office study, June 18, 2013, http://www.cbo.gov/publication/44225.

42. For an analysis of this issue, see Douglas Massey, “The new immigration and ethnicity in the United States,” Population and Development Review 21 (1995): 631–652.

43. UNDP, Human Development Report, 1992 (New York: Oxford University Press, 1992), p. 57.

44. On the emigration of Indian information technol- ogy workers, see “India’s plan to plug the brain drain,” Financial Times, April 24, 2000, p.17.

45. World Bank, “Migration and development briefs,” http://go.worldbank.org/R88ONI2MQ0.

46. For an excellent overview of these issues, see UNDP, Human Development Report, 2009, http:// hdr.undp.org/en.

47. For a discussion, see Simon Commander, Mari Kangasniemi, and L. Alan Winters, “The brain drain: Curse or boon? A survey of the literature,” in Chal- lenges to Globalization: Analyzing the Economics (Chi- cago: University of Chicago Press, 2004), pp. 235–272. See also C. Simon Fan and Oded Stark, “International migration and ’educated unemployment,’” Journal of Development Economics 83 (2007): 76–87.

48. A theoretical contribution to the literature on his- torical growth and its relevance to contemporary developing countries can be found in Marvin Goodfriend and John McDermott, “Early devel- opment,” American Economic Review 85 (1995): 116–133. Goodfriend and McDermott argue that long-term economic development involves four fundamental processes: the exploitation of increasing returns to specialization, the transition from household to market production, knowledge and human capital accumulation, and industri- alization. With regard to developing countries, they argue that “the continuing widespread use of primitive production processes alongside relatively modern techniques is the most striking feature of less-developed countries” (p. 129).

49. Douglass C. North, “Economic performance through time,” American Economic Review 84 (1994): 359–368, and Douglass C. North, Institutions, Institutional Change and Economic Performance (New York: Cam- bridge University Press, 1990). For a provocative analysis of the historical links between economic development and political development, including democratization and the extension of human and legal rights, drawing on economic theory and 500 years of the global historical record, see Acemo- glu and Robinson, Economic Origins of Dictatorship and Democracy, and Acemoglu and Robinson, Why Nations Fail, 2012.

50. In Chapters 3 and 4, we examine economic growth more after including the contending views about whether such diminishing returns apply to aggregate growth experience. For an appeal- ingly intuitive discussion of these two effects, see Eli Berman, “Does factor-biased technological change stifle international convergence? Evidence from manufacturing,” NBER Working Paper, rev. September 2000. Note, however, that other fac- tors such as institutional quality may be at least as important as capital per worker in explaining

107CHAPTER 2 Comparative Economic Development

income per capita, as you will see later in this chapter and in Chapters 3 and 4. On the long-term divergence between developed and developing nations, see Pritchett, “Divergence, big time.”

51. Note that earlier and longer periods tend to show more divergence due to “divergence big time” effect of inequalities growing since the start of the industrial era. Note also that there is evidence of convergence taking place for the years 2001 to 2007 (a shorter period than entertained in this lit- erature but an encouraging and intriguing short- term trend that will bear watching closely); data are awaited to determine if this continued or even amplified after the financial crisis. The sample criteria for the diagrams in Figure 2.8 were as follows. All data are constructed from the Penn World Table using PPP values (which extended through 2007 when the graphs were constructed in 2010). To be included, a country had to have data available in the PWT database for the sample period; by starting in 1980, a relatively small num- ber of countries had to be omitted. For the world diagram, six countries were excluded as 1980 base-period outliers that had very high income in that year due to a temporary oil price increase (Brunei, Qatar, UAE, Libya, Saudi Arabia, and Kuwait). The criterion for inclusion as a develop- ing country in the 1980 base year was classifica- tion as a low- or middle-income country in the 1980 World Bank’s World Development Report; use of this early classification (in the base year of the study) avoided the bias of excluding countries that had grown fast enough to become high- income countries during this period. An impli- cation of this criteria, however, is the exclusion of centrally planned and oil-exporting countries; these two groupings had separate classification categories in the 1980 WDR not based on income level, and for consistency, China is excluded from this group as a centrally planned economy. As it is only one data point, this exclusion does not affect the failure to find country convergence in this period. (But the finding of population-weighted convergence since 1989 is substantially driven by the rapid per capita income growth of China, where China is included; note that China is also included in the world sample in Figure 2.8a). For the purpose of the OECD convergence diagram,

the inclusion criteria was all original members plus Japan, Finland, Australia, and New Zealand, the four countries that joined after its founding but before 1973 (after which no new countries were admitted until Mexico in 1994), but with the exclusion of West Germany due to the statistical problem presented by its 1990 unification with East Germany.

52. For a detailed discussion, see J. Bradford De Long, “Productivity growth, convergence, and welfare: Comment,” American Economic Review 78 (1988): 1138–1154.

53. Figure from Human Development Report, 2005, ch. 1. For graphs showing relative versus absolute income convergence for China in relation to the United States, see Stephen C. Smith, http://www .gwu.edu/~iiep/G2/. Note that in the 2008-2012 period, continued rapid growth in China, in com- bination with very slow growth in a majority of high-income OECD countries, led to estimates that China achieved higher per capita GDP abso- lute gains than OECD high-income countries. (Such absolute inequality measures can be used to address other kinds of questions about conver- gence or divergence, but this is rarely done. For example, it could be used in the size distribution of income at national or even international scale; but the usual preferred property for inequality measures is to make relative income comparisons. Details are discussed in Chapter 5.)

54. UNDP, Human Development Report 2013, p.13.

55. Branko Milanovic, “True world income distribu- tion, 1988 and 1993: First calculation based on household surveys alone,” Economic Journal 112, (2002) 51–92.

56. See Dani Rodrik, “Unconditional convergence in manufacturing,” Quarterly Journal of Economics 128, No. 1 (2012): 165–204.

57. We thank Daron Acemoglu, Shahe Emran, Stanley Engerman, and Karla Hoff for their helpful com- ments on this section. Not all of the causal links described here are supported by the same type of evidence. Some are underpinned by widely (if not universally) accepted statistical (econo- metric) evidence. Other causal links emerge from historical studies. All links discussed are argued in the development economics literature to be

108 PART ONE Principles and Concepts

underlying factors leading to divergent devel- opment outcomes. The discussion follows the numbering of the arrows in Figure 2.11, which is arranged for concise display.

58. For very readable introductions to this research, see Daron Acemoglu, Simon Johnson, and James A. Robinson, “Understanding prosperity and poverty: Geography, institutions, and the reversal of fortune,” in Understanding Poverty, eds. Abhijit Banerjee, Roland Benabou, and Dilip Mookher- jee (New York: Oxford University Press, 2006), pp. 19–36, and Stanley L. Engerman and Kenneth L. Sokoloff, “Colonialism, inequality, and long-run paths of development,” in Understanding Poverty, pp. 37–62. See also Daron Acemoglu, Simon John- son, and James A. Robinson, “The colonial origins of comparative development: An empirical inves- tigation,” American Economic Review 91 (2001): 1369–1401, and Kenneth L. Sokoloff and Stanley L. Engerman, “History lessons: Institutions, fac- tor endowments, and paths of development in the New World,” Journal of Economic Perspectives 14 (2000): 217–232. For an excellent review of the work of these authors, see Karla Hoff, “Paths of institutional development: A view from economic history,” World Bank Research Observer 18 (2003): 205–226. See also Dani Rodrik, Arvind Subrama- nian, and Francesco Trebbi, “Institutions rule: The primacy of institutions over geography and inte- gration in economic development,” Journal of Eco- nomic Growth 9 (2004): 135–165, and Dani Rodrik and Arvind Subramanian, “The primacy of insti- tutions, and what this does and does not mean,” Finance and Development (June 2003), http://www .imf.org/external/pubs/ft/fandd/2003/06/pdf/ rodrik.pdf.

59. On the role of geography, see Diamond, Guns, Germs, and Steel; Gallup, Sachs, and Mellinger, “Geography and economic development”; Jeffrey D. Sachs, “Institutions don’t rule: Direct effects of geography on per capita income,” NBER Work- ing Paper No. 9490, 2003; and Jeffrey D. Sachs, “Institutions matter, but not for everything,” Finance and Development (June 2003), http://www .imf.org/external/pubs/ft/fandd/2003/06/pdf/ sachs.pdf. See also Douglas A. Hibbs and Ola Olsson, “Geography, biogeography and why some countries are rich and others poor,” Proceedings of

the National Academy of Sciences (2004): 3715–3740; for a discussion on landlocked status as it affects poor African economies, see Paul Collier, The Bot- tom Billion: Why the Poorest Countries Are Failing and What Can Be Done about It (Oxford: Oxford University Press, 2007), pp. 53–63, 165–166, and 179–180.

60. See Quamrul Ashraf and Oded Galor, “The ’Out of Africa’ hypothesis, human genetic diver- sity, and comparative economic development,” American Economic Review 103 (2013): 1-46. It is doubtful whether there could be any meaningful policy implications.

61. See Chapter 4 for an analysis of problems of coor- dination failure and the importance of mecha- nisms to correct it.

62. For accessible discussions, see North, Institutions, Institutional Change and Economic Performance; Justin Lin and Jeffrey Nugent, “Institutions and economic development,” Handbook of Economic Development, vol. 3A (Amsterdam: North Holland, 1995); Dani Rodrik, “Institutions for high-quality growth: What they are and how to acquire them,” Studies in Comparative International Development 35, No. 3 (September 2000): 3–31; and Acemoglu, Johnson, and Robinson, “Understanding prosper- ity and poverty.” Note that the quality of many of the institutions described in this paragraph of the text is correlated, and it is disputed which ones matter most and the degree to which they are substitutes for each other in spurring growth.

63. As an instrument for the types of institutions established (note that scholars have widely debated this instrument). For a discussion, with some important caveats, see Rodrik, Subramanian, and Trebbi, “Institutions rule.”

64. This is after the problem of simultaneity between income and institutions is controlled for by taking advantage of the exogeneity of initial settler mor- tality risk (other approaches using different data still find some role for geography; see the papers by Sachs in note 59). See Acemoglu, Johnson, and Robinson, “Colonial origins of comparative devel- opment.” The schema on page 1370 in their paper corresponds to links 3-10-18-21 or 3-10-19-22 in Figure 2.10 in this text. See also Daron Acemoglu, Simon Johnson, James A. Robinson, and Yunyong

109CHAPTER 2 Comparative Economic Development

Thaicharoen, “Institutional causes, macroeco- nomic symptoms: Volatility, crises and growth,” Journal of Monetary Economics 50 (2003): 49–123. For a summary, see Daron Acemoglu, “Root causes: A historical approach to assessing the role of institutions in economic development,” Finance and Development (June 2003), http://www.imf .org/external/pubs/ft/fandd/2003/06/pdf/Ace- moglu.pdf. It is also worth noting, however, that in the early colonial period, potential settlers who did wish to emigrate to Latin America and the Caribbean (and perhaps to some other colonies in later times) were sometimes restricted by immigra- tion rules. See Stanley L. Engerman and Kenneth L. Sokoloff, “Factor endowments, inequality, and paths of development among New World econo- mies,” Journal of LACEA Economia 3, No. 1 (Fall) (2002): 41–109. There is also some question about the use of largely eighteenth-century mortality data, which may possibly differ from earlier (but unavailable) mortality rates. These points may suggest some possible limitations to the mortal- ity data-based research, although the results show considerable robustness. For a debate, see David Y. Albouy, “The colonial origins of comparative development: An empirical investigation: Com- ment.” American Economic Review, 102, No. 6 (2012): 3059–-3076, and Acemoglu, Johnson, and Robinson, “The colonial origins of comparative development: An empirical investigation: Reply.” American Economic Review, 102, No. 6 (2012): 3077–3110. See also Rodrik et al., “Institutions rule,” and Pranab Bardhan, “Institutions matter, but which ones?” Economics of Transition 13 (2005): 499–532.

65. Sokoloff and Engerman, “History lessons”; Enger- man and Sokoloff, “Colonialism, inequality, and long-run paths of development.”

66. Engerman and Sokoloff, “Colonialism, inequality, and long-run paths of development.” On the role of labor scarcity in the development of institutions in North America, see David Galenson, “The set- tlement and growth of the colonies: Population, labor and economic development,” in The Cam- bridge Economic History of the United States, vol. 1, eds. Stanley L. Engerman and Robert Gallman (New York: Cambridge University Press, 1996).

67. See Daron Acemoglu, Simon Johnson, and James A. Robinson, “Reversal of fortune: Geography and

institutions in the making of the modern world income distribution,” Quarterly Journal of Econom- ics 118 (2002): 1231–1294. Although the reversal is now associated with this article, similar historical observations were a theme of the “dependency theory” literature, described in Chapter 3.

68. In fact, the Acemoglu-Johnson-Robinson theory could be said to turn dependency theory on its head. The neo-Marxist dependency theory (see Chapter 3) views development constraints as com- ing from foreign nationals, but in the Acemoglu et al. theory, the underlying development problem is the presence of extractive institutions, whether the extractors are nationals or foreigners, and the corrective is investment-encouraging institutions, whoever implements them. The preferred insti- tutions include some that are clearly non-Marxist, such as broader respect for private property rights. The implication of their argument is that it is at best no more important to get today’s rich countries to change their current behavior toward developing countries than it is to achieve reforms in local institutions, although former colonial powers might reasonably be asked to pay for costs of changing over to better domestic institutions, assuming that such change is possible. Inequality makes reform difficult to achieve.

69. This evidence is presented in Acemoglu, Johnson, and Robinson, “Reversal of fortune.” The evi- dence has been criticized by some economists on the grounds that measures of modern institutions actually show great variability rather than per- sistence and may follow rather than lead growth; see, for example, Edward L. Glaeser, Rafael La Porta, Florencio Lopez de Silanes, and Andrei Shleifer, “Do institutions cause growth?” Journal of Economic Growth 91 (2004): 271–303, who argue that human capital is a more fundamental fac- tor. But for a theoretical analysis of how change in specific political institutions is consistent with stability in economic institutions, see Daron Ace- moglu and James A. Robinson, “De facto political power and institutional persistence, American Eco- nomic Review 96 (2006): 326–330. For an empirical analysis providing evidence that education does not, in fact, lead to democracy within countries over time, see Daron Acemoglu, Simon Johnson, James A. Robinson, and Pierre Yared, “From

110 PART ONE Principles and Concepts

education to democracy?” American Economic Review 95 (2005): 44–49. Other critical commen- tary is found in Pranab K. Bardhan, “Institutions matter, but which ones?” Economics of Transition 13 (2005): 499–532.

70. The primary evidence for this is historical. See Landes, Wealth and Poverty of Nations. For exam- ple, the fragmentation of a continent divided by mountains, sea lanes, and rivers facilitated politi- cal competition that fueled institutional develop- ment. See also Diamond, Guns, Germs, and Steel.

71. See David Fielding and Sebastian Torres, “Cows and conquistadors: A contribution on the colo- nial origins of comparative development,” Journal of Development Studies 44 (2008): 1081–1099, and James Feyrer and Bruce Sacerdote, “Colonialism and modern income: Islands as natural experi- ments,” Review of Economics and Statistics 91 (2009): 245–262. Both build on the pioneering research of Acemoglu, Johnson, and Robinson.

72. Fielding and Torres, “Cows and conquistadors.” The neo-Europes are primarily the United States, Canada, Australia, and New Zealand.

73. See Feyrer and Sacerdote, “Colonialism and mod- ern income.” The authors use wind direction and wind speed as instruments for length and type of colonial experience of islands. They iden- tify a positive relationship between colonization length and both income and child survival rates. They also use their evidence to argue that “time spent as a colony after 1700 is more beneficial to modern income than years before 1700, consis- tent with a change in the nature of colonial rela- tionships over time.” Note, however, that some islands included in this research are still colonies, such as overseas French departments with large European populations, and that in other indepen- dent former colonies with high incomes, the orig- inal inhabitants were largely wiped out—facts that weaken the case for benefits of longer colo- nization from the viewpoint of those who were colonized. But on a positive historical note, Stan- ley Engerman pointed out that in the later colo- nial period, Europeans were often responsible for ending slavery in Africa (personal communica- tion with the authors).

74. Engerman and Sokoloff, “Colonialism, inequal- ity, and long-run paths of development.” For supporting econometric evidence on the nega- tive effects of inequality using an identification strategy inspired by the Engerman and Sokoloff hypothesis, see Box 2.2. See also William Easterly and Ross Levine, “Tropics, germs, and crops: The role of endowments in economic development,” Journal of Monetary Economics 50 (2003): 3–39. For a different argument, see Edward L. Glaeser, Gia- como Ponzetto, and Andrei Shleifer, “Why does democracy need education?” NBER Working Paper No. 12128, March 2006; however, see also Acemoglu et al., “From education to democracy?” For alternative perspectives, see Acemoglu and Robinson, Economic Origins of Dictatorship and Democracy. It remains unclear whether economic or political inequality is more fundamental, as politicians often amass wealth when their power is secure. For an interesting study suggesting that the latter is important, see Daron Acemo- glu, Maria Angelica Bautista, Pablo Querubin, and James A. Robinson, “Economic and political inequality in development: The case of Cundina- marca, Colombia,” June 2007, http://econ-www .mit.edu/faculty/download_pdf.php?id=1510.

75. Although in this century so far inequality has been rising in North America and falling some- what in some Latin American countries, the contrast remains extreme. For a set of excellent analyses on recent trends, see Luis F. López-Calva and Nora Lustig, eds., Declining Inequality in Latin America: A Decade of Progress? (Washington, D.C.: Brookings Institution, 2010).

76. Engerman and Sokoloff, “Colonialism, inequal- ity, and long-run paths of development.” See also Edward L. Glaeser, “Inequality,” in The Oxford Handbook of Political Economy, eds. Barry R. Wein- gast and Donald Wittman (New York: Oxford University Press, 2006), pp. 624–641.

77. See Glaeser et al., “Do institutions cause growth?”

78. Acemoglu et al., “From education to democracy?” esp. pp. 47–48. Evidence for the intuitive idea that migrants to the “neo-Europes” settled by Brit- ain embodied not just better institutions but also higher human capital levels is not well established; see Acemoglu, Johnson, and Robinson, “Colonial

111CHAPTER 2 Comparative Economic Development

origins of comparative development.” The effects of institutions held even when excluding these coun- tries. Another possible channel, recently introduced by Gregory Clark, is that institutions affect prefer- ences, which in turn directly or indirectly affect the quality of the workforce. For his provocative and controversial assessment, see A Farewell to Alms: A Brief Economic History of the World (Princeton, N.J.: Princeton University Press, 2007).

79. See, for example, Bardhan, “Institutions matter.” This article also argues some limitations of the empirical methods of Acemoglu and colleagues.

80. Glaeser et al., “Do institutions cause growth?”

81. See Jeffrey Frankel and David Romer, “Does trade cause growth?” American Economic Review 89 (1999): 379–399.

82. Not surprisingly, trade effects are complex. Geog- raphy can influence the pattern and amount of trade. And as countries develop and incomes rise, countries trade in greater amounts and in a wider range of goods. See Rodrik, Subramanian, and Trebbi, “Institutions rule.” They provide a dia- gram of the effects outlined in this paragraph in their Figure 1.

83. Of course, the effectiveness of each sector may also affect the effectiveness of the other sectors. This is not shown in the diagram.

84. Bardhan, “Institutions matter”; Rodrik, “Getting institutions right.” For a provocative analysis of the historical links between economic devel- opment and political development, including

democratization and the extension of human and legal rights, drawing on economic theory and 500 years of the global historical record, see Daron Acemoglu and James A. Robinson, Economic Ori- gins of Dictatorship and Democracy. For an insight- ful analysis of diverging development paths, see Kenneth L. Sokoloff and Stanley L. Engerman, “History lessons: Institutions, factor endow- ments, and paths of development in the New World,” Journal of Economic Perspectives 14 (2000): 217–232.

85. See, for example, UNDP, Human Development Report, 2005.

86. Dani Rodrik, “Institutions for high-quality growth: What they are and how to acquire them,” Studies in Comparative International Development 35, No. 3 (2000), 3–31, DOI: 10.1007/BF02699764, p. 5

87. For an elaboration of this point, see Chapter 8 and also Lawrence H. Summers and Vinod Thomas, “Recent lessons of development,” World Bank Research Observer 8 (1993): 241–254; Pam Woodall, “The global economy,” Economist, October 1, 1994, pp. 3–38; World Bank, World Development Indica- tors, 1998 (Washington, D.C.: World Bank, 1998), pp. 3–11; and UNDP, Human Development Report, 2003.

88. Similar conclusions can be found in Irma Adel- man and Cynthia Taft Morris, “Development his- tory and its implications for development theory,” World Development 25 (1997): 831–840.

112 PART ONE Principles and Concepts

Appendix 2.1

The Traditional Human Development Index (HDI)

Like the New HDI, the Traditional HDI ranks all countries on a scale of 0 (lowest human development) to 1 (highest human development). The Traditional HDI, the UNDP centerpiece until 2010, is still widely referenced, and in this Appen- dix we present it in detail with calculations and comparative examples. The Traditional HDI is based on three goals or end products of development, corre- sponding to health, education, and income: longevity as measured by life expec- tancy at birth, knowledge as measured by a weighted average of adult literacy (two-thirds) and gross school enrollment ratio (one-third), and standard of living as measured by real per capita gross domestic product adjusted for the differing purchasing power parity of each country’s currency to reflect cost of living and for the assumption of diminishing marginal utility of income. Using these three measures of development and applying a formula to data for 177 countries, the HDI ranks countries into four groups: low human development (0.0 to 0.499), medium human development (0.50 to 0.799), high human development (0.80 to 0.90), and very high human development (0.90 to 1.0).

Adjusted income is found by taking the log of current income. Then, to find the income index, one subtracts the log of 100 from the log of current income, on the assumption that real per capita income cannot possibly be less than $100 PPP.1

The difference gives the amount by which the country has exceeded this “lower goalpost.” To put this achievement in perspective, consider it in relation to the maximum that a developing country might reasonably aspire to over the coming generation. The UNDP sets this maximum at $40,000 PPP. So we then divide by the difference between the log of $40,000 and the log of $100 to find the country’s relative income achievement. This gives each country an index number that ranges between 0 and 1. For example, for the case of Bangladesh, whose 2007 PPP GDP per capita was estimated by the UNDP to be $1,241, the income index for that year is calculated as follows:

Income index = 3log (1,241) - log (100)4 3log (40,000) - log (100)4

= 0.420 (A2.1)

The effect of diminishing marginal utility is clear. An income of $1,241, which is just 3% of the maximum goalpost of $40,000, is already enough to reach more than two-fifths of the maximum value that the index can take. Note that a few countries have already exceeded the $40,000 PPP income target; in such cases, the UNDP assigned the maximum value of $40,000 PPP income, and so the country gets the maximum income index of 1.

To find the life expectancy (health proxy) index, the UNDP starts with a country’s current life expectancy at birth and subtracts 25 years. The latter is the lower goalpost, the lowest that life expectancy could have been in any coun- try over the previous generation. Then the UNDP divides the result by 85 years minus 25 years, or 60 years, which represents the range of life expectan- cies expected over the previous and next generations. That is, it is anticipated

113CHAPTER 2 Comparative Economic Development

that 85 years is a maximum reasonable life expectancy for a country to try to achieve over the coming generation. For example, for the case of Bangladesh, whose population life expectancy in 2007 was 65.7 years, the life expectancy index is calculated as follows:

Life expectancy index = 65.7 - 25 85 - 25

= 0.678 (A2.2)

Notice that no diminishing marginal utility of years of life are assumed; the same holds for the education index. The education index is made up of two parts, with two-thirds weight on literacy and one-third weight on school enrollment. Because gross school enrollments can exceed 100% (because of older students going back to school), this index is also capped at 100%. For the case of Bangladesh, adult literacy is estimated (rather uncertainly) at 53.5%, so

Adult literacy index = 53.5 - 0 100 - 0

= 0.535 (A2.3)

For the gross enrollment index, for Bangladesh it is estimated that 52.1% of its primary, secondary, and tertiary age population are enrolled in school, so the country receives the following value:

Gross enrollment index = 52.1 - 0 100 - 0

= 0.521 (A2.4)

Then, to get the overall education index, the adult literacy index is multi- plied by two-thirds and the gross enrollment index is multiplied by one-third. This choice reflects the view that literacy is the fundamental characteristic of an educated person. In the case of Bangladesh, this gives us

Education index = 2 3 1adult literacy index2 +

1 3 1gross enrollment index2

= 2 3 10.5352 +

1 3 10.5212 = 0.530 (A2.5)

In the final index, each of the three components receives equal, or one- third, weight. Thus,

HDI = 1 3 1income index2+

1 3 1life expectancy index2+

1 3 1education index2 (A2.6)

For the case of Bangladesh,

HDI = 1 3 10.4202 +

1 3 10.6782 +

1 3 10.5302 = 0.543 (A2.7)

One major advantage of the HDI is that it does reveal that a country can do much better than might be expected at a low level of income and that substan- tial income gains can still accomplish relatively little in human development.

Moreover, the HDI reminds us that by development, we clearly mean broad human development, not just higher income. Many countries, such as some of the higher-income oil producers, have been said to have experi- enced “growth without development.” Health and education are inputs into the national production function in their role as components of human capital,

114 PART ONE Principles and Concepts

meaning productive investments embodied in persons. Improvements in health and education are also important development goals in their own right (see Chapter 8). We cannot easily argue that a nation of high-income individ- uals who are not well educated and suffer from significant health problems that lead to their living much shorter lives than others around the globe has achieved a higher level of development than a low-income country with high life expectancy and widespread literacy. A better indicator of development disparities and rankings might be found by including health and education variables in a weighted welfare measure rather than by simply looking at income levels, and the HDI offers one very useful way to do this.

There are other criticisms and possible drawbacks of the HDI. One is that gross enrollment in many cases overstates the amount of schooling, because in many countries, a student who begins primary school is counted as enrolled without considering whether the student drops out at some stage. Equal (one- third) weight is given to each of the three components, which clearly has some value judgment behind it, but it is difficult to determine what this is. Note that because the variables are measured in very different types of units, it is difficult even to say precisely what equal weights mean. Finally, there is no attention to the role of quality. For example, there is a big difference between an extra year of life as a healthy, well-functioning individual and an extra year with a sharply limited range of capabilities (such as being confined to bed). Moreover, the qual- ity of schooling counts, not just the number of years of enrollment. Finally, it should be noted that while one could imagine better proxies for health and edu- cation, measures for these variables were chosen partly on the criterion that suf- ficient data must be available to include as many countries as possible.

Table A2.1.1 shows the 2009 Traditional Human Development Index (using 2007 data) for a sample of 24 developed and developing nations ranked from low to very high human development (column 3) along with their respective real GDP per capita (column 4) and a measure of the differential between the GDP per capita rank and the HDI rank (column 5). A positive number shows by how much a country’s relative ranking rises when HDI is used instead of GDP per capita, and a negative number shows the opposite. We see from Table A2.1.1 that the country with the lowest HDI (0.340) in 2007 was Niger, and the one with the highest (0.971) was Norway.

The HDI has a strong tendency to rise with per capita income, as wealthier countries can invest more in health and education, and this added human capital raises productivity. But what is so striking is that despite this expected pattern, there is still such great variation between income and broader mea- sures of well-being, as seen in Tables A2.1.1 and A2.1.2. For example, Senegal and Rwanda have essentially the same average HDI despite the fact that real income is 92% higher in Senegal. And Costa Rica has a higher HDI than Saudi Arabia, despite the fact that Saudi Arabia has more than double the real per capita income of Costa Rica. Many countries have an HDI significantly differ- ent from that predicted by their income. South Africa has an HDI of 0.683, but it ranks just 129th, 51 places lower than to be expected from its middle-income ranking. But similarly ranked São Tomé and Príncipe (number 131) ranks 17 places higher than expected from its income level.

For the countries listed in Table A2.1.2 with GDP per capita near $1,000, the HDI ranges dramatically from 0.371 to 0.543. Correspondingly, literacy rates

115CHAPTER 2 Comparative Economic Development

range from just 26% to 71%. Life expectancy ranges from only 44 to 61. Among countries with GDP per capita near $1,500, literacy ranges from 32% to 74% and enrollment, from 37% to 60%, with corresponding variations in the HDI. For the countries in Table A2.1.1 with GDP per capita near $2,000, the HDI ranges, from 0.511 to 0.710. Life expectancy ranges from 48 to 68. The literacy rate ranges from 56% to 99%. For countries listed in Table A2.1.1 with GDP per capita near $4,000, the HDI index ranges from 0.654 to 0.768. Life expectancy ranges from 65 to 74, and literacy rates range strikingly from 56% in Morocco to essentially universal literacy in Tonga. These dramatic differences show that the Human Development Index project is worthwhile. Ranking countries only by income—or for that matter only by health or education—causes us to miss important differences in countries’ development levels.

Average income is one thing, but sometimes even in a middle-income country, many people live in poverty. When the aggregate HDI for various countries was adjusted for income distribution, the relative rankings of many developing nations also changed significantly.2 For example, Brazil had such a highly unequal distribution that its ranking slipped, while Sri Lanka saw its HDI ranking rise due to its more egalitarian distribution.

TABLE A2.1.1 2009 Traditional Human Development Index for 24 Selected Countries (2007 Data)

Source: Data from United Nations Development Programme, Human Development Report, 2009, tab. 1.

Country Relative Ranking Human Development

Index (HDI) GDP Per Capita

(PPP, U.S. $) GDP Rank Minus

HDI Rank

Low Human Development     Niger 182 0.340 627 –6 Afghanistan 181 0.352 1,054 –17 Dem. Rep. Congo 176 0.389 298 5 Ethiopia 171 0.414 779 0 Rwanda 167 0.460 866 1 Côte d’Ivoire 163 0.484 1,690 –17 Malawi 160 0.493 761 12 Medium Human Development     Bangladesh 146 0.543 1,241 9 Pakistan 141 0.572 2,496 –9 India 134 0.612 2,753 –6 South Africa 129 0.683 9,757 –51 Nicaragua 124 0.699 2,570 6 Gabon 103 0.755 15,167 –49 China 92 0.772 5,383 10 Iran 88 0.782 10,955 –17 Thailand 87 0.783 8,135 –5 High Human Development     Saudi Arabia 59 0.843 22,935 –19 Costa Rica 54 0.854 10,842 19 Cuba 51 0.863 6,876 44 Chile 44 0.878 13,880 15 Very High Human Development     United Kingdom 21 0.947 35,130 –1 United States 13 0.956 45,592 –4 Canada 4 0.966 35,812 14 Norway 1 0.971 53,433 4

116 PART ONE Principles and Concepts

The HDI also ranges greatly for groups within countries. The impact of social exclusion can be seen vividly in Guatemala, where the Q’eqchi ethnic group had an HDI rank similar to Cameroon, and the Poqomchi ranked below Zimbabwe, as seen in Figure A2.1.1a. Regional differences across districts can be seen in Kenya, where the HDI of the capital area of Nairobi ranks as high as Turkey, but Kenya’s Turkana district’s HDI is lower than that of any country average, as shown in Figure 2.3b. Rural-urban differences are illustrated in China, where as Figure A2.1.1c shows, urban Shanghai’s HDI is nearly as high as that of Greece, while rural Gansu has an HDI on a par with India, and the HDI of rural Guizhou is below that of Cambodia. An earlier UN study found similarly that in South Africa, whites enjoy a high HDI level, while that for blacks was much lower.3

Among other things, the Traditional HDI had a large impact on encourag- ing conceptualization of development in a holistic way, elevating health and education to the same rank as income as development indicators; and broad- ening the types of measures, both individual and composite, that were calcu- lated and reported on a regular basis.

TABLE A2.1.2 2009 Human Development Index Variations for Similar Incomes (2007 Data)

Source: Data from United Nations Development Programme, Human Development Report, 2009, tab. 1.

Country GDP Per

Capita (U.S. $) HDI HDI Rank

Life Expectancy

(years) Adult

Literacy (%)

Combined Gross Enrollment

Ratio

GDP Per Capita near PPP $1,000     Madagascar 932 0.543 145 59.9 70.7 61.3 Haiti 1,140 0.532 149 61.0 62.1 52.1 Rwanda 866 0.460 167 49.7 64.9 52.2 Mali 1,083 0.371 178 48.1 26.2 46.9 Afghanistan 1,054 0.352 181 43.6 28.0 50.1 GDP Per Capita near PPP $1,500     Kenya 1,542 0.541 147 53.6 73.6 59.6 Ghana 1,334 0.526 152 56.5 65.0 56.5 Côte d’Ivoire 1,690 0.484 163 56.8 48.7 37.5 Senegal 1,666 0.464 166 55.4 41.9 41.2 Chad 1,477 0.392 175 48.6 31.8 36.5 GDP Per Capita near PPP $2,000     Kyrgyzstan 2,006 0.710 120 67.6 99.3 77.3 Laos 2,165 0.619 133 64.6 72.7 59.6 Cambodia 1,802 0.593 137 60.6 76.3 58.5 Sudan 2,086 0.531 150 57.9 60.9 39.9 Cameroon 2,128 0.523 153 50.9 67.9 52.3 Mauritania 1,927 0.520 154 56.6 55.8 50.6 Nigeria 1,969 0.511 158 47.7 72.0 53.0 GDP Per Capita near PPP $4,000     Tonga 3,748 0.768 99 71.7 99.2 78.0 Sri Lanka 4,243 0.759 102 74.0 90.8 68.7 Honduras 3,796 0.732 112 72.0 83.6 74.8 Bolivia 4,206 0.729 113 65.4 90.7 86.0 Guatemala 4,562 0.704 122 70.1 73.2 70.5 Morocco 4,108 0.654 130 71.0 55.6 61.0

117CHAPTER 2 Comparative Economic Development

FIGURE A2.1.1 Human Development Disparities within Selected Countries

Source: From Human Development Report, 2005, figs. 10–12. Reprinted with permission from the United Nations Development Programme.

0.8

0.7

India

Indonesia

Botswana

Cameroon Zimbabwe

0.6

0.5

0.4 (a) Large ethnic differences in HDI in Guatemala (b) Wide inequalities in human development

between districts in Kenya

(c) Rural-urban differences intensify regional disparities in China

H D

I, 2

0 0

4

0.8

0.7 Turkey

South Africa0.6

0.5

0.4

0.3

0.2

H D

I, 2

0 0

4

0.8

1.0

0.9 Greece

Urban

China

Rural

India Cambodia

Botswana

Urban Shanghai

Rural

Urban

Gansu

Rural

Urban

Guizhou

Rural

0.7

0.6

0.5

H D

I, 2

0 0

4

Niger Mali Busia

Kenya

Nairobi

Mombassa

Turkana

United Arab

Emirates

Q’eqchi

Achi

Ladino

Guatemala

Poqomchi

Notes

1. In fact, Lant Pritchett argues persuasively, consid- ering available country data and the cost of mini- mum nutrients, that $250 is a more realistic lower bound for per capita income. See Lant Pritchett, “Divergence, big time,” Journal of Economic Perspec- tives 11, No. 3 (1997): 3–17. The logarithms used in the Traditional HDI income index formula are common (base 10) logs rather than natural logs.

2. UNDP, Human Development Report, 1994 (New York: Oxford University Press, 1994).

3. All but the South Africa example are drawn from Human Development Report, 2006 (New York: Oxford University Press, 2006). An earlier Human Development Report gave South Africa an overall ranking of 0.666, with whites at 0.876 and blacks at 0.462.

PART ONE Principles and Concepts

Every nation strives after development. Economic progress is an essential com- ponent, but it is not the only component. As noted in Chapter 1, development is not purely an economic phenomenon. In an ultimate sense, it must encompass more than the material and financial side of people’s lives, to expand human freedoms. Development should therefore be perceived as a multidimensional process involving the reorganization and reorientation of entire economic and social systems. In addition to improvements in incomes and output, it typically involves radical changes in institutional, social, and administrative structures as well as in popular attitudes and even customs and beliefs. Finally, although development is usually defined in a national context, its more widespread real- ization may necessitate modification of the international economic and social system as well.

In this chapter, we explore the historical and intellectual evolution in scholarly thinking about how and why development does or does not take place. We do this by examining four major and often competing development theories. You will see that each offers valuable insights and a useful per- spective on the nature of the development process. Some newer models of development and underdevelopment draw eclectically on the classic theories, and we consider them in Chapter 4.

Approaches to the analysis of economic growth are introduced throughout this review of alternative theories of development and are then amplified in three chapter appendixes.

Classic Theories of Economic Growth and Development

There is no Economic Theory of Everything, —Robert Solow, Nobel laureate in economics

[In] modern economic growth…the rate of structural transformation of the economy is high.

—Simon Kuznets, Nobel laureate in economics

3

118

119CHAPTER 3 Classic Theories of Economic Growth and Development

3.1 Classic Theories of Economic Development: Four Approaches

The classic post–World War II literature on economic development has been dominated by four major and sometimes competing strands of thought: (1) the linear-stages-of-growth model, (2) theories and patterns of structural change, (3) the international-dependence revolution, and (4) the neoclassical, free- market counterrevolution. In recent years, an eclectic approach has emerged that draws on all of these classic theories.

Theorists of the 1950s and 1960s viewed the process of development as a series of successive stages of economic growth through which all countries must pass. It was primarily an economic theory of development in which the right quantity and mixture of saving, investment, and foreign aid were all that was necessary to enable developing nations to proceed along an economic growth path that had historically been followed by the more developed countries. Devel- opment thus became synonymous with rapid, aggregate economic growth.

This linear-stages approach was largely replaced in the 1970s by two competing schools of thought. The first, which focused on theories and patterns of struc- tural change, used modern economic theory and statistical analysis in an attempt to portray the internal process of structural change that a “typical” developing country must undergo if it is to succeed in generating and sustaining rapid eco- nomic growth. The second, the international-dependence revolution, was more radical and more political. It viewed underdevelopment in terms of international and domestic power relationships, institutional and structural economic rigidities, and the resulting proliferation of dual economies and dual societies both within and among the nations of the world. Dependence theories tended to emphasize external and internal institutional and political constraints on economic development. Emphasis was placed on the need for major new policies to eradicate poverty, to provide more diversified employment opportunities, and to reduce income inequal- ities. These and other egalitarian objectives were to be achieved within the context of a growing economy, but economic growth per se was not given the exalted status accorded to it by the linear-stages and structural-change models.

Throughout much of the 1980s and 1990s, a fourth approach prevailed. This neoclassical (sometimes called neoliberal) counterrevolution in economic thought emphasized the beneficial role of free markets, open economies, and the privati- zation of inefficient public enterprises. Failure to develop, according to this theory, was not due to exploitive external and internal forces as expounded by depen- dence theorists. Rather, it was primarily the result of too much government inter- vention and regulation of the economy. Today’s eclectic approach draws on all of these perspectives, and we will highlight the strengths and weaknesses of each.

3.2 Development as Growth and the Linear-Stages Theories

When interest in the poor nations of the world really began to materialize following World War II, economists in the industrialized nations were caught

120 PART ONE Principles and Concepts

off guard. They had no readily available conceptual apparatus with which to analyze the process of economic growth in largely agrarian societies that lacked modern economic structures. But they did have the recent experience of the Marshall Plan, under which massive amounts of U.S. financial and tech- nical assistance enabled the war-torn countries of Europe to rebuild and mod- ernize their economies in a matter of years. Moreover, was it not true that all modern industrial nations were once undeveloped agrarian societies? Surely their historical experience in transforming their economies from poor agricul- tural subsistence societies to modern industrial giants had important lessons for the “backward” countries of Asia, Africa, and Latin America. The logic and simplicity of these two strands of thought—the utility of massive injections of capital and the historical experience of the now developed countries—was too irresistible to be refuted by scholars, politicians, and administrators in rich countries, to whom people and ways of life in the developing world were often no more real than UN statistics or scattered chapters in anthropology books. Because of its emphasis on the central role of accelerated capital accu- mulation, this approach is often dubbed “capital fundamentalism.”

Rostow’s Stages of Growth

The most influential and outspoken advocate of the stages-of-growth model of development was the American economic historian Walt W. Rostow. According to Rostow, the transition from underdevelopment to development can be described in terms of a series of steps or stages through which all coun- tries must proceed. As Rostow wrote in the opening chapter of The Stages of Economic Growth:

This book presents an economic historian’s way of generalizing the sweep of modern history.…It is possible to identify all societies, in their economic dimen- sions, as lying within one of five categories: the traditional society, the pre- conditions for takeoff into self-sustaining growth, the take-off, the drive to maturity, and the age of high mass consumption.…These stages are not merely descriptive. They are not merely a way of generalizing certain factual observations about the sequence of development of modern societies. They have an inner logic and continuity.…They constitute, in the end, both a theory about economic growth and a more general, if still highly partial, theory about modern history as a whole.1

The advanced countries, it was argued, had all passed the stage of “takeoff into self-sustaining growth,” and the underdeveloped countries that were still in either the traditional society or the “preconditions” stage had only to follow a certain set of rules of development to take off in their turn into self- sustaining economic growth.

One of the principal strategies of development necessary for any takeoff was the mobilization of domestic and foreign saving in order to generate suf- ficient investment to accelerate economic growth. The economic mechanism by which more investment leads to more growth can be described in terms of the Harrod-Domar growth model,2 today often referred to as the AK model because it is based on a linear production function with output given by the capital stock K times a constant, often labeled A. In one form or another, it has frequently been applied to policy issues facing developing countries, such as in the two-gap model examined in Chapter 14.

Stages-of-growth model of development A theory of economic development, associated with the American economic historian Walt W. Rostow, according to which a country passes through sequential stages in achieving development.

Harrod-Domar growth model A functional economic rela- tionship in which the growth rate of gross domestic product (g) depends directly on the national net savings rate (s) and inversely on the national capital-output ratio (c).

121CHAPTER 3 Classic Theories of Economic Growth and Development

The Harrod-Domar Growth Model

Every economy must save a certain proportion of its national income, if only to replace worn-out or impaired capital goods (buildings, equipment, and materials). However, in order to grow, new investments representing net addi- tions to the capital stock are necessary. If we assume that there is some direct economic relationship between the size of the total capital stock, K, and total GDP, Y—for example, if $3 of capital is always necessary to produce an annual $1 stream of GDP—it follows that any net additions to the capital stock in the form of new investment will bring about corresponding increases in the flow of national output, GDP.

Suppose that this relationship, known in economics as the capital-output ratio, is roughly 3 to 1. If we define the capital-output ratio as k and assume further that the national net savings ratio, s, is a fixed proportion of national output (e.g., 6%) and that total new investment is determined by the level of total savings, we can construct the following simple model of economic growth:

1. Net saving (S) is some proportion, s, of national income (Y) such that we have the simple equation

S = sY (3.1)

2. Net investment (I) is defined as the change in the capital stock, K, and can be represented by ΔK such that

I = ΔK (3.2)

But because the total capital stock, K, bears a direct relationship to total national income or output, Y, as expressed by the capital-output ratio, c,3

it follows that K Y

= c

or ΔK ΔY

= c

or, finally,

ΔK = cΔY (3.3)

1/c is a measure of the efficiency of capital utilization.

3. Finally, because net national savings, S, must equal net investment, I, we can write this equality as

S = I (3.4)

But from Equation 3.1 we know that S = sY, and from Equations 3.2 and 3.3 we know that

I = ΔK = cΔY

It therefore follows that we can write the “identity” of saving equaling investment shown by Equation 3.4 as

S = sY = cΔY = ΔK = I (3.5)

Capital-output ratio A ratio that shows the units of capital required to produce a unit of output over a given period of time. Net savings ratio Savings expressed as a proportion of disposable income over some period of time.

122 PART ONE Principles and Concepts

or simply as

sY = c∆Y (3.6)

Dividing both sides of Equation 3.6 first by Y and then by c, we obtain the following expression:

∆Y Y

= s c

(3.7)

Note that the left-hand side of Equation 3.7, ∆Y>Y, represents the rate of change or rate of growth of GDP.

Equation 3.7, which is a simplified version of the famous equation in the Harrod-Domar theory of economic growth, states simply that the rate of growth of GDP (∆Y>Y) is determined jointly by the net national savings ratio, s, and the national capital-output ratio, c. More specifically, it says that in the absence of government, the growth rate of national income will be directly or positively related to the savings ratio (i.e., the more an economy is able to save—and invest—out of a given GDP, the greater the growth of that GDP will be) and inversely or negatively related to the economy’s capital-output ratio (i.e., the higher c is, the lower the rate of GDP growth will be). Equation 3.7 is also often expressed in terms of gross savings, sG, in which case the growth rate is given by

∆Y Y

= sG

c - δ (3.7’)

where δ is the rate of capital depreciation.4

The economic logic of Equations 3.7 and 3.7’ is very simple. To grow, economies must save and invest a certain proportion of their GDP. The more they can save and invest, the faster they can grow. But the actual rate at which they can grow for any level of saving and investment—how much additional output can be had from an additional unit of investment—can be measured by the inverse of the capital-output ratio, c, because this inverse, 1>c, is simply the output-capital or output-investment ratio. It follows that multiplying the rate of new investment, s = I>Y, by its productivity, 1>c, will give the rate by which national income or GDP will increase.

In addition to investment, two other components of economic growth are labor force growth and technological progress. The roles and functioning of these three components are examined in detail in Appendix 3.1. In the context of the Harrod-Domar growth model, labor force growth is not described explicitly. This is because labor is assumed to be abundant in a developing- country context and can be hired as needed in a given proportion to capital investments (this assumption is not always valid). In a general way, techno- logical progress can be expressed in the Harrod-Domar context as a decrease in the required capital-output ratio, giving more growth for a given level of investment, as follows from Equation 3.7 or 3.7’. This is obvious when we realize that in the longer run, this ratio is not fixed but can change over time in response to the functioning of financial markets and the policy environment. But again, the focus was on the role of capital investment.

123CHAPTER 3 Classic Theories of Economic Growth and Development

Obstacles and Constraints

Returning to the stages-of-growth theories and using Equation 3.7 of our simple Harrod-Domar growth model, we learn that one of the most funda- mental strategies of economic growth is simply to increase the proportion of national income saved (i.e., not consumed). If we can raise s in Equation 3.7, we can increase ∆Y>Y, the rate of GDP growth. For example, if we assume that the national capital-output ratio in some less developed country is, say, 3 and the aggregate net saving ratio is 6% of GDP, it follows from Equation 3.7 that this country can grow at a rate of 2% per year because

∆Y Y

= s c =

6% 3

= 2% (3.8)

Now if the national net savings rate can somehow be increased from 6% to, say, 15%—through some combination of increased taxes, foreign aid, and general consumption sacrifices—GDP growth can be increased from 2% to 5% because now

∆Y Y

= s c =

15% 3

= 5% (3.9)

In fact, Rostow and others defined the takeoff stage in precisely this way. Countries that were able to save 15 to 20% of GDP could grow (“develop”) at a much faster rate than those that saved less. Moreover, this growth would then be self-sustaining. The mechanisms of economic growth and development, therefore, would be simply a matter of increasing national savings and investment.

The main obstacle to or constraint on development, according to this theory, is the relatively low level of new capital formation in most poor coun- tries. But if a country wanted to grow at, say, a rate of 7% per year and if it could not generate savings and investment at a rate of 21% of national income (assuming that c, the final aggregate capital-output ratio, is 3) but could only manage to save 15%, it could seek to fill this “savings gap” of 6% through either foreign aid or private foreign investment.

Thus, the “capital constraint” stages approach to growth and development became a rationale and (in terms of Cold War politics) an opportunistic tool for justifying massive transfers of capital and technical assistance from the developed to the less developed nations. It was to be the Marshall Plan all over again, but this time for the nations of the developing world.

Necessary versus Sufficient Conditions: Some Criticisms of the Stages Model

Unfortunately, the mechanisms of development embodied in the theory of stages of growth did not always work. And the basic reason they didn’t work was not because more saving and investment isn’t a necessary condition for accelerated rates of economic growth but rather because it is not a sufficient condition. The Marshall Plan worked for Europe because the European countries receiving aid possessed the necessary structural, institutional, and

Necessary condition A con- dition that must be present, although it need not be in itself sufficient, for an event to occur. For example, capital formation may be a necessary condition for sustained eco- nomic growth (before growth in output can occur, there must be tools to produce it). But for this growth to continue, social, institutional, and attitudinal changes may have to occur.

Sufficient condition A condition that when present causes or guarantees that an event will or can occur; in economic models, a condition that logically requires that a statement must be true (or a result must hold) given other assumptions.

124 PART ONE Principles and Concepts

attitudinal conditions (e.g., well-integrated commodity and money markets, highly developed transport facilities, a well-trained and educated workforce, the motivation to succeed, an efficient government bureaucracy) to convert new capital effectively into higher levels of output. The Rostow and Harrod- Domar models implicitly assume the existence of these same attitudes and arrangements in underdeveloped nations. Yet, in many cases, they are lacking, as are complementary factors such as managerial competence, skilled labor, and the ability to plan and administer a wide assortment of development projects. There was also insufficient focus on another strategy for raising growth that is apparent from Equation 3.7: reducing the capital-output ratio, c, which entails increasing the efficiency with which investments generate extra output—a theme we take up later.

3.3 Structural-Change Models

Structural-change theory focuses on the mechanism by which underde- veloped economies transform their domestic economic structures from a heavy emphasis on traditional subsistence agriculture to a more modern, more urbanized, and more industrially diverse manufacturing and service economy. It employs the tools of neoclassical price and resource allocation theory and modern econometrics to describe how this transformation process takes place. Two well-known representative examples of the structural-change approach are the “two-sector surplus labor” theoretical model of W. Arthur Lewis and the “patterns of development” empirical analysis of Hollis B. Chenery and his coauthors.

The Lewis Theory of Economic Development

Basic Model One of the best-known early theoretical models of development that focused on the structural transformation of a primarily subsistence economy was that formulated by Nobel laureate W. Arthur Lewis in the mid-1950s and later modified, formalized, and extended by John Fei and Gustav Ranis.5 The Lewis two-sector model became the general theory of the development process in surplus-labor developing nations during most of the 1960s and early 1970s, and it is sometimes still applied, particularly to study the recent growth expe- rience in China and labor markets in other developing countries.6

In the Lewis model, the underdeveloped economy consists of two sectors: a traditional, overpopulated, rural subsistence sector characterized by zero marginal labor productivity—a situation that permits Lewis to classify this as surplus labor in the sense that it can be withdrawn from the traditional agri- cultural sector without any loss of output—and a high-productivity modern, urban industrial sector into which labor from the subsistence sector is grad- ually transferred. The primary focus of the model is on both the process of labor transfer and the growth of output and employment in the modern sector. (The modern sector could include modern agriculture, but we will call the sector “industrial” as a shorthand). Both labor transfer and modern- sector employment growth are brought about by output expansion in that sector. The speed with which this expansion occurs is determined by the rate

Structural-change theory The hypothesis that underde- velopment is due to underuti- lization of resources arising from structural or institutional factors that have their origins in both domestic and interna- tional dualism. Development therefore requires more than just accelerated capital for- mation.

Structural transformation The process of transforming an economy in such a way that the contribution to national income by the manufacturing sector eventually surpasses the contribution by the agricultural sector. More generally, a major alteration in the industrial composition of any economy.

Lewis two-sector model A theory of development in which surplus labor from the traditional agricultural sector is transferred to the modern industrial sector, the growth of which absorbs the surplus labor, promotes industrial- ization, and stimulates sus- tained development.

Surplus labor The excess supply of labor over and above the quantity demanded at the going free-market wage rate. In the Lewis two-sector model of economic devel- opment, surplus labor refers to the portion of the rural labor force whose marginal produc- tivity is zero or negative.

125CHAPTER 3 Classic Theories of Economic Growth and Development

of industrial investment and capital accumulation in the modern sector. Such investment is made possible by the excess of modern-sector profits over wages on the assumption that capitalists reinvest all their profits. Finally, Lewis assumed that the level of wages in the urban industrial sector was constant, determined as a given premium over a fixed average subsistence level of wages in the traditional agricultural sector. At the constant urban wage, the supply curve of rural labor to the modern sector is considered to be perfectly elastic.

We can illustrate the Lewis model of modern-sector growth in a two- sector economy by using Figure 3.1. Consider first the traditional agricultural sector portrayed in the two right-hand diagrams of Figure 3.1b. The upper diagram shows how subsistence food production varies with increases in labor inputs. It is a typical agricultural production function in which the total output or product (TPA) of food is determined by changes in the amount of

Production function A tech- nological or engineering rela- tionship between the quantity of a good produced and the quantity of inputs required to produce it.

FIGURE 3.1 The Lewis Model of Modern-Sector Growth in a Two-Sector Surplus-Labor Economy

L 1

KM 3 > KM 2 > KM 1

D1 (KM 1) = MPLM

D1

D2

D3

MPLA

MPLA

APLA

APLA

LA

D2 (KM 2)

D3 (KM 3)

TPM (KM1)

TPM (KM 2)

TPM (KM 3)

TPA (KA) TPA

TPM 3

TPM 2

TPM 1

WM SL

WA WA

L 2 L 3

L1 L 2

F G H

L 3

QLM LA

LA = WA

QLA

TPA

0

0 0

0

T o

ta l p

ro d

u ct

( m

a n

u fa

ct u

re s)

R e

a l w

a g

e (

= M

P L

M )

Quantity of labor (QLM) (thousands) Quantity of labor (QLA) (millions)

T o

ta l p

ro d

u ct

( fo

o d

) A

ve ra

g e

( m

a rg

in a

l) p

ro d

u ct

s TPM = f(LM, KM, tM) KM 3 > KM 2 > KM1

–

–

– TPA = f (LA, KA, tA)

– –

(a) Modern (industrial) sector (b) Traditional (agricultural) sector

Surplus labor

126 PART ONE Principles and Concepts

the only variable input, labor (LA), given a fixed quantity of capital, KA , and unchanging traditional technology, tA. In the lower-right diagram, we have the average and marginal product of labor curves, APLA and MPLA, which are derived from the total product curve shown immediately above. The quantity of agricultural labor (QLA) available is the same on both horizontal axes of the right-hand side of the figure and is expressed in millions of workers, as Lewis is describing an underdeveloped economy where much of the population lives and works in rural areas.

Lewis makes two assumptions about the traditional sector. First, there is surplus labor in the sense that MPLA is zero, and second, all rural workers share equally in the output so that the rural real wage is determined by the average and not the marginal product of labor (as will be the case in the modern sector). Metaphorically, this may be thought of as passing around the family rice bowl at dinnertime, from which each person takes an equal share (this need not be literally equal shares for the basic idea to hold). Assume that there are LA agri- cultural workers producing TPA food, which is shared equally as WA food per person (this is the average product, which is equal to TPA>LA). The marginal product of these LA workers is zero, as shown in the bottom diagram of Figure 3.1b; hence the surplus-labor assumption applies to all workers in excess of LA (note the horizontal TPA curve beyond LA workers in the upper-right diagram).

The upper-left diagram of Figure 3.1a portrays the total product (production function) curves for the modern industrial sector. Once again, output of, say, manufactured goods (TPM) is a function of a variable labor input, LM, for a given capital stock KM and technology, tM. On the horizontal axes, the quantity of labor employed to produce an output of, say, TPM1, with capital stock KM1, is expressed in thousands of urban workers, L1. In the Lewis model, the modern- sector capital stock is allowed to increase from KM1 to KM2 to KM3 as a result of the reinvestment of profits by industrial capitalists. This will cause the total product curves in Figure 3.1a to shift upward from TPM(KM1) to TPM(KM2) to TPM(KM3). The process that will generate these capitalist profits for reinvestment and growth is illustrated in the lower-left diagram of Figure 3.1a. Here we have modern-sector marginal labor product curves derived from the TPM curves of the upper diagram. Under the assumption of perfectly competitive labor markets in the modern sector, these marginal product of labor curves are in fact the actual demand curves for labor. Here is how the system works.

WA in the lower diagrams of Figures 3.1a and 3.1b represents the average level of real subsistence income in the traditional rural sector. WM in Figure 3.1a is therefore the real wage in the modern capitalist sector. At this wage, the supply of rural labor is assumed to be unlimited or perfectly elastic, as shown by the horizontal labor supply curve WMSL. In other words, Lewis assumes that at urban wage WM above rural average income WA, modern-sector employers can hire as many surplus rural workers as they want without fear of rising wages. (Note again that the quantity of labor in the rural sector, Figure 3.1b, is expressed in millions, whereas in the modern urban sector, Figure 3.1a, units of labor are expressed in thousands.) Given a fixed supply of capital KM1 in the initial stage of modern-sector growth, the demand curve for labor is deter- mined by labor’s declining marginal product and is shown by the negatively sloped curve D1(KM1) in the lower-left diagram. Because profit-maximizing modern-sector employers are assumed to hire laborers to the point where their

Average product Total output or product divided by total factor input (e.g., the average product of labor is equal to total output divided by the total amount of labor used to produce that output).

Marginal product The increase in total output resulting from the use of one additional unit of a variable factor of production (such as labor or capital). In the Lewis two-sector model, surplus labor is defined as workers whose marginal product is zero.

127CHAPTER 3 Classic Theories of Economic Growth and Development

marginal physical product is equal to the real wage (i.e., the point F of inter- section between the labor demand and supply curves), total modern-sector employment will be equal to L1. Total modern-sector output, TPM1, would be given by the area bounded by points 0D1FL1. The share of this total output paid to workers in the form of wages would be equal, therefore, to the area of the rectangle 0WMFL1. The balance of the output shown by the area WMD1F would be the total profits that accrue to the capitalists. Because Lewis assumes that all of these profits are reinvested, the total capital stock in the modern sector will rise from KM1 to KM2. This larger capital stock causes the total product curve of the modern sector to shift to TPM(KM2), which in turn induces a rise in the mar- ginal product demand curve for labor. This outward shift in the labor demand curve is shown by line D2(KM2) in the bottom half of Figure 3.1a. A new equi- librium modern-sector employment level will be established at point G with L2 workers now employed. Total output rises to TPM2 or 0D2GL2, while total wages and profits increase to 0WMGL2 and WMD2G, respectively. Once again, these larger (WMD2G) profits are reinvested, increasing the total capital stock to KM3, shifting the total product and labor demand curves to TPM(KM3) and to D3(KM3), respectively, and raising the level of modern-sector employment to L3.

This process of modern-sector self-sustaining growth and employment expansion is assumed to continue until all surplus rural labor is absorbed in the new industrial sector. Thereafter, additional workers can be withdrawn from the agricultural sector only at a higher cost of lost food production because the declining labor-to-land ratio means that the marginal product of rural labor is no longer zero. This is known as the “Lewis turning point.” Thus, the labor supply curve becomes positively sloped as modern-sector wages and employment continue to grow. The structural transformation of the economy will have taken place, with the balance of economic activity shifting from tra- ditional rural agriculture to modern urban industry.

Criticisms of the Lewis Model Although the Lewis two-sector development model is simple and roughly reflects the historical experience of economic growth in the West, four of its key assumptions do not fit the institutional and economic realities of most contemporary developing countries.

First, the model implicitly assumes that the rate of labor transfer and employment creation in the modern sector is proportional to the rate of modern-sector capital accumulation. The faster the rate of capital accumu- lation, the higher the growth rate of the modern sector and the faster the rate of new job creation. But what if capitalist profits are reinvested in more sophisticated laborsaving capital equipment rather than just duplicating the existing capital, as is implicitly assumed in the Lewis model? (We are, of course, here accepting the debatable assumption that capitalist profits are in fact reinvested in the local economy and not sent abroad as a form of “capital flight” to be added to the deposits of Western banks.) Figure 3.2 reproduces the lower, modern-sector diagram of Figure 3.1a, only this time the labor demand curves do not shift uniformly outward but in fact cross. Demand curve D2(KM2) has a greater negative slope than D2(KM1) to reflect the fact that additions to the capital stock embody laborsaving technical progress— that is, KM2 technology requires much less labor per unit of output than KM1 technology does.

Self-sustaining growth Economic growth that continues over the long run based on saving, investment, and complementary private and public activities.

128 PART ONE Principles and Concepts

We see that even though total output has grown substantially (i.e., 0D2EL1 is significantly greater than 0D1EL1), total wages (0WMEL1) and employment (L1) remain unchanged. All of the extra output accrues to capitalists in the form of profits. Figure 3.2 therefore provides an illustration of what some might call “antidevelopmental” economic growth—all the extra income and output growth are distributed to the few owners of capital, while income and employment levels for the masses of workers remain largely unchanged. Although total GDP would rise, there would be little or no improvement in aggregate social welfare measured, say, in terms of more widely distributed gains in income and employment.

The second questionable assumption of the Lewis model is the notion that surplus labor exists in rural areas while there is full employment in the urban areas. Most contemporary research indicates that there is little surplus labor in rural locations. True, there are both seasonal and geographic exceptions to this rule (e.g., at least until recently in parts of China and the Asian subcontinent, some Caribbean islands, and isolated regions of Latin America where land ownership is very unequal), but by and large, development economists today agree that Lewis’s assumption of rural surplus labor is generally not valid.

The third dubious assumption is the notion of a competitive modern-sector labor market that guarantees the continued existence of constant real urban wages up to the point where the supply of rural surplus labor is exhausted. Prior to the 1980s, a striking feature of urban labor markets and wage deter- mination in almost all developing countries was the tendency for these wages to rise substantially over time, both in absolute terms and relative to average rural incomes, even in the presence of rising levels of open modern-sector unemployment and low or zero marginal productivity in agriculture. Institu- tional factors such as union bargaining power, civil service wage scales, and

FIGURE 3.2 The Lewis Model Modified by Laborsaving Capital Accumulation: Employment Implications

D1

L1

D2

E S

A D1 (KM1)

WM

D2 (KM2)

Quantity of labor

0

R e

a l w

a g

e (

= M

P L

M )

129CHAPTER 3 Classic Theories of Economic Growth and Development

multinational corporations’ hiring practices tend to negate competitive forces in modern-sector labor markets in developing countries.

The fourth concern with the Lewis model is its assumption of dimin- ishing returns in the modern industrial sector. Yet there is much evidence that increasing returns prevail in that sector, posing special problems for devel- opment policymaking that we will examine in Chapter 4.

We study the Lewis model because, as many development specialists still think about development in this way either explicitly or implicitly, it helps stu- dents participate in the debates. Moreover, the model is widely considered rel- evant to recent experiences in China, where labor has been steadily absorbed from farming into manufacturing, and to a few other countries with similar growth patterns. The Lewis turning point at which wages in manufacturing start to rise was widely identified with China’s wage increases starting in 2010 (see the case study for Chapter 4).

However, when we take into account the laborsaving bias of most modern technological transfer, the existence of substantial capital flight, the wide- spread nonexistence of rural surplus labor, the growing prevalence of urban surplus labor, and the tendency for modern-sector wages to rise rapidly even where substantial open unemployment exists, we must acknowledge that the Lewis two-sector model—though valuable as an early conceptual portrayal of the development process of sectoral interaction and structural change and a description of some historical experiences, including some recent ones such as China—requires considerable modification in assumptions and analysis to fit the reality of most contemporary developing nations.

Structural Change and Patterns of Development

Like the earlier Lewis model, the patterns-of-development analysis of struc- tural change focuses on the sequential process through which the economic, industrial, and institutional structure of an underdeveloped economy is trans- formed over time to permit new industries to replace traditional agriculture as the engine of economic growth. However, in contrast to the Lewis model and the original stages view of development, increased savings and investment are perceived by patterns-of-development analysts as necessary but not suf- ficient conditions for economic growth. In addition to the accumulation of capital, both physical and human, a set of interrelated changes in the economic structure of a country are required for the transition from a traditional eco- nomic system to a modern one. These structural changes involve virtually all economic functions, including the transformation of production and changes in the composition of consumer demand, international trade, and resource use as well as changes in socioeconomic factors such as urbanization and the growth and distribution of a country’s population.

Empirical structural-change analysts emphasize both domestic and inter- national constraints on development. The domestic ones include economic constraints such as a country’s resource endowment and its physical and popu- lation size, as well as institutional constraints such as government policies and objectives. International constraints on development include access to external capital, technology, and international trade. Differences in development level among developing countries are largely ascribed to these domestic and

Patterns-of-development analysis An attempt to identify characteristic fea- tures of the internal process of structural transformation that a “typical” developing economy undergoes as it gen- erates and sustains modern economic growth and devel- opment.

130 PART ONE Principles and Concepts

international constraints. However, it is the international constraints that make the transition of currently developing countries differ from that of now industrialized countries. To the extent that developing countries have access to the opportunities presented by the industrial countries as sources of capital, technology, and manufactured imports, as well as markets for exports, they can make the transition at an even faster rate than that achieved by the indus- trial countries during the early periods of their economic development. Thus, unlike the earlier stages model, the structural-change model recognizes the fact that developing countries are part of an integrated international system that can promote (as well as hinder) their development.

The best-known model of structural change is the one based largely on the empirical work of the late economist Hollis B. Chenery and his colleagues, who examined patterns of development for numerous developing countries during the postwar period. (This approach also built on research by Nobel laureate Simon Kuznets on modern economic growth of developed countries.)7 Their empirical studies, both cross-sectional (among countries at a given point in time) and time-series (over long periods of time), of countries at different levels of per capita income led to the identification of several characteristic features of the development process. These included the shift from agricultural to industrial production, the steady accumulation of physical and human capital, the change in consumer demands from emphasis on food and basic necessities to desires for diverse manufactured goods and services, the growth of cities and urban indus- tries as people migrate from farms and small towns, and the decline in family size and overall population growth as children lose their economic value and parents substitute what is traditionally labeled child quality (education) for quantity (see Chapter 6), with population growth first increasing and then decreasing in the process of development. Proponents of this school often call for development specialists to “let the facts speak for themselves” rather than get bogged down in the arcana of theories such as the stages of growth. This is a valuable counter- balance to empty theorizing, but it also has its own limits.

Conclusions and Implications

The structural changes that we have described are the “average” patterns of development that Chenery and his colleagues observed among countries in time-series and cross-sectional analyses. The major hypothesis of the structural- change model is that development is an identifiable process of growth and change, whose main features are similar in all countries. However, as mentioned earlier, the model does recognize that differences can arise among countries in the pace and pattern of development, depending on their particular set of circum- stances. Factors influencing the development process include a country’s resource endowment and size, its government’s policies and objectives, the availability of external capital and technology, and the international trade environment.

One limitation to keep in mind is that by emphasizing patterns rather than theory, this approach runs the risk of leading practitioners to draw the wrong conclusions about causality—in effect, to “put the cart before the horse.” Observing developed-country patterns such as the decline of the share of the labor force in agriculture over time, many developing-country policymakers have been inclined to neglect that vital sector. But as you will see in Chapter 9,

131CHAPTER 3 Classic Theories of Economic Growth and Development

that is precisely the opposite conclusion to the one that should be drawn. Observing the important role of higher education in developed countries, policymakers may be inclined to emphasize the development of an advanced university system even before a majority of the population has gained basic literacy, a policy that has led to gross inequities even in countries at least nom- inally committed to egalitarian outcomes, such as Tanzania.

Empirical studies on the process of structural change lead to the conclusion that the pace and pattern of development can vary according to both domestic and international factors, many of which lie beyond the control of an indi- vidual developing nation. Yet despite this variation, structural-change econo- mists argue that one can identify certain patterns occurring in almost all coun- tries during the development process. And these patterns, they argue, may be affected by the choice of development policies pursued by governments in developing countries as well as the international trade and foreign-assistance policies of developed nations. Hence, structural-change analysts are basically optimistic that the “correct” mix of economic policies will generate beneficial patterns of self-sustaining growth. The international-dependence school to which we now turn is, in contrast, much less sanguine and is in many cases downright pessimistic.

3.4 The International-Dependence Revolution

During the 1970s, international-dependence models gained increasing support, especially among developing-country intellectuals, as a result of growing disen- chantment with both the stages and structural-change models. While this theory to a large degree went out of favor during the 1980s and 1990s, versions of it have enjoyed a resurgence in the twenty-first century as some of its views have been adopted, albeit in modified form, by theorists and leaders of the antiglo- balization movement.8 Essentially, international-dependence models view devel- oping countries as beset by institutional, political, and economic rigidities, both domestic and international, and caught up in a dependence and dominance relationship with rich countries. Within this general approach are three major streams of thought: the neocolonial dependence model, the false-paradigm model, and the dualistic-development thesis.

The Neocolonial Dependence Model

The first major stream, which we call the neocolonial dependence model, is an indirect outgrowth of Marxist thinking. It attributes the existence and con- tinuance of underdevelopment primarily to the historical evolution of a highly unequal international capitalist system of rich country–poor country relation- ships. Whether because rich nations are intentionally exploitative or uninten- tionally neglectful, the coexistence of rich and poor nations in an international system dominated by such unequal power relationships between the center (the developed countries) and the periphery (the developing countries) renders attempts by poor nations to be self-reliant and independent difficult and some- times even impossible.9 Certain groups in the developing countries (including landlords, entrepreneurs, military rulers, merchants, salaried public officials, and

Dependence The reliance of developing countries on developed-country economic policies to stimulate their own economic growth. Dependence can also mean that the developing countries adopt developed-country education systems, technology, economic and political systems, attitudes, consumption pat- terns, dress, and so on.

Dominance In international affairs, a situation in which the developed countries have much greater power than the less developed countries in decisions affecting important international economic issues, such as the prices of agricul- tural commodities and raw materials in world markets.

Neocolonial dependence model A model whose main proposition is that underdevelopment exists in developing countries because of continuing exploitative eco- nomic, political, and cultural policies of former colonial rulers toward less developed countries.

Underdevelopment An economic situation charac- terized by persistent low levels of living in conjunction with absolute poverty, low income per capita, low rates of economic growth, low con- sumption levels, poor health services, high death rates, high birth rates, dependence on foreign economies, and limited freedom to choose among activities that satisfy human wants.

Center In dependence theory, the economically developed world.

Periphery In dependence theory, the developing countries.

132 PART ONE Principles and Concepts

trade union leaders) that enjoy high incomes, social status, and political power constitute a small elite ruling class whose principal interest, knowingly or not, is in the perpetuation of the international capitalist system of inequality and conformity in which they are rewarded. Directly and indirectly, they serve (are dominated by) and are rewarded by (are dependent on) international special- interest power groups, including multinational corporations, national bilateral- aid agencies, and multilateral assistance organizations like the World Bank or the International Monetary Fund (IMF), which are tied by allegiance or funding to the wealthy capitalist countries. The elites’ activities and viewpoints often serve to inhibit any genuine reform efforts that might benefit the wider population and in some cases actually lead to even lower levels of living and to the per- petuation of underdevelopment. In short, the neo-Marxist, neocolonial view of underdevelopment attributes a large part of the developing world’s continuing poverty to the existence and policies of the industrial capitalist countries of the northern hemisphere and their extensions in the form of small but powerful elite or comprador groups in the less developed countries.10 Underdevelopment is thus seen as an externally induced phenomenon, in contrast to the linear-stages and structural-change theories’ stress on internal constraints, such as insufficient savings and investment or lack of education and skills. Revolutionary struggles or at least major restructuring of the world capitalist system is therefore required to free dependent developing nations from the direct and indirect economic control of their developed-world and domestic oppressors.

One of the most forceful statements of the international-dependence school of thought was made by Theotonio Dos Santos:

Underdevelopment, far from constituting a state of backwardness prior to capitalism, is rather a consequence and a particular form of capitalist development known as dependent capitalism.…Dependence is a conditioning situation in which the econ- omies of one group of countries are conditioned by the development and expansion of others. A relationship of interdependence between two or more economies or between such economies and the world trading system becomes a dependent rela- tionship when some countries can expand through self-impulsion while others, being in a dependent position, can only expand as a reflection of the expansion of the dom- inant countries, which may have positive or negative effects on their immediate devel- opment. In either case, the basic situation of dependence causes these countries to be both backward and exploited. Dominant countries are endowed with technological, commercial, capital and sociopolitical predominance over dependent countries—the form of this predominance varying according to the particular historical moment— and can therefore exploit them, and extract part of the locally produced surplus. Dependence, then, is based upon an international division of labor which allows industrial development to take place in some countries while restricting it in others, whose growth is conditioned by and subjected to the power centers of the world.11

A similar but obviously non-Marxist perspective was expounded by Pope John Paul II in his widely quoted 1988 encyclical letter (a formal, elaborate expression of papal teaching) Sollicitudo rei socialis (The Social Concerns of the Church), in which he declared:

One must denounce the existence of economic, financial, and social mechanisms which, although they are manipulated by people, often function almost auto- matically, thus accentuating the situation of wealth for some and poverty for the rest. These mechanisms, which are maneuvered directly or indirectly by the more

Comprador group In depen- dence theory, local elites who act as fronts for foreign investors.

133CHAPTER 3 Classic Theories of Economic Growth and Development

developed countries, by their very functioning, favor the interests of the people manipulating them. But in the end they suffocate or condition the economies of the less developed countries.

The False-Paradigm Model

A second and less radical international-dependence approach to development, which we might call the false-paradigm model, attributes underdevelopment to faulty and inappropriate advice provided by well-meaning but often uninformed, biased, and ethnocentric international “expert” advisers from developed-country assistance agencies and multinational donor organizations. These experts are said to offer complex but ultimately misleading models of development that often lead to inappropriate or incorrect policies. Because of institutional factors such as the central and remarkably resilient role of tradi- tional social structures (tribe, caste, class, etc.), the highly unequal ownership of land and other property rights, the disproportionate control by local elites over domestic and international financial assets, and the very unequal access to credit, these policies, based as they often are on mainstream, neoclassical (or perhaps Lewis-type surplus-labor or Chenery-type structural-change) models, in many cases merely serve the vested interests of existing power groups, both domestic and international.

In addition, according to this argument, leading university intellectuals, trade unionists, high-level government economists, and other civil servants all get their training in developed-country institutions where they are unwit- tingly served an unhealthy dose of alien concepts and elegant but inappli- cable theoretical models. Having little or no really useful knowledge to enable them to come to grips in an effective way with real development problems, they often tend to become unknowing or reluctant apologists for the existing system of elitist policies and institutional structures. In university economics courses, for example, this typically entails the perpetuation of the teaching of many “irrelevant” Western concepts and models, while in government policy discussions, too much emphasis is placed on attempts to measure capital- output ratios, increase savings and investment ratios, privatize and deregulate the economy, or maximize GDP growth rates. As a result, proponents argue that desirable institutional and structural reforms, many of which we have discussed, are neglected or given only cursory attention.

The Dualistic-Development Thesis

Implicit in structural-change theories and explicit in international-dependence theories is the notion of a world of dual societies, of rich nations and poor nations and, in the developing countries, pockets of wealth within broad areas of poverty. Dualism is the existence and persistence of substantial and even increasing divergences between rich and poor nations and rich and poor peoples on various levels. Specifically, although research continues, the tradi- tional concept of dualism embraces four key arguments:12

1. Different sets of conditions, of which some are “superior” and others “inferior,” can coexist in a given space. Examples of this element of dualism

False-paradigm model The proposition that developing countries have failed to develop because their devel- opment strategies (usually given to them by Western economists) have been based on an incorrect model of development, one that, for example, overstresses capital accumulation or market liber- alization without giving due consideration to needed social and institutional change.

Dualism The coexis- tence of two situations or phenomena (one desirable and the other not) that are mutually exclusive to dif- ferent groups of society—for example, extreme poverty and affluence, modern and traditional economic sectors, growth and stagnation, and higher education among a few amid large-scale illiteracy.

134 PART ONE Principles and Concepts

include Lewis’s notion of the coexistence of modern and traditional methods of production in urban and rural sectors; the coexistence of wealthy, highly educated elites with masses of illiterate poor people; and the dependence notion of the coexistence of powerful and wealthy indus- trialized nations with weak, impoverished peasant societies in the interna- tional economy.

2. This coexistence is chronic and not merely transitional. It is not due to a temporary phenomenon, in which case, time could eliminate the dis- crepancy between superior and inferior elements. In other words, the in- ternational coexistence of wealth and poverty is not simply a historical phenomenon that will be rectified in time. Although both the stages-of- growth theory and the structural-change models implicitly make such an assumption, to proponents of the dualistic development thesis, growing international inequalities seem to refute it.

3. Not only do the degrees of superiority or inferiority fail to show any signs of diminishing, but they even have an inherent tendency to increase. For example, the productivity gap between workers in developed countries and their counterparts in most developing countries seems to widen.

4. The interrelations between the superior and inferior elements are such that the existence of the superior elements does little or nothing to pull up the inferior element, let alone “trickle down” to it. In fact, it may actually serve to push it down—to “develop its underdevelopment.”

Conclusions and Implications

Whatever their ideological differences, the advocates of the neocolonial- dependence, false-paradigm, and dualism models reject the exclusive emphasis on traditional neoclassical economic theories designed to accelerate the growth of GDP as the principal index of development. They question the validity of Lewis-type two-sector models of modernization and industrialization in light of their questionable assumptions and developing- world history. They further reject the claims made by Chenery and others that there are well-defined empirical patterns of development that should be pursued by most poor countries. Instead, dependence, false-paradigm, and dualism theorists place more emphasis on international power imbalances and on needed fundamental economic, political, and institutional reforms, both domestic and worldwide. In extreme cases, they call for the outright expropriation of privately owned assets in the expectation that public asset ownership and control will be a more effective means to help eradicate absolute poverty, provide expanded employment opportunities, lessen income inequalities, and raise the levels of living (including health, education, and cultural enrichment) of the masses. Although a few radical neo-Marxists would even go so far as to say that economic growth and structural change do not matter, the majority of thoughtful observers recognize that the most effective way to deal with these diverse social problems is to accelerate the pace of economic growth through domestic and international reforms, accompanied by a judicious mixture of both public and private economic activity.

135CHAPTER 3 Classic Theories of Economic Growth and Development

Dependence theories have two major weaknesses. First, although they offer an appealing explanation of why many poor countries remain underde- veloped, they give no insight into how countries initiate and sustain devel- opment. Second and perhaps more important, the actual economic experience of developing countries that have pursued revolutionary campaigns of indus- trial nationalization and state-run production has been mostly negative.

If we are to take dependence theory at face value, we would conclude that the best course for developing countries is to become entangled as little as possible with the developed countries and instead pursue a policy of autarky, or inwardly directed development, or at most trade only with other devel- oping countries. But large countries that embarked on autarkic policies, such as China and, to a significant extent, India, experienced stagnant growth and ultimately decided to open their economies, China beginning this process after 1978 and India, after 1990. At the opposite extreme, economies such as Taiwan and South Korea, and China more recently, which have most empha- sized exports to developed countries, have grown strongly. Although in many cases close ties to metropolitan countries during the colonial period appar- ently produced damaging outcomes—as in Peru under Spain, the Congo under Belgium, India under Great Britain, and West Africa under France— in a majority of cases, this relationship appeared to have significantly altered during the postcolonial period. Clearly, however, conflicts of interest between the developed and developing worlds, such as took center stage at the Copen- hagen climate summit in December 2009 and have played a role in recent WTO and G20 meetings, are genuine and cannot be ignored.

We next consider the view that the keys to development are found in free markets. For perspective, as will be noted in later chapters, governments can succeed or fail just as markets can; the key to successful development perfor- mance is achieving a careful balance among what government can success- fully accomplish, what the private market system can do, and what both can best do working together.

While the international-dependence revolution in development theory was capturing the imagination of many Western and developing country scholars, a reaction was emerging in the late 1970s and early 1980s in the form of a neo- classical free-market counterrevolution. This very different approach would ultimately dominate Western (and to a lesser extent developing country) the- ories of economic development during the 1980s and early 1990s.

3.5 The Neoclassical Counterrevolution: Market Fundamentalism

Challenging the Statist Model: Free Markets, Public Choice, and Market-Friendly Approaches

In the 1980s, the political ascendancy of conservative governments in the United States, Canada, Britain, and West Germany came with a neoclassical counterrevolution in economic theory and policy. In developed nations, this counterrevolution favored supply-side macroeconomic policies, rational expec- tations theories, and the privatization of public corporations. In developing

Autarky A closed economy that attempts to be completely self-reliant.

Neoclassical counterrevolution The 1980s resurgence of neoclassical free-market ori- entation toward development problems and policies, counter to the interventionist depen- dence revolution of the 1970s.

136 PART ONE Principles and Concepts

countries, it called for freer markets and the dismantling of public ownership, statist planning, and government regulation of economic activities. Neoclas- sicists obtained controlling votes on the boards of the world’s two most pow- erful international financial agencies—the World Bank and the International Monetary Fund. In conjunction and with the simultaneous erosion of influence of organizations such as the International Labor Organization (ILO), the United Nations Development Programme (UNDP), and the United Nations Conference on Trade and Development (UNCTAD), which more fully represent the views of delegates from developing countries, it was inevitable that the neoconservative, free-market challenge to the interventionist arguments of dependence theorists would gather momentum.

The central argument of the neoclassical counterrevolution is that under- development results from poor resource allocation due to incorrect pricing policies and too much state intervention by overly active developing-nation governments. Rather, the leading writers of the counterrevolution school, including Lord Peter Bauer, Deepak Lal, Ian Little, Harry Johnson, Bela Balassa, Jagdish Bhagwati, and Anne Krueger, argued that it is this very state intervention in economic activity that slows the pace of economic growth. The neoliberals argue that by permitting competitive free markets to flourish, privatizing state-owned enterprises, promoting free trade and export expansion, welcoming investors from developed countries, and eliminating the plethora of government regulations and price distortions in factor, product, and financial markets, both economic efficiency and economic growth will be stimulated. Contrary to the claims of the dependence theorists, the neo- classical counterrevolutionaries argue that the developing world is underde- veloped, not because of the predatory activities of the developed world and the international agencies that it controls, but rather because of the heavy hand of the state and the corruption, inefficiency, and lack of economic incentives that permeate the economies of developing nations. What is needed, therefore, is not a reform of the international economic system, a restructuring of dualistic developing economies, an increase in foreign aid, attempts to control popu- lation growth, or a more effective development planning system. Rather, it is simply a matter of promoting free markets and laissez-faire economics within the context of permissive governments that allow the “magic of the market- place” and the “invisible hand” of market prices to guide resource allocation and stimulate economic development. They point both to the success of econ- omies like South Korea, Taiwan, and Singapore as “free-market” examples (although, as we shall see later, these Asian Tigers are far from the laissez-faire neoconservative prototype) and to the failures of the public-interventionist economies of Africa and Latin America.13

The neoclassical counterrevolution can be divided into three component approaches: the free-market approach, the public-choice (or “new political economy”) approach, and the “market-friendly” approach. Free-market analysis argues that markets alone are efficient—product markets provide the best signals for investments in new activities; labor markets respond to these new industries in appropriate ways; producers know best what to produce and how to produce it efficiently; and product and factor prices reflect accurate scarcity values of goods and resources now and in the future. Competition is effective, if not perfect; technology is freely available and nearly costless to absorb; information

Free markets The system whereby prices of com- modities or services freely rise or fall when the buyer’s demand for them rises or falls or the seller’s supply of them decreases or increases.

Free-market analysis Theo- retical analysis of the prop- erties of an economic system operating with free markets, often under the assumption that an unregulated market performs better than one with government regulation.

137CHAPTER 3 Classic Theories of Economic Growth and Development

is also perfect and nearly costless to obtain. Under these circumstances, any gov- ernment intervention in the economy is by definition distortionary and counter- productive. Free-market development economists have tended to assume that developing-world markets are efficient and that whatever imperfections exist are of little consequence.

Public-choice theory, also known as the new political economy approach, goes even further to argue that governments can do (virtually) nothing right. This is because public-choice theory assumes that politicians, bureaucrats, citizens, and states act solely from a self-interested perspective, using their power and the authority of government for their own selfish ends. Citizens use political influence to obtain special benefits (called “rents”) from gov- ernment policies (e.g., import licenses or rationed foreign exchange) that restrict access to important resources. Politicians use government resources to consolidate and maintain positions of power and authority. Bureaucrats and public officials use their positions to extract bribes from rent-seeking cit- izens and to operate protected businesses on the side. Finally, states use their power to confiscate private property from individuals. The net result is not only a misallocation of resources but also a general reduction in individual freedoms. The conclusion, therefore, is that minimal government is the best government.14

The market-friendly approach is a variant on the neoclassical counter- revolution associated principally with the 1990s writings of the World Bank and its economists, many of whom were more in the free-market and public-choice camps during the 1980s.15 This approach recognizes that there are many imperfections in developing-country product and factor markets and that governments do have a key role to play in facilitating the oper- ation of markets through “nonselective” (market-friendly) interventions— for example, by investing in physical and social infrastructure, health care facilities, and educational institutions, and by providing a suitable climate for private enterprise. The market-friendly approach also differs from the free-market and public-choice schools of thought by accepting the notion that market failures (see Chapters 4 and 11) are more widespread in devel- oping countries in areas such as investment coordination and environmental outcomes. Moreover, phenomena such as missing and incomplete infor- mation, externalities in skill creation and learning, and economies of scale in production are also endemic to markets in developing countries. In fact, the recognition of these last three phenomena gives rise to newer schools of development theory, the endogenous growth approach, to which we turn in Appendix 3.3 at the end of this chapter, and the coordination failure approach, discussed in Chapter 4.

Traditional Neoclassical Growth Theory

Another cornerstone of the neoclassical free-market argument is the assertion that liberalization (opening up) of national markets draws additional domestic and foreign investment and thus increases the rate of capital accumulation. In terms of GDP growth, this is equivalent to raising domestic savings rates, which enhances capital-labor ratios and per capita incomes in capital-poor developing countries.

Public-choice theory (new political economy approach) The theory that self-interest guides all individual behavior and that governments are inefficient and corrupt because people use gov- ernment to pursue their own agendas.

Market-friendly approach The notion historically pro- mulgated by the World Bank that successful development policy requires governments to create an environment in which markets can operate efficiently and to intervene only selectively in the economy in areas where the market is inefficient.

Market failure A market’s inability to deliver its theo- retical benefits due to the exis- tence of market imperfections such as monopoly power, lack of factor mobility, sig- nificant externalities, or lack of knowledge. Market failure often provides the justification for government intervention to alter the working of the free market.

Capital-labor ratio The number of units of capital per unit of labor.

138 PART ONE Principles and Concepts

The Solow neoclassical growth model in particular represented the seminal contribution to the neoclassical theory of growth and later earned Robert Solow the Nobel Prize in economics.16 It differed from the Harrod-Domar formulation by adding a second factor, labor, and introducing a third independent variable, technology, to the growth equation. Unlike the fixed-coefficient, constant- returns-to-scale assumption of the Harrod-Domar model, Solow’s neoclassical growth model exhibited diminishing returns to labor and capital separately and constant returns to both factors jointly. Technological progress became the residual factor explaining long-term growth, and its level was assumed by Solow and other neoclassical growth theorists to be determined exogenously, that is, independently of all other factors in the model.

More formally, the standard exposition of the Solow neoclassical growth model uses an aggregate production function in which

Y = K α1AL21 -α (3.10)

where Y is gross domestic product, K is the stock of capital (which may include human capital as well as physical capital), L is labor, and A represents the productivity of labor, which grows at an exogenous rate. For developed countries, this rate has been estimated at about 2% per year. It may be smaller or larger for developing countries, depending on whether they are stagnating or catching up with the developed countries. Because the rate of techno- logical progress is given exogenously (at 2% per year, say), the Solow neo- classical model is sometimes called an “exogenous” growth model, to be con- trasted with the endogenous growth approach (discussed in Appendix 3.3). In Equation 3.10, α represents the elasticity of output with respect to capital (the percentage increase in GDP resulting from a 1% increase in human and physical capital). Since α is assumed to be less than 1 and private capital is assumed to be paid its marginal product so that there are no external econ- omies, this formulation of neoclassical growth theory yields diminishing returns both to capital and to labor.

The Solow neoclassical growth model implies that economies will con- verge to the same level of income per worker “conditionally”—that is, other things equal, particularly savings rates, depreciation, labor force growth, and productivity. The Solow neoclassical growth model is examined in detail in Appendix 3.2.

According to traditional neoclassical growth theory, output growth results from one or more of three factors: increases in labor quantity and quality (through population growth and education), increases in capital (through saving and investment), and improvements in technology (see Appendix 3.1). Closed economies (those with no external activities) with lower savings rates (other things being equal) grow more slowly in the short run than those with high savings rates and tend to converge to lower per capita income levels. Open economies (those with trade, foreign investment, etc.), however, experience income convergence at higher levels as capital flows from rich countries to poor countries where capital-labor ratios are lower and thus returns on investments are higher. Consequently, by impeding the inflow of foreign investment, the heavy-handedness of many developing countries’ governments, according to neoclassical growth theory, will retard growth

Closed economy An economy in which there are no foreign trade transactions or other economic contacts with the rest of the world.

Open economy An economy that practices foreign trade and has extensive financial and nonfinancial contacts with the rest of the world.

Solow neoclassical growth model Growth model in which there are diminishing returns to each factor of pro- duction but constant returns to scale. Exogenous techno- logical change generates long- term economic growth.

139CHAPTER 3 Classic Theories of Economic Growth and Development

in the economies of the developing world. In addition, openness is said to encourage greater access to foreign production ideas that can raise the rate of technological progress.

Conclusions and Implications

Whereas dependence theorists (many, but not all, of whom were economists from developing countries) saw underdevelopment as an externally induced phenomenon, neoclassical revisionists (most, but not all, of whom were Western economists) saw the problem as an internally induced phenomenon of developing countries, caused by too much government intervention and bad economic policies. Such finger-pointing on both sides is not uncommon in issues so contentious as those that divide rich and poor nations.

But what of the neoclassical counterrevolution’s contention that free markets and less government provide the basic ingredients for development? On strictly efficiency (as opposed to equity) criteria, there can be little doubt that market price allocation usually does a better job than state intervention. The problem is that many developing economies are so different in structure and organization from their Western counterparts that the behavioral assump- tions and policy precepts of traditional neoclassical theory are sometimes questionable and often incorrect. Competitive free markets generally do not exist, nor, given the institutional, cultural, and historical context of many developing countries, would they necessarily be desirable from a long-term economic and social perspective (see Chapter 11). Consumers as a whole are rarely sovereign about what goods and services are to be produced, in what quantities, and for whom. Information is limited, markets are fragmented, and much of the economy in low-income countries is still nonmonetized.17

There are widespread externalities of both production and consumption as well as discontinuities in production and indivisibilities (i.e., economies of scale) in technology. Producers, private or public, have great power in deter- mining market prices and quantities sold. The ideal of competition is typi- cally just that—an ideal with little substance in reality. Although monopolies of resource purchase and product sale are pervasive in the developing world, the traditional neoclassical theory of monopoly also offers little insight into the day-to-day activities of public and private corporations. Decision rules can vary widely with the social setting so that profit maximization may be a low- priority objective, especially in state-owned enterprises, in comparison with, say, the creation of jobs or the replacement of foreign managers with local personnel. Finally, the invisible hand often acts not to promote the general welfare but rather to lift up those who are already well-off while failing to offer opportunities for upward mobility for the vast majority.

Much can be learned from neoclassical theory with regard to the impor- tance of elementary supply-and-demand analysis in arriving at “correct” product, factor, and foreign-exchange prices for efficient production and resource allocation. However, enlightened governments can also make effective use of prices as signals and incentives for influencing socially optimal resource allocations. Indeed, we will often demonstrate the usefulness of various tools of neoclassical theory in our later analysis of problems such as population growth, agricultural stagnation, unemployment and underemployment,

140 PART ONE Principles and Concepts

child labor, educational demands, the environment, export promotion versus import substitution, devaluation, project planning, monetary policy, microfinance, and economic privatization. Nevertheless, the reality of the institutional and political structure of many developing-world economies— not to mention their differing value systems and ideologies—often makes the attainment of appropriate economic policies based either on markets or on enlightened public intervention an exceedingly difficult endeavor. In an environment of widespread institutional rigidity and severe socioeconomic inequality, both markets and governments will typically fail. It is not simply an either-or question based on ideological leaning; rather, it is a matter of assessing each individual country’s situation on a case-by-case basis. Devel- oping nations need to adopt local solutions in response to local constraints.18

Development economists must therefore be able to distinguish between textbook neoclassical theory and the institutional and political reality of con- temporary developing countries.19 They can then choose the traditional neo- classical concepts and models that can best illuminate issues and dilemmas of development and discard those that cannot. Approaches to making these distinctions and choices in key policy applications will feature centrally in Parts Two and Three.

3.6 Classic Theories of Development: Reconciling the Differences

In this chapter, we have reviewed a range of competing theories and approaches to the study of economic development. Each approach has its strengths and weaknesses. The fact that there exists such controversy—be it ideological, theoretical, or empirical—is what makes the study of economic development both challenging and exciting. Even more than other fields of economics, development economics has no universally accepted doctrine or paradigm. Instead, we have a continually evolving pattern of insights and under- standings, reflecting in part improved data and emergence of new technol- ogies and new institutions, that together provide the basis for examining the possibilities of contemporary development of the diverse nations of Africa, Asia, and Latin America.

You may wonder how consensus could emerge from so much disagreement. Although it is not implied here that such a consensus exists today or will ever emerge when such sharply conflicting values and ideologies prevail, we do suggest that something of significance can be gleaned from each of the four approaches that we have described. For example, the linear-stages model emphasizes the crucial role that saving and investment play in promoting sus- tainable long-run growth. The Lewis two-sector model of structural change underlines the importance of transfers of resources from low-productivity to high-productivity activities in the process of economic development, attempting to analyze the many linkages between traditional agriculture and modern industry, and clarifying recent growth experiences such as that of China. The empirical research of Chenery and his associates seeks to doc- ument precisely how economies undergo structural change while identifying

141CHAPTER 3 Classic Theories of Economic Growth and Development

the numerical values of key economic parameters involved in that process. The thoughts of international-dependence theorists alert us to the importance of the structure and workings of the world economy and the many ways in which decisions made in the developed world can affect the lives of millions of people in the developing world. Whether or not these activities are deliber- ately designed to maintain developing nations in a state of dependence is often beside the point. The very fact of their dependence and their vulnerability to key economic decisions made in the capitals of North America, western Europe, or Japan (not to mention those made by the IMF and the World Bank) forces us to recognize the importance of some of the insights of the interna- tional-dependence school. The same applies to arguments regarding the dual- istic structures and the role of ruling elites in the domestic economies of the developing world.

Although a good deal of conventional neoclassical economic theory needs to be modified to fit the unique social, institutional, and structural circumstances of developing nations, there is no doubt that promoting efficient production and distribution through a proper, functioning price system is an integral part of any successful development process. Many of the arguments of the neoclassical counterrevolutionaries, especially those related to the inefficiency of state-owned enterprises and the failures of development planning (see Chapter 11), and the harmful effects of government-induced domestic and international price distortions (see Chapters 7, 12, and 15), are as well taken as those of the dependence and structuralist schools. By contrast, the unquestioning exaltation of free markets and open economies along with the universal disparagement of public-sector lead- ership in promoting growth with equity in the developing world is open to serious challenge. As the chapters in Parts Two and Three reveal, successful development requires a skillful and judicious balancing of market pricing and promotion where markets can exist and operate efficiently, along with intelligent and equity-oriented government intervention in areas where unfettered market forces would lead to undesirable economic and social outcomes. Great strides have been made in modern development economic analysis in clarifying the logic of how well-formulated government policy can facilitate the development of markets and shared growth, as will be explained in Chapter 4.

In summary, each of the approaches to understanding development has something to offer. Their respective contributions will become clear later in the book when we explore in detail both the origins of and possible solutions to a wide range of problems such as poverty, population growth, unemployment, rural development, international trade, and the environment. They also inform contemporary models of development and underdevel- opment, to which we turn in the next chapter.

142 PART ONE Principles and Concepts

142

Case Study 3

Schools of Thought in Context: South Korea and Argentina

Acloser examination of two countries confirms the conclusion that each of the first four broad approaches to development—stages of growth, structural patterns of development, dependence, and neoclassical—provides important insights about development processes and policy. South Korea and Argentina are reasonably well matched for such a comparison; for example, both are midsize in pop- ulation (41 million in Argentina and 50 million in South Korea in 2011), and both were long classified as middle-income countries. But South Korea, now designated by the World Bank as a high-income country with about $31,000 PPP in 2008, has nearly double the per capita income of Argentina, with about $17,000 PPP in 2011, whereas 30 years earlier the reverse was true. Can the four classic approaches to development explain this reversal?

South Korea Stages of Growth South Korea confirms some linear-stages views, albeit in a limited way. Its share of investment in national income has been among the highest in the world, and this is a crucial part of the explanation of the nation’s rapid ascent. To under- stand just how rapid this ascent has been, consider that the country did not even rate a mention in Ros- tow’s Stages of Economic Growth in 1960, when the book was published, and few of the “preconditions for takeoff” were in place. Investment has been very high since then, but as a share of GNI, the investment ratio, at 15%, was still below takeoff levels in 1965. Yet it rose dramatically to 37% of GNI by 1990 and remained close to 40% in the 2000–2007 period (though the ratio has fallen in the last few years). Still, South Korea’s ascent has seemed to epitomize Ros- tow’s notion of an economy in the midst of a “drive to

maturity,” well on its way toward mastering the range of currently available technologies; and appears to be entering an “age of high mass consumption.”

Rostow claimed that maturity is attained some 60 years after takeoff begins, but he never denied unique experiences for each country, and it may well be that the gap between traditional and advanced tech- nology can actually be crossed more quickly at later stages of development. The larger the productivity gap is between countries, the quicker income can grow once takeoff has been achieved. South Korea certainly meets the “maturity” criterion of becoming integrated with the world economy through new types of exports and imports. Although the fact that India, rather than South Korea, was picked by Rostow for takeoff shows the limits of the predictive powers of the stages theory, the case of South Korea nonetheless offers some confirmation of their value.

Structural Patterns South Korea also confirms some patterns-of-development structural-change models. In particular, South Korea’s rise over the past gen- eration has been characterized by rapidly increasing agricultural productivity, shifts of labor from agri- culture to industry, the steady growth of the capital stock and of education and skills, and the demo- graphic transition from high to low fertility. These changes have occurred while South Korea’s per capita income has grown by more than 7% annually for the whole 1965–1990 period. Even in the 1990– 2002 period, as a more mature economy and in the face of the Asian financial crisis of 1997–1998, the economy grew at a 5.8% rate. In 2002–2011, it grew at less than 4% on average, still substantially higher than most other high-income countries. In the late 1940s and 1950s, South Korea carried out a thorough- going land reform, so agriculture was not neglected;

142

143CHAPTER 3 Classic Theories of Economic Growth and Development

but otherwise its growth through rapid expansion of the percentage of the labor force in industry has broadly conformed with the Lewis model of devel- opment. After about 1970, productivity growth in agriculture also increased rapidly, owing in part to a successful integrated rural development program.

Dependence Revolution But South Korea poses a serious challenge to the dependence revolution models. Here is a poor country that became tied in with the international economy: It was strongly dependent in international relations—it was a Japanese colony until 1945 and thereafter wholly dependent on maintaining the goodwill of the United States for defense against invasion by North Korea. It received a large part of its national budget in the form of U.S. aid in the 1950s and both exported and imported a great deal from developed countries, especially the United States and Japan. The shape of the nation’s development was thus “conditioned” in large part by export opportunities to developed countries, and dependence theory would predict that retarded development opportunities should result. Yet South Korea today is an OECD member and is widely considered a “graduate” to developed- country status. Of course, dependence theorists could and do claim that South Korea is an exception because of the magnitude of aid it received and the self-interests of the advanced countries in seeing its full successful development because of its role as a bulwark against communism. And the Korean gov- ernment pursued some particular policies that the dependence school would by and large applaud, including carrying out an extremely active industrial upgrading policy, sharply limiting the role of multi- national corporations and deliberately establishing indigenous industries as an alternative, and using debt rather than direct foreign equity investment to finance extraordinary levels of investment. South Korea also implemented one of the most ambitious land reform programs in the developing world and placed strong emphasis on primary rather than university education, two policies of exceptional importance. But this does not explain how South Korea was able to adopt such policies to break out of dependence in the first place.

Neoclassical Counterrevolution South Korea likewise poses a strong challenge to the neoclas- sical counterrevolution models. The nation was highly interventionist at home and in international trade,

with the government making extensive use of devel- opment planning, using a wide range of tax breaks and incentives to induce firms to follow government direc- tives and interventions, setting individual company export targets, orchestrating efforts in various indus- tries to upgrade the average technological level, coor- dinating foreign technology licensing agreements, using monopoly power to get the best deal from com- peting multinationals, and generally inducing firms to move rapidly up the ladder of (dynamic) com- parative advantage (see Chapter 12). These policies addressed real technology and skill-raising market failure problems of development, and at least prior to the 1997 Asian currency crisis, from which Korea quickly recovered, very few cases of glaring gov- ernment failure can be pointed to in this experience. Of course, it does confirm that firms respond to economic incentives. But it may also be claimed with at least equal force that South Korea provides a compelling example of government’s role in overcoming coordi- nation failures, as examined in Chapter 4 and applied to South Korea in the end-of-chapter case study for Chapter 12.

Argentina In contrast, for Argentina, stages and patterns the- ories illuminate relatively little economic history, whereas the dependence revolution and neoclas- sical counterrevolution theories together offer important insights. It remains unclear whether Argentina has now relaunched onto a new growth episode following its 2002 default, as growth has been erratic, foreign exchange reserves falling, and political uncertainty returning.

Stages of Growth The history of Argentina poses a strong challenge to the linear-stages approach. Rostow defined takeoff as “the interval when the old blocks and resistances to steady growth are finally overcome.…Growth becomes its normal condition.” In 1870, Argentina ranked 11th in the world in per capita income (ahead of Germany); today, it is not even in the top 60. Although Rostow said that in determining a country’s stage, tech- nology absorption, not income per inhabitant, is what matters, he dated Argentina’s preconditions for takeoff as an extended period before 1914 and con- cluded that takeoff “in some sense” began in World War I, but “in the mid 1930s…a sustained takeoff was inaugurated, which by and large can now [1960] be

143

144 PART ONE Principles and Concepts

judged to have been successful,” concluding that “in Latin America the takeoff has been completed in two major cases (Mexico and Argentina).” Rostow attri- butes the fact that preconditions were there for some time before takeoff to excessive import of foreign capital over too long a period without increasing domestic savings. (But South Korea was also a heavy foreign borrower until recently.) Argentina certainly met Rostow’s criterion of developing manufacturing sectors at a rapid rate.

But now let’s look at what happened in Argentina since Rostow put the country forward as an example. According to World Bank data, Argentina had a neg- ative growth rate throughout the 1965–1990 period, and in the 1980s, domestic investment shrank at a −8.3% rate, falling back well below Rostow’s threshold takeoff investment levels. Although Argentina grew at 3.6% in 1990–2001, it defaulted on its debt in 2002, and the economy shrank 11%, followed by a recovery and resumed if erratic growth. Argentina’s share of investment in GDP from 2000 to 2007 was 17%, well under half that of South Korea. Like many other Latin American and African countries in the 1970s, 1980s, and 1990s, Argentina demonstrated that development progress is not irreversible and that sustained growth can come to an end. It remains unclear whether Argentina has now relaunched onto a new growth episode following its 2002 default, as growth has been erratic, foreign exchange reserves falling, and political uncertainty returning.

Structural Patterns Argentina did exhibit many of the usual structural patterns of development as agri- cultural productivity rose, industrial employment grew (albeit slowly), urbanization took place, fertility fell, and so on. But the fact that many structural regu- larities of development were observed even as living standards in the country stagnated illustrates some of the shortcomings of relying too much on selected pieces of data without the assistance of guiding theory on how the parts fit together.

Dependence Revolution In contrast to South Korea, the case of Argentina offers some vindi- cation for dependence theories in that the country relied to a large extent on exporting primary goods, and the real prices of these goods fell compared to imports. Multinational corporations played a large role, and Argentina was unable to create its own viable manufacturing export industries, ultimately having to submit to stringent structural-adjustment

programs, sell state industries to foreign com- panies, and other constraints. Dependence theorists can claim with some justification that Argentina’s conditioned development fell victim to developed- country economic interests, especially those of British and American corporations.

Neoclassical Counterrevolution But Argentina also offers some vindication for neoclassical coun- terrevolution theory in that faulty interventionist restrictions, inefficient state enterprise, bias against production for exports, and unnecessary red tape ended up hurting industry and entrepreneurship. Government policy consistently seemed to support privileged interests rather than broad goals of devel- opment, and government failure was usually worse than market failure in the country. In the mid-1990s, a large-scale liberalization and privatization program seemed to be beginning to reinvigorate growth in Argentina. Unfortunately, by 2002, four years of recession culminated in economic implosion as the economy collapsed under the weight of rising internal fiscal and external trade deficits, caused in part by the linking of the peso to a strong U.S. dollar. Dependence theorists claimed vindication. The recovery and com- paratively rapid growth since 2004, despite Argen- tina’s 2002 debt default, showed that single explana- tions for development success and failure are rarely adequate. Yet Argentina’s economic recovery remains vulnerable—for example, growth dropped from about 9% in 2010 and 2011 to under 2% in 2012—and political institutions remain somewhat unsettled.

Summary It is interesting that as South Korea provides a chal- lenge to both dependence and neoclassical theory— the starkest opposites in many ways—Argentina can be viewed more as a vindication for these two theories. And whereas South Korea serves more to confirm linear stages of growth and conclusions about structural patterns of development, Argentina poses challenges to their universal importance. Yet each of these four approaches has added some- thing vital to our understanding of development experiences and prospects in just these two coun- tries. South Korea also illustrates the role of gov- ernment in overcoming coordination failures, while Argentina illustrates how government can become part of a bad equilibrium, topics explored in depth in the next chapter. ■

144

145CHAPTER 3 Classic Theories of Economic Growth and Development

Sources Fishlow, Albert, et al. Miracle or Design? Lessons from the

East Asian Experience. Washington, D.C.: Overseas Development Council, 1994.

Heston, Alan, Robert Summers, and Bettina Aten. Penn World Table, version 6.3. Center for International Com- parisons of Production, Income and Prices, Univer- sity of Pennsylvania, August 2009, http://pwt.econ .upenn.edu/php_site/pwt63/pwt63_form.php.

Porter, Michael. Competitive Advantage of Nations. New York: Free Press, 1990.

Rodrik, Dani. “Coordination failures and government policy: A model with applications to East Asia and

Eastern Europe.” Journal of International Economics 40 (1996): 1–22.

Rostow, Walt W. The Stages of Economic Growth: A Non- Communist Manifesto. London: Cambridge University Press, 1960.

Smith, Stephen C. Industrial Policy in Developing Countries: Reconsidering the Real Sources of Expert-Led Growth. Washington, D.C.: Economic Policy Institute, 1991.

Thurow, Lester. Head to Head. New York: Morrow, 1992. World Bank. Korea: Managing the Industrial Transition.

Washington, D.C.: World Bank, 1987. World Bank, World Development Indicators, various years. World Bank, World Development Reports, various years.

Concepts for Review

Autarky Average product Capital-labor ratio Capital-output ratio Center Closed economy Comprador groups Dependence Dominance Dualism False-paradigm model Free market Free-market analysis

Harrod-Domar growth model Lewis two-sector model Marginal product Market failure Market-friendly approach Necessary condition Neoclassical counterrevolution Neocolonial dependence model Net savings ratio Open economy Patterns-of-development

analysis Periphery

Production function Public-choice theory Self-sustaining growth Solow neoclassical growth

model Stages-of-growth model

of development Structural-change theory Structural transformation Sufficient condition Surplus labor Underdevelopment

Questions for Discussion

1. Explain the essential distinctions among the stages-of-growth theory of development, the structural-change models of Lewis and Chenery, and the theory of international dependence in both its neo-Marxist and false-paradigm con- ceptualizations. Which model do you think pro- vides the best explanation of the situation in most developing nations? Explain your answer.

2. Explain the meaning of dualism and dual societies. Do you think that the concept of dualism ade- quately portrays the development picture in most developing countries? Explain your answer.

3. Some people claim that international dualism and domestic dualism are merely different manifestations

of the same phenomenon. What do you think they mean by this, and is it a valid conceptualization? Explain your answer.

4. What is meant by the term neoclassical counterrevo- lution? What are its principal arguments, and how valid do you think they are? Explain your answer.

5. Given the diversity of developing countries, do you think that there could ever be a single, unified theory of development? Explain your answer.

6. Is the neoclassical, free-market theory necessarily incompatible with dependence theory? How might these two approaches work together?

7. In what ways do developing countries depend on rich countries? In what ways is the opposite true?

145

146 PART ONE Principles and Concepts

Notes

1. The Stages of Economic Growth: A Non-Communist Manifesto, 3rd Edition by W. W. Rostow. Copy- right © 1960, 1971, 1990 Cambridge University Press. Reprinted with permission.

2. This model is named after two economists, Sir Roy Harrod of England and Professor Evesey Domar of the United States, who separately but concur- rently developed a variant of it in the early 1950s.

3. In traditional presentations, including of this text, the symbol k has been used for this capital-output ratio, rather than the symbol c as used here. However, we use the symbol c to make sure it is not misidentified with the use (also traditional) of the symbol k in the Solow growth model (which stands there for the capital-labor ratio), discussed later in this chapter. Note also that in practice a developing economy may utilize capital ineffi- ciently, that is, more than strictly required from an engineering standpoint.

4. To see this simply, note that Y K>c, so ΔY (1>c) Δ K. But Δ K by definition is net investment, IN, which is given by gross investment, IG, less allowance for depreciation, which in turn is given by the rate of depreciation times the capital stock, δK. That is, Δ K

IG 2 δ K. But gross investment is identically equal to gross savings, SG. So Δ Y (1>c)[SG 2 δK]. (Note that the gross savings rate, sG, is given by SG>Y.) Dividing both sides by Y and simplifying gives Δ Y> Y sG>c 2 δ, the result in the text.

5. W. Arthur Lewis, “Economic development with unlimited supplies of labour,” Manchester School 22 (1954): 139–191; John C. H. Fei and Gustav Ranis, Development of the Labor Surplus Economy: Theory and Policy (Homewood, Ill.: Irwin, 1964). See also Ragnar Nurkse, Problems of Capital For- mation in Underdeveloped Countries (New York: Oxford University Press, 1953).

6. For interesting applications to the recent growth experience of China, see Nazrul Islam and Kazuhiko Yokota, “Lewis growth model and China’s industri- alization,” Asian Economic Journal 22 (2008): 359–396; Xiaobo Zhang, Jin Yang, and Shenglin Wang, “China has reached the Lewis turning point,” IFPRI discussion paper No. 977 (Washington, D.C.: Inter- national Food Policy Research Institute, 2010). Cai

Fang, “A turning point for China and challenges for further development and reform,” Zhongguo shehui kexue (Social Sciences in China), 3 (2007) 4–12; and Huang Yiping and Jiang Tingsong, What Does the Lewis Turning Point Mean for China? A Com- putable General Equilibrium Analysis, China Center for Economic Research Working Paper 2010-03, March 2010. See also Gary Fields, “Dualism in the labor market: A perspective on the Lewis model after half a century,” Manchester School 72 (2004): 724–735. Research on models of structural change continues; for an interesting contribution, see Douglas Gollin, Stephen L. Parente, and Richard Rogerson, “The food problem and the evolution of international income levels,” Journal of Monetary Economics 54 (2007): 1230–1255.

7. See Hollis B. Chenery, Structural Change and Devel- opment Policy (Baltimore: Johns Hopkins Uni- versity Press, 1979); Hollis B. Chenery and Moshe Syrquin, Patterns of Development, 1950–70 (London: Oxford University Press, 1975); Moshe Syrquin, “Patterns of structural change,” in Handbook of Development Economics, eds. Hollis B. Chenery and T. N. Srinivasan (Amsterdam: Elsevier, 1989), vol. 1, pp. 205–273; and Hollis B. Chenery, Sherman Robinson, and Moshe Syrquin, Industrialization and Growth: A Comparative Study, (New York: Oxford University Press, 1986). For a succinct summary of research by Simon Kuznets, see his “Modern economic growth, findings and reflec- tions,” American Economic Review 63 (1973): 247–258.

8. See, for example, Sarah Anderson, John Cavanagh, Thea Lee, and the Institute for Policy Studies, Field Guide to the Global Economy (New York: New Press, 2000); Robin Broad, ed., Global Backlash: Citizen Initiatives for a Just World Economy (Lanham, Md.: Rowman & Littlefield, 2002); and John Gray, False Dawn: The Delusions of Global Capitalism (New York: New Press, 2000).

9. See Paul Baran, The Political Economy of Neo- Colonialism (London: Heinemann, 1975). An out- standing literature review is contained in Keith Griffin and John Gurley, “Radical analysis of imperialism, the Third World, and the transition to socialism: A survey article,” Journal of Economic

147CHAPTER 3 Classic Theories of Economic Growth and Development

Literature 23 (1985): 1089–1143. See also Ted C. Lewellen, Dependency and Development: An Intro- duction to the Third World (Westport, Conn.: Bergin & Garvey, 1995).

10. A provocative and well-documented application of this argument to the case of Kenya can be found in Colin Leys, Underdevelopment in Kenya: The Political Economy of Neo-Colonialism (London: Heinemann, 1975).

11. Theotonio Dos Santos, “The crisis of development theory and the problem of dependence in Latin America,” in Underdevelopment and Development, ed. Henry Bernstein (Harmondsworth, England: Penguin, 1973), pp. 57–80. See also Benjamin J. Cohen, The Question of Imperialism: The Political Economy of Dominance and Dependence (New York: Basic Books, 1973).

12. Hans W. Singer, “Dualism revisited: A new approach to the problems of dual societies in developing countries,” Journal of Development Studies 7 (1970): 60–61. Domestic dualism models continue to be developed. For example, see Arup Banerji and Sanjay Jain, “Quality dualism,” Journal of Development Economics 84 (2007): 234–250. For an interesting empirical study, see Niels- Hugo Blunch and Dorte Verner, “Shared sec- toral growth versus the dual economy model: Evidence from Côte d’Ivoire, Ghana, and Zimbabwe,” African Development Review 18, No. 3 (2006): 283–308; Jonathan Temple and Ludger Woessmann, “Dualism and cross-country growth regressions,” Journal of Economic Growth, 11, No. 3 (2006): 187–228.

13. For examples of the literature of neoclassical coun- terrevolutionaries, see Peter T. Bauer, Reality and Rhetoric: Studies in the Economics of Development (London: Weidenfield & Nicolson, 1984); Deepak Lal, The Poverty of Development Economics (Cam- bridge, Mass.: Harvard University Press, 1985); Ian Little, Economic Development: Theories, Policies, and International Relations (New York: Basic Books 1982); and any mid-1980s issue of the World Bank’s World Development Report and the International Monetary Fund’s Annual World Economic Outlook. An outstanding critique of this literature can be found in John Toye, Dilemmas of Development: Reflections on the Counter-Revolution in Development Theory and Policy (Oxford: Blackwell, 1987). See also

Ziya Onis, “The limits of neoliberalism: Toward a reformulation of development theory,” Journal of Economic Issues 29 (1995): 97–119; Lance Taylor, “The revival of the liberal creed: The IMF and the World Bank in a globalized economy,” World Devel- opment 25 (1997): 145–152; and Alexandro Portes, “Neoliberalism and the sociology of development: Emerging trends and unanticipated facts,” Popu- lation and Development Review 23 (1997): 229–259.

14. For a good explication of the tenets of the public- choice theory, see Merilee S. Grindle and John W. Thomas, Public Choices and Public Policy Change: The Political Economy of Reform in Developing Coun- tries (Baltimore: Johns Hopkins University Press, 1991). The classic article in the field is by Nobel laureate James M. Buchanan, “Social choice, democracy and free markets,” Journal of Political Economy 62 (1954): 114–123. For a critique, see Paul P. Streeten, “Markets and states: Against mini- malism,” World Development 21 (1993): 1281–1298, and Amartya Sen, “Rationality and social choice,” American Economic Review 85 (1995): 1–24.

15. See any of the 1990s World Development Reports. For a critique of this approach, see Ajit Singh, “State intervention and ‘market-friendly’ approach to development: A critical analysis of the World Bank theses,” in The States, Markets and Development, eds. Amitava K. Dutt, Kwan S. Kim, and Ajit Singh (London: Elgar, 1994). The World Bank has since added further nuance to its approach, for example, by encouraging use of the “growth diagnostics” approach described at the end of Chapter 4.

16. The Solow model is set forth in Robert Solow, “A contribution to the theory of economic growth,” Quarterly Journal of Economics 70 (1956): 65–94.

17. For a discussion of these and related issues, see Heinz W. Arndt, “’Market failure’ and under- development,” World Development 16 (1988): 210–229.

18. On identifying and addressing local constraints, see Ricardo Hausmann, Dani Rodrik, and Andrés Velasco, “Growth diagnostics,” One Economics, Many Recipes: Globalization, Institutions, and Eco- nomic Growth, by Dani Rodrik (Princeton, N.J.: Princeton University Press, 2007). See Chapter 4 for more details.

19. An additional approach, reviewed in Chapter 2, is the new institutionalism. It has been alternatively

148 PART ONE Principles and Concepts

viewed as a component of the neoclassical counter- revolution or of the postneoclassical mainstream development economics. The institutions include property rights, prices and market structures, money and financial institutions, firms and indus- trial organization, and relationships between gov- ernment and markets. The basic message of the new institutionalism is that even in a neoclassical world, the success or failure of development efforts will depend on the nature, existence, and proper func- tioning of a country’s fundamental institutions. The origins of the new institutionalism can be found in the theory of institutions pioneered by the work of Nobel laureate Ronald Coase. See Ronald H.

Coase, “The institutional structure of production,” American Economic Review 82 (1992): 713–719; Oliver E. Williamson, “The institutions and governance of economic development and reform,” in Proceedings of the World Bank Annual Conference on Development Economics, 1994 (Washington, D.C.: World Bank, 1995); and Jean-Jacques Laffont, “Competition, con- formation and development,” Annual World Bank Conference on Development Economics, 1998 (Wash- ington, D.C.: World Bank, 1999). Writings of Nobel laureate Douglass North have been particularly influential, notably Institutions, Institutional Change, and Economic Performance (New York: Cambridge University Press, 1990).

149CHAPTER 3 Classic Theories of Economic Growth and Development

Appendix 3.1

Components of Economic Growth

Three components of economic growth are of prime importance:

1. Capital accumulation, including all new investments in land, physical equipment, and human resources through improvements in health, edu- cation, and job skills

2. Growth in population and hence eventual growth in the labor force

3. Technological progress—new ways of accomplishing tasks

In this appendix, we look briefly at each.

Capital Accumulation

Capital accumulation results when some proportion of present income is saved and invested in order to augment future output and income. New factories, machinery, equipment, and materials increase the physical capital stock of a nation (the total net real value of all physically productive capital goods) and make it possible for expanded output levels to be achieved. These directly productive investments are supplemented by investments in what is known as social and economic infrastructure—roads, electricity, water and sanitation, communications, and the like—which facilitates and integrates economic activities. For example, investment by a farmer in a new tractor may increase the total output of the crops he can produce, but without adequate transport facilities to get this extra product to local commercial markets, his investment may not add anything to national food production.

There are less direct ways to invest in a nation’s resources. The instal- lation of irrigation systems may improve the quality of a nation’s agricultural land by raising productivity per hectare. If 100 hectares of irrigated land can produce the same output as 200 hectares of nonirrigated land using the same other inputs, the installation of such irrigation is the equivalent of doubling the quantity of nonirrigated land. Use of chemical fertilizers and the control of insects with pesticides may have equally beneficial effects in raising the productivity of existing farmland. All these forms of investment are ways of improving the quality of existing land resources. Their effect in raising the total stock of productive land is, for all practical purposes, indistinguishable from the simple clearing of hitherto unused arable land.

Similarly, investment in human resources can improve its quality and thereby have the same or even a more powerful effect on production as an increase in human numbers. Formal schooling, vocational and on-the-job training programs, and adult and other types of informal education may all be made more effective in augmenting human skills as a result of direct invest- ments in buildings, equipment, and materials (e.g., books, film projectors, personal computers, science equipment, vocational tools, and machinery such as lathes and grinders). The advanced and relevant training of teachers, as

Capital accumulation Increasing a country’s stock of real capital (net investment in fixed assets). To increase the production of capital goods necessitates a reduction in the production of consumer goods.

Capital stock The total amount of physical goods existing at a particular time that have been produced for use in the production of other goods and services.

Economic infrastructure The amount of physical and financial capital embodied in roads, railways, waterways, airways, and other transpor- tation and communications, plus other facilities such as water supplies, financial insti- tutions, electricity, and public services such as health and education.

150 PART ONE Principles and Concepts

well as good textbooks in economics, may make an enormous difference in the quality, leadership, and productivity of a given labor force. Improved health can also significantly boost productivity. The concept of investment in human resources and the creation of human capital is therefore analogous to that of improving the quality and thus the productivity of existing land resources through strategic investments.

All of these phenomena and many others are forms of investment that lead to capital accumulation. Capital accumulation may add new resources (e.g., the clearing of unused land) or upgrade the quality of existing resources (e.g., irrigation), but its essential feature is that it involves a trade-off between present and future consumption—giving up a little now so that more can be had later, such as giving up current income to stay in school.

Population and Labor Force Growth

Population growth, and the associated eventual increase in the labor force, have traditionally been considered a positive factor in stimulating economic growth. A larger labor force means more productive workers, and a large overall population increases the potential size of domestic markets. However, it is questionable whether rapidly growing supplies of workers in developing countries with a surplus of labor exert a positive or a negative influence on economic progress (see Chapter 6 for an in-depth discussion of the pros and cons of population growth for economic development). Obviously, it will depend on the ability of the economic system to absorb and productively employ these added workers—an ability largely associated with the rate and kind of capital accumulation and the availability of related factors, such as managerial and administrative skills.

Given an initial understanding of these first two fundamental components of economic growth and disregarding for a moment the third (technology), let us see how they interact via the production possibility curve to expand soci- ety’s potential total output of all goods. For a given technology and a given amount of physical and human resources, the production possibility curve por- trays the maximum attainable output combinations of any two commodities— say, rice and radios—when all resources are fully and efficiently employed. Figure A3.1.1 shows two production possibility curves for rice and radios.

Initial possibilities for the production of rice and radios are shown by the curve PP. Now suppose that without any change in technology, the quantity of physical and human resources were to double as a result of either invest- ments that improved the quality of the existing resources or investment in new resources—land, capital, and, in the case of larger families, labor. Figure A3.1.1 shows that this doubling of total resources will cause the entire pro- duction possibility curve to shift uniformly outward from PP to P′P′. More radios and more rice can now be produced.

Because these are assumed to be the only two goods produced by this economy, it follows that the gross domestic product (the total value of all goods and services produced) will be higher than before. In other words, the process of economic growth is under way.

Note that even if the country in question is operating with underuti- lized physical and human resources, as at point X in Figure A3.1.1, a growth

Production possibility curve A curve on a graph indicating alternative combinations of two commodities or categories of commodities (e.g., agri- cultural and manufactured goods) that can be produced when all the available factors of production are efficiently employed. Given available resources and technology, the curve sets the boundary between the attainable and the unobtainable.

151CHAPTER 3 Classic Theories of Economic Growth and Development

of productive resources can result in a higher total output combination, as at point X′, even though there may still be widespread unemployment and underutilized or idle capital and land. But note also that there is nothing deterministic about resource growth leading to higher output growth. This is not an economic law, as attested by the poor growth record of many con- temporary developing countries. Nor is resource growth even a necessary condition for short-run economic growth because the better utilization of idle existing resources can raise output levels substantially, as portrayed in the movement from X to X′ in Figure A3.1.1. Nevertheless, in the long run, the improvement and upgrading of the quality of existing resources and new investments designed to expand the quantity of these resources are principal means of accelerating the growth of national output.

Now, instead of assuming the proportionate growth of all factors of pro- duction, let us assume that, say, only capital or only land is increased in quality and quantity. Figure A3.1.2 shows that if radio manufacturing is a relatively large user of capital equipment and rice production is a relatively land-intensive process, the shifts in society’s production possibility curve will be more pronounced for radios when capital grows rapidly (Figure A3.1.2a) and for rice when the growth is in land quantity or quality (Figure A3.1.2b). However, because under normal conditions both products will require the use of both factors as productive inputs, albeit in different combinations, the production possibility curve still shifts slightly outward along the rice axis in Figure A3.1.2a when only capital is increased and along the radio axis in Figure A3.1.2b when only the quantity or quality of land resources is expanded.

Technological Progress

It is now time to consider the third, and to many economists the most important, source of economic growth, technological progress. In its simplest form, technological progress results from new and improved ways of accom- plishing traditional tasks such as growing crops, making clothing, or building

FIGURE A3.1.1 Effect of Increases in Physical and Human Resources on the Production Possibility Frontier

P

P P ′

X ′ X

R a

d io

s

Rice

P ′

0

Technological progress Increased application of new scientific knowledge in the form of inventions and inno- vations with regard to both physical and human capital.

152 PART ONE Principles and Concepts

a house. There are three basic classifications of technological progress: neutral, laborsaving, and capital-saving.

Neutral technological progress occurs when higher output levels are achieved with the same quantity and combinations of factor inputs. Simple innovations like those that arise from the division of labor can result in higher total output levels and greater consumption for all individuals. In terms of production possibility analysis, a neutral technological change that, say, doubles total output is conceptually equivalent to a doubling of all productive inputs. The outward-shifting production possibility curve of Figure A3.1.1 could therefore also be a diagrammatic representation of neutral technological progress.

By contrast, technological progress may result in savings of either labor or capital (i.e., higher levels of output can be achieved with the same quantity of labor or capital inputs). Computers, the Internet, automated looms, high- speed electric drills, tractors, mechanical ploughs—these and many other kinds of modern machinery and equipment can be classified as products of laborsaving technological progress. Technological progress since the late nineteenth century has consisted largely of rapid advances in laborsaving technologies for producing everything from beans to bicycles to bridges.

Capital-saving technological progress is a much rarer phenomenon. But this is primarily because most of the world’s scientific and technological research is conducted in developed countries, where the mandate is to save labor, not capital. In the labor-abundant (capital-scarce) developing coun- tries, however, capital-saving technological progress is what is needed most. Such progress results in more efficient (lower-cost) labor-intensive methods of production—for example, hand- or rotary-powered weeders and threshers, foot-operated bellows pumps, and back-mounted mechanical sprayers for small-scale agriculture. The indigenous development of low-cost, efficient, labor-intensive (capital-saving) techniques of production is one of the essential

FIGURE A3.1.2 Effect of Growth of Capital Stock and Land on the Production Possibility Frontier

P

P P ′

R a

d io

s

Rice

P ′

0

(a) Growth in capital stock

P

P P ′

R a

d io

s

Rice

P ′

0

(b) Growth in land resources

Neutral technological progress Higher output levels achieved with the same quantity or combination of all factor inputs.

Laborsaving technological progress The achievement of higher output using an unchanged quantity of labor inputs as a result of some invention (e.g., the com- puter) or innovation (such as assembly-line production).

Capital-saving technological progress Technological progress that results from some invention or inno- vation that facilitates the achievement of higher output levels using the same quantity of inputs of capital.

153CHAPTER 3 Classic Theories of Economic Growth and Development

ingredients in any long-run employment-oriented development strategy (see Appendix 5.1).

Technological progress may also be labor- or capital-augmenting. Labor- augmenting technological progress occurs when the quality or skills of the labor force are upgraded—for example, by the use of videotapes, televisions, and other electronic communications media for classroom instruction. Similarly, capital-augmenting technological progress results in the more productive use of existing capital goods—for example, the substitution of steel for wooden plows in agricultural production.

We can use our production possibility curve for rice and radios to examine two very specific examples of technological progress as it relates to output growth in developing countries. In the 1960s, agricultural scientists at the International Rice Research Institute in the Philippines developed a new and highly productive hybrid rice seed, known as IR-8, or “miracle rice.” These new seeds, along with later further scientific improvements, enabled some rice farmers in parts of South and Southeast Asia to double or triple their yields in a matter of a few years. In effect, this technological progress was “embodied” in the new rice seeds (one could also say it was “land-augmenting”), which per- mitted higher output levels to be achieved with essentially the same comple- mentary inputs (although more fertilizer and pesticides were recommended). In terms of our production possibility analysis, the higher-yielding varieties of hybrid rice could be depicted, as in Figure A3.1.3, by an outward shift of the curve along the rice axis with the intercept on the radio axis remaining essen- tially unchanged (i.e., the new rice seeds could not be directly used to increase radio production).

In terms of the technology of radio production, the invention of tran- sistors has probably had as significant an impact on communications as the development of the steam engine had on transportation. Even in the remotest parts of Africa, Asia, and Latin America, the transistor radio has become a prized possession. The introduction of the transistor, by obviating the need

Labor-augmenting technological progress Technological progress that raises the productivity of an existing quantity of labor by general education, on-the-job training programs, and so on.

Capital-augmenting technological progress Technological progress that raises the productivity of capital by innovation and inventions.

FIGURE A3.1.3 Effect of Technological Change in the Agricultural Sector on the Production Possibility Frontier

P

P P ′

R a

d io

s

Rice

0

154 PART ONE Principles and Concepts

for complicated, unwieldy, and fragile tubes, led to an enormous growth of radio production. The production process became less complicated, and workers were able to increase their total productivity significantly. Figure A3.1.4 shows that as in the case of higher-yielding rice seeds, the tech- nology of the transistor can be said to have caused the production possibility curve to rotate outward along the vertical axis. For the most part, the rice axis intercept remains unchanged (although perhaps the ability of rice paddy workers to listen to music on their transistor radio while working may have made them more productive!).

Conclusion

The sources of economic progress can be traced to a variety of factors, but by and large, investments that improve the quality of existing physical and human resources, increase the quantity of these same productive resources, and raise the productivity of all or specific resources through invention, inno- vation, and technological progress have been and will continue to be primary factors in stimulating economic growth in any society. The production pos- sibility framework conveniently allows us to analyze the production choices open to an economy, to understand the output and opportunity cost impli- cations of idle or underutilized resources, and to portray the effects on eco- nomic growth of increased resource supplies and improved technologies of production.

FIGURE A3.1.4 Effect of Technological Change in the Industrial Sector on the Production Possibility Frontier

P

P

P ′

R a

d io

s

Rice

0

155CHAPTER 3 Classic Theories of Economic Growth and Development

Appendix 3.2

The Solow Neoclassical Growth Model

The Solow neoclassical growth model, for which Robert Solow of the Massachu- setts Institute of Technology received the Nobel Prize, is probably the best-known model of economic growth.1 Although in some respects Solow’s model describes a developed economy better than a developing one, it remains a basic reference point for the literature on growth and development. It implies that economies will conditionally converge to the same level of income if they have the same rates of savings, depreciation, labor force growth, and productivity growth. Thus, the Solow model is the basic framework for the study of convergence across countries (see Chapter 2). In this appendix, we consider this model in further detail.

The key modification from the Harrod-Domar (or AK) growth model, con- sidered in this chapter, is that the Solow model allows for substitution between capital and labor. In the process, it assumes that there are diminishing returns to the use of these inputs.

The aggregate production function, Y = F(K, L) is assumed characterized by constant returns to scale. For example, in the special case known as the Cobb-Douglas production function, at any time t we have

Y1t2 = K1t2α1A1t2L1t221 - α (A3.2.1)

where Y is gross domestic product, K is the stock of capital (which may include human capital as well as physical capital), L is labor, and A(t) represents the productivity of labor, which grows over time at an exogenous rate.

Because of constant returns to scale, if all inputs are increased by the same amount, say 10%, then output will increase by the same amount (10% in this case). More generally,

γY = F1γK, γL)

where γ is some positive amount (1.1 in the case of a 10% increase). Because γ can be any positive real number, a mathematical trick useful in

analyzing the implications of the model is to set γ = 1>L so that

Y>L = f1K>L, 12 or y = f1k2 (A3.2.2)

Lowercase variables are expressed in per-worker terms in these equations. The concave shape of ƒ(k)—that is, increasing at a decreasing rate—reflects dimin- ishing returns to capital per worker, as can be seen in Figure A3.2.1.2 In the Harrod-Domar model, this would instead be a straight, upward-sloping line.

This simplification allows us to deal with just one argument in the production function. For example, in the Cobb-Douglas case introduced in Equation A3.2.1,

y = Akα (A3.2.3)

This represents an alternative way to think about a production function, in which everything is measured in quantities per worker. Equation A3.2.3 states that output per worker is a function that depends on the amount of capital per worker. The more capital with which each worker has to work, the more output that worker can produce. The labor force grows at rate n per year, say,

156 PART ONE Principles and Concepts

and labor productivity growth, the rate at which the value of A in the pro- duction function increases, occurs at rate λ. The total capital stock grows when savings are greater than depreciation, but capital per worker grows when savings are also greater than what is needed to equip new workers with the same amount of capital as existing workers have.

The Solow equation (Equation A3.2.4) gives the growth of the capital-labor ratio, k (known as capital deepening), and shows that the growth of k depends on savings sf(k), after allowing for the amount of capital required to service depre- ciation, δk, and after capital widening, that is, providing the existing amount of capital per worker to net new workers joining the labor force, nk. That is,

∆k = sf1k2 - 1δ + n2k (A3.2.4)

Versions of the Solow equation are also valid for other growth models, such as the Harrod-Domar model.

For simplicity, we are assuming for now that A remains constant. In this case, there will be a state in which output and capital per worker are no longer changing, known as the steady state. (If A is increasing, the corresponding state will be one in which capital per effective worker is no longer changing. In that case, the number of effective workers rises as A rises; this is because when workers have higher productivity, it is as if there were extra workers on the job.) To find this steady state, set ∆k = 0:

sf1k*2 = 1δ + n2k* (A3.2.5)

The notation k* means the level of capital per worker when the economy is in its steady state. That this equilibrium is stable can be seen from Figure A3.2.1.3

The capital per worker k* represents the steady state. If k is higher or lower than k*, the economy will return to it; thus k* is a stable equilibrium. This sta- bility is seen in the diagram by noting that to the left of k*, k < k*. Looking at the diagram, we see that in this case, 1n + δ2k < sf(k). But now looking at the Solow equation (Equation A3.2.4), we see that when (n + δ)k < sf(k), ∆k > 0. As a result, k in the economy is growing toward the equilibrium point k*. By similar reasoning to the right of k*, (n + δ)k > sf(k), and as a result, ∆k < 0

FIGURE A3.2.1 Equilibrium in the Solow Growth Model

(n + δ)k

k k*

y = f (k)

f (k)

sf (k)

157CHAPTER 3 Classic Theories of Economic Growth and Development

(again refer to Equation A3.2.4), and capital per worker is actually shrinking toward the equilibrium k*.4 Note that in the Harrod-Domar model, sf(k) would be a straight line, and provided that it was above the (n + δ)k line, growth in capital per worker—and output per worker—would continue indefinitely.

Equation (A3.2.5) has an interpretatation that the savings per worker, sf(k*), is just equal to δk*, the amount of capital (per worker) needed to replace depreciating capital, plus nk*, the amount of capital (per worker) that needs to be added due to population (labor force) growth.

The Solow model has a (single) equilibrium income per worker, again given by Equation (A3.2.5) above. In contrast, the Harrod Domar equilibrium is (constant, balanced) growth—there is no equilibrium income per worker. Essentially, this is because f(k)—and hence sf(k)—does not exhibit diminishing returns; rather, it is a straight line. That is, growth continues as long as the line sf(k) stays above the line (δ + n)k.

It is instructive to consider what happens in the Solow neoclassical growth model if we increase the rate of savings, s. A temporary increase in the rate of output growth is realized as we increase k by raising the rate of savings. We return to the original steady-state growth rate later, though at a higher level of output per worker in each later year. The key implication is that unlike in the Harrod-Domar (AK) analysis, in the Solow model an increase in s will not increase growth in the long run; it will only increase the equilibrium k*. That is, after the economy has time to adjust, the capital-labor ratio increases, and so does the output-labor ratio, but not the rate of growth. The effect is shown in Figure A3.2.2, in which savings is raised to s′. In contrast, in the Harrod- Domar model, an increase in s raises the growth rate. (This is because in the Harrod-Domar model, sf(k) becomes a straight line from the origin that does not cross (n + δ)k; and so, as we assume that sf(k) lies above (n + δ)k, growth continues at the now higher Harrod-Domar rate—a result that was repre- sented, for example, in the comparison of Equations 3.8 and 3.9.)

Note that the neoclassical growth model (Equation A3.2.5 and Figure A3.2.1) implies that while economies will (conditionally) converge to the same level of income per worker other things equal, it does not imply unconditional convergence. This can be seen clearly in Figure A3.2.2: We can interpret the alternative savings rates (s and s’) in the figure as corresponding to those pre- vailing in two different countries; the country with the higher savings rate converges to a higher equilibrium income per worker.

Note carefully that in the Solow model, an increase in s does raise equi- librium output per person—which is certainly a valuable contribution to development—just not the equilibrium rate of growth. And the growth rate does increase temporarily as the economy kicks up toward the higher equi- librium capital per worker. Moreover, simulations based on cross-national data suggest that if s is increased, the economy may not return even halfway to its steady state for decades.5 That is, for practical purposes of policymaking in developing countries, even if the Solow model is an accurate depiction of the economy, an increase in savings may substantially increase the growth rate for many decades to come. (Both theoretically and empirically, the link between the rate of savings and the rate of growth remains controversial.)

Finally, it is possible that the rate of savings (and hence investment) is posi- tively related to the rate of technological progress itself so that the growth of A depends on s. This could be the case if investment uses newer-vintage

158 PART ONE Principles and Concepts

capital and hence is more productive, if investment represents innovation in that it solves problems faced by the firm, and if other firms see what the investing firm has done and imitate it (“learning by watching”), generating externalities. This leads to a model between the standard Solow model and the endogenous growth models such as the one examined in Appendix 3.3.

FIGURE A3.2.2 The Long-Run Effect of Changing the Savings Rate in the Solow Model

f(k)

s'f (k)

k k* k*�

y

(n + δ)k

sf (k)

1. Robert M. Solow, “A contribution to the theory of economic growth,” Quarterly Journal of Economics 70 (1956): 65–94.

2. Note that the symbol k is used for K>L and not for K>Y, as it is used in many expositions (including previous editions of this text) of the AK or Harrod- Domar model.

3. Readers with more advanced mathematical training may note that Figure A3.2.1 is a phase diagram, which applies given that the Inada conditions hold: that the marginal product of k goes to infinity as k goes to zero and goes to zero as k goes to infinity (this follows from Inada conditions assumed sepa- rately for capital and labor inputs). This diminishing- returns feature drives results of the Solow model.

4. Note that in the Solow model with technological progress, that is, growth of A, the capital-labor ratio grows to keep pace with the effective labor force, which is labor power that is augmented by its increasing productivity over time.

5. See N. Gregory Mankiw, David Romer, and David N. Weil, “A contribution to the empirics of eco- nomic growth,” Quarterly Journal of Economics 107 (1992): 407–437. This article shows that when human capital is accounted for, as well as physical capital, the Solow model does a rather good job of explaining incomes and growth across countries. For a critical view, see William Easterly and Ross Levine, “It’s not factor accu- mulation: Stylized facts and growth models,” World Bank Economic Review 15 (2001): 177–219, with the reply by Robert M. Solow, “Applying growth theory across countries,” World Bank Economic Review 15 (2001): 283–288. For time- series evidence that the Solow model does a good job of explaining even the case of South Korean growth, see Edward Feasel, Yongbeom Kim, and Stephen C. Smith, “Investment, exports, and output in South Korea: A VAR approach to growth empirics,” Review of Devel- opment Economics 5 (2001): 421–432.

Notes

159CHAPTER 3 Classic Theories of Economic Growth and Development

Appendix 3.3

Endogenous Growth Theory

Motivation for Endogenous Growth Theory

The mixed performance of neoclassical theories in illuminating the sources of long-term economic growth has led to dissatisfaction with traditional growth theory. In fact, according to traditional theory, there is no intrinsic characteristic of economies that causes them to grow over extended periods of time. The literature is instead concerned with the dynamic process through which capital-labor ratios approach long-run equilibrium levels. In the absence of external “shocks” or technological change, which is not explained in the neoclassical model, all economies will converge to zero growth. Hence, rising per capita GNI is considered a temporary phenomenon resulting from a change in technology or a short-term equilibrating process in which an economy approaches its long-run equilibrium.

Any increases in GNI that cannot be attributed to short-term adjustments in stocks of either labor or capital are ascribed to a third category, commonly referred to as the Solow residual. This residual is responsible for roughly 50% of historical growth in the industrialized nations.1 In a rather ad hoc manner, neoclassical theory credits the bulk of economic growth to an exogenous or completely independent process of technological progress. Though intuitively plausible, this approach has at least two insurmountable drawbacks. First, using the neoclassical framework, it is impossible to analyze the determinants of technological advance because it is completely independent of the decisions of economic agents. And second, the theory fails to explain large differences in residuals across countries with similar technologies.

According to neoclassical theory, the low capital-labor ratios of devel- oping countries promise exceptionally high rates of return on investment. The free-market reforms impressed on highly indebted countries by the World Bank and the International Monetary Fund should therefore have prompted higher investment, rising productivity, and improved standards of living. Yet even after the prescribed liberalization of trade and domestic markets, many developing countries experienced little or no growth and failed to attract new foreign investment or to halt the flight of domestic capital. The frequently anomalous behavior of developing-world capital flows (from poor to rich nations) helped provide the impetus for the development of the concept of endogenous growth theory or, more simply, the new growth theory.

The new growth theory provides a theoretical framework for analyzing endogenous growth, persistent GNI growth that is determined by the system governing the production process rather than by forces outside that system. In contrast to traditional neoclassical theory, these models hold GNI growth to be a natural consequence of long-run equilibrium. The principal motivations of the new growth theory are to explain both growth rate differentials across countries and a greater proportion of the growth observed. More succinctly, endogenous growth theorists seek to explain the factors that determine the size of λ, the rate of growth of GDP that is left unexplained and exogenously deter- mined in the Solow neoclassical growth equation (i.e., the Solow residual).

Solow residual The pro- portion of long-term economic growth not explained by growth in labor or capital and therefore assigned primarily to exogenous technological change.

Endogenous growth theory (new growth theory) Economic growth gen- erated by factors within the production process (e.g., increasing returns or induced technological change) that are studied as part of a growth model.

160 PART ONE Principles and Concepts

Models of endogenous growth bear some structural resemblance to their neo- classical counterparts, but they differ considerably in their underlying assump- tions and the conclusions drawn. The most significant theoretical differences stem from discarding the neoclassical assumption of diminishing marginal returns to capital investments, permitting increasing returns to scale in aggregate production, and frequently focusing on the role of externalities in determining the rate of return on capital investments.2 By assuming that public and private investments in human capital generate external economies and productivity improvements that offset the natural tendency for diminishing returns, endog- enous growth theory seeks to explain the existence of increasing returns to scale and the divergent long-term growth patterns among countries. And whereas technology still plays an important role in these models, exogenous changes in technology are no longer necessary to explain long-run growth.

A useful way to contrast the new (endogenous) growth theory with tra- ditional neoclassical theory is to recognize that many endogenous growth theories can be expressed by the simple equation Y = AK, as in the Harrod- Domar model. In this formulation, A is intended to represent any factor that affects technology, and K again includes both physical and human capital. But notice that there are no diminishing returns to capital in this formula, and the possibility exists that investments in physical and human capital can generate external economies and productivity improvements that exceed private gains by an amount sufficient to offset diminishing returns. The net result is sus- tained long-term growth—an outcome prohibited by traditional neoclassical growth theory. Thus, even though the new growth theory reemphasizes the importance of savings and human capital investments for achieving rapid growth, it also leads to several implications for growth that are in direct con- flict with traditional theory. First, there is no force leading to the equilibration of growth rates across closed economies; national growth rates remain con- stant and differ across countries, depending on national savings rates and technology levels. Furthermore, there is no tendency for per capita income levels in capital-poor countries to catch up with those in rich countries with similar savings and population growth rates. A serious consequence of these facts is that a temporary or prolonged recession in one country can lead to a permanent increase in the income gap between itself and wealthier countries.

But perhaps the most interesting aspect of endogenous growth models is that they help explain anomalous international flows of capital that exacerbate wealth disparities between developed and developing countries. The poten- tially high rates of return on investment offered by developing economies with low capital-labor ratios are greatly eroded by lower levels of complementary investments in human capital (education), infrastructure, or research and development (R&D).3 In turn, poor countries benefit less from the broader social gains associated with each of these alternative forms of capital expen- diture.4 Because individuals receive no personal gain from the positive externalities created by their own investments, the free market leads to the accumulation of less than the optimal level of complementary capital. (We examine these issues further in Chapter 4.)

Where complementary investments produce social as well as private ben- efits, governments may improve the efficiency of resource allocation. They can do this by providing public goods (infrastructure) or encouraging private

Complementary investments Investments that complement and facilitate other productive factors.

161CHAPTER 3 Classic Theories of Economic Growth and Development

investment in knowledge-intensive industries, where human capital can be accumulated and subsequent increasing returns to scale generated. Unlike the Solow model, new growth theory models explain technological change as an endogenous outcome of public and private investments in human capital and knowledge-intensive industries. Thus, in contrast to the neoclassical counter- revolution theories examined in Appendix 3.2, models of endogenous growth suggest an active role for public policy in promoting economic development through direct and indirect investments in human capital formation and the encouragement of foreign private investment in knowledge-intensive indus- tries such as computer software and telecommunications.

The Romer Model

To illustrate the endogenous growth approach, we examine the Romer endog- enous growth model in detail because it addresses technological spillovers (in which one firm or industry’s productivity gains lead to productivity gains in other firms or industries) that may be present in the process of industri- alization. Thus, it is not only the seminal model of endogenous growth but also one of particular relevance for developing countries. We use a simplified version of Romer ’s model that keeps his main innovation—in modeling tech- nology spillovers—without presenting unnecessary details of savings deter- mination and other general equilibrium issues.

The model begins by assuming that growth processes derive from the firm or industry level. Each industry individually produces with constant returns to scale, so the model is consistent with perfect competition; and up to this point it matches assumptions of the Solow model. But Romer departs from Solow by assuming that the economy-wide capital stock, K, positively affects output at the industry level so that there may be increasing returns to scale at the economy-wide level.

It is valuable to think of each firm’s capital stock as including its knowledge. The knowledge part of the firm’s capital stock is essentially a public good, like A in the Solow model, that is spilling over instantly to the other firms in the economy. As a result, this model treats learning by doing as “learning by investing.” You can think of Romer ’s model as spelling out—endogenizing— the reason why growth might depend on the rate of investment (as in the Harrod-Domar model). In this simplification, we abstract from the household sector an important feature of the original model, in order to concentrate on issues concerning industrialization.5 Formally,

Yi = AK α i L

1 - α i K

β (A3.3.1)

We assume symmetry across industries for simplicity, so each industry will use the same level of capital and labor. Then we have the aggregate production function:

Y = AK α + βL1 - α (A3.3.2)

To make endogenous growth stand out clearly, we assume that A is con- stant rather than rising over time; that is, we assume for now that there is

Romer endogenous growth model An endogenous growth model in which technological spillovers are present; the economy-wide capital stock positively affects output at the industry level, so there may be increasing returns to scale at the econ- omy-wide level.

Public good An entity that provides benefits to all individuals simultaneously and whose enjoyment by one person in no way diminishes that of anyone else.

162 PART ONE Principles and Concepts

1. Oliver J. Blanchard and Stanley Fischer, Lectures on Macroeconomics (Cambridge, Mass.: MIT Press, 1989).

2. For a short history of the evolution of theoretical models of growth, see Nicholas Stern, “‘The deter- minants of growth,” Economic Journal 101 (1991): 122–134. For a more detailed but technical dis- cussion of endogenous growth models, see Robert

Barro and Xavier Sala-i-Martin, Economic Growth, 2nd ed. (Cambridge, Mass.: MIT Press, 2003), and Elhanan Helpman, “Endogenous macroeconomic growth theory,” European Economic Review 36 (1992): 237–268.

3. See Paul M. Romer, “Increasing returns and long- run growth,” Journal of Political Economy 94 (1986): 1002–1037; Robert E. Lucas, “On the mechanics of

no technological progress. With a little calculus,6 it can be shown that the resulting growth rate for per capita income in the economy would be

g - n = βn

1 - α - β (A3.3.3)

where g is the output growth rate and n is the population growth rate. Without spillovers, as in the Solow model with constant returns to scale, β = 0, and so per capita growth would be zero (without technological progress).7

However, with Romer ’s assumption of a positive capital externality, 1β 7 02, we have that g - n 7 0 and Y>L is growing. Now we have endog- enous growth, not driven exogenously by increases in productivity. If we also allowed for technological progress, so that λ in the Solow model is greater than zero, growth would be increased to that extent.8

Criticisms of Endogenous Growth Theory

An important shortcoming of the new growth theory is that it remains dependent on a number of traditional neoclassical assumptions that are often inappropriate for developing economies. For example, it assumes that there is but a single sector of production or that all sectors are symmetrical. This does not permit the crucial growth-generating reallocation of labor and capital among the sectors that are transformed during the process of structural change.9 Moreover, economic growth in developing countries is frequently impeded by inefficiencies arising from poor infrastructure, inadequate institutional structures, and imperfect capital and goods markets. Because endogenous growth theory overlooks these very influential factors, its applicability for the study of economic development is limited, especially when country-to-country comparisons are involved. For example, existing theory fails to explain low rates of factory capacity utilization in low-income countries where capital is scarce. In fact, poor incentive structures may be as responsible for sluggish GNI growth as low rates of saving and human capital accumulation. Allocational inefficiencies are common in economies undergoing the transition from traditional to commercialized markets. However, their impact on short- and medium-term growth has been neglected due to the new theory’s emphasis on the determinants of long-term growth rates. Finally, empirical studies of the predictive value of endogenous growth theories have to date offered only limited support.10

Notes

163CHAPTER 3 Classic Theories of Economic Growth and Development

economic development,” Journal of Monetary Eco- nomics 22 (1988): 3–42; and Robert Barro, “Gov- ernment spending in a simple model of endog- enous growth,” Journal of Political Economy 98 (1990): 5103–5125.

4. For a concise technical discussion of the impor- tance of human capital as a complementary input, see Robert B. Lucas, “Why doesn’t capital flow from rich to poor countries?” AEA Papers and Pro- ceedings 80 (1990): 92–96.

5. The specific functional form in Equation A3.3.1, known as Cobb-Douglas production functions, will be assumed for simplicity.

6. By the chain rule,

Y # =

dY dt

= 0Y 0K

0K 0t

+ 0Y 0L

0L 0t

By the exponent rule, we know that

0Y 0K

= A(α + β)K α+β- 1L1 -α

0Y 0L

= AK α+β(1 - α)L1 -α- 1

Combining these three equations, we have

Y # = dY>dt = 3AK α+βL1 -α4 c(α + β)

K #

K + (1 - α)

L #

L d

The first term in brackets in the preceding expression is of course output, Y. For a steady state, K

# >K, L

# >L, and Y

# >Y are all constant. From

earlier discussion of the Harrod-Domar and Solow models, we know that

K # = I - δK = sY - δK

where δ stands for the depreciation rate. Dividing this expression through by K, we have

K #

K =

sY K

- δ

For K # >K constant in the preceding expression, we

must have Y>K constant. If this ratio is constant, we have

K #

K =

Y #

Y = g, a constant growth rate

So from the expression for dY>dt above, for the aggregate production function, with L

# >L = n,

which is also a constant, we have

Y #

Y = (α + β)

K #

K + (1 - α)

L #

L S g

= (α + β)g + (1 - α)n S g - n

= c (1 - α) + (α + β) - 1

1 - (α + β) dn

which is Equation A3.3.3. This may also be expressed as

g = n(1 - α)

1 - α - β

7. Recall that there is no technological progress, so λ in the Solow model is zero.

8. In a more complex model, decisions about, and effects of, factors such as research and devel- opment investment can be modeled explicitly. Firms would decide on general investment and R&D investment. The effect of the latter on overall output would enter in a manner similar to K in Equation A3.3.1. For a discussion and references, see Gene M. Grossman and Elhanan Helpman, “Endogenous innovation in the theory of growth” in the symposium on new growth theory in the Journal of Economic Perspectives 8 (1994): 3–72.

9. Syed Nawab Haider Naqvi, “The significance of development economics,” World Development 24 (1996): 977.

10. For an excellent review and empirical critique of the new growth theory, see Howard Pack, “Endogenous growth theory: Intellectual appeal and empirical shortcomings,” Journal of Eco- nomic Perspectives 8 (1994): 55–72. See also articles by Paul M. Romer and Robert M. Solow in the same issue. For an argument that endogenous theory performs well in explaining differences in growth rates among countries, see Barro and Sala-i-Martin, Economic Growth. An excellent survey of quantitative growth research disputing this claim and indicating widening gaps between rich and poor countries can be found in Jonathan Temple, “The new growth evidence,” Journal of Economic Literature 37 (1999): 112–156.

After more than a half century of experience with attempting to encourage modern development, we have learned that development is both possible and extremely difficult to achieve. Thus, an improved understanding of impedi- ments and catalysts of development is of the utmost importance. Since the late 1980s, significant strides have been made in the analysis of economic devel- opment and underdevelopment. In some cases, ideas of the classic theories reviewed in Chapter 3 have been formalized, and in the process, their logi- cal structure and their significance for policy have been clarified and refined. At the same time, the analysis has also led to entirely new insights into what makes development so hard to achieve (as witnessed in sub-Saharan Africa) but also possible to achieve (as witnessed in East Asia). Indeed, this is what makes the study of economic development so very important: It does not hap- pen automatically; it requires systematic effort. But development is far from a hopeless cause; we know it can be done. Theory helps us think systematically about how to organize our efforts to help achieve development—a goal sec- ond to none in its importance to humanity.

In this chapter, we review a sample of some of the most influential of the new models of economic development. In some ways, these models show that development is harder to achieve, in that it faces more barriers than had previously been recognized. But greater understanding itself facilitates improvements in development strategy, and the new models have already influenced development policy and modes of international assistance. The

164

Contemporary Models of Development and Underdevelopment

Individuals need not make the right tradeoffs. And whereas in the past we thought the implication was that the economy would be slightly distorted, we now understand that the interaction of these slightly distorted behaviors may produce very large distortions. The consequence is that there may be multiple equilibria and that each may be inefficient.

—Karla Hoff and Joseph E. Stiglitz, Frontiers in Development Economics, 2002

Governments can certainly deter entrepreneurship when they try to do too much; but they can also deter entrepreneurship when they do too little.

—Dani Rodrik, One Economics, Many Recipes, 2007

4

165CHAPTER 4 Contemporary Models of Development and Underdevelopment

chapter concludes with a framework for appraising the locally binding constraints on the ability of a developing nation to further close the gap with the developed world.

The new research has broadened considerably the scope for modeling a market economy in a developing-country context. One of its major themes is incorporating problems of coordination among economic agents, such as among groups of firms, workers, or firms and workers together. Other key themes, often but not always in conjunction with the coordination problem, include the formal exploration of situations in which increasing returns to scale, a finer division of labor, the availability of new economic ideas or knowl- edge, learning by doing, information externalities, and monopolistic competi- tion or other forms of industrial organization other than perfect competition predominate. The new perspective frequently incorporates work in the “new institutional economics,” such as that of Nobel laureate Douglass C. North, and introduced in Chapter 2. All of these approaches depart to some degree from conventional neoclassical economics, at least in its assumptions of per- fect information, the relative insignificance of externalities, and the unique- ness and optimality of equilibria.1

4.1 Underdevelopment as a Coordination Failure

Many newer theories of economic development that became influential in the 1990s and the early years of the twenty-first century have emphasized complementarities between several conditions necessary for successful devel- opment. These theories often highlight the problem that several things must work well enough, at the same time, to get sustainable development under way. They also stress that in many important situations, investments must be undertaken by many agents in order for the results to be profitable for any individual agent. Generally, when complementarities are present, an action taken by one firm, worker, or organization increases the incentives for other agents to take similar actions.

Models of development that stress complementarities are related to some of the models used in the endogenous growth approach (described in Appendix 3.3), in ways we will point out later in the chapter, but the coordination failure approach has evolved relatively independently and offers some significant and distinct insights.2 Put simply, a coordination failure is a state of affairs in which agents’ inability to coordinate their behavior (choices) leads to an outcome (equilibrium) that leaves all agents worse off than in an alternative situation that is also an equilibrium. This may occur even when all agents are fully informed about the preferred alternative equilibrium: They simply cannot get there because of dif- ficulties of coordination, sometimes because people hold different expectations and sometimes because everyone is better off waiting for someone else to make the first move. This section spells out the meaning and implications of these per- spectives in detail, through both simple models and examples.

When complementarities are present, an action taken by one firm, worker, organization, or government increases the incentives for other agents to take

Binding constraint The one limiting factor that if relaxed would be the item that accelerates growth (or that allows a larger amount of some other tar- geted outcome).

Complementarity An action taken by one firm, worker, or organization that increases the incentives for other agents to take similar actions. Com- plementarities often involve investments whose return depends on other investments being made by other agents.

Economic agent An economic actor—usually a firm, worker, consumer, or government offi- cial—that chooses actions so as to maximize an objective; often referred to as “agents.”

Coordination failure A situ- ation in which the inability of agents to coordinate their behavior (choices) leads to an outcome (equilibrium) that leaves all agents worse off than in an alternative situa- tion that is also an equilibrium.

166 PART onE Principles and Concepts

similar actions. In particular, these complementarities often involve invest- ments whose return depends on other investments being made by other agents. In development economics, such network effects are common, and we consider some important examples later in this chapter, including the model of the big push, in which production decisions by modern-sector firms are mutually reinforcing, and the o-ring model, in which the value of upgrading skills or quality depends on similar upgrading by other agents. Curiously, such effects are also common in analyses of frontier technologies in devel- oped countries, particularly information technologies, in which the value of using an operating system, word-processing program, spreadsheet program, instant messaging, and other software or product standard depends on how many other users also adopt it. In both cases, the circular causation of posi- tive feedback is common.3 This framework may also be used in analyses of the middle-income trap, in which countries develop to a degree but chroni- cally fail to reach high-income status, often due to lack of innovation capacity.

An important example of a complementarity is the presence of firms using specialized skills and the availability of workers who have acquired those skills. Firms will not enter a market or locate in an area if workers do not possess the skills the firms need, but workers will not acquire the skills if there are no firms to employ them. This coordination problem can leave an economy stuck in a bad equilibrium—that is, at a low average income or growth rate or with a class of citizens trapped in extreme poverty. Even though all agents would be better off if workers acquired skills and firms invested, it might not be possible to get to this better equilibrium without the aid of government. As we will see, such coordination problems are also common in initial industrialization, as well as in upgrading skills and technologies, and may extend to issues as broad as chang- ing behavior to modern “ways of doing things.” Such problems are further com- pounded by other market failures, particularly those affecting capital markets.4

Another example typical of rural developing areas concerns the com- mercialization of agriculture. As Adam Smith already understood, special- ization is one of the sources of high productivity. Indeed, specialization and a detailed division of labor are hallmarks of an advanced economy. But we can specialize only if we can trade for the other goods and services we need. Producers must somehow get their products to markets while convincing distant buyers of their quality. As Shahe Emran and Forhad Shilpi stress, in the development of agricultural markets, middlemen play a key role by effectively vouching for the quality of the products they sell; they can do this because they get to know the farmers from whom they buy as well as the products. It is difficult to be an expert in the quality of many products, so in order for a specialized agricultural market to emerge, there needs to be a suf- ficient number of concentrated producers with whom a middleman can work effectively. But without available middlemen to whom the farmers can sell, they will have little incentive to specialize in the first place and will prefer to continue producing their staple crop or a range of goods primarily for per- sonal consumption or sale within the village. The result can be an underde- velopment trap in which a region remains stuck in subsistence agriculture.5

In many cases, the presence of complementarities creates a classic “chicken and egg” problem: Which comes first, the skills or the demand for skills? Often the answer is that the complementary investments must come at the same time,

Big push A concerted, economy-wide, and typically public policy–led effort to initiate or accelerate economic development across a broad spectrum of new industries and skills.

o-ring model An economic model in which produc- tion functions exhibit strong complementarities among inputs and which has broader implications for impediments to achieving economic devel- opment.

Middle-income trap A con- dition in which an economy begins development to reach middle-income status but is chronically unable to prog- ress to high-income status. Often related to low capacity for original innovation or for absorption of advanced technology, and may be com- pounded by high inequality.

Underdevelopment trap A poverty trap at the regional or national level in which underdevelopment tends to perpetuate itself over time.

167CHAPTER 4 Contemporary Models of Development and Underdevelopment

through coordination. This is especially true when, as is generally the case, there is a lag between making an investment and realizing the return on that invest- ment.6 In this case, even if, for some reason, all parties expect a change to a bet- ter equilibrium, they will still be inclined to wait until other parties have made their investments. Thus, there can be an important role for government policy in coordinating joint investments, such as between the workers who want skills that employers can use and the employers who want equipment that workers can use. Neither may be in a position (or find it in their self-interest) to take the first step; each may be better off waiting for the other parties to invest first.

As another example, a new or modernizing firm using new technologies may provide benefits to other firms that the adopting firm cannot capture; so each firm has an incentive to underinvest in the new technology unless a sufficient number of others invest. Some of these benefits may include raising demand for key industrial products such as steel, helping pay for the fixed costs of an essential infrastructure such as railroads or container ports, or learning from others’ experiences. We will take a closer look at this problem later in the chapter.

The new work expands the scope for potentially valuable government pol- icy interventions, but it does not take their success for granted. Rather, govern- ment itself is increasingly analyzed in contemporary development models as one of the components of the development process that may contribute to the problem as well as to the solution; government policy is understood as partly determined by (endogenous to) the underdeveloped economy (see Chapter 11). For example, a dictator such as Mobutu Sese Seko, the former ruler of the Dem- ocratic Republic of Congo when it was known as Zaire, may prefer to keep his country in an underdevelopment trap, knowing full well that as the economy develops, he will lose power. But rather than concluding that government gen- erally exacerbates underdevelopment rather than facilitates development (as in extreme versions of the neoclassical counterrevolution school), many devel- opment specialists look actively for cases in which government policy can still help, even when government is imperfect, by pushing the economy toward a self-sustaining, better equilibrium. Such deep interventions move an econ- omy to a preferred equilibrium or even to a higher permanent rate of growth in which there is no incentive to go back to the behavior associated with the bad equilibrium. In these cases, government has no need to continue the interventions, because the better equilibrium will be maintained automati- cally. Government can then concentrate its efforts on other crucial problems in which it has an essential role (e.g., in addressing problems of public health). This onetime-fix character of some multiple-equilibrium problems makes them worthy of special focus because they can make government policy that much more powerful in addressing problems of economic development. But it also makes the policy choices more momentous, because a bad policy today could mire an economy in a bad equilibrium for years to come.

In much of economics, such complementarities are not present. For exam- ple, in competitive markets, when there is excess demand, there is counter- pressure for prices to rise, restoring equilibrium. Whenever congestion may be present, these counterpressures are very strong: The more people there are fishing in one lake, the more fishers try to move to another lake that is less crowded; the more people there are using one road, the more commuters

Deep intervention A government policy that can move the economy to a pre- ferred equilibrium or even to a higher permanent rate of growth, which can then be self-sustaining so that the policy need no longer be enforced because the better equilibrium will then prevail without further intervention.

Congestion The opposite of a complementarity; an action taken by one agent that decreases the incentives for other agents to take similar actions.

168 PART onE Principles and Concepts

try to find an alternative route. But in the process of economic development, joint externalities are common: Underdevelopment begets underdevelopment, while processes of sustainable development, once under way, tend to stimulate further development.

Coordination problems are illustrated by the where-to-meet dilemma: Sev- eral friends know that they will all be in Buenos Aires on a certain day but have neglected to settle on a specific location within the city. Now they are out of communication and can arrive at a common meeting point only by chance or by very clever guessing. They want to meet and consider themselves better off if they can do so; there is no incentive to “cheat.” Thus, the where-to-meet prob- lem is quite different from that of prisoners’ dilemma, another problem often encountered in theories of economic development.7 But the fact that all gain from coordination does not make the where-to-meet dilemma easy to solve. There are many famous places in Buenos Aires: the Plaza de Mayo, the Cathe- dral, the colorful Caminito neighborhood, the Café Tortoni, the Cementerio de la Recoleta, even the casino. Only with luck would the friends end up making the same guesses and meeting in the same place. Arriving at, say, the center of Cami- nito and not finding the others there, one of our travelers might decide to try the Plaza de Mayo instead. But en route she might miss another of the other travel- ers, who at that moment might be on his way to check out the Cementerio. So the friends never meet. Something analogous happens when farmers in a region do not know what to specialize in. There may be several perfectly good products from which to choose, but the critical problem is for all the farmers to choose one so that middlemen may profitably bring the region’s produce to market.

The story may lose a bit of its power in the age of texting, cell phones, and e-mail. For example, as long as the friends have each other ’s contact informa- tion, they can come to an agreement about where to meet. Sometimes what seems at first a complex problem of coordination is really a simpler one of communication. But anyone who has tried to establish a meeting time by phone or e-mail with a large number of participants with no formal leader knows that this can be a slow and cumbersome process. Without a clear leader and with a large enough number of participants, no meeting place may be agreed to on short notice before it is too late. And in real economic problems, the people who need to “meet”—perhaps to coordinate investments—do not even know the identity of the other key agents.8 However, our example does point up possibilities for improved prospects for development with the advent of modern computing and telecommunications technology. Of course, peasant farmers may not have access to cell phones or e-mail (but see the case study for Chapter 11 on the Grameen Bank).

4.2 Multiple Equilibria: A Diagrammatic Approach

The standard diagram to illustrate multiple equilibria with possible coordi- nation failure is shown in Figure 4.1. This diagram, in one version or another, has become almost as ubiquitous in discussions of multiple equilibria as the famous supply-and-demand (“Marshallian scissors”) diagram in discussions of single equilibrium analysis.9

Where-to-meet dilemma A situation in which all parties would be better off cooperat- ing than competing but lack information about how to do so. If cooperation can be achieved, there is no subse- quent incentive to defect or cheat.

Prisoners’ dilemma A situation in which all parties would be better off cooper- ating than competing, but once cooperation has been achieved, each party would gain the most by cheating, provided that others stick to cooperative agreements—thus causing any agreement to unravel.

Multiple equilibria A con- dition in which more than one equilibrium exists. These equilibria sometimes may be ranked, in the sense that one is preferred over another, but the unaided market will not move the economy to the preferred outcome.

169CHAPTER 4 Contemporary Models of Development and Underdevelopment

The basic idea reflected in the S-shaped function of Figure 4.1 is that the ben- efits an agent receives from taking an action depend positively on how many other agents are expected to take the action or on the extent of those actions. For example, the price a farmer can hope to receive for his produce depends on the number of middlemen who are active in the region, which in turn depends on the number of other farmers who specialize in the same product.

How do we find the equilibria in this type of problem? In the Marshallian supply-and-demand scissors diagram, equilibrium is found where the supply and demand curves cross. In the multiple-equilibria diagram, equilibrium is found where the “privately rational decision function” (the S-shaped curve in Figure 4.1) crosses the 45-degree line. This is because in these cases, agents observe what they expected to observe. Suppose that firms expected no other firms to make invest- ments, but some firms did anyway (implying a positive vertical intercept in the diagram). But then, seeing that some firms did make investments, it would not be reasonable to continue to expect no investment! Firms would have to revise their expectations upward, matching their expectations to the level of investment they actually would see. But if firms now expected this higher level of investment, firms would want to invest even more. This process of adjustment of expectations would continue until the level of actual investment would just equal the level of expected investment: At that level, there would be no reason for firms to adjust their expecta- tions any further. So the general idea of an equilibrium in such cases is one in which all participants are doing what is best for them, given what they expect others to do, which in turn matches what others are actually doing. This happens when the function crosses the 45-degree line. At these points, the values on the x-axis and y-axis are equal, implying in our example that the level of investment expected is equal to the level that all agents find best (e.g., the profit-maximizing level).

D1

Y1

D2 D3

P ri

va te

ly r

a ti

o n

a l d

e ci

si o

n (e

.g .,

i n

d iv

id u

a l i

n ve

st m

e n

t le

ve l)

Expected decision by other agents (e.g., average investment level)

6125_04_FG001

FigUrE 4.1 Multiple Equilibria

170 PART onE Principles and Concepts

In the diagram, the function cuts the 45-degree line three times. Any of these points could be an equilibrium: That is what we mean by the possibil- ity of multiple equilibria. Of the three, D1 and D3 are “stable” equilibria. They are stable because if expectations were slightly changed to a little above or below these levels, firms would adjust their behavior—increase or decrease their investment levels—in a way to bring us back to the original equilibrium. Note that in each of these two stable equilibria, the S-shaped function cuts the 45-degree line from above—a hallmark of a stable equilibrium.

At the middle equilibrium at D2, the function cuts the 45-degree line from below, and so it is unstable. This is because in our example, if a little less investment were expected, the equilibrium would be D1, and if a little more, were expected, the equilibrium would move to D3. D2 could therefore be an equilibrium only by chance. Thus, in practice, we think of an unstable equi- librium such as D2 as a way of dividing ranges of expectations over which a higher or lower stable equilibrium will hold sway.

Typically, the S-shaped “privately rational decision function” first increases at an increasing rate and then at a decreasing rate, as in the diagram. This shape reflects what is thought to be the typical nature of complementarities. In general, some agents may take the complementary action (such as invest- ing) even if others in the economy do not, particularly when interactions are expected to be with foreigners, such as through exporting to other countries. If only a few agents take the action, each agent may be isolated from the oth- ers, so spillovers may be minimal. Thus, the curve does not rise quickly at first as more agents take the action. But after enough agents invest, there may be a snowball effect, in which many agents begin to provide spillover benefits to neighboring agents, and the curve increases at a much faster rate. Finally, after most potential investors have been positively affected and the most important gains have been realized, the rate of increase starts to slow down.

In many cases, the shape of the function in Figure 4.1 could be different, how- ever. For example, a very “wobbly” curve could cut the 45-degree line several times. In the case of telephone service, getting on e-mail or instant messaging, or buying a fax machine, where the value of taking the action steadily increases with the number of others in the network, the function may only increase at an increasing rate (like a quadratic or exponential function). Depending on the slope of the function and whether it cuts the 45-degree line, there can be a single equilibrium or multiple equilibria, including cases in which either no one ever adopts a new technology or virtually everyone does. In general, the value (util- ity) of the various equilibria (two in this case) is not the same. For example, it is very possible that everyone is better off in the equilibrium in which more people use the network. In this case, we say the equilibria are Pareto-ranked, with the higher rank to the equilibrium giving higher utility to everyone; in other words, moving to this equilibrium represents a Pareto improvement over the equilibrium with fewer users.

The classic example of this problem in economic development concerns coordinating investment decisions when the value (rate of return) of one investment depends on the presence or extent of other investments. All are better off with more investors or higher rates of investment, but the market may not get us there without the influence of certain types of government policy (but note that we may also not arrive at the preferred solutions if we have the wrong kinds of government policy). The difficulties of investment

Pareto improvement A situ- ation in which one or more persons may be made better off without making anyone worse off.

171CHAPTER 4 Contemporary Models of Development and Underdevelopment

coordination give rise to various government-led strategies for industrializa- tion that we consider both in this chapter and later in the text (see especially Chapter 12).

The investment coordination perspective helps clarify the nature and extent of problems posed when technology spillovers are present, such as seen in the Romer model described in Appendix 3.3.10 Given what was learned in examining endogenous growth theory about the possible relation between investment and growth, you can see that an economy can get stuck in a low growth rate largely because the economy is expected to have a low investment rate. Strategies for coordinating a change from a less produc- tive to a more productive set of mutually reinforcing expectations can vary widely, as the example in Box 4.1 and the findings in Box 4.2 illustrate. How- ever, changing expectations may not be sufficient if it is more profitable for

BOX 4.1 Synchronizing Expectations: resetting “Latin American Time”

Kaushik Basu and Jorgen Weibull argue that while the importance of culture is undeniable, the in- nateness of culture is not. They present a model that shows that punctuality may be “simply an equilib- rium response of individuals to what they expect oth- ers to do” and that the same society can benefit from a “punctual equilibrium” or get caught in a lateness equilibrium.

Estimates suggested that Ecuador lost between 4% and 10% of its GDP due to chronic lateness. As one commentator put it, “Tardiness feeds on itself, creating a vicious cycle of mañana, mañana.” Lately, Ecuador has tried to make up for lost time. Inspired by some in the younger generation who are fed up with “Latin American time,” government and busi- ness have joined in a private-sector-funded drive to get people to show up at their scheduled appointment times. The country has launched a national campaña contra la impuntualidad (campaign against lateness), coordinated by Participación Ciudadana (Citizen Par- ticipation). The result is a test of the idea that a society can consciously switch from a bad to a good equilib- rium through a change in expectations.

The campaign is a timely one. A newspaper is pub- lishing a list each day of officials who are late for public events. A popular poster for the campaign against late- ness describes the disease and says, “Treatment: Inject yourself each morning with a dose of responsibility,

respect and discipline. Recommendation: Plan, orga- nize activities and repair your watches.” Hundreds of public and private institutions have signed up to a promise to be punctual. A popular notice for meeting rooms in the style of hotel “Do Not Disturb” signs has been making the rounds. On one side it says, “Come in: You’re on time.” When the meeting begins at its scheduled time, it is turned around to the other side, which reads, “Do not enter: The meeting began on time.”

In Peru, a similar campaign is under way. If the campaign against lateness proves successful, it will be about more than time. If a social movement to change expectations about punctuality can be made to work, something similar might be tried around the world for fixing even more pernicious problems, such as public corruption.

Sources: Kaushik Basu and Jorgen Weibull, “Punctuality: A cultural trait as equilibrium,” in Economics for an Imperfect World: Essays in Honor of Joseph Stiglitz, ed. Richard Arnott et al. (Cambridge, Mass.: MIT Press, 2003); Scott Wilson, “In Ecuador, a timeout for tardiness drive promotes punctuality,” Washington Post Foreign Ser- vice, November 4, 2003, p. A22; “The price of lateness,” Economist, November 22, 2003, p. 67; “Punctuality pays,” New Yorker, April 5, 2004, p. 31. For an interesting cri- tique, see Andrew M. Horowitz, “The punctuality prisoners’ dilemma: Can current punctuality initiatives in low-income countries succeed?” Paper presented at the Northeast Universities Development Consortium Conference, Harvard University, October 2007.

BOX 4.2 FiNDiNgS Village Coordination and Monitoring for Better Health Outcomes

Chapter 4 explains the important role of improved information, shared expectations, and coordina- tion across agents in making development progress. Coordination across households potentially can im- prove outcomes, for example, by changing norms to- ward lower fertility and ending harmful practices, and enforcing noncorrupt and efficient public-service pro- visions. A recent study by Martina Björkman and Ja- kob Svensson shows how these mechanisms may work by drawing on evidence from a randomized control trial. The researchers found that initially, villagers had little information about the scope of health problems in their village compared with outside standards, nor about what to reasonably expect from government- funded health workers. The program provided villag- ers with the knowledge and resources to enable them to monitor health workers individually and through their community organization. This is important to do as a community because both information gath- ering and monitoring have features of public goods. The results suggest that such a program can improve the behavior of health workers and lead to measurably better health outcomes—all for apparently very mod- est cost outlays.

The study questions were whether the interven- tion caused an increased quantity and quality of health care provision; and whether this resulted in im- proved health outcomes. The researchers were check- ing for impacts along the hypothesized “accountabil- ity chain” that treatment communities became more involved in monitoring health workers and that the intervention changed the behavior of health workers. The initial intervention had three components: first, a meeting of villagers; second, a meeting with health care workers; and finally, a meeting including both groups. This was followed by a plan of action and monitoring organized by villagers.

Initially, a “report card” comparing performance of the local health facility with others was prepared. Then facilitators in conjunction with local commu- nity leaders and community-based organizations

organized a village meeting to hear and discuss the results and develop an action plan. (This is similar to the process of many community-based develop- ment activities in Africa and elsewhere.) Participation in the two-afternoon event was carefully planned to include—and hear from—diverse representatives to avoid elite capture. The facilitators “encouraged com- munity members to develop a shared view on how to improve service delivery and monitor the provider,” which were “summarized in an action plan.” In these meetings, researchers observed some common con- cerns that “included high rates of absenteeism, long waiting-time, weak attention of health staff, and dif- ferential treatment.”

The health facility meeting was a one-afternoon, all-staff event where facilitators contrasted the facil- ity’s information on service provision with findings from a household survey. Finally, an “interface meet- ing” was held with community representatives cho- sen at the community meetings and health workers, where rights, responsibilities, and suggestions for improvements were discussed, resulting in a “shared action plan…on what needs to be done, how, when and by whom.” Then, “after the initial meetings, the communities were themselves in charge of establish- ing ways of monitoring the provider.”

The program was associated with (and apparently caused) positive health outcomes, including relatively higher weights of infants, fewer deaths of children less than five years old, and greater utilization of health facilities. Evidence showed that as a result of the pro- gram, treatment practices also improved the “quality and quantity of health care provision,” suggesting that increases “are due to behavioral changes.” In par- ticular, equipment (such as a thermometer) was used more often; waiting time was reduced; clinic cleanli- ness improved; better information was provided to patients; appropriate supplements and vaccines for children were provided more often; and absenteeism by health workers declined. The program was esti- mated to improve health outcomes to a degree similar

172 PART onE Principles and Concepts

173CHAPTER 4 Contemporary Models of Development and Underdevelopment

a firm to wait for others to invest rather than to be a “pioneer” investor. In that case, government policy is generally needed in addition to a change of expectations. This explains why attention to the potential presence of mul- tiple equilibria is so important. Market forces can generally bring us to one of the equilibria, but they are not sufficient to ensure that the best equilibrium will be achieved, and they offer no mechanism to become unstuck from a bad equilibrium and move toward a better one.

A similar multiple-equilibria situation will be encountered in our analy- sis of the Malthus population trap in Chapter 6. In this population trap, fertility decisions need in effect to be coordinated across families—all are

to findings from high-impact medical trials. However, such trials assume the health system is working fine and only benefits from improved procedures and medications; in contrast, this approach focused on getting health workers to do what they were supposed to do in the first place.

Some checks confirmed the program more likely had its impact through community participation rather than other mechanisms, but it is still pos- sible that other mechanisms such as health workers responding to learning about patient rights rather than community pressure played some role; so we may not yet be certain how the program worked. This type of question is important to investigate because understanding mechanisms helps with designing other programs effectively.

Overall, the researchers surmised that “lack of rel- evant information and failure to agree on, or coordi- nate expectations of, what is reasonable to demand from the provider were holding back individual and group action to pressure and monitor the provider.”

The authors caution that: “Before scaling up, it is also important to subject the project to a cost-benefit analysis.…A back-of-the-envelope calculation sug- gests that….The estimated cost of averting the death of a child under five is around $300.” If this estimate holds up to more systematic analysis, it would be an unusually cost-effective program. The authors con- cluded by noting that “future research should ad- dress long-term effects, identify which mechanisms or combination of mechanisms that are important, and study the extent to which the results generalize to other social sectors.”

There remain some other questions. As hinted, it is uncertain whether these improvements can be sustained over time—at least without periodic outside facilita- tion—for example, if the initial interest for participants is in being part of a foreign-sponsored program and this motive fades over time, or if long-term threats to collec- tive organization including free riding and capture rear their heads. So it would be valuable to return to these villages to look at conditions after a few years. It is not clear yet how well or how cost-effectively this approach would work elsewhere—the “external validity” question again. Even if the program does indeed work through the mechanism of empowerment, as seems quite likely, the real powers that be may not have allowed such out- comes if material interests of rulers were threatened by the program. Moreover, as the researchers note, an approach that combined more monitoring from the top of the health ministry in combination with the bottom- up monitoring of communities, as done in this program, could have even larger positive impacts. Finally, people and their communities have limited time; so inducing a shift of time to the health system monitoring activity in this program could cause a decrease in the amount of other valuable community activities.

But in sum, this is an exemplary design and evalu- ation of a community-based development program that provides substantive evidence of what can work to improve health (and empowerment) of villagers in a low-income rural area.

Sources: Martina Björkman and Jakob Svensson, “Power to the People: Evidence from a Randomized Field Experi- ment on Community-Based Monitoring in Uganda,” Quarterly Journal of Economics, 124 (2), pp 735–769, May 2009; and supplementary appendix.

174 PART onE Principles and Concepts

better off if the average fertility rate declines, but any one family may be worse off by being the only one to have fewer children. We also see coordi- nation failures in processes of urbanization and other key elements of eco- nomic development.

In general, when jointly profitable investments may not be made without coordination, multiple equilibria may exist in which the same individuals with access to the same resources and technologies can find themselves in either a good or a bad situation. In the view of many development economists, it is very plausible that many of the least developed countries, including many in sub-Saharan Africa, are essentially caught in such circumstances. Of course, other problems are also present. For example, political pressures from potential losers in the modernization process can also prevent shifts to better equilibria. In addition, modern technology may not yet be available in the country. The technology transfer problem is another important concern in economic devel- opment. In fact, another problem illustrated by the graph in Figure 4.1 could be that the amount of effort each firm in a developing region expends to increase the rate of technology transfer depends on the effort undertaken by other firms; bringing in modern technology from abroad often has spillover effects for other firms. But the possibility of multiple equilibria shows that making better technology available is generally a necessary but not a sufficient condi- tion for achieving development goals.

4.3 Starting Economic Development: The Big Push

Whether an economy has been growing sustainably for some time or has been stagnant seems to make a very big difference for subsequent development. If growth can be sustained for a substantial time, say, a generation or more, it is much more unusual for economic development to later get off track for long (though, of course, there will be setbacks over the business cycle as the economy is affected by temporary shocks). Certainly, we have had too many disappointing experiences to assume, with Rostow, that once economic devel- opment is under way, it can in effect never be stopped. As noted in the case study in Chapter 3, a century ago, Argentina was regarded as a future pow- erhouse of the world economy, yet it later experienced relative stagnation for more than half a century. A look at the record, however, allows us to agree with Rostow at least in that it is very difficult to get modern economic growth under way in the first place and much easier to maintain it once a track record has been established.

Why should it be so difficult to start modern growth? Many models of development that were influential in earlier years, such as the Lewis model examined in Chapter 3, assume perfectly competitive conditions in the indus- trial sector. Under perfect competition, it is not clear why starting develop- ment would be so difficult, provided at least that the needed human capital is developed, the technology transfer problem is adequately addressed, and government provides other essential services. But development seems hard to initiate even when better technologies are available—they often go unused. Apparently, people do not have the incentives to put the new technology to

175CHAPTER 4 Contemporary Models of Development and Underdevelopment

work. Beyond this, perfect competition does not hold under conditions of increasing returns to scale. And yet, looking at the Industrial Revolution, it is clear that taking advantage of returns to scale has been key. Many develop- ment economists have concluded that several market failures work to make economic development difficult to initiate, notably pecuniary externalities, which are spillover effects on costs or revenues.

Perhaps the most famous coordination failures model in the development literature is that of the “big push,” pioneered by Paul Rosenstein-Rodan, who first raised some of the basic coordination issues.11 He pointed out several problems associated with initiating industrialization in a subsistence economy, of the type introduced in Chapter 1. The problem is easiest to perceive if we start with the simplifying assumption that the economy is not able to export. In this case, the question becomes one of who will buy the goods produced by the first firm to industrialize. Starting from a subsistence economy, no work- ers have the money to buy the new goods. The first factory can sell some of its goods to its own workers, but no one spends all of one’s income on a single good. Each time an entrepreneur opens a factory, the workers spend some of their wages on other products. So the profitability of one factory depends on whether another one opens, which in turn depends on its own potential prof- itability, and that in turn depends on the profitability of still other factories. Such circular causation should now be a familiar pattern of a coordination failure problem. Moreover, the first factory has to train its workers, who are accustomed to a subsistence way of life. The cost of training puts a limit on how high a wage the factory can pay and still remain profitable. But once the first firm trains its workers, other entrepreneurs, not having to recoup training costs, can offer a slightly higher wage to attract the trained workers to their own new factories. However, the first entrepreneur, anticipating this likeli- hood, does not pay for training in the first place. No one is trained, and indus- trialization never gets under way.

The big push is a model of how the presence of market failures can lead to a need for a concerted economy-wide and probably public-policy-led effort to get the long process of economic development under way or to accelerate it. Put differently, coordination failure problems work against successful industrializa- tion, a counterweight to the push for development. A big push may not always be needed, but it is helpful to find ways to characterize cases in which it will be.

Rosenstein-Rodan’s arguments became a major part of the way develop- ment economists thought about development problems in the 1950s and 1960s, and they have continued to be taught in development courses. But while some of the basic intuition has thus been around for decades, the approach received a huge boost following the 1989 publication of a technical paper by Kevin Murphy, Andrei Shleifer, and Robert Vishny, which for the first time demon- strated the formal logic of this approach more clearly.12 Its recent appeal is also due in part to its perceived value in explaining the success of the East Asian miracle economies, notably that of South Korea. One value of using a formal model is to get a clearer sense of when the need for coordination is more likely to present a serious problem. The approach of these authors was in turn simplified and popularized by Paul Krugman in his 1995 monograph, Development, Geography, and Economic Theory, and became the classic model of the new development theories of coordination failure of the 1990s.13

Pecuniary externality A positive or negative spillover effect on an agent’s costs or revenues.

176 PART onE Principles and Concepts

The Big Push: A Graphical Model

Assumptions In any model (indeed, in any careful thinking), we need to make some assumptions, sometimes seemingly large assumptions, to make any progress in our understanding. The analysis of the big push is no excep- tion to this rule. The assumptions we use for the big push analysis here can be relaxed somewhat, though at the expense of requiring more mathemati- cal technique, but it should be noted that we cannot relax our assumptions as much as we are accustomed to doing in simpler microeconomic problems, such as those that assume perfect competition. Here we cannot meaningfully assume perfect competition in the modern sector, where increasing returns to scale and hence natural monopoly, or at least monopolistic competition, pre- vail. To paraphrase Paul Krugman, if we think development has something significant to do with increasing returns to scale, then we will have to sacrifice some generality to address it. We will make six types of assumptions.

1. Factors. We assume that there is only one factor of production—labor. It has a fixed total supply, L.

2. Factor payments. The labor market has two sectors. We assume that workers in the traditional sector receive a wage of 1 (or normalized to 1, treating the wage as the numeraire; that is, if the wage is 19 pesos per day, we sim- ply call this amount of money “1” to facilitate analysis using the geometry in Figure 4.2). Workers in the modern sector receive a wage W 7 1 (that is, some wage that is greater than 1).

As a stylized fact, this wage differential is found in every developing country, even if it needs some explanation (see Chapter 7). The underlying reason for this differential may be a compensation for disutility of mod- ern factory types of work. If so, in equilibrium, workers would receive no net utility benefits from switching sectors during industrialization; but if economic profits are generated, this will represent a Pareto improvement (in this case because investors are better off and no one is worse off), and average income would rise (there can also be income redistribution so that everyone may be made better off, not just no one worse off). Moreover, if there is surplus labor in the economy or if modern wages are higher than opportunity costs of labor for some other reason,14 the social benefits of industrialization are all the greater.15 Finally, note that we are examining one example of a model in which a driving force for an underdevelop- ment trap is the relatively high wages that have to be paid in the modern sector. We do this because it is an approach that is easy to characterize graphically and that has received a lot of attention. As will be described later, however, high modern wages is only one circumstance in which a coordination problem may exist. In fact, we will see that there may be coordination failure problems even if modern-sector wages are no higher than those in the traditional sector.

3. Technology. We assume that there are N types of products, where N is a large number.16 For each product in the traditional sector, one worker produces one unit of output (this is a less stringent assumption than it

177CHAPTER 4 Contemporary Models of Development and Underdevelopment

appears because again we have a certain freedom in choosing our unit of measurement; if a worker produces three pairs of shoes per day, we call this quantity one unit). This is a very simple example of constant- returns-to-scale production. In the modern sector, there are increasing returns to scale. We want to introduce increasing returns in a very sim- ple way. Assume that no product can be produced unless a minimum of, say, F workers are employed. This is a fixed cost. Because we are keep- ing things simple to facilitate analysis of the core issues, we have not put capital explicitly in the model; thus the only way to introduce a fixed cost is to require a minimum number of workers. After that, there is a linear production function in which workers are more productive than those in the traditional sector. Thus labor requirements for producing any product in the modern sector take the form L = F + cQ, where c 6 1 is the marginal labor required for an extra unit of output. The trade-off is that modern workers are more productive, but only if a significant cost is paid up front. As this fixed cost is amortized over more units of output, average cost declines, which is the effect of increasing returns to scale. We assume sym- metry: The same production function holds for producing any product in the modern sector.

4. Domestic demand. We assume that each good receives a constant and equal share of consumption out of national income. The model has only one

Li

4671_FG.04.02.eps

L /NF0

Q1

W1W2

W3

Q2 B

A

Qi

Traditional- sector production function

Modern- sector production function

Labor

FigUrE 4.2 The Big Push

178 PART onE Principles and Concepts

period and no assets; thus there is no saving in the conventional sense. As a result, if national income is Y, then consumers spend an equal amount, Y/N, on each good.17

5. International supply and demand. We assume that the economy is closed. This makes the model easy to develop. The most important conclusions will remain when trade is allowed, provided that there are advantages to having a domestic market. These advantages likely include initial economies of scale and learning to achieve sufficient quality, favorable product characteristics, and better customer support before having to produce for distant and unknown consumers. These are very realistic considerations: Evidence suggests that export-led economies such as South Korea have benefited enormously from the presence of a sub- stantial domestic market to which early sales are directed.18 Moreover, export-led economies have benefited from an active industrial policy aimed at overcoming coordination failures (see Chapter 12). The points will also hold if there are necessary inputs that are not tradable, such as certain types of services. Alternative models focusing on infrastructure investments can also imply the need for a big push even with a fully open world economy.19

6. Market structure. We assume perfect competition in the traditional (cottage industry) sector, with free entry and no economic profits. Therefore, the price of each good will be 1, the marginal cost of labor (which is the only input). We assume that at most, one modern-sector firm can enter each market. This limitation is a consequence of increasing returns to scale. Given the assumptions about preferences, the monopolist faces unit-elas- tic demand, so if this monopolist could raise its price above 1, it would be profitable to do so.20 However, if price is raised above 1, competition from the traditional-sector producers will cause the modern-sector firm to lose all of its business. Therefore, the monopolist will also charge a price of 1 if it decides to enter the market.21 Because the monopolist charges the same price, it will monopolize this particular market if it enters but will also produce the same quantity that was produced by the traditional produc- ers. Because this firm is the only one using modern techniques and, in pro- ducing all other products, workers receive a wage of 1, national income will be essentially the same, so more units of output cannot be sold.22 We also assume that at the point the monopolist would choose to produce, it is able to produce at least as much output as the traditional producers for that same level of labor; otherwise, it would make no sense to switch out of the traditional techniques.

Conditions for Multiple Equilibria With these six assumptions, we can characterize cases that will require a big push. To begin, suppose that we have a traditional economy with no modern production in any market. A potential producer with modern technology (i.e., a technology like the one described previously, with fixed costs and increasing returns) considers whether it is profitable to enter the market. Given the size of the fixed cost, the answer depends on two considerations: (1) how much more efficient the modern

179CHAPTER 4 Contemporary Models of Development and Underdevelopment

sector is than the traditional sector and (2) how much higher wages are in the modern sector than in the traditional sector.

In Figure 4.2, production functions are represented for the two types of firms for any industry.23 The traditional producers use a linear technique with slope 1, with each worker producing one unit of output. The modern firm requires F workers before it can produce anything, but after that, it has a linear technique with slope 1/c 7 1. Price is 1, so revenues PQ can be read off the Q axis. For the traditional firm, the wage bill line lies coincident with the production line (both start at the origin and have a slope of 1). For the modern firm, the wage bill line has slope W 7 1. At point A, we see the output that the modern firm will produce if it enters, provided there are traditional firms operating in the rest of the economy. Whether the modern firm enters depends, of course, on whether it is profitable to do so.

Using Figure 4.2, first consider a wage bill line like W1 passing below point A. With this relatively low modern wage, revenues exceed costs, and the mod- ern firm will pay the fixed cost F and enter the market. In general, this out- come is more likely if the firm has lower fixed costs or lower marginal labor requirements as well as if it pays a lower wage. By assumption, production functions are the same for each good, so if a modern firm finds it profitable to produce one good, the same incentives will be present for producing all goods, and the whole economy will industrialize through market forces alone; demand is now high enough that we end up at point B for each product. This shows that a coordination failure need not always happen: It depends on the technology and prices (including wages) prevailing in the economy.

If a wage bill line like W2 holds, passing between points A and B, the firm would not enter if it were the only modern firm to do so in the economy because it would incur losses. But if modern firms enter in each of the mar- kets, then wages are increased to the modern wage in all markets, and income expands. We may assume that price remains 1 after industrialization. Note that the traditional technique still exists and would be profitable with a price higher than 1. So to prevent traditional firms from entering, modern firms can- not raise prices above 1.24 The modern firm can now sell all of its expanded output (at point B), produced by using all of its available labor allocation (L/N), because it has sufficient demand from workers and entrepreneurs in the other industrializing product sectors. As can be seen in Figure 4.2, with prevailing wage W2, point B is profitable after industrialization because it lies above the W2 line. Workers are also at least as well off as when they worked in the traditional sector because they can afford to purchase an additional quan- tity of goods in proportion to their increased wage,25 and they have changed sectors (from traditional to modern) voluntarily. All of the output is purchased because all of national income is spent on output; national income is equal to wages plus profits, the value of which is output of each product times the number of products N.26

Thus, with a prevailing wage like W2, there are two equilibria: one in which producers with modern techniques enter in all markets, and profits, wages, and output are higher than before; and one in which no modern producer enters, and wages and output remain lower. The equilibrium with higher output is unambiguously better, but in general, the market will not get there by itself.

180 PART onE Principles and Concepts

A final possibility is found in a wage bill line like W3, passing above point B. In this case, even if a modern producer entered in all product sectors, all of these firms would still lose money, so again the traditional technique would continue to be used. In general, whenever the wage bill line passes below point A, the market will lead the economy to modernize, and whenever it passes above A, it will not. The steeper (i.e., more efficient) the modern-sector production technique or the lower the fixed costs, the more likely it is that the wage bill will pass below the corresponding point A. If the line passes above B, it makes no sense to industrialize. But if the wage line passes between points A and B, it is efficient to industrialize, but the market will not achieve this on its own. Be sure to note that these are three different wages that might exist, depending on conditions in a particular economy at one point in time, not three wages that occur successively.

Again, the problematic cases occur when the wage bill line passes between A and B, thus creating two equilibria: one in which there is industrialization and the society is better off (point B) and one without industrialization (point A). However, the market will not get us from A to B because of a coordination failure.27 In this case, there is a role for policy in starting economic develop- ment. There is no easy test to determine where a traditional economy, such as Mozambique, is located on this continuum. But at least we can begin to under- stand why development often has not gotten under way, even when technol- ogy is available.

Note that in general, it is not necessary for all product sectors to indus- trialize to get a sufficient push for some to do so. It is only necessary that a sufficient number industrialize in order to generate enough national income (through the higher industrial wage and positive profits from the industri- alized product sectors) to make industrialization minimally profitable. Also note that each firm’s failure to take into account the impact of its investments on demand for other firms’ goods represents a very small distortion by itself. But when added up across all of the product sectors, the resulting distortion— namely, the failure to industrialize at all—is very large indeed.

We could also have cases of semi-industrialization, in which benefits or costs accrue in different amounts to different product sectors or in which there are different types of spillovers from firm to firm. For example, this is plau- sible when the level of required fixed costs declines the more product sectors industrialize, because there are more local examples from which to learn.28 With this alternative type of externality, no wage premium is necessary for multiple equilibria to be present. In this case, if there are clusters of two or more firms that have large effects on each other ’s fixed costs, F, but not on firms outside of the cluster, the result can be an equilibrium in which only the industries in this cluster change to modern techniques. Thus, in this circum- stance, we could have three or more equilibria; we could also have enclave economies, in which a modern sector exists side by side with traditional cot- tage industries in other product sectors.29

Notice that this model has not assumed the existence of any type of technological externality, in which the presence of one advanced firm can, through “learning by watching” other firms’ production methods or some similar effect, generate spillovers to other firms that can raise their productiv- ity as well as lower their costs. This is another type of market failure that can

Technological externality A positive or negative spillover effect on a firm’s production function through some means other than market exchange.

181CHAPTER 4 Contemporary Models of Development and Underdevelopment

also lead to inefficiently low investment; we considered one such possibility when we examined the Romer endogenous growth model in Appendix 3.3.

Other Cases in Which a Big Push May Be Necessary

The need for a big push can result from four conditions beyond those described previously.

1. Intertemporal effects. Even if the industrial wage rate is 1 (i.e., the same as the traditional-sector wage), multiple equilibria can occur if investment must be undertaken in the current period to get a more efficient produc- tion process in the next period.30 Investment in the first period depresses aggregate demand in the first period but increases it in the second (or later) period. But investment will be undertaken only if it is profitable, that is, if demand is expected to be high enough in the second period, and this may require that many product sectors invest simultaneously. Once again, however, the market does not ensure that industrialization will oc- cur, even when it is (Pareto-)preferred, because of pecuniary externalities. Again the source of the multiple equilibria is that one firm’s profits do not capture its external contribution to overall demand for modern-sector products because it also raises wage income in the future periods when other entering modern firms will be seeking to sell their own products. When there is a case for a big push, industrialization makes the society better off (is Pareto-preferred) because first-period income is decreased only by the fixed cost, but second-period income is sufficiently increased by both the wage and profits in other product sectors to more than offset this.31 Note once again that a part of the profits can, in principle, also be subject to income redistribution so that everyone may be made better off rather than just some people made better off and no one made worse off.

2. Urbanization effects. If some of the traditional cottage industry is rural and the increasing-returns-to-scale manufacturing is urban, urban dwellers’ demand may be more concentrated in manufactured goods (e.g., foods must be processed to prevent spoilage due to the time needed for trans- portation and distribution). If this is the case, one needs a big push to urbanization to achieve industrialization.32

3. Infrastructure effects. By using infrastructure, such as a railroad or a port, an investing modern firm helps defray the large fixed costs of that infra- structure. The existence of the infrastructure helps investing firms lower their own costs. But investing firms thereby contribute indirectly to lower- ing the costs of other firms (by lowering the average cost of infrastructure use). Infrastructure, such as roads, railroads, and ports, is not tradable; by definition, it is located in a particular region. And openness to foreign investment cannot always solve the problem because investors do not know whether firms will develop to make use of the infrastructure.33 The critical point is that when one product sector industrializes, it increases the size of the market for the use of infrastructure services that would be used by other product sectors and so makes the provision of these ser- vices more profitable. But it is also possible that efficient industrialization

182 PART onE Principles and Concepts

may not take place, even if the infrastructure is built, if other coordination problems are present.

4. Training effects. There is underinvestment in training facilities because en- trepreneurs know that the workers they train may be enticed away with higher wages offered by rival firms that do not have to pay these training costs. There is also too little demand by workers for training because they do not know what skills to acquire. (In addition to not knowing whether firms will make investments requiring these skills, people are not born with perfect information about their comparative advantage; basic edu- cation helps workers discover it.) This is part of the economic case for mandatory public education. Note that in this case, openness to trade can- not resolve the coordination failure unless there is free mobility of labor across borders, which has yet to develop perfectly even within the Eu- ropean Union, where there are few formal barriers to such mobility, and is far from emerging for any developing country. In any case, relying on expatriate skilled workers is hardly an adequate solution to a country’s own underdevelopment. Actually, infrastructure and trained workers are subsets of a general case of jointly used intermediate goods. Another ex- ample is joint research facilities for small firms in an “industrial district” (see Chapter 7).

Why the Problem Cannot Be Solved by a Super-Entrepreneur

Some readers may wonder, why can’t one agent solve the coordination failure problems by capturing all the rent? In other words, why not have a super- entrepreneur who enters into all of the markets that need to be coordinated and receives the profits from all of them? For some types of coordination failures, this solution is ruled out in advance. For example, regarding educa- tion and skill development, there is a legal constraint on bonded labor. But in terms of our industrialization problem, why can’t one agent become a super- entrepreneur in each of the N markets simultaneously? There are at least four significant theoretical answers and one decisive empirical answer.

First, there may be capital market failures. How could one agent assemble all the capital needed to play the super-entrepreneur role? Even if this were logistically imaginable, how would lenders have confidence in their invest- ments? In particular, how could a penalty for default be imposed?

Second, there may be costs of monitoring managers and other agents and designing and implementing schemes to ensure compliance or provide incentives to follow the wishes of the employer; these are often referred to as agency costs. Monitoring is too expensive once the scale of a firm gets too large. Even if the plan is to sell off the industries, these industries must be developed simultaneously. The super-entrepreneur is likely to know more about the firms than the potential buyers do. In other words, if the firm is so profitable, why would its owners be selling? Thus, potential purchasers of the industries face a problem of asymmetric information, often known as the “lemons problem.”34

Third, there may be communication failures. Suppose someone says to you, “I am coordinating investments, so work with me.” Should you do so?

Agency costs Costs of moni- toring managers and other employees and of designing and implementing schemes to ensure compliance or pro- vide incentives to follow the wishes of the employer.

Asymmetric information A situation in which one party to a potential transaction (often a buyer, seller, lender, or borrower) has more infor- mation than another party.

183CHAPTER 4 Contemporary Models of Development and Underdevelopment

How do you know this person will eventually be the coordinator? There is a potentially huge profit to be made by assuming the super-entrepreneur role, so many agents might wish to play it. If many try to claim the role, with which one should you coordinate? Even if each agent personally encounters only one pretender to the super-entrepreneur role, that pretender may still not be the right one (i.e., the coordinator with whom you can make money).

Fourth, there are limits to knowledge. Even if we stipulate that the econ- omy as a whole has access to modern technological ideas, this does not mean that one individual can gain sufficient knowledge to industrialize (or even gain enough knowledge about whom to hire to industrialize).

Finally, there is the empirical reason that no private agent has been observed playing the role of super-entrepreneur. Whether because of problems of moni- toring, knowledge, capital markets, or other diseconomies of scope, “solving” problems with ever-larger firms clearly provides no answer. For example, it is rare enough to find a firm producing steel and even a significant fraction of the products using steel, let alone one firm owning all the industries backwardly linked from steel or forwardly linked from steel-using industries to industries further down the production chain. Nor can the problem be solved by direct government production (at least without unacceptable cost), as the extreme case of the former Soviet Union demonstrates. Rather, public coordination of actions of private investors is generally needed to solve the problem, a com- mon interpretation of the role of industrial policy in East Asia.

in a Nutshell Thus we have seen that under some conditions, pecuniary exter- nalities associated with the development process can lead to multiple equilib- ria, which may create a case for a big push policy. Our main example (the moderate wage premium case) and each of the other examples have as a com- mon feature a process by which an investing (industrializing) firm captures only part of the contribution of its investment to the profits of other investing firms. In these examples, firms adopting increasing-returns-to-scale technolo- gies are having one or more of the following effects: raising total demand, shifting demand toward manufactured goods, redistributing demand toward the (later) periods in which other industrializing firms sell, reducing the fixed costs of later entrants, or helping defray the fixed costs of an essential infra- structure. Each of these has external beneficial effects on other industrializing firms.

4.4 Further Problems of Multiple Equilibria

Inefficient Advantages of Incumbency

The presence of increasing returns in modern industries can also create another kind of bad equilibrium. Once a modern firm has entered, it has an advantage over any rivals because its large output gives it low average costs. So if an even better modern technology becomes available to a potential rival, it may not be easy for the new technology to supplant the old. Even though the new technique has a lower per-unit cost for any given level of output, the firm with the old technique has an advantage because its large output lets it

184 PART onE Principles and Concepts

produce at a lower per-unit cost than that of the new technique, which starts out with a small customer base and a large fixed cost. As a result, firms may need access to significant amounts of capital to cover losses while they build their customer base. If capital markets do not work well, as they often do not in developing countries (see Chapter 15), the economy may be stuck with backward, less cost-effective industries.35

Behavior and Norms

Movement to a better equilibrium is especially difficult when it involves many individuals changing their behavior from one of rent seeking or corruption to honesty and the value of building a reputation to reap the gains from coopera- tion (e.g., with business partners). Your choice of partner may determine much. If you naively cooperate with an opportunistic, predator type, you may be worse off than by going it alone. Only by cooperating with other good-willed cooperators may you reach the best outcome. Moreover, past experience may lead people to expect opportunistic behavior at least among certain groups of potential business partners, which in turn raises the incentives for the potential partners to actually act that way. If there is nothing to be gained and some- thing to be lost by being honest, the incentives lie in being dishonest. On the other hand, in some settings, individuals take it on themselves to enforce norms rather than leaving this task to government. If many people work to enforce a norm such as honesty, each individual’s enforcement burden is relatively low. You can have equilibria where most people resist corruption, and so corruption is rare; and you can have equilibria where few resist corruption, and corruption is common.

We cannot rely on good organizations to prevail in competition if the rules of the game tend to reward the bad organizations. Rather, the critical impor- tance of policies for developing or reforming institutions is highlighted, such as reform of the framework of property rights, antitrust, clean government rules, and other laws, regulations, and industry association norms that set the rules of the game for economic life. Once the new behavior assumes the status of a norm, it is much easier to maintain. Some neoclassical theorists have at times implied that good institutions would be developed through the market mechanism. Bad institutions would be outcompeted by good institu- tions. But reform of institutions aiding and abetting coordination failure—for example, by permitting or encouraging corruption—is itself subject to coor- dination failure.

Once cooperative relationships (e.g., in business) become a norm, more people may adopt cooperative behavior. But norms of all kinds are subject to inertia. Although norms may have been adaptive when they originated, they are hard to change, even when they become dysfunctional. An example is a value such as that to be a good citizen (or a good Hindu, Muslim, Chris- tian, animist, etc.) one must have a large number of children. This value may have been adaptive at a premodern stage, but today it inhibits development. Another example may be to distrust anyone who is not a member of your fam- ily. This may be helpful in a tribal context, and caution is always advisable, but this extreme injunction hardly encourages the formation of successful business partnerships in a modern economy.

185CHAPTER 4 Contemporary Models of Development and Underdevelopment

Linkages

There are several ways to undertake a big push, encouraging the simulta- neous expansion of the modern sector in many industries. One strategy for solving coordination problems is to focus government policy on encouraging the development of industries with key backward or forward linkages. This could mean subsidies or quid pro quos for domestic industries to enter these key industries, as was done in South Korea; it could mean incentives for mul- tinational firms to enter in key industries and provide advanced training, a policy followed in Singapore; or it could mean establishing a few key public enterprises to act as pioneers in an industry (that could later be sold), as was done in South Korea and Taiwan.36 The theory of linkages stresses that when certain industries are developed first, their interconnections or linkages with other industries will induce or at least facilitate the development of new industries. Backward linkages raise demand for an activity, while forward linkages lower the costs of using an industry’s output; both may involve interactions between the size of the market and increasing returns to scale and hence pecuniary externalities. In other words, linkages are especially significant for industrialization strategy when one or more of the industries involved have increasing returns to scale, of which a larger market may take advantage. For example, when the manufacture of power looms expands, enabling a reduction in the price of power looms, there are forward linkage effects due to increased output of woven cloth made by the power looms. When increased demand for chemicals used in textile manufacture causes expansion of the chemical industry that enables it to produce at a larger scale and hence lower cost, a backward linkage can occur. Both examples illustrate a pecuniary externality effect (a lowering of cost) when there are increasing returns in the linked industry.

The linkage approach targets investment in a key linkage as a start to overcoming a coordination failure and generating positive feedback. Such a policy would select industries with a larger number of links to other indus- tries and greater strength of those links. In choosing among industries with several strong links (and passing a cost-benefit test), one policy would gen- erally select industries that have a smaller likelihood of private investment, because that is where the most intransigent bottlenecks are most likely to be found. If an investment is profitable, it is more likely that an entrepre- neur will come along to fill that niche.37 This observation provides a reason to interpret with some caution studies that show state-owned enterprises to be less efficient than private ones. If government systematically enters vital but less profitable industries because of their beneficial effects on develop- ment, it is unreasonable to hold these enterprises to the same profit standards as those of the private firms. This is certainly not to say that state-owned enterprises are generally as efficient as privately owned ones; in fact, there is much evidence to the contrary.38 We can say, however, that a blanket state- ment, such as has often been made in publications from agencies such as the World Bank, that government should never be in the business of production, even temporarily or in any industry, is sometimes unreasonable in the light of linkages and other strategic complementarities that a developing economy needs to address.

Linkages Connections between firms based on sales. A backward linkage is one in which a firm buys a good from another firm to use as an input; a forward linkage is one in which a firm sells to another firm. Such linkages are especially significant for industrialization strategy when one or more of the industries (product areas) involved have increasing returns to scale that a larger market takes advantage of.

186 PART onE Principles and Concepts

Inequality, Multiple Equilibria, and Growth

Other important work being done on growth and multiple equilibria addresses the impact of inequality on growth. The traditional view has been that some inequality may enhance growth because the savings of the rich are higher than those of the poor. If at least some savings to be mobilized for investment purposes must come from within a country, then according to this view, too high a degree of equality could compromise growth. However, the poor save at much higher rates than previously believed, when savings are properly measured to include expenditures on health, children’s education, and improvements on a home.

Moreover, where inequality is great, the poor may not be able to obtain loans because they lack collateral; indeed, one definition of what it means to be poor is to be entirely or mostly lacking in a source of collateral. Poor per- sons unable to get a loan to start a business due to such capital market imper- fections may get stuck in subsistence or wage employment, although they (and perhaps potential employees) could do much better if they had access to financing or if there were a more even distribution of income. For example, Abhijit Banerjee and Andrew Newman show that multiple equilibria, includ- ing equilibria involving outcomes with virtually all citizens enjoying high incomes and outcomes with predominantly low-income people, can exist when imperfect credit markets provide too few people with the opportunity to become entrepreneurs.39

Similarly, if the poor lack access to credit, they may not be able to obtain loans to finance otherwise very productive schooling. If the poor are unable to bequeath much to their next generation, families can be trapped in pov- erty from generation to generation; however, if schooling could somehow be achieved, they could escape from this poverty trap. It is best to keep in mind a rather expansive definition of what is meant by a transfer from parents to be used for human capital accumulation by their children. It is more than tuition and more than forgone wages or work on the farm to help the family because it goes well beyond the cost of formal schooling and may be thought of as the building of a whole array of “capabilities” (see Chapter 1) that one acquires almost as a simple by-product of growing up in an affluent, educated family.

In a formal model of this problem, Oded Galor and Joseph Zeira examined the implications of missing credit markets for growth and the distribution of both income and human capital. They developed an endogenous growth model that points up the importance of both human capital and distribution, and of the interaction between the two, for economic growth and develop- ment as well as for more short-term macroeconomic adjustments. Their analy- sis contains two critical assumptions: (1) imperfect capital markets, which, as will be described in detail in Chapter 15, is a typical condition of these mar- kets, and (2) indivisibilities in human capital investment, which means that markets treat investment in human capital as coming in discrete packages, such as a year of school, if not larger blocks, such as primary, secondary, and tertiary education. The second assumption does not seem unreasonable, both because of the nature of learning and because of the screening nature of mar- kets for human capital. A threshold level of knowledge is necessary before an employer will be willing to pay for it. Further, because education acts as a screen for inherent ability, as will be discussed in Chapter 8, we have the well-known “sheepskin effect”; that is, there is a very large jump in the return

Poverty trap A bad equilib- rium for a family, community, or nation, involving a vicious circle in which poverty and underdevelopment lead to more poverty and underde- velopment, often from one generation to the next.

187CHAPTER 4 Contemporary Models of Development and Underdevelopment

to human capital when an individual passes primary school and again when the person obtains a secondary school diploma and so on. This is not because the last course taken conveys so much more knowledge than the ones preced- ing it but because the degree itself is what enables the individual to prove that an entire regimen of requirements has been met. Note that indivisibilities in amounts of investment imply a region of increasing returns to scale, as in the fixed costs of the big push model. Once again, increasing returns play a key role in generating multiple equilibria.40 Empirically, many studies have found a negative impact of inequality on growth, especially for the period after 1980.41

4.5 Michael Kremer’s O-ring Theory of Economic Development

Another innovative and influential model that provides important insights into low-level equilibrium traps was provided by Michael Kremer.42 The notion is that modern production (especially in contrast to traditional crafts production) requires that many activities be done well together in order for any of them to amount to a high value. This is a form of strong complementar- ity and is a natural way of thinking about specialization and the division of labor, which along with economies of scale is another hallmark of developed economies in general and industrial production in particular. The name for Kremer ’s model is taken from the 1986 Challenger disaster, in which the failure of one small, inexpensive part caused the space shuttle to explode. The O-ring theory is interesting in part because it explains not only the existence of pov- erty traps but also the reasons that countries caught in such traps may have such exceptionally low incomes compared with high-income countries.

The O-Ring Model

The key feature of the O-ring model is the way it models production with strong complementarities among inputs. We start by thinking of the model as describing what is going on inside a firm, but as we will see, this model also provides valuable insights into the impact of complementarities across firms or industrial (product) sectors of the economy.

Suppose that a production process is broken down into n tasks. There are many ways of carrying out these tasks, which for simplicity we order strictly by level of skill, q, required, where 0 … q … 1. The higher the skill is, the higher the probability that the task will be “successfully completed” (which may mean, for example, that the part created in this task will not fail). Kremer ’s concept of q is quite flexible. Other interpretations may include a quality index for characteristics of the good: Consumers would be willing to pay more for higher-quality characteristics. For example, suppose that q = 0.95. Among other interpretations, this can mean (1) that there is a 95% chance that the task is completed perfectly, so the product keeps maximum value, and a 5% chance that it is completed so poorly that it has no value; (2) that the task is always completed well enough that it keeps 95% of its maximum value; or (3) that the product has a 50% chance of having full value and a 50% chance of an error reducing the value of the product to 90%. For simplicity, assume that

188 PART onE Principles and Concepts

the probability of mistakes by different workers is strictly independent. The production function assumed is a simple one: Output is given by multiplying the q values of each of the n tasks together, in turn multiplied by a term, say, B, that depends on the characteristics of the firm and is generally larger with a larger number of tasks. Suppose also that each firm hires only two workers. Then the o-ring production function looks like this:43

BF(q i q j) = q i q j (4.1)

That is, to make things simple, for this exposition we let the multiplier, B, equal 1. In addition to the form of the production function, we make three other significant types of simplifying assumptions: (1) Firms are risk-neutral, (2) labor markets are competitive, and (3) workers supply labor inelastically (i.e., they work regardless of the wage). If we consider capital markets, we assume that they are competitive as well. For now, we also assume that the economy is closed.

One of the most prominent features of this type of production function is what is termed positive assortative matching. This means that workers with high skills will work together and workers with low skills will work together. When we use the model to compare economies, this type of matching means that high-value products will be concentrated in countries with high-value skills. In this model, everyone will like to work with the more productive workers, because if your efforts are multiplied by those of someone else, as they are in Equation 4.1, you will be more productive when working with a more produc- tive person. In competitive markets, your pay is based on how productive you are. A firm with a higher-productivity worker can more afford to pay a higher wage and has the incentive to bid higher to do so, because the value of output will be higher with two productive workers, say, than with one low- and one high-productivity worker. As a result, there will be a strong tendency for the most productive workers to work together.

This can be seen easily if we imagine a four-person economy. Suppose that this economy has two high-skill qH workers and two low-skill qL workers. The four workers can be arranged either as matched skill pairs or unmatched skill pairs. Total output will always be higher under a matching scheme because

q2H + qL2 7 2qHqL (4.2)

Recall that (x − y)2 7 0 for any x that is not the same as y, so let x stand for qH and y stand for qL. Then x

2 + y2 7 2xy, the same as in Equation 4.2. (Or try this by plugging in any values qH 7 qL.) This generalizes to larger numbers of workers in the firms and the economy; the result is that workers sort out by skill level.44

Because total value is higher when skill matching rather than skill mixing takes place, the firm that starts with high-productivity workers can afford to bid more to get additional high-productivity workers, and it is profitable to do so. Of course, every firm would like to hire the most productive worker, but it would be in that worker’s interest to team up with other high-productivity workers. Think of firms being formed while workers try to determine for which firm they want to work. After the high-productivity workers pair off, they are out of the picture. The less productive workers are then stuck with each other. If there are many classes of skill or productivity, first the highest-skill workers

o-ring production function A production function with strong complementarities among inputs, based on the products (i.e., multiplying) of the input qualities.

189CHAPTER 4 Contemporary Models of Development and Underdevelopment

get together, then the next highest, and so on, such that skill matching results as a cascading process. For example, a symphony orchestra will be adversely affected as a whole by hiring one single poor performer. So an otherwise excel- lent orchestra has every incentive to bid the most for an outstanding performer to replace the poor performer. Similarly, the best jazz performers play and record together rather than each leading a group of poorer players. The restau- rant with the very best chef also hires mature, highly trained, full-time waiters, while a fast-food restaurant does not hire a famous chef.

This sorting process is perhaps most vividly easy to remember by analogy to Nobel laureate Gary Becker ’s famous “marriage market” model, which is a somewhat different case45 but offers some additional intuition. If prospec- tive spouses care only about attractiveness, every man wants to marry the most attractive woman, and every woman wants to marry the most attractive man, so the most attractive man and woman will marry. They are now out of the picture, so next, the second most attractive man and woman marry. This process continues until the least attractive man and woman marry. Of course, beauty is in the eye of the beholder, and most people care about things besides attractiveness in a mate such as kindness, intelligence, wealth, beliefs, inter- ests, commitment, and sense of humor; but the marriage model serves as a memorable analogy. The result in the business world is that some firms and workers, even an entire low-income economy, can fall into a trap of low skill and low productivity, while others escape into higher productivity.

Although this model may seem abstract, a numerical example can show how the firms with high-skill workers can and will pay more to get other high- skill workers or will have more incentive to upgrade skills among existing workers. Suppose that there are six workers; three have q = 0.4 and are grouped together in equilibrium, while the other three have q = 0.8. Now suppose that the q of one of the workers in the first firm rises from 0.4 to 0.5 (perhaps due to training). Similarly, suppose the q of one worker in the second firm rises from 0.8 to 1.0. In each case, we have a 25% increase in the quality of one worker. As you may expect, a 25% increase in the quality of one worker leads to a 25% increase in output quality. But starting from a higher level of quality, that 25% clearly translates into a much larger point increase: In the example, the first firm goes from (0.4)(0.4)(0.4) = 0.064 to (0.4)(0.4)(0.5) = 0.080; this is a difference of 0.080 − 0.064, which is a point change of 0.016; and 0.016/0.064 = 0.25, which is a 25% increase. For the second firm, we move from (0.8)(0.8)(0.8) = 0.512 to (0.8) (0.8)(1.0) = 0.640; the change in this case is 0.128, which is again 25%. However, the point value of the increase is much greater—eight times greater—for a dou- bled point-value investment (0.2 in the second firm versus 0.1 in the first firm). If a firm can increase quality in percentage terms at constant marginal cost or even a not too quickly rising cost, there is a virtuous circle in that the more the firm upgrades overall, the more value it obtains by doing so. Accordingly, wages will increase at an increasing rate as skill is steadily raised. As Kremer shows, the O-ring model is consistent with competitive equilibrium.

The O-ring result of positive assortative matching relies on some rather strong assumptions. How important are each of these, and how much can they be relaxed? Two points are crucial: (1) Workers must be sufficiently imperfect substitutes for each other, and (2) we must have sufficient complementarity of tasks. As long as these conditions hold, the basic results will follow.

190 PART onE Principles and Concepts

To see why workers must be imperfect substitutes, suppose they were per- fect substitutes. Specifically, suppose there are two skill levels, qL and qH = 2qL, so every qH worker can be replaced by two qL workers with no other change. Thus qH workers will be paid twice the amount that qL workers are paid. We can draw no predictions about what combination of worker skill levels a firm—or an economy—will use, so we can learn nothing about low-skill-level equilibrium traps. In fact, there is empirical evidence for imperfect substitut- ability across worker types in firms.

To see why we must have complementarity of tasks, suppose that there were two tasks indexed by g and h but with no complementarity between them. To be specific, suppose that our qH worker is hired for the g task, and a qL worker is hired for the h task; then

F(q Hq L) = g (q H) + h (q L)

Here skills are imperfect substitutes for each other, because only one type of worker can be hired for each task (i.e., no two-for-one type of substitution is possible here). However, because tasks are not complementary, the optimal choice of skill for the g task is independent of that of the h task, and again no strategic complementarities are present.46

Implications of the O-Ring Theory

The analysis has several important implications:

• Firms tend to employ workers with similar skills for their various tasks.

• Workers performing the same task earn higher wages in a high-skill firm than in a low-skill firm.

• Because wages increase in q at an increasing rate, wages will be more than proportionally higher in developed countries than would be predicted from standard measures of skill.

• If workers can improve their skill level and make such investments, and if it is in their interests to do so, they will consider the level of human capital investments made by other workers as a component of their own decision about how much skill to acquire. Put differently, when those around you have higher average skills, you have a greater incentive to acquire more skills. This type of complementarity should by now be a familiar condition in which multiple equilibria can emerge; it parallels issues raised in our analysis of the big push model. Kremer shows that a graph like Figure 4.1 can apply to choices about how much skill to acquire.

• One can get caught in economy-wide, low-production-quality traps. This will occur when there are (quite plausibly) O-ring effects across firms as well as within firms. Because there is an externality at work, there could thus be a case for an industrial policy to encourage qual- ity upgrading, as some East Asian countries have undertaken in the past (see Chapter 12, section 12.6, and its end-of-chapter case study of South Korea). This could be relevant for a country trying to escape the middle-income trap.

191CHAPTER 4 Contemporary Models of Development and Underdevelopment

• O-ring effects magnify the impact of local production bottlenecks because such bottlenecks have a multiplicative effect on other production.

• Bottlenecks also reduce the incentive for workers to invest in skills by lowering the expected return to these skills.

Following Kremer, consider a simple illustration of these bottleneck effects. Suppose that n tasks are required to produce a good. Let q be the standard skill level of these n tasks. But now let the actual skill level of two workers be cut in half in all firms. With an O-ring production function, output would fall by 75% (the result of cutting output in half once and then again). But then the marginal product of quality also falls by 75% for all the remaining n − 2 tasks, and thus so does the incentive to invest in increasing skill. The strong assump- tion of our simple O-ring production function may overstate the case, but the point that strategic complementarities can cause low-skill equilibria remains.

As workers reduce their planned skill investments, this further reduces the level of skill in the economy and thereby lowers further the incentive to invest in skill. To some extent, such bottlenecks could be ameliorated by inter- national trade and investment, because foreign inputs and investors provide an alternative source of inputs from outside the bottlenecked economy. One explanation of why economies that have cut themselves off from the interna- tional economy, such as India or China before the 1980s, have not fared as well as those that are more integrated, such as South Korea, could well be their failure to take advantage of foreign inputs or investments; the O-ring analysis helps explain why the impact could be so great. Trade cannot solve all prob- lems of industrialization, but the O-ring model helps explain why trade can play a key role as a part of an industrialization strategy.

The model also has implications for the choice of technology. When skill is scarce, a firm is less likely to choose a technique with higher value but compli- cated production technology with many tasks, because the costs of doing any one of those tasks poorly are magnified. In this way, the value of production is increas- ing in the complexity of the product, assuming that the product is completed suc- cessfully. Given positive assortative matching, firms producing products or using technologies that must be deployed at large scale or many steps will be induced to employ high-quality employees. Mistakes are costly to firms with large numbers of workers and production steps; therefore, such firms place exceptional value on high-quality, skilled workers who are unlikely to make mistakes.47 This indicates one reason why rich countries with high-skill workers tend to have larger firms and specialize in more complex products; it also helps explain why firm size and wages are positively correlated within and across countries.

Finally, under some additional assumptions, the model can also help explain the international brain drain. It is often observed that when a worker of any given skill moves from a developing to a developed country, he or she immediately receives a higher wage for using those same skills. A version of the O-ring model is one way of explaining this.

Thus Kremer ’s O-ring model points out many of the implications of strong complementarities for economic development and the distribution of income across countries. As Kremer concludes, “If strategic complementarity is suf- ficiently strong, microeconomically identical nations or groups within nations could settle into equilibria with different levels of human capital.”48

192 PART onE Principles and Concepts

4.6 Economic Development as Self-Discovery

In simple models with perfect information, it is assumed that firms, and devel- oping economies as a whole, already know their comparative advantage. But individuals must discover their own comparative advantage in labor markets; for example, no one is born knowing they are well suited to become an econo- mist or international development specialist. Somewhat analogously, nations must learn what activities are most advantageous to specialize in. As Ricardo Hausmann and Dani Rodrik show, this is a complex task—and one prone to market failure.49 It is not enough to tell a developing nation to specialize in “labor-intensive products,” because even if this were always true, there are a vast number of such products in the world economy of today, and underly- ing costs of production of specific products can differ greatly from country to country. So it is socially valuable to discover that the true direct and indirect domestic costs of producing a particular product or service in a given country are low or can be brought down to a low level. It is valuable in part because once an activity is shown to be profitable, it can usually be imitated, at least after some lag, spawning a new domestic industry. An example is the ready- made garment industry in Bangladesh, which spread from the first pioneers as dozens of entrepreneurs entered the market. But as markets are eventu- ally open to competing firms, they will take away potential profits from the original innovator. And since, due to this information externality, innovators do not reap the full returns generated by their search for profitable activities, there will be too little searching for the nation’s comparative advantage—too much time carrying on with business as usual and too little time devoted to “self-discovery.” The term self-discovery somewhat whimsically expresses the assumption that the products in question have already been discovered by someone else (either long ago, or recently in a developed economy); what remains to be discovered is which of these products a local economy is rela- tively good at making.

Hausmann and Rodrik also point out another market failure: There can be too much diversification after the point where the nation discovers its most advantageous products to specialize in. This is because there may be an extended period in which entry into the new activity is limited. Hausmann and Rodrik conclude that in the face of these market failures, government policy should counteract the distortions by encouraging broad investments in the modern sector in the discovery phase. In fact, they also argue that policy should in some cases work to rationalize production afterward, encouraging movement out of higher-cost activities and into the lower-cost activities, paring down industries to the ones with the most potential for the economy. The authors draw parallels with some of the successful export and industrial policy experiences of East Asia, a topic to which we will return in Chapter 12.

The authors note three “building blocks” of their theory: There is uncer- tainty about what products a country can produce efficiently; there is a need for local adaptation of imported technology so that it cannot be used productively “off the shelf”; and once these two obstacles have been overcome, imitation is often rapid (reducing the profitability of pioneers). They present a number of case examples that show the reasonableness of each of these assumptions

Information externality The spillover of information— such as knowledge of a production process—from one agent to another, without intermediation of a market transaction; reflects the public good characteristic of infor- mation (and susceptibility to free riding)—it is neither fully excludable from other uses, nor nonrival (one agent’s use of information does not pre- vent others from using it).

193CHAPTER 4 Contemporary Models of Development and Underdevelopment

in practice, such as the unexpected emergence of the information technology industry in India and the surprising differences in the exports from various countries with similar apparent comparative advantages, such as Bangladesh (hats but not bedsheets) and Pakistan (bedsheets but not hats); the history of local adaptations of various types of Western technology in East Asia (such as shipbuilding in South Korea); and the rapid diffusion of new products and techniques in the local economy (often facilitated by the movement of per- sonnel across firms), as seen in the growth of the cut-flower export industry in Colombia.

4.7 The Hausmann-rodrik-Velasco growth Diagnostics Framework

Encouraging efficient investment and widespread entrepreneurship plays a prominent role in accelerating growth and promoting development more broadly. But the once popular idea of finding a “one size fits all” policy for eco- nomic development is now generally recognized as a myth. Different countries face different binding constraints on achieving faster rates of growth and eco- nomic development. A key mission for economic development specialists is to help determine the nature of the constraints for each country. Ricardo Haus- mann, Dani Rodrik, and Andrés Velasco (HRV) propose a growth diagnostics decision tree framework for zeroing in on a country’s most binding constraints on economic growth. HRV explain that targeting the most binding constraint has important advantages over other approaches to policy selection.50

If a developing nation experiences a relatively low level of private invest- ment and entrepreneurship, what steps should it take? The basic decision tree for addressing this question is seen in Figure 4.3, with arrows leading to the ten bottom boxes (that is, the boxes from which no arrows extend further). At the first stage of the tree, the analyst seeks to divide countries between those for which the main problem is a low underlying rate of return and those for which the problem is an abnormally high cost of finance. Let us consider the former case first, following the left arrow pointing to Low return to economic activity.

Low returns to investors may be due to the fact that there are intrinsi- cally low underlying social returns to economic activities. Alternatively, low returns may be caused by what is termed low private appropriability, meaning limited ability of investors to reap an adequate share of the rewards of their otherwise profitable investments. Considering these cases in turn, low social returns may be caused by one of three factors.

First, as noted in Chapter 2, poor geography such as tropical pests, moun- tains, and other physical barriers, distance to world markets, and landlocked status (which may render port access politically dubious or economically costly) may limit the ability of a low-income country to initiate and sustain economic development, especially when other compounding factors are pres- ent. When these constraints are most binding, development policy must initially focus on strategies for overcoming them. Second, low human capital— skills and education as well as health of workers—are complementary with other factors in production, affecting the returns to economic activity. For

Growth diagnostics A deci- sion tree framework for iden- tifying a country’s most bind- ing constraints on economic growth.

Social returns The profit- ability of an investment in which both costs and benefits are accounted for from the perspective of the society as a whole.

194 PART onE Principles and Concepts

example, if economic returns are most affected by lack of literacy and numer- acy, this becomes a development policy priority. (The importance of health and education was also stressed in Chapter 2, and this will be examined in depth in Chapter 8.) Third, every developing nation must provide the vital infrastructure needed to achieve and sustain a modern economy, beginning with basic physical structures such as roads, bridges, railroads, ports, telecom- munications, and other utilities. With bad infrastructure, otherwise high-return economic activities may prove unprofitable. In some countries, inadequate and imbalanced infrastructure is the main factor preventing an acceleration of growth, and in such cases, policies focusing on providing it would boost investment and growth the most.

But the problem may lie not with the underlying social return to economic activities but with low appropriability, meaning that investors cannot reap an adequate share of returns to investment. We get to low appropriability from the right arrow emanating from Low return to economic activity. In turn, appro- priability problems can be due to either government failures or market failures. In the HRV diagram, government failures are divided between micro risks and macro risks. Micro risks address fundamental institutional weaknesses such as inadequacy of property rights, government corruption, and exces- sively high effective taxation. That is, the return to economic activity may be high enough, but elites rather than investors may capture a large fraction of the returns and make investments unattractive. Despite the difficulty of effectively reforming institutions when reform threatens the interests of elites (see Chapter 2), such reform must become the development priority when micro risks are binding. As the case study of China at the end of this chapter

Bad international finance

Low social returns

Poor geography

Low human capital

Low domestic

saving

Poor intermediation

Macro risks: financial, monetary,

fiscal instability

Micro risks: property rights,

corruption, taxes

Information externalities:

“self-discovery”

Coordination externalities

Bad infrastructure

Government failures

Market failures

Bad local finance

Low return to economic activity

Low appropriability

Possible causes

High cost of finance

Source: Ricardo Hausmann, Dani Rodrik, and Andrés Velasco, “Getting the diagnosis right,” Finance and Development 43 (2006), available at http://www.inf.org/external/ pubs/ft/fandd/2006/03/hausmann.htm. Reprinted with permission.

FigUrE 4.3 Hausmann-rodrik-Velasco growth Diagnostics Decision Tree

Problem: Low levels of private investment and entrepreneurship

195CHAPTER 4 Contemporary Models of Development and Underdevelopment

demonstrates, reform can sometimes be accomplished in stages through tran- sitional institutions. Appropriability may also be limited by macro risks—the failure of government to provide financial, monetary, and fiscal stability.

The fundamental problem may also be large-scale market failures of the type stressed in this chapter. These may include the self-discovery problems pointed up by Hausmann and Rodrik and reviewed in section 4.6. They may also take the form of coordination externalities, such as seen in the big push model of underdevelopment, examined in section 4.3. Other types of market failure and government failure are examined in Chapter 11.

In yet other cases, the main problem may not be underlying low rates of return but rather an abnormally high cost of finance. The possibilities are out- lined following the right arrow from the top box in Figure 4.3 to High cost of finance. Here the problem may be bad international finance—inadequate access to foreign sources of capital or problems with debt, examined in Chapter 13; or the problem may reside in bad local finance, due either to low availability of loanable funds through domestic financial markets, traced to low domestic sav- ing, or to poor intermediation owing to an inadequate or overregulated banking system that is unable or unwilling to channel funds to the economic activi- ties with high returns. These also lead to other policy challenges, examined in Chapter 15.

In sum, one size does not fit all in development policy. Economic develop- ment strategies focusing on resource mobilization through foreign assistance and other capital flows, along with increased domestic national saving, can be most effective when domestic returns are both high and privately appropri- able. In contrast, strategies focusing on market liberalization and opening up the economy can be most effective when social returns are high and the most serious obstacle to private appropriation are government-imposed excessive taxes and restrictions. Finally, strategies focusing on industrial policy (elabo- rated on in Chapter 12) can be most effective when private returns are low, not because of what a government does (errors of commission), but because of what a government does not do (errors of omission).

HRV illustrate their approach with case studies of El Salvador, Brazil, and the Dominican Republic. They argue that each case exhibits a different “diag- nostic signal” of the most binding constraint, as seen in Box 4.3. HRV stress that an approach to development strategy that determines one or two policy priorities on this diagnostic basis will be more effective than pursuing a long laundry list of institutional and governance reforms that may not be targeted toward the most binding constraints.

It is often difficult to observe a binding constraint directly. In practice, growth diagnostics usually involves some economic detective work. To evalu- ate whether a proposed constraint is binding, a growth diagnostician looks for evidence on its implications. If the constraint is excessive taxation, we can expect to see high movement into the informal sector or underground econ- omy. If the constraint is infrastructure, we can expect to see significant conges- tion. If the constraint is education, we can expect to see high rates of return to education. In general, the analyst looks for economic behavior consistent with agents trying to get around a constraint.

Growth diagnostics is also subject to some limitations and criticisms. One implicit assumption is that development can be equated with growth, which

196 PART onE Principles and Concepts

in turn is held back by investment. This is a useful analytical assumption for this and a range of other purposes, but it does not and cannot provide a com- plete understanding of development purposes, mechanisms, and constraints. And of course, it is often not a simple matter to find a single binding con- straint. There can be uncertainty about the “position” of each constraint in the economy, so we can only make a probabilistic assessment of which one is binding. If there are important complementarities between two investments,

BOX 4.3 FiNDiNgS Three Country Case Study Applications of growth Diagnostics

El Salvador

HRV argue that this economy is constrained by a lack of productive ideas. The binding constraint is a lack of innovation and demand for investment to replace the traditional cotton, coffee, and sugar sectors, or low “self-discovery.” So the best strategy focus for El Sal- vador would be to encourage more entrepreneurship and development of new business opportunities.

Brazil

HRV identify the country’s binding constraint as lack of sufficient funds to invest despite an abundance of productive ideas. They argued that private returns in Brazil are high, and therefore other flaws (inad- equate business environment, a low supply of infra- structure, high taxes, high prices for public services, weak contract enforcement and property rights, and inadequate education) are not as binding in Brazil. So investment is instead constrained by Brazil’s inability to mobilize sufficient domestic and foreign savings to finance needed investments at reasonable interest rates. Although Brazil could increase national savings to a degree by reducing government expenditure, this might not be politically feasible. If so, HRV suggest that higher taxes and user fees and lower infrastruc- ture and human capital subsidies might work. “If the country can move to a faster growth path and if waste does not grow with GDP, it may outgrow its burdens and gradually improve its tax and spending system as fiscal resources become more abundant.” In subse-

quent work, Hausmann has emphasized the impor- tance of “creating a financially viable state that does not over-borrow, over-tax or under-invest” to success- fully raise domestic savings.

Dominican republic

HRV conclude that the Dominican Republic is con- strained by core public goods in product sectors key for growth. The country began a new reform sequence dur- ing the 1980s, after it could no longer rely on sugar and gold exports. It followed a narrow strategy of investing in needed public goods for two emerging product (or service) sectors with high potential, tourism and ma- quila assembly manufacturing. The keys were security and infrastructure near the main tourist destinations and special trade policy benefits for the light manufac- turing assembly (maquila) sector. As the economy grew from these sources, other constraints were hit, notably in the financial sector; getting past them (particularly a costly financial crisis) was bumpy, but the binding con- straints stayed or became visible, so policymakers could focus on relaxing them to keep growth going.

Sources: Ricardo Hausmann, Dani Rodrik, and Andrés Velasco, “Growth diagnostics,” in One Economics, Many Recipes: Globalization, Institutions, and Economic Growth, by Dani Rodrik (Princeton, N.J.: Princeton University Press, 2007), ch. 2; Ricardo Hausmann, “In search of the chains that hold Brazil back,” October 31, 2008, http://papers. ssrn.com/sol3/papers.cfm?abstract_id=1338262. An excel- lent practicum is found in “Doing Growth Diagnostics in Practice: A ’Mindbook.’” See http://www.cid.harvard.edu/ cidwp/177.html. The World Bank offers a set of growth diagnostics exercises at its Web site, http://web.world- bank.org/.

197CHAPTER 4 Contemporary Models of Development and Underdevelopment

combining them (in some sense) should be considered. Further, the fact that one constraint is not binding today does not mean that we can neglect it when there are long gestation periods before current investments become produc- tive. For example, consider investments in education: Students require several years of schooling followed by experience before these investments become productive. So although education may not be binding for a particular coun- try such as Bolivia at a particular point in time, this does not mean that it will not become binding at a later time; in response, we may need to make invest- ments today. Clearly, identifying and addressing constraints that are likely to become binding in the future is even more challenging than targeting today’s more visible bottlenecks.

Growth diagnostics has already had an effect on the work of develop- ment agencies. For example, the Inter-American Development Bank (IDB), the regional development bank for the western hemisphere, has been com- missioning growth diagnostic studies of many member economies while training staff and nationals in the skills needed to conduct their own growth diagnostics. World Bank economists have applied the method in a dozen country pilot studies in Africa, Asia, and Latin America. And develop- ing country scholars have applied the approach to their own countries. Although growth diagnostics might be criticized as “more art than science,” at the very least this new approach forces the analyst to focus on country- specific circumstances and thus to get to know the individual country very well. This is one of the reasons that growth diagnostics offers a valuable complement to econometric studies.

4.8 Conclusions

The important point is not that people keep doing inefficient things. This is not in itself very surprising. The deeper point is that people keep doing ineffi- cient things because it is rational to keep doing them, and it will remain ratio- nal as long as others keep doing inefficient things. This leads to a fundamental problem of coordination failure. Sometimes firms and other economic agents will be able to coordinate to achieve a better equilibrium on their own. But in many cases, government policy and aid will be necessary to overcome the resulting vicious circles of underdevelopment.

The purpose of economic development theory is not only to understand underdevelopment but also to devise effective policies to redress it. The analy- sis of coordination failure problems in this chapter affirmed that early develop- ment theorists such as Paul Rosenstein-Rodan identified important potential problems that are ignored in conventional competitive equilibrium models.51 The new perspectives offer some important overall lessons for policy, but they are not simple lessons with easy applicability, and indeed they present some- thing of a two-edged sword. On one side, the analysis shows that the potential for market failure, especially as it affects the prospects for economic devel- opment, is broader and deeper than had been fully appreciated in the past. Rather than the small “deadweight triangle losses” of conventional economic

198 PART onE Principles and Concepts

analysis of monopoly, pollution externalities, and other market failures, coor- dination failure problems can have more far-reaching effects and consequently much greater costs.52 For example, the interactions of slightly distorted behav- iors by potential investors failing to consider the income effects of the wages they pay may produce very large distortions, such as the outright failure to industrialize. This makes the potential benefit of an active role for government larger in the context of multiple equilibria.

The coordination failures that may arise in the presence of complemen- tarities highlight potential policies for deep interventions that move the economy to a preferred equilibrium or even to a higher permanent rate of growth that can then be self-sustaining. For example, once a big push has been undertaken, government coordination may no longer be necessary. The unaided market can often maintain industrialization once it is achieved, even when it cannot initiate or complete the process of industrialization. For another example, we will see in Chapter 8 that in some cases, the presence of child labor represents a kind of bad equilibrium among the families with children who work, one that might be fixed with appropriate policy. After successfully abolishing child labor, it is possible that the regulations will not have to be actively enforced to keep child labor from making a resurgence (because most parents send their children to work only because they have to). If there is no incentive to go back to the behavior associated with the bad equilibrium, government has no need to continue the interventions. Instead, government can concentrate its efforts on other crucial problems in which it has an essential role (e.g., problems of public health). This onetime-fix character of some multiple-equilibria problems makes them worthy of spe- cial focus because they can make government policy much more powerful in addressing problems of economic development. Among other implications, the prospect of deep interventions can mean that the costs of implementing policy can be reduced and that carefully targeted development assistance could have more effective results.

The other edge of the sword, however, is that with deep interventions, the potential costs of a public role become much larger. Policy choices are more momentous because a bad policy today could push an economy into a bad equilibrium for years to come. This is because government can be a major part of the problem, playing a key role in perpetuating a bad equilibrium such as a high-corruption regime, in part because some government officials and politi- cians may benefit personally from it. Bad policy can even initiate a move to a worse equilibrium than a country began with. To expect government to be the source of reform that moves the economy to a better equilibrium in countries where government has been part of the complex nexus of a bad equilibrium can be naive. For example, as the 2001 Nobel laureate Joseph Stiglitz pointed out, development officials should have been more suspicious of corrupt government officials’ embracing of the World Bank’s doctrine of thorough- going privatization in the late 1980s and early 1990s. Why would corrupt officials have done so if they benefited from a stream of rents captured from public enterprises? The answer, Stiglitz suggests, is that these officials found that by corrupting the process of privatization, they could get not only a stream of corrupt rents from the annual operations of the enterprise but also

199CHAPTER 4 Contemporary Models of Development and Underdevelopment

a share of the present discounted value of the whole future operations of the enterprise.53 The results of corrupt privatization in Russia in particular have been devastating for its economy, preventing it from enjoying the benefits of the market and potentially keeping it in a suboptimal equilibrium for many years to come. Even when a government is not corrupt, the potential impact of a well-intentioned but flawed government policy is much greater when it can push the economy to a fundamentally different equilibrium, which may be difficult to reverse. This is all the more problematic in the many cases in which “history matters” in a developing economy—that is, when past condi- tions determine what is possible today.

Both government failure and market failure (including coordination prob- lems and information externalities) are real, but public- and private-sector con- tributions to development are also vital. Therefore, we need to work toward the development of institutions in which actors in the public and private sec- tors have incentives to work productively together (directly and indirectly) in such a way as to create the conditions necessary to break out of poverty traps. In achieving this goal, the international community also has a vital role to play, providing ideas and models and serving as a catalyst for change, as well as providing some of the necessary funding.

The growth diagnostics approach is a valuable tool for domestic and international analysts who start with a detailed understanding of a develop- ing country; it can be helpful in identifying binding constraints on national growth and the policy priorities to address them.

In sum, the contributions of the new theories of development reviewed in this chapter include a better understanding of the causes and effects of pov- erty traps, achieved by more precisely pinning down roles of different types of strategic complementarities, explaining the role of expectations, clarify- ing the importance of externalities, illuminating the potential scope for deep interventions, and improving our understanding of both the potential role of government and the constraints on the effectiveness of that role—when gov- ernment itself becomes a player in an underdevelopment trap. Finally, the new approaches point out more clearly the real potential contributions of outside development assistance that extend beyond provision of capital to modeling new ways of doing things.

As democratic government spreads in the developing world, the new understandings of underdevelopment traps can make for a more effective guide to policy design than was available even a few years ago. As Karla Hoff has aptly summarized, “Governments fail, even in democracies, just as markets do. But a positive development of recent years is to try more limited interventions to harness the spillovers among agents, and to try to sequence policy reforms in a way that makes it more likely for good equilibria to emerge.” 54

In Parts Two and Three, as we consider pressing issues affecting devel- oping countries today, we will be using the insights provided by both the classic theories and the new models of development and underdevelop- ment to inform our understanding of both the nature of the problems faced and the potential benefits and pitfalls of policies designed to help over- come them.

200 PART onE Principles and Concepts

An Extraordinary Performance

From 1978 to 2011, the economy of China grew at an average rate of close to 9% a year, an unprece- dented achievement for any economy in history, let alone the world’s most populous nation, with over 19% of global population. China’s income per cap- ita by 2012 was approaching six times what it was in 1978, when reforms began. Growth was three times the rate that would be considered respectable by the recent standards of most low-income countries.

China has also experienced the world’s most dramatic reductions in poverty. The World Bank’s most recent estimate is that just 12% of China’s population lives on less than $1.25 per day (27% below $2 per day). This means that hundreds of millions fewer people were living in extreme pov- erty in a span of just three decades. Reductions in extreme poverty in China are far faster and greater than anywhere else in the world.

Debate on Sources of Success For such a stunning record, the roots of China’s suc- cess remain a source of disagreement. The Chinese experience seems to change everything—but does it? And if so, in what ways? Success has a thou- sand fathers, and all the major traditional and new schools of thought on development want to claim China as their most important case in point. China is hailed as an example of the benefits of markets, trade, and globalization. Yet by conventional mea- sures, institutions in China remain quite weak. For example, the World Bank’s 2013 “Ease of Doing Busi- ness” index ranks China poorly, at No. 96—worse than Russia, Mongolia, Zambia, or Serbia. Manufac- tured exports are a key to China’s growth, and market incentives have played a primary motivational role in

business decisions. But China has also adopted activ- ist industrial policies, pushing exports of increasingly higher skill and technology content, and it embarked on its period of rapid growth around 1980, more than a decade before significant trade liberalization. But often overlooked is that China’s agricultural productivity growth was also very high. Moreover, much of China’s growth in the 1980s and early 1990s was due to rural township and village enterprises, which had a quasi-cooperative and quasi-municipally owned character. There has been less privatization of state-owned enterprises than in most developing countries. In the meantime, countries in Africa, Latin America, and elsewhere that have most closely fol- lowed the free-market model have generally not done particularly well. While all schools may find some- thing in China to let them claim it as vindication of their favored development policies, it is also clear that if China were performing dismally, each could (and likely would) find reasons why its own theories, including free-market theory, predicted such a failure.

There have been many special explanations for China’s remarkable success. Many of them contain part of the truth, but such dramatic success is more than the sum of these parts. Let us review some of the explanations.

Regional “Demonstrations.” The presence of regional “demonstration” models has been crucial. Japan was emulated by other countries in the East Asian region. Hong Kong provided an additional example for China, as did China’s archrival Taiwan. Taiwan, Hong Kong, and South Korea focused on export-oriented industrialization strategy at a time when world trade was growing rapidly (see the end-of-chapter case studies for Chapters 12 and 13).

Case Study 4

Understanding a Development Miracle: China

200

201CHAPTER 4 Contemporary Models of Development and Underdevelopment

Leveraging the Lure of a Billion Consumers By the late 1980s, the locus of regional growth shifted to China as investors began to pour investments into China in large part because of the allure of its eventual market of more than 1.3 billion consum- ers. Government played off potential investors who wanted access to China’s consumers, demanding and getting extensive technology transfer, public and private Chinese business partnerships, local content, and other concessions in exchange for the right to sell to Chinese citizens. Although the market was limited at first by both low incomes and government policies, early investors found high incentives to export from several special eco- nomic zones on the southeast coast. These inves- tors discovered that China offered very cheap labor with unusually high skills and work habits for its income level.

Export-Led Investment and Growth Once early investments built up a sufficient critical mass, agglomeration benefits of concentrated economic activity kicked in (see Chapter 7). The more pro- ducers located in China, the greater the benefits for an increasing number of suppliers to operate there. At this point, investments started to feed on them- selves in a cumulative causation. In the meantime, when wages began to rise, companies could set up production farther west, or migrants from the west could move to the new industrial centers. Given China’s population of hundreds of millions of low- income farmers, expectations were formed that this process of wage restraint could continue for an extended time—although a string of wage increases beginning in 2010 began to challenge these expec- tations as financial analysts argued that the Lewis turning point had been reached (see Chapter 3).

Coordination After the bloody crackdown on Tiananmen Square protests in 1989, there was con- siderable doubt about whether the reforms would continue and therefore whether investment and growth would remain high (making other invest- ments profitable). The Chinese leader Deng Xiaop- ing paid a 1991 visit to the southern China regions that had been leading in growth and reform and proclaimed, “You should be bolder and develop faster.” A rapid burst of investment and growth, as well as policy reform, followed his speech and its subsequent publication. It has been suggested that in effect this served to coordinate expectations and

led to the shift from a lower-growth to a higher- growth equilibrium. But much more generally, the government of China has used its centralized authority to coordinate investments across indus- tries. Moreover, government negotiation of licenses and other business agreements helped ensure that China got more favorable deals than many other developing countries that relied on private com- pany-level business transactions, although in this, the role model lessons from South Korea was also a benefit.

Health and Education Investments The cen- tral planning of China’s first decades after its 1949 Communist revolution were by most measures a failure. Industry was highly inef- ficient. As many as 30 million people died in a late-1950s famine caused by poor central plan- ning decisions and political pressures that led party and government officials to regularly overstate the harvest prospects. As Amartya Sen stresses, famines rarely occur in democratic coun- tries with a free press. Such disasters were only partly offset by the early and ongoing emphasis on basic health and education in China and then on reductions of fertility through China’s one- child policy (see the case study for Chapter 6). But these basic first steps on education, health, and eventually fertility helped set the stage for growth and poverty reduction when later com- bined with market incentives. One of the results is the apparently higher educational and skill level of factory workers for given wages in China in comparison to its competitor countries.

Productivity Growth There has been consider- able debate about whether rapid growth in other East Asian countries is the result of capital accu- mulation or productivity gains. Alwyn Young, Paul Krugman, and others have concluded that South Korea and other Asian Tigers grew more from investing heavily in capital assets such as machinery and factories than by improved worker efficiency. Wing Thye Woo concluded that most of China’s growth came from the reallocation of labor, particularly from agriculture to other activities, and that sustainable total factor productivity progress was much lower, on the order of 2% per year.

But for the case of China, Zuliu Hu and Mohsin Khan concluded that productivity gains explained more than 42% of China’s growth in the formative

201

202 PART onE Principles and Concepts

1979–1994 period and that productivity had over- taken investment by the early 1990s as the largest source of growth. This was considered surprising, in part because of the breathtaking pace of capital investments in China. But on the other hand, when China’s rapid growth began in the late 1970s in the areas close to Hong Kong, while it was clear that a large volume of investment funds was flowing from capital-abundant Hong Kong (a British crown colony at the time) to capital-scarce China, the big- ger story was the flow of productive ideas over the Hong Kong border, a barrier that had long pre- vented the transfer of both capital and know-how. Of these two factors, it often seemed that the ideas were more important than the finance.

There is widespread concern that by now, China has entered an investment bubble stage in which many investments are of dubious quality, particu- larly in real estate and some infrastructure and industrial sectors. Even so, the rapid pace of devel- opment in China has been unprecedented.

Recent research by Xiaodong Zhu, Loren Brandt, and their coauthors has provided new documenta- tion that productivity growth, rather than mere fac- tor accumulation, has been a very important source of China’s rapid growth of output. In particular, Zhu has presented well-regarded evidence that productivity growth in the nonagricultural, non- state-owned sector is the most important source of growth in China. Noting that productivity is still well below that of the United States, he argues that there should still be significant opportunities for productivity to continue to grow rapidly in China by adopting foreign technology, learning best pro- duction practices, and improving institutions and policies, particularly to allocate capital more effi- ciently.

In another study, Ashoka Mody and Fang-Yi Wang of the World Bank examined the causes of industrial growth in China and concluded:

…much of the action came from region-specific influ- ences and regional spillovers. Regional influences included the open-door policies and special economic zones that successfully attracted investments from over- seas Chinese to particular locations. Existing regional strengths, especially high-quality human capital and infrastructure, also contributed to growth. Our results illuminate the interplay between conditions conducive for growth—for example, the contribution of foreign

expertise is greatly enhanced by available human capi- tal. China made judicious use of the advantages of back- wardness by targeting areas that were less developed and less encumbered by the legacy of existing institu- tions, although it was fortunate in this regard that the backward regions were in close proximity to Hong Kong and Taiwan.

Thus, the China case also illuminates complemen- tarities, a recurrent theme of this chapter.

Reform on the Margin As examined in detail in Chapter 2, developing inclusive institutions that protect property rights and enforce contracts, and place checks on execu- tive authority and the power of elites such as through the rule of law, have demonstrated impor- tance in long-term economic development. China appears to be an outlier, in that such protections are demonstrably weak. Yet it is extremely difficult to navigate the course from bad to good institutions. It is rarely possible to follow a straight line on the map, as a vortex of obstacles are encountered, and the ship of state itself may be the cause of many of the problems. The process of getting the institutions right is one of starting with a clear understanding of both formal and informal local rules, and mov- ing toward an eventual goal even when it cannot be seen clearly—in the presence of initial and then newly emerging constraints and opportunities, chartering what may seem to outsiders as large deviations off-course. A metaphor used by the post- Mao paramount leader Deng Xiaoping may also reflect in part this type of step-by-step, graduated process—“crossing the river by feeling the stones.”

In China, the way that market incentives were introduced and used seems to have been almost as important as the fact that they were introduced at all. One of the most important features of the past quarter century of economic history in China has been the very gradual implementation of reforms. China’s approach has been the opposite of that of many eastern European countries such as Russia and Poland, which opted for a “big bang,” a sud- den comprehensive changeover to a free-market economy. (Hungary and Slovenia are two coun- tries in that region that pursued a more gradualist strategy.) China has introduced new and transi- tional institutions that exist side by side with pre- vious institutions of central planning for extended

202

203CHAPTER 4 Contemporary Models of Development and Underdevelopment

periods. In the former Soviet Union and eastern Europe, central planning was abolished almost immediately, and economic depression, with drops in output of up to 50%, ensued before gradual recovery. In contrast, China kept the central plan- ning system partially intact for an extended period. Previous quotas for buyers and sellers at fixed planned prices were maintained. Reform was instead introduced on the margin. After filling their quotas, producers were free to buy and sell at mar- ket-determined prices; resales were generally not prohibited. This “dual-track” system simulated the allocational efficiency of a more competitive market economy and created strong incentives for firms to improve efficiency and increase output, in a man- ner less threatening to the status quo.

Moreover, while in other transition and devel- oping countries state-owned enterprises (SOEs) were sold off to private investors fairly quickly, in China these remained in government hands for an extended period. The government tried to reform them internally, with limited success. But at the same time, China has allowed and encouraged a new, more efficient sector to grow up around them. In recent years, China has privatized or closed many of the smaller SOEs. Many larger SOEs con- tinue to operate in a relatively inefficient manner, and some economists have suspected for years that their accumulating indebtedness will eventually pose significant financial risks to the economy. But the counterargument proposes that if the economy can continue to grow rapidly, it is also possible that China may stay ahead of this problem with- out experiencing a financial crash. Eventually, as employment opportunities continue to expand, more of the larger SOEs can be privatized or closed.

Further, for the first nearly two decades of reform, from the late 1970s to the mid-1990s, at the local level, township and village enterprises (TVEs) were encouraged. The TVEs were vaguely owned by local government, but their private entrepreneurs and employees held “vaguely defined” property rights, as Martin L. Weitzman and Chenggang Xu termed them. These TVEs accounted for a very large share of industrial output growth in China. Finally, after the Chinese economy had grown nearly fourfold, the majority of these TVEs were privatized in the late 1990s—by this point the private entrepreneurs had triumphed (or their underlying control became

clarified). But the TVEs played a unique role in spur- ring growth and spreading the benefits of develop- ment to rural areas.

Reforms in the late 1970s and 1980s favored agriculture and entrepreneurship in the rural areas where most of the poor lived, and poverty fell as income rose. From at least the early to mid-1990s, the terms of trade shifted toward industry and urban areas. Yasheng Huang makes a strong case that this represented an important turning point, associated with growing inequality and other seri- ous challenges.

Still, strong average growth continued through many changes. As outlined by Yingyi Qian, China’s transitional institutions have served a dual pur- pose: to improve efficiency while compensating the losers (and thereby preserving legitimacy or at least reducing the chance of political backlash). Provided that the quotas were enforced—and for the most part they seem to have been in the transition in China—the dual-track allocation system protected the interests of those who had benefited from and planned on receiving inputs at fixed, low prices. As a result, these agents did not oppose or undermine reforms and indeed could benefit further to the degree they could learn to produce more efficiently and operate in markets effectively. The system was largely phased out many years later, after the eco- nomic landscape had changed dramatically.

The vague local-government ownership of the TVEs provided protection for investors who feared government hostility toward private property and worried about expropriation. The impression that these companies were owned by the township or village protected the de facto private owners. Once reform proceeded to a certain point, these de facto owners were able to “take off the red hat,” as the saying went in China, and assume full ownership in exchange for considerations to local government, and taxes replaced direct revenue transfer out of the TVEs. Qian shows how similar arguments apply to fiscal and financial reforms. Under the reforms, local government continued to have a responsibility to provide revenue to the central government, but local government was allowed to keep a large share of collections on the margin before local and cen- tral revenue collection was fully separated. Govern- ment also allowed anonymous banking accounts for a long transition period, to credibly constrain

203

204 PART onE Principles and Concepts

the ability of the government to arbitrarily impose high individual taxes on successful entrepreneurs; Qian judged the program a success despite the fact that this diverges from what is considered normal best practice in advanced Western countries.

Yingyi Qian’s insightful explanation is:

The difference between China and Russia is not at all that China has established best-practice institutions and Russia has not. The difference lies in the institutions in transition.…The real challenge in reform facing transi- tion and developing countries is not so much know- ing where to end up, but searching for a feasible path toward the goal. Therefore, it focuses on transitional institutions, not best-practice institutions.…The gen- eral principle of efficiency-improving and interest-com- patible institutional change is simple, but the specific forms and mechanisms of transitional institutions often are not. Successful institutional forms usually are not a straightforward copy of best-practice institutions. They need not be and sometimes should not be. They need not be because room exists for efficiency improvement that does not require fine tuning at the beginning. They should not be because the initial conditions are country- and context-specific, requiring special arrangements.… Understanding these mechanisms sometimes needs an appeal to the counterintuitive second best argument, which states that removing one distortion may be coun- terproductive in the presence of another distortion.

Finally, for peasants in parts of China where the rural sector has done well, earlier land reforms have been among the causes—with the revolution setting the stage and the late-1970s reforms giving greater incentives to individual farmers. Land reform has been notoriously difficult to implement in other parts of the world. Remittances from migrant work- ers have fueled a service-sector boom in some rural areas, and prices received by farmers have gener- ally risen, particularly near urban areas.

China’s Coming Challenges China’s successes do need to be kept in perspective. Since 1980, China has grown about 4½ times faster than the United States, as measured by per capita output. As a result, China has been closing the rela- tive gap in living standards. In 1980, China’s income per person was only 2% of that in the United States, but by 2012, it had grown to over 15%. But even if China’s output per person continued to grow at its unprecedented recent rate of 8.4% and the United

States at its long-run rate of just 1.9%, China would still not catch up until close to 2040.

A high rate of domestic saving is associated with a trade surplus. Savings have been extremely high and rising in China. As of 2011, China was saving nearly half of its national income—an astounding and unprecedented rate compared to the coun- try’s own past rates (already a high 35% in 1990) and in relation to the high rates that have gener- ally prevailed in East Asia. Such high rates are not consistent with the pivot toward increasing local consumption as an engine of growth.

It is now generally accepted in China and inter- nationally that continuing to grow at such high rates is essentially impossible. Before China grew rapidly, South Korea did so, and before South Korea, Japan did. The later a country starts modern economic growth, the faster it can grow because the distance from traditional methods to the frontier technology of the day grows greater over time. But the pace of catch-up generally slows as an economy gets closer to the technology frontier and needs to innovate. Policymakers in China are actively preparing for this challenge. Despite its extraordi- nary record to date and considerable resources at its disposal, the substantial challenges that China faces in its attempt to reach developed country sta- tus should not be underestimated. There are some other limits and caveats to China’s success and to the lessons that other countries can learn from it.

Poverty and Vulnerability Life can indeed be harder than ever for the millions remaining in extreme poverty, such as rural peasants in some parts of the country facing the loss of security; offi- cial corruption, including reports of official land grabs from peasants; rising local taxes; and mini- mal improvements in technology or skills. At the same time, despite the growth in average wages, inequality in China—once quite low—has been rising dramatically; inequality has now reached approximately the same level as in the United States, worst among the developed countries.

Environment and Pollution Moreover, the environmental crisis in China is reaching epic pro- portions. A majority of the most polluted cities in the world are located in China, and health problems are growing. Water resource problems, erosion, and loss of habitat undermine the prospects for sustain- able development. The extreme air pollution is now

204

205CHAPTER 4 Contemporary Models of Development and Underdevelopment

causing not just misery but deaths and other serious and growing health problems. This reached historic proportions in the so-called Beijing “airpocalypse” of January 2013, when pollution indicators exceeded 40 times World Health Organization standards; many other cities such as Tianjin and Harbin have been severely affected. There are very few historical precedents for prolonged pollution exposure of this magnitude. But a 2013 joint study by China, United States, and Israel university researchers estimated that air pollution in China has already decreased live expectancy north of the Huai River by an amazing 5.5 years, including increases in lung cancer, heart attack, and stroke.

Moreover, China’s looming water shortages threaten to curtail industry, coal production, and agriculture. Some of China’s environmental challenges result from global climate change; but many if not most result from poor national management of the environment. Although China produces about one-tenth of global output, it consumes nearly one-fifth of the world’s energy production. Coal accounts for more than 70% of China’s electricity production. Coal generates more greenhouse gases than any other significant energy source. Coal production also uses a lot of water. The rapid expansion of coal use is placing major demands on China’s increasingly scarce water supply, adding to the growing demands stemming from irrigation and expanding cities. China is now the world’s larg- est emitter of greenhouse gases such as CO2, and emis- sions have been growing rapidly (see Chapter 10).

Product and worker safety Since 2007, highly publicized scandals concerning the safety of food, drugs, and other consumer products threatened the international public image of Chinese-made prod- ucts. Indeed, product safety standards are low, and their regulation is lax. Foreign and local investors, and government, all share in the blame. China’s regulatory institutions will need to catch up with the progress made in other aspects of national economic development.

Avoiding the Middle-Income Trap Chinese offi- cials and researchers are also concerned about sus- ceptibility to the “middle-income trap” and are engaging discussions with Latin American coun- tries on this topic; Huang Yiping and Jiang Tingsong stressed that what “really trapped many Latin Amer- ica and Middle East middle-income countries was lack of innovation capability. They failed to move up

the industrial ladder beyond resource-based activi- ties. This will also be the real test for China.” As the IMF concluded in its October 2013 World Economic Outlook, “There is strengthening conviction that China will grow more slowly over the medium term than in the recent past.” The alternative is probably wasteful and unsustainable investment that would result in serious economic crises. The question for China will be how it can maintain somewhat more modest but historically still high growth, of perhaps 6.5%, sustainably over the next three decades. An economy growing at this rate must have a different structure of investment than an economy growing at 10% (as China did in 2010). Making these adjust- ments will not be easy. Developing innovative capac- ity will be an important part of the answer; first steps are being taken, but better institutions may be needed to sustain the momentum.

Addressing Structural Imbalances There are sev- eral other imbalances in China’s economy that may lead to problems going forward. The World Bank pointed out in its 2013 Global Economic Prospects that “ongoing rebalancing efforts remain a priority, as does engineering a gradual decline in its unsustain- ably high investment rate.” The report also stressed that “should investments prove unprofitable, the servicing of existing loans could become problem- atic—potentially sparking a sharp uptick in nonper- forming loans that could require state intervention.”

China’s very large export surplus has come under great criticism, as this was widely argued to be one of the underlying causes of the global finan- cial crisis. One cause of the surpluses is probably the undervaluation of China’s exchange rate, esti- mated to be at least 20%. Undervaluation has been used by a number of East Asian economies as an industrial strategy for encouraging expansion of the manufacturing sector (notably in the 1960s and 1970s by South Korea and Taiwan; see Chapter 12), but those economies were much smaller than that of China. Note, however, that as recently as 2009 analysts estimated rates of overvaluation of up to 40%, approximately double the estimates of just four years later; and the external surplus as a share of GDP has decreased to a correspondingly degree since then (see Chapter 12 for details on measure- ment and analysis of international trade). Indeed, by 2013, some manufacturers found themselves struggling to adapt to a less overvalued currency.

205

206 PART onE Principles and Concepts

Inevitably, more China-based firms will engage in direct foreign investment in their export destina- tion countries such as the United States, just as Japan and South Korea did before them; but this will be a drawn-out process due to China’s still relatively low (if strongly growing) average productivity level, and also probably national security worries arising in Europe and Japan as well as in the United States.

Another factor in the large trade surplus is Chi- na’s high rate of savings, mentioned earlier, where the savings rate, long well above international averages, increased dramatically in the 1998–2010 period (when it peaked at approximately 49% of national income).

In parallel, investment as a share of GDP, long over 40%, reached an unprecedented 48% by 2010, before moderating slightly. Part of the uptick in recent years was due to an active response to the 2008 global economic crisis. The adjustment to sustainable investment and growth rates will be extremely difficult to accomplish without major and possibly prolonged disruptions. Yet in one sense the scope of the problem may also be somewhat exag- gerated by the way national statistics are prepared, which as Jun Zhang and Tian Zhu argued in a 2013 study does not account for hidden consumption by the growing number of high-income citizens, the rent-equivalent consumption of owner-occupied housing, and reported corporate expenses that are actually more like private consumption. On the other hand, on the other side of the balance sheet, the extraordinarily high investment—with evidence showing that a significant amount of it is at very low productivity—has not been challenged, and these statistics (including international trade and finance data) must be considered and better understood as a whole.

The huge indebtedness of the state-owned enter- prise (SOE) sector and other public debt (such as local government loans using land collateral) is thought by some financial commentators to be likely to eventually lead to a significant financial crisis— though other analysts argue that China can “grow its way out of” these problems.

For years, analysts have expressed concerns about the risk of “bubbles” (see Chapter 13) developing in financial and housing markets, due to very high rates of debt-financed investment. The potential problem has only worsened: just from 2008 to 2013, the over-

all credit to GDP ratio increased from about 120% in 2008 to close to 200%. This again connects to Chi- na’s unprecedentedly high investment rate and low consumption rate; adjustment is now essential, and indeed has begun, but the extent of the imbalances suggests that the transition from investment-led to consumption-led growth will be unusually long and difficult, and is very unlikely to be entirely smooth.

Political Weaknesses There are also political weak- nesses. On the one hand, some analysts make a case for the strengths of more authoritarian regimes, at least in early stages of development and when leadership fosters a developmental state. But on the other hand, this may make for a less flexible response to changing circumstances and difficul- ties in escaping a possible middle-income trap. Some leaders in China have called for urgent political reforms. And the dramatically wors- ened inequality in China may undermine not just political stability but ultimately opportunities for future growth (for details on the challenges of ris- ing inequality for growth and development, see Chapter 5).

Relatedly, China will need to find a way to con- tinue its ongoing institutional reforms, whether through implementing new and productive transi- tional institutions or more fundamental change. In their 2012 book, Why Nations Fail, Daron Acemoglu and James Robinson make an extraordinary argu- ment that institutional weaknesses will ultimately stall development in China. In their argument, insti- tutions in China closely resemble the ”extractive” political systems of other failed states where crony capitalism is the norm, vested interests are pro- tected, and potentially disruptive entrepreneurs are blocked. They conclude that growth in China will ultimately be “unlikely to translate into sustained economic development.” While relatively few ana- lysts think the challenges are this steep, undoubt- edly the needed reforms will be politically difficult to undertake.

The much-anticipated economic and social policy changes announced at the Third Plenum in November 2013 promised a further “unleash- ing” of market forces with their rhetoric promo- tion from a “basic” to “decisive” role—even if some details were left vague, and some predictions for reform were not realized (particularly in the field of finance). But left unambiguous was that the

206

207CHAPTER 4 Contemporary Models of Development and Underdevelopment

Communist Party monopoly on political control remained unchallenged—and indeed seemed to have been reinforced.

Finally, despite the extraordinary economic growth in China, Richard Easterlin has found that improvements in happiness and satisfaction in the country simply have not kept pace, particularly among the bottom third.

Managing Urbanization The scope of urban- ization in China has been called the largest migra- tion in human history, and indeed it has been breathtaking. For the first time in its history, China has become a more urban than rural society, with the halfway mark believed to have been crossed sometime in 2011; as recently as 1980, more than 80% of Chinese citizens lived in rural areas. Before 2030, China may reach the “Urban Billion” mark. Chongqing featured in the vignette in Chapter 1, growing from 200,000 in the 1930s, to about 2 mil- lion during the Cultural Revolution of the 1970s, to now over 30 million people in the metropolitan area. In the south, Shenzhen was transformed from a fishing village near Hong Kong to another mega- city in just a couple of decades. But conditions of ordinary people in many cities do not correspond to the media images of postmodern skyscrapers, as most are moving to large tracts of sometimes bleak, uniform apartment buildings, crawling in epic traf- fic jams through a vast urban sprawl—and indeed inhaling the “breathtaking” air pollution—in a pic- ture simultaneously of public overinvestment in some areas and underinvestment in others.

Demographic Challenges China also has a rapidly aging population. For the last decade of the twentieth century and first decade and a half of the twenty-first century, China has benefited from a demographic dividend (see Chapter 6), in which by global standards an unusually large fraction of its population has been of working age (neither too young nor too old to be active in the workforce). This “dividend” occurs in the process of economic development after the drop in births per woman but before the previous larger cohorts retire, allowing for rapid income growth. China is now entering a phase in which a large fraction of its working population will begin to retire. One challenge is the need to implement a modern pension system. Another is to respond to

a shrinking workforce and the need to support a large retired population. It is a challenge common to many modern societies but may be particularly acute in China due to its one-child policy that has been in effect since about 1980, which has greatly accelerated the demographic transition. There was a slight relaxation of this policy at the Third Ple- num in November 2013, to allow urban families for which either husband or wife is an only child to have a second child (previously this was allowed only if each was an only child). But this change may have very limited impact on fertility because of the high cost of raising children in China’s cities. The very high ratio of males to females (see Chap- ter 8) remains another serious demographic chal- lenge that may lead to continued distortions.

There are several explanations of China’s histor- ically unprecedented high savings rates (approach- ing 50% by some measures), but many of them relate to the unusual demographic challenges; they include “life-cycle” saving for retirement by an aging population that lacks social security, pre- cautionary savings due to increased income uncer- tainty because of fears about catastrophic family events such as major illnesses or layoffs, poor financial intermediation, and—in an influential new theory of Shang-Jin Wei and Xiaobo Zhang— competitive saving by parents of sons who now greatly outnumber daughters due to China’s growing sex-ratio imbalance and compete for pro- spective wives by offering larger houses and other wealth. High savings may be associated with the apparent property bubble that some economists in China believe has become dangerous—yet China has demonstrated a capacity for managing chal- lenges, and considerable reserves for addressing crises.

Other Limits to Emulating China’s Policies There are other limits to the lessons of China’s growth for other developing countries. China is quite homogeneous, overwhelmingly populated by members of the Han ethnic group. In Africa and other parts of the world, ethnic diversity is associ- ated with slower growth, though only in countries that also have incomplete or nonexistent political freedoms. Clearly, China is lacking in many free- doms. There may be limits to the ability of other countries to carry out China’s brand of centrally

207

208 PART onE Principles and Concepts

designed and implemented policies for transition and directed growth when either broader demo- cratic freedoms are in place or greater ethnic diver- sity is present. Finally, China, like much of the rest of East Asia, has a relatively poor endowment of natural resources. Many development specialists have concluded that this lack is actually more of a benefit than a drawback. Natural resource abun- dance encourages political infighting for control over the revenues, while manufacturing success is more important when a country does not have nat- ural resources to fall back on. It requires more ini- tiative and more efforts to upgrade technology and skill. In terms of geographic advantages, East Asia is also much less plagued than Africa and other developing regions by problems such as malaria and other tropical diseases for which medicines are not readily available, the difficulties and disadvan- tages of tropical agriculture, and the problems of landlocked countries.

The experience of China assures us that the East Asian miracle is not a fluke due to special local factors in economies such as South Korea and Taiwan. It gives us much greater confidence when we say that “real development is possible.” On the other

hand, there are clear limits to the ability of other developing regions to emulate the success of China. Not only do other developing countries differ in geography, demography, institutions, and allure to foreign investors, but also other regions may find themselves starved for investments that are redi- rected to China while remaining unable to compete with China’s impressive combination of low wages, high skills and know-how, and agglomeration of economic activity. Some East Asian countries have greatly benefited from the surge in import demand from China. The commodity price boom of recent years, which has stimulated demand in several countries in Africa, is significantly attributable to growth in China. And China itself has a good chance of continued high, albeit moderated growth, provided it manages the next phase of its transi- tion carefully. In the meantime, many developing countries that have hoped to rely more on manu- factured exports view the success of China as much as a threat as an opportunity. Growth in China will continue to be a central theme in the global devel- opment drama—both in its huge economic impact and the policy debate spurred by its extraordinary achievements. ■

Acemoglu, Daron, and James Robinson, Why Nations Fail: The Origins of Power, Prosperity, and Poverty. New York: Crown Business, 2012

Asian Development Bank. The National Accounts of the People’s Republic of China: Measurement Issues, Recent Developments, and the Way Forward. Manila, Philippines: Asian Development Bank, 2007.

Brandt, Loren, and Xiaodong R. Zhu. “Distribution in a decentralized economy: Growth and infla- tion in China under reform.” Journal of Political Economy 108 (2000): 422–439.

Byrd, William, and Lin Qingsong, eds. China’s Rural Industry: Structure, Development, and Reform. New York: Oxford University Press, 1990.

Chen, B., and Y. Feng. “Determinants of economic growth in China: Private enterprise, education, and openness.” China Economic Review 11 (2000): 1–15.

Chen, Shaohua, and Martin Ravallion. “China’s (uneven) progress against poverty.” Journal of Development Economics 82 (2007): 1–42.

———. “Learning from success: Understanding China’s (uneven) progress against poverty.” Finance and Development 41 (2004): 16–19.

Easterlin, Richard A. “When growth outpaces hap- piness.” New York Times, September 27, 2012, http://www.nytimes.com/2012/09/28/opin- ion/in-china-growth-outpaces-happiness.html?_ r=0.

Financial Times, numerous issues: the authors acknowledge the general contribution of the in- depth daily reporting on China by the Financial Times in informing the development of this case study over more than a decade.

George Washington University, Institute for Inter- national Economic Policy. “G2 at GW: Annual Conferences on China’s Economic Development and U.S.-China Economic Relations”, 2008–2013; “virtual conference volumes” at http://www .gwu.edu/~iiep/G2_at_GW/.

Sources

208

209CHAPTER 4 Contemporary Models of Development and Underdevelopment

Hu, Zuliu, and Mohsin S. Khan. “Why is China growing so fast?” IMF Working Paper No. 96/75, 1996.

Huang, Yasheng, Capitalism with Chinese Characteristics. New York: Cambridge University Press, 2008.

Islam, Nazrul, and Kazuhiko Yokota. “Lewis growth model and China’s industrialization.” Asian Economic Journal 22 (2008): 359–396.

Krugman, Paul. “The myth of Asia’s miracle.” For- eign Affairs 73, No. 6 (1994): 62–78.

Lau, Lawrence J., Yingyi Qian, and Gérard Roland. “Pareto-improving economic reforms through dual-track liberalization.” Economics Letters 55 (1997): 285–292.

———. “Reform without losers: An interpretation of China’s dual-track approach to transition.” Journal of Political Economy 108 (2000): 120–143.

Lu, D. “Industrial policy and resource allocation: Implications of China’s participation in glo- balization.” China Economic Review 11 (2000): 342–360.

Miller, Tom. China’s Urban Billion: The Story Behind the Biggest Migration in Human History. London: Zed Books, 2012.

Mody, Ashoka, and Fang-Yi Wang. “Explaining industrial growth in coastal China: Economic reforms…and what else?” World Bank Economic Review 11 (1997): 293–325.

Qian, Yingyi. “How reform worked in China.” In In Search of Prosperity: Analytic Narratives on Eco- nomic Growth, ed. Dani Rodrik. Princeton, N.J.: Princeton University Press, 2003.

Rawski, Thomas. “Measuring China’s recent GDP growth: Where do we stand?” October 2002. http://www.pitt.edu/~tgrawski/papers2002/ measuring.pdf.

Romer, Paul M. “Idea gaps and object gaps in eco- nomic development.” Journal of Monetary Eco- nomics 32 (1993): 543–573.

———. “Two strategies for economic development: Using ideas versus producing ideas.” Proceedings of the Annual World Bank Conference on Develop- ment Economics, 1992. Washington D.C. World Bank, 1993, pp. 63–91.

Sen, Amartya. Development as Freedom. New York: Knopf, 1999.

Smith, Stephen C. “Employee participation in China’s township and village enterprises.” China Economic Review 6 (1995): 157–167.

Smith, Stephen C. “Industrial policy and export suc- cess: Third World development strategies recon- sidered.” In U.S. Trade Policy and Global Growth, ed. Robert Blecker. New York: Sharpe, 1996.

Smith, Stephen C., and Yao Pan, “US-China Eco- nomic Relations,” in Oxford Companion to the Eco- nomics of China, eds.: Shenggen Fan, Ravi Kanbur, Shang-Jin Wei, and Xiaobo Zhang. Oxford Uni- versity Press.

United Nations Development Programme. Human Poverty Report, 2000. New York: United Nations, 2000.

Vogel, Ezra. One Step Ahead in China. Cambridge, Mass.: Harvard University Press, 1989.

Wei, Shang-Jin, and Xiaobo Zhang. “The competi- tive saving motive: Evidence from rising sex ratios and savings rates in China.” NBER Work- ing Paper No. 15093, 2009.

Weitzman, Martin L., and Chenggang Xu. “Chi- nese township and village enterprises as vaguely defined cooperatives.” Journal of Comparative Eco- nomics 20 (1994): 121–145.

Woetzel, Jonathan, et al. Preparing for China’s Urban Billion. McKinsey Global Institute, February 2009.

Woo, Wing Thye. “Chinese economic growth: Sources and prospects.” In The Chinese Economy, ed. Michel Fouquin and Françoise Lemoine. London: Economica, 1998.

Yiping, Huang, and Jiang Tingsong. “What does the Lewis turning point mean for China? A comput- able general equilibrium analysis.” China Center for Economic Research Working Paper No. E2010005, March 2010.

Young, Alwyn. “The tyranny of numbers: Con- fronting the statistical realities of the East Asian growth experience.” Quarterly Journal of Econom- ics 110 (1995): 641–680.

Zhang, Jun, and Tian Zhu. “Re-Estimating China’s Underestimated Consumption.” Working paper, China Europe International Business School, September 7, 2013.

Zhou, Ning, Yunshi Wang, and Lester Thurow. “The PRC’s real economic growth rate.” Unpublished report. Cambridge, Mass.: Massachusetts Institute of Technology, 2002.

Zhu, Xiaodong, “Understanding China’s growth: Past, present and future. Journal of Economic Per- spectives (Fall 2012).

209

210 PART onE Principles and Concepts

Concepts for review

Agency costs Asymmetric information Big push Binding constraint Complementarity Congestion Coordination failure Deep intervention

Economic agent Growth diagnostics Information externality Linkages Middle-income trap Multiple equilibria O-ring model O-ring production function

Pareto improvement Pecuniary externality Poverty trap Prisoners’ dilemma Social returns Technological externality Underdevelopment trap Where-to-meet dilemma

Questions for Discussion

1. Can you think of additional examples of comple- mentarities from everyday life? Does the S-shaped curve of Figure 4.1 shed any light on them? Do you think your examples help as a metaphor for economic development problems?

2. What role do you think international trade and foreign investment can play in solving some of the problems identified in the big push model? In the O-ring model? What limitations to your argu- ments can you think of?

3. The word trap suggests that there may be a way to escape. Do you think developing countries can escape all of the traps described in this chapter? Which ones would be most difficult to escape? How could the developed world be of assistance in these cases? Could developed countries do more?

4. Why might high levels of inequality lead to lower rates of growth and development? Why might it be difficult to get out of this kind of trap?

5. Why is the government sometimes a part of the problem of coordination failure rather than the solution? Does this make the problem hopeless? What could be done in this case?

6. One of the characteristics of some developing economies is the relatively low level of trust of people outside one’s extended family. How might the models explored in this chapter shed light on this problem?

7. Can you think of an example of O-ring production from everyday life? Do you think your example is a good metaphor for development problems?

8. Modern economic models sometimes require strong assumptions. What do you think are some of the trade-offs between a more rigorous, logically cohesive model with strong assump- tions but clear inferences and a description of problems followed by a verbal discussion of possible implications? Do you think the two approaches can be used together to inform each other?

9. As you read later chapters, think about whether the models described in this chapter are useful in shedding additional light on the nature of the problems considered. Some of the later prob- lems you might consider are child labor, poor health and nutrition among the poor, high fer- tility, environmental degradation, availability of credit for the poor, urbanization, protectionism in international trade by developed and devel- oping countries, reform of government, and land reform.

10. Select a developing country that interests you and search for evidence suggesting which factors are the binding constraint on growth. (For inspira- tion, see the sources in Box 4.3.)

11. What kinds of market failures are present in the economic self-discovery framework, and how may they be overcome?

12. Consider the most recent economic performance in China. To what extent do you think it confirms, and to what extent calls for adjustments in, the analysis in the China case study?

211CHAPTER 4 Contemporary Models of Development and Underdevelopment

Notes

1. See Karla Hoff and Joseph E. Stiglitz, “Modern economic theory and development,” in Frontiers in Development Economics, eds. Gerald M. Meier and Joseph E. Stiglitz (New York: Oxford Univer- sity Press, 2000). The Hoff and Stiglitz epigram (header quote) is drawn from this source, p. 390.

2. For example, the two approaches have converged when low-growth paths resulting from a coor- dination failure have been explicitly examined within an endogenous growth framework. See Oded Galor and Joseph Zeira, “Income distribu- tion and macroeconomics,” Review of Economic Studies 60 (1993): 35–52.

3. For an insightful discussion of how many of the perspectives of this approach are applied to “new economy” issues, see Carl Shapiro and Hal Var- ian, Information Rules: A Strategic Guide to the Net- work Economy (Boston: Harvard Business School Press, 1999).

4. The problems cannot be solved even by perfect labor contracting (which is generally impossible in any case) if there is a risk of involuntary sepa- rations between firms and their employees (e.g., firm bankruptcies or death or serious illness of an employee). For a particularly insightful formal model, see Daron Acemoglu, “Training and inno- vation in an imperfect labour market,” Review of Economic Studies 64 (1997): 445–464.

5. For an interesting formal model of this problem with supporting empirical evidence from rural Bangladesh, see Shahe Emran and Forhad Shilpi, “Marketing externalities, multiple equilibria, and market development,” a paper presented at the Northeast Universities Development Con- ference, Boston University, September 2001. See also Shahe Emran and Forhad Shilpi, “The extent of the market and stages of agricultural special- ization,” Canadian Journal of Economics 45, No. 3 (2012): 1125–1153.

6. Alicia Adsera and Debraj Ray, “History and co- ordination failure,” Journal of Economic Growth 3 (1998): 267–276; Debraj Ray, Development Econom- ics (Princeton, N.J.: Princeton University Press, 1998), ch. 5.

7. For an introductory overview of the prisoners’ dilemma problem, see Robert Gibbons, Game Theory for Applied Economists (Princeton, N.J.: Princeton University Press, 1992), pp. 2–7.

8. Even under perfect information conditions, how- ever, coordination can remain a problem.

9. Technically, Figure 4.1 assumes that agents are homogeneous and depicts a symmetrical Nash equilibrium, but this can be generalized to cases in which agents differ. An example of an upward- sloping supply curve intersecting a downward- sloping demand curve—to produce a single equi- librium—can be seen in Figure 5.5, for the case of a labor market.

10. Technically, what is depicted is a set of symmetri- cal Nash equilibria. The S-shaped curve is the re- action curve of a representative agent to the aver- age behavior of the other agents.

11. Paul Rosenstein-Rodan, “Problems of industrial- ization of eastern and southeastern Europe,” Eco- nomic Journal 53 (1943): 202–211.

12. Kevin M. Murphy, Andrei Shleifer, and Robert W. Vishny, “Industrialization and the big push,” Journal of Political Economy 97 (1989): 1003–1026.

13. Paul Krugman, Development, Geography, and Eco- nomic Theory (Cambridge, Mass.: MIT Press, 1995), ch. 1. For an alternative exposition and an algebraic development of the model, see Kaushik Basu, Analytical Development Economics (Cambridge, Mass.: MIT Press, 1997), pp. 17–33.

14. One reason could be an efficiency wage effect, in which workers work harder to avoid being fired when paid a high wage, thereby raising produc- tivity enough to pay for the higher wage.

15. We are assuming that modern-sector workers would be changing the sectors (from traditional to modern) in which they work voluntarily; that is, they are not slave labor.

16. In the formal model of Murphy, Shleifer, and Vishny, there is a continuum of products, but that need not concern us here.

17. This consumption pattern means that there is unit- elastic demand; this is the type of demand function

212 PART onE Principles and Concepts

that follows from a Cobb-Douglas utility function with equal preference weights for all goods, such as a utility function given by the products of the amounts of each type of good consumed. Techni- cally, Murphy, Schleifer, and Vishny assume that there is one representative consumer who supplies all labor and receives all profits and, with their other assumptions, set up the model so that Figure 4.2 and other parts of the analysis can be thought of either as the economy as a whole or as any particular market, but these considerations need not concern us here.

18. See, for example, Hollis B. Chenery, Sherman Robinson, and Moshe Syrquin, Industrializa- tion and Growth: A Comparative Study (New York: Oxford University Press, 1986).

19. For work in this field, see, for example, Andrés Rodriguez-Clare, “The division of labor and eco- nomic development,” Journal of Development Eco- nomics 49 (1996): 3–32. Rodriguez-Clare starts with three plausible conditions that have had wide the- oretical and empirical support since Adam Smith in the first two cases and Alfred Marshall in the third: There are productivity gains from the divi- sion of labor, the division of labor is limited by the extent of the market, and, as explained in Chapter 7, efficiency gains are derived from the proxim- ity of suppliers and users of certain inputs. Given these assumptions, Rodriguez-Clare then shows that a small, open economy may be caught in an underdevelopment trap in which a “shallow divi- sion of labor” (i.e., a low variety of specialized in- puts) is self-reinforcing. This in turn leads to a low rate of return to capital, so foreign investment or domestic capital accumulation may not material- ize to help solve the problem. For another illustra- tion, see Dani Rodrik, “Coordination failures and government policy: A model with applications to East Asia and Eastern Europe,” Journal of Interna- tional Economics 40 (1996): 1–22. See also Murphy, Shleifer, and Vishny, “Industrialization,” sec. 6.

20. Recall from microeconomics that we can write mar- ginal revenue as P(Q)[1 - 1/h], where P is price and h is the (absolute value of) price elasticity of demand. With unit elasticity, h = 1; then note that this producer has positive constant marginal costs. Therefore, profits may be indefinitely increased by decreasing output and raising price accordingly.

21. In other words, the producer acts as a limit-pricing monopolist.

22. Wages have risen to w, but this product sector is by definition a very small part of the economy, so we can ignore income effects, which are negligible.

23. The graph was suggested by Krugman. See Devel- opment, Geography, and Economic Theory.

24. Thus conditions for monopoly limit pricing are still present.

25. With a price of 1, the quantity of goods purchased by workers is equal to the wage bill.

26. To see this, note that after the big push, total wages in the economy are w2(L/N)N, and total profits are [1Q2 − w2(L/N)]N. Summing these, we get 1Q2N, the value of total output.

27. Expressed differently, the problem is that market failure is present. In particular, as Krugman points out, the interaction between a firm’s internal economies of scale and the existence of perfectly elastic labor supplies at low wages together generate pecuniary externalities that inhibit the entry of modern firms. In other words, by gen- erating an increase in aggregate demand, each firm makes a contribution to a mutually profit- able big push to industrialization, even though individually each firm would lose money by in- dustrializing alone. Thus, although the economies of scale are internal to the firm, when combined with the presence of a traditional sector paying low wages, de facto external pecuniary positive externalities are generated. Again, this is because each firm’s production has the effect of raising other firms’ revenues, making them more profit- able. A simple characterization of the conundrum is that if there is only one modern firm, profits are greater in the traditional sector, but if there is a modern firm in every activity, profits are greater in the modern sector.

28. Formally, F = F(N), where F is falling as N rises.

29. For details of one insightful formal model that casts the big push model in relatively accessible algebraic terms, see Stefano Paternostro, “The poverty trap: The dual externality model and its policy implications,” World Development 25 (1997): 2071–2081.

30. Note that formally, in this case, efficient means “la- borsaving,” but the point is more general.

31. As Murphy, Schleifer, and Vishny show, there is also a plausible equilibrium condition that an increased interest-rate effect is not too large.

213CHAPTER 4 Contemporary Models of Development and Underdevelopment

32. Openness to trade will not resolve this problem be- cause the development of cities in other countries does not generally assist with national development. Urbanization is discussed further in Chapter 7.

33. In principle, if it is known that a sufficient number of modern firms will enter, the infrastructure prob- lem can be solved by using perfect price discrimi- nation, but if firms have different fixed costs that are not observable by the infrastructure provider or if perfect price discrimination is not possible for some other reason, the infrastructure may not be built, even when it is efficient to do so. See Murphy, Shleifer, and Vishny, “Industrialization,” sec. 6. For an accessible algebraic derivation using a diagram similar to Figure 4.2, see Pranab Bardhan and Chris Udry, Development Microeconomics (New York: Oxford University Press, 1999), pp. 208–211.

34. The term lemons derives from poor-quality used cars. As is well known, new cars lose a significant part of their value as soon as they leave the show- room. This is because the mere fact that a car is offered for sale is taken as valuable information about the car in itself. People wanting to buy a car are generally not expert mechanics, so they need to search for some shorthand information to help them decide what a car is worth; obviously, own- ers of a poor-quality car are more likely to offer it for sale. Analogies to this “lemons problem” have many other applications in economics, such as in financial markets (see Chapter 15). Also see George Akerlof, “The market for lemons,” Quar- terly Journal of Economics 84 (1970): 488–500.

35. For an excellent survey of some of the new de- velopments covered in this section, see Hoff and Stiglitz, “Modern economic theory and develop- ment.” Another good discussion of this and related topics is found in Ray, Development Economics, ch. 5.

36. See Alice Amsden, Asia’s Next Giant: South Korea and Late Industrialization (Oxford: Oxford Uni- versity Press, 1989) and The Rise of the Rest (New York: Oxford University Press, 2001); Carl J. Dahl- man, Bruce Ross-Larson, and Larry E. Westphal, “Managing technical development: Lessons from the newly industrializing countries,” World De- velopment 15 (1987): 759–775; Richard Luedde- Neurath, Import Controls and Export-Oriented De- velopment: A Reassessment of the South Korean Case (Boulder, Colo.: Westview Press, 1986); Howard Pack and Larry E. Westphal, “Industrial strategy

and technological change: Theory versus real- ity,” Journal of Development Economics 22 (1986): 87–128; Joseph Stern et al., Industrialization and the State: The Korean Heavy and Chemical Industry Drive (Cambridge, Mass.: Harvard University Press, 1995); Gordon White, ed., Developmental States in East Asia (New York: St. Martin’s Press, 1988); and Stephen C. Smith, “Industrial policy and export success: Third World development strategies recon- sidered,” in U.S. Trade Policy and Global Growth, ed. Robert Blecker (New York: Sharpe, 1996), pp. 267–298. On linkages, see also Masahisa Fujita, Paul Krugman, and Anthony J. Venables, The Spa- tial Economy: Cities, Regions, and International Trade (Cambridge, Mass.: MIT Press 1999).

37. This perspective helps account for the popularity of input-output analysis in development planning and policy formulation, especially in earlier years, although it is an imperfect tool for this purpose (see Chapter 11).

38. For some evidence, see William L. Megginson and Jeffry M. Netter, “From state to market: A survey of empirical studies on privatization,” Journal of Economic Literature 39 (2001): 321–390.

39. See Abhijit V. Banerjee and Andrew F. Newman, “Occupational choice and the process of devel- opment,” Journal of Political Economy 101 (1993): 274–298.

40. Galor and Zeira’s model rests on an alternative way to characterize imperfect capital markets— that the rate of interest for borrowers is greater than that for lenders. One can verify the reason- ableness of this assumption with a brief visit to any bank. The model is a simple two-period, over- lapping-generations model. See Galor and Zeira, “Income distribution and macroeconomics.”

41. Torsten Persson and Guido Tabellini, “Is inequality harmful for growth?” American Economic Review 84 (1994): 600–621; see also Chapter 5 of this text.

42. Michael Kremer, “The O-ring theory of economic development,” Quarterly Journal of Economics 108 (1993): 551–575. A good exposition of the model, which provides alternative proofs to the ones found in Kremer, is found in Basu, Analytical Development Economics.

43. More generally, there are n tasks; for simplic- ity, we continue to assume that one, and only one, worker must perform each of the n tasks,

214 PART one Principles and Concepts

but conceptually, n should be thought of as tasks rather than number of workers. If, and only if, all tasks are performed successfully, output per worker is given by B, which is given in value terms (or if thought of in quantity terms, price is normalized to 1). Conventional capital, k, may also be used (if not, simply set k = 1), which is in- troduced in the formula, with diminishing returns (of course, capital might also be of varying qual- ity). Expected output y is given by

E(y) = Ka a q n

i = 1 qibnB

In general, we must multiply by n because other- wise the firm can only lose value by adding more differentiated tasks. In the O-ring theory, Kremer analyzes what happens when B = B(n), where B' (n) 7 0, as a way of endogenizing technology choice.

44. For a more formal and more general demonstra- tion that firms would choose to employ workers of the same skill level (or as close to identical as pos- sible), let us continue the example from note 43. A necessary condition for a maximum with respect to each of the labor qualities q is

dw(qi) dqi

‚ dy

dqi = a q

j ≠ i qjb nBKa

This equation tells us that in equilibrium, the value of the marginal product of skill is equal to the marginal cost of skill in wage payments. In other words, the firm finds that the added value of replacing one worker with another with higher skill while leaving the skill levels of all other workers constant is equal to the resulting increase in the wage bill. Next, note that the second deriv- ative, or the derivative of the marginal product of skill for the ith worker with respect to the skill level of the other workers, is positive; that is,

d2y

dqi d = a q

j ≠ i qjb = nBK a 7 0

This positive cross-derivative indicates that firms with high-skilled workers in all but one task receive

the greatest benefits from having a high-skilled worker in the remaining task, and so they can and would bid the most for high-skilled workers.

45. Technically, this type of marriage market match- ing process does not depend on the presence of positive cross-derivatives, as in note 44, but re- sults only from individual preferences, along with the assumption of nontransferable utility (mean- ing that there can be no side payments). Thus there are two types of situations in which positive assortative matching may occur.

46. See Michael Kremer and Eric Maskin, “Wage in- equality and segregation by skill,” NBER Working Paper No. 5718, 1996.

47. See Kremer, “O-ring theory,” for a formal statement of this result and for extensions to cases of endoge- nous skill investments under imperfect information.

48. Ibid., p. 574. The multiple equilibrium analysis is found on pp. 564–571.

49. Ricardo Hausmann and Dani Rodrik, “Economic development as self-discovery,” Journal of Develop- ment Economics 72 (2003): 603–633. A related and insightful earlier analysis was provided by Karla Hoff, “Bayesian learning in an infant industry model,” Journal of International Economics 43 (1997): 409–436.

50. Ricardo Hausmann, Dani Rodrik, and Andrés Velasco, “Growth diagnostics,” in One Econom- ics, Many Recipes: Globalization, Institutions, and Economic Growth, by Dani Rodrik (Princeton, N.J.: Princeton University Press, 2007), ch. 2.

51. Paul Krugman, Development, Geography, and Economic Theory, (Cambridge: MIT Press, 1995.)

52. Hoff and Stiglitz, “Modern economic theory and development.”

53. Seminar presentation by Joseph E. Stiglitz at the World Bank, May 27, 1999; and ibid., p. 421.

54. Karla Hoff, “Beyond Rosenstein-Rodan: The mod- ern theory of coordination problems in develop- ment,” in Annual World Bank Conference on Devel- opment Economics, 1999 (Washington, D.C.: World Bank, 2000), p. 146.

PA R T T W O Problems and Policies: Domestic

Chapters 1 and 2 introduced the problem that despite significant improvements over the past half century, extreme poverty remains widespread in the develop- ing world. In 2010, more than 1.2 billion people lived on less than $1.25 per day at 2005 U.S. purchasing power parity (2013 World Bank estimate). Some 2.4 billion— more than one-third of the world’s population—lived on less than $2 a day. As you will see in the next few chapters, often these impoverished people suffer from un- dernutrition and poor health, have little or no literacy, live in environmentally de- graded areas, have little political voice, are socially excluded, and attempt to earn a meager living on small and marginal farms (or as day laborers) or in dilapidated urban slums. In this chapter, we set the stage with an in-depth examination of the problems of poverty and of highly unequal distributions of income.

That development requires a higher gross national income (GNI), and hence sustained growth, is clear. The basic issue, however, is not only how to make GNI grow but also who would make it grow: the few or the many. If it were the rich, it would most likely be appropriated by them, and progress against poverty would be slow, and inequality would worsen. But if it were generated by the many, they would be its principal beneficiaries, and the fruits of economic growth would be shared more evenly. Thus, many developing countries that had experienced relatively high rates of economic growth by historical standards discovered that such growth often brought little in the way of significant benefits to their poor.

216

Poverty, Inequality, and Development

No society can surely be flourishing and happy, of which by far the greater part of the numbers are poor and miserable.

—Adam Smith, 1776

Viewed through the lens of human development, the global village appears deeply divided between the streets of the haves and those of the have-nots.

—United Nations Development Programme, Human Development Report, 2006

Social protection directly reduces poverty and helps make growth more pro-poor. —Organization for Economic Cooperation and Development, 2010

The coincidence of severe and persistent poverty and hunger indicates the presence of poverty traps—conditions from which individuals or groups cannot emerge without the help of others.

—International Food Policy Research Institute, 2007

The World Bank Group has adopted two new goals: end extreme poverty by 2030 and boost shared prosperity by maximizing income growth for the poorest 40 percent in every country.

—Jim Yong Kim, President, World Bank, 2013

5

217CHAPTER 5 Poverty, Inequality, and Development

Because the elimination of widespread poverty and high and even growing income inequality are at the core of all development problems and in fact define for many people the principal objective of development policy, we begin Part Two by focusing on the nature of the poverty and inequality problem in devel- oping countries. Although our main focus is on economic poverty and inequali- ties in the distribution of incomes and assets, it is important to keep in mind that this is only part of the broader inequality problem in the developing world. Of equal or even greater importance are inequalities of power, prestige, status, gender, job satisfaction, conditions of work, degree of participation, freedom of choice, and many other dimensions of the problem that relate more to our second and third components of the meaning of development, self-esteem, and freedom to choose. As in most social relationships, we cannot really separate the economic from the noneconomic manifestations of inequality. Each reinforces the other in a complex and often interrelated process of cause and effect.

After introducing appropriate measures of inequality and poverty, we de- fine the nature of the poverty and income distribution problem and consider its quantitative significance in various developing nations. We then examine in what ways economic analyses can shed light on the problem and explore possible alternative policy approaches directed at the elimination of poverty and the reduction of excessively wide disparities in the distributions of in- come in developing countries. A thorough understanding of these two funda- mental economic manifestations of underdevelopment provides the basis for analysis in subsequent chapters of more specific development issues, includ- ing population growth, education, health, rural development, environmental degradation and climate change, and foreign assistance.

In this chapter, therefore, we will examine the following critical questions about the relationship among economic growth, income distribution, and poverty:

1. How can we best measure inequality and poverty?

2. What is the extent of relative inequality in developing countries, and how is this related to the extent of absolute poverty?

3. Who are the poor, and what are their economic characteristics?

4. What determines the nature of economic growth—that is, who benefits from economic growth, and why?

5. Are rapid economic growth and more equal distributions of income com- patible or conflicting objectives for low-income countries? To put it another way, is rapid growth achievable only at the cost of greater inequalities in the distribution of income, or can a lessening of income disparities con- tribute to higher growth rates?

6. Do the poor benefit from growth, and does this depend on the type of growth a developing country experiences? What might be done to help the poor benefit more?

7. What is so bad about extreme inequality?

8. What kinds of policies are required to reduce the magnitude and extent of absolute poverty?

218 PART Two Problems and Policies: Domestic

We begin the chapter by defining inequality and poverty, terms that are com- monly used in informal conversation but need to be measured more precisely to provide a meaningful understanding of how much progress has already been made, how much remains to be achieved, and how to set incentives for government officials to focus on the most pressing needs. You will see that the most important measures of poverty and inequality used by development economists satisfy properties that most observers would agree are of funda- mental importance. After a discussion of why attention to inequality as well as poverty is important, we then use the appropriate measures of poverty and inequality to evaluate the welfare significance of alternative patterns (or “ typologies”) of growth. After reviewing the evidence on the extent of poverty and inequality in the developing world, we conclude with an overview of the key issues in poverty policy. Some important principles of effective poverty policies are considered, together with some initial examples of programs that have worked well in practice. We conclude the chapter with a comparative case study of Ghana and Côte d’Ivoire, which illustrates, issues of the quality of growth and the difficulties of achieving it.

5.1 Measuring Inequality

In this section, we define the dimensions of the income distribution and pov- erty problems and identify some similar elements that characterize the prob- lem in many developing nations. But first we should be clear about what we are measuring when we speak about the distribution of income and absolute poverty.

Economists usually distinguish between two principal measures of income distribution for both analytical and quantitative purposes: the personal or size distribution of income and the functional or distributive factor share distribu- tion of income.

Size Distributions

The personal or size distribution of income is the measure most commonly used by economists. It simply deals with individual persons or households and the total incomes they receive. The way in which they received that in- come is not considered. What matters is how much each earns irrespective of whether the income is derived solely from employment or comes also from other sources such as interest, profits, rents, gifts, or inheritance. Moreover, the locational (urban or rural) and occupational sources of the income (e.g., agriculture, manufacturing, commerce, services) are ignored. If Ms. X and Mr. Y both receive the same personal income, they are classified together irrespec- tive of the fact that Ms. X may work 15 hours a day as a doctor while Mr. Y doesn’t work at all but simply collects interest on his inheritance.

Economists and statisticians therefore like to arrange all individuals by as- cending personal incomes and then divide the total population into distinct groups, or sizes. A common method is to divide the population into succes- sive quintiles (fifths) or deciles (tenths) according to ascending income levels and then determine what proportion of the total national income is received

Personal distribution of income (size distribution of income) The distribution of income according to size class of persons—for example, the share of total income accruing to the poorest specific per- centage or the richest specific percentage of a population— without regard to the sources of that income.

Quintile A 20% proportion of any numerical quantity. A population divided into quin- tiles would be divided into five groups of equal size.

Decile A 10% portion of any numerical quantity; a popu- lation divided into deciles would be divided into ten equal numerical groups.

219CHAPTER 5 Poverty, Inequality, and Development

by each income group. For example, Table 5.1 shows a hypothetical but fairly typical distribution of income for a developing country. In this table, 20 indi- viduals, representing the entire population of the country, are arranged in order of ascending annual personal income, ranging from the individual with the lowest income (0.8 units) to the one with the highest (15.0 units). The total or national income of all individuals amounts to 100 units and is the sum of all en- tries in column 2. In column 3, the population is grouped into quintiles of four individuals each. The first quintile represents the bottom 20% of the population on the income scale. This group receives only 5% (i.e., a total of 5 money units) of the total national income. The second quintile (individuals 5 through 8) receives 9% of the total income. Alternatively, the bottom 40% of the population (quintiles 1 plus 2) is receiving only 14% of the income, while the top 20% (the fifth quintile) of the population receives 51% of the total income.

A common measure of income inequality that can be derived from col- umn 3 is the ratio of the incomes received by the top 20% and bottom 40% of the population. This ratio, sometimes called a Kuznets ratio after Nobel laure- ate Simon Kuznets, has often been used as a measure of the degree of inequal- ity between high- and low-income groups in a country. In our example, this inequality ratio is equal to 51 divided by 14, or approximately 3.64.

To provide a more detailed breakdown of the size distribution of income, decile (10%) shares are listed in column 4. We see, for example, that the bot- tom 10% of the population (the two poorest individuals) receives only 1.8% of the total income, while the top 10% (the two richest individuals) receives 28.5%. Finally, if we wanted to know what the top 5% receives, we would divide

Table 5.1 Typical Size Distribution of Personal Income in a Developing Country by Income Shares—Quintiles and Deciles

    Share of Total Income (%)

Individuals Personal Income

(money units)

Quintiles

Deciles

1 2 3 4 5 6 7 8 9

10 11 12 13 14 15 16 17 18 19 20

Total (national income)

0.8 1.0 1.4 1.8 1.9 2.0 2.4 2.7 2.8 3.0 3.4 3.8 4.2 4.8 5.9 7.1

10.5 12.0 13.5 15.0

100.0

      5       9      

13      

22      

51 100

  1.8

  3.2

  3.9

  5.1

  5.8

  7.2

  9.0

  13.0

  22.5

  28.5

100.0

Income inequality The dis- proportionate distribution of total national income among households.

220 PART Two Problems and Policies: Domestic

the total population into 20 equal groups of individuals (in our example, this would simply be each of the 20 individuals) and calculate the percentage of total income received by the top group. In Table 5.1, we see that the top 5% of the population (the twentieth individual) receives 15% of the income, a higher share than the combined shares of the lowest 40%.

Lorenz Curves

Another common way to analyze personal income statistics is to construct what is known as a Lorenz curve.1 Figure 5.1 shows how it is done. The numbers of income recipients are plotted on the horizontal axis, not in absolute terms but in cumulative percentages. For example, at point 20, we have the lowest (poorest) 20% of the population; at point 60, we have the bottom 60%; and at the end of the axis, all 100% of the population has been accounted for. The vertical axis shows the share of total income received by each percentage of population.

It is also cumulative up to 100%, meaning that both axes are the same length. The entire figure is enclosed in a square, and a diagonal line is drawn from the lower left corner (the origin) of the square to the upper right corner. At every point on that diagonal, the percentage of income received is exactly equal to the percentage of income recipients—for example, the point halfway along the length of the diagonal represents 50% of the income being distributed to exactly 50% of the population. At the three-quarters point on the diagonal, 75% of the income would be distributed to 75% of the population. In other words, the diagonal line in Figure 5.1 is representative of “perfect equality” in size distribution of income. Each percentage group of income recipients is receiving

Lorenz curve A graph depicting the variance of the size distribution of income from perfect equality.

FIgure 5.1 The lorenz Curve

E

F

G

H

I

D C

B A

Percentage of income recipients

100

90

80

70

60

50

40

30

20

10

0 10 20 30 40 50 60 70 80 90 100

Line of equality

Lorenz curve

P e

rc e

n ta

g e

o f

in co

m e

6125_05_FG001

221CHAPTER 5 Poverty, Inequality, and Development

that same percentage of the total income; for example, the bottom 40% receives 40% of the income, while the top 5% receives only 5% of the total income.2

The Lorenz curve shows the actual quantitative relationship between the percentage of income recipients and the percentage of the total income they did in fact receive during, say, a given year. In Figure 5.1, we have plotted this Lorenz curve using the decile data contained in Table 5.1. In other words, we have divided both the horizontal and vertical axes into ten equal segments corresponding to each of the ten decile groups. Point A shows that the bottom 10% of the population receives only 1.8% of the total income, point B shows that the bottom 20% is receiving 5% of the total income, and so on for each of the other eight cumulative decile groups. Note that at the halfway point, 50% of the population is in fact receiving only 19.8% of the total income.

The more the Lorenz line curves away from the diagonal (line of perfect equality), the greater the degree of inequality represented. The extreme case of perfect inequality (i.e., a situation in which one person receives all of the na- tional income while everybody else receives nothing) would be represented by the congruence of the Lorenz curve with the bottom horizontal and right-hand vertical axes. Because no country exhibits either perfect equality or perfect in- equality in its distribution of income, the Lorenz curves for different countries will lie somewhere to the right of the diagonal in Figure 5.1. The greater the degree of inequality, the greater the bend and the closer to the bottom horizon- tal axis the Lorenz curve will be. Two representative distributions are shown in Figure 5.2, one for a relatively equal distribution (Figure 5.2a) and the other for a relatively unequal distribution (Figure 5.2b). (Can you explain why the Lorenz curve could not lie above or to the left of the diagonal at any point?)

FIgure 5.2 The greater the Curvature of the lorenz line, the greater the relative Degree of Inequality

100

P e

rc e

n ta

g e

o f

in co

m e

Line of equality

Lorenz curve

0 100

Percentage of population

(a) A relatively equal distribution

100

P e

rc e

n ta

g e

o f

in co

m e

Line of equality

Lorenz curve

0 100

Percentage of population

(b) A relatively unequal distribution

6125_05_FG002

222 PART Two Problems and Policies: Domestic

Gini Coefficients and Aggregate Measures of Inequality

A final and very convenient shorthand summary measure of the relative de- gree of income inequality in a country can be obtained by calculating the ratio of the area between the diagonal and the Lorenz curve divided by the total area of the half-square in which the curve lies. In Figure 5.3, this is the ratio of the shaded area A to the total area of the triangle BCD. This ratio is known as the Gini concentration ratio or Gini coefficient, named after the Italian statisti- cian who first formulated it in 1912.

Gini coefficients are aggregate inequality measures and can vary anywhere from 0 (perfect equality) to 1 (perfect inequality). In fact, as you will soon dis- cover, the Gini coefficient for countries with highly unequal income distribu- tions typically lies between 0.50 and 0.70, while for countries with relatively equal distributions, it is on the order of 0.20 to 0.35. The coefficient for our hypothetical distribution of Table 5.1 and Figure 5.1 is approximately 0.44—a relatively unequal distribution.

Four possible Lorenz curves such as might be found in international data are drawn in Figure 5.4. In the “Lorenz criterion” of income distribution, whenever one Lorenz curve lies above another Lorenz curve, the economy corresponding to the upper Lorenz curve is more equal than that of the lower curve. Thus, economy A may unambiguously be said to be more equal than economy D. Whenever two Lorenz curves cross, such as curves B and C, the Lorenz criterion states that we “need more information” or additional assump- tions before we can determine which of the underlying economies is more equal. For example, we might argue on the grounds of the priority of address- ing problems of poverty that curve B represents a more equal economy, since the poorest are richer, even though the richest are also richer (and hence the middle class is “squeezed”). But others might start with the assumption that

FIgure 5.3 estimating the gini Coefficient

P e

rc e

n ta

g e

o f

in co

m e

B C

D

Percentage of population

Lorenz curve

Line of equality

Gini coefficient = total area BCD

6125_05_FG003

A

shaded area A

Gini coefficient An aggregate numerical measure of income inequality ranging from 0 (perfect equality) to 1 (perfect inequality). It is mea- sured graphically by dividing the area between the perfect equality line and the Lorenz curve by the total area lying to the right of the equality line in a Lorenz diagram. The higher the value of the coefficient is, the higher the inequality of income distribution; the lower it is, the more equal the distri- bution of income.

223CHAPTER 5 Poverty, Inequality, and Development

an economy with a stronger middle class is inherently more equal, and those observers might select economy C.

One could also use an aggregate measure such as the Gini coefficient to decide the matter. As it turns out, the Gini coefficient is among a class of meas- ures that satisfy four highly desirable properties: the anonymity, scale inde- pendence, population independence, and transfer principles.3 The anonymity principle simply means that our measure of inequality should not depend on who has the higher income; for example, it should not depend on whether we believe the rich or the poor to be good or bad people. The scale independence principle means that our measure of inequality should not depend on the size of the economy or the way we measure its income; for example, our inequality measure should not depend on whether we measure income in dollars or in cents or in rupees or rupiahs or for that matter on whether the economy is rich on average or poor on average—because if we are interested in inequality, we want a measure of the dispersion of income, not its magnitude (note that mag- nitudes are very important in poverty measures). The population independence principle is somewhat similar; it states that the measure of inequality should not be based on the number of income recipients. For example, the economy of China should be considered no more or less equal than the economy of Vi- etnam simply because China has a larger population than Vietnam. Finally, we have the transfer principle (sometimes called the Pigou-Dalton principle after its creators); it states that, holding all other incomes constant, if we transfer some income from a richer person to a poorer person (but not so much that the poorer person is now richer than the originally rich person), the resulting new income distribution is more equal. If we like these four criteria, we can measure the Gini coefficient in each case and rank the one with the larger Gini as more unequal. However, this is not always a perfect solution. For example, the Gini coefficient can, in theory, be identical for two Lorenz curves that cross; can you see why by looking at curves B and C in Figure 5.4? And sometimes different

Percentage of income recipients

P e

rc e

n ta

g e

o f

in co

m e

0

100

100

A

B C D

6125_05_FG004.

FIgure 5.4 Four Possible lorenz Curves

224 PART Two Problems and Policies: Domestic

inequality measures that satisfy our four properties can give different answers as to which of two economies are more unequal.4

Note that a measure of dispersion common in statistics, the coefficient of variation (CV), which is simply the sample standard deviation divided by the sample mean, is another measure of inequality that also satisfies the four criteria. Although the CV is more commonly used in statistics, the Gini coefficient is often used in studies of income and wealth distribution due to its convenient Lorenz curve interpretation. Note, finally, that we can also use Lorenz curves to study inequality in the distribution of land, in education and health, and in other assets.

Functional Distributions

The second common measure of income distribution used by economists, the functional or factor share distribution of income, attempts to explain the share of total national income that each of the factors of production (land, labor, and capital) receives. Instead of looking at individuals as separate enti- ties, the theory of functional income distribution inquires into the percentage that labor receives as a whole and compares this with the percentages of total income distributed in the form of rent, interest, and profit (i.e., the returns to land and financial and physical capital). Although specific individuals may receive income from all these sources, that is not a matter of concern for the functional approach.

A sizable body of theoretical literature has been built up around the con- cept of functional income distribution. It attempts to explain the income of a factor of production by the contribution that this factor makes to produc- tion. Supply and demand curves are assumed to determine the unit prices of each productive factor. When these unit prices are multiplied by quantities employed on the assumption of efficient (minimum-cost) factor utilization, we get a measure of the total payment to each factor. For example, the supply of and demand for labor are assumed to determine its market wage. When this wage is then multiplied by the total level of employment, we get a measure of total wage payments, also sometimes called the total wage bill.

Figure 5.5 provides a simple diagrammatic illustration of the traditional theory of functional income distribution. We assume that there are only two factors of production: capital, which is a fixed (given) factor, and labor, which is the only variable factor. Under competitive market assumptions, the demand for labor will be determined by labor’s marginal product (i.e., additional work- ers will be hired up to the point where the value of their marginal product equals their real wage). But in accordance with the principle of diminishing marginal products, this demand for labor will be a declining function of the numbers employed. Such a negatively sloped labor demand curve is shown by line DL in Figure 5.5. With a traditional, neoclassical, upward-sloping la- bor supply curve SL, the equilibrium wage will be equal to WE and the equi- librium level of employment will be LE. Total national output (which equals total national income) will be represented by the area 0RELE.

5 This national income will be distributed in two shares: 0WEELE going to workers in the form of wages and WERE remaining as capitalist profits (the return to owners of cap- ital). Hence, in a competitive market economy with constant-returns-to-scale

Functional distribution of income (factor share distribu- tion of income) The distri- bution of income to factors of production without regard to the ownership of the factors.

Factors of production Resources or inputs required to produce a good or a service, such as land, labor, and capital.

225CHAPTER 5 Poverty, Inequality, and Development

production functions (a doubling of all inputs doubles output), factor prices are determined by factor supply and demand curves, and factor shares always combine to exhaust the total national product. Income is distributed by func- tion—laborers are paid wages, owners of land receive rents, and capitalists ob- tain profits. It is a neat and logical theory in that each and every factor gets paid only in accordance with what it contributes to national output, no more and no less. In fact, as you may recall from Chapter 3, this model of income distribution is at the core of the Lewis theory of modern-sector growth based on the reinvestment of rising capitalist profits.

Unfortunately, the relevance of the functional theory is greatly diminished by its failure to take into account the important role and influence of nonmar- ket forces such as power in determining these factor prices—for example, the role of collective bargaining between employers and trade unions in the set- ting of modern-sector wage rates, and the power of monopolists and wealthy landowners to manipulate prices on capital, land, and output to their own personal advantage. Appendix 5.1 examines the economic implications of fac- tor price distortions, and we return to consider their implications for policy at the end of this chapter.

The Ahluwalia-Chenery Welfare Index (ACWI)

A final approach to accounting for the distribution of income in assessing the quality of growth is to value increases in income for all individuals but to assign a higher weight to income gains by lower-income individuals than to gains by higher-income individuals. Perhaps the best-known example is the Ahluwalia- Chenery Welfare Index (ACWI), which is explained in Appendix 5.2.

W a

g e

r a

te

Profits

Wages

Employment

0

R

EWE

LE

SL

DL = MPL

6125_05_FG005

Figure 5.5 Functional income Distribution in a Market economy: An illustration

226 PART Two Problems and Policies: Domestic

5.2 Measuring absolute Poverty

Now let’s switch our attention from relative income shares of various percen- tile groups within a given population to the fundamentally important ques- tion of the extent and magnitude of absolute poverty in developing countries.

Income Poverty

In Chapter 2, we defined the extent of absolute poverty as the number of peo- ple who are unable to command sufficient resources to satisfy basic needs. They are counted as the total number living below a specified minimum level of real income—an international poverty line. That line knows no national boundaries, is independent of the level of national per capita income, and takes into account differing price levels by measuring poverty as anyone liv- ing on less than $1.25 a day or $2 per day in PPP dollars. Absolute poverty can and does exist, therefore, as readily in New York City as it does in Kolkata, Cairo, Lagos, or Bogotá, although its magnitude is likely to be much lower in terms of percentages of the total population.

Absolute poverty is sometimes measured by the number, or “headcount,” H, of those whose incomes fall below the absolute poverty line, Yp. When the headcount is taken as a fraction of the total population, N, we define the headcount index, H/N (also referred to as the “headcount ratio”). The pov- erty line is set at a level that remains constant in real terms so that we can chart our progress on an absolute level over time. The idea is to set this level at a standard below which we would consider a person to live in “absolute human misery,” such that the person’s health is in jeopardy.

Of course, to define a minimum health standard that is invariant across historical epochs is an impossibility, in part because technology changes over time. For example, today we have 15-cent oral rehydration therapy packets that can save the life of a child in Malawi. Not long ago, the death of a child after a diarrheal disease would be taken as a sad but inevitable part of life, whereas today we regard such a death as a catastrophic moral failure of the international community. We simply come as close as we can to establishing a reasonable minimum standard that might hold over a few decades so that we can estimate more carefully how much progress we have made on a (more) absolute rather than a (highly) relative scale.

Certainly one would not accept the international poverty level of $1.25 a day in an unquestioning way when planning local poverty work. One practical strategy for determining a local absolute poverty line is to start by defining an adequate basket of food, based on nutritional requirements from medical studies of required calories, protein, and micronutrients. Then, using local household survey data, one can identify a typical basket of food purchased by households that just barely meet these nutritional requirements. One then adds other expen- ditures of this household, such as clothing, shelter, and medical care, to deter- mine the local absolute poverty line. Depending on how these calculations are done, the resulting poverty line may come to more than $1.25 per day at PPP.

However, simply counting the number of people below an agreed-on poverty line has serious limitations. For example, if the poverty line is set at U.S. $450 per person, it makes a big difference whether most of the absolute

Absolute poverty The situation of being unable or only barely able to meet the subsistence essentials of food, clothing, and shelter.

Headcount index The proportion of a country’s population living below the poverty line.

227CHAPTER 5 Poverty, Inequality, and Development

poor earn $400 or $300 per year. Both are accorded the same weight when calculating the proportion of the population that lies below the poverty line; clearly, however, the poverty problem is much more serious in the latter in- stance. Economists therefore attempt to calculate a total poverty gap (TPG) that measures the total amount of income necessary to raise everyone who is below the poverty line up to that line. Figure 5.6 illustrates how we can meas- ure the total poverty gap as the shaded area between poverty line, PV, and the annual income profile of the population.

Even though in both country A and country B, 50% of the population falls below the same poverty line, the TPG in country A is greater than in country B. Therefore, it will take more of an effort to eliminate absolute poverty in country A.

The TPG—the extent to which the incomes of the poor lie below the pov- erty line—is found by adding up the amounts by which each poor person’s income, Yi, falls below the absolute poverty line, Yp, as follows:

TPG = a H

i = 1 1Yp - Yi2 (5.1)

We can think of the TPG in a simplified way (i.e., no administrative costs or general equilibrium effects are accounted for) as the amount of money per day it would take to bring every poor person in an economy up to our defined minimum income standards. On a per capita basis, the average poverty gap (APG) is found by dividing the TPG by the total population:

APG = TPG

N (5.2)

Often we are interested in the size of the average poverty gap in relation to the poverty line, so we would use as our income shortfall measure the normalized pov- erty gap (NPG): NPG = APG/Yp; this measure lies between 0 and 1 and so can be useful when we want a unitless measure of the gap for easier comparisons.

Total poverty gap (TPG) The sum of the difference between the poverty line and actual income levels of all people living below that line.

Country A

TPG TPG

Country B

P PV V

A n

n u

a l i

n co

m e

A n

n u

a l i

n co

m e

0 50 100 0 50 100 Percentage of populationPercentage of population

(b) A relatively small poverty gap(a) A relatively large poverty gap

6125_05_FG006

FIgure 5.6 Measuring the Total Poverty gap

228 PART Two Problems and Policies: Domestic

Another important poverty gap measure is the average income shortfall (AIS), which is the total poverty gap divided by the headcount of the poor: AIS = TPG/H. The AIS tells us the average amount by which the income of a poor person falls below the poverty line. This measure can also be divided by the poverty line to yield a fractional measure, the normalized income shortfall (NIS): NIS = AIS/Yp.

The Foster-greer-Thorbecke Index We are also often interested in the de- gree of income inequality among the poor, such as the Gini coefficient among those who are poor, Gp, or alternatively, the coefficient of variation (CV) of incomes among the poor, CVp. One reason that the Gini or CV among the poor can be important is that the impact on poverty of economic shocks can differ greatly, depending on the level and distribution of resources among the poor. For example, if the price of rice rises, as it did in 1998 in Indonesia, low-in- come rice producers, who sell a little of their rice on local markets and whose incomes are slightly below the absolute poverty line, may find that this price rise increases their incomes to bring them out of absolute poverty. On the other hand, for those with too little land to be able to sell any of the rice they grow and who are net buyers of rice on markets, this price increase can greatly worsen their poverty. Thus, the most desirable measures of poverty would also be sensitive to the distribution of income among the poor.

As is the case with inequality measures, there are criteria for a desirable poverty measure that are widely accepted by development economists: the anonymity, population independence, monotonicity, and distributional sen- sitivity principles. The first two principles are very similar to the properties we examined for inequality indexes: Our measure of the extent of poverty should not depend on who is poor or on whether the country has a large or small population. The monotonicity principle means that if you add income to someone below the poverty line, all other incomes held constant, poverty can be no greater than it was.6 The distributional sensitivity principle states that, other things being equal, if you transfer income from a poor person to a richer person, the resulting economy should be deemed strictly poorer. The headcount ratio measure satisfies anonymity, population independence, and monotonicity, but it fails on distributional sensitivity. The simple headcount fails even to satisfy the population independence principle.

A well-known poverty index that in certain forms satisfies all four criteria is the Foster-Greer-Thorbecke (FGT) index, often called the Pα class of pov- erty measures.7 The Pα index is given by

Pα = 1 N

a H

i = 1 a

Yp - Yi Yp

b α (5.3)

where Yi is the income of the ith poor person, Yp is the poverty line, and N is the population. Depending on the value of α, the Pα index takes on different forms. If α = 0, the numerator is equal to H, and we get the headcount ratio, H/N. Unfortunately, this measure is the same whether those in poverty earn 90 cents per day or 50 cents per day, so it cannot reveal the depth of poverty.

If α = 1, we get the normalized (per capita) poverty gap. An alternative formula that can be derived for P1 is given by P1 = (H/N)*(NIS), that is, the headcount ratio (H/N) times the normalized income shortfall (NIS). So, P1 has

Foster-Greer-Thorbecke (FGT) index A class of measures of the level of absolute poverty.

229CHAPTER 5 Poverty, Inequality, and Development

the properties that poverty goes up whenever either the fraction of people in poverty goes up or the fractional income deficits (poverty depth) go up (or both)—in general, this makes it a better measure than P0.

If α = 2, we account for poverty severity, in that the impact on measured poverty of a gain in income by a poor person increases in relation to the square of the distance of the person from the poverty line. For example, raising the income of a person from a household living at half the per capita poverty line by, say, one penny per day would have five times the impact on poverty re- duction as would raising by the same amount the income of a person living at 90% of the poverty line; this differing magnitude results from squaring the poverty gaps, so the P2 measure captures the severity of poverty.

As a numerical example of the calculation of P2, consider an 8-person econ- omy with a poverty line of 1, and a hypothetical income distribution of: (0.6, 0.6, 0.8, 0.8, 2, 2, 6, 6). The headcount is 4, because two people have incomes of 0.6 and two people have incomes of 0.8; but the others have incomes above the poverty line. Using these numbers, we can find the P2 level of poverty from equation 5.3:

P2 = (1/8)[0.4 2 + 0.42 + 0.22 + 0.22] = (1/8) [0.16 + 0.16 + 0.04

+ 0.04] = 0.4/8 = 0.05

Note that P2 can be expressed in an alternative form to add further intuition. If α = 2, the resulting measure, P2, can be rewritten as

8

P2 = a H N b 3NIS2 + 11 - NIS221CVP224 (5.4)

As Equation 5.4 shows, P2 contains the CVp measure, and it satisfies all four of the poverty axioms.9 Clearly, P2 increases whenever H/N, NIS, or CVp increases. Note from the formula that there is a greater emphasis on the distribution of income among the poor (CVp) when the normalized income shortfall is small and a lesser emphasis when the NIS is large.

The P2 poverty measure, also known as the squared poverty gap index, has become a standard of income poverty measure used by the World Bank and other agencies, and it is used in empirical work on income poverty because of its sensitivity to the depth and severity of poverty. Mexico uses the P2 poverty measure to allocate funds for education, health, and welfare programs for the poor (in particular in the Progresa/Oportunidades Program, described at the end of Chapter 8), in accordance with the regional intensity of poverty.10

Another reason to prefer P2 (or at least P1) over P0 is that standard headcount measures also have the perverse property of creating incentives for officials to focus efforts on the poor who are closest to the poverty line—because that is the easiest and cheapest way for them to demonstrate progress. We encountered a version of this problem in Chapter 1—a critique of the Millennium Develop- ment Goals focus on reducing the fraction of those living below the poverty line.

Values of P0 and P2 for selected developing countries are found in Table 5.6 later in this chapter.

Person-equivalent Headcounts Although P1 and P2 are more informative measures, which provide better incentives to poverty programs than P0, many agencies (including U.S. Agency for International Development—USAID)

230 PART Two Problems and Policies: Domestic

continue to report progress primarily if not exclusively in terms of P0 head- count measures—apparently responding to public and legislative expec- tations to discuss poverty in terms of numbers of people. Given a political need to feature “headline” headcount measures, a partial improvement is to convert changes in the poverty gap into its headcount-equivalent (based on the initial average income shortfall). If aid agencies featured a supplementary headcount-equivalent, they could report in terms of numbers of people while accounting for changes in poverty depth. Estimates using this approach show progress against poverty in many countries is significantly greater than re- vealed using conventional headcount measures alone. 11

Multidimensional Poverty Measurement Poverty cannot be adequately measured with income alone, as Amartya Sen’s capability framework, exam- ined in Chapter 1, makes apparent. To fill this gap, Sabina Alkire and James Foster have extended the FGT index to multiple dimensions.12

As always, the first step in measuring poverty is to know which people are poor. In the multidimensional poverty approach, a poor person is identified through what is called the “dual cutoff method”: first, the cutoff levels within each of the dimensions (analogous to falling below a poverty line such as $1.25 per day if income poverty were being addressed) and second, the cutoff of the number of dimensions in which a person must be deprived (below the line) to be deemed multidimensionally poor. Using calculations analogous to the single-dimensional Pα index, the multidimensional Mα index is constructed. The most basic measure is the fraction of the population in multidimensional poverty—the multidimensional headcount ratio HM.

The most common measure in practice is M0, the adjusted headcount ratio, which uses ordinal data and is similar conceptually to the poverty gap P1 (which again can be expressed as the headcount ratio times the normalized in- come shortfall). M0 may be represented by the product of the multidimensional headcount ratio times the average fraction of dimensions in which the poor are deprived (or “average intensity of poverty” A, that is, M0 = HM *A. (In contrast to the simple multidimensional headcount ratio, the adjusted multidimensional headcount ratio satisfies the desirable property (called “dimensional monoto- nicity”) that if the average fraction of deprivations increases, so does M0).

In applied studies, proxy measures, called indicators, are used for each of the selected dimensions. Details of the way this measure has been constructed and ap- plied in the UNDP Multidimensional Poverty Index and findings across countries are reported in Section 5.4, when we apply the poverty measures to examine the extent of poverty in different countries and regions. Another wisely used applica- tion is the Women’s Empowerment in Agriculture Index, referred to in Chapter 9.

5.3 Poverty, Inequality, and Social Welfare

What’s So Bad about Extreme Inequality?

Throughout this chapter, we are assuming that social welfare depends posi- tively on the level of income per capita but negatively on poverty and nega- tively on the level of inequality, as these terms have just been defined. The

231CHAPTER 5 Poverty, Inequality, and Development

problem of absolute poverty is obvious. No civilized people can feel satisfied with a state of affairs in which their fellow humans exist in conditions of such absolute human misery, which is probably why every major religion has em- phasized the importance of working to alleviate poverty and is at least one of the reasons why international development assistance has the nearly univer- sal support of every democratic nation. But it may reasonably be asked, if our top priority is the alleviation of absolute poverty, why should relative inequality be a concern? We have seen that inequality among the poor is a critical factor in understanding the severity of poverty and the impact of market and policy changes on the poor, but why should we be concerned with inequality among those above the poverty line?

There are three major answers to this question. First, extreme income ine- quality leads to economic inefficiency. This is partly because at any given aver- age income, the higher the inequality is, the smaller the fraction of the popula- tion that qualifies for a loan or other credit. Indeed, one definition of relative poverty is the lack of collateral. When low-income individuals (whether they are absolutely poor or not) cannot borrow money, they generally cannot adequately educate their children or start and expand a business. Moreover, with high in- equality, the overall rate of savings in the economy tends to be lower, because the highest rate of marginal savings is usually found among the middle classes. Although the rich may save a larger dollar amount, they typically save a smaller fraction of their incomes, and they almost always save a smaller fraction of their marginal incomes. Landlords, business leaders, politicians, and other rich elites are known to spend much of their incomes on imported luxury goods, gold, jewelry, expensive houses, and foreign travel or to seek safe havens abroad for their savings in what is known as capital flight. Such savings and investments do not add to the nation’s productive resources; in fact, they represent substan- tial drains on these resources. In short, the rich do not generally save and invest significantly larger proportions of their incomes (in the real economic sense of productive domestic saving and investment) than the middle class or even the poor.13 Furthermore, inequality may lead to an inefficient allocation of as- sets. As you will see in Chapter 8, high inequality leads to an overemphasis on higher education at the expense of quality universal primary education, which not only may be inefficient but is also likely to beget still more inequality in incomes. Moreover, as you will see in Chapter 9, high inequality of land owner- ship—characterized by the presence of huge latifundios (plantations) alongside tiny minifundios that are incapable of supporting even a single family—also leads to inefficiency because the most efficient scales for farming are family and medium-size farms. The result of these factors can be a lower average income and a lower rate of economic growth when inequality is high.14

The second reason to be concerned with inequality above the poverty line is that extreme income disparities undermine social stability and solidarity. Also, high inequality strengthens the political power of the rich and hence their economic bargaining power. Usually this power will be used to encour- age outcomes favorable to themselves. High inequality facilitates rent seeking, including actions such as excessive lobbying, large political donations, bribery, and cronyism. When resources are allocated to such rent-seeking behaviors, they are diverted from productive purposes that could lead to faster growth. Even worse, high inequality makes poor institutions very difficult to improve,

232 PART Two Problems and Policies: Domestic

because the few with money and power are likely to view themselves as worse off from socially efficient reform, and so they have the motive and the means to resist it (see Chapter 2). Of course, high inequality may also lead the poor to support populist policies that can be self-defeating. Countries with ex- treme inequality, such as El Salvador and Iran, have undergone upheavals or extended civil strife that have cost countless lives and set back development progress by decades. High inequality is also associated with pathologies such as higher violent crime rates. In sum, with high inequality, the focus of politics often tends to be on supporting or resisting the redistribution of the existing economic pie rather than on policies to increase its size (Chapter 11 examines these concerns in more detail).15

Finally, extreme inequality is generally viewed as unfair. The philosopher John Rawls proposed a thought experiment to help clarify why this is so.16 Suppose that before you were born into this world, you had a chance to se- lect the overall level of inequality among the earth’s people but not your own identity. That is, you might be born as Bill Gates, but you might be born as the most wretchedly poor person in rural Ethiopia with equal probability. Rawls calls this uncertainty the “veil of ignorance.” The question is, facing this kind of risk, would you vote for an income distribution that was more equal or less equal than the one you see around you? If the degree of equality had no effect on the level of income or rate of growth, most people would vote for nearly perfect equality. Of course, if everyone had the same income no mat- ter what, there would be little incentive to work hard, gain skills, or innovate. As a result, most people vote for some inequality of income outcomes, to the extent that these correspond to incentives for hard work or innovation. But even so, most vote for less inequality than is seen in the world (or in virtually any country) today. This is because much of the inequality we observe in the world is based on luck or extraneous factors, such as the inborn ability to kick a football or the identity of one’s great-grandparents.

For all these reasons, for this part of the analysis we will write welfare, W, as

W = W1Y, I, P2 (5.5)

where Y is income per capita and enters our welfare function positively, I is inequality and enters negatively, and P is absolute poverty and also enters negatively. These three components have distinct significance, and we need to consider all three elements to achieve an overall assessment of welfare in devel- oping countries. (A similar framework can be applied to health and education.)

Dualistic Development and Shifting Lorenz Curves: Some Stylized Typologies

As introduced by Gary Fields, Lorenz curves may be used to analyze three limiting cases of dualistic development:17

1. The modern-sector enlargement growth typology, in which the two-sector economy develops by enlarging the size of its modern sector while main- taining constant wages in both sectors. This is the case depicted by the

233CHAPTER 5 Poverty, Inequality, and Development

Lewis model in Chapter 3. It corresponds roughly to the historical growth pattern of Western developed nations and, to some extent, the pattern in East Asian economies such as China, South Korea, and Taiwan.

2. The modern-sector enrichment growth typology, in which the economy grows but such growth is limited to a fixed number of people in the mod- ern sector, with both the numbers of workers and their wages held con- stant in the traditional sector. This roughly describes the experience of many Latin American and African economies.

3. The traditional-sector enrichment growth typology, in which all of the ben- efits of growth are divided among traditional-sector workers, with little or no growth occurring in the modern sector. This process roughly de- scribes the experiences of countries whose policies focused on achieving substantial reductions in absolute poverty even at very low incomes and with relatively low growth rates, such as Sri Lanka, and the state of Kerala in southwestern India.

Using these three special cases and Lorenz curves, Fields demonstrated the validity of the following propositions (reversing the order just presented):

1. In the traditional-sector enrichment typology, growth results in higher in- come, a more equal relative distribution of income, and less poverty. Tradi- tional-sector enrichment growth causes the Lorenz curve to shift uniformly upward and closer toward the line of equality, as depicted in Figure 5.7.

2. In the modern-sector enrichment growth typology, growth results in higher in- comes, a less equal relative distribution of income, and no change in poverty.

100

P e

rc e

n ta

g e

o f

in co

m e

0 100 Percentage of income recipients

6125_05_FG007

FIgure 5.7 Improved Income Distribution under the Traditional- Sector enrichment growth Typology

234 PART Two Problems and Policies: Domestic

Modern-sector enrichment growth causes the Lorenz curve to shift down- ward and farther from the line of equality, as shown in Figure 5.8.

3. Finally, in the case of Lewis-type, modern-sector enlargement growth, abso- lute incomes rise and absolute poverty is reduced, but the Lorenz curves will always cross, indicating that we cannot make any unambiguous state- ment about changes in relative inequality: It may improve or worsen. Fields shows that if, in fact, this style of growth experience is predominant, in- equality is likely first to worsen in the early stages of development and then to improve. The crossing of the Lorenz curves is demonstrated in Figure 5.9.

The explanation for the crossing in Figure 5.9 is as follows: The poor who remain in the traditional sector have their incomes unchanged, but these incomes are now a smaller fraction of the larger total, so the new Lorenz curve, L2, lies below the old Lorenz curve, L1, at the lower end of the income distri- bution scale. Each modern-sector worker receives the same absolute income as before, but now the share received by the richest income group is smaller, so the new Lorenz curve lies above the old one at the higher end of the income distribution scale. Therefore, somewhere in the middle of the distribution, the old and new Lorenz curves must cross.18

These three typologies offer different predictions about what will hap- pen to inequality in the course of economic growth. With modern-sector enrichment, inequality rises steadily, while under traditional-sector enrich- ment, inequality falls steadily. Under modern-sector enlargement, inequality first rises and then falls;19 if this admittedly highly stylized process of devel- opment were occurring, we would not be concerned about the temporary rise in inequality, because in addition to being temporary, it would be reflecting a

FIgure 5.8 Worsened Income Distribution under the Modern-Sector enrichment growth Typology

100

P e

rc e

n ta

g e

o f

in co

m e

0 100 Percentage of income recipients

6125_05_FG008

235CHAPTER 5 Poverty, Inequality, and Development

process in which citizens are, one by one, achieving incomes above the abso- lute poverty line.20

These observations tell us that we have to qualify our conclusion that a rise in inequality is inherently bad. In some cases, inequality may increase on a tem- porary basis due to causes that will eventually make everyone better off and ultimately lower inequality. However, with modern-sector enrichment growth, the increase in inequality is not later reversed, and the poor do not escape their poverty.21 So we need to be careful about drawing conclusions from short-run changes in economic statistics before we know more about the underlying changes in the real economy that have given rise to these statistics. The process of modern-sector enlargement growth suggests a possible mechanism that can give rise to Kuznets’s “inverted-U” hypothesis, so we turn to this question next.

Kuznets’s Inverted-U Hypothesis

Simon Kuznets suggested that in the early stages of economic growth, the dis- tribution of income will tend to worsen; only at later stages will it improve.22 This observation came to be characterized by the “inverted-U” Kuznets curve because a longitudinal (time-series) plot of changes in the distribution of in- come—as measured, for example, by the Gini coefficient—seemed, when per capita GNI expanded, to trace out an inverted U-shaped curve in some of the cases Kuznets studied, as illustrated in Figure 5.10.

Explanations as to why inequality might worsen during the early stages of economic growth before eventually improving are numerous. They almost always relate to the nature of structural change. Early growth may, in accord- ance with the Lewis model, be concentrated in the modern industrial sector, where employment is limited but wages and productivity are high.

FIgure 5.9 Crossing lorenz Curves in the Modern-Sector enlargement growth Typology

100 P

e rc

e n

ta g

e o

f in

co m

e

0 100 Percentage of income recipients

L1 L2

6125_05_FG009

Kuznets curve A graph reflecting the relationship between a country’s income per capita and its inequality of income distribution.

236 PART Two Problems and Policies: Domestic

As just noted, the Kuznets curve can be generated by a steady process of modern-sector enlargement growth as a country develops from a traditional to a modern economy. Alternatively, returns to education may first rise as the emerging modern sector demands skills and then may fall as the supply of ed- ucated workers increases and the supply of unskilled workers falls. So while Kuznets did not specify the mechanism by which his inverted-U hypothesis was supposed to occur, it could in principle be consistent with a sequential pro- cess of economic development. But as shown earlier, traditional- and modern- sector enrichment would tend to pull inequality in opposing directions, so the net change in inequality is ambiguous, and the validity of the Kuznets curve is an empirical question.

Disregarding the merits of the methodological debate, few development economists would argue that the Kuznets sequence of increasing and then de- clining inequality is inevitable. There are now enough case studies and specific examples of countries such as Taiwan, South Korea, Costa Rica, and Sri Lanka to demonstrate that higher income levels can be accompanied by falling and not rising inequality. It all depends on the nature of the development process.

evidence on the Inverted-u Hypothesis Let us look at data collected from 18 countries on the percentage shares in total national income going to differ- ent percentile groups (see Table 5.2). Though methods of collection, degree of coverage, and specific definitions of personal income may vary from coun- try to country, the figures recorded in Table 5.2 give a first approximation of the magnitude of income inequality in developing countries. For example, we see that in Zambia, the poorest 20% (first quintile) of the population receives only 3.6% of the income, while the highest 10% and 20% (fifth quintile) receive 38.9% and 55.2%, respectively. By contrast, in a relatively equal developed country like Japan, the poorest 20% receives a much higher 10.6% of the in- come, while the richest 10% and 20% get only 21.7% and 35.7%, respectively.

FIgure 5.10 The “Inverted-u” Kuznets Curve

0.75

0.50

0.35

0.25

0

Gross national income per capita

G in

i co

e ff

ic ie

n t

6125_05_FG010

237CHAPTER 5 Poverty, Inequality, and Development

The income distribution of the United States, a relatively less equal developed country, is given for comparison in Table 5.2.

Consider now the relationship, if any, between levels of per capita income and degree of inequality. Are higher incomes associated with greater or lesser inequality, or can no definitive statement be made? Table 5.3 on page 240 pro- vides data on income distribution in relation to per capita GNI for a sampling of countries, arranged from lowest to highest in terms of per capita income. What clearly emerges from Table 5.3 is that per capita incomes are not necessarily re- lated to inequality. The very poorest countries, such as Ethiopia, may have low inequality simply because there is so little income. But even very poor countries such as Mozambique and Zambia have extremely high inequality by interna- tional standards. Although many high-inequality Latin American countries are found in the middle-income range, this range also includes countries such as Egypt and Indonesia, as well as eastern European countries, with low inequal- ity. High-income countries do tend to be somewhat more equal than middle- income countries, but again, there is wide variation in inequality levels. In re- cent years, there has even been a tendency for inequality to rise in high-income countries and to fall at least somewhat in several Latin American countries.

In fact, the Kuznets curve that is seen in the data is now understood to be partially a statistical fluke resulting from the fact that for extraneous histori- cal reasons, most Latin American countries just happen to have both a middle level of income and a high level of inequality (see Box 5.1).

Detailed longitudinal studies of developing countries show a very mixed pattern. Juan Luis Lonondro found an inverted U for Colombia, but Harry Oshima found no particular pattern among several Asian countries.23 In fact, for many

Table 5.2 Selected Income Distribution estimates

Source: based on World Bank, World Development Indicators, 2010. (Washington, D.C.: World Bank, 2010), tab. 2.9.

    Quintile    

Country

lowest 10%

1st

2nd

3rd

4th

5th

Highest 10%

Year

Bangladesh Brazil China Colombia Costa Rica Guatemala Honduras India Jamaica Namibia Pakistan Peru Philippines South Africa Tanzania Zambia Japan United States

4.3 1.1 2.4 0.8 1.6 1.3 0.7 3.6 2.1 0.6 3.9 1.3 2.4 1.3 3.1 1.3 4.8 1.9

9.4 3.0 5.7 2.3 4.4 3.4 2.5 8.1 5.2 1.5 9.1 3.6 5.6 3.1 7.3 3.6

10.6 5.4

12.6 6.9 9.8 6.0 8.5 7.2 6.7

11.3 9.0 2.8

12.8 7.8 9.1 5.6

11.8 7.8

14.2 10.7

16.1 11.8 14.7 11.0 12.7 12.0 12.1 14.9 13.8 5.5

16.3 13.0 13.7 9.9

16.3 12.8 17.6 15.7

21.1 19.6 22.0 19.1 19.7 19.5 20.4 20.4 20.9 12.0 21.3 20.8 21.2 18.8 22.3 20.6 22.0 22.4

40.8 58.7 47.8 61.6 54.6 57.8 58.4 45.3 51.2 78.3 40.5 54.8 50.4 62.7 42.3 55.2 35.7 45.8

26.6 43.0 31.4 45.9 38.6 42.4 42.2 31.1 35.6 65.0 26.5 38.4 33.9 44.9 27.0 38.9 21.7 29.9

2005 2007 2005 2006 2007 2006 2006 2005 2004 1993 2005 2007 2006 2000 2001 2005 1993 2000

238 PART Two Problems and Policies: Domestic

bOX 5.1 The latin america effect

Gary Fields and George Jakubson used a combi-nation of both cross-sectional and longitudinal (time-series) data to consider whether the inverted-U could result from the Latin American effect and how patterns might differ across countries. Figure 5.11 plots a combination of data from the 35 countries in Fields and Jakubson’s data set, where reliable estimates of the Gini coefficient have been available for vari- ous developing countries at different points in time. The inverted-U relationship, tracing the triangles, is a computer-generated parabola that best fits the data under standard statistical criteria. Observations on Latin American countries are circled: All of the high- est-inequality countries in their data come from that region. Statistically, when the Latin American identity of the country is controlled for, the inverted-U drawn

in Figure 5.11 tends to disappear in this data set and others as well.24

So the question is, what happens over time? In Figure 5.12 on page 239, selected countries from the data in Figure 5.11 have been isolated. As can be seen, the data from Brazil, which have the label 1 in the diagram, do plainly show an inverted-U pattern. Data from Hong Kong and Singapore, in contrast, labeled 4 and 5 in the diagram, appear to reflect a U-shaped pat- tern. But when these separate experiences are merged into one picture, the eyes (and the computer) mislead- ingly trace an inverted U in the data taken as a whole. This reinforces the great importance of understanding what gives rise to the statistical patterns in the data rather than taking them at face value.

0.63 0.61 0.59 0.57 0.55 0.53 0.51 0.49 0.47 0.45 0.43 0.41 0.39 0.37 0.35 0.33 0.31 0.29 0.27

0 1 2 3 4 5 6 GNI per capita (thousands of 1980 U.S. dollars)

G in

i co

e ff

ic ie

n t

Actual data Predicted relationship Latin American country

FIgure 5.11 Kuznets Curve with latin american Countries Identified

Source: Gary S. Fields, Distribution and Development: A New Look at the Developing World (Cambridge, Mass.: MIT Press, 2001), ch. 3, p. 46. © 2001 Massachusetts Institute of Technology, by permission of The MIT Press.

239CHAPTER 5 Poverty, Inequality, and Development

countries, there is no particular tendency for inequality to change in the pro- cess of economic development. Inequality seems to be a rather stable part of a country’s socioeconomic makeup, altered significantly only as a result of a substantial upheaval or systematic policies. East Asia achieved its relatively low inequality largely from exogenous forces: the U.S. occupation of Japan, the Nationalist takeover of Taiwan, and the expulsion of the Japanese from South Korea. In all three cases, land reform that had far-reaching effects on inequality was implemented (we examine land reform in Chapter 9). But inequality can be gradually reduced through well-implemented policies to promote pro-poor growth over time. With regressive policies, inequality may rise over time.

Growth and Inequality

Having examined the relationship between inequality and levels of per capita income, let us look now briefly at the relationship, if any, between economic growth and inequality. During the 1960s and 1990s, per capita growth in East Asia averaged 5.5% while that of Africa declined by 0.2%, yet both Gini coef- ficients remained essentially unchanged. Once again, it is not just the rate but

FIgure 5.12 Plot of Inequality Data for Selected Countries

Source: Gary S. Fields, Distribution and Development: A New Look at the Developing World (Cambridge, Mass.: MIT Press, 2001), ch. 3, p. 44. © 2001 Massachusetts Institute of Technology, by permission of The MIT Press.

1 1 1

1

1 1

2

2

2 2

4

4 4 4

5

5

5

3 3 3 3 3 3 3

0 1 2 3 4 5 6 GNI per capita (thousands of 1980 U.S. dollars)

0.61 0.59 0.57 0.55 0.53 0.51 0.49 0.47 0.45 0.43 0.41 0.39 0.37 0.35 0.33

G in

i co

e ff

ic ie

n t

1 = Brazil 2 = Costa Rica 3 = Pakistan 4 = Hong Kong 5 = Singapore

6125_05_FG012

240 PART Two Problems and Policies: Domestic

also the character of economic growth (how it is achieved, who participates, which sectors are given priority, what institutional arrangements are designed and emphasized, etc.) that determines the degree to which that growth is or is not reflected in improved living standards for the poor. Clearly, it is not neces- sary for inequality to increase for higher growth to be sustained.

5.4 absolute Poverty: extent and Magnitude

Like so much in economic development, the critical problem of eradicating ab- solute poverty is one of bad news and good news—of a glass that may be seen as either half empty or half full.

It is extremely difficult to arrive at a tight estimate of the extent of global poverty at any point in time. Major World Bank reports issued within a couple of years of each other have provided estimates of the dollar-a-day headcount that differ by tens of millions of people. This reflects the difficulty of the task. Another difficulty is determining the most appropriate cutoff income for ex- treme poverty. The $1-a-day line was first set in 1987 dollars, and for years the standard was $1.08 in 1993 U.S. purchasing power parity. In 2008, the equiva- lent line was reset at $1.25 at 2005 U.S. purchasing power. This (along with

Character of economic growth The distributive implications of economic growth as reflected in such factors as participation in the growth process and asset ownership.

Table 5.3 Income and Inequality in Selected Countries

Source: data from World Bank, World Development Indicators, 2010 (Washington, D.C.: World Bank, 2010), tabs. 1.1 and 2.9.

Country

Income Per Capita (u.S. $, 2008)

gini Coefficient

Survey Year for gini Calculation

low Income Ethiopia Mozambique Nepal Cambodia Zambia lower Middle Income India Cameroon Bolivia Egypt Indonesia upper Middle Income Namibia Bulgaria South Africa Argentina Brazil Mexico upper Income Hungary Spain Germany United States Norway

  280 380 400 640 950

1,040 1,150 1,460 1,800 1,880

  4,210 5,490 5,820 7,190 7,300 9,990

  12,810 31,930 42,710 47,930 87,340

  29.8 47.1 47.3 40.7 50.7

36.8 44.6 57.2 32.1 37.6

  74.3 29.2 57.8 48.8 55.0 51.6

  30.0 34.7 28.3 40.8 25.8

  2005 2003 2004 2007 2005

2005 2001 2007 2005 2007

  1993 2003 2000 2006 2007 2008

  2004 2000 2000 2000 2000

241CHAPTER 5 Poverty, Inequality, and Development

improved estimates of prices faced by the poor) resulted in an increase in the estimated number of the poor but did not change the conclusion that the num- ber in poverty has been falling markedly since 1990, most conspicuously due to progress in China. Even as updated to today’s dollars, the poverty line is to some degree arbitrary (although it has corresponded roughly to what many developing countries use and is at least related to expenditures of people who barely meet minimum nutrition).

The most recent systematic poverty estimates (available as of early 2014) show that in 2010 some 1.22 billion people lived below $1.25 per day, and some 2.36 billion below $2 per day (see Figure 5.13). The number of people living in $1.25 per day income poverty fell from about 1.94 billion in 1981 – a 37% reduction in the headcount. The drop in the number living on less than $2 per day was much smaller – under 8% - but this more modest de- cline was partly due to people whose incomes actually had crossed above the $1.25 per day, though still remained below $2 per day. These achieve- ments in reducing the number of people living in poverty are all the more impressive when we note that world population rose by 2.39 billion people (53%) between 1981 and 2010 (UN estimates). Thus the headcount ratio (frac- tion) living on less than $1.25 per day fell to about 18% by 2010 – approaching half (55%) of its 1990 level of 33%. Thus, the MDG of halving $1.25 per day poverty was close to having been met by 2010; and preliminary estimates show that this goal had been met – and indeed exceeded – by the end of 2013. Global and regional poverty trends are summarized in Figure 5.13. Note that the numbers of the poor who live in sub-Saharan Africa rose steadily through- out this three-decade period; but the headcount of the poor declined in other regions.

FIgure 5.13 global and regional Poverty Trends, 1981–2010

Source: Figure drawn using data from PovcalNet/World Bank; data downloaded 13 February 2014 from http://iresearch .worldbank.org/PovcalNet/index.htm?1.

0

0.5

1

1.5

2

2.5

3

1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2010

P eo

p le

li vi

n g

in P

o ve

rt y

(b il

li o

n s)

People living on less than $1.25 per day, other developing regions

People living on less than $1.25 per day, East Asia & Paci�c

People living on less than $1.25 per day, South Asia

People living on more than $1.25 and less than $2.00 per day all developing regions

People living on less than $1.25 per day, Sub- Saharan Africa

242 PART Two Problems and Policies: Domestic

The incidence of extreme poverty is very uneven around the developing world. Household survey–based estimates are regarded as the most accurate ways to estimate poverty incidence. Table 5.4 provides some survey-based poverty estimates by region at the $1.25 and $2 poverty lines. As can be seen, poverty incidence is very high in both South Asia, with about 40% below $1.25 per day, and in sub-Saharan Africa, with 51% below. But poverty severity is far higher in sub-Saharan Africa, with a squared poverty gap index P2 (in per- centage terms) at 11.05, far above that of South Asia at 3.64. Table 5.5 provides estimates for some specific countries in Africa, Asia, and Latin America at the $1.25 and $2 poverty lines. It can be seen that about 44% of India’s 2004 rural population lived below the $1.25-a-day poverty line, while almost 80% lived on less than $2 per day. In contrast, less than 36% of its urban population lived on less than $1.25 per day, although about 66% still lived on less than $2 per day.

Unfortunately, sub-Saharan Africa has shown far less progress than other developing regions. While the fraction living in poverty has fallen somewhat in the last decade, the headcount of individuals living in poverty rose dra- matically in the 1981–2010 period, from about 205 million to about 414 million (World Bank, 2013). The concentration of poverty may make it more difficult to redress. In most countries in other regions, the poverty gap has fallen along with the poverty headcount. But between 1981 and 2010, the average income of the extremely poor hardly increased in sub-Saharan Africa, remaining near an appalling 70 cents per person per day.

The Multidimensional Poverty Index (MPI)

The MPI is the most prominent application of multidimensional poverty measurement; it incorporates three dimensions at the household level: health, education, and wealth.

Table 5.4 regional Poverty Incidence, 2010

Source: data from World Bank, “PovcalNet,” http://iresearch.worldbank.org/PovcalNet, accessed 13 February 2014.

region Headcount ratio (P0) Poverty gap (P1) Squared Poverty gap (P2)

regional aggregation at $1.25 per Day East Asia and the Pacific Europe and Central Asia Latin America and the Caribbean Middle East and North Africa South Asia Sub-Saharan Africa Total regional aggregation at $2 per Day East Asia and the Pacific Europe and Central Asia Latin America and the Caribbean Middle East and North Africa South Asia Sub-Saharan Africa Total

  12.48 0.66 5.53 2.41

31.03 48.47 20.63

29.14 2.27

10.18 11.55 65.8 69.31 40.08

  2.82 0.21 2.89 0.55 7.09

20.95 6.3

9.42 0.64 4.67 2.66

22.86 35.22 15.32

  0.93 0.13 2.12 0.23 2.36

11.85 2.92  

4.05 0.3 3.13 0.99

10.19 22.03 7.79

243CHAPTER 5 Poverty, Inequality, and Development

Table 5.5 Income Poverty Incidence in Selected Countries

Country

Year

Per Capita Monthly Income

(2005 PPP)

Headcount ratio (%)

Poverty gap (%)

Squared Poverty gap (%)

gini Index

(%)

Incidence at $1.25 a Day; Poverty line at 38 (monthly equivalent) Bangladesh Benin Brazil Burkina Faso China—Rural China—Urban Côte d’Ivoire Guatemala* Honduras* India—Rural India—Urban Indonesia—Rural Indonesia—Urban Madagascar Mexico Mozambique Nicaragua* Nigeria Pakistan Peru Philippines Rwanda Senegal

2005 2003 2007 2003 2005 2005 2002 2006 2006 2004 2004 2005 2005 2005 2006 2002 2005 2003 2004 2006 2006 2000 2005

48.27 52.77

346.64 46.85 71.34

161.83 101.11 191.7

184.45 49.93 62.43 62.79 89.1

44.82 330.37 36.58

151.18 39.46 65.76

216.82 98.99 33.76 66.86

50.47 47.33 5.21

56.54 26.11 1.71

23.34 12.65 18.19 43.83 36.16 24.01 18.67 67.83 0.65

74.69 15.81 64.41 22.59 7.94

22.62 76.56 33.5

14.17 15.73 1.26

20.27 6.46 0.45 6.82 3.83 8.19

10.66 10.16 5.03 4.06

26.52 0.13 35.4 5.23

29.57 4.35 1.86 5.48

38.21 10.8

5.20 6.97 0.44 9.38 2.26 0.24 2.87 1.63 5.00 3.65 3.80 1.61 1.29

13.23 0.05

20.48 2.54 17.2 1.28 0.61 1.74

22.94 4.67

33.22 38.62 55.02 39.6

35.85 34.8

48.39 53.69 55.31 30.46 37.59 29.52 39.93 47.24 48.11 47.11 52.33 42.93 31.18 49.55 44.04 46.68 39.19

Incidence at $2 a Day; Poverty line at 60.84 (monthly equivalent) Bangladesh Benin Brazil Burkina Faso China—Rural China—Urban Côte d’Ivoire Guatemala* Honduras* India—Rural India—Urban Indonesia—Rural Indonesia—Urban Madagascar Mexico Mozambique Nicaragua* Nigeria Pakistan Peru Philippines Rwanda Senegal

2005 2003 2007 2003 2005 2005 2002 2006 2006 2004 2004 2005 2005 2005 2006 2002 2005 2003 2004 2006 2006 2000 2005

48.27 52.77

346.64 46.85 71.34

161.83 101.11 191.7

184.45 49.93 62.43 62.79 89.1

44.82 330.37 36.58

151.18 39.46 65.76

216.82 98.99 33.76 66.86

80.32 75.33 12.70 81.22 55.63 9.38

46.79 25.71 29.73 79.53 65.85 61.19 45.85 89.62 4.79

90.03 31.87 83.92 60.32 18.51 45.05 90.3

60.37

34.35 33.51 4.15

39.26 19.47 2.12

17.62 9.63

14.15 30.89 25.99 19.55 14.85 46.94 0.96

53.56 12.26 46.89 18.75 5.95

16.36 55.69 24.67

17.55 18.25 1.85

22.58 8.94 0.81 8.78 4.84 8.91

14.69 12.92 8.27 6.39 28.5 0.31

36.00 6.44 30.8 7.66 2.54 7.58 38.5

12.98

33.22 38.62 55.02 39.60 35.85 34.8

48.39 53.69 55.31 30.46 37.59 29.52 39.93 47.24 48.11 48.07 52.33 42.93 31.18 49.55 44.04 44.11 39.19

Source: data from World Bank, “PovcalNet,” http://iresearch.worldbank.org/PovcalNet.

Income is imperfectly measured, but even more important, the advantages provided by a given amount of income greatly differ, depending on circum- stances. To capture this idea, the United Nations Development Programme (UNDP) used its Human Poverty Index26 from 1997 to 2009.

244 PART Two Problems and Policies: Domestic

In 2010, the UNDP replaced the HPI with its Multidimensional Poverty Index (MPI); by building up the index from the household level, the MPI takes into account that there are negative interaction effects when people have multiple deprivations—worse poverty than can be seen by simply adding up separate deprivations for the whole country, then taking averages, and only then combining them.

The index’s creators report that they selected the three dimensions (health, education, and standard of living) and each of their corresponding indicators because they reflect problems often mentioned by the poor, they have been long considered important by the development community particularly as re- flected in the Millennium Development Goals (see Chapter 1), and they are well established philosophically as human rights or basic needs; naturally, reliable data also had to be available for enough countries when selecting spe- cific indicators for the index.

With respect to health, two indicators—whether any child has died in the family and whether any adult or child in the family is malnourished—are weighted equally (so each counts one-sixth toward the maximum possible deprivation in the MPI). Regarding education also, two indicators—whether not even one household member has completed five years of schooling and whether any school-age child is out of school for grades one through eight—are given equal weight (so again, each counts one-sixth toward the MPI). Finally, in terms of standard of living, equal weight is placed on six deprivations (each counting one-eighteenth toward the maximum possible): lack of electricity, insufficiently safe drinking water, inadequate sanitation, inadequate flooring, unimproved cooking fuel, and lack of more than one of five assets—telephone, radio, television, bicycle, and motorbike or similar vehicle.

Calculating deprivation in this way, individuals are then identified as “multi- dimensionally poor” when their family is deprived by a “weighted sum” of 0.3 or more (3 out of 10 points as calculated in practice). For concreteness, consider three examples of families whose members would be classified as multidimensionally poor. First, a person would get a value of 33% and thus be considered poor by hav- ing a child in the family who was malnourished, while at the same time the most ed- ucated person in the family received only three years of schooling. Second, a multi- dimensionally poor person might live in a household that had experienced a child’s death and was also deprived in at least three of the six living standard indicators, which also would sum to 1/6 + 1/18 + 1/18 + 1/18 = 1/3, or 33%. Third, they could live in a household that was deprived in the other three living standard indicators and in which there was a school-age child not attending school. But if there were no health or education deprivations, a person would have to live in a family which was deprived in all six standard-of-living indicators to be deemed poor. Thus, the MPI approach identifies the very poor by measuring a range of important house- hold deprivations directly, rather than only indirectly through income, then build- ing the index from household measures up to the aggregate measure. Rather than using already aggregated statistics in an index, the approach takes into account the multiplied or interactive harm done when multiple deprivations are experienced by individuals in the same family. In essence, the approach assumes that an individual’s lack of capability in one area can to a degree be made up for by other capabilities— but only to a degree. (Put differently, capabilities are treated as substitutes up to a point but then as complements.) This greatly augments measures used previously.

Multidimensional Poverty Index (MPI) A poverty mea- sure that identifies the poor using dual cutoffs for levels and numbers of deprivations, and then multiplies the per- centage of people living in poverty times the percent of weighted indicators for which poor households are deprived on average.

245CHAPTER 5 Poverty, Inequality, and Development

Finally, the actual MPI for the country (or region or group) is computed with the adjusted headcount ratio; as noted previously, a convenient way to express the resulting value is the product of the headcount ratio, HM (the per- centage of people living in multidimensional poverty) and the average inten- sity of deprivation, A (the percentage of weighted indicators for which poor households are deprived on average). The adjusted headcount ratio, HMA, is a special case of the broader class of multidimensional poverty measures devel- oped by Sabira Alkire and James Foster introduced earlier; HMA is readily cal- culated, and it also satisfies some desirable properties, including dimensional monotonicity, meaning that when a person deemed poor becomes deprived in another indicator, he or she is deemed even poorer.27

In its 2013 Human Development Report, the UNDP presents the MPI for 104 developing countries, based on the currently available data; some examples are given in Table 5.6. Brazil and Mexico have very low MPI levels of just 0.011 and 0.015, respectively, while the world’s most impoverished country for which data were available to compute the MPI, Niger, ranks 104th, with an MPI value of 0.642. The UNDP reports that there are nearly 1.6 billion peo- ple living in multidimensional poverty—several hundred million more than the estimated number living on an income of less than $1.25 per day. At the broadest level, the results are not out of line with what one might expect; sub-Saharan Africa has the highest proportion of people living in poverty, and South Asia has the largest number of people living in poverty.

The poorest country is Niger, the only country with an MPI higher than 0.6. Six other countries had an MPI higher than 0.5, all in sub-Saharan Africa: Ethiopia, Mali, Burkina Faso, Burundi, Mozambique, and Guinea (available earlier data also show Angola, the Central African Republic, and Somalia with an MPI greater than 0.5).

Countries outside Africa with high levels of multidimensional poverty for their regions include Bangladesh (with an MPI of 0.292), Cambodia (0.212), Haiti (0.299), Honduras (0.159), India (0.283), Lao PRD (0.267), Nepal (0.217) Pakistan (0.264), Timor-Leste (0.360), and Yemen (0.283).

The results show that knowing income poverty is not enough if our con- cern is with multidimensional poverty. For example, multidimensionally, Bangladesh is substantially less poor and Pakistan substantially poorer than would be predicted by these countries’ income poverty (this finding may be related to some of the comparisons in the end-of-chapter case study in Chapter 2). In Africa, Ethiopia is far more multidimensionally poor and Tanzania much less so than predicted by income poverty. Most Latin American countries studied rank worse on multidimensional poverty than on income poverty, but Colombia’s income and MPI poverty ranks are about the same.

The severity of poverty in Africa is also highlighted by some of the findings. In Guinea, Mali, and Niger, more than 50% are poor and live in a household in which at least one child has died. In Mozambique, Guinea, Burundi, Mali, Ethiopia, Burkina Faso, and Niger, more than 50% live in a poor household where no one has completed five years of education. Outside of Africa, 39% in India and 37% in Bangladesh live in a poor household where at least one child or woman is undernourished.28

Different regions in the same country can have very different MPIs. In Kenya, the MPI for Nairobi is close to that of Brazil. Central Kenya’s MPI is similar to

246 PART Two Problems and Policies: Domestic

Table 5.6 Multidimensional Poverty Index, Data for 2007–2011

Country and Survey Year MPI Percent Poor Thousands Poor Poverty Intensity (A)

Bangladesh 2007 (D) Brazil 2006 (N) Burundi 2005 (M) Bolivia, PS 2008 (D) Burkina Faso 2010 (D) Cambodia 2010 (D) Colombia 2010 (D) Congo, DR 2010 (M) Côte d’Ivoire 2005 (D) Dominican Republic 2007 (D) Egypt 2008 (D) Ethiopia 2011 (D) Ghana 2008 (D) Guinea 2005 (D) Haiti 2005/2006 (D) Honduras 2005/2006 (D) India 2005/2006 (D) Indonesia 2007 (D) Kenya 2008/2009 (D) Lao PRD 2006 (M) Liberia 2007 (D) Mali 2006 (D) Mexico 2006 (N) Madagascar 2008/2009 (D) Malawi 2010 (D) Mozambique 2009 (D) Nepal 2011 (D) Niger 2006 (D) Nigeria 2008 (D) Pakistan 2006/2007 (D) Peru 2008 (D) Philippines 2008 (D) Rwanda 2010 (D) Senegal 2010/2011 (D) Sierra Leone 2008 (D) South Africa 2008 (N) Tanzania, 2010 (D) Timor-Leste 2009/2010 (D) Uganda 2011 (D) Vietnam 2010/2011 (M) Yemen 2006 (M)

0.292 57.8 83,207 50.4 0.011 2.7 5,075 39.3 0.530 84.5 6,128 62.7 0.089 20.5 1,972 43.7 0.535 84.0 13,834 63.7 0.212 45.9 6,415 46.1 0.022 5.4 2,500 40.9 0.392 74.0 48,815 53.0 0.353 61.5 11,083 57.4 0.018 4.6 439 39.4 0.024 6.0 4,699 40.7 0.564 87.3 72,415 64.6 0.144 31.2 7,258 46.2 0.506 82.5 7,459 61.3 0.299 56.4 5,346 53.0 0.159 32.5 2,281 48.9 0.283 53.7 612,203 52.7 0.095 20.8 48,352 45.9 0.229 47.8 18,863 48.0 0.267 47.2 2,757 56.5 0.485 83.9 3,218 57.7 0.558 86.6 11,771 64.4 0.015 4.0 4,313 38.9 0.357 66.9 13,463 53.3 0.334 66.7 9,633 50.1 0.512 79.3 18,127 64.6 0.217 44.2 13,242 49.0 0.642 92.4 12,437 69.4 0.310 54.1 83,578 57.3 0.264 d 49.4 d 81,236 d 53.4 d 0.066 15.7 4,422 42.2 0.064 13.4 12,083 47.4 0.350 69.0 6,900 50.8 0.439 74.4 7,642 58.9 0.439 77.0 4,321 57.0 0.057 13.4 6,609 42.3 0.332 65.6 28,552 50.7 0.360 68.1 749 52.9 0.367 69.9 24,122 52.5 0.017 4.2 3,690 39.5 0.283 52.5 11,176 53.9

Key: D indicates data are from Demographic and Health Surveys, M indicates data are from Multiple Indicator Cluster Surveys, d indicates lower bound estimate, and N indicates data are from national surveys. Not all indicators were available for all countries; caution should thus be used in cross-country comparisons. Where data are missing, indicator weights are adjusted to total 100%.

Source: UNDP, Human Development Report, 2013, pp. 160–161.

that of Bolivia. And northeastern Kenya has a worse MPI even than Niger. There are also great inequalities across ethnic groups in Kenya, with 29% of the Embu considered multidimensionally poor, compared with a staggering 96% of the Turkana and Masai peoples. Great inequalities are also found in India, in which indigenous (“tribal”) peoples and low-ranked (“scheduled”) castes are far poorer than people from high-ranking castes. In the Delhi and Kerala regions, just 14 to 16% are MPI poor, but in Jharkhand and Bihar, 77 to 81% are MPI poor. Finally, changes in the MPI over time are examined for three countries: Ghana saw its MPI halved from 0.29 to 0.14; Bangladesh saw its MPI reduced by a more modest 22%; and in Ethiopia, the MPI fell by 16% in the periods studied.

247CHAPTER 5 Poverty, Inequality, and Development

As with all indexes, the MPI has some limitations. As mentioned, data are from the household rather than the individual level (such as whether any child of school age is out of school or whether any family member is under- nourished). It does not fully distinguish between past and present conditions (because its measure is whether a child has ever died). It does not distinguish differences within households (such as who may use the bicycle or whether the undernourished individuals are females). Proxies are imperfect; for exam- ple, nourishment does not capture micronutrient deficiencies. Sometimes a person has to be labeled nondeprived if data are missing, so the numbers may understate poverty somewhat. Education considers only inputs such as enrolling or attending for five years, not outputs such as being able to read. And the choice of basic assets is questionable; for example, even where a radio and a simple bicycle are present, a woman may have just one dress and the children may sleep on a rough concrete floor.

The MPI provides a new and fundamentally important way to measure poverty, to help us understand how poverty levels differ across and within countries, and also how the dimensions (or composition) of poverty can differ greatly in different settings. Ultimately, this should assist with better design and targeting of programs and policies and help us evaluate their performance more quickly and effectively.

For now, because of the way living standards and human development sur- veys are conducted, most of the usable data is at the household level, making it difficult to “drill down” to the individual level. Household data are far better than what used to be available; in fact, the availability of household data has already had a substantial impact on improving the study of development economics. It is a great improvement to be able to focus on what is happening at the family rather than the national level. Well-designed income poverty measures such as P2 will always be used for many purposes; but the MPI is likely to help usher in an era in which multidimensional poverty is examined in most assessments.

Chronic Poverty Research suggests that approximately one-third of all peo- ple who are income poor at any one time are chronically (always) poor. Andrew McKay and Bob Baulch provide a well-regarded “guesstimate” that about 300 to 420 million people were chronically poor at the $1-per-day level in the late 1990s. The other two-thirds are made up of families that are vulnerable to pov- erty and become extremely poor from time to time. These may be divided be- tween families usually poor but occasionally receiving enough income to cross the poverty line and families usually nonpoor but occasionally experiencing a shock that knocks them temporarily below the poverty line. Chronic poverty is concentrated in India, where the largest numbers are found, and in Africa, where the severity of poverty among the chronically poor is greatest.29

Problems of the poorest of the poor pose particular challenges. Ultrapoverty differs from conventional poverty in terms of depth (degree of deprivation), length (duration of time), and breadth (the number of dimensions, such as il- literacy and malnutrition).30 The mutual reinforcement among the different dimensions of poverty can potentially result in multiple mutually reinforcing poverty traps. This makes ultrapoverty a more difficult problem to address than conventional poverty, which can more often be redressed with simpler solutions such as microfinance (see Chapter 15) plus business training. The

248 PART Two Problems and Policies: Domestic

chronic nature and severity of ultrapoverty also make short-term policies more problematic. Poverty innovators such as Fazle Hasan Abed have concluded that conventional programs have often not reached the ultra-poor. An income- based definition of ultrapoverty is living on half the dollar-a-day poverty line, or 54 cents per day in 1993 dollars. According to International Food Policy Research Institute (IFPRI) estimates, 162 million people live below this stark income level, generally with malnutrition and other destitute conditions. The IFPRI study concluded:

poverty just below $1 a day has fallen faster than poverty below 50 cents a day, suggesting that it has been easier to reach those living closer to the dollar-a-day line rather than those living well below it. . . .The slow progress of poverty reduc- tion for the world’s most deprived indicates the presence of poverty traps, or conditions from which the poorest individuals or groups cannot emerge without outside assistance.31

Some NGOs have responded to this problem, such as BRAC’s Targeting the Ultra-Poor Program and Grameen’s Beggars Program, both introduced in the case study for Chapter 11.

The prospect for ending poverty depends critically on two factors: first, the rate of economic growth—provided it is undertaken in a shared and sustain- able way—and second, the level of resources devoted to poverty programs and the quality of those programs.

Growth and Poverty

Are the reduction of poverty and the acceleration of growth in conflict? Or are they complementary? Traditionally, a body of opinion held that rapid growth is bad for the poor because they would be bypassed and marginal- ized by the structural changes of modern growth. Beyond this, there had been considerable concern in policy circles that the public expenditures required for the reduction of poverty would entail a reduction in the rate of growth. The concerns that concentrated efforts to lower poverty would slow the rate of growth paralleled the arguments that countries with lower inequality would experience slower growth. In particular, if there were redistribution of income or assets from rich to poor, even through progressive taxation, the concern was expressed that savings would fall. However, while the middle class gen- erally has the highest savings rates, the marginal savings rates of the poor, when viewed from a holistic perspective, are not small. In addition to finan- cial savings, the poor tend to spend additional income on improved nutrition, education for their children, improvements in housing conditions, and other expenditures that, especially at poverty levels, represent investments rather than consumption. There are at least five reasons why policies focused toward reducing poverty levels need not lead to a slower rate of growth—and indeed could help to accelerate growth.

First, widespread poverty creates conditions in which the poor have no access to credit, are unable to finance their children’s education, and, in the absence of physical or monetary investment opportunities, have many children as a source of old-age financial security. Moreover, lack of credit denies people liv- ing in poverty of opportunities for entrepreneurship that could otherwise help

249CHAPTER 5 Poverty, Inequality, and Development

to spur growth. Together these factors cause per capita growth to be less than what it would be if there were less poverty.

Second, a wealth of empirical data bears witness to the fact that unlike the historical experience of the now developed countries, the rich in many contem- porary poor countries are generally not noted for their frugality or for their desire to save and invest substantial proportions of their incomes in the local economy.

Third, the low incomes and low levels of living for the poor, which are mani- fested in poor health, nutrition, and education, can lower their economic productivity and thereby lead directly and indirectly to a slower-growing economy. Strategies to raise the incomes and levels of living of the poor will therefore contribute not only to their material well-being but also to the productivity and income of the economy as a whole.32 (These issues are considered further in Chapter 8.)

Fourth, raising the income levels of the poor will stimulate an overall increase in the demand for locally produced necessity products like food and clothing, whereas the rich tend to spend more of their additional incomes on imported luxury goods. Rising demand for local goods provides a greater stimulus to local production, lo- cal employment, and local investment. Such demand thus creates the conditions for rapid economic growth and a broader popular participation in that growth.33

Fifth, a reduction of mass poverty can stimulate healthy economic expansion by acting as a powerful material and psychological incentive to widespread public par- ticipation in the development process. By contrast, wide income disparities and substantial absolute poverty can act as powerful material and psychological disincentives to economic progress. They may even create the conditions for an ultimate rejection of progress by the masses, impatient at the pace of pro- gress or its failure to alter their material circumstances.34 We can conclude, therefore, that promoting rapid economic growth and reducing poverty are not mutually conflicting objectives.35

That dramatic reductions in poverty need not be incompatible with high growth is seen both in case studies and in the cross-national comparisons of data. Countries where poverty has been reduced the most tend to have had sustained growth; at the same time, growth does not guarantee poverty reduc- tion. Over the past 30 years, China has experienced the highest growth rate in the world and also the most dramatic reductions in poverty. The headcount of the poor in China fell from 634 million in 1981 to 128 million in 2004, with the corresponding headcount ratio falling from 64% to 10%. This did not occur merely as a result of high growth. Policies actively encouraged modern-sector enlargement. Moreover, China has worked with the World Bank and other de- velopment agencies to improve its poverty reduction programs and has built on its long-standing efforts to provide at least minimal education and health care for its people as a firm foundation for long-term progress. Although the plight of many peasants has worsened in recent years, especially in interior regions, and inequality has greatly increased, the positive overall results of China’s efforts to fight extreme poverty are apparent. Recent dramatic reduc- tions of poverty in Vietnam have followed a similar pattern.

Richer countries strongly tend to have low levels of absolute poverty. Through one means or another—the availability of employment and entre- preneurship opportunities and greater public and NGO assistance—people who live in rich countries tend to escape from poverty. Among developing countries, there is evidence that countries with faster overall rates of per

250 PART Two Problems and Policies: Domestic

capita income growth also tend on average to have faster rates of per capita income growth among those in the bottom quintile of the income distribution, though the proportions vary widely. While we cannot passively count on even sustainable growth by itself to end absolute poverty, ending poverty can be greatly facilitated through wise and shared stewardship of the various resources provided by growth.36

Certainly, the relationship between economic growth and progress among the poor does not by itself indicate causality. Some of the effect probably runs from improved incomes, education, and health among the poor to faster over- all growth (as suggested by some of the arguments listed previously). Moreo- ver, as we have noted, poverty reduction is possible without rapid growth. But whatever the causality, it is clear that growth and poverty reduction are entirely compatible objectives.

5.5 economic Characteristics of High-Poverty groups

So far we have painted a broad picture of the income distribution and poverty problem in developing countries. We have argued that the magnitude of abso- lute poverty results from a combination of low per capita incomes and highly unequal distributions of that income. Clearly, for any given distribution of in- come, the higher the level of per capita income is, the lower the numbers of the absolutely poor. But higher levels of per capita income are no guarantee of lower levels of poverty. An understanding of the nature of the size distribu- tion of income is therefore central to any analysis of the poverty problem in low-income countries.

But painting a broad picture of absolute poverty is not enough. Before we can formulate effective policies and programs to attack poverty at its source, we need some specific knowledge of these high-poverty groups and their economic characteristics.37

Rural Poverty

Perhaps the most valid generalizations about the poor are that they are dis- proportionately located in rural areas, that they are primarily engaged in ag- ricultural and associated activities, that they are more likely to be women and children than adult males, and that they are often concentrated among minor- ity ethnic groups and indigenous peoples. Data from a broad cross section of developing nations support these generalizations. We find, for example, that about two-thirds of the very poor scratch out their livelihood from subsist- ence agriculture either as small farmers or as low-paid farmworkers. Some of the remaining one-third are also located in rural areas but engaged in petty services, and others are located on the fringes and in marginal areas of urban centers, where they engage in various forms of self-employment such as street hawking, trading, petty services, and small-scale commerce. On the aver- age, we may conclude that in Africa and Asia, about 80% of all target poverty groups are located in the rural areas, as are about 50% in Latin America. Some data for specific countries are provided in Table 5.7.

251CHAPTER 5 Poverty, Inequality, and Development

It is interesting to note, in light of the rural concentration of absolute pov- erty, that the majority of government expenditures in most developing coun- tries over the past several decades has been directed toward the urban area and especially toward the relatively affluent modern manufacturing and com- mercial sectors. Whether in the realm of directly productive economic invest- ments or in the fields of education, health, housing, and other social services, this urban modern-sector bias in government expenditures is at the core of many of the development problems that will be discussed in succeeding chap- ters. We need only point out here that in view of the disproportionate number of the very poor who reside in rural areas, any policy designed to alleviate poverty must necessarily be directed to a large extent toward rural develop- ment in general and the agricultural sector in particular (we will discuss this matter in detail in Chapter 9).

Women and Poverty

Women make up a substantial majority of the world’s poor. If we compared the lives of the inhabitants of the poorest communities throughout the devel- oping world, we would discover that virtually everywhere women and chil- dren experience the harshest deprivation. They are more likely to be poor and malnourished and less likely to receive medical services, clean water, sanita- tion, and other benefits.38 The prevalence of female-headed households, the

Table 5.7 Poverty: rural versus urban

Source: data from World Bank, World Development Indicators, 2010 (Washington, D.C.: World Bank, 2010), tab. 2.7.

   Percentage below National Poverty line

region and Country

Survey Year

rural Population

urban Population

National Population

Sub-Saharan africa Benin Burkina Faso Cameroon Malawi Tanzania Uganda Zambia asia Bangladesh India Indonesia Uzbekistan Vietnam latin america Bolivia Brazil Dominican Republic Guatemala Honduras Mexico Peru

  2003 2003 2007 2005 2001 2006 2004

  2005 2000 2004 2003 2002

  2007 2003 2007 2006 2004 2004 2004

  46.0 52.4 55.0 55.9 38.7 34.2 72.0

  43.8 30.2 20.1 29.8 35.6

  63.9 41.0 54.1 72.0 70.4 56.9 72.5

  29.0 19.2 12.2 25.4 29.5 13.7 53.0

  28.4 24.7 12.1 22.6 6.6

  23.7 17.5 45.4 28.0 29.5 41.0 40.3

  39.0 46.4 29.9 52.4 35.7 31.1 68.0

  40.0 28.6 16.7 27.2 28.9

  37.7 21.5 48.5 51.0 50.7 47.0 51.6

252 PART Two Problems and Policies: Domestic

lower earning capacity of women, and their limited control over their spouses’ income all contribute to this disturbing phenomenon. In addition, women have less access to education, formal-sector employment, social security, and government employment programs. These facts combine to ensure that poor women’s financial resources are meager and unstable relative to men’s.

A disproportionate number of the ultrapoor live in households headed by women, in which there are generally no male wage earners. Because the earning potential of women is considerably below that of their male counter- parts, women are more likely to be among the very poor. In general, women in female-headed households have less education and lower incomes. Further- more, the larger the household is, the greater the strain on the single parent and the lower the per capita food expenditure.

A portion of the income disparity between male- and female-headed households can be explained by the large earnings differentials between men and women. In addition to the fact that women are often paid less for performing similar tasks, in many cases they are essentially barred from higher-paying occupations. In urban areas, women are much less likely to obtain formal employment in private companies or public agencies and are frequently restricted to illegal, low-productivity jobs. The illegality of piece- work, as in the garment industry, prevents it from being regulated and ren- ders it exempt from minimum-wage laws or social security benefits. Even when women receive conventional wage payments in factory work, minimum wage and safety legislation may be flagrantly ignored. Similarly, rural women have less access to the resources necessary to generate stable incomes and are frequently subject to laws that further compromise earning potential. Legisla- tion and social custom often prohibit women from owning property or signing financial contracts without a husband’s signature. With a few notable excep- tions, government employment or income-enhancing programs are accessible primarily if not exclusively by men, exacerbating existing income disparities between men and women.

But household income alone fails to describe the severity of women’s rela- tive deprivation. Because a higher proportion of female-headed households are situated in the poorest areas, which have little or no access to government-spon- sored services such as piped water, sanitation, and health care, household mem- bers are more likely to fall ill and are less likely to receive medical attention. In addition, children in female-headed households are less likely to be enrolled in school and more likely to be working in order to provide additional income.

The degree of economic hardship may also vary widely within a household. We have already discussed the fact that GNI per capita is an inadequate measure of development because it fails to reflect the extent of absolute poverty. Likewise, household income is a poor measure of individual welfare because the distribu- tion of income within the household may be quite unequal. In fact, among the poor, the economic status of women provides a better indication of their own welfare, as well as that of their children. Existing studies of intrahousehold re- source allocation clearly indicate that in many regions of the world, there exists a strong bias against females in areas such as nutrition, medical care, education, and inheritance. Moreover, empirical research has shown that these gender bi- ases in household resource allocation significantly reduce the rate of survival among female infants. This is one reason why recorded female-male sex ratios

253CHAPTER 5 Poverty, Inequality, and Development

are so much below their expected values, primarily in Asian countries, that well over 100 million girls and women are said to be “missing.”39 The favor shown to- ward boys in part reflects the fact that men are perceived to have a greater poten- tial for contributing financially to family survival. This is not only because well- paying employment for women is unavailable but also because daughters are often married to families outside the village, after which they become exclusively responsible to their in-laws and thus cease contributing to their family of origin.

The extent of these internal biases is strongly influenced by the economic status of women. Studies have found that where women’s share of income within the home is relatively high, there is less discrimination against girls, and women are better able to meet their own needs as well as those of their children. When household income is marginal, most of women’s income is contributed toward household nutritional intake. Since this fraction is considerably smaller for men, a rise in male earnings leads to a less than proportionate increase in the funds available for the provision of daily needs. It is thus unsurprising that programs designed to increase nutrition and family health are more effective when targeting women than when targeting men. In fact, significant increases in total household income do not necessarily translate into improved nutritional status (see Chapter 8). The persistence of low levels of living among women and children is common where the economic status of women remains low. Box 5.2 provides some views of the poor on gender relations.

Women’s control over household income and resources is limited for a number of reasons. Of primary importance is the fact that a relatively large proportion of the work performed by women is unremunerated—for example, collecting firewood and cooking—and may even be intangible, as with parent- ing. Women’s control over household resources may also be constrained by the fact that many women from poor households are not paid for the work they perform in family agriculture or business. It is common for the male head of household to control all funds from cash crops or the family business, even though a significant portion of the labor input is provided by his spouse. In addition, in many cultures, it is considered socially unacceptable for women to contribute significantly to household income, and hence women’s work may remain concealed or unrecognized. These combined factors perpetuate the low economic status of women and can lead to strict limitations on their control over household resources.

Development policies that increase the productivity differentials between men and women are likely to worsen earnings disparities as well as further erode women’s economic status within the household. Since government pro- grams to alleviate poverty frequently work almost exclusively with men, they tend to exacerbate these inequalities. In urban areas, training programs to in- crease earning potential and formal-sector employment are generally geared to men, while agricultural extension programs promote male-dominated crops, frequently at the expense of women’s vegetable plots (see Chapter 9). Studies have shown that development efforts can actually increase women’s workload while at the same time reduce the share of household resources over which they exercise control. Consequently, women and their dependents re- main the most economically vulnerable group in developing countries.

The fact that the welfare of women and children is strongly influenced by the design of development policy underscores the importance of integrating

254 PART Two Problems and Policies: Domestic

women into development programs. To improve living conditions for the poorest individuals, women must be drawn into the economic mainstream. This would entail increasing female participation rates in educational and training programs, formal-sector employment, and agricultural extension pro- grams. It is also of primary importance that precautions be taken to ensure that women have equal access to government resources provided through schooling, services, employment, and social security programs. Legalizing informal-sector employment where the majority of the female labor force is employed would also improve the economic status of women.

The consequences of declines in women’s relative or absolute economic status have both ethical and long-term economic implications. Any process of growth that fails to improve the welfare of the people experiencing the greatest hardship, broadly recognized to be women and children, has failed to accom- plish one of the principal goals of development. In the long run, the low status of women is likely to translate into slower rates of economic growth. This is true because the educational attainment and future financial status of children are much more likely to reflect those of the mother than those of the father. Thus, the benefits of current investments in human capital are more likely to be passed on to future generations if women are successfully integrated into the growth process. And considering that human capital is perhaps the most important prerequisite for growth, education and enhanced economic status for women are critical to meeting long-term development objectives. (We ex- amine these issues in greater detail in Chapter 8.)

As feminist development economists have often expressed it, official pov- erty programs cannot simply “add women and stir.” Women-centered poverty strategies often require us to challenge basic assumptions. The harsher con- ditions for women and women’s crucial role in a community’s escape from poverty mean that involvement of women cannot be left as an afterthought

bOX 5.2 Problems of gender relations in Developing Countries: Voices of the Poor

Sister, if you don’t beat them, they’ll stop being good. And if they’re good and you beat them, they’ll stay that way.

—a man in bangladesh

When my husband died, my in-laws told me to get out. So I came to town and slept on the pavement.

—a middle-aged widow in Kenya

When I was working, I used to decide. When she is working, she owns her money and does any- thing she wishes.

—a man from Vila Junqueira, brazil

Problems have affected our relationship. The day my husband brings in money, we are all right

together. The day he stays at home [out of work], we are fighting constantly.

—a woman from el gawaber, egypt

The unemployed men are frustrated because they can no longer play the part of family providers and protectors. They live on the money made by their wives and feel humiliated because of this.

—an elderly woman from uchkun, Kyrgyzstan

When a woman gives her opinion, they [men] make fun of her and don’t pay attention. If women go to a meeting, they don’t give their opinion.

—a woman in las Pascuas, bolivia

255CHAPTER 5 Poverty, Inequality, and Development

but will be most effective if it is the first thought—and the consistent basis for action—when addressing poverty.

Ethnic Minorities, Indigenous Populations, and Poverty

A final generalization about the incidence of poverty in the developing world is that it falls especially heavily on minority ethnic groups and indigenous popula- tions. We pointed out in Chapter 2 that some 40% of the world’s nation-states have more than five sizable ethnic populations, one or more of which faces seri- ous economic, political, and social discrimination. In recent years, domestic con- flicts and even civil wars have arisen out of ethnic groups’ perceptions that they are losing out in the competition for limited resources and job opportunities. The poverty problem is even more serious for indigenous peoples, whose numbers exceed 300 million in over 5,000 different groups in more than 70 countries.40

Although detailed data on the relative poverty of minority ethnic and in- digenous peoples are difficult to obtain (for political reasons, few countries wish to highlight these problems), researchers have compiled data on the poverty of indigenous people in Latin America.41 The results clearly demon- strate that a majority of indigenous groups live in extreme poverty and that being indigenous greatly increases the chances that an individual will be malnourished, illiterate, in poor health, and unemployed. For example, the research has shown that in Mexico, over 80% of the indigenous population is poor, compared to 18% of the nonindigenous population. Table 5.8 shows that similar situations exist in countries such as Bolivia, Guatemala, and Peru (not to mention Native American populations in the United States and Canada). Moreover, a 2006 World Bank study confirmed that all too little progress had been made. Whether we speak of Tamils in Sri Lanka, Karens in Myanmar, Untouchables in India, or Tibetans in China, the poverty plight of minorities is as serious as that of indigenous peoples.

Poor Countries Finally, it should be noted that the poor come from poor coun- tries. Although this may seem like a trivial observation, it is actually a useful note of optimism. The negative relationship between poverty and per capita income sug- gests that if higher incomes can be achieved, poverty will be reduced, if only be- cause of the greater resources that countries will have available to tackle poverty problems and the growth of civil society and the voluntary sector. Unfortunately,

Table 5.8 Indigenous Poverty in latin america

Sources: Data for the left side of the table from George Psacharopoulos and Harry A. Patrinos, “Indigenous people and poverty in Latin America,” Finance and Development 31 (1994): 41, used with permission; data for the right side of the table from Gillette Hall and Harry A. Patrinos, eds., Indigenous Peoples, Poverty, and Human Development in Latin America, 1994–2004 (New York: Palgrave Macmillan, 2006).

Population below the Poverty line (%), early 1990s Change in Poverty (%), Various Periods

Country Indigenous Nonindigenous Period Indigenous Nonindigenous

Bolivia Guatemala Mexico Peru

64.3 86.6 80.6 79.0

48.1 53.9 17.9 49.7

1997–2002 1989–2000 1992–2002 1994–2000

0 −15

0 0

−8 −25 −5 +3

256 PART Two Problems and Policies: Domestic

as noted earlier, a high level of absolute poverty can also retard a country’s growth prospects. Moreover, many of the poorest countries in sub-Saharan Africa experi- enced outright declines in per capita income throughout the 1980s and 1990s and in some cases during the first decade of this century. Among those that are growing, at current growth rates it would take decades to reach the levels of income at which poverty tends to be eradicated. After all, Brazil, which has been solidly middle-in- come for decades, still has 8% of its population living on less than $1.25 per day. In- come poverty, malnutrition, low school attendance, and child labor in Brazil finally showed a substantial decline after the turn of this century, when antipoverty and social safety net programs were greatly expanded (see the case study at the end of Chapter 1). We can conclude that higher national incomes greatly facilitate poverty reduction, while at the same time, poverty still needs to be addressed directly.

5.6 Policy Options on Income Inequality and Poverty: Some basic Considerations

Areas of Intervention

Developing countries that aim to reduce poverty and excessive inequalities in their distribution of income need to know how best to achieve their aim. What kinds of economic and other policies might governments in developing countries adopt to reduce poverty and inequality while maintaining or even accelerating economic growth rates? As we are concerned here with moderat- ing the size distribution of incomes in general and raising the income levels of people living in poverty, it is important to understand the various deter- minants of the distribution of income in an economy and see in what ways government intervention can alter or modify their effect. The main focus of this section is on the relationship between income inequality and poverty. We examine the effects of policies and programs involving nonincome aspects of poverty in the subsequent chapters in Part Two—particularly with respect to health, nutrition, and education in Chapter 8.

We can identify four broad areas of possible government policy interven- tion, which correspond to the following four major elements in the determina- tion of a developing economy’s distribution of income.

1. Altering the functional distribution—the returns to labor, land, and capital as determined by factor prices, utilization levels, and the consequent shares of national income that accrue to the owners of each factor.

2. Mitigating the size distribution—the functional income distribution of an economy translated into a size distribution by knowledge of how owner- ship and control over productive assets and labor skills are concentrated and distributed throughout the population. The distribution of these asset holdings and skill endowments ultimately determines the distribution of personal income.

3. Moderating (reducing) the size distribution at the upper levels through progressive taxation of personal income and wealth. Such taxation increases government

257CHAPTER 5 Poverty, Inequality, and Development

revenues that decrease the share of disposable income of the very rich—rev- enues that can, with good policies, be invested in human capital and rural and other lagging infrastructure needs, thereby promoting inclusive growth. (An individual or family’s disposable income is the actual amount available for expenditure on goods and services and for saving.)

4. Moderating (increasing) the size distribution at the lower levels through public expenditures of tax revenues to raise the incomes of the poor either di- rectly (e.g., by conditional or unconditional cash transfers) or indirectly (e.g., through public employment creation such as local infrastructure projects or the provision of primary education and health care). Such pub- lic policies raise the real income levels of the poor above what their per- sonal income levels would otherwise be, and, as will become clear in later chapters, can do so sustainably when they build the capabilities and assets of people living in poverty.

Altering the Functional Distribution of Income through Relative Factor Prices

Altering the functional distribution is a traditional economic approach. It is ar- gued that as a result of institutional constraints and faulty government policies, the relative price of labor in the formal, modern, urban sector is higher than what would be determined by the free interplay of the forces of supply and de- mand. For example, the power of trade unions to raise minimum wages to artifi- cially high levels (higher than those that would result from supply and demand) even in the face of widespread unemployment is often cited as an example of the “distorted” price of labor. From this it is argued that measures designed to reduce the price of labor relative to capital (e.g., through market-determined wages in the public sector or public wage subsidies to employers) will cause employers to substitute labor for capital in their production activities. Such fac- tor substitution increases the overall level of employment and ultimately raises the incomes of the poor, who have been excluded from modern-sector employ- ment and typically possess only their labor services. Put differently, artificially increased modern-sector wages reduce the rate of modern-sector enlargement growth, thus harming the poor. (For details of this analysis, see Appendix 5.1.)

However, in recent years, some scholars and practitioners, particularly from the developing world, argue that the impact of minimum wages on poverty is more nuanced in theory and practice, particularly when the pos- sibility of income sharing among the poor is accounted for. In India, the Self- Employed Women’s Association argues that minimum wages have beneficial effects even on informal-sector workers. And research by Darryl McLeod and Nora Lustig concludes that higher minimum wages are correlated with reduc- tions in poverty.42 Thus, actual impacts may vary, depending on local circum- stances. These qualifications are particularly relevant for relatively low-skill and informal activities, such as garment stitching, beedi rolling, and incense rolling, in which workers have commonly held very low bargaining power, often due to monopsony, if not extramarket forces.

In addition, often the price of capital equipment is “institutionally” set at artificially low levels (below what supply and demand would dictate)

Disposable income The income that is available to households for spending and saving after personal income taxes have been deducted.

258 PART Two Problems and Policies: Domestic

through various public policies such as investment incentives, tax allowances, subsidized interest rates, overvalued exchange rates, and low tariffs on capital goods imports such as tractors and automated equipment relative to tariffs set on consumer goods. If these special privileges and capital subsidies were removed so that the price of capital would rise to its true “scarcity” level, pro- ducers would have a further incentive to increase their utilization of the abun- dant supply of labor and lower their uses of scarce capital. Moreover, owners of capital (both physical and financial) would not receive the artificially high economic returns they now enjoy.

Because factor prices are assumed to function as the ultimate signals and incentives in any economy, correcting these prices (i.e., lowering the relative price of labor and raising the relative price of capital) would, in general, not only increase productivity and efficiency but also reduce inequality by pro- viding more wage-paying jobs for currently unemployed or underemployed unskilled and semiskilled workers. It would also lower the artificially high incomes of owners of capital. Removal of such factor-price distortions would therefore go a long way toward combining more growth, efficiently generated, with higher employment, less poverty, and greater equality (a more detailed analysis is presented in Appendix 5.1).

We may conclude that there is much merit to the traditional factor-price dis- tortion argument and that correcting prices should contribute to a reduction in poverty and an improved distribution of income. How much it actually con- tributes will depend on the degree to which firms and farms switch to more labor-intensive production methods as the relative price of labor falls and the relative price of capital rises. These are important empirical questions, the an- swers to which will vary from country to country. Moreover, recent research would suggest that a close study of local conditions is needed before conclud- ing that all minimum wages cause increases in poverty in all circumstances.

Modifying the Size Distribution through Increasing Assets of the Poor

Given correct resource prices and utilization levels for each type of produc- tive factor (labor, land, and capital), we can arrive at estimates for the total earnings of each asset. But to translate this functional income into personal in- come, we need to know the distribution and ownership concentration of these assets among and within various segments of the population. Here we come to what is probably the most important fact about the determination of in- come distribution within an economy: The ultimate cause of the unequal dis- tribution of personal incomes in most developing countries is the unequal and highly concentrated patterns of asset ownership (wealth) in these countries. The principal reason why 20% of their population often receives over 50% of the national income (see Table 5.2) is that this 20% probably owns and con- trols well over 90% of the productive and financial resources, especially physi- cal capital and land but also financial capital (stocks and bonds) and human capital in the form of better education and health. Correcting factor prices is certainly not sufficient to reduce income inequalities substantially or to elimi- nate widespread poverty where physical and financial asset ownership—and education—are highly concentrated.

Asset ownership The own- ership of land, physical capital (factories, buildings, machin- ery, etc.), human capital, and financial resources that gener- ate income for owners.

259CHAPTER 5 Poverty, Inequality, and Development

It follows that the second and perhaps more important line of policy to re- duce poverty and inequality is to focus directly on reducing the concentrated control of assets, the unequal distribution of power, and the unequal access to ed- ucational and income-earning opportunities that characterize many developing countries. A classic case of such redistribution policies as they relate to the rural poor, who comprise 70% to 80% of the target poverty group, is land reform. The basic purpose of land reform is to transform tenant cultivators into smallholders who will then have an incentive to raise production and improve their incomes. But as we explain in Chapter 9, land reform may be a weak instrument of income redistribution if other institutional and price distortions in the economic system prevent small farm holders from securing access to much needed critical inputs such as credit, fertilizers, seeds, marketing facilities, and agricultural education. Similar reforms in urban areas could include the provision of commercial credit at affordable rates (rather than through traditional, high-interest moneylenders) to small entrepreneurs (microcredit—for details, see Chapter 15 and the case study on the Grameen Bank at the end of that chapter) so that they can expand their business and provide more jobs to local workers.

In addition to the redistribution of existing productive assets, dynamic re- distribution policies could be gradually pursued. For example, governments at least in developing countries that are growing could facilitate the transfer of a certain proportion of annual savings and investments to low-income groups so as to bring about a more gradual and perhaps politically more acceptable redistribution of additional assets as they accumulate over time. This is what is often meant by the expression “redistribution from growth.” Whether such a gradual redistribution from growth is any more possible than a redistribu- tion of existing assets is a moot point, especially in the context of very unequal power structures. But some form of asset redistribution, whether static or dy- namic, seems to be a necessary condition for any significant reduction of pov- erty and inequality in most developing countries.

Human capital in the form of education and skills is another example of the unequal distribution of productive asset ownership. Public policy should therefore promote wider access to educational opportunities (for girls as well as boys) as a means of increasing income-earning potential for more people. But as in the case of land reform, the mere provision of greater access to addi- tional education is no guarantee that the poor will be better off unless comple- mentary policies—for example, the provision of more productive employment opportunities for the educated—are adopted to capitalize on this increased human capital. The relationship among education, employment, and develop- ment is discussed further in Chapter 8.

People living in poverty tend to have common problems, but the prevalent forms of deprivation and social exclusion can differ considerably even across regions within a country. Policymakers need to have a strong knowledge base. Essential to the process is a means to find out and utilize what the poor know about their own conditions of poverty. Practitioners stress that the more that people living in poverty are engaged in setting the agenda, the more effec- tive programs to increase their assets and capabilities tend to be. But attention must be given to different segments of the local poor communities, as different priorities are often found between men and women, between ethnic groups, and between castes.

Redistribution policies Policies geared to reducing income inequality and expanding economic opportunities in order to promote development, includ- ing income tax policies, rural development policies, and publicly financed services.

Land reform A deliberate attempt to reorganize and transform existing agrarian systems with the intention of improving the distribution of agricultural incomes and thus fostering rural development.

260 PART Two Problems and Policies: Domestic

Progressive Income and Wealth Taxes

Any national policy attempting to improve the living standards of the bot- tom 40% must secure sufficient financial resources to transform paper plans into program realities. The major source of such development finance is the direct and progressive taxation of both income and wealth. Direct progressive income taxes focus on personal and corporate incomes, with the rich required to pay a progressively larger percentage of their total income in taxes than the poor. Taxation on wealth (the stock of accumulated assets and income) typi- cally involves personal and corporate property taxes but may also include progressive inheritance taxes. In either case, the burden of the tax is designed to fall most heavily on the upper-income groups.

In reality, in many developing countries (and some developed countries), the gap between what is supposed to be a progressive tax structure and what different income groups actually pay can be substantial. Progressive tax struc- tures on paper often turn out to be regressive taxes in practice, in that the lower- and middle-income groups often end up paying a proportionally larger share of their incomes in taxes than the upper-income groups. The reasons for this are simple. The poor are often taxed at the source of their incomes or expenditures (by withholding taxes from wages, general poll taxes, or indirect taxes levied on the retail purchase of goods such as cigarettes and beer). By contrast, the rich derive by far the largest part of their incomes from the return on physical and financial assets, which often go unreported. They often also have the power and ability to avoid paying taxes without fear of government reprisal. Policies to enforce progressive rates of direct taxation on income and wealth, especially at the highest levels, are what are most needed in this area of redistribution activ- ity. (See Chapter 15 for a further discussion of taxation for development.)

Direct Transfer Payments and the Public Provision of Goods and Services

The direct provision of tax-financed public consumption goods and services to the very poor is another potentially important instrument of a comprehen- sive policy designed to eradicate poverty. Examples include public health pro- jects in rural villages and urban fringe areas, school lunches and preschool nutritional supplementation programs, and the provision of clean water and electrification to remote rural areas. Direct money transfers and subsidized food programs for the urban and rural poor, as well as direct government poli- cies to keep the prices of essential foodstuffs low, represent additional forms of public consumption subsidies.

Direct transfers and subsidies can be highly effective, but they need to be de- signed carefully. Four significant problems require attention. First, when resources for attacking poverty are limited—as they always are—they need to be directed to people who are genuinely poor. Second, it is important that beneficiaries not become unduly dependent on the poverty program; in particular, we do not want to give the poor less incentive to build the assets, such as education, that can enable them to stay out of poverty. But a “safety net” can also be valuable to encourage the poor to accept a more entrepreneurial attitude toward their mi- croenterprises. This is much more possible when the poor do not fear that their

Progressive income tax A tax whose rate increases with increasing personal incomes.

Regressive tax A tax struc- ture in which the ratio of taxes to income tends to decrease as income increases.

Indirect taxes Taxes levied on goods ultimately pur- chased by consumers, includ- ing customs duties (tariffs), excise duties, sales taxes, and export duties.

Public consumption All current expenditures for pur- chases of goods and services by all levels of government, including capital expenditures on national defense and security.

Subsidy A payment by the government to producers or distributors in an industry to prevent the decline of that industry, to reduce the prices of its products, or to encour- age hiring.

261CHAPTER 5 Poverty, Inequality, and Development

children will suffer terrible consequences if their small businesses fail. Third, we do not want to divert people who are productively engaged in alternative eco- nomic activities to participate in the poverty program instead. Finally, poverty policies are often limited by resentment from the nonpoor, including those who are working hard but are not very far above the poverty line themselves.

When a subsidy of goods consumed by the poor is planned, it should be targeted to the geographic areas where the poor are found and should em- phasize goods that nonpoor people do not consume. This helps conserve re- sources for the program and minimizes efforts by nonpoor people to benefit from the program. For example, nutritional supplements can be provided for any woman who brings her baby to the neighborhood poverty program center located in villages and neighborhoods with a high incidence of absolute pov- erty. Although more affluent mothers could use the program, few would risk the stigma of venturing into the poorer villages and neighborhoods, let alone the center itself. The nutritional supplements help poor mothers and their small children stay healthy and thus help break the cycle of poverty.

In addition, it may be useful to impose a work requirement before food aid is provided. This is done in the well-known Bangladesh Food for Work Program and in the Maharashtra Employment Guarantee Scheme in India. More recently, the government of India has introduced a nationwide pro- gram to guarantee 100 days of employment to at least one family member each year; early reports suggest that the program has provided substan- tial benefits. In programs such as these, the poor are put to work building infrastructure, such as roads from outlying areas (where the poor live) to market towns, that will ultimately benefit the poor and others in the region. Although the administrative costs are generally higher and the skills of the workers significantly lower than would be the case with a commercially procured construction contract, in many cases these valuable infrastructure projects would never be tackled at all in the absence of the program. The high work requirement and very modest payment discourage the nonpoor from participating, thus conserving resources. This characteristic is known as the “screening” function of workfare programs. These requirements also help preserve the program’s political sustainability: When people see that the poor are getting “a hand up rather than a handout,” the programs tend to attract wider public support.

In sum, we can say that workfare, such as the Food for Work Program, represents a better policy than welfare or direct handouts when the following criteria are met:

• The program does not reduce or seriously undermine incentives for the poor to acquire human capital and other assets.

• There are greater net benefits of the work output of the program.

• It is harder to screen the poor without the workfare requirement.

• There is lower opportunity cost of time for poor workers (so the economy loses little output when they join the workfare program).

• There is higher opportunity cost of time for nonpoor workers (so they won’t avail themselves of the benefits).

workfare program A pov- erty alleviation program that requires program beneficiaries to work in exchange for ben- efits, as in a food-for-work program.

262 PART Two Problems and Policies: Domestic

• The fraction of the population living in poverty is smaller (so the extra costs of a universal welfare program would be high).

• There is less social stigma attached to participating in a workfare pro- gram, so the poor do not suffer undue humiliation and are less deterred from seeking the help that their families need (otherwise, a discreet wel- fare transfer may be preferable to a highly visible workfare program).43

The poor often have low bargaining power in their communities, and while it is difficult politically to increase this power, well-designed programs can accomplish this indirectly by providing improved “outside options” such as guaranteed public employment programs when they are needed.

We will be continuing our examination of policies for poverty reduction throughout the remainder of this text. Appropriate agricultural development policies represent a crucial strategy for attacking poverty because such a high fraction of the poor are located in rural areas and engaged in agricultural pur- suits. Strategies for agricultural development are examined in Chapter 9. In addition, the poor in urban as well as rural areas suffer from degraded envi- ronmental conditions, which lower opportunities for economic growth and also worsen the health of the poor; these problems are examined in Chapter 10.

Another set of viable policies involve targeted poverty programs to increase the capabilities and human and social capital of the poor. An important ex- ample centers on helping the poor develop their microenterprises, on which a large fraction of the nonagricultural poor depend for their survival. It has been found that credit is the binding constraint for many of these tiny firms. By building up the working capital and other assets of microenterprises, the poor can improve their productivity and incomes. The microfinance strategy for ac- complishing this goal, as exemplified by the Grameen Bank of Bangladesh, is examined in Chapter 15. In addition, relatively new approaches to attacking poverty focus on an integrated approach to achieving higher incomes together with improved education, health, and nutrition among the poor, notably, con- ditional cash transfer (CCT) programs that transfer incomes to poor families conditional on behaviors such as keeping their children in school; these ap- proaches are considered in Chapter 8 and its case study. Finally, strategies to assist the development of the urban informal sector are examined in Chapter 7.

5.7 Summary and Conclusions: The Need for a Package of Policies

To summarize our discussion of alternative policy approaches to the problems of poverty and inequality in development, the need is not for one or two iso- lated policies but for a “package” of complementary and supportive policies, including the following four basic elements.44

1. A policy or set of policies designed to correct factor price distortions (under- pricing capital or overpricing modern-sector skilled wages) so as to ensure that market or institutionally established prices provide accurate signals and incentives to both producers and resource suppliers. Correcting distorted prices should contribute to greater productive efficiency, more employment,

263CHAPTER 5 Poverty, Inequality, and Development

and less poverty. The promotion of indigenous technological research and development of efficient, labor-intensive methods of production may also be valuable. (For a further analysis of factor price distortions, see Appendix 5.1.)

2. A policy or set of policies designed to bring about far-reaching structural changes in the distribution of assets, power, and access to education and asso- ciated income-earning (employment) opportunities. Such policies go beyond the realm of markets and touch on the whole social, institutional, cultural, and political fabric of the developing world. But such fundamental structural changes and substantive asset redistributions, whether immediately achieved (e.g., through public-sector interventions) or gradually introduced over time (through redistribution from growth), will increase the chances of improving significantly the living conditions of the masses of rural and urban poor.

3. A policy or set of policies designed to modify the size distribution of in- come at the upper levels through the enforcement of legislated progres- sive taxation on incomes and wealth; and at the same time, providing the poor with direct transfer payments and the expanded provision of publicly provided consumption goods and services, including workfare programs. The net effect is to create a social “safety net” for people who may be bypassed by the development process.

4. A set of targeted policies to directly improve the well-being of the poor and their communities, which goes beyond safety net schemes, to offer programs that build capabilities and human and social capital of the poor, such as microfinance, health, education, agricultural development, envi- ronmental sustainability, and community development and empower- ment programs, as described throughout this text. These can be carried out either by government or by nongovernmental organizations through local and international support.

While providing a focus on ending extreme poverty and mitigating harmful inequality, such policies can be designed to encourage and accelerate inclusive economic growth targeted at the poor, while keeping in mind the inherently multidimensional nature of poverty. Key examples include growth-supporting investments in education, nutrition, health, and infrastructure that raise the incomes of those in the bottom deciles of the income distribution. Chapters 2 through 4 considered the sources of economic growth and basic policies to iden- tify constraints and maintain growth that benefit people living in poverty. Ad- ditional supporting trade, macro, and financial policies are examined in more detail in Chapters 13 through 15. But when it is not inclusive, growth by itself is insufficient to eliminate extreme poverty, at least in any time frame that a nation—let alone people living in poverty—will find acceptable. So encourage- ment of inclusive growth goes hand in hand with active policies and programs to reduce poverty and to prevent nonpoor people from falling into poverty.

Though the task of ending extreme poverty will be difficult, it is possible, if we can only muster the will. As noted by James Speth, the executive director of the United Nations Development Programme, “Poverty is no longer inevi- table. The world has the material and natural resources, the know-how and the people to make a poverty-free world a reality in less than a generation. This is not woolly idealism but a practical and achieveable goal.”45

264264

ghana’s development has exceeded expecta-tions—at least after many disappointments. Côte d’Ivoire (CIV) started with many apparent ad- vantages, but on many economic measures, Ghana has closed the development gap that existed be- tween itself and CIV at independence.

It is recommended that you read Chapters 2 and 5 in conjunction with this case. These country illustrations provide further interpretation of the more general research discussed in those chapters.

A Natural Comparative Case Study Ghana and Côte d’Ivoire border each other in West Africa. Their land area is similar in size at 92,456 square miles (239,450 km2) and 124,502 square miles (322,458 km2), respectively. Their populations are also similar, with 25.5 million people in Ghana and 20.6 million in Côte d’Ivoire in 2012. Becoming indepen- dent within three years of each other and also sharing similar geographies, these adjoining countries make for a natural comparison. One of the most striking dif- ferences is that Ghana was part of the British Empire from 1821 to 1957, and CIV was a French colony from 1842 until 1960. (Note, however, that full colonial rule took a long time to become established throughout the territories of these countries; the French were still fighting to extend their presence into the early years of the twentieth century.)

How did these colonial histories matter? Did their influences extend after independence, affect- ing later development policies for good or ill? Or have other, internal factors been more decisive? Can this help us to better understand why it is so challenging to sustain high growth, eliminate pov- erty and hunger, and to achieve other Millennium Development Goals?

The experiences of a half-century after inde- pendence illustrate some of the opportunities for and threats to development. This case study raises thought-provoking questions and presents the types of information one would weigh in address- ing this and other comparative country studies. This case illustrates how the frameworks and many- country statistical studies of Chapters 2 and 5 can be applied to understanding development experi- ences in comparative perspective. The richness of culture and nuances of complex political histories are abstracted to feature some broad approaches and findings in development economics in a short space. Readers are encouraged to explore these leading African nations in detail.

Poverty and Human Development As reported in the UNDP’s 2013 Human Development Report, Ghana is considered a medium human development country, with a New Human Development Index (NHDI) value of 0.558, while CIV is considered a low human development country, with an NHDI of just 0.432. Ghana’s performance is 22 positions higher than predicted by income, while CIV’s is 9 positions lower. In the 1990 Human Development Report, when the original HDI was introduced, the numbers were 0.393 for CIV and 0.360 for Ghana. Both made prog- ress, but Ghana much more so. CIV’s Multidimen- sional Poverty Index (MPI) as reported in the 2013 Human Development Report is very high at 0.353, while Ghana’s MPI is substantially lower at 0.144. And the 2009 Human Development Report Human Poverty In- dex (see note 11) for CIV was 0.374, ranking 29 places lower in the country rankings based on human pov- erty than income poverty (the fraction under $1.25 per day). This suggested that what the UNDP termed human poverty is relatively worse in CIV than even

Case Study 5

Institutions, Inequality, and Incomes: ghana and Côte d’Ivoire

265

its income poverty would suggest. Ghana’s HPI was significantly better, at 0.281 (with its ranking as pre- dicted by its income poverty).

These outcomes would have surprised many who wrote at the time of independence. In 1960, Ghana had a real GDP per capita of just $594, far behind CIV’s $1,675; but in 2007, according to the Penn World Table, Ghana had reached $1,653—a gain of 278% and nearly enough to close its original deficit—while CIV increased to $2,228, a modest gain of just 33% after 47 years. In 2011, Ghana’s estimated income per capita PPP of $1,830 surpassed CIV’s level of $1,780 (2013 World Development Indicators).

Ghana has a life expectancy of 64, while that of CIV is only 55 (2012 PRB estimates); in 1960, life ex- pectancy in CIV was 51, to Ghana’s 46, a dramatic reversal. In 2011, under-5 mortality was 115 in CIV and a still high but significantly lower 78 in Ghana.

Aysit Tansel showed that by 1987, Ghana was well ahead of CIV in mean years of schooling by each gen- der and across age groups. By 2008, the adult literacy rate reached 65.0% in Ghana versus 48.7% in CIV.

Highly credible information on the extent of ex- treme poverty in these countries is difficult to find, but it is not doubted that at the time of independ- ence, poverty was far higher in Ghana. Using 1987 surveys, the World Bank put dollar-a-day poverty at just 3.28% in CIV that year but 46.51% in Ghana; a comparable figure for Ghana (from a 1998 study) was 36% and for CIV (2002) was 16%. The most recent available World Bank estimates are 28.6% below $1.25 per day in Ghana (2006 survey) and 23.8% (2008 survey) in CIV (2013 World Develop- ment Indicators). It appears that over time, poverty has fallen in Ghana and risen in CIV.

Progress in both countries is small in compari- son to East Asia; but the differences between these countries are substantial. How can we begin to un- derstand such differences? Sometimes even recent changes in the patterns of development can have long historical roots, and we consider this first.

Long-Run Factors in Comparative Development The Colonial Impact and the Legacy of Institu- tions The Portuguese built a fortress on the coast of Ghana in 1482 and named it Elmina (“The Mine”). Later, the British named this area the Gold Coast, as it was known until independence in 1957.

*According to the AJR dataset, which is based on the work of the historian Philip Curtin, the other highest-mortality colonies were Togo, Gambia, Mali, and Nigeria. By contrast, the death rate was just 14.9 in Hong Kong, and 17.7 in Malaysia and Singapore. Settler mortality was used as an in- strument for early institutions in the literature (see Chapter 2), and we examine two countries with identical settler mortalities giving attention to additional elements. (Conclusions of the research cited here are based on multicountry statistical analysis, not on case studies; we are taking such research as a starting point for issues to consider when conducting more in-depth comparative case studies).

Côte d’Ivoire (Ivory Coast) received its name from the French. These names apparently reflect how the colonial powers viewed the territories: as “coasts” rather than nations; as commodities for trade rather than people, or simply as a mine. The colo- nialists’ priority of resources over people could not have been more obvious. Ghana suffered earlier and more from the impact of the slave trade. But CIV also suffered ill treatment, including a brutal campaign by the French to subdue the “interior” in the late nineteenth and early twentieth centuries and impose forced labor. How do we understand this terrible colonial experience and its possible af- termath? Settler mortality rates, which is correlated with the establishment of extractive institutions by the colonial power with long-term pernicious ef- fects (see Chapter 2, section 2.7), was stunningly high in these two countries, each with an estimated 668 deaths per 1,000 per year, among the highest in the study by Acemoglu, Johnson, and Robin- son (AJR); for comparison, the rate was just 15.5 in South Africa.*

Institutional Quality The expectation is that inherited institutions should be particularly bad in these two countries because colonialists would have had little incentive to protect property rights, encourage investment, or allow broad access to economic opportunities or political participation; instead, in stark terms, the incentive was to steal or have others steal for you. In their data for cur- rent institutional quality, the “average protection against expropriate risk” was 6.27 in Ghana and 7.00 in CIV, compared to a range from 3.50 in the Democratic Republic of Congo (known as Zaire at the time) to 10.00 in the United States—better, though not spectacularly better, investor protection. But a range of recent studies give higher marks to Ghana. Although all-country rankings of institu- tional quality should be used with caution, as they

265

can contain subjective elements; when a group of independently produced indicators with different focuses all point in the same direction, they be- come suggestive (though still never substituting for careful country-specific appraisal). Regarding corruption perceptions, according to Transparency International, in 2012 Ghana ranked 64th and CIV 130th out of 176 countries ranked. Regarding “ease of doing business,” the World Bank–International Finance Corporation 2010 rankings of 183 countries listed Ghana as 92nd (7th in sub-Saharan Africa) and CIV as 168th (32nd in the region). Regarding democracy, for 2012 the Economist listed Ghana (ranked 78th of 167) as a “flawed democracy” (two steps above authoritarian) and CIV (ranked 136th) as authoritarian. And on current property rights protections, a 2013 ranking sponsored by the Wall Street Journal and Heritage Foundation placed Ghana at 50 on a scale of 100 and CIV just 25. Critics point out limits and flaws of these various rankings, but they are consistent. So this, too, must be better understood. Is it because things had gotten so bad in Ghana that reform became the only option?

Ethnolinguistic Fractionalization Another fea- ture associated in the economics literature with low incomes and growth is ethnolinguistic fractionaliza- tion, with some social scientists also pointing out the potential dangers of religious fractionalization. In fact, both countries are fairly highly fractionalized, but CIV more so. Both countries have an Akan majority (45% in Ghana and 42% in CIV) and many smaller groups. In Ghana, the population is 69% Christian and 16% Muslim, but in CIV, adherents are much more evenly divided, with 39% Muslim and 33% Christian. Al- though scholars debate the proper way to measure fractionalization, seven main measures are used, with CIV higher on six, in some cases substantially higher.* CIV was torn by civil war in 2002, which has split the country, and the opportunistic use of fractionalization by political figures is an important factor.

Population Patterns of population growth are of- ten considered an important aspect of development, as discussed in Chapter 6. At independence in 1960,

266

*For example, according to the 1997 basic Easterly-Levine (ELF) measure, CIV was rated 0.86 and Ghana 0.71, with the range in Africa from 0.04 for Burundi to 0.9 for Congo and Uganda. On the widely cited 2003 Alesina et al. alternative measure, CIV is 0.82 and Ghana 0.67 in a range from 0 to 0.93. These are the usual baseline measures, but one measure of the seven points in the other direction: the 1999 measure of Fearon, on which CIV is 0.78 and Ghana 0.85.

the population of CIV was just 3.6 million, so it grew about 5½ times by 2007. In contrast, Ghana’s popula- tion was already nearly 7 million in 1960, so it grew by less than 31>3 times in the same period. Even now, the total fertility rate is a high 4.0 in Ghana but sig- nificantly higher at 4.9 in CIV, with one extra birth per woman. While just 8% of married women of childbearing age use modern contraceptives in CIV, 17% do in Ghana—still a small fraction but more than twice the incidence of CIV (the gap remains, at 24% to 13%, when considering both traditional and modern methods). High birth rates generally hinder economic development. Faster population growth is associated with slower per capita income growth and slower improvement in other development in- dicators; lower fertility increases family incentives and resources for education. But the geographic distribution of population does not seem to have particularly strong political implications. For ex- ample, Jeffrey Herbst classifies both Ghana and CIV as among just 7 of 40 sub-Saharan African countries with a “neutral political geography.”

Extreme Inequality As explained in Chapter 5 (and introduced in Chapter 2), extreme inequality can retard the development process. The most re- cent World Bank Gini coefficient estimates for CIV and Ghana do not differ significantly (at 0.42 and 0.43). But Arnim Langer points out that the com- bination of relatively high and rising inequality in CIV, coupled with rising ethnic tensions that politi- cal actors had deliberately made worse, led to the conflict that broke out there in the early 2000s (ethnic inequalities as a factor in conflict is consid- ered in Chapter 14, section 14.5).

Common Law versus Civil Law? As a former British colony, Ghana’s legal system is based on common law, while the legal system in CIV is based on French civil law. Since the late 1990s, the view that common law legal systems provide a better foundation than civil law systems for the develop- ment of the financial system has been very influ- ential. Authors in this literature such as Rafael La Porta and his colleagues argue either that common law better protects property rights, better enforces contracts, offers more predictability, or that it is bet- ter able to adapt to changes in economic conditions. Investment is generally necessary for economic growth (Chapters 3 and 4), and the development of an effective financial system encourages investment

267

(Chapter 15). Some evidence supports the predic- tion that civil law countries will experience less fi- nancial development and lower rates of investment. But differences between French and British institu- tions besides the legal system may be important.

French versus British Rule? The British Empire is commonly considered to have preferred indirect rule, relying on its ability to dominate local traditional political systems rather than to create new ones (pos- sibly related to common law tradition). In contrast, the French are said to have tended to employ direct rule of their colonies, introducing their own central- ized administrative structures, perhaps related to their own legal and historical traditions. Tactics might well have been similar regardless of the colonizer if condi- tions strongly favored central rule or indirect rule. But where starting conditions were similar in both colo- nies and when local advantages of either centraliza- tion or decentralization were not strong, a centralized French strategy and a decentralized British strategy might plausibly have been expected.

The evidence does reflect a more decentralized rule in British Ghana and more centralized rule in French Côte d’Ivoire. But if centralized rule is then transmitted to the postcolonial regime, the result can be a state with too few checks and balances. Decentralized rule, in contrast, provides better in- centives and checks against large-scale government corruption (see Chapter 11). The postcolonial record is complex but shows continued strong tendencies toward centralization in CIV, although the after- math of civil strife increases uncertainty about the future course (indeed there is some risk that CIV may face a prolonged period as a failed state). As Catherine Boone notes in her richly detailed study of both countries, the case of Ghana is subtle with initial but far from fully successful postcolonial government attempts at more centralization, prob- ably in part to wrest a larger share of agricultural revenues, but in 1992 there was a reinstatement of at least a ceremonial role—and unofficially a much larger role—for chiefs and other traditional village governance. This built on long traditions that were not systematically undermined under the British the way they were under the French.

Finally, some observers view postindependence CIV as having a more dependent relationship with France. Besides colonial rule having negative effects in general, close CIV dependence on its former co-

lonial ruler may have been a hindrance to its eco- nomic and political growth and development over the long run. In contrast, Ghana diversified more of its international relations, perhaps giving it some- what higher bargaining power in pursuing its na- tional development interests.

Education Some scholars consider education of central importance in explaining economic growth; Edward Glaeser and coauthors even argue that improved education can result in improved insti- tutions. Educational attainment was abysmal in both nations at the time of independence. One of the most striking postcolonial differences between the countries is the higher level of educational at- tainment in Ghana, where there have been greater investments in education. In the early years after independence, there was strong policy attention to providing basic education in some of the poorer areas in Ghana. In 2010, according to the 2013 Human Development Report (HDR), the mean years of schooling was almost 3 years higher in Ghana (at 7.0) than in CIV (at 4.2). Moreover, expected school- ing is now 11.4 years in Ghana, compared with only 6.5 years in CIV. Education is intrinsically valuable, as reflected in the HDI; it has apparently been a fac- tor in faster growth and may even figure in later institutional improvements. Ghana has also had recent success scaling up basic health insurance.

Development Policies Development policies are often framed by a country’s underlying eco- nomic institutions; this can place constraints on the types of beneficial reforms and policies that a country can successfully implement. The failure of a country to implement otherwise obvious policies (such as investing in quality primary education) may not reflect failures of understanding as much as the realities of political constraints. But when achieved, well-designed and implemented policies can have very positive effects on development outcomes; bad policies can have disastrous consequences.

Policies in Ghana Both nations started as (and still are) largely agrarian economies, with over half of the labor force working in rural areas. But the two countries have had somewhat different policy trajectories. The general scholarly view is that in the first quarter century after independence, Ghana chose many poorly conceived and often cor- rupt interventionist policies. Early policies have been described as oriented toward urban industry,

with inefficiently implemented import substitu- tion to replace manufactured imports with locally produced ones (see Chapter 12). But one policy as- sociated with the early rule of Kwame Nkrumah through 1966 was an emphasis on basic education, which may have left an enduring legacy through difficult subsequent swings. After disastrous poli- cies and extreme instability, including coups in the mid-1960s to early 1980s, Ghana underwent a policy transformation to become a favorite country of liberalization promoters in the World Bank and elsewhere in the 1980s.

The development process is complex and rarely proceeds linearly. In Ghana, there was relative deterioration from independence until the early 1980s; much of its economic growth took place from the mid-1980s to the present. For example, cocoa had long been an important part of Ghana’s economy, but it went into decline when state mar- keting boards (described in Chapter 9) limited the price farmers received for cocoa, so as to subsi- dize industrialization. After farmers were allowed to receive a much higher price and technical as- sistance was offered, output greatly increased, particularly in two spurts in the late 1980s and early 2000s. Fertilizer use and improved varieties have diffused among farmers (diffusion in Ghana for the case of pineapples is examined in Findings Box 9.1 in Chapter 9). Cocoa growing now pro- vides a basic livelihood for over 700,000 farmers in Ghana.

By the early 1990s, World Bank analysts such as Ishrat Husain were pointing to Ghana as a country that had been doing a better job at following and implementing more of its recommended market- friendly policies than countries such as CIV.

A reason given for large-scale reform in Ghana (and in explaining other countries as well) is that things got so bad that there became no choice but to embrace reform. Naturally, when according to lo- cal conditions things become so bad that continued resistance to change is futile, something changes— perhaps not always for the better. Ghana became a classic example for proponents of the controversial view that duress “causes” reform. A criticism, to paraphrase Dani Rodrik, is that it is not clear how much duress is enough to “cause” reform; and as a result, it is not very convincing when analysts

simply claim that a reform did not happen because the situation must not have been bad enough.

Policies in Côte d’Ivoire In contrast, CIV expe- rienced relatively faster growth in the 1960s and 1970s and then decline from 1980 to the present (recently more pronounced due to civil conflict). In- stitutions that appear to perform serviceably for two decades can have underlying weaknesses that later emerge—for example, politicians treat weaknesses as a political opportunity or the system proves to have too little flexibility as new challenges emerge.

CIV is widely viewed as having started down a more market-based, export-oriented path in a way that should have helped the rural agricultural sec- tor, where most of the population and most people living in poverty were located. But this did not pre- vent elites from extracting what they could from the rural areas. In fact, there were a number of policy lurches. An apparently favorable tactic might have been an early policy of effectively trying to keep all the ethnic groups engaged in and benefiting from growth in the national economy. There were large migrations into CIV, for example, including the forced labor brought into CIV from Burkina Faso (Upper Volta) by the French in the early 1940s. A more ethnically based politics in the late 1990s is viewed by specialists in the politics of CIV as help- ing to precipitate the disaster of regional and ethnic conflict in the 2000s.

Enduring Questions By 1990, Ghana was al- ready being deemed a “success story” by the World Bank and others. Is it because the nation followed the right policies? And if so, what explains why Ghana chose good policies and CIV did not? How much benefit can be attributed to the volume of aid itself?

CIV fell into a period of severe conflict in 2002– 2007; many lives were lost, and resources continue to be diverted into managing the problems, with perceptions of prospects still damaged. French military involvement reflected France’s ongoing unique relationship with CIV. In contrast, Ghana has so far remained stable. Why? And can it con- tinue to remain stable? It remains to be seen how well Ghana comes through its recent discovery and production of oil. In principle, new resources can help reduce poverty, directly and indirectly. But for many countries, a “resource curse” has resulted

268

269

from political conflict over resource revenues and an overspecialized or even “hollowed out” economy (see Chapter 14).

Have leadership differences mattered for de- velopment of these countries? Socialist Kwame Nkrumah constructively supported education but diverted resources from cocoa exports to local in- dustry, leading to economic disaster; under duress, socialist Jerry Rawlings embraced market-oriented policy reforms that led to short-term pain but long- term gain. Subsequent leaders have been pragmatic and at least have done relatively little harm and perhaps some good. CIV’s capitalist President Félix Houphouët-Boigny, backed by France (“Françaf- rique”), seemed early on to be leading his country to economic success but stole billions from the pub- lic purse and led the country to ruin while clinging to power for 33 years until his death in 1993. Subse- quent leadership has impressed few observers. Of course, extraordinary leadership in government or civil society can play a decisive role in the course of development—think of Nelson Mandela in South Africa or Muhammad Yunus in Bangladesh. But in ordinary experience, is leadership the key, or is it underlying institutions? Or popular movements? Education? Imported ideas and technology? These remain enduring questions, and answers may de- pend on local circumstances.

As an examination of just two countries to il- lustrate more general evidence in the literature, it cannot be concluded beyond doubt that institu- tions set up by Great Britain in Ghana and France

in Côte d’Ivoire had a dominant effect on the suc- cesses and failures of these nations in subsequent poverty reduction and economic growth. But there is support for factors identified in the large-sam- ple statistical studies introduced in Chapter 2, no- tably institutions, inequality, and at least indirectly education. Colonial institutions apparently had negative effects, and within colonization, the de- gree of decentralization under colonial rule appar- ently also mattered. The reemergence of more de- centralized governance in Ghana since 1992 may be related to less damaging British governance practices in this respect. At the same time, history is not destiny; Ghana has made progress that was not well predicted by instruments for colonial in- stitutions. Nor are things necessarily bleak for CIV. Institutions and inequality are highly resistant to change. But the global trend is toward continued progress in human development, and other Afri- can nations such as Rwanda have made enormous economic strides that were very difficult to imag- ine just a few years earlier. But in CIV, the standoff following contaminated presidential elections in 2010 led to what has been called the Second Ivo- rian Civil War. Rather than simply blame CIV, it may be possible to trace the shape of policymak- ing to underlying institutions– doing so may be a way to help address deeper constraints. It is to be hoped that the international community can play a constructive role in facilitating improvements in CIV’s underlying institutions, as well as to help se- cure the peace.

Sources The authors would like to thank Gina Lambright, David

Shinn, and Jennifer Spencer for their comments on the first draft of this case study, and Andrew Klein and Kevin Salador for their research assistance.

Acemoglu, Daron, Simon Johnson, and James A. Robin- son. “The colonial origins of comparative develop- ment: An empirical investigation.” American Econom- ics Review 91 (2001): 1360–1401.

Adjibolosoo, Senyo. “Ghana at fifty years old: A critical review of the historical genesis of why Ghanaians are where they are today.” Review of Human Factor Studies. 13 (2007): 6–40.

Alesina, Alberto, and Eliana La Ferrara. “Ethnic diver- sity and economic performance.” Journal of Economic Literature 43 (2005): 762–800.

Alkire, Sabina, and Maria Emma Santos “Acute Multi- dimensional Poverty: A New Index for Developing Countries.” OPHI Working Paper 38, 2010, at http:// www.ophi.org.uk/acute-multidimensional-poverty-a- new-index-for-developing-countries/

Beck, Thorsten, Asli Demirguc-Kunt, and Ross Levine. “Law and finance: Why does legal origin matter?” Journal of Comparative Economics (2003): 663–675.

Blunch, Niels-Hugo, and Dorte Verner. “Shared sectoral growth versus the dual economy model: Evidence from Côte d’Ivoire, Ghana, and Zimbabwe.” Develop- ment Review 18 (2006): 283–308.

Boone, Catherine. Political Topographies of the African State. Territorial Authority and Institutional Choice. Cambridge: Cambridge University Press, 2003.

Crook, Richard, Simplice Affou, Daniel Hammond, Adja F. Vanga, and Mark Owusu-Yeboah. “The law, legal institutions and the protection of land rights in Ghana and Côte d’Ivoire: Developing a more effective and equitable system.” IDS Research Report 58. Brighton, U.K.: Institute of Development Studies at the Univer- sity of Sussex, 2007.

Easterly, William. “Inequality does cause underdevelop- ment.” Journal of Development Economics 84 (2007): 755–776.

Engelbert, Pierre. “Pre-colonial institutions, post-colonial states, and economic development in tropical Africa.” Political Research Quarterly (2000): 7–36.

———. State Legitimacy and Development in Africa. Boulder, Colo.: Rienner, 2000.

Firmin-Sellers, Kathryn. “Institutions, context, and outcomes: Explaining French and British rule in West Africa.” Comparative Politics 32 (2000): 253–272.

Glaeser, Edward L., Rafael La Porta, Florencio Lopez de Silanes, and Andrei Shleifer. “Do institutions cause growth?” Journal of Economic Growth 9 (2004): 271–303.

Herbst, Jeffrey. States and Power in Africa. Comparative Les- sons in Authority and Control. Princeton, N.J.: Princeton University Press, 2000.

Heston, Alan, Robert Summers, and Bettina Aten. Penn World Table, vers. 6.3. Center for International Com- parisons of Production, Income, and Prices at the University of Pennsylvania, August 2009.

Husain Ishrat. “Why do some adjust more successfully than others? Lessons from seven African countries.” World Bank Africa Regional Office Policy Research Working Paper No. 1364, Office of the Chief Economist, 1994.

Husain, Ishrat, and Rashid Faruqee. Adjustment in Africa: Lessons from Country Case Studies. Washington, D.C.: World Bank, 1994.

La Porta, Rafael, Florencio Lopez-de-Silanes, Andrei Shleifer, and Robert W. Vishny. “Law and finance.” Journal of Political Economy 106 (1998) 1113–1155.

Langer, Arnim, “Horizontal Inequalities and Violent Conflict” Côte d’Ivoire Country Paper, 2005. http:// hdr.undp.org/en/reports.

MacLean, Lauren Morris. “Constructing a social safety net in Africa: An institutionalist analysis of colonial rule and state social policies in Ghana and Côte d’Ivoire.” Studies in Comparative International Develop- ment 37 (2002): 64–90.

Office of the Chief Economist, Africa Region, World Bank. Yes, Africa Can: Success Stories from a Dynamic Continent. Washington, D.C.: World Bank, 2009.

Population Council. “Côte d’Ivoire, 1998–99 results from the Demographic and Health Survey source.” Studies in Family Planning 34 (2003): 53–57.

Population Reference Bureau. World Population Data Sheet, 2012.

———. “Progress in Reducing Adolescent Childbearing: Ghana’s Success Story,” A PRB ENGAGE Snap- shot, 2013 http://www.prb.org/Journalists/Web- casts/2013/ghana-adolescent-childbearing-engage- short.aspx.

Rodrik, Dani. “Understanding economic policy reform.” Journal of Economic Literature 34 (1996): 9–41.

Stewart, Frances. “Horizontal inequalities: A neglected dimension of development.” QEH Working Paper Series, No. 81. Oxford: University of Oxford, 2002.

Tansel, Aysit. “Schooling attainment, parental education, and gender in Côte d’Ivoire and Ghana.” Economic Development and Cultural Change 45 (1997): 825–856.

Tsikata, Fui S. “National mineral policies in a changing world: The vicissitudes of mineral policy in Ghana.” Resources Policy 23 (1997): 9–14.

United Nations Development Programme. Human Development Report, 2009 and 2010. New York: Oxford University Press, 2009 and 2010.

White, Howard. “Using household survey data to meas- ure educational performance: The case of Ghana.” Social Indicators Research 74 (2005): 395–422.

World Bank. “PovcalNet.” http://iresearch.worldbank .org/PovcalNet.

———. World Development Indicators, 2010, 2013. Washington, D.C.: World Bank, 2010, 2013.

Yimam, Arega. Social Development in Africa, 1950–1985: Methodological Perspectives and Future Prospects. Brook- field, Vt.: Avebury, 1990.

270

271CHAPTER 5 Poverty, Inequality, and Development

Concepts for review

Absolute poverty Asset ownership Character of economic growth Decile Disposable income Elasticity of factor substitution

(Appendix 5.1) Factor price distortions

(Appendix 5.1) Factors of production Foster-Greer-Thorbecke

(FGT) index

Functional distribution of income Gini coefficient Headcount index Income inequality Indirect taxes Kuznets curve Land reform Lorenz curve Multidimensional poverty

index (MPI) Neoclassical price incentive

model (Appendix 5.1)

Personal distribution of income Progressive income tax Public consumption Quintile Redistribution policies Regressive tax Subsidy Total poverty gap (TPG) Workfare programs

Questions for Discussion

1. Most development economists now seem to agree that the level and rate of growth of GNI and per capita income do not provide sufficient measures of a country’s development. What is the essence of their argument? Give some examples.

2. Distinguish between size and functional distributions of income in a nation. Which do you conclude is the more appropriate concept? Explain your answer.

3. What is meant by absolute poverty? What measures of income poverty are favored by development economists? How do income poverty measures differ from the UNDP’s Multidimensional Poverty Index? Why should we be concerned with the mea- surement of poverty in developing nations?

4. What are the principal economic characteristics of high-poverty groups? What do these characteris- tics tell us about the possible nature of a poverty- focused development strategy?

5. Describe Kuznets’s inverted-U hypothesis. Discuss the conceptual merits and limitations of this hypoth- esis for contemporary developing countries.

6. In the text, when we examined statistics from a wide range of developing countries, we found that growth does not guarantee poverty reduction; while higher income is clearly associated with less poverty, economies can even reach upper-middle-income status but continue to struggle with a quite high in- cidence of extreme poverty. What does this tell us about the importance of the character of a nation’s growth process and about its institutional structure?

7. What is the relationship between a Lorenz curve and a Gini coefficient? Give some examples of how Lorenz curves and Gini coefficients can be used as summary measures of equality and in- equality in a nation’s distribution of income.

8. “The major determinant of a country’s income distribution is its distribution of productive and income-earning assets.” Explain the meaning of this statement, giving examples of different kinds of productive and income-earning assets.

9. Are rapid economic growth (as measured by ei- ther GNI or per capita GNI) and a more equal dis- tribution of personal income necessarily conflict- ing objectives? Summarize the arguments both for and against the presumed conflict of objectives, and state and explain your own view.

10. How might inequality lead to faster growth or de- velopment? How might it lead to slower growth or development?

11. Is progress being made in the fight against poverty? Why or why not?

12. What types of poverty policies have proved effective?

13. Economic growth is said to be a necessary but not sufficient condition to eradicate absolute poverty and reduce inequality. What is the reasoning be- hind this argument?

14. Outline the range of major policy options for a de- veloping country to alter and modify its size distri- bution of national income. Which policies do you believe are absolutely essential? Explain your answer.

272 PART Two Problems and Policies: Domestic

appendix 5.1

appropriate Technology and employment generation: The Price Incentive Model

Choice of Techniques: An Illustration

The basic proposition of the neoclassical price incentive model is quite sim- ple and in the best tradition of the neoclassical theory of the firm. Following the principle of economy, producers (firms and farms) are assumed to face a given set of relative factor prices (e.g., of capital and labor) and to use the com- bination of capital and labor that minimizes the cost of producing a desired level of output. They are further assumed to be capable of producing that out- put with a variety of technological production processes, ranging from highly labor-intensive to highly capital-intensive methods. Thus, if the price of capital is very expensive relative to the price of labor, a relatively labor-intensive pro- cess will be chosen. Conversely, if labor is relatively expensive, our economiz- ing firm or farm will use a more capital-intensive method of production—it will economize on the use of the expensive factor, which in this case is labor.

The conventional economics of technical choice is portrayed in Figure A5.1.1. Assume that the firm, farm, industry, or economy in question has only two tech- niques of production from which to choose: technique or process 0A, which requires larger inputs of (homogeneous) capital relative to (homogeneous) labor, and technique or process 0B, which is relatively labor-intensive. Points F and G

Neoclassical price incentive model A model whose main proposition is that if market prices are to in- fluence economic activities in the right direction, they must be adjusted to remove factor price distortions by means of subsidies, taxes, or the like so that factor prices may reflect the true opportunity cost of the resources being used.

Figure a5.1.1 Choice of Techniques: The Price Incentive Model

L

C a

p it

a l

K ′

Q1

Q1′

Labor

F

G �

A

B

K

L′0

�

6125_05_01_APP001

273CHAPTER 5 Poverty, Inequality, and Development

represent unit output levels for each process, and the line Q1 FGQ’1 connecting F and G is therefore a unit-output isoquant. (Note that in the traditional neoclas- sical model, an infinite number of such techniques or processes are assumed to exist so that the isoquant or equal-product line takes on its typical convex curvature.)

According to this theory, optimum (least-cost) capital-labor combinations (efficient or appropriate technologies) are determined by relative factor prices. Assume for the moment that market prices of capital and labor reflect their scar- city or shadow values and that the desired output level is Q1 in Figure A5.1.1. If capital is cheap relative to labor (price line KL), production will occur at point F using capital-intensive process 0A. Alternatively, if the market prices of labor and capital are such that labor is the relatively cheap (abundant) factor (line K’ L’), optimal production will occur at point G, with the labor-intensive tech- nique, 0B, chosen. It follows that for any technique of production currently in use, a fall in the relative price of labor, all other things being equal, will lead to a substi- tution of labor for capital in an optimal production strategy. (Note that if capital- intensive process 0A “dominates” labor-intensive process 0B—that is, if technology 0A requires less labor and less capital than 0B for all levels of output—then for any factor price ratio, the capital-intensive technique will be chosen.)

Factor Price Distortions and Appropriate Technology

Given that most developing countries are endowed with abundant supplies of labor but possess very little financial or physical capital, we would naturally expect production methods to be relatively labor-intensive. But in fact we of- ten find production techniques in both agriculture and industry to be heav- ily mechanized and capital-intensive. Large tractors and combines dot the ru- ral landscape of Asia, Africa, and Latin America, while people stand idly by. Gleaming new factories with the most modern and sophisticated automated machinery and equipment are a common feature of urban industries, while idle workers congregate outside the factory gates. Surely, this phenomenon could not be the result of a lesser degree of economic rationality on the part of farmers and manufacturers in developing countries.

The explanation, according to the price incentive school, is simple. Because of a variety of structural, institutional, and political factors, the actual market price of labor is higher and that of capital is lower than their respective true scarcity, or shadow, values dictate. In Figure A5.1.1, the shadow price ratio would be given by line K’ L’, whereas the actual (distorted) market price ra- tio is shown by line KL. Market wage structures are relatively high because of trade union pressure, politically motivated minimum-wage laws, an increasing range of employee fringe benefits, and the high-wage policies of multinational corporations. In former colonial nations, high-wage structures are often relics of expatriate remuneration scales based on European levels of living and “hard- ship” premiums. By contrast, the price of (scarce) capital is kept artificially low by a combination of liberal capital depreciation allowances, low or even negative real interest rates, low or negative effective rates of protection on capital goods imports, tax rebates, and overvalued foreign-exchange rates (see Chapter 13).

The net result of these factor price distortions is the encouragement of in- appropriate capital-intensive methods of production in both agriculture and

Factor price distortions Situations in which factors of production are paid prices that do not reflect their true scarcity values (i.e., their competitive market prices) be- cause of institutional arrange- ments that tamper with the free working of market forces of supply and demand.

274 PART Two Problems and Policies: Domestic

manufacturing. Note that from the private-cost-minimizing viewpoint of indi- vidual firms and farms, the choice of a capital-intensive technique is correct. It is their rational response to the existing structure of price signals in the market for factors of production. However, from the viewpoint of society as a whole, the social cost of underutilized capital, and especially labor, can be very sub- stantial. Government policies designed to “get the prices right”—that is, to remove factor price distortions—contribute not only to more employment but also to a better overall utilization of scarce capital resources through the adop- tion of more appropriate technologies of production.

The Possibilities of Labor-Capital Substitution

The actual employment impact of removing factor price distortions will de- pend on the degree to which labor can be substituted for capital in the pro- duction processes of various developing-world industries. Economists refer to this as the elasticity of factor substitution and define it roughly as the ratio of the percentage change in the proportion of labor used relative to capital (labor-capital or L/K ratio) compared to a given percentage change in the price of capital relative to labor (PK/PL). Algebraically, the elasticity of substitution, ηLK can be defined as follows:

ηLK = d1L>K2 1L>K2

d1PK>PL2 1PK>PL2 (A5.1.1)

For example, if the relative price of capital rises by 1% in the manufacturing sector and the labor-capital ratio rises as a result by, say, 1.5%, the elasticity of substitution in the manufacturing industry will be equal to 1.5. If PK/PL falls by, say, 10% while L/K falls by only 6%, the elasticity of substitution for that industry will be 0.6. Relatively high elasticities of substitution (ratios greater than about 0.7) are indicative that factor price adjustments can have a sub- stantial impact on levels and combinations of factor utilization. In such cases, factor price modifications may be an important means of generating more employment opportunities.

In general, most empirical studies of the elasticity of substitution for man- ufacturing industries in less developed countries reveal coefficients in the range 0.5–1.0. These results indicate that a relative reduction in wages (either directly or by holding wages constant while letting the price of capital rise) of, say, 10% will lead to a 5% to 10% increase in employment.

Elasticity of factor substitution A measure of the degree of substitutability between factors of produc- tion in any given production process when relative factor prices change.

275CHAPTER 5 Poverty, Inequality, and Development

appendix 5.2 The ahluwalia-Chenery Welfare Index

The necessity of reorienting development priorities away from an exclusive preoccupation with maximizing rates of GNI growth and toward broader so- cial objectives such as the eradication of poverty and the reduction of exces- sive income disparities is now widely recognized throughout the developing world. Figures for GNI per capita give no indication of how national income is actually distributed and who is benefiting most from the growth of produc- tion. We have seen, for example, that a rising level of absolute and per capita GNI can camouflage the fact that the poor are no better off than before.

The calculation of the rate of GNI growth is largely a calculation of the rate of growth of the incomes of the upper 40% of the population, who re- ceive a disproportionately large share of the national product. Therefore, the GNI growth rates can be a very misleading index of improved welfare. To give an extreme example, suppose that an economy consisted of only 10 people and that 9 of them had no income at all and the tenth received 100 units of income. The GNI for this economy would be 100 and per capita GNI would be 10. Now suppose that everyone’s income increased by 20% so that GNI rose to 120 while per capita income grew to 12. For the 9 individuals with no income before and still no income now (1.20 * 0 = 0), such a rise in per capita income would provide no cause for rejoicing. The one rich individual still would have all the income. And GNI, instead of being a welfare index of society as a whole, is merely measuring the welfare of a single individual!

The same line of reasoning applies to the more realistic situation where incomes are very unequally distributed, although not perfectly unequal as in our example. Taking the figures from Table 5.1, where we divided the popula- tion into quintiles that received 5%, 9%, 13%, 22%, and 51% income shares, respectively, we found that these income shares are a measure of the relative economic welfare of each income class and that the rate of income growth in each quintile is a measure of the economic welfare growth of that class. We can approximate the growth in the total welfare of society as the simple weighted sum of the growth of income in each class. This is in fact what the rate of GNI growth measures—the weights applied to each income class are their respec- tive shares of national income. To be specific, in the case of a population di- vided into quintiles according to rising income levels, we have

G = w1 g1 + w2 g2 + w3 g3 + w4 g4 + w5 g5 (A5.2.1)

where G = a weighted index of growth of social welfare, gi = the growth rate of income of the ith quintile (where the i quintiles are ordered 1, 2, 3, 4, and 5 in our example), and wi = the “welfare weight” of the ith quintile (in our example, w1 = 0.05, w2 = 0.09, w3 = 0.13, w4 = 0.22, and w5 = 0.51). As long as the weights add up to unity and are nonnegative, our overall measure of the growth of social welfare, G, must fall somewhere between the maximum and minimum income growth rates in the various quintiles. In the extreme case of all income accruing to one individual or one group of individuals in the highest

276 PART Two Problems and Policies: Domestic

quintile and where the welfare weights are the income shares (as they are with GNI growth calculations), Equation A5.2.1 would be written as

G = 0g1 + 0g2 + 0g3 + 0g4 + 1g5 = 1g5 (A5.2.2)

The growth of social welfare would therefore be associated exclusively with the growth of incomes of the top quintile of the population!

In the example derived from Table 5.1, the GNI-share-weighted index of social welfare would be written as

G = 0.05g1 + 0.09g2 + 0.13g3 + 0.22g4 + 0.51g5 = 1g5 (A5.2.3)

Now suppose that the income growth rate of the bottom 60% of the popu- lation was zero (g1 = g2 = g3 = 0) while that of the top 40% was 10% (g4 = g5 = 0.10). Equation A5.2.3 could then be written as

G = 0.05102 + 0.09102 + 0.1310.102 + 0.2210.102 + 0.5110.102 = 0.073 (A5.2.4)

and the social welfare index would rise by more than 7%, which is the rate of growth of GNI (i.e., GNI would rise from 100 in Table 5.1 to 107.3 if the incomes of the 4th and 5th quintiles grew by 10%). Thus, we have an illustra- tion of a case where GNI rises by 7.3%, implying that social well-being has increased by this same proportionate amount even though 60% of the popu- lation is no better off than before. This bottom 60% still has only 5, 13, and 22 units of income, respectively. Clearly, the distribution of income would be worsened (the relative shares of the bottom 60% would fall) by such a respect- able growth rate of GNI.

The numerical example given by Equation A5.2.4 illustrates our basic point. The use of the growth rate of GNI as an index of social welfare and as a method of comparing the development performance of different countries can be misleading, especially where countries have markedly different distribu- tions of income. The welfare weights attached to the growth rates of differ- ent income groups are unequal, with a heavy social premium being placed on the income growth of the highest-quintile groups. In the example of Equation A5.2.3, a 1% growth in the income of the top quintile carries more than 10 times the weight of a 1% growth in the lowest quintile (0.51 compared with 0.05) because it implies an absolute increment that is 10 times larger. In other words, using the measure of GNI growth as an index of improvements in so- cial welfare and development accords to each income group a welfare valua- tion that corresponds to its respective income share (i.e., a 1% increase in the income of the richest 20% of the population is implicitly assumed to be more than 10 times as important to society as a 1% increase in the income of the bot- tom 20%). It follows that the best way to maximize social welfare growth is to maximize the rate of growth of the incomes of the rich while neglecting the poor! If ever there was a case for not equating GNI growth with development, this example should provide a persuasive illustration.

277CHAPTER 5 Poverty, Inequality, and Development

Constructing a Poverty-Weighted Index of Social Welfare

An alternative to using a simple GNI growth rate or distributive share index of social welfare would be to construct an equal-weights or even a poverty- weighted index. Such indexes might be especially relevant for countries con- cerned with the elimination of poverty as a major development objective. As the name indicates, an equal-weights index weights the growth of income in each income class not by the proportion of total income in that class but rather by the proportion of the total population—that is, all people are treated (weighted) equally. In an economy divided into quintiles, such an index would give a weight of 0.2 to the growth of income in each quintile. So a 10% increase in the income of the lowest 20% of the population would have the same bearing on the overall measure of social welfare improvements as a 10% increase in the top 20% group or in any other quintile group, even though the absolute increase in income for the bottom group would be much smaller than for the upper groups.

Using an equal-weights index in our example of a 10% income growth of the top two quintiles with the bottom three remaining static, we would have

G = 0.20g1 + 0.20g2 + 0.20g3 + 0.20g4 + 0.20g5 (A5.2.5)

or, inserting growth rates for g1, through g5,

G = 0.20102 + 0.20102 + 0.20102 + 0.201102 + 0.2010.102 = 0.04 (A5.2.6)

Social welfare would increase by only 4%, compared to the 7.3% increase recorded by using the distributive shares or GNI growth rate index. Even though recorded GNI still grew by 7.3%, this alternative welfare index of de- velopment would show only a 4% rise.

Finally, consider a developing country that is genuinely and solely con- cerned with improving the material well-being of, say, the poorest 40% of its population. Such a country might wish to construct a poverty-weighted index of development, which places “subjective” social values on the income growth rates of only the bottom 40%. In other words, it might arbitrarily place a wel- fare weight on w1 of 0.60 and on w2 of 0.40 while giving w3, w4, and w5 zero weights. Using our same numerical example, the social welfare growth index for this country would be given by the expression

G = 0.60g1 + 0.40g2 + 0g3 + 0g4 + 0g5 (A5.2.7)

which, when substituting g1 = g2 = g3 = 0 and g4 = g5 = 0.10, becomes

G = 0.60102 + 0.40102 + 0102 + 010.102 + 010.102 = 0 (A5.2.8)

The poverty-weighted index therefore records no improvement in social wel- fare (no development), even though recorded GNI has grown by 7.3%!

Although the choice of welfare weights in any index of development is purely arbitrary, it does represent and reflect important social value judgments about goals and objectives for a given society. It would certainly be interest- ing to know, if this were possible, the real implicit welfare weights of the

278 PART Two Problems and Policies: Domestic

various development strategies of different developing countries. Our main point, however, is that as long as the growth rate of GNI is explicitly or implic- itly used to compare development performances, we know that a “wealthy weights” index is actually being employed.

To put some real-world flavor into the discussion of alternative indexes of improvements in economic welfare and to illustrate the usefulness of different weighted growth indexes in evaluating the economic performance of various countries, consider the data in Table A5.2.1 compiled by Montek Ahluwalia and Hollis Chenery. The table shows the growth of income in 12 countries as mea- sured first by the rate of growth of GNI (GNI weights), second by an equal- weights index, and third by a poverty-weighted index where the actual weights assigned to income growth rates of the lowest 40%, the middle 40%, and the top 20% of the population are 0.6, 0.4, and 0.0, respectively. Some interesting conclu- sions emerge from a review of the last three columns of Table A5.2.1:

1. Economic performance as measured by equal-weights and poverty- weighted indexes was notably worse in some otherwise high-GNI-growth countries like Brazil, Mexico, and Panama. Because these countries all ex- perienced a deterioration in income distribution and a growing concentra- tion of income growth in the upper groups over this period, the equal- weights and poverty-weighted indexes naturally show a less impressive development performance than the simple GNI measure.

2. In five countries (Colombia, Costa Rica, El Salvador, Sri Lanka, and Taiwan), the weighted indexes show a better performance than GNI growth, be- cause the relative income growth of lower-income groups proceeded more rapidly over the period in question in those five countries than that of the higher-income groups.

Table a5.2.1 Income Distribution and Growth in 12 Selected Countries

Sources: International Bank for Reconstruction and Development/The World Bank: Redistribution with Growth: An Approach to Policy. Copyright © 1974 by The World Bank. Reprinted with permission.

  Income growth annual Increase in Welfare Country

upper 20%

Middle 40%

lowest 40%

gNI Weights

equal Weights

Poverty Weights

Brazil Colombia Costa Rica El Salvador India Mexico Panama Peru Philippines South Korea Sri Lanka Taiwan

6.7 5.2 4.5 3.5 5.3 8.8 8.8 3.9 5.0

12.4 3.1 4.5

3.1 7.9 9.3 9.5 3.5 5.8 9.2 6.7 6.7 9.5 6.3 9.1

3.7 7.8 7.0 6.4 2.0 6.0 3.2 2.4 4.4

11.0 8.3

12.1

5.2 6.2 6.3 5.7 4.2 7.8 8.2 4.6 5.5

11.0 5.0 6.8

4.1 7.3 7.4 7.1 3.3 6.5 6.7 4.4 5.4

10.7 6.5 9.4

3.5 7.8 7.8 7.4 2.5 5.9 5.2 3.8 5.2

10.5 7.6

11.1

279CHAPTER 5 Poverty, Inequality, and Development

Notes

1. The Lorenz curve is named for Max Otto Lorenz, an American economist who in 1905 devised this convenient and widely used diagram to show the relationship between population groups and their respective income shares.

2. A more precise definition of perfect equality would take into account the age structure of a population and expected income variations over the life cycle of all households within that popula- tion. See Morton Paglin, “The measurement and trend of inequality: A basic revision,” American Economic Review 65 (1975): 598–609.

3. For the details, see Gary S. Fields, Distribution and Development: A New Look at the Developing World (Cambridge, Mass.: MIT Press, 2001), ch. 2.

4. For more details on this and an alternative exposi- tion of inequality properties, see Amartya Sen and James E. Foster, On Economic Inequality, expanded ed. (Oxford: Clarendon Press, 1997).

5. The sum of all workers’ marginal product must equal total gross national income (GNI). Mathematically, GNI is simply the integral of the marginal product curve between 0 and LE. This is because the marginal product function is the derivative of the GNI curve: GNI = ƒ(L, K); MPL = ƒ’(L).

6. If measured poverty is always strictly lower after such transfers, this property is called strong mo- notonicity. The headcount ratio satisfies monoto- nicity but not strong monotonicity.

7. For technical details, see James Foster, Joel Greer, and Erik Thorbecke, “A class of decomposable pov- erty measures,” Econometrica 52 (1984): 761–766.

8. For proof that Equation 5.4 follows from Equation 5.3, see Foster, Greer, and Thorbecke, “A class of

decomposable poverty measures,” Cornell Uni- versity Discussion Paper No. 242, 1981.

9. It is similar in spirit to the Sen index, S 5 (H/N) [NIS 1 (1 2 NIS)Gp], where Gp stands for the Gini coefficient among the poor. For the technical de- tails and derivations of the P2 and S poverty meas- ures, see Sen and Foster, On Economic Inequality, pp. 165–194, and ibid.

10. For the same reason, the P2 measure has now be- come part of the Mexican constitution (chap. 5, art. 34). Interview with Erik Thorbecke, Cornell Chronicle, May 11, 2000.

11. For example, Uganda saw impressive reductions in poverty between 1999 and 2009, but the head- count decreased by only 1.9 million people. By the person-equivalent measure, poverty fell by 4.4 million poor person-equivalents. This measure adjusts for poverty depth, but still does not reflect poverty severity. For more details on the measure, along with applications to data from a number of developing countries, see Tony Castleman, James E. Foster, and Stephen C. Smith, “Person-Equivalent Poverty Measures,” paper presented at the Brook- ings Institution, February 12, 2013.

12. The Alkire-Foster method, as it has come to be known, reduces to the FGT index when poverty is measured with just one dimension. See Sabina Alkire and James Foster, “Counting and multidi- mensional poverty measurement,” Journal of Pub- lic Economics 95, No. 7 (2011): 476–487. For further intuition, see also Alkire and Foster “Understand- ings and misunderstanding of multidimensional poverty,” Journal of Economic Inequality 9(2), pp. 289–314.

3. In three countries (Peru, the Philippines, and South Korea), little change in income distribution during the period in question resulted in little varia- tion between the GNI measure and the two alternative weighted indexes of social welfare.

We may conclude, therefore, that a useful summary measure of the degree to which economic growth is biased toward the relative improvement of high- income or low-income groups is the positive or negative divergence between a weighted social welfare index and the actual growth rate of GNI.

280 PART Two Problems and Policies: Domestic

13. Various UN studies on sources of savings in de- veloping nations show that small farmers and individuals seem to be among the highest savers. See Andrew Mason, “Savings, economic growth and demographic change,” Population and Devel- opment Review 14 (1988): 113–144.

14. Two technical articles that address the mecha- nisms by which higher inequality may lead to lower growth or incomes are Abhijit V. Banerjee and Andrew F. Newman, “Occupational choice and the process of development,” Journal of Political Economy 101 (1993): 274–298, and Oded Galor and Joseph Zeira, “Income distribution and macroeco- nomics,” Review of Economic Studies 60 (1993): 35–52. See also Fields, Distribution and Development, ch. 10. The empirical literature remains mixed, however.

15. See, for example, Torsten Persson and Guido Tabellini, “Is inequality harmful for growth?” American Economic Review 84 (1994): 600–621, and Alberto Alesina and Dani Rodrik, “Distributive politics and economic growth,” Quarterly Journal of Economics 109 (1994): 465–490. On the connec- tion to violent crime, see Morgan Kelly, “Inequal- ity and crime,” The Review of Economics and Statistics 82, No. 4 (2000), pp. 530–539.

16. John Rawls, A Theory of Justice (Cambridge, Mass.: Belknap Press, 1971).

17. This approach was developed by Gary S. Fields, Poverty, Inequality and Development (Cambridge: Cambridge University Press, 1980), pp. 46–56.

18. Ibid., p. 52.

19. This can perhaps be visualized most easily by considering a traditional economy in which everyone is “equally poor,” each claiming their share of, say, 50 cents per day. If the absolute pov- erty line is $1.25 per day, all are in absolute pov- erty. Then modernization begins, and the modern sector absorbs workers one by one, where the wage is, say, $2 per day. Starting from the line of perfect equality, the Lorenz curve bows out more and more until nearly half the people are in the modern sector. At that point, as more go to the modern sector, the Lorenz curve is less bowed in until finally everyone has been absorbed into the modern sector and all once again have equal in- comes but now at a higher level of $2 per day. In the process, all of the people have been pulled out

of poverty. (Try this as an exercise, plotting the Lorenz curves as this process takes place for an eight-person economy.) This exercise is adapted from Fields, ibid.

20. In fact, some would go further and say that an in- crease in relative inequality is not objectionable as long as everyone has a higher income, even though the rich get a larger share of the gains, even in pro- portion to their larger starting income. This situa- tion is called “first-order stochastic dominance” in the literature. However, even in this case, incomes might be increased even more with less inequality.

21. Of course, in real economies, all three of these growth typologies may take place at the same time, and the net result may be little or no change in inequality. Or in more unfortunate cases, with economies with negative growth, like many of those in sub-Saharan Africa in the 1980s and 1990s, there may be modern- and traditional-sector impoverishment, accompa- nied by a shrinking modern sector.

22. Simon Kuznets, “Economic growth and income inequality,” American Economic Review 45 (1955): 1–28, and “Quantitative aspects of the economic growth of nations,” Economic Development and Cultural Change 11 (1963): 1–80. One of the cross- sectional studies supporting the Kuznets hypoth- esis is Montek S. Ahluwalia, Nicholas G. Carter, and Hollis B. Chenery, “Growth and poverty in developing countries,” Journal of Development Economics 16 (1979): 298–323. Studies arguing against the hypothesis include Ashwani Saith, “Development and distribution: A critique of the cross-country U-hypothesis,” Journal of Develop- ment Economics 13 (1983): 367–382, and Sudhir Anand and S. M. R. Kanbur, “The Kuznets process and the inequality-development relationship,” Journal of Development Economics 40 (1993): 25–42.

23. The parabola plotted results from an ordinary least-squares regression. Fields reports results showing that in using a country fixed-effect specification, the estimated inverted U flips to an estimated U-pattern. For details, see Fields, Distri- bution and Development, ch. 3 (pp. 42–43).

24. Ibid., p. 35.

25. The 2008 and 2010 estimates are reported in the 2013 World Development Indicators. For an over- view on the $1.25 a day estimation, see Martin

281CHAPTER 5 Poverty, Inequality, and Development

Ravallion, Shaohua Chen, and Prem Sangraula, New Evidence on the Urbanization of Global Poverty (Washington, D.C.: World Bank, 2007), and Martin Ravallion, Shaohua Chen, and Prem Sangraula, “Dollar a Day Revisited,” World Bank, Policy Research Working Paper No. 4620, May 2008.

26. The HPI measured three deprivations—of life (as the percentage of people unlikely to live beyond 40 years of age), of basic education (as the per- centage of adults who are illiterate), and of overall economic provisioning (as the percentage of peo- ple without access to safe water plus the percent- age of children underweight for their age), giving them equal weight in a manner analogous to the original HDI. The 2009 HDR report ranked 135 countries from lowest to highest HPI and found this could differ substantially from income pov- erty rankings and the old HDI ranking. Since the HPI value indicates the proportion of the popula- tion adversely affected by the three deprivations, a higher HPI reflects greater deprivation. In the report, Côte d’Ivoire ranked 29 places higher (worse) in the country rankings based on income poverty than on human poverty; Morocco ranked 50 places higher; Iran, 44 higher; Algeria, 19 higher; Ethiopia, 30 higher. The implication is that human poverty is worse in these countries than headcount ratio in- come poverty measures indicate. In contrast, some of the countries that perform better on the human poverty ranking include Nigeria, 11 places lower; Ghana, 18 lower; Madagascar, 14 lower; Bolivia, 21 lower; and Tanzania, 37 lower. The MPI is strongly preferred because it aggregates up from the house- hold level and allows for interactions of poverty dimensions; an index like the HPI may be used be- cause it is familiar, can be applied to a larger num- ber of countries, and can be extrapolated further back in time and at more frequent intervals.

27. The MPI was introduced in the 2010 Human De- velopment Report (New York: United Nations Development Programme, 2010); for details, see Sabina Alkire and Maria Emma Santos, Acute Multidimensional Poverty: A New Index for Devel- oping Countries, Human Development Research Paper No. 2010/11 (New York: United Nations Development Programme, 2010). The MPI is based on the increasingly used Alkire-Foster Method (AFM); for an introduction, see Sabina Alkire and

James Foster, “Counting and multidimensional poverty measurement,” in Oxford Poverty and Human Development Initiative, Working Paper 07, 2008. Forthcoming Journal of Public Economics.

28. UNDP Human Development Report, 2010.

29. See Chronic Poverty Research Centre, Chronic Pov- erty Report, 2004–05, http://www.chronicpoverty .org/resources/cprc_report_2004-2005_contents .html, and Andrew McKay and Bob Baulch, “How many chronically poor people are there in the world? Some preliminary estimates,” CPRC Working Paper No. 45, Chronic Poverty Research Centre, 2003.

30. We may also note that greater spatial concentra- tion of poverty—a higher percentage of people in a given region who are poor—is an additional consideration for how ultrapoverty differs.

31. See International Food Policy Research Institute, The World’s Most Deprived (Washington: D.C.: IFPRI, 2007).

32. Partha Dasgupta and Debraj Ray, “Inequality as a determinant of malnutrition and unemployment policy,” Economic Journal 97 (1987): 177–188.

33. An empirical study of variables explaining growth in developing countries during the 1960–1973 pe- riod provide support for the argument that poli- cies designed to promote better distribution and reduce poverty are, on balance, growth-stimulat- ing rather than growth-retarding. See Norman L. Hicks, “Growth vs. basic needs: Is there a trade- off?” World Development 7 (1979): 985–994.

34. For empirical evidence on how improved distri- bution can increase domestic demand, promote political stability, and generate higher growth rates, see Alberto Alesina and Roberto Perotti, “The political economy of growth: A critical sur- vey of the recent literature,” World Bank Economic Review 8 (1994): 351–371, and Alberto Alesina and Dani Rodrik, “Distributive policies and economic growth,” Quarterly Journal of Economics 109 (1994): 465–490.

35. See World Bank, World Development Report, 2000/2001 (New York: Oxford University Press, 2000). See also World Bank, World Development Report, 1990 (New York: Oxford University Press, 1990); Albert Fishlow, “Inequality, poverty, and growth: Where do we stand?” in Proceedings of

282 PART Two Problems and Policies: Domestic

the World Bank Annual Conference on Development Economics, 1995, eds. Michael Bruno and Boris Pleskovic (Washington, D.C.: World Bank, 1996); Nancy Birdsall, David Ross, and Richard Sabot, “Inequality and growth reconsidered: Lessons from East Asia,” World Bank Economic Review 9 (1995): 477–508; and George R. G. Clarke, “More evidence on income distribution and growth,” Journal of Development Economics 47 (1995): 403–427.

36. A well-known study is David Dollar and Aart Kraay, “Growth is good for the poor,” Journal of Economic Growth 7 (2002): 195–225. They find that on average, incomes of the bottom 20% grow about as fast as the overall average. However, critiques of the generality of this claim of rough proportionality have been summarized by the University of Manchester Chronic Poverty Re- search Center in its Chronic Poverty Report 2004/05: “It does not allow for variation around the aver- age (which is known to be significant), it uses a relative concept of poverty, the data set used has been criticized, it does not consider poverty depth, and researchers using different economet- ric methods with the same data have produced contradictory findings.” Clearly it is possible and sometimes does occur that inequality can increase with growth enough to offset any gains for the poor, including some cases in which rapid growth increases the incentive and opportunity of theft of natural resources from poor communities. The es- sential point is that growth is not guaranteed to automatically end absolute poverty or do so in an acceptable time frame, so targeted policies are generally also needed.

37. For a classic overview of the nature, magni- tude, and incidence of poverty in the developing world, see World Bank, World Development Report, 2000/2001.

38. For a comprehensive analysis of how poverty directly affects women’s lives in developing countries, see Irene Tinker, Persistent Inequalities: Women and World Development (New York: Oxford University Press, 1990); Judith Bruce and Daisy Dwyer, eds., A Home Divided: Women and Income in the Third World (Stanford, Calif.: Stanford Uni- versity Press, 1988); Janet Momsen, Women and

Development in the Third World (New York: Rout- ledge, 1991); and Diane Elson, “Gender-aware analysis and development economics,” Journal of International Development 5 (1993): 237–247.

39. Amartya Sen, “Missing women,” British Medi- cal Journal 304 (1992): 587–588. A well-regarded 2003 analysis conclude that about 100 million or more women are “missing” in Asia alone. Stephan Klasen and Claudia Wink, “Missing Women: Re- visiting the Debate,” Feminist Economics, 9 (2–3), 2003, 263–299.

40. The International Fund for Agricultural Devel- opment provides basic statistics and links to key resources on indigenous peoples and develop- ment at http://www.iFad.org/pub/factsheet/ ip/e.pdf.

41. See, for example, Haeduck Lee, The Ethnic Di- mension of Poverty and Income Distribution in Latin America (Washington, D.C.: World Bank, 1993); George Psacharopoulos and Harry A. Patri- nos, “Indigenous people and poverty in Latin America,” Finance and Development 31 (1994): 41– 43; and Gillette Hall and Harry Anthony Patrinos, eds., Indigenous Peoples, Poverty and Human De- velopment in Latin America; 1994–2004 (New York: Palgrave Macmillan, 2006).

42. Darryl McLeod and Nora Lustig, “Minimum wages and poverty in developing countries: Some empirical evidence,” in Labor Markets in Latin America: Combining Social Protection with Market Flexibility (Washington, D.C.: Brookings Institu- tion, 1997). An interesting theoretical contribution is found in Gary S. Fields and Ravi Kanbur, “Min- imum wages and poverty with income-sharing,” Journal of Economic Inequality 5 (2007): 135–147. Details of SEWA’s in-house studies on minimum wages for poor informal workers are found at http://www.sewaresearch.org.

43. For the classic analytical treatment of the work- fare-versus-welfare problem, see Timothy J. Bes- ley and Stephen Coate, “Workfare versus welfare: Incentive arguments for work requirements in poverty alleviation programs,” American Economic Review 82 (1992): 249–261.

44. For other discussions of poverty policies, see Arne Bigsten, “Poverty, inequality and development,”

283CHAPTER 5 Poverty, Inequality, and Development

in Surveys in Development Economics, ed. Norman Gemmell (Oxford: Blackwell, 1987), pp. 157–163; Jagdish N. Bhagwati, “Poverty and public pol- icy,” World Development 16 (1988): 539–555; World Bank, World Development Report, 2000/2001; Irma Adelman and Sherman Robinson, “Income distri- bution and development,” in Handbook of Develop- ment Economics, vol. 2, eds. Hollis B. Chenery and

T. N. Srinivasan (Amsterdam: Elsevier, 1989), pp. 982–996; and Paul P. Streeten, Strategies for Human Development: Global Poverty and Unemploy- ment (Copenhagen: Handelshojskølens Forlag, 1994).

45. James Speth, “Foreword,” in United Nations Development Programme, Human Development Report, 1997, p. iii.

6.1 The Basic Issue: Population Growth and the Quality of Life

In 2013, the world’s population reached about 7.2 billion people. In that year, the United Nations Population Division projected that population would rise to about 8.1 billion in 2025 and reach about 9.6 billion by the year 2050. The overwhelming majority of that population will inhabit the developing world. What will be the economic and social implications for development if such projections are realized? Is this scenario inevitable, or will it depend on the success or failure of development efforts? Finally, even more significant, is rapid population growth per se as serious a problem as many people believe, or is it a manifestation of more fundamental problems of underdevelopment and the unequal utilization of global resources between rich and poor nations, as others argue?

In this chapter, we examine many of the issues relating population growth to economic development. We begin, however, by looking at historical and recent population trends and the changing geographic distribution of the world’s people. After explaining basic demographic concepts, we present some well-known economic models and hypotheses regarding the causes and consequences of rapid population growth in contemporary developing countries. Controversies surrounding the significance of the population factor in general and these models and hypotheses in particular are then explored.

284

Population Growth and Economic Development: Causes, Consequences, and Controversies

Economic development may be far from “the best contraceptive” [that it is sometimes described as]…. On the other hand, social development—especially women’s education and employment—can be very effective indeed.

—Amartya Sen, Nobel laureate in economics

6

285CHAPTER 6 Population Growth and Economic Development

Finally, we evaluate a range of alternative policy options that developing countries may wish to adopt to influence the size and growth of their popu- lations, as well as ways in which industrialized countries can contribute to a more manageable global population and resource environment. Population policies in China and India, the nations with the largest populations in the world, are the focus of this chapter ’s case study.

Every year, more than 75 million people are being added to the world’s population. Almost all of this net population increase—97%—is in developing countries. Increases of such magnitude are unprecedented. But the prob- lem of population growth is not simply a problem of numbers. It is a prob- lem of human welfare and of development, as defined in Chapter 1. Rapid population growth can have serious consequences for the well-being of all humanity. If development entails the improvement in people’s levels of living—their incomes, health, education, and general well-being—and if it also encompasses their capabilities, self-esteem, respect, dignity, and free- dom to choose, then the really important question about population growth is this: How does the contemporary population situation in many devel- oping countries contribute to or detract from their chances of realizing the goals of development, not only for the current generation but also for future generations? In addressing this central issue, we examine the reasons and consequences for the positive relationship between poverty and family size. More broadly, we examine what drives high population growth in develop- ing (particularly low-income) countries, why population growth in general subsequently falls as countries grow and develop, and the causes and impli- cations of these patterns.

6.2 Population Growth: Past, Present, and Future

World Population Growth throughout History

For most of human existence on earth, humanity’s numbers have been few. When people first started to cultivate food through agriculture some 12,000 years ago, the estimated world population was no more than 5 million (see Table 6.1). Two thousand years ago, world population had grown to nearly 250 million, less than a fifth of the population of China today. From year 1 on our calendar to the beginning of the Industrial Revolution around 1750, it tri- pled to 728 million people, less than three-quarters of the total number living in India today. During the next 200 years (1750–1950), an additional 1.7 billion people were added to the planet’s numbers. But in just four decades thereafter (1950–1990), the earth’s human population more than doubled again, bringing the total figure to around 5.3 billion. The world entered the twenty-first cen- tury with over 6 billion people.

As seen in Figure 6.1, in 1950 about 1.7 billion people lived in develop- ing countries, representing about two-thirds of the world total; by 2050, the population of less developed countries will reach over 8 billion, nearly seven-eighths of the world’s population. In the corresponding period,

286 PART TWO Problems and Policies: Domestic

the population of the least developed countries will increase by tenfold, from about 200 million to 2 billion people. In contrast, the population of the developed countries will grow very little between now and 2050, even accounting for immigration from developing countries.

Turning from absolute numbers to percentage growth rates, for almost the whole of human existence on earth until approximately 300 years ago, population grew at an annual rate not much greater than zero (0.002%, or 20 per million). Naturally, this overall rate was not steady; there were many ups and downs as a result of natural catastrophes and variations in growth rates among regions. By 1750, the population growth rate had accelerated to 0.3% per year. By the 1950s, the rate had again accelerated, tripling to about 1.0% per year. It continued to accelerate until around 1970, when it peaked at 2.35%.1 Today the world’s population growth rate remains at a historically high rate of nearly 1.2% per year, but the rate of increase is slow- ing. However, the population growth rate in Africa is still an extremely high 2.3% per year. (Note that estimates of population numbers and growth rates differ according to research methods, but the broad trends are similar across major studies.)

The relationship between annual percentage increases and the time it takes for a population to double in size, or doubling time,2 is shown in the rightmost column of Table 6.1 (calculation of doubling time is explained in endnote 2). We see that before 1650, it took nearly 36,000 years, or about

Doubling time Period that a given population or other quantity takes to increase by its present size.

Year

Estimated Population (millions)

Estimated Annual Increase in the

Intervening Period (%)

Doubling Time

(years)

10,000 b.c.e. 5     1 c.e. 250 0.04 1,733 1650 545 0.04 1,733 1750 728 0.29 239 1800 906 0.45 154 1850 1,171 0.53 130 1900 1,608 0.65 106 1950 2,576 0.91 76 1970 3,698 2.09 33 1980 4,448 1.76 39 1990 5,292 1.73 40 2000 6,090 1.48 47 2010 6,892 1.22 57

2050 (projected) 9,600 0.98 71

TABLE 6.1 Estimated World Population Growth

Sources: Population Reference Bureau, World Population Data Sheet (Washington, D.C.: Population Reference Bureau, 2010 and previous annuals); Warren S. Thompson and David T. Lewis, Population Problems, 5th ed. (New York: McGraw-Hill, 1965), p. 384; United Nations, Demographic Yearbook for 1971 (New York: United Nations, 1971); United Nations, Report on the World Social Situation, 1997 (New York: United Nations, 1997), p. 14; and United Nations Population Division, World Population Prospects: The 2012 Revision. New York: United Nations (2013). An alternate system of broadly comparable and earlier estimates is found in Michael Kremer, “Population growth and technological change: One million B.C. to 1990,” Quarterly Journal of Economics 108 (1993): 681–716.

287CHAPTER 6 Population Growth and Economic Development

1,400 generations, for the world population to double. Today it would take about 58 years, or two generations, for world population to double at cur- rent growth rates. Moreover, whereas it took 1,750 years to add 480 million people to the world’s population between year 1 and the onset of the Indus- trial Revolution, this same number of people is today being added in less than 7 years.

The reason for the sudden change in overall population trends is that for almost all of recorded history, the rate of population change, whether up or down, had been strongly influenced by the combined effects of famine, dis- ease, malnutrition, plague, and war—conditions that resulted in high and fluctuating death rates. In the twentieth century, such conditions came increas- ingly under technological and economic control. As a result, human mortal- ity (the death rate) is now lower than at any other point in human existence. It is this decline in mortality resulting from rapid technological advances in modern medicine, improved nutrition, and the spread of modern sanitation measures throughout the world, particularly within the past half-century, that has resulted in the unprecedented increases in world population growth, especially in developing countries. In short, population growth today is pri- marily the result of a rapid transition from a long historical era characterized by high birth and death rates to one in which death rates have fallen sharply but birth rates, especially in the least developed countries, have fallen more slowly from their historically high levels.

Structure of the World’s Population

The world’s population is very unevenly distributed by geographic region, by fertility and mortality levels, and by age structures.

Geographic Region More than three-quarters of the world’s people live in developing countries; fewer than one person in four lives in an economically developed nation. Figure 6.2 shows the regional distribution of the world’s population as it existed in 2010 and as it is projected for 2050.

0

10

P o

p u

la ti

o n

( b

il li

o n

s)

2050

Less Developed Countries

1970 1990

More Developed Countries

Least Developed Countries

2010 20301950

1

2

3

4

5

6

7

8

9

Source: Population Reference Bureau World Population Data Sheet 2012, page 4; data are drawn from United Nations Population Division, World Population Prospects: The 2010 Revision (2011), medium-variant estimates.

FIGuRE 6.1 World Population Growth, 1950–2050

288 PART TWO Problems and Policies: Domestic

World population distribution is put into dramatic perspective by the map in Figure 6.3. Attention is drawn to the large size of India in comparison with Europe. China is bordered on the north and west by a thin strip of land that represents Russia. Mexico looms very large in comparison with Canada—a dramatic reversal of conventional maps; taken together, even the Caribbean islands are larger than Canada. Bangladesh, smaller in size than the state of Wisconsin, is larger than Germany and France combined. In Africa, the prominence of Nigeria stands out. Indonesia, which gets comparatively little

Northern America 5%

Europe 11%

Latin America 8%

Africa 15%

Northern America 5%

Europe 8%

Latin America 7%

Africa 22%

(a) Total population 2010: 6.89 billion (b) Total population 2050: 9.5 billion

Asia and the Pacific 61%

Asia and the Pacific 58%

Source: Data from Population Reference Bureau, World Population DataSheet, 2010.

FIGuRE 6.2 World Population Distribution by Region, 2010 and 2050

Source: worldmapper.org:http://www.worldmapper.org/display.php?selected=2).

FIGuRE 6.3 Map with Country Sizes Proportional to Their Fraction of World Population

289CHAPTER 6 Population Growth and Economic Development

international attention, dwarfs its neighbor Australia while appearing nearly as large as the United States.

Fertility and Mortality Trends The rate of population increase is quanti- tatively measured as the percentage yearly net relative increase (or decrease, in which case it is negative) in population size due to natural increase and net international migration. Natural increase simply measures the excess of births over deaths or, in more technical terms, the difference between fertility and mortality. Net international migration is of very limited, though growing, importance today (although in the nineteenth and early twentieth centuries it was an extremely important source of population increase in North America, Australia, and New Zealand and corresponding relative decrease in west- ern Europe). Population increases in developing countries therefore depend almost entirely on the difference between their crude birth rates (or simply birth rates) and death rates.

Recall from Chapter 2 that most developing nations have birth rates rang- ing from 15 to 45 per 1,000. By contrast, in almost all developed countries, the rate is less than 15 per 1,000. Moreover, developing country birth rates today are still often higher than they were in preindustrial western Europe. But there has been a substantial decline in fertility over the past three decades, not only in countries like Taiwan, South Korea, and China, where rapid eco- nomic and social development have taken place, but also in nations where economic growth has been less rapid, including Mexico and Bangladesh, and in some where growth has stagnated, such as Zimbabwe. The total fertility rate (TFR)—the average number of children a woman would have, assuming that current age-specific birth rates remain constant throughout her childbear- ing years—has fallen dramatically in many countries since 1970, as the exam- ples in Table 6.2 demonstrate, but remains high in sub-Saharan Africa (5.1 in 2012) and western Asia (2.9). Niger with 7.1 and Afghanistan with 6.2 were among the highest in the world.3

Modern vaccination campaigns against malaria, smallpox, yellow fever, and cholera as well as the proliferation of public health facilities, clean water supplies, improved nutrition, and public education have all worked together over the past three decades to lower death rates by as much as 50% in parts of Asia and Latin America and by over 30% in much of Africa and the Middle East. Death rates have fallen for all age groups. Nevertheless, the average life span remains about 12 years greater in the developed countries. This gap has been sharply reduced in recent decades. For example, in 1950, life expectancy at birth for people in developing countries averaged 35 to 40 years, compared with 62 to 65 years in the developed world. Considerable progress has been made on reducing the under-5 mortality rate. For example, according to UN compilations between 1990 and 2008, it fell from 121 per 1,000 to 74 per 1,000 in South Asia, from 73 to 38 per 1,000 in Southeast Asia and from 52 to 23 per 1,000 in Latin America and the Caribbean. Although the under-5 mortality rate declined from 184 to 144 per 1,000 in sub-Saharan Africa in this period, progress in the region continued to lag. In 2009, because of still relatively high under-5 mortality rates and the AIDS epidemic, sub-Saharan Africa had the lowest life expectancy, 51 years, while in the high-income countries, life expec- tancy at birth averaged nearly 78 years. In East Asia and Latin America, life

Rate of population increase The growth rate of a population, calculated as the natural increase after adjusting for immigration and emigration.

Natural increase The differ- ence between the birth rate and the death rate of a given population.

Net international migration The excess of persons migrat- ing into a country over those who emigrate from that country.

Crude birth rate The num- ber of children born alive each year per 1,000 population (often shortened to birth rate).

Death rate The number of deaths each year per 1,000 population.

Total fertility rate (TFR) The number of chil- dren that would be born to a woman if she were to live to the end of her childbearing years and bear children in accordance with the prevail- ing age-specific fertility rates.

Life expectancy at birth The number of years a newborn child would live if subjected to the mortality risks prevail- ing for the population at the time of the child’s birth.

Under-5 mortality rate Deaths among children between birth and 5 years of age per 1,000 live births.

290 PART TWO Problems and Policies: Domestic

expectancies have now reached an impressive 74 and 73 years, respectively. Finally, note that there remains a biological susceptibility for old people to die at higher rates than young people due to aging. Although death rates of children and younger people are higher on average in a developing country with rapid population growth, the fact that their populations are so youthful on average explains why they may have an overall population-average death rate that is lower than that of a developed country with a much older average population. You may notice this possibly unexpected relationship when you look at demographic statistics.

Some of the striking population projections issued by the United Nations in 2013 are reported in Box 6.1.

Age Structure and Dependency Burdens Population is relatively youth- ful in the developing world. As of 2011, children under the age of 15 con- stitute more than 40% of the total population of the low-income countries, 32% of the lower-middle income countries, but just 17% of high-income countries.4 In countries with such an age structure, the youth dependency ratio—the proportion of youths (under age 15) to economically active adults (ages 15 to 64)—is very high. Thus, the workforce in developing countries must support almost twice as many children as it does in the wealthier countries. In the United Sates, the workforce age group (15 to 64) amounts to about 67% of the total population, with 20% under age 15 and 13% over age 65 as of 2011; the corresponding ratios in the United Kingdom are similar: 66%, 18%, and 17% respectively. In the euro area, some 19% of the population is over age 65; and in Japan nearly one-quarter of the popu- lation already has reached age 65. The main problems in more developed countries relate more to their low population growth and old-age depen- dents (over age 65). By contrast, in sub-Saharan Africa, the economically active workforce makes up about 54% of the total population (just 3% of the population is over age 65) as of 2011. In general, the more rapid the population growth rate is, the greater the proportion of dependent chil- dren in the total population and the more difficult it is for people who

Youth dependency ratio The proportion of young people under age 15 to the working population aged 16 to 64 in a country.

TABLE 6.2 Fertility Rate for Selected Countries, 1970 and 2009

Sources: World Bank, World Development Report, 1994 (New York: Oxford University Press, 1994), tab. 26; Population Reference Bureau, World Population Data Sheet (Washington, D.C.: Population Reference Bureau, 2012).

  Total Fertility Rate

Country 1970 2012

Bangladesh 7.0 2.3 Colombia 5.3 2.1 Indonesia 5.5 2.3 Jamaica 5.3 2.1 Mexico 4.9 2.3 Thailand 5.5 1.6 Zimbabwe 7.7 4.1

291CHAPTER 6 Population Growth and Economic Development

are working to support those who are not. This phenomenon of youth dependency also leads to an important concept, the hidden momentum of population growth.

The Hidden Momentum of Population Growth

Perhaps the least understood aspect of population growth is its tendency to continue even after birth rates have declined substantially. Population growth has a built-in tendency to continue, a powerful momentum that, like a speed- ing automobile when the brakes are applied, tends to keep going for some time before coming to a stop. In the case of population growth, this momen- tum can persist for decades after birth rates drop.

Hidden momentum of population growth The phenomenon whereby popu- lation continues to increase even after a fall in birth rates because the large existing youthful population expands the population’s base of potential parents.

BOX 6.1 FInDInGS The 2012 Revised United Nations Population Projections

Here is a summary of some of the main findings found in the UN’s World Population Prospects 2012 Revision, published in June 2013.

• World population is now projected at 8.1 bil- lion by 2025; and 9.6 billion by 2050.

• Most population growth will continue to oc- cur in developing regions where population will grow from 5.9 billion in 2013 to about 8.2 billion in 2050.

• “Give or take a billion”: The projections de- pend on assumptions—the 2050 population could turn out to be as little as 8.3 billion or as many as 10.9 billion.

• Most population growth will occur in Africa.

• The 49 least developed countries are pro- jected to double in size from 900 million in 2013 to 1.8 billion in 2050.

• Beyond Africa, projected population growth in the rest of world is just over 10% for 2013–2100.

• New projected total population is higher, particularly after 2075 because: • Current fertility level estimates are higher

in some countries with better information (in particular, in 15 high-fertility sub- Saharan African countries, estimated births

per woman were adjusted upwards more than 5%).

• In some cases, the actual level of fertility appears to have risen in recent years.

• In other cases, the previous estimate was too low.

• Other projections include: • Developed region population will be

little changed at 1.3 billion–even with immigration.

• India will become the world’s most popu- lous country, passing China around 2028, when each will have about 1.45 billion people.

• The population of Nigeria could pass that of the United States by 2050; by 2100 it could rival China as the second most popu- lous country.

• By 2100, several other countries are pro- jected to have populations over 200 mil- lion: Indonesia Tanzania, Pakistan, Congo, Ethiopia, Uganda, and Niger.

Source: United Nations Population Division, World Popu- lation Prospects: The 2012 Revision. New York: United Nations, Department of Economic and Social Affairs, 13 June 2013; downloaded from www.unpopulation.org. For a summary see http://www.un.org/apps/news/story. asp?NewsID=45165#.UlAkZmRVRz0.

292 PART TWO Problems and Policies: Domestic

There are two basic reasons for this. First, high birth rates cannot be altered substantially overnight. The social, economic, and institutional forces that have influenced fertility rates over the course of centuries do not simply evaporate at the urging of national leaders. We know from the experience of European nations that such reductions in birth rates can take many decades. Consequently, even if developing countries assign top priority to the limita- tion of population growth, it will still take many years to lower national fertil- ity to desired levels.

The second and less obvious reason for the hidden momentum of population growth relates to the age structure of many developing countries’ populations. Figure 6.4 illustrates the great difference between age structures in less developed and more developed countries by means of two population pyramids for 2010. Each pyramid rises by five-year age intervals for both males and females, with the total number in each age cohort measured on the horizontal axis. Panel A (the left and middle panels) show population pyramids for developed and developing countries, respectively (the age scale is that listed between these two figures). Expressed in millions of people, rather than percentages, the fig- ure clearly reveals that most future population growth will take place in the developing world. The steeper bottom rungs for the developing world as a whole, in contrast to a very low-income country such as Ethiopia (right panel),

Population pyramid A graphic depiction of the age structure of the population, with age cohorts plotted on the vertical axis and either population shares or numbers of males and females in each cohort on the horizontal axis.

Ethiopia: Rapid Growth

Age 100+ 95–99 90–94 85–89 80–84 75–79 70–74 65–69 60–64 55–59 50–54 45–49 40–44 35–39 30–34 25–29 20–24 15–19 10–14

5–9 0–4

10 10864202468 Share of population (%), 2005

Developing Countries

300 200 200100 100 3000 Population (millions), 2010

Male Female

85+ 80–84 75–79 70–74 65–69 60–64 55–59 50–54 45–49 40–44 35–39 30–34 25–29 20–24 15–19 10–14

5–9 0–4

Age

300 200 200100 100 3000 Population (millions), 2010

Developed Countries

Source: Graphs detailing Developed Countries from World Population Data Sheet of the Popula- tion Reference Bureau, Inc., by Population Reference Bureau. Copyright 2010 by Population Reference Bureau, Inc. Reproduced with permission of Population Reference Bureau, Inc., via Copyright Clearance Center. Graph detailing Ethiopia from Population Bulletin 62 (2007), p. 6. Reprinted with permission from Population Reference Bureau, Inc.

FIGuRE 6.4 Population Pyramids: All Developed and Developing Countries and the Case of Ethiopia

293CHAPTER 6 Population Growth and Economic Development

reflect the large declines in population growth in lower-middle income devel- oping countries over the past quarter century, and particularly in China (see the case study at the end of this chapter). For developed countries, in the con- temporary period the population in middle cohorts is typically greater than that of young cohorts; this is partly but certainly not exclusively viewed as a transitional feature of a period in which women have been delaying births until later in life.

From the Ethiopia pyramid (Panel B) expressed as share of population, young people greatly outnumber their parents (the age scale in this case is found to the right of the figure). When their generation reaches adulthood, the number of potential parents will inevitably be much larger than at present. It follows that even if these new parents have only enough children to replace themselves (two per couple, as compared with their parents, who may have had four or more children), the fact that the total number of couples having two children is much greater than the number of couples who previously had more children means that the total population will still increase substantially before leveling off.5

Panel A also focuses attention on the fact that some age brackets are increas- ing in size in some countries, while they are decreasing in others. This reflects that in the demographic transition, the fraction of the population of working age first rises and then falls. On the one hand, countries where the fraction of prime working-age citizens is rising face a potential crisis if many remain unemployed, as this is associated with inequality and (especially among males) social unrest, not to mention the potential output loss. On the other hand, this rise is also an important window of opportunity for strong income and productivity gains, referred to as the demographic dividend—a period in which there are fewer children to support, a larger fraction of women join or remain in the workforce for longer periods of time, and there are more avail- able resources to invest in human capital (see Chapter 8).

In contrast, where the fraction of people of working age is falling as a result of population aging, the resources needed for old-age support are increasing. This is already a challenge for most high-income countries. Leading up to this period, a higher savings rate is required; but then allowing more immigration can also help. The transition is likely to pose an even greater challenge for some middle-income countries with big drops in fertility ahead of previous historical patterns, most notably China (see the case study at the end of the chapter), but also in several other Asian countries.6

6.3 The Demographic Transition

The process by which fertility rates eventually decline to low and stable levels has been portrayed by a famous concept in economic demography called the demographic transition.

The demographic transition attempts to explain why all contemporary developed nations have more or less passed through the same three stages of modern population history. Before their economic modernization, these countries for centuries had stable or very slow-growing populations as a result of a combination of high birth rates and almost equally high death rates. This was stage 1. Stage 2 began when modernization, associated with better

Demographic transition The phasing-out process of popu- lation growth rates from a vir- tually stagnant growth stage, characterized by high birth rates and death rates through a rapid-growth stage with high birth rates and low death rates to a stable, low-growth stage in which both birth and death rates are low.

294 PART TWO Problems and Policies: Domestic

public health methods, healthier diets, higher incomes, and other improve- ments led to a marked reduction in mortality that gradually raised life expec- tancy from under 40 years to over 60 years. However, the decline in death rates was not immediately accompanied by a decline in fertility. As a result, the growing divergence between high birth rates and falling death rates led to sharp increases in population growth compared to past centuries. Stage 2 thus marks the beginning of the demographic transition (the transition from stable or slow-growing populations first to rapidly increasing numbers and then to declining rates). Finally, stage 3 was entered when the forces and influences of modernization and development caused the beginning of a decline in fertility; eventually, falling birth rates converged with lower death rates, leaving little or no population growth.

This process implies movement from a relatively high number of births per woman to a population replacement fertility level that can be calculated to reach about 2.05 to 2.1 births per woman when nearly all women survive to the mean age of childbearing, as they do in developed countries. In develop- ing countries with much lower survival rates, replacement fertility can be well over 3 births per woman.7

Figure 6.5 depicts the three historical stages of the demographic transition in western Europe. Before the early nineteenth century, birth rates hovered around 35 per 1,000, while death rates fluctuated around 30 per 1,000. This resulted in population growth rates of around 5 per 1,000, or less than 0.5% per year. Stage 2, the beginning of western Europe’s demographic transition, was initiated around the first quarter of the nineteenth century by slowly fall- ing death rates as a result of improving economic conditions and the gradual development of disease and death control through modern medical and pub- lic health technologies. The decline in birth rates (stage 3) did not really begin until late in the nineteenth century, with most of the reduction many decades occurring after modern economic growth had begun and long after death rates began their descent. But since the initial level of birth rates was generally low

Replacement fertility The number of births per woman that would result in stable population levels.

1780 1840 1890 1910 0

10

20

40

Year

A ct

u al

b ir

th a

n d

d ea

th r

at es

p er

1 ,0

00 in

h ab

it an

ts

Birth rate Death rate

20101850

30

Stage 1

Future

Stage 2 Stage 3

FIGuRE 6.5 The Demographic Transition in Western Europe

295CHAPTER 6 Population Growth and Economic Development

in western Europe as a result of either late marriage or celibacy, overall rates of population growth seldom exceeded the 1% level, even at their peak. By the end of western Europe’s demographic transition in the second half of the twentieth century, the relationship between birth and death rates that marked the early 1800s had reversed, with birth rates fluctuating and death rates remaining fairly stable or rising slightly. This latter phenomenon was simply due to the older age distributions of contemporary European populations. The patterns of the demographic transition in Europe are clear, though research continues to better identify the causal factors at work.8

Figure 6.6 shows the population histories of contemporary developing countries, which contrast with those of western Europe and fall into two patterns.

Birth rates in many developing countries today are considerably higher than they were in preindustrial western Europe. This is because women tend to marry at an earlier age. As a result, there are both more families for a given population size and more years in which to have children. In the 1950s and 1960s, stage 2 of the demographic transition occurred throughout most of the developing world. The application of highly effective imported modern medical and public health technologies caused death rates in developing countries to fall much more rap- idly than in nineteenth-century Europe. Given their historically high birth rates (still over 35 per 1,000 in many countries), this has meant that stage 2 of the demographic transition has been characterized by peak population growth rates well in excess of 2.0% per annum in most developing countries.

With regard to stage 3, we can distinguish between two broad classes of developing countries. In case A in Figure 6.6, modern methods of death

1900 1950

Case B

Case A

0

10

20

40

A ct

u al

b ir

th a

n d

d ea

th r

at es

p er

1 ,0

00 in

h ab

it an

ts

Death rate

Birth rate

Case B

1970 2000

30

Stage 1

Future

Stage 2 Stage 3

50

Case A

Year

FIGuRE 6.6 The Demographic Transition in Developing Countries

Source: Based on National Academy of Sciences, The Growth of World Population (Washington, D.C.: National Academy of Sciences, 1963), p. 15.

296 PART TWO Problems and Policies: Domestic

control, combined with rapid and widely distributed rises in levels of living, have resulted in death rates falling as low as 10 per 1,000 and birth rates also falling rapidly, to levels between 12 and 25 per 1,000. These countries, includ- ing Taiwan, South Korea, Costa Rica, China, Cuba, Chile, and Sri Lanka, have thus entered stage 3 of their demographic transition and have experienced rapidly falling rates of overall population growth.

But some developing countries fall into case B of Figure 6.6. After an initial period of rapid decline, death rates have failed to drop further, largely because of the persistence of widespread absolute poverty and low levels of living and more recently because of the AIDS epidemic. Moreover, the continuance of still quite high birth rates as a result of these low levels of living causes overall population growth rates to remain relatively high. These countries, including many of those in sub-Saharan Africa and the Middle East, are still in stage 2 of their demographic transition. Though fertility is declining, it remains very high in these parts of the world.

The important question, therefore, is this: When and under what con- ditions are developing nations likely to experience falling birth rates and a slower expansion of population? To answer this question, we need to ask a prior one. What are the principal determinants or causes of high fertility rates in developing countries, and can these determinants of the “demand” for children be influenced by government policy? To try to answer this criti- cal question, we turn to a very old and famous classical macroeconomic and demographic model, the Malthusian “population trap,” and a contemporary and highly influential neoclassical microeconomic model, the household the- ory of fertility.

6.4 The Causes of High Fertility in Developing Countries: The Malthusian and Household Models

The Malthusian Population Trap

More than two centuries ago, the Reverend Thomas Malthus put forward a theory of the relationship between population growth and economic devel- opment that is influential today. Writing in his 1798 Essay on the Principle of Population and drawing on the concept of diminishing returns, Malthus pos- tulated a universal tendency for the population of a country, unless checked by dwindling food supplies, to grow at a geometric rate, doubling every 30 to 40 years.9 At the same time, because of diminishing returns to the fixed factor, land, food supplies could expand only at a roughly arithmetic rate. In fact, as each member of the population would have less land to work, his or her marginal contribution to food production would actually start to decline. Because the growth in food supplies could not keep pace with the burgeon- ing population, per capita incomes (defined in an agrarian society simply as per capita food production) would have a tendency to fall so low as to lead to a stable population existing barely at or slightly above the subsistence

297CHAPTER 6 Population Growth and Economic Development

level. Malthus therefore contended that the only way to avoid this condition of chronic low levels of living or absolute poverty was for people to engage in “moral restraint” and limit the number of their progeny. Hence, we might regard Malthus, indirectly and inadvertently, as the father of the modern birth control movement.

Modern economists have given a name to the Malthusian idea of a popula- tion inexorably forced to live at subsistence levels of income. They have called it the low-level equilibrium population trap or, more simply, the Malthusian population trap. Diagrammatically, the basic Malthusian model can be illus- trated by comparing the shape and position of curves representing population growth rates and aggregate income growth rates when these two curves are each plotted against levels of per capita income. An example of this is pre- sented in Figure 6.7.

On the vertical axis, we plot numerical percentage changes, both positive and negative, in the two principal variables under consideration (total pop- ulation and aggregate income). On the horizontal axis are levels of per cap- ita income. Figure 6.7 depicts the basic ideas. The x-axis shows the level of income per capita. The y-axis shows two rates—of population growth and of total income growth. Per capita income growth is, by definition, the difference between income growth and population growth—hence the vertical difference between these two curves. Thus, as we saw in Chapter 3 in our discussion of the Harrod-Domar (or AK) model, whenever the rate of total income growth is greater than the rate of population growth, income per capita is rising; this corresponds to moving to the right along the x-axis. Conversely, whenever the rate of total income growth is less than the rate of population growth, income per capita is falling, moving to the left along the x-axis. When these rates are equal, income per capita is unchanging. We can then explore the shapes of population growth and growth of income to understand potential implica- tions of this relationship.

Malthusian population trap The threshold population level anticipated by Thomas Malthus (1766–1834) at which population increase was bound to stop because life- sustaining resources, which increase at an arithmetic rate, would be insufficient to support human population, which would increase at a geometric rate.

Income per capita

G ro

w th

r a

te s

S T

Total income growth rate (∆Y/Y)

Population growth rate (∆P/P)

FIGuRE 6.7 The Malthusian Population Trap

298 PART TWO Problems and Policies: Domestic

First consider population growth. When income is very low, say, below $250 per year at purchasing power parity, nutrition is so poor that people become susceptible to fatal infectious diseases; pregnancy and nursing become problematic; and, ultimately, outright starvation may occur. This is shown on the left in Figure 6.7. But after this minimum level of income per capita is reached, population begins to grow, eventually reaching a peak rate (perhaps at 3% to 4% per year); and then the population growth rate begins to fall until at last a fairly stable population is reached (a growth rate close to zero). Note that this pattern of population growth first increasing and then decreasing as per capita income rises corresponds to the pattern of the demographic transi- tion, explained in section 6.3.

In Figure 6.7, total income growth becomes greater as the economy devel- ops (and income per capita rises). An economic reason for this positive rela- tionship is the assumption that savings vary positively with income per capita. Countries with higher per capita incomes are assumed to be capable of generating higher savings rates and thus more investment. Again, given a Harrod-Domar-type model of economic growth (see Chapter 3), higher sav- ings rates mean higher rates of aggregate income growth. Eventually, how- ever, growth levels off at a maximum. (Incomes of middle-income countries might grow fastest as they borrow technology to catch up—not shown in this diagram—but these higher rates cannot be continued once the technology frontier is reached.)

As drawn, the curves first cross at a low level of income, labeled S (for subsistence). This is a stable equilibrium: If per capita income levels become somewhat larger than (to the right of) S, it is assumed that population size will begin to increase in part because higher incomes improve nutrition and reduce death rates. But then, as shown in the figure, population is growing faster than income (the ΔP/P curve is vertically higher than the ΔY/Y curve), so income per capita is falling, and we move to the left along the x-axis. The arrow pointing in the direction of S from the right therefore shows per capita income falling back to this very low level. On the other hand, if income per capita were a little less than S, the total income curve would be above the population growth curve and so income per capita would be rising. This cor- responds to a move to the right along the x-axis. Thus, our conclusion is that point S represents a stable equilibrium (much as in our study of stable equilib- ria in Figure 4.1). This very low population growth rate along with a very low income per person is consistent with the experience of most of human history prior to the modern era.10

According to modern-day neo-Malthusians, poor nations will never be able to rise much above their subsistence levels of per capita income unless they initiate preventive checks (birth control) on their population growth. In the absence of such preventive checks, Malthusian positive checks (star- vation, disease, wars) on population growth will inevitably provide the restraining force. However, if per capita income can somehow reach a thresh- old level, labeled T in Figure 6.7, from that point population growth is less than total income growth, and thus per capita income grows continually, at a rate such as 2% per year (the approximate U.S. per capita growth rate from 1870 to 2010).

299CHAPTER 6 Population Growth and Economic Development

Countries or regions in such a population trap can also escape it by achiev- ing technological progress that shifts the income growth rate curve up at any level of per capita income. And it may be able to achieve changes in economic institutions and culture (“social progress”) that shifts the population growth curve down. In this way, the population trap equilibrium is eliminated alto- gether, and the economy is able to proceed with self-sustaining growth. An example of such a result is depicted in Figure 6.8. Total income growth is now greater than population growth at each level of per capita income. As a result, income per capita now grows steadily.

We have examined strategies for accelerating income growth in Chap- ters 3 (including its appendices) and 4, and we will examine specific growth policies further in Chapters 7, 9, 12, and 14. The main focus of the remain- der of this chapter is on changes in economic institutions, economic power in households, and cultural norms, to reduce fertility to maintain popula- tion growth below income growth, and eventually to achieve population stability.

In addition to the classic Malthusian model, the multiple equilibrium analysis of Chapter 4, Figure 4.1, is also relevant to understanding high-fertility traps. In the diagram, we can take the x-axis to represent (expected) fertility and the y-axis, the family’s own fertility decision. The upward-sloping response (along the S-shaped curve) of the individual family fertility decision to aver- age fertility may be caused by at least two important complementarities—a basis for possible multiple equilibria. First, if others have high fertility, this may increase the number of formal-sector job seekers without (proportionally) increasing the number of (higher-paying) formal-sector jobs. Each family may feel it needs a larger number of children to raise the probability that at least

0

1

3

5

Income per capita (Y/P)

G ro

w th

r a

te (

% )

Population growth rate (∆P/P)2

–1

4 Total income growth rate (∆Y/Y)

FIGuRE 6.8 How Technological and Social Progress Allows nations to Avoid the Population Trap

300 PART TWO Problems and Policies: Domestic

one child will get a modern job. In addition, families often follow local social norms about fertility and tend to model their own behavior on the behavior of others in their community.

It is plausible that the resulting positively sloped response curve also has an S-shape, similar to the one in Figure 4.110a. If the fertility response curve cuts the 45-degree line from above at least twice, then there are at least two stable equilibria (see Chapter 4, section 4.2): one with high and another with low levels of average fertility.11 Some findings on the effects of changing norms on fertility decisions is presented in Box 6.2.

BOX 6.2 FInDInGS Social Norms and the Changing Patterns of Fertility in Bangladesh

In this chapter, we describe an idea—presented in part by Partha Dasgupta—that social norms play a role in setting an equilibrium fertility rate: If families followed local customs about fertility—modeling their own behavior on that of their neighbors—the com- munity might be trapped at a higher fertility rate than would prevail if they could manage a change in so- cial expectations. The idea was also a starting point for empirical research by Kaivan Munshi and Jacques Myaux on the uneven transition to lower fertility in rural developing areas.

Munshi and Myaux applied their research to the experience of the Matlab area of Bangladesh. Fertil- ity reduction varied greatly across apparently similar villages. In addition, response to the same family- planning program also varied greatly in the magni- tude of their effects and time lags before these effects were realized. Data on fertility collected twice annu- ally over an 11-year period offered a unique chance to learn about this process. (The data set included con- traceptive use and demographic and socioeconomic characteristics for all women living in all 70 villages in the Matlab area who took part in the program and were followed throughout the 11-year period.)

Munshi and Myaux offered an explanation for widely varying local patterns: “Most societies have traditionally put norms into place to regulate fertility. When the economic environment changes, individu- als gradually learn through their social interactions

about the new reproductive equilibrium that will emerge in their community.” In this case, the change was in the availability of modern contraception. There is likely some proportion of people who will be perpetually resistant to contraception; the remainder will be open-minded about it but may not want to behave too differently than local norms dictate. Un- til this process plays out, people will not know how many of their neighbors will be firmly resistant to change and thus whether contraceptive use will ul- timately be socially acceptable overall. Munshi and Myaux propose that families’ uncertainty about what potential new equilibrium (what level of contracep- tive prevalence) in their villages will emerge leads to caution, giving rise to slow and different rates of fertility transition in otherwise apparently similar vil- lages. They developed a model to demonstrate the underlying logic of this explanation and concluded that social norms do make a difference; the process of moving to a better equilibrium can be slow. In some cases, movement out of the high fertility equilibrium (too high for many who are stuck there) can be pre- vented indefinitely.

In rural Bangladesh, which has a large majority Muslim population but also a minority Hindu popu- lation, social norms correspond to religious groups. Women are secluded generally (through purdah) and almost never interact with anyone (including women) from another religious group.

301CHAPTER 6 Population Growth and Economic Development

In this context, the researchers studied an “exoge- nous economic intervention”—a thorough, long-term family-planning program introduced throughout the village areas, studied and promoted door to door to each religious group with equal intensity. This is the kind of quasi experiment needed to understand the effects of social interactions, a process of wide im- portance in development economics and one that presents great challenges for econometrics (statistical analysis). The authors examined the data and showed that a woman’s contraception use “respond[s] strongly to contraceptive prevalence within their own religious group in the village, cross-religion effects are entirely absent in the data.” This held despite the fact that “all individuals in the village have access to the same family-planning inputs” and even when the people are otherwise very similar. Thus, the findings are “con- sistent with the view that changing social norms are driving changes in reproductive behavior in these communities. “As in the model, uncertainty about the ultimate prevalence of contraception use “is slowly

resolved over time as women in the village interact sequentially with each other from one period to the next, which explains the gradual change in contracep- tive prevalence that we see in the data, as well as the convergence to different levels of contraceptive use across communities.”

As societies gain the possibility of modern eco- nomic development, advantages of smaller family sizes grow both for families and for the societies of which they are a part. But multiple equilibria are pos- sible. Many in communities with full knowledge of and access to contraception may still perpetuate high fertility rates when social norms and sanctions to contrary behavior prevail. Addressing situations like these requires attention to social aspects of the devel- opment process.

Source: Kaivan Munshi and Jacques Myaux, “Social norms and the fertility transition,” Journal of Development Eco- nomics 80 (2006): 1–38. For further background on the issues involved, see also Partha Dasgupta, An Inquiry into Well-Being and Destitution (New York: Oxford University Press, 1993).

Criticisms of the Malthusian Model

The Malthusian population trap provides a theory of the relationship between population growth and economic development. Unfortunately, it is based on a number of simplistic assumptions and hypotheses that do not stand the test of empirical verification. We can criticize the population trap on two major grounds.

First, the model ignores the enormous impact of technological progress in offsetting the growth-inhibiting forces of rapid population increases. As we saw in Chapter 2, the history of modern economic growth has been closely associated with rapid technological progress in the form of a continuous series of scientific, technological, and social inventions and innovations. Increasing rather than decreasing returns to scale have been a distinguishing feature of the modern growth epoch. While Malthus was basically correct in assum- ing a limited supply of land, he did not—and in fairness could not at that time—anticipate the manner in which technological progress could augment the availability of land by raising its quality (its productivity) even though its quantity might remain roughly the same.

In terms of the population trap, rapid and continuing technological prog- ress can be represented by an upward shift of the income growth (total prod- uct) curve so that at all levels of per capita income, it is vertically higher than

302 PART TWO Problems and Policies: Domestic

the population growth curve. This is shown in Figure 6.8. As a result, per cap- ita income will continue to grow over time. All countries therefore have the potential of escaping the Malthusian population trap.

The second basic criticism of the trap focuses on its assumption that national rates of population increase are directly (positively) related to the level of national per capita income. According to this assumption, at rela- tively low levels of per capita income, we should expect to find population growth rates increasing with increasing per capita income. But research indi- cates that there appears to be no clear correlation between population growth rates and levels of per capita income. As a result of modern medicine and public health programs, death rates have fallen rapidly and have become less dependent on the level of per capita income. Moreover, birth rates seem to show no rigid relationship with per capita income levels. Fertility rates vary widely for countries with the same per capita income, especially below $1,000. It is not so much the aggregate level of per capita income that mat- ters for population growth but rather how that income is distributed. It is the level of household income, not the level of per capita income, that seems to matter most.

In sum, Malthusian and neo-Malthusian theories as applied to contem- porary developing nations have severely limited relevance for the following reasons:

1. They do not take adequate account of the role and impact of technological progress.

2. They are based on a hypothesis about a macro relationship between popu- lation growth and levels of per capita income that does not stand up to empirical testing of the modern period.

3. They focus on the wrong variable, per capita income, as the principal determinant of population growth rates. A much better and more valid approach to the question of population and development centers on the microeconomics of family size decision making in which individual, and not aggregate, levels of living become the principal determinant of a family’s decision to have more or fewer children.

We continue to study the Malthusian trap even though evidence shows that it is not currently relevant for three main reasons: First, because many people still believe it holds in poor countries today, despite the recent evidence; and people working in the development field should understand the model and the elements of it that do not currently apply so that they can engage the debate effectively. Second, because it seems clear that such traps have occurred in the historical past and may have been factors in population collapses, includ- ing in the pre-Columbian Americas. Third—as we will explore in the remain- der of this chapter—the fact that this model no longer applies underlines the importance of factors that can prevent its emergence. These include efforts to continue steady and sustainable rises in agricultural productivity; moreover, they encompass increases in women’s empowerment and freedom to choose— along with their incomes—which reduce the old-age security motive behind high fertility.

303CHAPTER 6 Population Growth and Economic Development

The Microeconomic Household Theory of Fertility

In recent years, economists have begun to look more closely at the microeco- nomic determinants of family fertility in an attempt to provide a better theo- retical and empirical explanation for the observed falling birth rates associated with stage 3 of the demographic transition. In doing this, they have drawn on the traditional neoclassical theory of household and consumer behavior for their basic analytical model and have used the principles of economics and optimization to explain family size decisions.

The conventional theory of consumer behavior assumes that an individ- ual with a given set of tastes or preferences for a range of goods (a “utility function”) tries to maximize the satisfaction derived from consuming these goods subject to his or her own income constraint and the relative prices of all goods. In the application of this theory to fertility analysis, children are considered as a special kind of consumption (and in developing coun- tries, particularly low-income countries, investment) good so that fertility becomes a rational economic response to the consumer ’s (family’s) demand for children relative to other goods. The usual income and substitution effects are assumed to apply. That is, if other factors are held constant, the desired number of children can be expected to vary directly with household income (this direct relationship may not hold for poor societies; it depends on the strength of demand for children relative to other consumer goods and to the sources of increased income, such as female employment), inversely with the price (cost) of children, and inversely with the strength of tastes for other goods relative to children. Mathematically, these relationships can be expressed as follows:

Cd = f (Y, Pc , Px , tx), x = 1, g, n (6.1)

where Cd, the demand for surviving children (an important consideration in low-income societies where infant mortality rates are high), is a function of the given level of household income (Y), the “net” price of children (the difference between anticipated costs, mostly the opportunity cost of a mother ’s time, and benefits, potential child income and old-age support, Pc), the prices of all other goods (Px), and the tastes for goods relative to children (tx). Under standard neoclassical conditions, we would expect the following (expressed both math- ematically and in words):

0 Cd /0 Y > 0 The higher the household income, the greater the demand for children.

0 Cd /0 Pc < 0 The higher the net price of children, the lower the quantity demanded.

0 Cd /0 Px > 0 The higher the prices of all other goods relative to children, the greater the quantity of children demanded.

0 Cd /0 tx < 0 The greater the strength of tastes for goods relative to children, the fewer children demanded.

Figure 6.9 provides a simplified diagrammatic presentation of the microeco- nomic theory of fertility. The number of desired (surviving) children, Cd, is

Microeconomic theory of fertility The theory that family formation has costs and benefits that determine the size of families formed.

304 PART TWO Problems and Policies: Domestic

measured along the horizontal axis, and the total quantity of goods consumed by the parents, Gp, is measured on the vertical axis.

Household desires for children are expressed in terms of an indifference map representing the subjective degree of satisfaction derived by the parents for all possible combinations of commodities and children. Each individual indifference curve portrays a locus of commodity-child combinations that yield the same amount of satisfaction. Any point (or combination of goods and chil- dren) on a “higher” indifference curve—that is, on a curve farther out from the origin—represents a higher level of satisfaction than any point on a lower indif- ference curve. But each indifference curve is a “constant satisfaction” locus.

In Figure 6.9, only four indifference curves, I1 to I4, are shown; in theory, there is an infinite set of such curves, filling the whole quadrant and covering all possible commodity-child combinations. The household’s ability to “pur- chase” alternative combinations of goods and children is shown by the budget constraint line, ab. Thus, all combinations on or below line ab (within the tri- angular area 0ab) are financially attainable by the household on the basis of its perceived income prospects and the relative prices of children and goods, as represented by the slope of the ab budget constraint. The steeper the slope of the budget line, the higher the price of children relative to goods.

According to the demand-based theory of fertility, the household chooses from among all attainable combinations the one combination of goods and children that maximizes family satisfaction on the basis of its subjectively determined preferences. Diagrammatically, this optimal combination is rep- resented by point f, the tangency point between the budget constraint, ab, and indifference curve, I2. Therefore, C3 children and G2 goods will be demanded.

c

a

e

0

Number of children desired (Cd)

G o

o d

s co

n su

m e

d b

y p

a re

n ts

( G

p )

g

h

f

G3

G2

G1

C1 C2 C3

I1

I2 I3

I4

a′

b″ b′bd

FIGuRE 6.9 Microeconomic Theory of Fertility: An Illustration

305CHAPTER 6 Population Growth and Economic Development

A rise in family income, represented in Figure 6.9 by the parallel outward shift of the budget line from ab to a′b′, enables the household to attain a higher level of satisfaction (point h on curve I4) by consuming more of both commodi- ties and children—that is, if children, like most commodities, are assumed to be normal goods (demand for them rises with income), an important if in low- income countries where children are often in demand primarily as a source of future financial security. Note that as income rises, parents may spend more on each child, preferring a smaller number of children, each of higher “qual- ity,” for example, healthier and better educated.

Similarly, an increase in the price (opportunity cost) of children relative to other goods will cause households to substitute commodities for children. Other factors (namely, income and tastes) being constant, a rise in the relative price of children causes the household utility-maximizing consumption combi- nation to occur on a lower indifference curve, as shown by the movement of the equilibrium point from f to e when the budget line rotates around point a to ab″.

Note, finally, that if there is a simultaneous increase in household income and net child price as a result of, say, expanding female employment opportunities and a rise in wages, coupled with a tax on children beyond a certain number per family, there will be both an outward shift and downward rotation of the budget constraint line of Figure 6.9 to, say, dashed line cd. The result is a new util- ity-maximizing combination that includes fewer children per family (point g compared with point f ). In other words, higher levels of living for low-income families in combination with a relative increase in the price of children (whether brought about directly by fiscal measures or indirectly by expanded female employment opportunities) will motivate households to have fewer children while still improving their welfare. This is just one example of how the eco- nomic theory of fertility can shed light on the relationship between economic development and population growth as well as suggest possible lines of policy.

The Demand for Children in Developing Countries

The economic theory of fertility assumes that the household demand for children is determined by family preferences for a certain number of surviving (usually male) children (i.e., in regions of high mortality, parents may produce more chil- dren than they actually desire in the expectation that some will not survive), by the price or “opportunity cost” of rearing these children, and by levels of family income. Children in poor societies are seen partly as economic investment goods in that there is an expected return in the form of both child labor and the provi- sion of financial support for parents in old age.12 However, in many developing countries, there is a strong intrinsic psychological and cultural determinant of family size, so the first two or three children should be viewed as “consumer” goods for which demand may not be very responsive to relative price changes.

The choice mechanism in the economic theory of fertility as applied to developing countries is assumed, therefore, to exist primarily with regard to the additional (“marginal”) children who are considered as investments. In deciding whether or not to have additional children, parents are assumed to weigh private economic benefits against private costs, where the principal benefits are the expected income from child labor, usually on the farm, and eventual financial support for elderly parents. Balanced against these benefits are the two principal elements of cost: the opportunity cost of the mother ’s

306 PART TWO Problems and Policies: Domestic

time (the income she could earn if she were not at home caring for her children) and the cost of educating children—the financial trade-off between having fewer “high-quality,” high-cost, educated children with high-income- earning potential versus more “low-quality,” low-cost, uneducated children with much lower earning prospects.

Using the same thought processes as in the traditional theory of consumer behavior, the theory of family fertility concludes that when the price or cost of children rises as a result of, say, increased educational and employment opportunities for women or a rise in school fees or the establishment of mini- mum-age child labor laws or the provision of publicly financed old-age social security schemes, parents will demand fewer additional children, substitut- ing, perhaps, quality for quantity or a mother ’s employment income for her child-rearing activities. It follows that one way to induce families to desire fewer children is to raise the price of child rearing by, say, providing greater educational opportunities and a wider range of higher-paying jobs for young women.

Recent research on household behavior has led to a major improvement of this theory. Households in developing countries generally do not act in a “uni- tary” manner, depicted with this traditional model. Instead, men and women have different objective functions; for example, husbands may prefer to have more children than wives. Household behavior is then explained as a result of bargaining between husbands and wives. Although the broad impacts we have just described continue to hold, the process includes increased bargain- ing power of women. Nonunitary, bargaining-based models of household behavior also improve our understanding of otherwise puzzlingly inefficient household behaviors, such as higher investment in husbands’ farm plots than wives’ farm plots even when a more even investment could lead to higher family incomes.13

Some Empirical Evidence Statistical studies in a broad spectrum of developing countries have provided support for the economic theory of fertil- ity.14 For example, it has been found that high female employment opportuni- ties outside the home and greater female school attendance, especially at the primary and secondary levels, are associated with significantly lower levels of fertility. As women become better educated, they tend to earn a larger share of household income and to produce fewer children. Moreover, these studies have confirmed the strong association between declines in child mortality and the subsequent decline in fertility. Assuming that households desire a target number of surviving children, increased female education and higher levels of income can decrease child mortality and therefore increase the chances that the firstborn will survive. As a result, fewer births may be necessary to attain the same number of surviving children. This fact alone underlines the impor- tance of educating women and improving public health and child nutrition programs in reducing fertility levels.

Implications for Development and Fertility

All of the foregoing can be summarized by saying that the effect of social and economic progress in lowering fertility in developing countries will be

307CHAPTER 6 Population Growth and Economic Development

the greatest when the majority of the population and especially the very poor share in its benefits. Specifically, birth rates among the very poor are likely to fall where the following socioeconomic changes come to pass:

1. An increase in the education of women and a consequent improvement in their role and status

2. An increase in female nonagricultural wage employment opportunities, which raises the price or cost of their traditional child-rearing activities

3. A rise in family income levels through the increased direct employment and earnings of a husband and wife or through the redistribution of income and assets from rich to poor

4. A reduction in infant mortality through expanded public health pro- grams and better nutritional status for both mother and child, and better medical care

5. The development of old-age and other social security systems outside the extended family network to lessen the economic dependence of parents, especially women, on their offspring

6. Expanded schooling opportunities so that parents can better substitute child “quality” for large numbers of children

In short, expanded efforts to make jobs, education, and health more broadly available to poverty groups in general and women in particular will not only contribute to their economic and psychic well-being (i.e., to their development) but also contribute substantially to their motivation for smaller families (i.e., their freedom to choose), which is vital to reducing population growth rates. Where such motivation exists, well-executed family-planning programs can then be an effective tool.15 But before discussing policy issues and what government might or might not do, we should point out that while there seems to be considerable agreement regarding the determinants or causes of population growth, substantial disagreement and controversy remain regarding its consequences.

6.5 The Consequences of High Fertility: Some Conflicting Perspectives

For many years, development economists and other social scientists have debated the seriousness of the consequences of rapid population growth.16 On the one hand, we must recognize that population growth is not the only, or even the primary, source of low levels of living, eroding self-esteem, and limited freedom in developing nations. On the other hand, it would be equally naive to think that rapid population growth in many countries and regions is not a serious intensifier and multiplier of those integral components of underdevel- opment, especially the first and third. The following discussion summarizes some of the main arguments for and against the idea that the consequences of rapid population growth lead to serious development problems. It then

Family-planning programs Public programs designed to help parents plan and regulate their family size.

308 PART TWO Problems and Policies: Domestic

considers whether some consensus can be reached so that specific policy goals and objectives can be postulated.17

It’s Not a Real Problem

We can identify three general lines of argument on the part of people who assert that population growth is not a cause for concern:

• The problem is not population growth but other issues. • Population growth is a false issue deliberately created by dominant rich-

country agencies and institutions to keep developing countries in their dependent condition.

• For many developing countries and regions, population growth is in fact desirable.

Other Issues Many observers from both rich and poor nations argue that the real problem is not population growth per se but one or all of the following four issues.

1. Underdevelopment. If correct strategies are pursued and lead to higher levels of living, greater self-esteem, and expanded freedom, population will take care of itself. Eventually, it will disappear as a problem, as it has in all of the present economically advanced nations. According to this ar- gument, underdevelopment is the real problem, and development should be the only goal. With it will come economic progress and social mech- anisms that will more or less automatically regulate population growth and distribution. As long as people in developing countries remain im- poverished, uneducated, and unhealthy and the social safety net remains weak, the large family will constitute the only real source of social secu- rity (i.e., parents will continue to be denied the freedom to choose a small family if they so desire). Some proponents of the underdevelopment ar- gument then conclude that birth control programs will surely fail, as they have in the past, when there is no motivation on the part of poor families to limit their size.

2. World Resource Depletion and Environmental Destruction. Population can only be an economic problem in relation to the availability and uti- lization of scarce natural and material resources. The fact is that devel- oped countries, with less than one-quarter of the world’s population, consume almost 80% of the world’s resources. In terms of the depletion of the world’s limited resources, therefore, the addition of another child in the developed countries is as significant as the birth of many times as many additional children in the underdeveloped countries. Accord- ing to this argument, developed nations should curtail their excessively high consumption standards instead of asking less developed nations to restrict their population growth. The latter ’s high fertility is really due to their low levels of living, which are in turn largely the result of the over- consumption of the world’s scarce resources by rich nations. This com- bination of rising affluence and extravagant consumption habits in rich

309CHAPTER 6 Population Growth and Economic Development

countries and among rich people in poor countries, and not population growth, should be the major world concern. We will analyze issues of the environment and development in Chapter 10.

3. Population Distribution. According to this third argument, it is not the number of people per se that is causing population problems but their distribution in space. Many regions of the world (e.g., parts of sub-Saha- ran Africa) and many regions within countries (e.g., the northeastern and Amazon regions of Brazil) are viewed as underpopulated in terms of avail- able or potential resources. Others simply have too many people concen- trated in too small an area (e.g., central Java or most urban concentrations). Governments should therefore strive not to moderate the rate of popula- tion growth but rather to bring about a more natural spatial distribution of the population in terms of available land and other productive resources.

4. Subordination of Women. Perhaps most important, as noted previously, women often bear the disproportionate burdens of poverty, poor educa- tion, and limited social mobility. In many cases, their inferior roles, low status, and restricted access to birth control are manifested in their high fertility. According to this argument, population growth is a natural out- come of women’s lack of economic opportunity. If women’s health, edu- cation, and economic well-being are improved along with their role and status in both the family and the community, this empowerment of women will inevitably lead to smaller families and lower population growth.

It’s a Deliberately Contrived False Issue

The second main line of argument denying the significance of population growth as a major development problem is closely allied to the neocolonial dependence theory of underdevelopment discussed in Chapter 3. Basically, it is argued that the overconcern in the rich nations with the population growth of poor nations is really an attempt by the former to hold down the devel- opment of the latter in order to maintain an international status quo that is favorable to the rich nations’ self-interests. Rich nations are pressuring poor nations to adopt aggressive population control programs, even though they themselves went through a period of sizable population increase that acceler- ated their own development processes.

A radical neo-Marxist version of this argument views population control efforts by rich countries and their allied international agencies as racist or genocidal attempts to reduce the relative or absolute size of the poor, largely nonwhite populations of the world who may someday pose a serious threat to the welfare of the rich, predominantly white societies. Worldwide birth con- trol campaigns are seen as manifestations of the fears of the developed world in the face of a possible radical challenge to the international order by the peo- ple who are its first victims.

It’s a Desirable Phenomenon

A more conventional economic argument is that of population growth as an essential ingredient to stimulate economic development. Larger populations

310 PART TWO Problems and Policies: Domestic

provide the needed consumer demand to generate favorable economies of scale in production, to lower production costs, and to provide a sufficient and low-cost labor supply to achieve higher output levels. Population “revi- sionist” economists of the neoclassical counterrevolution school argue, for example, that free markets will always adjust to any scarcities created by pop- ulation pressures.18 Such scarcities will drive up prices and signal the need for new cost-saving production technologies. In the end, free markets and human ingenuity (Julian Simon’s “genius” as the “ultimate resource") will solve any and all problems arising from population growth. This revisionist viewpoint was clearly in contrast with the traditional “orthodox” argument that rapid population growth had serious economic consequences that, if left uncor- rected, would slow economic development.

At the other end of the political spectrum, it has been argued by some developing-world neo-Marxist pronatalists that many rural regions in devel- oping countries are in reality underpopulated in the sense that much unused but arable land could yield large increases in agricultural output if only more people were available to cultivate it. Many regions of tropical Africa and Latin America and even parts of Asia are said to be in this situation. With respect to Africa, for example, some observers have noted that many regions had larger populations in the remote past than after independence.19 Their rural depopu- lation resulted not only from the slave trade but also from compulsory military service, confinement to reservations, and the forced-labor policies of former colonial governments. For example, the sixteenth-century Kongo kingdom is said to have had a population of approximately 2 million. But by the time of the colonial conquest, which followed 300 years of slave trade, the population of the region had fallen to less than one-third of that figure. After independence, parts of the Democratic Republic of Congo (formerly known as the Belgian Congo and later as Zaire) had barely caught up to the sixteenth-century num- bers.20 Other regions of western and eastern Africa provide similar examples— at least in the eyes of advocates of rapid population growth in Africa.

In terms of ratios of population to arable land (land under cultivation, fal- low land, pastures, and forests), Africa south of the Sahara is said by these sup- porters of population expansion to have a total of 1.4 billion arable hectares. Land actually being cultivated amounts to only a fraction of this potential. Thus, only 12% of all potential arable land is under cultivation, and this low rural population density is viewed as a serious drawback to raising agricul- tural output.21 Similar arguments have been expounded with regard to such Latin American countries as Brazil and Argentina.

Three other noneconomic arguments, each found to some degree in a wide range of developing countries, complete the “population growth is desirable” viewpoint. First, many countries claim a need for population growth to protect currently underpopulated border regions against the expansionist intentions of neighboring nations. Second, there are many ethnic, racial, and religious groups in less developed countries whose attitudes favoring large family size have to be protected for both moral and political reasons. Finally, military and political power are often seen as dependent on a large and youthful population.

Many of these arguments have a certain realism about them—if not in fact, then at least in the perceptions of vocal and influential individuals in both the developed and developing worlds. The important point is that they represent

311CHAPTER 6 Population Growth and Economic Development

a considerable range of opinions and viewpoints and therefore need to be seriously weighed against the counterarguments of theorists who believe that rapid population growth is indeed a real and important problem for underde- veloped countries. Let us now look at some of these counterarguments.

It Is a Real Problem

Positions supporting the need to curtail population growth because of the neg- ative economic, social, and environmental consequences are typically based on one of the following three arguments.

The Extremist Argument: Population and Global Crisis The extreme version of the population-as-problem position attempts to attribute almost all of the world’s economic and social evils to excessive population growth. Unrestrained population increase is seen as the major crisis facing humankind today. It is regarded as the principal cause of poverty, low levels of living, malnutrition, ill health, environmental degradation, and a wide array of other social problems. Value-laden and incendiary terms such as population bomb and population explosion are tossed around. Indeed, dire predictions of world food catastrophes and ecological disaster are often attributed almost entirely to the growth in population numbers.22 Such an extreme position leads some of its advocates to assert that “world” (i.e., developing country) population stabilization or even decline is the most urgent contemporary task, even if it requires severe and coercive measures such as compulsory sterilization to control family size in some of the most densely populated developing coun- tries, such as India and Bangladesh.

The Theoretical Argument: Population-Poverty Cycles and the need for Family-Planning Programs The population-poverty cycle theory is the main argument advanced by economists who hold that too rapid popula- tion growth yields negative economic consequences and thus should be a real concern for developing countries. Advocates start from the basic proposition that population growth intensifies and exacerbates the economic, social, and psychological problems associated with the condition of underdevelopment. Population growth is believed to retard the prospects for a better life for the already born by reducing savings rates at the household and national levels. It also severely draws down limited government revenues simply to provide the most rudimentary economic, health, and social services to the additional people. This, in turn, further reduces the prospects for any improvement in the levels of living of the existing generation and helps transmit poverty to future generations of low-income families.

Because widespread absolute poverty and low levels of living are thus seen as a major cause of large family size, and large families retard economic growth, it follows that economic and social development is a necessary condi- tion for bringing about an eventual slowing or cessation of population growth at low levels of fertility and mortality. But according to this argument, it is not a sufficient condition—that is, development provides people with the incen- tives and motivations to limit their family size, but family-planning programs

Population-poverty cycle A theory to explain how poverty and high population growth become reinforcing.

312 PART TWO Problems and Policies: Domestic

are needed to provide them with the technological means to avoid unwanted pregnancies. Even though countries such as France, Japan, the United States, Great Britain, and, more recently, Taiwan and South Korea were able to reduce their population growth rates without widespread family-planning clinics, it is argued that the provision of these services will enable other countries desir- ing to control excessive population growth to do so more rapidly than if these family-planning services were not available.

A Simple Model A basic model that economists use to demonstrate these adverse consequences of rapid population growth is a simplification of the standard Solow-type neoclassical growth equation.23 Using the standard production function, Y = f(K,L,R,T) —that is, output is a function of capital, labor, resources, and technology—and holding the resource base fixed, we can derive the result that

y - l = a(k - l) + t (6.2)

where y = rate of GNI growth ΔY/Y, l = rate of labor force (population) growth ΔL/L, k = rate of growth of the capital stock ΔK/K, a = capital elas- ticity of output (usually found to be constant), and t = the effect of techno- logical change (the Solow residual in empirical studies of sources of economic growth).

Assuming constant returns to scale, Equation 6.2 simply states that the rate of per capita income growth (y − l) is directly proportional to the rate of growth of the capital-labor ratio (k − l) plus the residual effects of technologi- cal progress (including improved human and physical capital). Therefore, in the absence of technological change, the higher the rate of population growth (l), the more rapid the rate of capital stock growth (k) must be and thus the greater the concomitant savings and investment rate just to maintain constant levels of per capita income. Moreover, because k may not be independent of l , as is traditionally assumed in neoclassical growth models, but may in fact be inversely related due to the reduced savings impact implied by the higher dependency burden effects of rapid population growth, it follows that the negative economic impact of population growth may even be greater than these models imply. Finally, if low incomes induce poor families to have more children as a source of cheap labor and old-age security, then we have another vicious circle in progress—poor people have large families partly to compen- sate for their poverty, but large families mean greater population growth, higher dependency burdens, lower savings, less investment, slower economic growth, and ultimately greater poverty. In an extreme case, a neo-Malthusian population trap can emerge. Population growth is thus seen as both a cause and a consequence of underdevelopment!

However, keep in mind that, as you saw in Chapters 3 and 4, population growth can tell only part of the story of economic growth. In this regard, Wil- liam Easterly argued that “even if population growth lowered per capita growth one for one (the general view of the population alarmists), this would explain only about one-third of the variation in per capita growth.”24 Growth in productivity, especially as spurred by structural transformation of the economy (Chapter 3), is usually more important in economic development outcomes.

313CHAPTER 6 Population Growth and Economic Development

Other Empirical Arguments: Seven negative Consequences of Population Growth According to the latest empirical research, the potential negative consequences of population growth for economic development can be divided into seven categories: its impact on economic growth, poverty and inequality, education, health, food, the environment, and international migration.25

1. Economic Growth. Evidence shows that although it is not the culprit be- hind economic stagnation, rapid population growth lowers per capita in- come growth in most developing countries, especially those that are al- ready poor, dependent on agriculture, and experiencing pressures on land and natural resources.

2. Poverty and Inequality. Even though aggregate statistical correlations between measures of poverty and population growth at the national level are often inconclusive, at the household level the evidence is strong and compelling. The negative consequences of rapid population growth fall most heavily on the poor because they are the ones who are made land- less, suffer first from cuts in government health and education programs, and bear the brunt of environmental damage. Poor women once again bear the greatest burden of government austerity programs, and another vicious circle ensues. To the extent that large families perpetuate poverty, they also exacerbate inequality.

3. Education. Although the data are sometimes ambiguous on this point, it is generally agreed that large family size and low incomes restrict the oppor- tunities of parents to educate all their children. At the national level, rapid population growth causes educational expenditures to be spread more thinly, lowering quality for the sake of quantity. This in turn feeds back on economic growth because the stock of human capital is reduced by rapid population growth.

4. Health. High fertility harms the health of mothers and children. It increases the health risks of pregnancy, and closely spaced births have been shown to reduce birth weight and increase child mortality rates.

5. Food. Feeding the world’s population is made more difficult by rapid population growth—a large fraction of developing country food require- ments are the result of population increases. New technologies of produc- tion must be introduced more rapidly, as the best lands have already been cultivated. International food relief programs become more widespread.

6. Environment. Rapid population growth contributes to environmental degradation in the form of forest encroachment, deforestation, fuelwood depletion, soil erosion, declining fish and animal stocks, inadequate and unsafe water, air pollution, and urban congestion (see Chapter 10).

7. International Migration. Many observers consider the increase in inter- national migration, both legal and illegal, to be one of the major conse- quences of developing countries’ population growth. Though many factors spur migration (see Chapter 7), an excess of job seekers (caused by rapid

314 PART TWO Problems and Policies: Domestic

population growth) over job opportunities is surely one of them. However, unlike the first six consequences listed here, some of the economic and social costs of international migration fall on recipient countries, increasingly in the developed world. It is not surprising, therefore, that this issue has recently taken on political importance in North America and Europe (see Chapter 2).

Goals and Objectives: Toward a Consensus

In spite of what may appear to be seriously conflicting arguments about the positive and negative consequences of population growth, a common ground has emerged on which many people on both sides of the debate can agree. This position is characterized succinctly by Robert Cassen:

After decades of controversy over the issue of population policy, there is a new international consensus among and between industrial and developing countries that individuals, countries, and the world at large would be better off if population were to grow more slowly. The consequences of rapid population growth should be neither exaggerated nor minimized. Some past expressions of alarm have been counterproductive, alienating the very audiences they were intended to persuade; at the same time, claims that population growth was not all that important have had the effect of diminishing a proper concern for the subject.26

The following three propositions constitute the essential components of this intermediate or consensus opinion.

1. Population growth is not the primary cause of low levels of living, ex- treme inequalities, or the limited freedom of choice that characterize much of the developing world. The fundamental causes of these prob- lems must be sought, rather, in the plight of poor families, especially women, and the failure of other aspects of domestic and international de- velopment policy.

2. The problem of population is not simply one of numbers but involves the quality of life and material well-being. Thus, developing country popula- tion size must be viewed in conjunction with developed-country affluence in relation to the quantity, distribution, and utilization of world resources, not just in relation to developing countries’ indigenous resources.

3. Rapid population growth does serve to intensify problems of underdevel- opment and to make prospects for development that much more remote. As noted, the momentum of growth means that, barring catastrophe, the population of developing countries will increase dramatically over the coming decades, no matter what fertility control measures are adopted now. It follows that high population growth rates, though not the princi- pal cause of underdevelopment, are nevertheless important contributing factors in specific countries and regions of the world.

In view of these three propositions, we may conclude that the following three policy goals and objectives might be included in any realistic approach to the issue of population growth in developing countries.

315CHAPTER 6 Population Growth and Economic Development

1. In countries or regions where population size, distribution, and growth are viewed as an existing or potential problem, the primary objective of any strategy to limit further growth must deal not only with the population vari- able per se but also with the underlying social and economic conditions of underdevelopment. Problems such as absolute poverty, gross inequality, widespread unemployment (especially among women), limited female access to education, malnutrition, and poor health facilities must be given high pri- ority. Their amelioration is both a necessary concomitant of development and a fundamental motivational basis for the expanded freedom of the indi- vidual to choose an optimal—and in many cases, smaller—family size.

2. To bring about smaller families through development-induced motiva- tions, family-planning programs providing both the education and the technological means to regulate fertility for people who wish to regulate it should be established.

3. Developed countries should help developing countries achieve their lowered fertility and mortality objectives, not only by providing contra- ceptives and funding family-planning clinics, but also, even more impor- tant, by curtailing their own excessive depletion of nonrenewable world resources through programs designed to cut back on the unnecessary consumption of products that intensively use such resources; by making genuine commitments to eradicating poverty, illiteracy, disease, and mal- nutrition in developing countries as well as their own; and by recognizing in both their rhetoric and their international economic and social dealings that development is the real issue, not simply population control.

6.6 Some Policy Approaches

In view of these broad goals and objectives, what kinds of economic and social policies might developing and developed-country governments and inter- national assistance agencies consider to bring about long-term reductions in the overall rate of world population growth? Three areas of policy can have important direct and indirect influences on the well-being of present and future world populations:

1. General and specific policies that developing country governments can initiate to influence and perhaps even control their population growth and distribution

2. General and specific policies that developed-country governments can initiate in their own countries to lessen their disproportionate consump- tion of limited world resources and promote a more equitable distribution of the benefits of global economic progress

3. General and specific policies that developed-country governments and international assistance agencies can initiate to help developing countries achieve their population objectives

Let us deal with each of these areas in turn.

316 PART TWO Problems and Policies: Domestic

What Developing Countries Can Do

Earlier discussions have led to the conclusion that the principal variables influencing the demand for children at the family level are the ones most closely associated with the concept of development as we have defined it in Chapter 1. Thus, certain development policies are particularly crucial in the transition from a high-growth to a low-growth population. These policies aim at eliminating absolute poverty; lessening income inequalities; expand- ing educational opportunities, especially for women; providing increased job opportunities for both men and women; bringing the benefits of modern preventive medicine and public health programs, especially the provision of clean water and sanitation, to the rural and urban poor; improving maternal and child health through more food, better diets, and improved nutrition so as to lower infant mortality; and creating a more equitable provision of other social services to wide segments of the population. Again, it is not numbers per se or parental irrationality that is at the root of the “population problem.” Rather, it is the pervasiveness of absolute poverty and low levels of living that provide the economic rationale for large families and burgeoning populations. And it is the spillover effects or negative social externalities of these private parental decisions (e.g., for education, health care, food supplies, environment and resource degradation, job creation, overall growth, and income distribu- tion) that provide the strictly economic efficiency justification (in terms of “market failure” arguments) for government intervention in population mat- ters. Clearly, there are noneconomic justifications as well.

Although long-run development policies of the kind just outlined are essential to ultimate population stabilization, there are five more specific poli- cies that developing country governments might try to adopt to lower birth rates in the short run.27

First, they can try to persuade people to have smaller families through the media and the educational process, both formal (school system) and informal (adult education).

Second, they can enhance family-planning programs to provide health and contraceptive services to encourage the desired behavior. Such publicly spon- sored or officially supported programs now exist in most developing coun- tries. Today only a few countries do not have such publicly sponsored or officially endorsed family-planning programs. However, there remains sub- stantial unmet demand for contraceptives, as seen in Box 6.3.

Third, they can deliberately manipulate economic incentives and disincentives for having children—for example, through the elimination or reduction of maternity leaves and benefits, the reduction or elimination of financial incen- tives, or the imposition of financial penalties for having children beyond a certain number; the establishment of old-age social security provisions and minimum-age child labor laws; the raising of fees and elimination of heavy public subsidies for higher education; and the subsidization of smaller families through direct money payments. Although some form of population-related incentive or disincentive schemes now exist in over 30 developing countries, Singapore, India, Bangladesh, South Korea, and China have been especially prominent in experimenting with policies to reduce family size. For example, Singapore allocated scarce public housing without giving consideration to

317CHAPTER 6 Population Growth and Economic Development

family size. It also limited paid maternity leave to a maximum of two children, scaling the delivery fee according to number of children and reducing income tax relief from five to three children. In 1984, it even went so far as to give special priority in school admission to all children born to women with uni- versity degrees while penalizing non-degree-holding women with more than two children. The presumed but dubious rationale was that educated women have brighter children whose births should be encouraged while discourag- ing the less educated (and presumably less intelligent) women from bearing more children. But fertility fell so dramatically that by 2004, the city-state had introduced incentives to increase fertility (as with Japan and Europe, relaxed controls on immigration would be more cost-effective). China has by far the most comprehensive set of state-enforced incentives and disincentives; they are described in the case study at the end of this chapter.

Fourth, governments can attempt to coerce people into having smaller fami- lies through the power of state legislation and penalties. For obvious reasons, few governments would attempt to engage in such coercion; not only is it often morally repugnant and politically unacceptable, but it is also almost always extremely difficult to administer. The defeat of Indian Prime Minister Indira Gandhi’s government in 1977 was largely due to popular resentment of the government’s forced-sterilization program.

Finally, no policy measures will be successful in controlling fertility unless efforts are made to raise the social and economic status of women and hence create

BOX 6.3 FInDInGS Contraceptives Need and Use in Developing Countries, 2003 to 2012

Jacqueline Darroch and Susheela Singh analyzed the use and need for contraceptives in developing countries, using data from comparable national sur- veys for married and unmarried women ages 15 to 49 in 2003, 2008, and 2012. Darroch and Singh esti- mated numbers and percentages of women wanting to avoid pregnancy, according to whether they were us- ing modern contraceptives, or using either no method or only a traditional method. They found that “the number of women wanting to avoid pregnancy and therefore needing effective contraception increased substantially,” from 716 million in 2003 to 867 mil- lion in 2012. Most of the increase corresponded to population growth. The percentage of women wish- ing to avoid pregnancy also rose, from 54% in 2003 to 57% in 2012. At the same time, the “use of modern contraceptive methods also increased, and the overall proportion” of all women ages 15-49 “with “unmet

need for modern methods among those wanting to avoid pregnancy decreased,” from 29% in 2003, to 26% in 2012 (although the number rose from 210 mil- lion to 222 million). The unmet need for modern con- traceptives among those wanting to avoid pregnancy remained very high, “especially in sub-Saharan Africa (53 million [60%] of 89 million), south Asia (83 mil- lion [34%] of 246 million), and western Asia (14 mil- lion [50%] of 27 million).” The authors maintained that, “to meet the unmet need for modern contracep- tion, countries need to increase resources, improve access to contraceptive services and supplies, and provide high-quality services and large-scale public education interventions to reduce social barriers.”

Source: Jacqueline Darroch and Susheela Singh. “Trends in contraceptive need and use in developing countries in 2003, 2008, and 2012: An analysis of national surveys.” The Lancet 381 (May 18, 2013): 1756–1762.

318 PART TWO Problems and Policies: Domestic

conditions favorable to delayed marriage and lower marital fertility.28 A cru- cial ingredient in any program designed to lower fertility rates is the increased education of women, followed by the creation of jobs for them outside the home. The availability of income-earning opportunities can lead young women to delay marriage by enabling them to become economically self- sufficient and therefore in a better position to exercise control over their choice of partner and the timing of marriage. It can also reduce family pressures for early marriage by allowing women to make a contribution to parental house- hold income. An independent source of income also secures a stronger posi- tion for married women in the household, reducing their dependence on other family members, particularly male offspring, for economic security. Fur- thermore, it enables women to consider the opportunity costs of additional children when childbearing competes with income-generating activities. In general, the availability of outside sources of income offers women genuine alternatives to early marriage and frequent childbearing, which are often moti- vated by their lack of resources. An additional benefit of employment outside the home is that it reduces women’s isolation, which is often an impediment to the provision of family-planning services, and can increase their household bargaining power.29

The importance of these policies to improve the role and status of women was underlined at the 1994 Cairo International Conference on Population and Development, where emphasis was placed on the general empowerment of women, especially in the area of reproductive choice. The Cairo Program of Action summarized this position in the following manner:

The empowerment and autonomy of women and the improvement of their politi- cal, social, economic and health status . . . [are] essential for the achievement of sus- tainable development and . . . for the long-term success of population programs. Experience shows that population and development programs are most effective when steps have simultaneously been taken to improve the status of women.30

What the Developed Countries Can Do

When we view the problems of population from the perspective of global resources and the environment, as we should, the question of the relationship between population size and distribution and the depletion of many nonre- newable resources in developed and underdeveloped countries assumes major importance. In a world where 4.5% of the population, located in one coun- try, the United States, accounts for nearly one-fifth of the annual world total energy use, we are clearly not dealing only or even primarily with a problem of population numbers when it comes to environment and resources. We must also be concerned with the impact of rising affluence and the very unequal worldwide distribution of incomes on the depletion of many nonrenewable resources such as petroleum, certain basic metals, and other raw materials essential for economic growth. The use of fossil fuel energy to power private automobiles, operate home and office air conditioners, and so on in the devel- oped nations remains the major contributor of carbon dioxide (CO2) gases into the atmosphere and to the phenomenon of greenhouse global warming (see Chapter 10).31 It also means that there is potentially that much less to fertilize

Reproductive choice The concept that women should be able to determine on an equal status with their husbands and for themselves how many children they want and what methods to use to achieve their desired family size.

319CHAPTER 6 Population Growth and Economic Development

small family farms in the less developed nations. Alternatively, it means that poor families will have to pay more to obtain these valuable resource inputs.

Many similar examples could be given of the gross inequalities in global resource use. Perhaps more important, one could cite innumerable instances of the unnecessary and costly waste of many scarce and nonrenewable resources by the affluent developed nations. The point, therefore, is that any worldwide program designed to engender a better balance between resources and people by limiting developing-country population growth through social interven- tion and family planning must also include the responsibility of rich nations to simplify their own consumption demands and lifestyles. Such changes would free resources that could then be used by poor nations to generate the social and economic development essential to slowing population growth.

In addition to simplifying lifestyles and consumption habits, one other positive (if unlikely) internal policy that rich nations could adopt to mitigate current world population problems would be to liberalize the legal conditions for the international immigration of poor, unskilled workers and their fami- lies from Africa, Asia, and Latin America to North America, Europe, Japan, and Australia. The international migration of peasants from Europe to North America, Australia, and New Zealand in the nineteenth and early twentieth centuries was a major factor in moderating the problems of underdevelop- ment and population pressure in European countries. No such safety valve or outlet exists today for developing countries. In fact, what few outlets existed have over the past two decades been progressively closed. Yet clearly, many labor-scarce societies could benefit economically from international migration, and the benefits to developing countries would be enormous. For example, the United Nations has estimated that legal barriers to international migration from the developing to the developed world cost developing nations at least $250 billion a year.32

How Developed Countries Can Help Developing Countries with Their Population Programs

There are a number of ways in which the governments of rich countries and multilateral donor agencies can help the governments of developing coun- tries achieve their population policy objectives sooner. The most important of these concerns the willingness of rich countries to be of genuine assistance to poor countries in their development efforts, particularly in sub-Saharan Africa. Such genuine support would consist not only of expanded public and private financial assistance but also of improved trade relations, such as tar- iff- and quota-free access to developed-country markets, more appropriate technology transfers, assistance in developing indigenous scientific research capacities, better international commodity-pricing policies, and a more equi- table sharing of the world’s scarce natural resources. (These and other areas of international economic relations between rich and poor countries will be examined in Part Three.)

There are two other activities more directly related to fertility moderation in which rich-country governments, international donor agencies, and pri- vate nongovernmental organizations (NGOs) can play an important assisting role. The first is the area of research into the technology of fertility control, the

320 PART TWO Problems and Policies: Domestic

contraceptive pill, modern intrauterine devices (IUDs), voluntary sterilization procedures, and, particularly in the age of AIDS, effective barrier contracep- tion. Research has been going on in this area for a number of years, almost all of it financed by international donor organizations, private foundations, and aid agencies of developed countries. Further efforts to improve the effective- ness of this low-cost contraceptive technology while minimizing the health risks should be encouraged.

The second area includes financial assistance from developed countries for family-planning programs, public education, and national population policy research activities in the developing countries. This has traditionally been the primary area of developed-country assistance in the field of population. Total resources devoted to these activities have risen dramatically. It remains an open question, however, whether such resources (especially those allocated to premature family-planning programs) might not have been more effectively used to achieve their fertility goals had they instead been devoted directly to helping low-income countries to raise the levels of living of their poorest people. As pointed out earlier, it is of little value to have sophisticated family- planning programs when people are not motivated to reduce family size.

We conclude with a note of optimism. Fertility rates in many of the poorest countries, such as Bangladesh and most of the countries in sub-Saharan Africa, have experienced an impressive decline. Population experts have lowered their estimates of world population growth for coming decades. In no small part, this decline is the result of more widespread availability of family plan- ning. This change helps set the stage for an opportunity for successful devel- opment efforts in the coming years, but developed countries need to do their part in providing expanded development assistance, especially efforts focused on the need and opportunity to greatly reduce the incidence of poverty, which remains the biggest cause of high rates of fertility.

321

Case Study 6

Population, Poverty, and Development: China and India

Two of the world’s fastest growing economies, China and India, also happen to be the world’s two most populous nations, with some 1.35 billion and 1.22 billion people, respectively. Both countries continue to grow, albeit at slower paces. Accord- ing to the 2012 UN Population Division’s medium- variant projections, by 2030 India will become the world’s most populous nation, with 1.48 billion people. The United Nations projects a population of 1.45 billion in China by 2030, which is then pro- jected to fall to about 1.28 billion by about 2065. In contrast, India’s population is projected to continue growing until about 2065, reaching a peak of about 1.64 billion around 2065 before its population finally starts to decline.

India’s 2013 population of more than 1.2 billion is well over triple the number at independence, despite introducing the world’s first family- planning policy in 1950. At 1.35 billion, China’s population remains larger, but its highly restric- tive one-child policy, despite being fairly success- ful at slowing fertility, has apparently been less successful than approaches based on women’s empowerment and education in some parts of India, such as the state of Kerala. What can we learn about population and development from the world’s most populous countries?

In India, it is common to hear the view that “everything is growing faster in China than India, except for population.” India, which had well under two-thirds of China’s population half a century ago, is projected to surpass China’s population by 200 million people by 2050. Like most developing countries, both countries’ populations grew rapidly when their mortality rates fell and their birth rates fell much more slowly. Both countries have viewed

population pressures as threatening prospects for future development.

It is well known that as incomes rise, fertility falls, due largely to the increased opportunity cost of women’s time. The causality between fertility and growth runs in both directions. China’s rapid economic growth since about 1980 has also been attributed in part to its lower fertility rate. India’s increased growth rate since about 1990 may also be related to its more moderate decline in fertil- ity. Both may reflect in part the “demographic dividend” examined earlier in the chapter. Thus, population policy can potentially play an important role in setting the stage for growth. Moreover, to the degree that we accept Nobel economics laure- ate Amartya Sen’s view that development is free- dom, the greater opportunities available to young women when fertility is reduced or delayed is itself a key indicator of development success, and popu- lation policy can help realize these goals.

Population Policy in China China has been the world’s most populous nation for centuries. After the Communist takeover in 1949, Chinese leaders led by Mao Zedong took a broadly pronatalist stance, believing that a com- munist society could solve any population prob- lems and that a larger population would mean a more powerful country. Mao (whom China’s lead- ers still call “60% right” about policy) went so far as to send advocates of population control to jail. However, in the face of famine in the late 1950s, these policies moderated.

In 1980, China initiated a tough new drive to deter births, with a goal of lowering the annual birth rate to 1% during the decade. Stringent and often

draconian measures to achieve that goal were intro- duced in 1982 and 1983 as the Chinese government adopted a policy of one child per family. Social and political pressures to limit family size to one child included requiring women to appeal to the neigh- borhood committee or council for formal permis- sion to become pregnant. Although first births were routinely approved, second births were usually approved only if the first child had a serious birth defect or if the woman had remarried. Economic incentives included giving priority to one-child families in housing, medical care, and education. Mothers of two or more children were often denied promotions, and steep fines, sometimes in excess of 10 times China’s per capita income, were levied for second and third children. Although a grow- ing number of exceptions have been introduced in recent years, notably allowing a second child if the first child is a girl, and allowing a second child if both parents are themselves only children; at the Third Plenum in 2013, it was announced that the policy would be relaxed further, with a second child permitted if either parent is an only child (subject to verification by the government). Despite these adjustments, the policy remains probably the most restrictive in the world.

Given such rigid national policies and a strong cultural preference for boys, it is not surprising that there have been many reports of girls receiv- ing less medical attention and also of selective abortion of female fetuses and even female infanti- cide (“gendercide”).

Male-to-female ratios are higher than the nor- mal level in many Asian countries, and gender bias is at least partly to blame. Amartya Sen’s pioneer- ing 1992 research estimated that 44 to 50 million women were already “missing” in China, depend- ing on whether the comparison is to Western coun- tries or to Africa. The most recent data confirm that these trends have continued, with Stephan Klasen and Claudia Wink calculating that well over 6% of women are “missing” in China. It is estimated that in 2010, there were 106 males for every 100 females in China overall; and in a trend pointing to a wors- ening of the problem, close to 118 boys were born for every 100 girls. Such balance is all but unprec- edented in recorded world history. Of course, these current cultural preferences may change with fur- ther economic development. In fact, this ratio is

now falling, albeit very, slightly, from a recent peak ratio of 120 boys to 100 girls, according to official government data.

The full impact of China’s population control programs is uncertain. Only time will tell whether the benefits of reduced population growth achieved through severe social and economic pressures for one-child families will be worth the cost of a harsh break with traditional family norms and percep- tions regarding the value of children. Resistance in rural areas, where well over 60% of the population still resides, was apparently so widespread that in August 1988, when the Chinese government discov- ered to its surprise that the population had already passed the 1 billion mark, it decided to increase its enforcement of the one-child norm in rural as well as urban areas. However, popular opposition again caused it to relax its stringent controls slightly and to focus more on elevating the status of women and providing greater old-age security.

By the mid-1990s, China’s fertility rate reached 1.9 births per woman, and it fell further to below 1.6 by 2013. This rate is below replacement level and is consistent with a slow long-term decline in population growth. Because of population momen- tum, China’s population has continued to grow as larger, younger cohorts replace smaller, older ones. However, the country’s largest cohorts are now passing out of their childbearing years. The popula- tion growth rate has slowed dramatically, and the population is not expected to exceed 1.4 billion at its peak before starting to fall.

In practice, many families have two children rather than one, and others in rural areas, including ethnic minorities exempted from the one-child pol- icy, have more than two children. But fertility rates are extremely low in the urban areas to which an increasing share of the rural population is moving. Typical estimates suggest that upward of 250 mil- lion fewer people were born in China than would have been born without the one-child policy—an enormous impact. There are now concerns that China will have to reevaluate the policy to prevent too high a dependency ratio of retired to working adults.

The apparent success of China’s tough fertility policies has led some observers to see advantages of dictatorship rather than democracy in spurring development. But in fact there are several ways

322

323

in which the lack of a free press in particular and democracy more generally has held back China’s development. In Mao’s “Great Leap Forward,” at least 30 million people died due to poor govern- ment decisions and incentives for bureaucrats to send overly optimistic reports from the field. Demo- cratic India, by contrast, has not had a famine since independence in 1947. Amartya Sen attributes China’s lead in economic growth to its massive investments in health and education, which India has lacked. Dictatorship can be good or bad for fertility programs or any other aspects of develop- ment. But the risks of a very bad outcome are prob- ably much lower with democracy.

Successful population control in China comes with its own risks and unintended consequences as well as substantial rewards. By 2050, China will have almost twice as many people above age 50 as below age 20. In addition, while fertility has fallen, preference for boys over girls has actually intensi- fied. Many Chinese families seem to feel that if they are to have only one child, it should be a boy, to carry on the family name and help support the parents in their old age. A 2007 report from China’s State Population and Family Planning Commission concluded that the country may have about 30 mil- lion more men than women of marriageable age by 2020 and warned that the result could be social instability. A 2009 study by economists Shang-Jin Wei and Xiaobo Zhang provided robust evidence that China’s recent new surge in savings is caused in large part by competitive investments in housing and other wealth accumulation by families seeking to attract brides for their sons. Such a savings surge even has potential implications for global imbal- ances (see Chapters 12 and 13).

The high fraction of the population now of work- ing age has provided, in China’s case at least, a “demographic dividend.” But the next phase of the demographic transition is likely to pose major challenges for China, with its big drops in fertil- ity ahead of historical patterns in other countries; hence the saying that “China must get rich before it gets old.” But as far as is known, no society has ever faced such rapid population aging. By 2013, the labor force in China had already begun to slowly shrink.

In sum, although rapid economic growth and coercion and incentives in family planning account

for part of China’s drop in fertility, other factors include female literacy, improved child health, and greater economic opportunities for women. These have also been factors in the strong success in fertil- ity reduction in the Indian state of Kerala.

Population Policy in India In 1949, India became the first country to implement a national family-planning program. It has proved to be relatively ineffective and has proceeded in fits and starts. By the early 1970s, observers were becoming increasingly alarmed by the very high rate of population growth in India.

When Prime Minister Indira Gandhi tried to implement drastic population control in 1975–1977, a period during which she seized dictatorial pow- ers, it was a failure. Reports of forced sterilizations, sometimes in mass “sterilization camps,” and other coercive measures ended up giving family plan- ning a bad reputation in many areas of the country. Indeed, public revulsion toward these coercive fer- tility policies helped bring the “emergency” period to an end more quickly, and when elections were held in 1977, Gandhi was voted out of office. Her return to power in the elections of 1980 was made possible in part by her commitment not to rein- troduce coercive birth control policies. Years later, villagers in some parts of India avoided health workers out of fear of forced sterilization.

However, family planning did become more widely practiced. Some of the acceptance of limits on family size reflected rising income among the close to 250 million middle-class Indians and some- what improved conditions among a significant frac- tion of the poor. Some of it reflected modest moves back to policy incentives to encourage smaller fami- lies. There have been variations from state to state. In Madhya Pradesh, individuals who had a third or subsequent child after January 2001 were banned from running for election to village council posts, spurring considerable controversy. In 2004, an uproar over reported higher fertility among Mus- lims than among Hindus—reports that turned out to be greatly exaggerated—revealed the continuing political sensitivity of the issue.

As fertility has fallen, a preference for boys over girls has developed, particularly in the “Hindi belt” in northern India. The result is a “missing women” problem parallel to China’s. Stronger male bias is

actually found in the better-off states of India, and researchers Jean Drèze, Anne-Catherine Guio, and Mamta Murthi found that “female disadvantage in child survival is significantly lower in districts with higher poverty levels.”

P. N. Mari Bhat and A. J. Francis Zavier analyzed data from the National Family Health Survey and estimated that “in northern India, girls currently constitute about 60% of the unwanted births and that the elimination of unwanted fertility has the potential to raise the sex ratio at birth to 130 boys per 100 girls.” Such a dramatic imbalance seems likely to lead to future social stress. As of 2010, the ratio of males to females in India as a whole had reached 108 to 100, one of the highest in the world; the ratio at birth is now approximately 112 to 100. But this imbalance is not inevitable—social devel- opment can make all the difference.

Kerala, a state on India’s southwest coast that has emphasized poverty reduction and human devel- opment, is an important case in point. By the mid- 1990s, Kerala’s fertility rate had fallen to just 1.7 births per woman and has remained low—still 1.7 in 2010 (Indian Planning Commission)—implying a slowly falling population over time (in the absence of in-migration). Thus, Kerala’s fertility rate was until recently less than that of China, but unlike China, the dramatic reductions in fertility in Kerala were achieved without coercion, let alone China’s huge direct economic incentives for lowered fertil- ity. In India overall, the fertility rate is 2.5, and in Bihar, a socially backward state, the fertility rate in 2010 was 3.7, similar to that of Pakistan. Overall, there are actually slightly more females than males in Kerala.

Norms of behavior can be highly influential, and multiple equilibria resulting from differ- ent expected norms of behavior are possible, as explored in Chapter 4 and applied to population norms in this chapter in section 6.4. There has been a slow but steady movement in attitudes toward the notion that a happy family is a small family in the India of today. Amartya Sen has observed that sharp declines in the rate of fertility in India in literate states, particularly Kerala and Tamil Nadu, was greatly influenced by public discourse on the negative impacts of high fertility. Discussions have emphasized problems caused both for young women and for communities as a whole.

In addition, and especially more recently, greater awareness on the part of rural women of urban norms of women’s empowerment, facili- tated by village television and the Internet, may have made a big impact, proving that cultural awareness can be powerful. Robert Jensen and Emily Oster provide some evidence on the power of television in India.

While television, billboard, and other advertis- ing in India has promoted family planning, and there is some evidence that these campaigns can have some positive impact on their own, such efforts have been far more successful when the social climate has changed enough to be receptive to the message. This helps explain why nongov- ernmental organizations working for comprehen- sive rural development have often apparently had more success than many government programs. In Kerala, if the official campaigns supporting small families have seemed more effective than elsewhere, it is largely because both social and economic conditions on the ground changed previ- ously or simultaneously. More than 85% of women in Kerala are literate, which means they have more power in the household and opportunities in the workforce as well as the ability to read print mate- rials about fertility and family planning. Some of Kerala’s success is due to the traditionally higher status of women in the local culture. But there is no reason that Kerala’s success cannot be dupli- cated elsewhere in India if there is the political and social will.

Sen concluded that Kerala’s impressive results in fertility reduction were achieved through active public dialogue that resulted ultimately in the emer- gence of new social attitudes and values—and that such dialogues on this sensitive subject were pos- sible only because of the very high level of female literacy in the state. Indeed, Sen pointed out that female literacy in Kerala was unmatched by any of China’s provinces.

The success of Kerala suggests that fertility reduction may depend not on rapid economic growth or even, in its absence, on draconian governmental policies but rather on grassroots human development that emphasizes women’s empowerment, in which civil society plays a lead- ing role. ■

324

Sources Acharya, Keya. “Sterilisation in India.” Contempo-

rary Review 279 (2001): 26. Amartya Sen, ”What’s the point of a development

strategy,” paper written for the United Nations Committee on Development Strategy and Management of the Market Economy, May 1996, page 20.

Barro, Robert J. Determinants of Economic Growth: A Cross-Country Empirical Study. Cambridge, Mass.: MIT Press, 1997.

“Can advertising create social change?” Busi- nessline, January 20, 2000, p. 1.

Dasgupta, Partha. “The population problem: The- ory and evidence.” Journal of Economic Literature 33 (1995): 1879–1902.

Drèze, Jean, Anne-Catherine Guio, and Mamta Murthi. “Mortality, fertility, and gender bias in India: A district-level analysis.” Population and Development Review 21 (1995): 745–782.

Drèze, Jean, and Mamta Murthi. “Fertility, educa- tion, and development: Evidence from India.” Population and Development Review 27 (2001): 33–63.

Jensen, Robert, and Emily Oster. “The power of TV: Cable television and women’s status in India.” Quarterly Journal of Economics 124 (2009): 1057–1094.

Klasen, Stephan and Claudia Wink, “Missing women: Revisiting the debate.” Feminist Eco- nomics, 9 (2–3), 2003, 263–299

Kremer, Michael. “Population growth and techno- logical change: One million b.c. to 1990.” Quarterly Journal of Economics 108 (1993): 681–716.

Mari Bhat, P. N., and Francis Zavier, A. J. “Fertil- ity decline and gender bias in northern India.” Demography 40 (2003): 637–657.

McElroy, Marjorie, and Dennis Tao Yang. “Carrots and sticks: Fertility effects of China’s population policies.” American Economic Review 90 (May 2000): 389–392.

Pritchett, Lant H. “Desired fertility and the impact of population policies.” Population and Develop- ment Review 20 (1994): 1–55.

Sen, Amartya. Development as Freedom. New York: Knopf, 1999.

Sen, Amartya. “Missing women.” British Medical Journal 304 (1992): 587–588.

Sunil, T. S., V. K. Pillai, and A. Pandey. “Do incen- tives matter? Evaluation of a family planning program in India.” Population Research and Policy Review 18 (1999): 563–577.

Wei, Shang-Jin and Xiaobo Zhang. “The competi- tive saving motive: Evidence from rising sex ratios and savings rates in China.”, Journal of Political Economy 119 (2011): 511–564.

Concepts for Review

Crude birth rate Death rate Demographic transition Doubling time Family-planning programs Hidden momentum of

population growth

Life expectancy at birth Malthusian population trap Microeconomic theory of fertility Natural increase Net international migration Population-poverty cycle Population pyramid

Rate of population increase Replacement fertility Reproductive choice Total fertility rate (TFR) Under-5 mortality rate Youth dependency ratio

325

326 PART TWO Problems and Policies: Domestic

notes

1. The 1970s marked the apogee in the history of world population growth. By the end of the decade, rates had begun to decline in a large number of developing countries, and it became clear that the pace of world population growth had peaked. For some evidence of this turning point, see Bernard Berelson, W. Parker Mauldin, and Sheldon Segal, “Population: Current status and policy options,” Social Science and Medicine 14c (1980): 71–97, and World Bank, World Development Report, 1984 (New York: Oxford University Press, 1984), ch. 4.

2. A convenient shorthand method of calculating doubling time is simply to divide any growth rate into the number 70. For example, something (an asset, population, GNI, etc.) growing at 2% per year will double its value in approximately 35 years. You may recall from algebra that the doubling time of a value (such as the real GNI of an economy) growing at rate p% per year may be found with the formula [1 + p/100]T = 2. Taking natural logs of each side, T ln[1 + p/100] ln 2. The natural log of 2 is approximately 0.7. On the

Questions for Discussion

1. Population growth in developing nations has pro- ceeded at unprecedented rates over the past few decades. Compare and contrast the present rate of population growth in less developed countries with that of the modern developed nations during their early growth years. What has been the major factor contributing to rapid developing country population growth since the Second World War? Explain your answer.

2. What is the relationship between the age structure of a population and its dependency burden? Is the dependency burden higher or lower in develop- ing countries? Why?

3. Explain the notion of the hidden momentum of population growth. Why is this an important concept for projecting future population trends in different developing nations?

4. Describe briefly the theory of the demographic transition. At what stage in this transition do most developing countries seem to be? Explain your answer.

5. How does the microeconomic theory of fertility relate to the theory of consumer choice? Do you think that economic incentives and disincentives influence family size decisions? Explain your answer, giving some specific examples of such incentives and disincentives.

6. “The world population problem is not just a mat- ter of expanding numbers but also one of rising affluence and limited resources. It is as much a problem caused by developed nations as it is one deriving from developing countries.” Comment on this statement.

7. List and briefly describe the principal causes of high population growth in developing countries and the major consequences.

8. Explain why fertility rates are falling much more rapidly in some developing countries than in others.

9. Outline and comment briefly on some of the argu- ments against the idea that population growth is a serious problem in developing nations.

10. Outline and comment briefly on some of the argu- ments in support of the idea that population growth is a serious problem in developing nations.

11. Outline and comment briefly on the various pol- icy options available to developing countries’ governments in their attempt to modify or limit the rate of population growth.

12. Suppose that a study finds that there is comple- mentarity in fertility decisions. What would this mean? What are the possible implications?

13. What aspects of population policy alternatives— including their strengths and weaknesses—are illustrated by the cases of China and India?

327CHAPTER 6 Population Growth and Economic Development

left- hand side, for small p, ln[1 + p/100] is approxi- mately equal to p/100. Substituting, Tp/100 = 0.7, or T = 70/p. For example, for reasonably small values of growth such as 4%, simply divide 70 by the percentage growth: After about 70/4 = 17.5 years, national income would double. As an addi- tional approximation, to find the growth of income per capita, simply subtract the rate of population growth. So if population is growing at 2% per year, in this example, income per capita would be grow- ing at 4% − 2% = 2% per year, and income per capita would double in approximately 70/2 = 35 years.

3. Population Reference Bureau, World Population Data Sheet, 2012 (Washington, D.C.: Population Reference Bureau, 2012).

4. The World Bank, World Bank World Development Indicators 2013 (Washington, D.C.: The World Bank), tab. 2.1.

5. For more discussion, see John Bongaarts, “Popu- lation policy options in the developing world,” Science 263 (1994): 771–776.

6. For an interesting reverse-population-alarmist perspective, see Philip Longman, “Think again: Global aging,” Foreign Policy (2012).

7. Replacement fertility may be approximated by the value of TFRR ≈ (1+SRB)/p(AM), where TFRR represents the replacement value for the total fertility rate, SRB represents the ratio of male to female births, and p(AM) represents the prob- ability of surviving to the mean age of the fertility schedule. See Samuel Preston, Patrick Heuveline, and Michel Guillo, Demography: Measuring and Modeling Population Processes (Oxford: Blackwell, 2001). Note that with gender balance and high female survival to mean fertility schedule (close to 30 years of age), the TFRR is close to 2.1. But when survival proportions are among the lowest in the world—close to 0.60 in the cases of Afghan- istan, Burundi, and Sierra Leone; replacement total fertility rates above 3.3 are implied. Under these conditions, a fertility rate of 2.1 would actu- ally result in population decline. See Thomas J. Espenshade, Juan Carlos Guzman, and Charles F. Westoff, “The surprising global variation in replacement fertility,” Population Research and Policy Review 22, No. 5-6 (2003): 575–583, who

calculate that the replacement rate across coun- tries ranges from 2.05 to 3.43.

8. See Timothy W. Guinnane, “The historical fertil- ity transition: A guide for economists,” Journal of Economic Literature 49, No. 3 (2011): 589–614.

9. A geometric progression is simply a doubling (or some other multiple) of each previous number, as in 1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 1,024, and so on. Like compound interest, geometric progres- sions have a way of reaching large numbers very rapidly.

10. Recent supporting evidence is found in Quamrul Ashraf and Oded Galor, “Dynamics and stagna- tion in the Malthusian epoch,"American Economic Review 101, No. 5 (2011): 2003–2041. The authors find that “technological superiority and higher land productivity had significant positive effects on population density but insignificant effects on the standard of living, during the time period 1–1500 CE.”

At lower levels of expected fertility, the slope of the S-shaped curve may increase at an increasing rate due to the presence of older children who can take care of younger siblings, together with a stronger family response to the negative impact of average fertility on wages and/or the decreased probability that any one child will gain formal-sector employment. But at higher levels of expected fertility, the slope of the S-shaped curve may increase at a decreasing rate due to worsening availability of education and health and increasing costs of raising additional children in relation to the benefits of doing so.

11. This interpretation is from Partha Dasgupta, An Inquiry into Well-Being and Destitution (New York: Oxford University Press, 1993), and is discussed in Pranab Bardhan and Chris Udry, Development Microeconomics (New York: Oxford University Press, 1999), p. 25.

12. The classic contribution is Simon Kuznets, Fertility Differentials between Less Developed and Developed Regions: Components and Implications (New Haven, Conn.: Economic Growth Center, Yale University, 1974).

13. See Chapter 9, and see Christopher Udry, “Gen- der, agricultural production, and the theory of the household,” Journal of Political Economy 104 (1996): 1010–1046.

10a.

328 PART TWO Problems and Policies: Domestic

14. See, for example, Nancy Birdsall, “Economic approaches to population growth,” in Handbook of Development Economics, vol. 1, eds. Hollis B. Chen- ery and T. N. Srinwasan (Amsterdam: Elsevier, 1988), pp. 478–542; Jean Drèze, Anne-Catherine Guio, and Mamta Murthi, “Mortality, fertility, and gender bias in India: A district-level analysis,” Population and Development Review 21 (1995): 745– 782; and Partha Dasgupta, “The population prob- lem: Theory and evidence,” Journal of Economic Literature 33 (1995): 1879–1902.

15. For empirical evidence that low fertility results mostly from economic, social, cultural, and edu- cational improvements in a population and only slightly from the availability of family-planning programs, see Lant H. Pritchett, “Desired fertility and the impact of population policies,” Population and Development Review 20 (1994): 1–55.

16. For an analysis of this conflict, see Jason L. Finkle and Barbara Crane, “The politics of Bucharest: Population, development, and the new inter- national economic order,” Population and Devel- opment Review 1 (1975): 87–114. Although this conflict was less visible in the Second World Pop- ulation Conference held in Mexico City in August 1984 and was a minor issue beneath that of repro- ductive choice and the empowerment of women at the Third Conference held in Cairo in 1994, it remained prominent in the discussions of many developing-world delegates.

17. For a more detailed discussion of these divergent opinions, see Michael S. Teitelbaum, “Population and development: Is a consensus possible?” For- eign Affairs 52 (1974): 749–757. See also Timothy King and Allen Kelley, The New Population Debate: Two Views on Population Growth and Economic Development (Washington, D.C.: Population Refer- ence Bureau, 1985), and Robert H. Cassen, Popu- lation Policy: A New Consensus (Washington, D.C.: Overseas Development Council, 1994).

18. See, for example, Colin Clark, “The ’popula- tion explosion’ myth,” Bulletin of the Institute of Development Studies 1 (1969); Julian Simon, The Ultimate Resource (Princeton, N.J.: Princeton Uni- versity Press, 1981); Nick Eberstadt, “Population and economic growth,” Wilson Quarterly (Winter 1986), pp. 95–129; and National Research Council, Population Growth and Economic Development: Policy

Questions (Washington, D.C.: National Academy Press, 1986).

19. Samir Amin, “Underpopulated Africa,” paper presented at the African Population Conference, Accra, Ghana, December 1971.

20. Ibid., fn. 2.

21. Ibid., p. 3. Of course, in the decades after these arguments were promulgated, population did dramatically increase in these regions. For another perspective on long-run benefits of greater pop- ulation density via faster technological prog- ress, see Michael Kremer, “Population growth and technological change: One million b.c. to 1990,” Quarterly Journal of Economics 108 (1993): 681–716.

22. For example, see Paul R. Ehrlich and Anne H. Ehrlich, Population, Resources, and Environment: Issues in Human Ecology, 2nd ed. (New York: Free- man, 1972); Lester R. Brown, In the Human Inter- est: A Strategy to Stabilize World Population (New York: Norton, 1974); and Paul R. Ehrlich and Anne H. Ehrlich, The Population Explosion (New York: Simon & Schuster, 1990).

23. We are grateful to Professor Harold Votey for suggesting this illustration. Details on the Solow model are found in Chapter 3 and Appendix 3.2.

24. William Easterly made the very basic argument in 1999 that “population growth does not vary enough across countries to explain variations in per capita growth. GDP per capita growth var- ies between - 2 and + 7 percent for all countries between 1960 and 1992. Population growth var- ies only between 1 and 4 percent.” Easterly, The Elusive Quest for Growth (Cambridge, Mass.: MIT Press, 1999), p. 92.

25. For a detailed review of this evidence, see Cassen, Population Policy, pp. 14–22; Dennis A. Ahlburg et al., Population and Economic Development: A Report to the Government of the Commonwealth of Australia (Canberra: Australian International Development Assistance Bureau, 1994); and Geoffrey McNicoll, “Effects of population growth: Visions and revi- sions,” Population and Development Review 21 (1995): 307–340. As the Ahlburg report demon- strates, not all of these consequences are unambig- uously negative. Much depends on the particular country and its demographic situation.

329CHAPTER 6 Population Growth and Economic Development

26. Robert Cassen, Population Policy, p. 12.

27. See Birdsall, “Economic approaches to population growth,” pp. 523–529.

28. Sousan Abadian, “Women’s autonomy and its impact on fertility,” World Development 24 (1996): 1793–1809. See also Shireen J. Jeejeebhoy, Women’s Education, Autonomy, and Reproductive Behavior: Experiences from Developing Countries (Oxford: Clarendon Press, 1995).

29. See Fenohasina Maret-Rakotondrazaka, “The effect of working outside the home on women’s empowerment in Nigeria,” Working Paper, George Washington University, 2014.

30. United Nations, International Conference, para. 4.1. See also Nancy Folbre, “Engendering economics:

New perspectives on women, work, and demo- graphic change,” in Proceedings of the World Bank Annual Conference on Development Economics, 1995, eds. Michael Bruno and Boris Pleskovic (Washington, D.C.: World Bank, 1996).

31. The United Nations Population Fund’s State of the World’s Population 2009 edition (New York: United Nations, 2009) examines relationships between population and climate change. Cited energy data are from World Resources Institute, World Resources, 2005 (New York: Oxford University Press, 2005), tab. 7.

32. More detail on the scale and benefits of remit- tances from international migration is provided in Chapter 14.

330

Title Here, Flush Left0

In this chapter, we focus on one of the most complex and nuanced dilemmas of the development process: the phenomenon of massive and historically unprec- edented movements of people from the rural countryside to the burgeoning cities of Africa, Asia, and Latin America. In Chapter 6, we documented the extraordinary increase in world and especially developing-country popula- tions over the past few decades. According to a 2013 UN estimate, by 2050 the world population is expected to reach 9.6 billion people, and nowhere will population growth be more dramatic than in the cities of the developing world. Indeed, according to estimates by the UN Population Division, for the first time in human history, in 2009 globally “the number of people living in urban areas (3.42 billion) had surpassed the number living in rural areas (3.41 billion).” The global urban majority is now widening with each passing year.1

After reviewing trends and prospects for overall urban population growth, we examine in this chapter the potential role of cities—both the modern sector and the urban informal sector—in fostering economic development. We then turn to a well-known theoretical model of rural-urban labor transfer in the context of rapid growth and high urban unemployment. In the final section, we evaluate various policy options that governments in developing countries may wish to pursue in their attempts to moderate the heavy flow of rural-to- urban migration and to ameliorate the serious unemployment problems that continue to plague their crowded cities. We also examine how the great poten- tial dynamism and productivity of developing cities can be better harnessed for

330

Urbanization and Rural-Urban Migration: Theory and Policy

Cities will increasingly become the main players in the global economy. —Kofi Annan, former secretary general of the United Nations and

Nobel laureate for Peace

The global economy is led by metropolitan economies. —Brookings Institution, Global MetroMonitor, 2010

More than one billion persons are living in slums. With rapid urbanization it is expected that in the next two decades there will be nearly two billion new urban residents, 90% of them in developing countries.

—World Bank, Addressing the Urbanization Challenge, 2013

Any strategy for a less desperate and more deliberate urbanization must include efforts to improve public services in rural areas.

—World Bank, World Development Report, 2009

7

331CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

rapid and more inclusive economic development. This chapter ’s case study looks at patterns of migration in India and Botswana.

7.1 Urbanization: Trends and Living Conditions

While the world as a whole became majority urban in 2009, even the devel- oping world is expected to become majority urban before 2020 (although the United Nations projects that the least developed countries will not reach this milestone until after 2050). Currently, most urban growth has occurred in cit- ies in Asia and Africa. Indeed, in 2012 the United Nations projected that the urban population of Africa will grow from 414 million in 2012 to over 1.2 bil- lion by 2050; and the urban population of Asia will grow from 1.9 billion to 3.3 billion. Thus taken together, the United Nations projects that Asia and Africa will account for some 86% of the global urban population increase in this period. In fact, there will be so much rural-to-urban migration in Asia that its rural population will actually decline in this period, as seen in Figure 7.1.2

Urbanization rates increase whenever urban population growth exceeds rural population growth. The positive association between urbanization and

FIGURE 7.1 Changes in Urban and Rural Population by Major Areas between 2011 and 2050 (in millions)

Source: United Nations, “Africa and Asia to lead urban population growth in the next four decades,”press release, http://esa.un.org/unup/pdf/WUP2011_Press-Release.pdf. Reproduced by permission of United Nations Publications.

–600 Africa OceaniaNorthern

America Latin America

and the Caribbean

EuropeAsia

–400

–200

0

200

400

600

800

1,000

1,200

1,400

1,600

851

295

52

–72

178

–24

110

–11

14 4

1,414

–480

Urban Population Rural Population

332 PART TWO Problems and Policies: Domestic

per capita income is one of the most obvious and striking “stylized facts” of the development process. Urbanization rates increase whenever urban population growth exceeds rural population growth. Generally, the more developed the country, measured by per capita income, the greater the share of population living in urban areas. The black linear fit line in Figure 7.2 shows urbanization versus the log of 2010 GNI per capita; the highest-income countries, such as Japan, are also among the most urbanized, while the very poorest countries, such as Burundi, are among the least urbanized. Urbanization is proceed- ing rapidly. According to UN projections, there will be almost 5 billion urban dwellers by 2030, nearly five-eighths of projected world population for that year. The projected 2030 urban population of Africa of 748 million will be larger than the entire 685 million population of Europe.

At the same time, while individual countries become more urbanized as they develop, today’s poorest countries are far more urbanized than today’s developed countries were when they were at a comparable level of development, as measured by income per capita. Returning to Figure 7.2, the dashed blue linear fit line shows the relationship between income per capita and urbaniza- tion that prevailed in 1960. A comparison of the two lines reveals that for any given income in 2010, a country that had the identical income in 1960 was sig- nificantly less urbanized.

FIGURE 7.2 Relationship between Urbanization and Per Capita GDP, 2010 with Comparison to Relationship in 1960

Sources: Luc Christiaensen, Remi Jedwab, Peter Lanjouw, and Harris Selod, “Urbanization and Poverty Reduction,” draft working paper, 2014 – special thanks to Remi Jedwab. Data sources: Maddison (2008), United Nations (2011) and World Bank (2013). Note: This figure shows the relationship between the urbanization rate and the log per capita of GDP, in 2005 PPP, for 119 developing countries in 2010. The solid line is a linear fit for the data in 2010. The dashed line is a linear fit for the data in 1960 (the scatter plot is not shown for 1960).

76 12111098

0

U rb

an iz

at io

n R

at e

in 2

01 0

(% )

20

40

60

80

100

Log Per Capita GDP in 2010 (PPP, cst 2005$)

Linear Fit in 2010 Linear Fit in 1960

333CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

In recent decades urbanization has continued in nearly all developing countries, even those that have experienced only minimal industrialization. Figure 7.3 shows urbanization over time and across income levels over the quarter century from 1970 to 1995. Each line segment represents the trajectory of one country, starting from the solid dots, which represent the 1970 income and urbanization level for a given country and ending at the end of the line segments (marked by a diamond), which represent the corresponding 1995 income and urbanization level for the same country. Although the World Bank caption to the figure stated that “urbanization is closely associated with eco- nomic growth,” the figure may also be interpreted as showing that urbaniza- tion is occurring everywhere, at high and low levels of income and whether growth is positive or negative. Even when the lines point to the left, indicat- ing shrinking incomes per capita over the period, they still generally point upward, indicating that urbanization continued. In short, urbanization is hap- pening everywhere in the world, although at differing rates.

Thus, it becomes clear that urbanization is not driven solely by income. In addition, some countries with approximately the same income level are signif- icantly more or less urbanized, partly due to differing domestic policies. So we need to consider urbanization carefully—is it only correlated with economic development, or is causation also at work?

Indeed, one of the most significant of all modern demographic phenomena is the rapid growth of cities in developing countries. In 1950, some 275 million people were living in cities in the developing world, 38% of the 724 million total urban population; by 2010, the world’s urban population had surpassed 3.4 billion, with over three-quarters of all urban dwellers living in metropoli- tan areas of low- and middle-income countries.

While in a significant number of cases the speed at which the share of urban population has increased in developing countries in the late twentieth and early twenty-first century is not much faster than in many of the devel- oped countries when they were urbanizing in the late nineteenth century,

FIGURE 7.3 Proportion of Urban Population by Region, 1970–1995

Source: The United Nations is the author of the original material. World Urbanization Prospects: The 2009 Revision. © 2009 United Nations. Reproduced with permission.

100 90 80 70 60 50 40 30 20 10

0 100 1,000 10,000 100,000

GDP capita (1987 U.S. $)

U rb

an p

o p

u la

ti o

n

(% o

f t o

ta l p

o p

u la

ti o

n )

1970 1995

334 PART TWO Problems and Policies: Domestic

nonetheless shares of urban population are being reached, particularly in Africa, at lower levels of per capita income than at a comparable stage in developed countries (again see Figure 7.2). Relatedly, urbanization in Africa is not associated with industrialization, as it was in the now-developed coun- tries. Moreover, in most regions of the developing world, because population is so much larger, the sheer numbers of people coming into the city is unprece- dented. Also unprecedented are the very large sizes of individual cities at such low levels of income per capita. The largest cities in developed countries in the past were much smaller than the large cities of developing countries today.

Although a majority of developing-country urban growth will be found in cities of less than 5 million people, it is also the case that population growth in cities of over 5 million in population is more rapid than growth of smaller cities (under 500,000) in the developing world. In fact, according to the UN, by 2025, only about half the urban population will be in cities with less than a half mil- lion people, the lowest fraction ever. Moreover, the developing world is also coming to dominate the world’s largest cities, including the megacities with over 10 million inhabitants. Figure 7.4 provides a map locating megacities, the

FIGURE 7.4 Megacities: Cities with 10 Million or More Inhabitants

Source: Data drawn from United Nations Population Division, World Urbanization Prospects: The 2011 Revision (New York: United Nations, 2011), at http://esa.un.org/unup/pdf/WUP2011_Highlights.pdf

São Paulo BRAZIL 14.8 19.9 23.2

New York UNITED STATES 16.2 16.1 20.4 23.6

Mexico City

MEXICO

15.3 20.4 24.6

Jakarta INDONESIA

12.8

Manila PHILIPPINES

11.9 16.3

Osaka JAPAN

11.0 11.5 12.0

Shanghai CHINA 20.2 28.4

Beijing CHINA 15.6 22.6

Tokyo JAPAN

23.3 32.5 37.2 38.7

Seoul KOREA

10.5

Wuhan CHINA

12.7

Tianjin CHINA

11.9Los Angeles UNITED STATES

10.9 13.4 15.7

Rio de Janeiro BRAZIL 12.0 13.6

Kolkata INDIA

10.9 14.4 18.7

Mumbai INDIA

12.4 19.7 26.6

Karachi PAKISTAN

13.9 20.2

Lahore PAKISTAN

11.2

Delhi INDIA

22.7 32.9

Chennai INDIA

12.8

Bangalore INDIA

13.2

London UK

10.3

Dhaka BANGLADESH

15.4 22.9

Buenos Aires ARGENTINA

10.5 13.5 15.5

Lima PERU 11.5Bogotá

COLOMBIA 11.4

Hyderabad INDIA

11.6 Chicago

USA

11.4

Krung Thep THAILAND

11.2

Shenzhen CHINA 10.6 15.5

Chongqing CHINA 13.6

Guangzhou CHINA

Guangzhou CHINA 10.8 15.5

1970 1990 2011

Population in millions

2025 (projected)

Istanbul TURKEY 11.3 14.9

Paris FRANCE

10.6 12.2

Lagos NIGERIA 11.2 18.9

Kinshasa DEM. REP. CONGO

14.5

Cairo EGYPT 11.2 14.7

Moscow RUSSIA

11.6 12.6

335CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

largest urban agglomerations in the world containing a population of at least 10 million people. As the figure shows, in 1970, there were only 2 megacities, but by 1990, there were 10, and by 2011, there were 23 such metropolises. Of these, 18 (over three-quarters) were located in the developing world. By 2025, 30 of the 37 megacities (more than 80%) will be in developing countries.

Based on numbers of people, the small and medium cities in developing countries have added more residents than the megacities. But while the num- ber living in cities of fewer than 500,000 will more than double (grow by 2.4 times) from 1970 to 2025, the number in megacities will increase by 16 times, from 39 million to 630 million. Figure 7.5 presents total urban populations in millions by different city sizes for 1970, 1990, and 2011, with projections to 2025. In 2011, more people lived in megacities of over 10 million than in cities from 5 to 10 million people in size. In principle, a megacity could offer large agglomer- ation economies, although congestion costs may rise rapidly. Another potential downside is that megacities tend to be more capital intensive, which does not match with the comparative advantage of most developing countries. Mega- cities, particularly in low-income countries, may also have outsized social and health problems. The relative balance of these factors is likely to differ across countries depending on the forces that led these cities to reach their megascales.

Moreover, as Figure 7.6 shows, going forward almost all of the incre- ments to the world’s population will be accounted for by the growth of urban areas as migrants continue to stream into the cities from rural areas and as

FIGURE 7.5 Total Population in Millions by City Size Class, 1970, 1990, 2011 and 2025

Source: United Nations Population Division, World Urbanization Prospects: The 2011 Revision (New York: United Nations, 2011), http://esa.un.org/unup/pdf/WUP2011_Highlights.pdf.

0 Fewer Than

500,000 10 Million

or more 5 to 10 Million

1 to 5 Million

500,000 to 1 Million

2,000

833

1,333

1,849

1,966

128 206

365

516

244

456

776

1,129

109142

283 402

39 145

359

630

1970 1990 2011 2025

500

1,000

T o

ta l p

o p

u la

ti o

n (

m il

li o

n ) 1,500

336 PART TWO Problems and Policies: Domestic

urbanization rates in the developing world continue to approach those of the developed world.

A central question related to the unprecedented size of these urban agglom- erations is how these cities will cope—economically, environmentally, and politically—with such high and rapidly rising concentrations of people. While it is true that cities offer the cost-reducing advantages of agglomeration econo- mies and economies of scale and proximity, as well as numerous economic and social externalities (e.g., skilled workers, cheap transport, social and cul- tural amenities), for many analysts the social costs of increasingly overloading of housing and social services, not to mention increased crime, pollution, and congestion, can outweigh these historical urban advantages.3

Along with the rapid spread of urbanization and the urban bias in devel- opment strategies has come this prolific growth of huge slums and shanty- towns. From the favelas of Rio de Janeiro and the pueblos jovenes of Lima to the bustees of Kolkata and the bidonvilles of Dakar, such makeshift communities have been growing rapidly. Today, at least one billion people live in urban slum settlements, representing nearly one-third of the urban population in all developing countries.

Figure 7.7 shows the annual growth of urban and slum populations in the 1990–2001 period, drawn from the 2006 UN Millennium Development Goals Report. As the Report summarized:

Sub-Saharan Africa is the world’s most rapidly urbanizing region, and almost all of this growth has been in slums, where new city residents face overcrowding,

Urban bias The notion that most governments in develop- ing countries favor the urban sector in their development policies, thereby creating a widening gap between the urban and rural economies.

FIGURE 7.6 Estimated and Projected Urban and Rural Population of the More and Less Developed Regions, 1950–2050

Source: Millennium Development Goal Report 2010. © 2010 United Nations. Reproduced with permission of United Nations Publications.

6,000

5,000

4,000

3,000

2,000

1,000

0

P o

p u

la ti

o n

( m

il li

o n

s)

20501950 1960 1970 1980 1990 2000 2010 2020 2030 2040

Less developed regions - Urban population Less developed regions - Rural population

More developed regions - Urban population More developed regions - Rural population

337CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

inadequate housing, and a lack of water and sanitation. In Western Asia, as well, most of the urban growth is occurring in slums. The rapid expansion of urban areas in Southern and Eastern Asia is creating cities of unprecedented size and complexity and new challenges for providing a decent environment for the poor. Northern Africa is the only developing region where the quality of urban life is improving.”

The importance of addressing this problem has been enshrined in the Mil- lennium Development Goals (MDGs —see Chapter 1), in which Target 11 of Goal 7 commits “to improve the lives of at least 100 million slum dwell- ers by the year 2020.” Yet even though this number seems likely to be met, it represents only about one-tenth of all urban slum dwellers as of 2013! It is extremely difficult to provide reliable projections of slum populations further into the future because much depends on uncertain future policies and eco- nomic growth rates, as well as the extent of migration that occurs in response to growth and policy change. Allowing that simple extrapolation of trends would tend to significantly overstate the problem, UN-Habitat has noted that such trends would point to a slum population of as large as 3 billion people in 2050.

FIGURE 7.7 Annual Growth of Urban and Slum Populations, 1990–2001

Source: Adapted from United Nations, Millennium Development Goals Report, 2006 (New York: United Nations, 2006), p. 20. Reproduced by permission of United Nations Publications.

R a

te o

f a

n n

u a

l g ro

w th

( %

)

1.3

3.8

2.7 3.0

2.3

3.4

4.6

2.5 2.2

1.3

–0.2

4.5

2.2

2.9

5

4

3

2

1

0

Slum population

Urban population

So ut

he as

ter n A

sia

We ste

rn As

ia

So ut

he rn

As ia

Ea ste

rn As

ia

Su b-

Sa ha

ran Af

ric a

No rth

ern Af

ric a

La tin

Am eri

ca an

d t he

Ca rib

be an

338 PART TWO Problems and Policies: Domestic

Although population growth and accelerated rural-urban migration are chiefly responsible for the explosion in urban shantytowns, part of the blame rests with governments. Their misguided urban-planning policies and out- moded building codes often mean that a majority of new urban housing is “illegal.” For example, colonial-era building codes in Nairobi, Kenya, made it impossible to build an “official” house for less than $3,500. The law has also required every dwelling to be accessible by car. As a result, two-thirds of Nairobi’s land has been occupied by 10% of the population, while many slum dwellings cannot legally be improved. Similarly, in Manila, Philippines, a large majority of the population has historically been too poor to be able to buy or rent an officially “legal” house.4 In fact, a widely held belief in some developing countries is that governments have intentionally sought to make the lives of new migrants as miserable as possible, hoping this will be an effec- tive deterrent to prospective migrants; but when people come to cities any- way, slums are the inevitable result. But often even government’s best efforts to neglect, discriminate against, or even destroy slums are not enough to can- cel out the many other distortions in disregarded, economically stagnant, or socially oppressive rural areas.

Statistics show that rural migrants constitute anywhere from 35% to 60% of recorded urban population growth. About three-quarters of developing coun- tries responding to UN surveys indicated that they had initiated policies to slow down or reverse their accelerating trends in rural-urban migration, and/or desire to do so.5

A critical issue that needs to be addressed is the extent to which national governments can formulate development policies that can have a definite impact on trends in and the character of urban growth. It is clear that the emphasis on industrial modernization, technological sophistication, and met- ropolitan growth created a substantial geographic imbalance in economic opportunities and contributed significantly to the accelerating influx of rural migrants into urban areas. Is it possible or even desirable now to attempt to influence these trends by pursuing a different set of population and develop- ment policies? With birth rates declining in many developing countries, rapid urban growth and accelerated rural-urban migration will undoubtedly be one of the most important development and demographic issues of the coming decades. And in urban areas, the growth and development of the informal sec- tor, as well as its role and limitations for labor absorption and economic prog- ress, will assume increasing importance.

Before examining other problems and policy approaches in developing- country cities more closely, let us first consider the potential advantages offered by cities. Urban areas have played a highly constructive role in the economies of today’s developed countries, and they offer huge and still largely untapped potential to do the same for developing countries. A detailed look at the informal sector in developing cities will give an idea of its potential as an engine of growth. We also consider in more detail what has been dif- ferent—and what has gone wrong—with urban development and the exces- sively rapid pace of rural-urban migration in many developing countries. We conclude with a look at constructive policies to help cities foster successful urban development while at the same time giving more balanced treatment to development in rural areas.

Rural-urban migration The movement of people from rural villages, towns, and farms to urban centers (cities) in search of jobs.

339CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

7.2 The Role of Cities

What explains the strong association between urbanization and development? To a large degree, cities are formed because they provide cost advantages to producers and consumers through what are called agglomeration economies. As noted by Walter Isard, these agglomeration economies come in two forms. Urbanization economies are effects associated with the general growth of a concentrated geographic region. Localization economies are effects captured by particular sectors of the economy, such as finance or automobiles, as they grow within an area. Localization economies often take the form of backward and forward linkages of the type introduced in Chapter 4. When transpor- tation costs are significant, users of the outputs of an industry may benefit from a nearby location to save on these costs. This benefit is a type of forward linkage. In addition, firms of the same or related industries may benefit from being located in the same city, so they can all draw on a large pool of workers with the specific skills used in that sector or from specialized infrastructure. This is a type of backward linkage. Workers with specialized skills appropri- ate to the industry prefer to be located there as well so that they can easily find a new job or be in a position to take advantage of better opportunities.

Industrial Districts

An economic definition of a city is “an area with relatively high population density that contains a set of closely related activities.” Firms often also pre- fer to be located where they can learn from other firms doing similar work. Learning takes place in both formal relationships, such as joint ventures, and informal ones, such as from tips learned in evening social clubs or over lunch. These spillovers are also agglomeration economies, part of the benefits of what Alfred Marshall called “industrial districts,” and they play a big role in Michael Porter ’s “clusters” theory of competitive advantage.6 Firms located in such industrial districts also benefit from the opportunity to contract out work easily when an unusually large order materializes. Thus, a firm of modest size does not have to turn down a big job due to lack of capacity, an arrangement that provides “flexible specialization.”7 Further, firms may wish to operate in well-known districts for the marketing advantages of locating where company procurers and household consumers of their goods know to shop to get the best selection.

It may not matter as much where such industrial districts are located as that they somehow got an early start there, perhaps because of a histori- cal accident. For example, in the United States, many innovative computer firms located in Silicon Valley, California, simply because other such firms were already located there. Analogously, suppliers to shoe firms located in the Sinos Valley in southern Brazil and in Guadalajara in Mexico because so many shoe firms located in those regions. Some of the benefits are gained sim- ply by the fact of location—Khalid Nadvi has termed this “passive collective efficiency”—but other benefits must be achieved through collective action, such as developing training facilities or lobbying government for needed infrastructure as an industry rather than as individual firms (“active collec- tive efficiency”).8

Agglomeration economies Cost advantages to producers and consumers from location in cities and towns, which take the forms of urbanization economies and localization economies.

Urbanization economies Agglomeration effects associ- ated with the general growth of a concentrated geographic region.

Localization economies Agglomeration effects cap- tured by particular sectors of the economy, such as finance or autos, as they grow within an area.

340 PART TWO Problems and Policies: Domestic

A growing body of evidence shows that industrial clusters are increasingly common in developing countries, at stages of industrial development rang- ing from cottage industry to advanced manufacturing techniques, and appear to be significant factors in emerging industrial competitiveness. Neverthe- less, the dynamism of these clusters has varied widely. Some of the identi- fied districts are traditional clusters of artisans that have shown little ability to innovate, export, or expand. But such groupings often remain one-family microenterprises with little division of labor or use of modern techniques. Pro- ducers in a village are better off sharing a common specialization than produc- ing a random assortment of goods, in part because intermediaries work with villages with a high concentration of producers in their sector. But such tra- ditional producers sometimes benefit little from “internal” divisions of labor within the firm, producing a largely complete product within the household and remaining at very low productivity and incomes. For example, a small town in Kenya may have a dozen or more families fabricating wheelbarrows, each family starting with timber and a few simple purchased metal inputs and producing a final product for sale. Nevertheless, clustering can generate more specialized employment in the rural nonfarm sector, as in the rural hand-loom weaver clusters of Ethiopia, in which microentrepreneurs share a work space, take part in a finer division of labor, and benefit from trade credits for work- ing capital. Researchers also found that better electricity reliability and other infrastructure that are available to a cluster lead to better firm performance; in particular, “producers in electrified towns work longer hours than those in towns without electricity.”9

In some cases, traditional township specializations have evolved into more developed clusters, with still modest-size but somewhat larger firms using a more detailed division of labor, such as a group of wheelbarrow producers with some specialization, each employing a few workers. Eventually, the cluster might expand in scope and become a low-tech metal products industrial district selling products throughout the country as the town grows into a small city. These clusters are reminiscent of the industrial districts of developed countries but require that sufficient financing be gathered to invest in core firms using somewhat larger-scale capital goods. But note that clusters of some sophistica- tion can emerge in an otherwise fairly rural but densely populated area. As manufacturing has progressed in China, there has been a dramatic emergence of specialized clusters, to the point where they have now become pervasive, as detailed in Box 7.1.

As Hermine Weijland found in her study of Java, Indonesia, “It needs only a few fortunate years of market expansion to create gains from externalities and joint action.”10 She cites as examples local clusters that have upgraded and now competitively produce such goods as roof tiles, rattan furniture, cast metal, and textiles. Similarly, Dorothy McCormick concluded from a study of six representative clusters in Africa that “groundwork clusters prepare the way; industrializing clusters begin the process of specialization, differentia- tion, and technological development; and complex industrial clusters produce competitively for wider markets.”11 In some cases, the evidence suggests that coordination failures are not overcome, and so there may be a role for gov- ernment policy in encouraging the upgrading of clusters. In other cases, it is the government itself that shares blame for cluster stagnation when it enforces

341CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

irrational and stifling regulations, which are far more damaging than the usual policy of benign neglect toward nascent clusters in the informal sector. Examples of clusters in developing countries that are widely considered suc- cessful include surgical instruments in Sialkot, Pakistan, and software in the Bengaluru (Bangalore) area in India. Clusters of all kinds, however, and par- ticularly those producing for the local market, face substantial challenges from globalization and trade liberalization.

BOX 7.1 FINDINGS The Emergence of Industrial Districts or Clusters in China

Prior to the 1980s, industry in China was stateowned, and factories were dispersed geographically for military defense. Beginning in 1980, Special Eco- nomic Zones such as Shenzhen were created to attract foreign firms in many industries; domestic firms sold inputs to them, but not as clusters. Township and village enterprises (TVEs) then emerged, initiated outside of local governments but “vaguely owned” by them. TVE managers usually tried a variety of activities, and early 1990s field research found little evidence that firms in the same or related industries were locating in close proximity to each other. But starting in the mid-1990s, TVEs rapidly privatized, and a combination of competition, responses to credit constraints, an abundance of entrepreneurial talent, and supportive local policies led to the emer- gence of localized industrial clusters. But like other Chinese institutions (see the case study in Chapter 4), some may ultimately prove “transitional.”

The Zhili Township children’s garment cluster stud- ied by Fleisher and colleagues saw “a significant rise in specialization and outsourcing among firms.” Median investment to start a business more than doubled, but bank loans remained unnecessary as many entrepre- neurs generated sufficient savings. Accordingly, many firms entered, and after 2000, wages rose and profit- ability fell. In response, firms selling directly to mar- kets sought to “signal their commitment to product quality”—nearly half by establishing trademarks and nearly a fifth achieving International Organization for Standardization (ISO) certification. Meanwhile, quality of subcontractors was “monitored by their outsourcing

partners.” Social capital is critical, Fleisher and col- leagues concluded: “Clustering within established communities where long-time relationships among family and neighbors prevail offers an institutional substitute for court enforcement of contractual rela- tionships among borrowers and lenders and between outsourcing firms and their subcontractors.” They also reported that “township government has imposed safety regulations in response to major industrial acci- dents” and helped “prevent a destructive ‘race to the bottom’ in terms of product quality and employee safety” where markets failed to do so.

From firm surveys in the Puyuan cashmere sweater district, Ruan and Zhang found that state-owned banks rarely gave loans to small and medium-size enterprises. But small firms borrowed from relatives and friends and gave and received credit from buyers and sellers, so clusters lowered “capital barriers to entry through the division of labor, enabling individuals to choose the appropriate type of specialization according to their capital portfolio,” while a deeper division of labor allowed “people with different talents and endow- ments to find their own positions.” Similar conclusions followed from a study of the world’s largest footwear cluster in Wenzhou.

With a detailed analysis of 1995 and 2004 firm cen- sus data, Long and Zhang confirmed that “China’s rapid industrialization is marked by increased clustering.” Their research supported the conclusion that cluster- ing of firms relaxed credit constraints through “two mechanisms: (1) within a cluster, finer division of labor lowers the capital barriers to entry, and (2) closer

342 PART TWO Problems and Policies: Domestic

proximity makes the provision of trade credit among firms easier.” They found that clusters use more “entre- preneurs and labor, and less . . . capital, compared to nonclustered large factories” and thus followed com- parative advantage. They noted that clusters could be useful in countries facing a “scarcity of capital and an inefficient financial system.” However, they cautioned, “clustering may be a second-best solution to the financ- ing problem when the local conditions do not permit easy access to regular financing.” Thus clustering, like TVEs, might be a transitional form until financial mar- kets deepened, formal contract enforcement could be provided, and larger investments would be needed.

Sources: Fleisher, Belton, Dinghuan Hu, William McGuire, and Xiaobo Zhang. “The evolution of an industrial cluster in China.” China Economic Review 21, No. 3 (September 2010): 456–469; Huang, Zuhui, Xiaobo Zhang, and Yun- wei Zhu. “The role of clustering in rural industrialization: A case study of Wenzhou’s footwear industry.” China Economic Review 19 (2008): 409–420; Long, Cheryl, and Xiaobo Zhang. “Cluster-based industrialization in China: Financing and performance.” IFPRI Discussion Paper No. 937. Washington, D.C.: International Food Policy Re- search Institute, 2009; Ruan, Jianqing, and Xiaobo Zhang. “Credit constraints, organizational choice, and returns to capital: Evidence from a rural industrial cluster in China.” IFPRI Discussion Paper No. 830. Washington, D.C.: In- ternational Food Policy Research Institute, 2008; Ruan, Jianqing, and Xiaobo Zhang. “Finance and cluster-based industrial development in China.” Economic Development and Cultural Change 58 (2009): 143–164.

Again, not all of the collective efficiency advantages of an industrial dis- trict are realized through passive location. Others are actively created by joint investments and promotional activities of the firms in the district. One fac- tor determining the dynamism of a district is the ability of its firms to find a mechanism for such collective action. While the government can provide financial and other important services to facilitate cluster development, social capital is also critical, especially group trust and a shared history of success- ful collective action, which requires time to develop. Government can help by bringing parties together and helping them gain experience in cooperating on more modest goals before tackling larger ones, but social capital normally grows organically in an economic community and cannot be created by fiat. Even with collective action to supplement passive benefits of agglomeration, traditional clusters may not survive in their current form into more advanced stages of industrialization. Nonetheless, as Hubert Schmitz and Khalid Nadvi note, even if transitional, districts in the informal sector may still play a crucial role in mobilizing underused human and financial resources. They argue that clustering enables entrepreneurs to focus on selected stages of the production process, while other producers focus on their own specialized stages. Thus, even though the overall capital needs of a cluster may be too large for individ- ual investors, each small producer individually needs only raise rather mod- est quantities of investment and working capital.12

Statistical estimates show that benefits of agglomeration can be quite sub- stantial in practice. For example, studies have demonstrated that “if a plant moves from a location shared by 1,000 workers employed by firms in the same industry to one with 10,000 such workers, output will increase an average of 15%, largely because the pool of specialized workers and inputs deepens.” Moreover, “productivity rises with city size, so much so that a typical firm will see its productivity climb 5% to 10% if city size and the scale of local industry double.” 13

Social capital The productive value of a set of social institu- tions and norms, including group trust, expected coopera- tive behaviors with predictable punishments for deviations, and a shared history of success- ful collective action, that raises expectations for participation in future cooperative behavior.

343CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

Efficient Urban Scale

Localization economies do not imply that it would be efficient for all of a country’s industries to be located together in a single city. These economies extend across closely related industries, such as those with strong backward and forward linkages, but there are fewer productivity benefits for unrelated industries to locate together. One notable exception is the potential spillover from technological progress in one industry to its adaptation for different uses in another industry. But there are also some important congestion costs. The higher the urban density, the higher the costs of real estate. It is much more expensive to build vertically than horizontally, increasingly so as skyscraper scale is reached, so that when market forces work properly, tall buildings are built primarily when urban land costs become high. (Note that skyscrapers and other buildings of monumental scale are sometimes built for political show rather than for economic efficiency, such as the world’s tallest build- ings in Dubai, United Arab Emirates; Shanghai, China; Mecca, Saudi Arabia; Taipei, Taiwan; and Kuala Lumpur, Malaysia.) In large urban areas, workers may find themselves with longer and longer commutes and greater transpor- tation costs and may demand higher wages to cover these costs. In addition, the costs of infrastructure such as water and sewer systems are higher in con- centrated urban areas. In theory, if costs of transportation of finished goods are high and consumers wish to be located in the largest city to avoid pay- ing those transportation costs as much as possible, economic activity could become indefinitely concentrated within a city (called the “black hole” effect), but it is generally much less costly to improve the transportation system of a country than to pay the costs of maintaining a gargantuan urban complex. Under competitive forces, and other things being equal, if workers are mobile, a worker in a large city with higher wages but higher costs of living (such as higher housing prices) is no better off in real material terms than a worker with comparable education, experience, ability, and health in a small city who has lower wages and lower costs of living.14

Thus, the concentrating, or “centripetal,” forces of urban agglomeration econ- omies are opposed by the dispersing, or “centrifugal,” forces of diseconomies featuring increasing costs with greater concentration, because some of the factors of production, most obviously land, are not mobile. We can “create” more cen- tral city land by building skyscrapers, but only to a certain scale and only at substantial cost. Thus, it is normal for an economy to have a range of cities, with sizes dependent on the scale of the industries it sponsors and the extent of agglomeration economies found for that industry or cluster of industries.

Two well-known theories of city size are the urban hierarchy model (cen- tral place theory) and the differentiated plane model.15 In the urban hierarchy model, originated by August Losch and Walter Christaller, plants in various industries have a characteristic market radius that results from the interplay of three factors: economies of scale in production, transportation costs, and the way the demand for land is spread over space. The larger the economies of scale in production and the lower the transportation costs, the larger the radius of territory that will be served by that industry to minimize costs. In contrast, if the price of real estate is bid up to high levels in the resulting cities, this will tend to create smaller radii. As a result, small cities contain activities

Congestion An action taken by one agent that decreases the incentives for other agents to take similar actions. Com- pare to the opposite effect of a complementarity.

344 PART TWO Problems and Policies: Domestic

with short market radii, while large cities emerge to contain activities of both small and large radii. Generally speaking, activities of a national scope, such as government and finance, will be located in a single city (though not neces- sarily the same large city because of the effect of congestion costs). Clearly, the urban hierarchy approach applies better to nonexport industries than to export industries. When countries have different specializations in the inter- national market or are at different stages of economic development, the size distribution of cities may potentially differ. For example, a developing coun- try that still overwhelmingly specializes in agriculture might reasonably have one or two large cities serving national industries such as finance and gov- ernment and many smaller towns serving local agricultural areas. A coun- try with a highly differentiated manufacturing and service base might have a large number of medium-size cities.

In the differentiated plane model, originated by Alfred Weber, Walter Isard, and Leon Moses, the limited number of transportation routes linking the indus- tries within an economy plays a key role. The model predicts urban concentra- tions at the points where the scarce transportation routes cross, called “internal nodes.” The hierarchy of urban sizes depends on the pattern of nodes and the industrial mix. Primary processing industries have few inputs and are usually located near the source of the primary resource. However, there will also be incentives for industries with strong backward or forward linkages to locate in the same city.

Of course, there is nothing inherently wrong with very large cities per se—even megacities have some special productive advantages in a global economy.15a But the distortions that have led to the outsize cities prevalent in developing countries have been costly and problematic.

7.3 The Urban Giantism Problem

In the case of developing countries, the main transportation routes are often a legacy of colonialism. Theorists of the dependence school (see Chapter 3) have compared colonial transportation networks to drainage systems, emphasizing ease of extraction of the country’s natural resources. In many cases, the capital city will be located near the outlet of this system on the seacoast. This type of transportation system is also called a “hub-and-spoke system,” which is especially visible when the capital city is located in the interior of the country. Many nations inherited a hub-and-spoke system from colonial times, includ- ing many in Africa and Latin America, which also facilitated movement of troops from the capital to the outlying towns to suppress revolts.

The differentiated plane approach emphasizes the lasting impact of histori- cal accidents. In this case, it helps explain where the most oversized cities are found in the developing world and suggests where policies of urban decentral- ization may be most helpful. Note that not all countries inherited such a hub- and-spoke system; Germany did not; the United States did not, in part because it is the result of the merger of 13 separate British colonies, which retained some measure of local autonomy, as do the federal states of Germany. The recent development of the United States makes the emergence of cities such as Atlanta from the crossing of transportation routes especially clear, but the

345CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

same principle has applied elsewhere over longer historical periods. Of course, as nations become wealthy, they generally build better transportation systems.

Sometimes one urban core becomes too large to keep the costs of the indus- tries located there to a minimum. In developed countries, other cores are often developed within the broad metropolitan region, enabling the region as a whole to continue to receive benefits of agglomeration while lowering some of the costs; or new cities may develop in entirely different parts of the country. But this creation of new urban cores does not happen automatically if there are advantages to locating where other firms and residents are already present. This is another chicken-and-egg coordination problem of the type described in Chapter 4. Who will be the pioneer if it is less costly to stay where you are and wait for other pioneers to settle in the new city first? In economic terms, the agglomeration economies of cities are externalities, which must somehow be internalized or the market will fail. How can this be done?

In the United States, developers frequently internalize the externality by creating a new “edge city” within a metropolitan area, financing and build- ing a new center where land is still relatively inexpensive, perhaps 10 to 50 kilometers from the original urban core. This takes place within a context of public oversight in the form of zoning regulations and inducements such as tax breaks. In developing countries, however, capital markets generally do not work well enough for this process of development to take place. In Europe, the public sector plays a much larger role in coordinating new towns and large developments.

In developing countries, however, governments are less involved in the dis- persal of economic activity to more manageable sizes or, if they are involved, are often less effective. For example, government may seek to disperse industry without regard to the nature of agglomeration economies, giving incentives for dispersal but no attention to clustering relevant industries together, a problem seen in industrial parks in Pakistan. And all too often, the incentives are for firms to concentrate in the capital city or other “urban giants.” A key problem of countries such as Peru and Argentina is that their giant capitals suffer from enormous levels of congestion, but adequate midsize cities that might provide alternative locations for growth are lacking. A well-designed infrastructure development program, including more efficient links between medium-size cit- ies and better roads, utilities, and telecommunications within these cities, can help alleviate this problem.

A more detailed comparison of North and South America is instructive. The largest urban area in the United States, the New York metropolitan area, has about 6% of the national population. Toronto, the largest metropolitan area in Canada, has about 5 million residents, some 15% of the Canadian population. But Mexico City holds nearly one-fifth of the population of Mexico, Monte- video nearly half of the population of Uruguay, Lima over one-quarter of the population of Peru, and Buenos Aires and Santiago close to a third of the pop- ulations of Argentina and Chile, respectively.16

First-City Bias

A form of urban bias that has often caused considerable distortions might be termed first-city bias. The country’s largest or first (“first-place”) city receives a

346 PART TWO Problems and Policies: Domestic

disproportionately large share of public investment and incentives for private investment in relation to the country’s second-largest city and other smaller cities. As a result, the first city receives a disproportionately—and ineffi- ciently—large share of population and economic activity.

Table 7.1 shows the largest and second-largest cities in the United States, Canada, and major Latin American countries. Notice that in all of the outsized capital cities—Buenos Aires, Santiago, Mexico City, and Lima—the first city also serves as the capital. Some other developing countries have remarkably outsized first cities, notably Thailand, where Bangkok has a population about 20 times the size of the second city. Further examples can be found in the Phil- ippines (where Manila has over 7 times the population of the second city) and Congo (where Kinshasa has more than 5 times the second city’s population). There are at least 10 other examples of relatively large first (primate) cities in developing nations with sizable populations.17

Causes of Urban Giantism

Why have first cities often swelled to such a large multiple of second cities in developing countries? Overall, urban giantism probably results from a com- bination of a hub-and-spoke transportation system and the location of the political capital in the largest city. This is further reinforced by a political cul- ture of rent seeking and the capital market failures that make the creation of new urban centers a task that markets cannot complete. Other more detailed explanations also generally involve unfortunate consequences of political economy (see Chapter 11). One argument, featured in the work of Paul Krug- man, stresses that under import substitution industrialization (see Chapter 12), with a high level of protection, there is much less international trade, and population and economic activity have an incentive to concentrate in a single city, largely to avoid transportation costs. Thus, firms wish to set up opera- tions in the city where the most consumers already live, which attracts more people to the region in search of jobs and perhaps lower prices (made pos- sible because there are fewer transport costs to be passed on to consumers and perhaps by economies of larger store size and specialized sales districts);

TABLE 7.1 Population of the Largest and Second-Largest Cities in Selected Countries (Millions)

Source: From UN World Urbanization Prospects 2009 Revision.

Country Largest-City Population Second-Largest-City Population Ratio

Canada Toronto, 5.035 Montreal, 3.603 1.40

United States New York, 18.727 Los Angeles, 12.303 1.52

Argentina Buenos Aires, 12.551 Cordoba, 1.423 8.82

Brazil São Paulo, 18.647 Rio de Janeiro, 11.368 1.64

Chile Santiago, 5.605 Valparaiso, 0.837 6.70

Mexico Mexico City, 18.735 Guadalajara, 4.057 4.62

Peru Lima, 8.081 Arequipa, 0.732 11.04

Note: Definitions of city size differ across studies.

347CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

this concentration in turn attracts still more firms and consumers in a circle of causation. However, when trade barriers are reduced, the incentive to focus production on the home market is also reduced, and exporters and their sup- pliers have much less incentive to be located in the country’s biggest popula- tion center. This moves production toward ports and borders, or elsewhere in the country, to escape the excessive congestion costs of the largest city.18

Another explanation for urban giants focuses on the consequences of dictators’ efforts to remain in power. As Figure 7.8 shows, on average, a much larger share of a country’s urbanized population (37%) lives in the first city in unstable dictatorships than in stable democracies (23%). In interpreting this finding, Alberto Ades and Edward Glaeser argue that unstable dictatorships (fearing overthrow) must provide “bread and circuses” for the first city (usu- ally the capital) to prevent unrest; this extreme urban bias in turn attracts more migrants to the favored city and a still larger need for bread and circuses. It should be noted that although the authors attempt to control for reverse cau- sality, it may still be the case that unstable dictatorships also tend to emerge in countries with high first-city concentrations.19

In the developing world, until recently, relatively few countries were effec- tive democracies. Until the democratization waves beginning in the 1980s, most developing countries had authoritarian governments of one form or another. To remain in power and prevent popular uprisings and coups, which were gener- ally thought to be most threatening when launched from the capital city, gov- ernments had an incentive to “buy off” the population of the largest city. This focus of national government spending on the capital city is the bread-and- circuses effect, recalling the phrasing of “rent-sharing” policies in ancient Rome in its period of expansion. The availability of better opportunities, whether the equivalent of the grain handouts in ancient Rome or jobs, wages, infrastructure,

FIGURE 7.8 Politics and Urban Concentration

Source: Data from Alberto F. Ades and Edward L. Glaeser, “Trade and circuses: Explain- ing urban giants,” Quarterly Journal of Economics 110 (1995): 196. Copyright © 1995 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology. Note: N = number of countries in group.

50

45

35

40

30

25

20

15

10

5

0

P o

p u

la ti

o n

li vi

n g

i n

la rg

e st

c it

y (%

o f

u rb

a n

p o

p u

la ti

o n

)

23%

Stable democracies

(N = 2 4 )

35%

Unstable democracies

(N = 6 )

30%

Stable dictatorships

(N = 1 6 )

37%

Unstable dictatorships

(N = 3 9 )

348 PART TWO Problems and Policies: Domestic

and other government services concentrated in the capital city of many of today’s developing countries, attracts an ever-growing migrant population, in turn leading to larger precautionary government spending as the fear of politi- cal instability grows.

Another political economy factor contributes to capital city giantism: It becomes advantageous for firms to be located where they have easy access to government officials, to curry political favor from a regime that can be induced to give companies special favors for a price or that simply demands bribes to function at all. The resulting first-city giantism may be viewed as a form of underdevelopment trap, which may be escaped fully only with a return to democratic rule together with a better balance of incentives to compete for exports as well as home consumption. Democracy does not eliminate political benefits of location in the national capital, but while lobbyists still congregate in the political capital, there may be less incentive for production to become overconcentrated there. Moreover, a free press tends to expose corruption and generate public pressure to root it out, as recent experience in many democra- tizing countries in Latin America and East Asia makes clear.

The explanations for urban giantism—production for the home market in the face of high protection and transport costs, few adequate smaller cities as alternative locations for firms reflecting infrastructure patterns, location of the capital in the largest city, and the political logic of unstable dictatorships—are complementary and help explain some of the advantages of democracies with more balanced economic policies, including well-planned investments in infra- structure. Such countries are able to avoid some of the costs of urban giantism.

Finally, special factors may lead to high costs of doing business elsewhere in the country. There is an incentive to locate in the capital where personal secu- rity is highest in countries in or emerging from conflict such as the Democratic Republic of Congo. And firms may be responding primarily to costs and risks resulting from extortion, greater corruption, or civil unrest in rural areas and small cities, as well as bad infrastructure. The swelling of the urban giant can therefore also be a symptom of binding constraints on development elsewhere in the country that growth diagnosticians can learn from (see Chapter 4). This may suggest priority policies to help overcome a nation’s particular problems of high costs of operating outside the primate city. In recent years Mexico City has been growing more slowly than the Mexican population as a whole, so that its share of the national population is also slowly becoming reduced.

With our better understanding of the causes of outsized primate cities, it becomes clear that this feature is not inevitable. Indeed, if trends toward greater democracy, reduced incidence of coups, increased outward-looking policies, and improved prospects of solving and preventing civil conflicts are maintained, the ratios of largest to second-largest cities where urban giantism has prevailed are likely to continue to decrease.

7.4 The Urban Informal Sector

As noted in Chapter 3, a focus of development theory has been on the dual- istic nature of developing countries’ national economies—the existence of a modern urban capitalist sector geared toward capital-intensive, large-scale

349CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

production and a traditional rural subsistence sector geared toward labor- intensive, small-scale production. This dualistic analysis has also been applied specifically to the urban economy, which has been decomposed into a formal and an informal sector.

The existence of an unorganized, unregulated, and mostly legal but unregis- tered informal sector was recognized in the 1970s, following observations in sev- eral developing countries that massive additions to the urban labor force failed to show up in formal modern-sector unemployment statistics. The bulk of new entrants to the urban labor force seemed to create their own employment or to work for small-scale family-owned enterprises. The self-employed were engaged in a remarkable array of activities, ranging from hawking, street vending, letter writing, knife sharpening, and junk collecting to selling fireworks, prostitution, drug peddling, and snake charming. Others found jobs as mechanics, carpenters, small artisans, barbers, and personal servants. Still others were highly successful small-scale entrepreneurs with several employees (mostly relatives) and higher incomes. Some could even eventually graduate to the formal sector, where they became legally registered, licensed, and subject to government labor regulations. With the unprecedented rate of growth of the urban population in developing countries expected to continue and with the increasing failure of the rural and urban formal sectors to absorb additions to the labor force, more attention is being devoted to the role of the informal sector in serving as a panacea for the growing unemployment problem.

The informal sector continues to play an important role in developing coun- tries, despite decades of benign neglect and even outright hostility. In many developing countries, about half of the employed urban population works in the informal sector. Figure 7.9 shows the relative importance of informal unem- ployment in selected cities. Most of these cities reflect the typical range of infor- mal-sector employment share, from about 30% to 70%. (The only exception is Ljubljana, a virtually developed city near Austria and Italy.) We find a similar pattern of high informal-sector employment in cities throughout the develop- ing world. For example, in India, the urban informal sector comprises 28.5% of employment in Kolkata, 46.5% in Ahmedabad, 49.5% in Mumbai, 53.8% in Chennai, 61.4% in Delhi, and 65.5% in Bangaluru.

The informal sector is characterized by a large number of small-scale pro- duction and service activities that are individually or family-owned and use simple, labor-intensive technology. They tend to operate like monopolistically competitive firms with ease of entry, excess capacity, and competition driv- ing profits (incomes) down to the average supply price of labor of potential new entrants. The usually self-employed workers in this sector have less for- mal education, are generally unskilled, and lack access to financial capital. As a result, worker productivity and income tend to be lower in the informal sector than in the formal sector. Moreover, workers in the informal sector do not enjoy the measure of protection afforded by the formal modern sector in terms of job security, decent working conditions, and old-age pensions. Many workers entering this sector are recent migrants from rural areas unable to find employ- ment in the formal sector. Their motivation is often to obtain sufficient income for survival, relying on their own indigenous resources to create work. As many members of the household as possible are involved in income-generating activities, including women and children, and they often work very long

Informal sector The part of the urban economy of developing countries charac- terized by small competitive individual or family firms, petty retail trade and services, labor-intensive methods, free entry, and market-determined factor and product prices.

350 PART TWO Problems and Policies: Domestic

hours. A large fraction inhabit shacks and small cinder-block houses that they themselves have built in slums and squatter settlements, which generally lack minimal public services such as electricity, water, drainage, transportation, and educational and health services. Many are vulnerable to cyclones (hurricanes), storm surges, mudslides, and other disasters caused by extreme weather—of the type predicted to substantially worsen with climate change (see Chapter 10). Others are even less fortunate, homeless, and living on the pavements. They find sporadic temporary employment in the informal sector as day laborers and hawkers, but their incomes are insufficient to provide even the most rudi- mentary shelter.

Policies for the Urban Informal Sector

In terms of its relationship with other sectors, the informal sector is linked with the rural sector in that it allows excess labor to escape from extreme rural pov- erty and underemployment, although under living and working conditions and for incomes that are often not much better. It is closely connected with the formal urban sector: The formal sector depends on the informal sector for cheap inputs and wage goods for its workers, and the informal sector in turn depends on the growth of the formal sector for a good portion of its income and clientele.

FIGURE 7.9 Importance of Informal Employment in Selected Cities

Source: UN-Habitat, “State of the World’s Cities, 2001,” http://www.unchs.org/Istanbul+5/statereport.htm. Reprinted with permission.

Ljubljana, Slovenia Moscow, Russian Fed.

Aden, Yemen Jakarta, Indonesia

Montevideo, Uruguay Belgrade, Serbia

Bishkek, Kyrgyzstan Zagreb, Croatia

Veliko Tarnovo, Bulgaria Cuenca, Ecuador

Nouakchott, Mauritania Entebbe, Uganda

Recife, Brazil Córdoba, Argentina

Harare, Zimbabwe Douala, Cameroon

Guayaquil, Ecuador Marinilla, Colombia Dhaka, Bangladesh

N’Djamena, Chad León, Nicaragua

Lilongwe, Malawi Surabaya, Indonesia

Conakry, Guinea

0 10 20 30 40 50 60 70 80 Share of the employed population in the informal sector (%)

351CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

Informal-sector incomes have remained persistently higher than those in the poorest rural regions, despite the continued flow of rural-urban migra- tion. The Nobel laureate Sir Arthur Lewis in the 1950s viewed traditional-sec- tor workers, petty traders such as newspaper hawkers, as unproductive and essentially engaged in distractions from the main urban work of industrializa- tion. But if wages are persistently higher in very competitive activities such as urban informal work than in rural work, this likely reflects higher productivi- ties as well. Consequently, a revisionist view espousing the constructive role of cities (that includes their informal sectors) in economic development has taken hold. This approach has been championed by the Dar es Salaam–based UN-Habitat, in its “State of the World’s Cities” reports.20 The 2001 report systematically criticized what it termed the “anti-urban bias of the develop- ment agencies.” Acting on the strong development tradition beginning with the Lewis skepticism of the urban informal sector, developed with the Todaro migration model (examined later in this chapter) emphasizing the negative consequences of urban bias for both efficiency and equity, continuing with the influential work of the integrated rural development school of the 1970s and recast and reemphasized under the Wolfensohn and subsequent presidencies at the World Bank, development agencies have indeed stressed rural develop- ment rhetorically. Many scholars have concluded, however, that this rhetoric often goes untranslated into real resources for the rural areas so that any pro- rural bias of development agencies is typically little more than a partial cor- rection to the overriding forces for urban bias. However, the renewed focus on the development role of cities is an important trend. Besides UN-Habitat, the World Bank and other agencies have placed increasing emphasis on improved urban development.21 The new focus is on how to make cities in developing countries more dynamic engines of growth and more livable environments, and it promises to be one of the more important streams of emerging research and policy analysis in economic development in coming years. In any case, while medium-size cities undoubtedly deserve greater attention for the con- structive role they play in the development process, this does not obviate the problem of overconcentration of activities in first-city urban giantism.

The important role that the informal sector plays in providing income opportunities for the poor is clear. There is some question, however, as to whether the informal sector is merely a holding ground for people awaiting entry into the formal sector and as such is a transitional phase that must be made as comfortable as possible without perpetuating its existence until it is itself absorbed by the formal sector or whether it is here to stay and should in fact be promoted as a major source of employment and income for the urban labor force, or some combination. The answer is likely to differ by country. A 2012 study by Isabel Günther and Andrey Launov found that for the case of Cote d’Ivoire, about half of those working in the informal sector fell into each category of “opportunity” or “last resort.”22

In support of the latter view, the formal sector in developing countries often has a small base in terms of output and employment. To absorb future additions to the urban labor force, the formal sector must be able to gener- ate employment at a very high rate. This means that output must grow at an even faster rate, since employment in this sector increases less than propor- tionately in relation to output. This sort of growth seems highly unlikely in

352 PART TWO Problems and Policies: Domestic

view of current trends. Thus, the burden on the informal sector to absorb more labor will continue to grow unless other solutions to the urban unemployment problem are provided. But young people face increasingly difficult job pros- pects, as can be seen in Figure 7.10.

The informal sector has demonstrated its ability to generate employment and income for the urban labor force. As pointed out earlier, it is already absorbing an average of 50% of the urban labor force. Some studies have shown the informal sector generating almost one-third of urban income.

Several other arguments can be made in favor of promoting the informal sector. First, scattered evidence indicates that the informal sector generates sur- pluses even in a hostile policy environment that denies it access to the advan- tages offered to the formal sector, such as credit, foreign exchange, and tax concessions. Thus, the informal sector’s surplus could provide an impetus to growth in the urban economy. Second, as a result of its low capital intensity, only a fraction of the capital needed in the formal sector is required to employ a worker in the informal sector, offering considerable savings to developing countries so often plagued with capital shortages. Third, by providing access to training and apprenticeships at substantially lower costs than provided by for- mal institutions and the formal sector, the informal sector can play an important role in the formation of human capital. Fourth, the informal sector generates demand for semiskilled and unskilled labor, whose supply is increasing in both

FIGURE 7.10 Youth Unemployment Rates, 1995 and 2005

Source: Adapted from United Nations, Millennium Development Goals Report, 2006 (New York: United Nations, 2006), p. 24. Reproduced by permission of United Nations Publications.

Yo u

th u

n e

m p

lo ym

e n

t ra

te (

% )

33.9 34.5

30

35

40

25

20

15

10

5

0

1995

2005

19.4 18.1

CI S (

for me

r S ov

iet U

nio n)

9.7

17.0

So ut

he as

ter n A

sia

14.2 15.2

La tin

Am eri

ca an

d t he

Ca rib

be an

9.4 11.3

So ut

he rn

As ia

7.2 7.8

Ea ste

rn As

ia

7.9 6.6

Oc ea

nia

15.8 13.8

De ve

lop ed

re gio

ns

12.1 13.7

Wo rld

18.0 18.3

Su b-

Sa ha

ran Af

ric a

20.8

23.6

We ste

rn As

ia

No rth

ern Af

ric a

353CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

relative and absolute terms and is unlikely to be absorbed by the formal sector with its increasing demands for a skilled labor force. Fifth, the informal sector is more likely to adopt appropriate technologies and make use of local resources, allowing for a more efficient allocation of resources. Sixth, the informal sector plays an important role in recycling waste materials, engaging in the collection of goods ranging from scrap metals to cigarette butts, many of which find their way to the industrial sector or provide basic commodities for the poor. Finally, promotion of the informal sector would ensure an increased distribution of the benefits of development to the poor, many of whom are concentrated in the informal sector.

Promotion of the informal sector is not, however, without its disadvan- tages. One of the major disadvantages in promoting the informal sector lies in the strong relationship between rural-urban migration and labor absorp- tion in the informal sector. Migrants from the rural sector have both a lower unemployment rate and a shorter waiting period before obtaining a job in the informal sector. Promoting income and employment opportunities in the informal sector could therefore aggravate the urban unemployment problem by attracting more labor than either the desirable parts of the informal or the formal sector could absorb. Furthermore, there is concern over the environ- mental consequences of a highly concentrated informal sector in the urban areas. Many informal-sector activities cause pollution and congestion (e.g., pedicabs) or inconvenience to pedestrians (e.g., hawkers and vendors). More- over, increased densities in slums and low-income neighborhoods, coupled with poor urban services, could cause enormous problems for urban areas. Any policy measures designed to promote the informal sector must be able to cope with these various problems. Finally, it is an almost universal observa- tion that when regular formal-sector employment becomes available, many informal-sector microentrepreneurs switch sectors to take these jobs—clear evidence of “revealed preference.”

The International Labor Organization has made some general suggestions as to what sorts of measures might be adopted to promote the informal sector. To begin with, governments will have to abandon their hostility toward the informal sector and adopt a more positive and sympathetic posture. For exam- ple, in Latin America, although improving in many cases, bureaucratic red tape and an inordinate number of administrative procedures needed to reg- ister a new business result in delays of up to 240 days in Ecuador, 310 days in Venezuela, and 525 days in Guatemala. Until recently, Brazil, Mexico, and Chile all required more than 20 applications before a company could be approved to do business. Such procedures not only cause excessive delays but can also inflate the costs of doing business by up to 70% annually. So informal- sector businesses simply skirt the law. Fortunately, there has been progress in improving these policies; the 2013 Doing Business annual report explains that, “in the past 8 years the start-up process received more attention from policy makers than any other area of business regulation tracked by Doing Business— through 368 reforms in 149 economies. These worldwide efforts reduced the average time to start a business from 50 days to 30 and the average cost from 89% of income per capita to 31%. But other metrics indicate less progress.”23

Because access to skills plays an important role in determining the struc- ture of the informal sector, governments should facilitate training in the areas

354 PART TWO Problems and Policies: Domestic

that are most beneficial to the urban economy. In this way, the government can play a role in shaping the informal sector so that it contains production and service activities that provide the most value to society. Specifically, such mea- sures might promote legal activities and discourage illegal ones by providing proper skills and other incentives. They could also generate taxes that now go unpaid.

The lack of capital is a major constraint on activities in the informal sector. The provision of credit would therefore permit these enterprises to expand, produce more profit, and hence generate more income and employment. Microfinance institutions have been leading the way in providing enhanced credit access (see Chapter 15). Access to improved technology would have similar effects. Providing infrastructure and suitable locations for work (e.g., designating specific areas for stalls) could help alleviate some of the environ- mental and congestion consequences of an expanded informal sector. Finally, better living conditions must be provided, if not directly, then by promoting growth of the sector on the fringes of urban areas or in smaller towns where the population will settle close to its new area of work, away from the urban density. Promotion of the informal sector outside the urban areas may also help redirect the flow of rural-urban migration, especially if carried out in conjunction with the policies discussed later in this chapter.

Women in the Informal Sector

In some regions of the world, women predominate among rural-urban migrants and may even comprise the majority of the urban population. Though historically many of these women are simply accompanying their spouses, a growing number of women in Latin America, Asia, and Africa migrate to seek economic opportunity. With the exception of the export enclaves of East Asia and a few other cities, where everything from computers to clothing and run- ning shoes are manufactured, only a small minority of these migrants is able to find employment in the formal sector, which is generally dominated by men. As a consequence, women often represent the bulk of the informal-sector labor supply, working for low wages at unstable jobs with no employee or social security benefits. The increase in the number of single female migrants has also contributed to the rising proportion of urban households headed by women, which tend to be poorer, experience tighter resource constraints, and retain relatively high fertility rates. The changing composition of migration flows has important economic and demographic implications for many urban areas of the developing world.

As UN-Habitat noted for its State of Women in Cities 2012/2013:

Urban women supposedly enjoy greater social, economic, political opportunities and freedoms than their rural counterparts. However, the notable gender gaps in labor and employment, decent work, pay, tenure rights, access to and accumula- tion of assets, personal security and safety and representation in formal structures of urban governance, show that women are often the last to benefit from the pros- perity of cities.24

Because members of female-headed households are generally restricted to low-productivity, informal-sector employment and experience higher

355CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

dependency burdens, they are more likely to be poor and malnourished and less likely to obtain formal education, health care, or clean water and sanitation, often remaining effectively excluded from government services. Dropout rates among children from households headed by women are much higher because the children are more likely to be working to contribute to household income.

Many women run small business ventures or microenterprises that require little or no start-up capital and often involve the marketing of homemade foodstuffs and handicrafts. Though women’s restricted access to capital leads to high rates of return on their tiny investments, the extremely low capital- labor ratios confine women to low-productivity undertakings. Studies in Latin America and Asia have found that where credit is available to women with informal-sector microenterprises, repayment rates have equaled or exceeded those for men (see Chapter 15) . And because women are able to make more productive use of capital and start from a much lower investment base, their rates of return on investments often surpass those for men.

Despite the impressive record of these credit programs, they remain lim- ited. The majority of institutional credit is still channeled through formal-sec- tor agencies, and as a result, women generally find themselves ineligible for even small loans. Government programs to enhance income in poor house- holds will inevitably neglect the neediest households so long as governments continue to focus on formal-sector employment of men and allocation of resources through formal-sector institutions. To solve the plight of poor urban women and their children, it is imperative that efforts be made to integrate women into the economic mainstream. Ensuring that women benefit from development programs will require that women’s special circumstances be considered in policy design.

The legalization and economic promotion of informal-sector activities, where the majority of the urban female labor force is employed, could greatly improve women’s financial flexibility and the productivity of their ventures. However, to enable women to reap these benefits, governments must repeal laws that restrict women’s rights to own property and conduct financial trans- actions. Likewise, barriers to women’s direct involvement in technical training programs and extension services must be eradicated. Finally, the provision of affordable child care and family-planning services would lighten the burden of women’s reproductive roles and permit them a greater degree of economic participation.

7.5 Migration and Development

As noted earlier in the chapter, rural-urban migration has been dramatic, and urban development plays an important role in economic development. Rates of rural-urban migration in developing countries have exceeded rates of urban job creation and thus have surpassed greatly the absorption capacity of both industry and urban social services.

Migration worsens rural-urban structural imbalances in two direct ways. First, on the supply side, internal migration disproportionately increases the growth rate of urban job seekers relative to urban population growth, which

356 PART TWO Problems and Policies: Domestic

itself is at historically unprecedented levels because of the high proportion of well-educated young people in the migrant system. Their presence tends to swell the urban labor supply while depleting the rural countryside of valu- able human capital. Second, on the demand side, urban job creation is gener- ally more difficult and costly to accomplish than rural job creation because of the need for substantial complementary resource inputs for most jobs in the industrial sector. Moreover, the pressures of rising urban wages and compul- sory employee fringe benefits in combination with the unavailability of appro- priate, more labor-intensive production technologies means that a rising share of modern-sector output growth is accounted for by increases in labor produc- tivity. Together this rapid supply increase and lagging demand growth tend to convert a short-run problem of resource imbalances into a long-run situation of chronic urban surplus labor.

But the impact of migration on the development process is much more per- vasive than its exacerbation of urban unemployment and underemployment. In fact, the significance of the migration phenomenon in most developing coun- tries is not necessarily in the process itself or even in its impact on the sectoral allocation of human resources. Rather, its significance lies in its implications for economic growth in general and for the character of that growth, particularly its distributional manifestations.

We must therefore recognize that migration in excess of job opportunities is both a symptom of and a contributor to underdevelopment. Understand- ing the causes, determinants, and consequences of internal rural-urban labor migration is thus central to understanding the nature and character of the development process and to formulating policies to influence this process in socially desirable ways. A simple yet crucial step in underlining the centrality of the migration phenomenon is to recognize that any economic and social policy that affects rural and urban real incomes will directly or indirectly influence the migration process. This process will in turn tend to alter the pat- tern of sectoral and geographic economic activity, income distribution, and even population growth. Because all economic policies have direct and indi- rect effects on the level and growth of urban or rural incomes, or both, they all will have a tendency to influence the nature and magnitude of the migration stream. Some policies may have a more direct and immediate impact, such as wages and income policies and employment promotion programs. There are other policies that, though less obvious, may in the long run be no less important. Included among these policies, for example, would be land tenure arrangements; commodity pricing policies; credit allocation; taxation; export promotion; import substitution; commercial policies; the geographic distri- bution of social services; the nature of public investment programs; attitudes toward private foreign investors; the organization of population and family- planning programs; the structure, content, and orientation of the educational system; the functioning of labor markets; and the nature of public policies toward international technology transfer and the location of new industries. There is thus a clear need to recognize the central importance of internal and, for many countries, even international migration and to integrate the two-way relationship between migration and population distribution on the one hand and economic variables on the other into a more comprehensive framework designed to improve development policy formulation.

357CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

In addition, we need to understand better not only why people move and what factors are most important in their decision-making process but also what the consequences of migration are for rural and urban economic and social development. If all development policies affect migration and are affected by it, which are the most significant, and why? What are the policy options and trade-offs among different and sometimes competing objectives (e.g., curtailing internal migration and expanding educational opportuni- ties in rural areas)? Part of our task in the following sections will be to seek answers to these and other questions relating to migration, unemployment, and development.

Migration patterns are complex. The most important type of migration from the standpoint of long-run development is rural-urban migration, but a great deal of rural-rural, urban-urban, and even urban-rural migration also takes place. Rural-urban migration is most important because the popula- tion share of cities is growing, despite the fact that fertility is much lower in urban areas, and the difference is accounted for by rural-urban migration. It is also important because of the potential development benefits of economic activity of cities, due to agglomeration economies and other factors. However, urban-rural migration is important to understand because it usually occurs when hard times in cities coincide with increases in output prices from the country’s cash crops, as occurred in Ghana not long ago. Thus, the overall picture is one of a remarkable amount of “churning,” or continuous move- ments of people within developing countries, especially over short distances. These movements contradict the popular image of stasis in traditional societ- ies. The composition of internal migration for several countries is shown in Figure 7.11.

In addition to wage differentials, age, and education, migration is also explained partly by relocation upon remarrying; prior emigration of family members; distance and costs of relocation; occurrence of famine, disease, vio- lence, and other disasters; and relative standing in the origin community, with those lower on the social order more likely to migrate. Migration can also be a form of portfolio diversification for families who seek to settle some members in areas where they may not be affected by economic shocks in the same way as if they had stayed at home.25

7.6 Toward an Economic Theory of Rural-Urban Migration

The economic development of western Europe and the United States was closely associated with the movement of labor from rural to urban areas. For the most part, with a rural sector dominated by agricultural activities and an urban sector focusing on industrialization, overall economic development in these countries was characterized by the gradual reallocation of labor out of agriculture and into industry through rural-urban migration, both internal and international. Urbanization and industrialization were in essence synon- ymous. This historical model served as a blueprint for structural change in developing countries, as evidenced, for example, by the original Lewis theory of labor transfer (see Chapter 3).

358 PART TWO Problems and Policies: Domestic

But the overwhelming evidence of the past several decades, when devel- oping nations witnessed a massive migration of their rural populations into urban areas despite rising levels of urban unemployment and underemploy- ment, lessens the validity of the Lewis two-sector model of development.26 An explanation of the phenomenon, as well as policies to address the resulting problems, must be sought elsewhere. One theory to explain the apparently paradoxical relationship of accelerated rural-urban migration in the context of rising urban unemployment has come to be known as the Todaro migration model and in its equilibrium form as the Harris-Todaro model.27

A Verbal Description of the Todaro Model

Starting from the assumption that migration is primarily an economic phe- nomenon, which for the individual migrant can be a quite rational decision despite the existence of urban unemployment, the Todaro model postulates that migration proceeds in response to urban-rural differences in expected

Todaro migration model A theory that explains rural-urban migration as an economically rational process despite high urban unemployment. Migrants calculate (present value of) urban expected income (or its equivalent) and move if this exceeds average rural income.

Harris-Todaro model An equilibrium version of the Todaro migration model that predicts that expected incomes will be equated across rural and urban sec- tors when taking into account informal-sector activities and outright unemployment.

FIGURE 7.11 Components of Migration in Selected Countries

Source: World Development Report, 1999–2000, “Internal migration and urbanization: Recent contributions and new evidence,” by Robert E. B. Lucas. Copyright 1999 by the World Bank.

100

80

60

40

20

0

Bo tsw

an a

Su da

n (N

or th

)

So ut

h Ko

re a

Et hi

op ia

Ho nd

ur as

In di

a

Th ai

la nd

Cô te

d’ Iv

oi re

Pe ru

Gh an

a

P e

rc e

n t

Rural-urban Rural-rural Urban-urban Urban-rural

359CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

income rather than actual earnings. The fundamental premise is that migrants consider the various labor market opportunities available to them in the rural and urban sectors and choose the one that maximizes their expected gains from migration.

In essence, the theory assumes that members of the labor force, both actual and potential, compare their expected incomes for a given time horizon in the urban sector (the difference between returns and costs of migration) with pre- vailing average rural incomes and migrate if the former exceeds the latter. (See Appendix 7.1 for a mathematical formulation.)

Consider the following illustration. Suppose that the average unskilled or semiskilled rural worker has a choice between being a farm laborer (or work- ing his own land) for an annual average real income of, say, 50 units or migrat- ing to the city, where a worker with his skill or educational background can obtain wage employment yielding an annual real income of 100 units. The more commonly used economic models of migration, which place exclusive emphasis on the income differential factor as the determinant of the decision to migrate, would indicate a clear choice in this situation. The worker should seek the higher-paying urban job. It is important to recognize, however, that these migration models were developed largely in the context of advanced industrial economies and hence implicitly assume the existence of full or near-full employment. In a full-employment environment, the decision to migrate can be based solely on the desire to secure the highest-paid job wher- ever it becomes available. Simple economic theory would then indicate that such migration should lead to a reduction in wage differentials through the interaction of the forces of supply and demand, in areas of both emigration and immigration.

Unfortunately, such an analysis is not realistic in the context of the insti- tutional and economic framework of most developing nations. First, these countries are beset by a chronic unemployment problem, which means that a typical migrant cannot expect to secure a high-paying urban job immediately. In fact, it is much more likely that on entering the urban labor market, many uneducated, unskilled migrants will either become totally unemployed or will seek casual and part-time employment as vendors, hawkers, repairmen, and itinerant day laborers in the urban traditional or informal sector, where ease of entry, small scale of operation, and relatively competitive price and wage determination prevail. In the case of migrants with considerable human capital in the form of a secondary or university certificate, opportunities are much better, and many will find formal-sector jobs relatively quickly. But they con- stitute only a small proportion of the total migration stream. Consequently, in deciding to migrate, the individual must balance the probabilities and risks of being unemployed or underemployed for a considerable period of time against the positive urban-rural real income differential. The fact that a typical migrant who gains a modern-sector job can expect to earn twice the annual real income in an urban area than in a rural environment may be of little consequence if the actual probability of his securing the higher-paying job within, say, a one-year period is one chance in five. Thus, the actual probability of his being successful in securing the higher-paying urban job is 20%, and therefore his expected urban income for the one-year period is in fact 20 units and not the 100 units that an urban worker in a full-employment environment

360 PART TWO Problems and Policies: Domestic

would expect to receive. So with a one-period time horizon and a probability of success of 20%, it would be irrational for this migrant to seek an urban job, even though the differential between urban and rural earnings capacity is 100%. However, if the probability of success were 60% and the expected urban income therefore 60 units, it would be entirely rational for our migrant with his one-period time horizon to try his luck in the urban area, even though urban unemployment may be extremely high.

If we now approach the situation by assuming a considerably longer time horizon—a more realistic assumption, especially in view of the fact that the vast majority of migrants are between the ages of 15 and 24—the decision to migrate should be represented on the basis of a longer-term, more permanent income cal- culation. If the migrant anticipates a relatively low probability of finding regular wage employment in the initial period but expects this probability to increase over time as he is able to broaden his urban contacts, it would still be rational for him to migrate, even though expected urban income during the initial period or periods might be lower than expected rural income. As long as the present value of the net stream of expected urban income over the migrant’s planning horizon exceeds that of the expected rural income, the decision to migrate is justifiable.

Rather than equalizing urban and rural wage rates, as would be the case in a competitive model, we see that rural-urban migration in our model equates rural and urban expected incomes. For example, if average rural income were 60 and urban income were 120, a 50% urban unemployment rate would be necessary before further migration would no longer be profitable. Because expected incomes are defined in terms of both wages and employment prob- abilities, it is possible to have continued migration despite the existence of siz- able rates of urban unemployment. In our example, migration would continue even if the urban unemployment rate were 30% to 40%.

A Diagrammatic Presentation

This process of achieving an unemployment equilibrium between urban expected wages and average rural income rather than an equalized rural- urban wage, as in the traditional neoclassical free-market model, can also be explained by a diagrammatic portrayal of the basic Harris-Todaro model. This is done in Figure 7.12.28 Assume only two sectors, rural agriculture and urban manufacturing. The demand for labor (the marginal product of labor curve) in agriculture is given by the negatively sloped line AA. Labor demand in manu- facturing is given by MM′ (reading from right to left). The total labor force is given by line OAOM. In a neoclassical, flexible-wage, full-employment market economy, the equilibrium wage would be established at W*AW

* M , with OAL

* A

workers in agriculture and OML * M workers employed in urban manufacturing.

All available workers are therefore employed. But what if urban wages are institutionally determined (inflexible down-

ward) as assumed by Todaro at a level WM, which is at a considerable distance above W*A? If for the moment we continue to assume that there is no unem- ployment, OMLM workers would get urban jobs, and the rest, OALM, would have to settle for rural employment at OAW

** A wages (below the free-market

level of OAW * A). So now we have an urban-rural real wage gap of WM - W**A ,

with WM institutionally fixed. If rural workers were free to migrate (as they

Present value The dis- counted value at the present time of a sum of money to be received in the future.

361CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

are almost everywhere except China), then despite the availability of only OMLM jobs, they are willing to take their chances in the urban job lottery. If their chance (probability) of securing one of these favored jobs is expressed by the ratio of employment in manufacturing, LM, to the total urban labor pool, LUS, then the expression

WA = LM LUS 1WM2 (7.1)

shows the probability of urban job success necessary to equate agricultural income WA with urban expected income 1LM>LUS2 1WM2, thus causing a potential migrant to be indifferent between job locations. The locus of such points of indifference is given by the qq′ curve in Figure 7.12.29 The new unem- ployment equilibrium now occurs at point Z, where the urban-rural actual wage gap is WM - WA, OALA workers are still in the agricultural sector, and OMLM of these workers have modern (formal)-sector jobs paying WM wages. The rest, OMLA - OMLM, are either unemployed or engaged in low-income informal- sector activities. This explains the existence of urban unemployment and the private economic rationality of continued rural-to-urban migration, despite this high unemployment. However, although it may be privately rational from a cost-benefit perspective for an individual to migrate to the city despite high unemployment, it can, as will soon become clear, be socially very costly.

There are many ways to extend the model; here we mention four. First, Equation 7.1 simplifies by assuming that those who migrate and do not get a modern job receive no income; but if they instead receive urban informal-sector

A

M' A

q

**

*

* *

*

OA OM

WA

WA

WA

WM

WM

LA

LUS

LA LM LM

E

Z

q

M

A g

ri cu

lt u

ra l w

a g

e r

a te

M a

n u

fa ct

u ri

n g

w a

g e

r a

te

FIGURE 7.12 The Harris-Todaro Migration Model

362 PART TWO Problems and Policies: Domestic

income, we modify expected income accordingly.30 Second, note that if instead of assuming that all urban migrants are the same, we incorporate the reality of different levels of human capital (education), we can understand why a higher proportion of the rural educated migrate than the uneducated—because they have a better chance (a higher probability) of earning even higher urban wages than unskilled migrants.

Third, we often observe that migrants from the same rural region tend to settle in common cities, even the same neighborhoods of cities, that are rela- tively distant from the migrants’ place of origin. In a model proposed by Wil- liam Carrington, Enrica Detragiache, and Tara Vishwanath, earlier migrants create a positive externality for later potential migrants from their home region by lowering their costs of moving by helping with resettlement and lowering their probability of unemployment by providing them with jobs or information about available jobs. Thus, the search for employment, selection into the migration decision, and forward-looking behavior may all be incorpo- rated into an equilibrium migration model.31

Fourth, the Todaro and Harris-Todaro models are relevant to develop- ing countries even if the wage is not fixed by institutional forces, such as a minimum wage. Recent theoretical research on rural-urban migration has confirmed that the emergence of a high modern-sector wage alongside unem- ployment or an urban traditional sector as seen in these models can also result from market responses to imperfect information, cost of labor turnover, efficiency wage payments, and other common features of labor markets.32

To sum up, the Todaro migration model has four basic characteristics:

1. Migration is stimulated primarily by rational economic considerations of relative benefits and costs—mostly financial but also psychological.

2. The decision to migrate depends on expected rather than actual urban-rural real-wage differentials, where the expected differential is determined by the interaction of two variables, the actual urban-rural wage differential and the probability of successfully obtaining employment in the urban sector.

3. The probability of obtaining an urban job is directly related to the urban employment rate and thus inversely related to the urban unemployment rate.

4. Migration rates in excess of urban job opportunity growth rates are not only possible but also rational and even likely in the face of wide urban- rural expected income differentials. High rates of urban unemployment are therefore inevitable outcomes of the serious imbalance of economic oppor- tunities between urban and rural areas in most underdeveloped countries.

Five Policy Implications

Although the Todaro theory might at first seem to devalue the critical impor- tance of rural-urban migration by portraying it as an adjustment mechanism by which workers allocate themselves between rural and urban labor markets, it does have important policy implications for development strategy with regard to wages and incomes, rural development, and industrialization.

First, imbalances in urban-rural employment opportunities caused by the urban bias, particularly first-city bias, of development strategies must be

Labor turnover Worker separations from employers, a concept used in theory that the urban-rural wage gap is partly explained by the fact that urban modern-sector employers pay higher wages to reduce labor turnover rates and retain trained and skilled workers.

Efficiency wage The notion that modern-sector urban employers pay a higher wage than the equilibrium wage rate in order to attract and retain a higher-quality work- force or to obtain higher pro- ductivity on the job.

363CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

reduced. Because migrants are assumed to respond to differentials in expected incomes, it is vitally important that imbalances between economic opportu- nities in rural and urban sectors be minimized. When urban wage rates rise faster than average rural incomes, they stimulate further rural-urban migra- tion in spite of rising levels of urban unemployment. This heavy influx of peo- ple into urban areas not only gives rise to socioeconomic problems in the cities but may also eventually create problems of labor shortages and lack of entre- preneurship in rural areas. Thus, policy distortions that induce more rapid rural-to-urban migration than would otherwise occur generally reduce overall social welfare.

Second, urban job creation is an insufficient solution for the urban unem- ployment problem. The traditional (Keynesian) economic solution to urban unemployment (the creation of more urban modern-sector jobs without simul- taneous attempts to improve rural incomes and employment opportunities) can result in the paradoxical situation in which more urban employment leads to higher levels of urban unemployment! Once again, the imbalance in expected income-earning opportunities is the crucial concept. Because migra- tion rates are assumed to respond positively to both higher urban wages and higher urban employment opportunities (or probabilities), it follows that for any given positive urban-rural wage differential (in most developing coun- tries, urban wages are typically three to four times as large as rural wages), higher urban employment rates will widen the expected differential and induce even higher rates of rural-urban migration. For every new job created, two or three migrants who were productively occupied in rural areas may come to the city. Thus, if 100 new jobs are created, there may be as many as 300 new migrants and therefore 200 more urban unemployed. Hence, a policy designed to reduce urban unemployment may lead not only to higher levels of urban unemployment but also to lower levels of agricultural output due to induced migration.

Third, indiscriminate educational expansion will lead to further migration and unemployment. The Todaro model also has important policy implications for curtailing public investment in higher education. The heavy influx of rural migrants into urban areas at rates much in excess of new employment oppor- tunities necessitates rationing in the selection of new employees. Although within each educational group such selection may be largely random, many observers have noted that employers tend to use educational attainment or number of years of completed schooling as the typical rationing device. For the same wage, they will hire people with more education in preference to those with less, even though extra education may not contribute to better job performance. Jobs that could formerly be filled by those with a primary education (sweepers, messengers, filing clerks, etc.) now require second- ary training; those formerly requiring a secondary certificate (clerks, typists, bookkeepers, etc.) must now have a university degree. It follows that for any given urban wage, if the probability of success in securing a modern-sector job is higher for people with more education, their expected income differential will also be higher, and they will be more likely to migrate to the cities. The basic Todaro model therefore provides an economic explanation for the observed fact in most developing countries that rural inhabitants with more education are more likely to migrate than those with less.

Induced migration Process in which the creation of urban jobs raises expected incomes and induces more people to migrate from rural areas.

364 PART TWO Problems and Policies: Domestic

Fourth, wage subsidies and traditional scarcity factor pricing can be coun- terproductive. As noted in Chapter 5 and Appendix 5.1, a standard economic policy prescription for generating urban employment opportunities is to eliminate factor price distortions by using “correct” prices, perhaps imple- mented by wage subsidies (fixed government subsidies to employers for each worker employed) or direct government hiring. Because actual urban wages generally exceed the market or “correct” wage as a result of a variety of insti- tutional factors, it is often argued that the elimination of wage distortions through price adjustments or a subsidy system will encourage more labor- intensive modes of production. Although such policies can generate more labor-intensive modes of production, they can also lead to higher levels of unemployment in accordance with our argument about induced migration. The overall welfare impact of a wage subsidy policy when both the rural and urban sectors are taken into account is not immediately clear. Much will depend on the level of urban unemployment, the size of the urban-rural expected-income differential, and the magnitude of induced migration as more urban jobs are created.

Finally, programs of integrated rural development should be encouraged. Policies that operate only on the demand side of the urban employment picture, such as wage subsidies, direct government hiring, elimination of fac- tor price distortions, and employer tax incentives, are probably far less effec- tive in the long run in alleviating the unemployment problem than policies designed directly to regulate the supply of labor to urban areas. Clearly, how- ever, some combination of both kinds of policies is most desirable.

Conceptually, it may be useful to think of cities and their surrounding rural areas as integrated systems. There are significant complementarities between town and country (see Chapter 9). Agricultural and raw materials grown and extracted in rural areas are inputs for urban industry. Although there is some urban agriculture, most food consumed in urban areas is grown in agricul- tural regions. Towns are needed to allow sufficient agglomeration economies, as well as economies of scale, to produce and exchange many goods and ser- vices that are needed in rural areas. In turn, when rural incomes grow, mar- kets for urban manufacturers expand. People come from their rural residences to work in the city by the day or the week. City residents temporarily migrate to nearby agricultural regions during peak planting and harvesting seasons. Thus, rural-urban linkages are extensive. And while investment in urban areas can accelerate migration to cities, investment in agriculture can raise produc- tivity and incomes, making labor redundant, and also accelerate migration. As a result, for policy purposes, it may make a great deal of sense to take account of rural impacts when devising urban policies and vice versa.

At the same time, as globalization proceeds (see Chapter 12), cities tend to trade more with other cities, often in distant parts of the world, and less with nearby rural areas. Moreover, cities generally get the upper hand when urban and rural areas are treated as a bloc, reinforcing urban bias. And rural hin- terlands, far from significant cities and from the attention of distant govern- ments, whether national or regional, often suffer from benign neglect at best and systematic exploitation at worst, such as forced sale of food at low prices. Thus, rural areas need to retain their own autonomy, and poverty programs need to be tailored to the needs of rural citizens.

Wage subsidy A govern- ment financial incentive to private employers to hire more workers, as through tax deductions for new job creation.

365CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

Every effort must be made to broaden the economic base of the rural econ- omy. The present unnecessary economic incentives for rural-urban migration can be minimized through creative and well-designed programs of integrated rural development. These should focus on both farm and nonfarm income generation, employment growth, health care delivery, educational improve- ment, infrastructure development (electricity, water, roads, etc.), and the provision of other rural amenities. Successful rural development programs adapted to the socioeconomic and environmental needs of particular coun- tries and regions seem to offer the only viable long-run solution to the prob- lem of excessive rural-urban migration.

To assert, however, that there is an urgent need for policies designed to curb the excessive influx of rural migrants is not to imply an attempt to reverse what some observers have called inevitable historical trends. Rather, the implication of the Todaro migration model is that there is a growing need for a policy package that does not exacerbate these historical trends toward urbanization by artificially creating serious imbalances in economic opportu- nities between urban and rural areas.

7.7 Conclusion: A Comprehensive Urbanization, Migration, and Employment Strategy

Developing-country cities are projected to grow by more than 2 billion people over the next three decades. This presents enormous challenges for the devel- oping world, but at the same time important economic development oppor- tunities. The pattern of urban settlements tends to be very persistent, so the quality of planning now for this enormous transformation will have ramifica- tions for decades to come.

Based on long-term trends, comparisons with developed countries, and still- strong individual incentives, continued urbanization and rural-urban migration are probably inevitable. Urban bias spurs migration, but focused investment in agriculture raises rural productivity sufficiently to require less labor; a major- ity of alternative types of employment expansion tend to be concentrated in urban areas because of agglomeration effects. Moreover, as education increases in rural areas, workers gain the skills they need, and perhaps the rising aspi- rations, to seek employment in the city. But the pace of rural-urban migration is still often excessive from the social viewpoint. At various points throughout this chapter, we have looked at possible policy approaches designed to improve the very serious migration and employment situation in developing countries. We conclude with a summary of what appears to be the growing consensus of most economists on the shape of a comprehensive migration and employment strategy.33 These elements reflect the complex and nuanced nature of the topic, with potentially excessive migration relative to urban opportunities partly due to low productivity, poor rural institutions, and harsh social conditions; and the great and still not fully tapped opportunities for urban dynamism as an engine of economic development. We consider 10 key elements:

1. Creating an appropriate rural-urban economic balance. A more appropriate balance between rural and urban economic opportunities appears to be

366 PART TWO Problems and Policies: Domestic

indispensable to ameliorating both urban and rural unemployment prob- lems and to slowing the pace of excessive rural-urban migration. The main thrust of this activity should be in the integrated development of the rural sector, the spread of rural nonfarm employment opportunities, improved credit access, better agricultural training, the reorientation of social invest- ments toward rural areas, improving rural infrastructure, and addressing shortcomings of rural institutions (including corruption, discrimination, and stratification), the presence of which has the effect of raising the cost of delaying out-migration.

2. Expansion of small-scale, labor-intensive industries. The composition or “prod- uct mix” of output has obvious effects on the magnitude (and in many cases the location) of employment opportunities, because some products (often basic consumer goods) require more labor per unit of output and per unit of capital than others. Expansion of these mostly small-scale and labor- intensive industries in both urban and rural areas can be accomplished in two ways: directly, through government investment and incentives and improved access to credit, particularly for activities in the urban informal sector, and indirectly, through income redistribution (either directly or from future growth) to the rural poor, whose structure of consumer demand is both less import-intensive and more labor-intensive than that of the rich. Under the right conditions, such enterprises can agglomerate as industrial districts in ways that can generate exports, as pointed to by the findings on China in Box 7.1. Policies that effectively discourage clustering of special- ized activities are likely to be harmful.

3. Eliminating factor price distortions. There is ample evidence to demonstrate that correcting factor price distortions—primarily by eliminating various capital subsidies and curtailing the growth of urban wages through mar- ket-based pricing—would increase employment opportunities and make better use of scarce capital resources. But by how much or how quickly these policies would work is not clear. Moreover, their migration implica- tions would have to be ascertained. Correct pricing policies by themselves are insufficient to fundamentally alter the present employment situation.34

4. Choosing appropriate labor-intensive technologies of production. One of the prin- cipal factors inhibiting the success of any long-run program of employment creation in both urban industry and rural agriculture is the almost complete technological dependence on (typically laborsaving) machinery and equip- ment from the developed countries. Domestic and international efforts can help reduce this dependence by developing technological research and adaptation capacities in developing countries. Such efforts might first be linked to the development of small-scale, labor-intensive rural and urban enterprises. They could focus on developing low-cost, labor-intensive methods of meeting rural infrastructure needs, including roads, irrigation and drainage systems, and essential health and educational services. This is an area where scientific and technological assistance from the developed countries could prove extremely helpful.

5. Modifying the linkage between education and employment. The emergence of the phenomenon of the educated unemployed is calling into question the

367CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

appropriateness of the massive quantitative expansion of educational sys- tems, especially at the higher levels. Formal education has become the ration- ing tunnel through which all prospective jobholders must pass. Although a full discussion of educational problems and policies must await the next chapter, one way to moderate the excessive demand for additional years of schooling (which in reality is a demand for modern-sector jobs) would be for governments, often the largest employers, to base their hiring practices and their wage structures on other criteria. Moreover, the creation of attrac- tive economic opportunities in rural areas would make it easier to redirect educational systems toward the needs of rural development. At present, many of the skills needed for development remain largely neglected.

6. Reducing population growth. This is most efficiently accomplished through reductions in absolute poverty and inequality, particularly for women, along with the expanded provision of family-planning and rural health services. The labor force size for the next two decades is already determined by today’s birth rates, and hidden momentum of population growth applies as well to labor force growth. Together with the demand policies identified in points 1 through 5, the population and labor supply reduc- tion policies described in this chapter provide an essential ingredient in any strategy to combat the severe employment problems that developing countries face now and in future years.

7. Decentralizing authority to cities and neighborhoods. Experience shows that decentralization of authority to municipalities is an essential step in the improvement of urban policies and the quality of public services. Local con- ditions vary greatly among small and large cities, as well as across different national regions, and policies need to be designed to reflect these differences. Local officials have greater information about evolving local conditions; and when officials are held accountable for local fiscal performance and know they must answer to recipients of the services they provide, they also have greater incentives to carry out their responsibilities effectively. Decentraliza- tion, with increased authority of cities and regions, has been a major interna- tional trend in the organization of government (see Chapter 11) .

8. Leveraging untapped opportunities for urban dynamism. With strong, pro-poor rural development policies in place, many developing countries in Africa, Asia, and Latin America can still make gains in harnessing the growth potential of developing-country cities, with ongoing attention to prepar- ing for its possible migration implications.

9. Addressing the desperate poverty needs of the poor now living in urban slum conditions. As poor rural residents continue to migrate to urban areas, there is a growing phenomenon of the “urbanization of global poverty,” even if more than half of the poor will be found in rural areas for the next decades. As Martin Ravallion, Shaohua Chen, and Prem Sangraula concluded, “By fostering economic growth, urbanization helped reduce absolute poverty in the aggregate but did little for urban poverty.”35 For poor residents in slum communities, basic protection is needed. These residents face disease and death from unsanitary conditions and increasing vulnerability to severe weather events and other disasters. These citizens urgently need a

368 PART TWO Problems and Policies: Domestic

basic safety net, let alone an improvement in the actively hostile policies that have prevailed in many developing nations and regions by denying property rights (which has allowed the seizing of land and the demolish- ing of housing) and other forms of discrimination. A change in basic poli- cies can lead to large improvements in living conditions in slums.

10. Anticipating and assisting the new “climate migrants.”In a related point, one major response to climate change is rural-to-urban migration (see Chapter 10), section 10.3). This needs to be anticipated and planned for. A critical part of the solution is more effective rural development, from better access to sus- tainable irrigation to improved rural institutions. But “climate migrants” are already arriving in developing-country cities, and many of them end up on land that is highly vulnerable to disasters brought about by extreme weather, such as mudslides following heavy rains. As the World Bank has proposed:

In facilitating migration as a response to climate impacts, it is better to formulate integrated migration and development policies that address the needs of volun- tary migrants and support their entrepreneurial abilities and technical skills. To the extent possible, policies should discourage settlement of migrants in areas with high exposure to persistent climate hazards . . . forward-looking plans should identify alternative sites, apply compensation formulas that allow migrants to relocate and develop new sources of livelihoods, and build public and social infra- structure for community life.36

We return to the topic of rural development in Chapter 9 and environment and development in Chapter 10.

We conclude by noting that while a much higher urban share of population is inevitable, the tempo and pattern of urbanization will be key determinants of whether the deeper objectives of economic development are achieved. China and India, which together account for over one-third of the world’s population, are in the midst of their most rapid migration and urbanization period. Several African and other Asian countries are entering this stage. Because of fixed costs, including infrastructure and land use patterns, the quality of policies toward urbanization and migration that are implemented now are thus of momentous importance for the character of economic development for many decades to come.

369

About half of the world’s population lives in cities; by 2025, nearly two-thirds will live in urban ar- eas. Most of the urban growth is taking place in the developing world. The patterns of this growth and its implications are complex. Urban population growth in the developing world is far more rapid than popu- lation growth generally; about half the urban growth is accounted for by migrants from rural areas. Un- checked urbanization of the developing world is placing a strain on infrastructure and public health and threatens social stability. Shantytowns and simi- lar makeshift settlements represent over one-third of developing-country urban residences. About half of the urban labor force works in the informal sector of low-skilled, low-productivity, often self-employed jobs in petty sales and services. Still, this sector may generate up to a third of urban income and features a low capital intensity, low-cost training, waste recy- cling, and employment creation. What drives migra- tion? The cases of India and Botswana are instructive in showing the value of the probabilistic theory of migration and suggesting ways of extending it.

The scale of urbanization in these countries is dra- matic. The UN Population Division projected in 2013 that India will surpass China as the world’s largest nation in 2028, when India reaches a population of 1.45 billion; due largely to migration, the growth of the urban population will be much faster than that of the rural population. Botswana is a small country but represents one of Africa’s relatively few long- term success stories (see the case study for Chapter 14); and as of 2012, its urbanization rate had already reached well over 60%, compared with an average of under one-third in sub-Saharan Africa as a whole.

Any economic or social policy that affects rural and urban incomes will influence migration; this, in turn,

will affect sectoral and geographic economic activity, income distribution, and even population growth. Before the Todaro and Harris-Todaro migration mod- els were introduced, migration was widely viewed as irrational or driven by noneconomic motivations, sometimes attributed to the lure of the “bright city lights.” Noneconomic factors do influence migra- tion decisions, but economic factors are now under- stood to be primary. In the economic version of the bright-city-lights theory, people rationally migrated on the basis of costs and benefits. In this approach, it was assumed that if migrants appeared to be worse off, this was because other benefits were being over- looked, with the effect of making the migrants feel better off (or raising their overall utility).

The Todaro migration models postulate that observed migration is individually rational but that migrants respond to urban-rural differences in expected rather than actual earnings. Urban mod- ern-sector earnings are much higher than rural earn- ings, which may in turn be even higher than urban traditional-sector earnings. Migration occurs until average or expected rather than actual incomes are equal across regions, generating equilibrium unem- ployment or underemployment in the urban tradi- tional sector. The extension of the model to consider equilibrium and effects of actions such as increases in wages and probability of employment in the urban areas, undertaken by Harris and Todaro, shows that under some conditions, notably elastic supply of labor, creation of employment opportuni- ties in cities can actually lead to an increase in unem- ployment by attracting more migrants than there are new jobs. Despite being individually rational, exten- sive rural-urban migration generates social costs for crowded cities, while excessive migration also

Case Study 7

Rural-Urban Migration and Urbanization in Developing Countries: India and Botswana

370

imposes external costs on the rural areas emptied of better-educated, more venturesome young people, as well as external costs on urban infrastructure and lost output.

One set of relevant migration and employment policies emphasizes rural development, rural basic- needs strategies, elimination of factor price distor- tions, appropriate technology choice, and appropriate education. Each is intended to increase the incentives for rural residents to remain in rural areas rather than migrate to cities. But even if rural development is successful, fewer rural laborers will ultimately be needed, and demand for products of the cities will grow, which will fuel migration anyway. So other pol- icies seek to influence the pace and pattern of urban development to gain the most benefits for the fewest costs from migration that is probably inevitable.

India provides an interesting setting for a case study because future urban migration is potentially so vast and because a number of interesting stud- ies have been undertaken there. Botswana offers a good counterpoint because it has been the subject of some of the most interesting empirical research and represents one of the most rapidly urbanizing African countries as well as one of its most impor- tant success stories.

India The growth of Delhi has been extraordinary: In 1950, Delhi was not even among the world’s 30 largest cit- ies, but by 2013 its population had soared to become second in size only to Tokyo.

One of the most detailed studies of rural-urban migration, providing some tests of the Todaro migration models and depicting the characteristics of migrants and the migration process, is Biswajit Banerjee’s Rural to Urban Migration and the Urban Labor Market: A Case Study of Delhi.

Everyone who has been to a major city in a devel- oping country has noticed the sharp inequality between residents with modern-sector jobs and those working in the informal sector. But can the informal sector be seen as a temporary way station on the road to the formal sector, or can the barriers between these sectors be explained by education and skill require- ments that informal-sector workers cannot hope to meet? Banerjee found that the idea of segmented formal-informal rural labor markets could be sub- stantiated statistically. After carefully controlling for

human capital variables, Banerjee was still left with earnings in the formal sector 9% higher than in the informal sector that were not explained by any stan- dard economic factor. Even so, the earnings differ- ences found in India were not nearly so dramatic as implied in some of the migration literature.

In much of the literature on urbanization, the typi- cal laborer is characterized as self-employed or work- ing on some type of piecework basis. But Banerjee found that only 14% of his informal-sector sample worked in nonwage employment. Interestingly, aver- age monthly incomes of nonwage workers were 47% higher than those of formal-sector workers.

Banerjee argued that entry into nonwage employ- ment was not easy in Delhi. Some activities required significant skills or capital. Those that did not were often controlled by cohesive “networks” of operators that controlled activities in various enterprises. Entry barriers to self-employment in petty services were probably lower in other developing-country cities.

Consistent with these findings, Banerjee found that mobility from the informal to the formal sec- tor was low: There was little evidence that more than a very small minority of informal-sector work- ers were actively seeking jobs in the formal sectors, and only 5% to 15% of rural migrants in the infor- mal sector had moved over to the formal sector in a year’s time.

Moreover, the rate of entrance into the formal sector from the informal sector was just one-sixth to one-third that of the rate of direct entry into the urban formal sector from outside the area.

Informal-sector workers tended to work in the same job almost as long as those in the formal sector; the average informal-sector worker had worked 1.67 jobs over a period of 61 months in the city, while formal-sector workers averaged 1.24 jobs over an urban career of 67 months.

Banerjee’s survey data suggested that a large num- ber of informal-sector workers who had migrated to the city were attracted to the informal rather than the formal sector, coming to work as domestic servants, informal construction laborers, and salespeople. Of those who began nonwage employment upon their arrival, 71% had expected to do so. The fact that only a minority of informal-sector workers continued to search for formal-sector work was taken as further evidence that migrants had come to Delhi expressly to take up informal-sector work.

371

Workers who appear underemployed may not consider themselves as such, may perceive no pos- sibility of moving into the modern sector, may be unable to effectively search for modern-sector work while employed in the informal sector, and hence do not create as much downward pressure on modern- sector wages as it may at first appear. This may be one factor keeping modern-sector wages well above informal-sector wages for indefinite periods of time despite high measured urban underemployment.

One reason for this focus on the informal sector was concluded to be the lack of contacts of informal- sector workers with the formal sector. About two- thirds of direct entrants into the formal sector and nearly as many of those switching from the infor- mal to the formal sector found their jobs through personal contacts. This overwhelming importance of contacts explained why some 43% of Banerjee’s sample migrated after receiving a suggestion from a contact, which suggests that job market informa- tion can become available to potential migrants without their being physically present in the city. An additional 10% of the sample had a prearranged job in the city prior to migration.

Finally, the duration of unemployment following migration is usually very short. Within one week, 64% of new arrivals had found employment, and although a few were unemployed for a long period, the average waiting time to obtain a first job was just 17 days.

Banerjee also found that migrants kept close ties to their rural roots. Some three-quarters of the migrants visited their villages of origin and about two-thirds were remitting part of their urban incomes, a sub- stantial 23% of income on average. This indicates that concern for the whole family appeared to be a guiding force in migration. It also suggests a source of the rapid flow of job market information from urban to rural areas.

In a separate study, A. S. Oberai, Pradhan Prasad, and M. G. Sardana examined the determinants of migration in three states in India—Bihar, Kerala, and Uttar Pradesh. Their findings were consistent with the ideas that migrants often have a history of chronic underemployment before they migrate, migrate only as a measure of desperation, and have the expectation of participating in the informal urban sector even in the long run. Remittances were found to be substantial, and considerable levels of return migration were also documented, among

other evidence of continued close ties of migrants to their home villages.

But Banerjee’s fascinating findings do not necessar- ily represent a challenge to the applicability of Harris- Todaro or other “probabilistic migration models.” Instead, they suggest that they need to be extended to accommodate the apparently common pattern of migrating with the ultimate aim of urban informal- sector employment. As Ira Gang and Shubhashis Gangopadhyay have noted, one can modify the model to include in the urban area not only a formal sector but also a highly paid informal sector, as well as a low-paid (or unemployed) sector. In this case, people will migrate looking for either a formal-sector job or a high-paying informal-sector job. This seems to be consistent with Banerjee’s evidence. The assump- tion that keeps the essence of the probabilistic mod- els intact is that the wage of the formal urban sector exceeds the high-paying informal wage, which in turn exceeds the agricultural wage, which in turn exceeds the low-paying informal (or unemployed) wage. In fact, if rural wages remain below all urban opportuni- ties, this suggests that we are well out of equilibrium, and much additional migration must occur before expected incomes can be equalized across sectors. The particular formulations of the Todaro models are really no more than examples of a general principle: that migrants go where they expect in advance to do better, not where they do better after the fact. The basic ideas of the Todaro models do not depend on a particular notion of an informal or a formal sector.

Oded Stark’s ideas on a family’s use of migration can be a useful supplement to the Todaro models and may apply to some of Banerjee’s findings. In his view, a family will send members to different areas as a “portfolio diversification” strategy, to reduce the risk that the family will have no income. This approach is useful to explain any observed migration from higher- to lower-wage areas and into higher-wage areas but not necessarily the area with the highest expected wage. The basic idea of the Todaro models still applies, but this approach looks at families rather than individuals and stresses risk aversion.

Other studies have shown that the Todaro migra- tion models have held up well without modifi- cation in other parts of the world. A survey by Deepak Mazumdar confirmed that the evidence is overwhelming that migration decisions are made according to rational economic motivations.

372

Sources Banerjee, Biswajit. “The role of the informal sector

in the migration process: A test of probabilistic migration models and labour market segmenta- tion for India.” Oxford Economic Papers 35 (1983): 399–422.

Banerjee, Biswajit. Rural to Urban Migration and the Urban Labor Market: A Case Study of Delhi. Mumbai: Himalaya Publishing House, 1986.

Cole, William E., and Richard D. Sanders. “Internal migration and urban employment in the Third World.” American Economic Review 75 (1985): 481–494.

Corden, W. Max, and Ronald Findlay. “Urban unemployment, intersectoral capital mobility, and development policy.” Economica 42 (1975): 37–78.

Gang, Ira N., and Shubhashis Gangopadhyay. “A model of the informal sector in development.” Journal of Economic Studies 17 (1990): 19–31.

———. “Optimal policies in a dual economy with open unemployment and surplus labour.” Oxford Economic Papers 39 (1987): 378–387.

Harris, John, and Michael P. Todaro. “Migration, unemployment, and development: A two-sector analysis.” American Economic Review 60 (1970): 126–142.

Lucas, Robert E. B. “Emigration to South Africa’s mines.” American Economic Review 77 (1987): 313–330.

———. “Migration amongst the Batswana.” Economic Journal 95 (1985): 358–382.

Mazumdar, Deepak. “Rural-urban migration in devel- oping countries.” In Handbook of Regional and Urban Economics, vol. 2. New York: Elsevier, 1987.

Oberai, A. S., Pradhan Prasad, and M. G. Sardana. Determinants and Consequences of Internal Migra- tion in India: Studies in Bihar, Kerala and Uttar Pradesh. Delhi: Oxford University Press, 1989.

Stark, Oded. The Migration of Labor. Cambridge, Mass.: Blackwell, 1991.

Botswana A study of migration behavior conducted by Robert E. B. Lucas in Botswana addressed such problems in one of the most careful empirical studies of migra- tion in a developing country. His econometric model consisted of four groups of equations—for employ- ment, earnings, internal migration, and migration to South Africa. Each group was estimated from micro- economic data on individual migrants and nonmi- grants. Very detailed demographic information was used in the survey.

Rural migrants in Botswana moved to five urban centers (they would be called towns rather than cities in many parts of the world) as well as to neighboring South Africa. Lucas found that unad- justed urban earnings were much higher than rural earnings—68% higher for males—but these differ- ences became much smaller when schooling and experience were controlled for.

Lucas’s results confirm that the higher a person’s expected earnings and the higher the estimated probability of employment after a move to an urban center, the greater the chances that the person will migrate. And the higher the estimated wage and probability of employment for a person in his or her

home village, the lower the chances that the person will migrate. This result was very “robust”—not sensitive to which subgroups were examined or the way various factors were controlled for—and sta- tistically significant. It represents clear evidence in support of Todaro’s original hypothesis.

Moreover, Lucas estimated that at current pay dif- ferentials, the creation of one job in an urban center would draw more than one new migrant from the rural areas, thus confirming the Harris-Todaro effect. Earnings were also found to rise significantly the longer a migrant had been in an urban center, hold- ing education and age constant. But the reason was because of increases in the rate of pay rather than in the probability of modern-sector employment.

Taken together, the best-conducted studies of urbanization confirm the value of probabilistic migration models as the appropriate place to start seeking explanations of rural-to-urban migration in developing countries. But these studies underscore the need to expand these explanations of migration, considering that many people today migrate to par- ticipate in the informal rather than the formal urban sector and that workers may face a variety of risks in different settings. ■

373

Stark, Oded, and David Levhari. “On migration and risk in LDCs.” Economic Development and Cultural Change 31, (1982): 191–196.

Todaro, Michael P. “A model of labor migration and urban unemployment in LDCs.” American Economic Review 59 (1969): 138– 148.

UN-Habitat, “State of the World’s Cities, 2001,” http://www.unchs.org/Istanbul+5/86.pdf.

United Nations. An Urbanizing World: Global Report on Human Settlements. Report presented to the Habitat II conference, Istanbul, 1996.

United Nations Population Division. World Urban- ization Prospects: The 1999 Revision. New York: United Nations, 2000.

Concepts for Review

Agglomeration economies Congestion Efficiency wage Harris-Todaro model Induced migration

Informal sector Labor turnover Localization economies Present value Rural-urban migration

Social capital Todaro migration model Urban bias Urbanization economies Wage subsidy

Questions for Discussion

1. Why might the problem of rapid urbanization be a more significant population policy issue than cur- tailing population growth rates over the next two decades for most developing countries? Explain your answer.

2. Describe briefly the essential assumptions and major features of the Todaro model of rural-urban migration. One of the most significant implica- tions of this model is the paradoxical conclusion that government policies designed to create more urban employment may in fact lead to more urban unemployment. Explain the reasons for such a paradoxical result.

3. “The key to solving the serious problem of exces- sive rural-urban migration and rising urban unem- ployment and underemployment in developing countries is to restore a proper balance between urban and rural economic and social opportuni- ties.” Discuss the reasoning behind this statement, and give a few specific examples of government

policies that would promote a better balance between urban and rural economic and social opportunities.

4. For many years, the conventional wisdom of development economics assumed an inherent conflict between the objectives of maximizing output growth and promoting rapid industrial employment growth. Might these two objectives be mutually supportive rather than conflicting? Explain your answer.

5. What is meant by the expression “getting prices right”? Under what conditions will eliminating factor price distortions generate substantial new employment opportunities? (Be sure to define factor price distortions.)

6. The informal sector has become a very large part of the urban economy. Distinguish between the urban formal and informal sectors, and discuss both the positive and the negative aspects of the informal urban labor market.

374 PART TWO Problems and Policies: Domestic

7. Why are primary cities—generally the capital— often disproportionately large in many developing countries? Which factors can be addressed with better policies?

8. What is an industrial district? How might govern- ments of developing countries help them succeed?

9. Suppose that potential migrants make decisions only based on comparisons of their expected incomes. Now suppose the rural wage is $1 per day. Urban modern sector employment can be obtained with 0.25 probability and pays $3 per day. The urban traditional sector pays $0.40 per day. Using this information, and making assump- tions as needed, can you make a prediction about whether there will be any rural-to-urban or urban- to-rural migration? Explain your reasoning, stat- ing explicitly any simplifying assumptions, and show all work. Consider an approach that cal- culates an expected income in the urban sector of 0.25(3) + (0.75)(0.40) = 1.05; and note that this

exceeds the rural wage of 1—would you predict that there will be rural-to-urban migration? What simplifying assumptions are needed to make this a valid conclusion? Now, what would the urban traditional sector daily income have to be to induce no net rural-urban migration? If wages in all sectors are inflexible, what else adjusts in this model to lead to equilibrium (how much does it adjust and what is the intuition)?

10. Explain the concept of urban bias. What policies are associated with it, and what are their likely effects on urban and rural areas?

11. Now explain the economic benefits of concentra- tion of economic activity in cities. How are vari- ous costs of doing business likely to be affected? Why are some of the potential benefits of urban- ization lost when congestion becomes substantial? What policies are likely to strengthen or weaken the opportunities to take advantage of the eco- nomic benefits of cities?

375CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

Appendix 7.1

A Mathematical Formulation of the Todaro Migration Model

Consider the following mathematical formulation of the basic Todaro model discussed in this chapter. Individuals are assumed to base their decision to migrate on considerations of income maximization and what they perceive to be their expected income streams in urban and rural areas. It is further assumed that the individual who chooses to migrate is attempting to achieve the prevailing average income for his or her level of education or skill attain- ment in the urban center of his or her choice. Nevertheless, the migrant is assumed to be aware of the limited chances of immediately securing wage employment and the likelihood that he or she will be unemployed or under- employed for a certain period of time. It follows that the migrant’s expected income stream is determined by both the prevailing income in the modern sector and the probability of being employed there, rather than being under- employed in the urban informal sector or totally unemployed.

If we let V(0) be the discounted present value of the expected “net” urban- rural income stream over the migrant’s time horizon; Yu(t) and Yr(t) the aver- age real incomes of individuals employed in the urban and the rural economy, respectively; n the number of time periods in the migrant’s planning horizon; and r the discount rate reflecting the migrant’s degree of time preference, then the decision to migrate or not will depend on whether

V102 = L n

t =0 3p1t2Yu1t2 - Yr1t24e-rtdt - C102 (A7.1.1)

is positive or negative, where C(0) represents the cost of migration and p(t) is the probability that a migrant will have secured an urban job at the average income level in period t.

In any one time period, the probability of being employed in the mod- ern sector, p(t), will be directly related to the probability π of having been selected in that or any previous period from a given stock of unemployed or underemployed job seekers. If we assume that for most migrants the selection procedure is random, then the probability of having a job in the modern sector within x periods after migration, p1x2, is p112 = π112 and p122 = π112 + 31 - π1124 π122 so that

p1x2 = p1x - 12 + 31 - p1x - 124π1x2 (A7.1.2)

or

p1x2 = π112 + a x

t =2 π1t2 q

t - 1

s =1 31 - π1s24 (A7.1.3)

where π1t2 equals the ratio of new job openings relative to the number of accumulated job aspirants in period t.

376 PART TWO Problems and Policies: Domestic

It follows from this probability formulation that for any given level of Yu(t) and Yi(t), the longer the migrant has been in the city, the higher his or her probability p of having a job and the higher, therefore, his or her expected income in that period.

Formulating the probability variable in this way has two advantages:

1. It avoids the “all or nothing” problem of having to assume that the migrant either earns the average income or earns nothing in the periods immediately following migration. Consequently, it reflects the fact that many underem- ployed migrants will be able to generate some income in the urban informal or traditional sector while searching for a regular job.

2. It modifies somewhat the assumption of random selection, since the prob- ability of a migrant’s having been selected varies directly with the time the migrant has been in the city. This permits adjustments for the fact that longer-term migrants usually have more contacts and better information systems so that their expected incomes should be higher than those of newly arrived migrants with similar skills.

Suppose that we now incorporate this behavioristic theory of migration into a simple aggregate dynamic equilibrium model of urban labor demand and supply in the following manner. We once again define the probability π of obtaining a job in the urban sector in any one time period as being directly related to the rate of new employment creation and inversely related to the ratio of unemployed job seekers to the number of existing job opportunities, that is:

π = λN

S - N (A7.1.4)

where λ is the net rate of urban new job creation, N is the level of urban employment, and S is the total urban labor force. If w is the urban real wage rate and r represents average rural real income, then the expected urban-rural real-income differential d is

d = wπ - r (A7.1.5)

or, substituting Equation A7.1.4 into Equation A7.1.5,

d = w λN

S - N - r (A7.1.6)

The basic assumption of our model once again is that the supply of labor to the urban sector is a function of the urban-rural expected real-income differ- ential, that is,

S = fs1d2 (A7.1.7)

If the rate of urban job creation is a function of the urban wage w and a policy parameter a, such as a concentrated governmental effort to increase employment through a program of import substitution, both of which operate on labor demand, we have

377CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

λ = fd1w; a2 (A7.1.8)

where it is assumed that 0λ>0a 7 0. If the growth in the urban labor demand is increased as a result of the governmental policy shift, the increase in the urban labor supply is

0S 0a

= 0S 0d

0d 0λ

0λ 0a

(A7.1.9)

Differentiating Equation A7.1.6 and substituting into Equation A7.1.9, we obtain

0S 0a

= 0S 0d

w N

S - N # 0λ 0a

(A7.1.10)

The absolute number of urban employed will increase if the increase in labor supply exceeds the increase in the number of new jobs created, that is, if

0S 0a

7 01λN2

0a =

N0λ 0a

(A7.1.11)

Combining Equations A7.1.10 and A7.1.11, we get

0S 0d

w N

S - N # 0λ 0a

7 N0λ 0a

(A7.1.12)

or

0S/S 0d/d

7 d w # S - N

S (A7.1.13)

or, finally, substituting for d:

0S/S 0d/d

7 wπ - r

w # S - N

S (A7.1.14)

Equation A7.1.14 reveals that the absolute level of unemployment will rise if the elasticity of urban labor supply with respect to the expected urban-rural income differential (0S/S)/(0d/d)—what has been called elsewhere the “migra- tion response function”—exceeds the urban-rural differential as a proportion of the urban wage times the unemployment rate, (S – N)/S. Alternatively, Equation A7.1.14 shows that the higher the unemployment rate, the higher must be the elasticity to increase the level of unemployment for any expected real-income differential. But note that in most developing nations, the inequality in Equation A7.1.14 will be satisfied by a very low elasticity of supply when realistic figures are used. For example, if the urban real wage is 60, average rural real income is 20, the probability of getting a job is 0.50, and the unemployment rate is 20%, then the level of unemployment will increase if the elasticity of urban labor sup- ply is greater than 0.033; that is, substituting into Equation A7.1.14, we get

0S/S 0d/d

= 10.5 * 602 - 20

60 10.202 =

2 60

= 0.033 (A7.1.15)

Note that before one can realistically predict what the impact of a policy to generate more urban employment will be on the overall level of urban unemploy- ment, one needs solid estimates of the empirical value of this elasticity coeffi- cient prevailing in particular developing nations.

378 PART TWO Problems and Policies: Domestic

Notes

1. The estimate of 9.6 billion people was announced in June 2013; see United Nations World Population Prospects: The 2012 Revision (New York: United Nations, Department of Economic and Social Affairs, June 13, 2013). The mid-2009 estimate for the point when the urban population became the majority globally is found in United Nations World Urbanization Prospects: The 2011 Revision, 2012; the quote is from the UN Population Division 2009 chart at: http://www.un.org/en/development/ desa/population/publications/urbanization/ urban-rural.shtml.

2. See United Nations World Urbanization Prospects: The 2011 Revision, released April 5, 2012.

3. A well-known comment along these lines was made in 1984 by former World Bank president Robert McNamara, who expressed his skepti- cism that huge urban agglomerations could be made to work at all: “These sizes are such that any economies of location are dwarfed by costs of congestion. The rapid population growth that has produced them will have far outpaced the growth of human and physical infrastructure needed for even moderately efficient economic life and orderly political and social relationships, let alone amenity for their residents.” See Robert S. McNamara, “The population problem: Time bomb or myth?” Foreign Affairs 62 (1984): 1107–1131. For additional arguments on problems caused by rapid urban population growth, see Bertrand Renaud, National Urbanization Policy in Developing Countries (New York: Oxford University Press, 1981). A less concerned viewpoint is expressed in Jeffrey G. Williamson, “Migration and urbaniza- tion,” in Handbook of Development Economics, vol. 1, eds. Hollis B. Chenery and T. N. Srinivasan (Amsterdam: Elsevier, 1988), pp. 426–465.

4. United Nations Population Fund, Population, Resources, and the Environment (New York: United Nations, 1991), p. 61.

5. United Nations Population Division, World Popu- lation Monitoring, 1987 (New York: United Nations, 1988). Those results were reiterated in the Program of Action of the 1994 International Conference on Population and Development, para. 9.1. More

recently, the United Nations reported in 2006 that nearly three-quarters of developing-country offi- cials indicated a strong desire to implement poli- cies that would reduce rural-to-urban migration or to take actions to reverse rural-urban migration trends. See United Nations Population Division, World Urbanization Prospects: The 2005 Revision.

6. See Michael Porter, The Competitive Advantage of Nations (New York: Free Press, 1990); his theory is reviewed further in Chapter 12. Marshall intro- duced the industrial districts concept in his 1890 Principles of Economics.

7. See Michael Piore and Charles Sabel, The Second Industrial Divide (New York: Basic Books, 1984).

8. See Khalid Nadvi, “Collective efficiency and col- lective failure: The response of the Sialkot Surgical Instrument Cluster to global quality pressures,” World Development 27 (1999): 1605–1626.

9. Gezahegn Ayele, Lisa Moorman, Kassu Wamisho, and Xiaobo Zhang, “Infrastructure and cluster development,” International Food Policy Research Institute Discussion Paper No. 980, 2009.

10. The significance of industrial districts in devel- oping countries is difficult to pin down, in part because such clusters overlap traditional politi- cal jurisdictions for which data are collected. An excellent source on this topic is Hubert Schmitz and Khalid Nadvi, eds., “Introduction: Cluster- ing and industrialization,” World Development 27 (1999): 1503–1514. See also Khalid Nadvi, “Collec- tive efficiency and collective failure: The response of the Sialkot Surgical Instrument Cluster to global quality pressures,” World Development 27 (1999): 1605–1626. Hermine Weijland, “Microenterprise clusters in rural Indonesia: Industrial seedbed and policy target,” in ibid., p. 1519.

11. Dorothy McCormick, “African enterprise and industrialization: Theory and reality,” in ibid., pp. 1531–1551.

12. Schmitz and Nadvi, “Introduction” ibid., pp. 1505–1506,

13. World Bank, World Development Report, 1999–2000 (New York: Oxford University Press, 2000), ch. 6.

14. Ibid.

379CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

15. For an introductory overview of urban economics, see, for example, Arthur M. O’Sullivan, Urban Economics, 5th ed. (New York: McGraw-Hill/ Irwin, 2002). Formal models of some of these ideas can be found in Masahisa Fujita, Paul Krugman, and Anthony J. Venables, The Spatial Economy: Cit- ies, Regions, and International Trade (Cambridge, Mass.: MIT Press, 1999). We would like to thank Anthony Yezer for his very helpful suggestions on these sections.

15a. For a discussion, see World Bank, World Develop- ment Report 2009: Reshaping Economic Geography.

16. In this comparison, it is no accident that a relatively modest scale of the largest city tends to be found in countries in which the political capital is not found in the largest city, as will be explained shortly. This has been true in Canada and the United States nearly since their founding; it is more recently true in Brazil, where urban growth has been diverted to the new capital, Brasilia, which was inaugurated in 1960 and has reached a population approach- ing 4 million. Comparative advantage and geog- raphy are other important factors; continent-sized countries are more plausible settings for multiple major hubs, as are also found in China and India. The picture also changes somewhat if one consid- ers what the United Nations termed megaregions in a 2010 report, which include Hong Kong–Shen- zhen–Guangzhou in China and Rio de Janeiro–São Paulo in Brazil.

17. With the exception of France and Britain, most ratios in Europe are small. Examples—Italy: Rome, 3.4 million; Milan, 2.9 million. Germany: Berlin, 3.4 million; Hamburg, 1.7 million. Netherlands: Rotterdam and Amsterdam, 1.0 million each. Portugal: Lisbon, 2.7 million; Porto, 1.3 million. Spain: Madrid, 5.4 million; Barcelona, 4.8 million. Other sizable developing countries where ratios of largest to second-largest city are relatively higher include Indonesia (about 4), Ethiopia (over 8), Afghanistan (over 6), and Côte d’Ivoire (over 6). Egypt, Iran, Iraq, Kenya, Nigeria, and Bangladesh all have ratios of about 3. Some ratios are higher with alternative metropolitan area estimates.

18. For example, while Mexico City continues to expand, it has a smaller share of industry than in decades past. A major reason is the growing concentration of export industries in northern

Mexico along the U.S. border, especially follow- ing implementation of NAFTA and, even more recently, the move of some low-skill industries to southern Mexico.

19. Alberto F. Ades and Edward L. Glaeser, “Trade and circuses: Explaining urban giants,” Quarterly Journal of Economics 110 (1995): 195–227. Urban concentration is defined as the average share of urbanized population living in the main city from 1970 to 1985. Stable countries are defined as those whose average number of revolutions and coups is below the worldwide median. Dictatorships are countries whose average Gastil democracy and freedoms index for the period is higher than 3. See also Rasha Gustavsson, “Explaining the phe- nomenon of Third World urban giants: The effects of trade costs,” Journal of Economic Integration 14 (1999): 625–650.

20. UN-Habitat’s annual “State of the World’s Cities” reports are available at http://www.unhabitat.org.

21. See World Bank, World Development Report, 2008–2009 (New York: Oxford University Press, 2008), on the often unrealized role of agriculture in development (discussed in Chapter 9); UN- Habitat on new developing-country perspec- tives on urbanization at http://www.unhabitat. org; and the World Bank on realizing more of the potential benefits of cities at http://www. worldbank.org/urban. See also World Bank, World Development Report 2009: Reshaping Eco- nomic Geography (New York: Oxford University Press, 2009).

22. For the 2012 CIV study, see Isabel Günther and Andrey Launov, “Informal employment in devel- oping countries: Opportunity or last resort?” Journal of Development Economics 97, No. 1 (2012): 88–98; the authors use a parametric identifica- tion strategy. For a concise review of the overall debate, see Cathy A. Rakowski, “Convergence and divergence in the informal sector debate: A focus on Latin America, 1984–92,” World Devel- opment 22 (1994): 501–516. See also Donald C. Mead and Christian Morrisson, “The informal sector elephant,” World Development 24 (1996): 1611–1619, and Edward Funkhauser, “The urban informal sector in Central America: Household survey evidence,” World Development 24 (1996): 1737–1751.

380 PART TWO Problems and Policies: Domestic

23. For updates on these and related indices, see Inter- national Finance Corporation, Doing Business 2013, Smarter Regulations for Small and Medium-Size Enterprises, http://www.doingbusiness.org/~/ media/GIAWB/Doing%20Business/Documents/ Annual-Reports/English/DB13-full-report.pdf.

24. UN-Habitat noted this for its State of Women in Cit- ies 2012/2013, http://www.unhabitat.org/pmss/ listItemDetails.aspx?publicationID=3457.

25. See Robert E. B. Lucas, “Internal migration and urbanization: Recent contributions and new evi- dence,” background paper for World Bank, World Development Report, 1999–2000.

26. Although the rate of rural-urban migration slowed during the 1980s, especially in Latin America and sub-Saharan Africa, as a result of declining urban real wages and fewer formal-sector employment opportunities, the actual number of migrants con- tinued to increase.

27. See Appendix 7.1 and Michael P. Todaro, “A model of labor migration and urban unemploy- ment in less developed countries,” American Economic Review 59 (1969): 138–148, and John R. Harris and Michael P. Todaro, “Migration, unem- ployment, and development: A two-sector analy- sis,” American Economic Review 60 (1970): 126–142.

28. This graph was first introduced in W. Max Corden and Ronald Findlay, “Urban unemployment, inter- sectoral capital mobility, and development policy,” Economica 42 (1975): 59–78. It reflects Harris and Tod- aro,” Migration, unemployment, and development.”

29. Note that qq’ is a rectangular hyperbola, a unitary- elasticity curve showing a constant urban wage bill; that is, LM ¥ WM is fixed.

30. That is, if informal-sector income is greater than zero, we add to expected urban income (on the right side of Equation 7.1) the informal-sector wage WUI times the probability of receiving it: WUI(1 – LM/LUS), where (1 – LM/LUS) is the prob- ability of not receiving the preferred urban formal wage. We can further distinguish wages and prob- abilities of receiving them in this period, or in a more general model in future periods; for a fully developed model, see Appendix 7.1.

31. William J. Carrington, Enrica Detragiache, and Tara Vishwanath, “Migration with endogenous

moving costs,” American Economic Review 86 (1996): 909–930.

32. Whereas the Todaro model focuses on the insti- tutional determinants of urban wage rates above the equilibrium wage, several later analysts have sought to explain this phenomenon by focusing on the high costs of labor turnover (the so-called labor turnover model) in urban areas and the notion of an efficiency wage; an above-equilib- rium urban wage enables employers to secure a higher-quality workforce and greater productiv- ity on the job. For a review of these various mod- els, see Joseph E. Stiglitz, “Alternative theories of wage determination and unemployment in LDCs: The labor turnover model,” Quarterly Journal of Economics 88 (1974): 194–227, and Janet L. Yel- len, “Efficiency wage models of unemployment,” American Economic Review 74 (1984): 200–205. For evidence of the existence and importance of an institutionally determined urban-rural wage gap, see Francis Teal, “The size and sources of economic rents in a developing country manu- facturing labour market,” Economic Journal 106 (1996): 963–976. In an influential study, Valerie Bencivenga and Bruce Smith make the alterna- tive assumption that urban modern firms do not know the productivity of migrants but that some potential migrants from rural areas are highly productive and others are unproductive within formal-sector (say, industrial) firms. In this sce- nario, firms will be motivated through competi- tive forces to (in effect) offer migrants a package of a wage and a probability of employment. Mod- ern-sector firms hire labor until their marginal products are equal to the resulting high wage rate, and unemployment ensues. Moreover, if modern- sector labor demand increases, both modern- and traditional-sector workforces expand proportion- ately, inducing additional migration. See Valerie R. Bencivenga and Bruce D. Smith, “Unemploy- ment, migration, and growth,” Journal of Political Economy 105 (1997): 582–608. An alternative per- spective in the economics-of-information frame- work, based on moral hazard problems, is offered by Hadi S. Esfahani and Djavad Salehi-Ifsahani, “Effort observability and worker productivity: Toward an explanation of economic dualism,” Economic Journal 99 (1989): 818–836.

381CHAPTER 7 Urbanization and Rural-Urban Migration: Theory and Policy

33. On problems of job creation, see World Bank, World Development Report 2012. For other perspec- tives on migration and urbanization policy, see, for example, Gary S. Fields, “Public policy and the labor market in less developed countries,” in The Theory of Taxation for Developing Countries, eds. David P. Newbery and Nicholas Stern (New York: Oxford University Press, 1987); Charles M. Becker, Andrew M. Hammer, and Andrew R. Morrison, Beyond Urban Bias in Africa: Urban- ization in an Era of Structural Adjustment (Ports- mouth, N.H.: Heinemann, 1994), chs. 4–7; David Turnham, Employment and Development: A New Review of Evidence (Paris: Organization for Eco- nomic Coordination and Development, 1993), pp. 245–253; Paul P. Streeten, Strategies for Human Development: Global Poverty and Unemployment (Copenhagen: Handelshøjskolens Forlag, 1994), pp. 50–64; and Cedric Pugh, “Poverty and prog- ress: Reflections on housing and urban policies in

developing countries, 1976–96,” Urban Studies 34 (1997): 1547–1595.

34. The literature has also examined strategies to eliminate excessive migration through wage subsidies; these would prove expensive and dif- ficult to administer, but their analysis has yielded interesting insights into the nature of the Harris- Todaro migration model. See, for example, Ira Gang and Shubhashis Gangopadhyay, “Optimal policies in a dual economy with open unemploy- ment and surplus labour,” Oxford Economic Papers 39 (1987): 378–387, which also contains references to important earlier work.

35. Martin Ravallion, Shaohua Chen, and Prem San- graula, “New evidence on the urbanization of global poverty,” World Bank Research Working Paper 4199, 2008.

36. World Bank World Development Report, 2010, p. 110.

8.1 The Central Roles of Education and Health

Education and health are basic objectives of development; they are important ends in themselves. Health is central to well-being, and education is essen- tial for a satisfying and rewarding life; both are fundamental to the broader notion of expanded human capabilities that lie at the heart of the meaning of development (see Chapter 1). At the same time, education plays a key role in the ability of a developing country to absorb modern technology and to develop the capacity for self-sustaining growth and development. More- over, health is a prerequisite for increases in productivity, and successful education relies on adequate health as well. Thus, both health and educa- tion can also be seen as vital components of growth and development—as inputs to the aggregate production function. Their dual role as both inputs and outputs gives health and education their central importance in economic development.

It is hard to overstate how truly dramatic the improvements in world health and education have been. In 1950, some 280 of every 1,000 children in the developing world as a whole died before their fifth birthday. By 2011, that number had fallen to 95 per 1,000 in low-income countries and 46 per 1,000 in middle-income countries (though now compared with 6 per 1,000 in high- income countries and just 4 in many European countries).1 Some important killers have been completely or nearly eradicated. Smallpox used to kill more

382

Human Capital: Education and Health in Economic Development

To end poverty and boost shared prosperity, countries need robust, inclusive economic growth. And to drive growth, they need to build human capital through investments in health, education and social protection for all their citizens.

—Jim Yong Kim, World Bank President, 2013

My work on human capital began with an effort to calculate both private and social rates of return to men, women, blacks, and other groups from investments in different levels of education.

—Gary Becker, Nobel laureate in economics

The slow improvement in the health status of our people has been a matter of great concern. There is no denying the fact that we have not paid adequate attention to this dimension of development thus far.

—Manmohan Singh, prime minister of India, 2005

8

383CHAPTER 8 Human Capital: Education and Health in Economic Development

than 5 million people every year; the virus no longer exists outside a few laboratory samples. Major childhood illnesses such as rubella and polio have been largely controlled through the use of vaccines. In addition, recent decades have witnessed a historically unprecedented extension of literacy and other basic education to a majority of people in the developing world. The United Nations reports that although there were still a staggering 775 million illiterate people aged 15 or older in the world in 2010, the good news is that 82% of all people are literate today, compared to just 63% as recently as 1970.2 But almost two-thirds of the world’s illiterate people are women.

Despite such outstanding achievements, developing countries continue to face great challenges as they seek to continue to improve the health and edu- cation of their people. The distribution of health and education within coun- tries is as important as income distribution; life expectancy may be quite high for better-off people in developing countries but far lower for the poor. Child mortality rates in developing countries remain more than 10 times higher than those found in the rich countries. These deaths generally result from conditions that are easily treatable, including millions who continue to die needlessly each year from dehydration caused by diarrhea. If child death rates in developing countries fell to those prevailing in the developed countries, the lives of nearly 7 million children would be saved each year. Many children who survive none- theless suffer chronic problems of malnutrition, debilitating parasitic infections, and other recurrent illnesses. Problems caused by lack of key micronutrients such as iodine, as well as protein, affect up to 2 billion people, but children are particularly vulnerable. Whereas citizens in Europe, North America, or Japan have more than 12 years of schooling on average, the average citizen in sub- Saharan Africa and South Asia spends less than five years in school—before taking account of teacher absenteeism and making no adjustment for the lack of schoolbooks and other resources even when a teacher is present. The “voices of the poor” in Box 8.1 convey some of the impact of deprivation in health and education on people’s lives.

In this chapter, we examine the roles of education and health in economic development. These two human capital issues are treated together because of their close relationship. There are dual impacts of the effects of health spend- ing on the effectiveness of the educational system and vice versa; and when we speak of investing in a person’s health and investing in a person’s education, we are after all talking about the same person. We then consider the relation- ships between income on the one hand and health and education on the other. Despite their close relationship, you will see that higher household income is no guarantee of improved health and education: Human capital must be given direct attention in its own right, even in economies that are growing rapidly. Health and education may be distributed very unequally, just as income and wealth are. But improved health and education help families escape some of the vicious circles of poverty in which they are trapped. Finally, we take a close look at educational and health systems in developing countries, to iden- tify the sources of the severe inequalities and inefficiencies that continue to plague them. The evidence reveals that investments in human capital have to be undertaken with both equity and efficiency for them to have their potential positive effects on incomes.

Literacy The ability to read and write.

Human capital Productive investments embodied in human persons, including skills, abilities, ideals, health, and locations, often resulting from expenditures on educa- tion, on-the-job training pro- grams, and medical care.

384 PART Two Problems and Policies: Domestic

BOX 8.1 Health and Education: Voices of the Poor

If you don’t have money today, your disease will take you to your grave.

—An old woman from Ghana

The children keep playing in the sewage. —Sacadura Cabral, Brazil

In the hospitals, they don’t provide good care to the indigenous people like they ought to; be- cause of their illiteracy they treat them badly. . . . They give us other medicines that are not for the health problem you have.

—A young man from La Calera, Ecuador

The school was OK, but now it is in shambles; there are no teachers for weeks. . . . There is no safety and no hygiene.

—Vila Junqueira, Brazil

If parents do not meet these payments, which are as high as 40 to 50 rupees per month, the teach- ers were reported to beat the student or submit a failing grade for her/him.

—Pakistan (“Voice of the Poor”)

Teachers do not go to school except when it is time to receive salaries.

—Nigeria (“Voice of the Poor”)

Before everyone could get health care, but now everyone just prays to God that they don’t get sick because everywhere they just ask for money.

—Vares, Bosnia and Herzegovina

Education and Health as Joint Investments for Development

Health and education are closely related in economic development.3 On one hand, greater health capital may improve the return to investments in educa- tion, in part because health is an important factor in school attendance and in the formal learning process of a child. A longer life raises the return to invest- ments in education; better health at any point during working life may in effect lower the rate of depreciation of education capital. On the other hand, greater education capital may improve the return to investments in health, because many health programs rely on basic skills often learned at school, including personal hygiene and sanitation, not to mention basic literacy and numeracy; education is also needed for the formation and training of health personnel. Finally, an improvement in productive efficiency from investments in educa- tion raises the return on a lifesaving investment in health. Box 8.2 summarizes the linkages between investments in health and education.

The past half-century or so has witnessed unprecedented advances in human capital. Health and education levels improved in both developed and developing countries, but by most measures, they have improved more rapidly in developing countries. As a result, there has been some international con- vergence in these measures. Only in sub-Saharan Africa, where life expectan- cies fell due to the AIDS crisis, has some doubt been cast on the trend toward catching up in health. As primary enrollments rise in developing countries, education is catching up, though some observers believe that the quality gap may be larger than ever. Even though the health and education gap between developed and developing countries remains large and further improvements may prove difficult, the progress to date has been unmistakable.4

385CHAPTER 8 Human Capital: Education and Health in Economic Development

Improving Health and Education: Why Increasing Income Is Not Sufficient

Health and education levels are much higher in high-income countries. There are good reasons to believe that the causality runs in both directions: With higher income, people and governments can afford to spend more on educa- tion and health, and with greater health and education, higher productivity and incomes are possible. Because of these relationships, development policy needs to focus on income, health, and education simultaneously. This conclu- sion is parallel to our conclusion in Chapter 5 that we need a multipronged strategy to address the stubborn problems of absolute poverty.

People will spend more on human capital when income is higher. But the evidence shows clearly that even if we were able to raise incomes with- out a large improvement in health and education, we could not count on that income increase being used to adequately invest in children’s education and health. The market will not solve this problem automatically, and in many cases, household consumption choices themselves may lead to a surprisingly small link between income and nutrition, especially for children.5 The income elasticity of the demand for calories (that is, the percentage change in calo- ries consumed for a percentage change in family incomes) among low-income people range from near zero to about 0.5, depending on the region and the sta- tistical strategy used by the researchers.6 This less than proportional response is due to two factors: Income is spent on other goods besides food, and part of the increased food expenditures is used to increase food variety without necessarily increasing the consumption of calories. If the relationship between

BOX 8.2 Linkages between Investments in Health and Education

• Health and education are investments made in the same individual.

• Greater health capital may raise the return on investment in education for several reasons:

Health is an important factor in school attendance.

Healthier children are more successful in school and learn more efficiently.

Deaths of school-age children also increase the cost of education per worker.

Longer life spans raise the return to investments in education.

Healthier individuals are more able to productively use education at any point in life.

• Greater education capital may raise the return to investment in health in the follow- ing ways:

Many health programs rely on skills learned in school (including literacy and numeracy).

Schools teach basic personal hygiene and sanitation.

Education is needed for the formation and training of health personnel.

Education leads to delayed childbearing, which improves health.

• Improvements in productive efficiency from investment in education raise the return on a lifesaving investment in health.

386 PART Two Problems and Policies: Domestic

income and nutrition is indeed quite low, as some studies suggest, then development policies that emphasize increasing incomes of the poor with- out attention to the way these additional resources are expended within the family may not lead to improved health, and successful development more generally, at least not very quickly.7 As discussed further in Chapter 15 and its case study, credit for microenterprises has been one of the most popular poverty alleviation strategies in recent years. In this case, credit may help the poor improve their nutrition, for example, because seasonal price fluctuations are also shown to be an important determinant of calorie consumption along with average income among the very poor, but credit will not be sufficient if nutrition remains inadequate and does not improve automatically with higher income.

Moreover, calories are not the same as nutrition, and the nutrition of earn- ers is not the same as the nutrition of their children. The income elasticity of “convenience” foods is greater than unity.8 An increase in income frequently allows families in developing countries to switch consumption from nutritious foods such as beans and rice to nonnutritious “empty calories” such as candy and soda, which may be perceived as modern and symbolic of economic suc- cess.9 A major problem is that poor health (e.g., diarrheal diseases) can negate the health advantages of better nutrition.10

There is considerable evidence that the better the education of the mother, the better the health of her children (see Figure 2.5 on page 61 and Box 8.3).11 There are still opportunities for improving health through simple activities in school that have not been utilized.

Health status, once attained, also affects school performance, as has been shown in studies of many developing countries. Better health and nutrition lead to earlier and longer school enrollment, better school attendance, and more effective learning.12 Thus, to improve enrollments and the effectiveness of schooling, we must improve the health of children in developing countries.

BOX 8.3 FINDINGS Mothers’ Health Knowledge Is Crucial for Raising Child Health

Usually, formal education is needed in comple-mentary relationship with ongoing access to cur- rent information. Paul Glewwe found in an analysis of data from Morocco that a mother’s basic health knowledge had a positive effect on her children’s health. Several mechanisms were possible, such as that “formal education directly teaches health knowl- edge to future mothers; literacy and numeracy skills acquired in school assist future mothers in diagnosing and treating child health problems; and exposure to modern society from formal schooling makes women more receptive to modern medical treatments.” But,

Glewwe concludes, “Mother’s health knowledge alone appears to be the crucial skill for raising child health. In Morocco, such knowledge is primarily obtained outside the classroom, although it is obtained using literacy and numeracy skills learned in school.... Teaching of health knowledge skills in Moroccan schools could substantially raise child health and nutrition in Morocco.”

Source: Based on Paul Glewwe, “Why does mother’s schooling raise child health in developing countries? Evidence from Morocco,” Journal of Human Resources 34 (1999): 124–159.

387CHAPTER 8 Human Capital: Education and Health in Economic Development

Indeed, advances in statistical methods are showing that the links from health to educational attainment in developing countries are stronger than had been believed (see Box 8.4). These effects are large for both boys and often espe- cially for girls.13

Finally, there are other important spillover benefits to investment in one’s health or education. An educated person provides benefits to people around him or her, such as reading for them or coming up with innovations that ben- efit the community.14 As a result, there are significant market failures in edu- cation. Moreover, a healthy person is not only less contagious but also can

BOX 8.4 FINDINGS School Impact of a Low-Cost Health Intervention

A study in the Busia district in Kenya conducted by Edward Miguel of the University of California at Berkeley and Michael Kremer of Harvard University showed that inexpensive “deworming” drugs to elimi- nate parasitic infections in children are also very cost- effective in increasing school attendance. The order in which schools received the treatments was random- ized, enabling identification of the causal effect of treatments by comparing outcomes with the not-yet- treated schools.

Their baseline survey showed 92% of schoolchil- dren were infected with at least one parasite, and 28% had at least three infections. A moderate to heavy infection was present in 31%. In fact, the prevalence was probably worse because “heavily infected children were more likely to be absent from school on the day of the survey.”

As a result of the deworming, absenteeism decreased by about one-quarter (7 percentage points). Younger children typically had suffered more infection, and they now attended 15 more school days per year on average; older children attended about 10 more days. The program cost per additional year of schooling was about $3.50, much less than the alternative methods used to increase school participation. Treated children also had lower anemia, somewhat reduced reported illness, and better height-for-age scores.

Children can spread parasitic infections across school districts, notably when they swim in the same lake. This explains the study’s finding that curing worm infections also led to substantial benefits for

neighboring school districts that had not yet been dewormed—a classic externality. Reduced infection can also benefit adults, who can work more days.

Although academic test scores did not increase sig- nificantly, this may have been due to the larger school class size that resulted from greater participation rates, thereby increasing the student-to-teacher ratio. Eval- uated over the course of the student’s lifetime, the deworming drugs are not only inexpensive but also yield a very high rate of return, with the implied pres- ent discounted value of wage gains of more than $30 per treated child. The net benefit of the program is greater than the cost of hiring additional teachers to keep this ratio from rising—though this does depend on the political will to do so.

Despite its large benefits, families in impoverished Busia are very sensitive to the price of deworming treatments, suggesting that subsidies will be needed for some time. As one might hope from such clear findings, this study has had a substantial impact on health priorities of developing-country governments and international agencies, and deworming programs are expanding in many countries. This study’s clear findings from carefully designed methods provide one of the important impetuses to the recent emphasis on and progress in deworming in schools.

Source: Edward Miguel and Michael Kremer, “Worms: Identifying impact on education and health in the pres- ence of treatment externalities,” Econometrica 72 (2004): 159–217. On deworming activities, see the links at http:// www.dewormtheworld.org.

388 PART Two Problems and Policies: Domestic

benefit the community in many ways that a sick person cannot. Because of such spillover effects, the market cannot be counted on to deliver the socially efficient levels of health and education. Thus, as the World Health Organi- zation (WHO) concluded, “Ultimate responsibility for the performance of a country’s health system lies with government.”15 Developing-country officials are drawing lessons from the many studies showing the interrelationships among health, education, and incomes and are devising integrated strategies. The case study of Mexico at the end of this chapter provides an important example.

8.2 Investing in Education and Health: The Human Capital Approach

The analysis of investments in health and education is unified in the human capital approach. Human capital is the term economists often use for educa- tion, health, and other human capacities that can raise productivity when increased. An analogy is made to conventional investments in physical capi- tal: After an initial investment is made, a stream of higher future income can be generated from both expansion of education and improvements in health. As a result, a rate of return can be deduced and compared with returns to other investments. This is done by estimating the present discounted value of the increased income stream made possible by these investments and then comparing it with their direct and indirect costs. Of course, health and educa- tion also contribute directly to well-being. For example, education increases empowerment and autonomy in major matters in life, such as capacity for civic engagement, making decisions concerning one’s own health care, and freedom to choose one’s own spouse over an arranged marriage.16 But the basic human capital approach focuses on the indirect ability to increase well- being by increasing incomes. In this section, we will generally illustrate points with educational investments, but the same principles apply to health invest- ments.

The impact of human capital investments in developing countries can be quite substantial. Figure 8.1 shows the age-earnings profiles by levels of education in Venezuela.17 The chart shows how incomes vary over the life cycle for people with various levels of education. Note that those with higher levels of education start full-time work at a later age, but as is shown, their incomes quickly outpace those who started working earlier. But such future income gains from education must be compared with the total costs incurred to understand the value of human capital as an investment. Education costs include any direct tuition or other expenditures specifically related to educa- tion, such as books and required school uniforms, and indirect costs, primarily income forgone because the student could not work while in school.

Formally, the income gains can be written as follows, where E is income with extra education, N is income without extra education, t is year, i is the discount rate, and the summation is over expected years of working life:

a Et - Nt11 + i2t (8.1)

Discount rate In present- value calculations, the annual rate at which future values are decreased to make them comparable to values in the present.

389CHAPTER 8 Human Capital: Education and Health in Economic Development

An analogous formula applies to health (such as improved nutritional status), with the direct and indirect cost of resources devoted to health compared with the extra income gained in the future as a result of higher health status.

Figure 8.2 provides a typical schematic representation of the trade-offs involved in the decision to continue in school.18 It is assumed that the indi- vidual works from the time he or she finishes school until he or she is unable to work, retires, or dies. This is taken to be 66 years. Two earnings profiles are presented—for workers with primary school but no secondary education and for those with a full secondary (but no higher) education. Primary gradu- ates are assumed to begin work at age 13, and secondary graduates, at age 17. For an individual in a developing country deciding whether to go on from primary to secondary education, four years of income are forgone. This is the indirect cost, as labeled in the diagram. The child may work part time, a pos- sibility ignored here for simplicity, but if so, only part of the indirect-cost area applies. There is also a direct cost, such as fees, school uniforms, books, and other expenditures that would not have been made if the individual had left school at the end of the primary grades. Over the rest of the person’s life, he or she makes more money each year than would have been earned with only a primary education. This differential is labeled “Benefits” in the diagram. Before comparing costs with benefits, note that a dollar today is worth more to an individual than a dollar in the future, so those future income gains must be discounted accordingly, as is done in Equation 8.1. The rate of return will be higher whenever the discount rate is lower, the direct or indirect costs are lower, or the benefits are higher.

This analysis was performed from the individual’s point of view in the three right-hand columns of Table 8.1. Notice that in sub-Saharan Africa, the

FIGuRE 8.1 Age-Earnings Profiles by Level of Education: Venezuela

Source: International Bank for Reconstruction and Development/The World Bank: The Profitability of Investment in Education: Concepts & Methods by George Psacharopoulos, 1995. Reprinted with permission.

300

250

200

150

100

50

0 10 25 40 55

Age

A n

n u

a l i

n co

m e

( th

o u

sa n

d s

o f

b o

li va

rs )

University Secondary Primary No education

6125_08_FG001

390 PART Two Problems and Policies: Domestic

private rate of return to primary education is over 37%! Despite this extraordi- nary return, many families do not make this investment because they have no ability to borrow even the meager amount of money that a working child can bring into the family—the topic of the next section. Note that the higher rates of return for developing countries reflect that the income differential between those with more and less schooling is greater on average than for the devel- oped countries.

The first three columns of Table 8.1 indicate the social rate of return. This is found by including the amount of public subsidy for the individual’s edu- cation as part of the direct costs, because this is part of the investment from the social point of view (and also by considering pretax rather than after-tax incomes). Details of the calculations are presented in note 19.19 It should be noted that these social returns are probably understated because they do not take into account the externality that educated people confer on oth- ers (e.g., being able to read for other family members and coworkers), not to mention other individual and social benefits such as increased autonomy and civic participation, being able to communicate more effectively, making more informed choices, and even being taken more seriously in public discus- sions, as stressed by Amartya Sen. Figure 8.2 can also be used to illustrate the benefit-cost trade-off from the public policy point of view by including fiscal costs and social welfare benefits; that is, adding the social costs of education such as subsidies to the direct costs part of the costs area below the x-axis and

FIGuRE 8.2 Financial Trade-Offs in the Decision to Continue in School

Earnings

Direct costs

Secondary graduates

Primary graduates

Benefits

Indirect costs

Direct costs

Age661713

391CHAPTER 8 Human Capital: Education and Health in Economic Development

adding in any net spillover benefits to the benefits area (not shown are such benefits as occur before graduation or after retirement).19a

8.3 Child Labor

Child labor is a widespread problem in developing countries. When children under age 15 work, their labor time disrupts their schooling and, in a major- ity of cases, prevents them from attending school altogether. Compounding this, the health of child workers is significantly worse, even accounting for their poverty status, than that of children who do not work; physical stunting among child laborers is very common. In addition, many laboring children are subject to especially cruel and exploitative working conditions.

The International Labor Organization (ILO), a UN body that has played a leading role on the child labor issue,20 reported in its 2010 quadrennial report on child labor that as of 2008, there was a total of 306 million children between ages 5 and 17 doing some kind of work, but about one-third of this is con- sidered permissible work based on national laws and existing ILO conven- tions. But 215 million are classified as “child laborers” because they “are either under the minimum age for work or above that age [through age 17] and engaged in work that poses a threat to their health, safety, or morals, or are subject to conditions of forced labor.” This number is down about 3% from the 222 million estimated for 2004. There are over 9 million child laborers between the ages of 5 and 11, nearly a third of them doing hazardous work. More than half of child laborers, some 115 million children, are still exposed to hazardous work. More than half of all child laborers live in Asia and the Pacific, but sub- Saharan Africa has the highest rate of child labor. Among children doing haz- ardous work, over 48 million live in Asia and the Pacific, nearly 39 million live in sub-Saharan Africa, and over 9 million live in Latin America. Child labor remains a problem in the Arab states, where the issue has been largely ignored until very recently. And major progress has been made in some countries such as Brazil and some regions such as Kerala in India.

Working conditions are often horrendous; the ILO reports that some of its surveys show that more than half of child laborers toil for nine or more hours

TabLe 8.1 Returns to Investment in education by Level, Regional averages (%)

Source: George G. Psacharopoulos and Harry A. Patrinos, “Returns to investment in education: A further update,” Education Economics 12, No. 2 (August 2004), tab. 1.

Social Private

Region Primary Secondary Higher Primary Secondary Higher

Asiaa 16.2 11.1 11.0 20.0 15.8 18.2 MENAb 15.6 9.7 9.9 13.8 13.6 18.8 Latin America/Caribbean 17.4 12.9 12.3 26.6 17.0 19.5 OECD 8.5 9.4 8.5 13.4 11.3 11.6 Sub-Saharan Africa 25.4 18.4 11.3 37.6 24.6 27.8 World 18.9 13.1 10.8 26.6 17.0 19.0 a Non-OECD. b Europe/Middle East/North Africa, Non-OECD

Note: How these rates of return were calculated is explained in detail in note 19 at the end of this chapter.

392 PART Two Problems and Policies: Domestic

per day. The worst forms of child labor endanger health or well-being, involv- ing hazards, sexual exploitation, trafficking, and debt bondage. In a 2011 pub- lication, the ILO reported that every year, about 22,000 children die as a result of work-related accidents. Clearly, child labor is not an isolated problem but a widespread one, especially in Africa and South Asia.

Nevertheless, it is not obvious that an immediate ban on all forms of child labor is always in the best interests of the child. Without work, a child may become severely malnourished; with work, school fees as well as basic nutri- tion and health care may be available. But there is one set of circumstances under which both the child laborer and the family as a whole may be unam- biguously better off with a ban on child labor: multiple equilibria. Kaushik Basu has provided such an analysis, and we shall first consider his simple model, which shows how this problem may arise.21

To model child labor, we make two important assumptions: First, a house- hold with a sufficiently high income would not send its children to work. As one might hope, there is strong evidence that this is true, at least most of the time. Second, child and adult labor are substitutes. In fact, children are not as productive as adults, and adults can do any work that children can do. This assertion is not an assumption; it is a finding of many studies of the produc- tivity of child laborers in many countries. It is important to emphasize this, because one rationalization for child labor often heard is that children have special productive abilities, such as small fingers, that make them important for the production of rugs and other products. However, there is no support for this view. In essentially every task that has been studied, including carpet weaving, adult laborers are significantly more productive. As a result, we can consider the supply of adult and child labor together in an economic analysis of the problem.

The child labor model is graphed in Figure 8.3. On the x-axis, we have the supply of labor in adult equivalents. Because we are interested in understand- ing the impact of the demand for labor, in a graph it is best to consider homo- geneous units of labor. So if a child laborer is γ times as productive as an adult worker, we consider one child the productive equivalent of γ adult workers. According to our assumptions, γ 6 1. For example, if a child laborer is half as productive as an adult worker, γ = 0.5.

We start with the assumption that in the region in question, all (unskilled) adults work, regardless of the wage. This gives us a perfectly inelastic, vertical adult labor supply curve, called AA′ in the diagram. Highly inelastic supply is a very reasonable assumption among families so poor that their children must work. While the parents may not have modern-sector jobs, every adult is involved in some type of activity to help the family survive. This adult sup- ply AA′ is simply the number of unskilled adults. To understand the total labor supply curve, consider what happens if the wage falls. If the wage falls below wH, then some families find they are poor enough that they have to send their children to work. At first wages are still high enough so that this affects only a few families and children, reflected in the fact that the S-shaped curve just below wH is still quite steep. As the wage continued to fall, more fami- lies would do the same, and labor supply expands along the S-shaped curve, which becomes flatter as smaller drops in the wage lead many more families to send their children to work. If a wage of wL were reached, all of the children

393CHAPTER 8 Human Capital: Education and Health in Economic Development

would work. At this point, we are on the vertical line labeled TT′, which is the aggregate labor supply of all the adults and all the children together. This sum is the number of adults plus the number of children, multiplied by their lower productivity, γ 6 1. (An S shape in the middle portion is likely, but the analysis holds even if this is a straight line.) The resulting supply curve for children and adults together is very different from the standard ones that we usually con- sider in basic microeconomics, such as the upward-sloping supply curve seen in Figure 5.5 (in Chapter 5), but it is highly relevant for the developing-country child labor context. To summarize, as long as the wage is above wH, the supply curve is along AA′; if the wage is below wL, the supply curve is along TT′, and in between, it follows the S-shaped curve between the two vertical lines.

Now consider the labor demand curve, DL; if demand is inelastic enough to cut the AA′ line above wH and also cut the TT′ line below wL, there will be two stable equilibria, labeled E1 and E2, in the diagram.

22 When there are two equi- libria, if we start out at the bad equilibrium E2, an effective ban on child labor will move the region to the good equilibrium E1. Moreover, once the economy has moved to the new equilibrium, the child labor ban will be self-enforcing, because by assumption, the new wage is high enough for no family to have to send its children to work. If poor families coordinate with each other and refuse to send their children to work, each will be better off; but in general, with a large number of families, they will be unable to achieve this.23

Figure 8.3 Child Labor as a Bad equilibrium

Source: From Kaushik Basu, “Child labor: Cause, consequence, and cure, with remarks on international labor standards,” Journal of Economic Literature 37 (1999): 1101. Reprinted with the permission of the author and the American Economic Association.

0 A�

A T

T � L

E2

E1

w A

wE1

wE2

wH

wL

D L

6125_08_FG003

394 PART Two Problems and Policies: Domestic

Banning child labor when there is an alternative equilibrium in which all children go to school might seem like an irresistible policy, but note that while all the families of child laborers are better off, employers may now be worse off, because they have to pay a higher wage. Thus, employers may use political pressure to prevent enactment of child labor laws. In this sense, child labor, even its worst forms, could actually be Pareto-optimal—a discovery that should remind us that Pareto optimality is sometimes a very weak condition on which to base development policy! In the same sense, many other problems of underdevelopment, including extreme poverty itself, may at times also be Pareto-optimal, in that solving these problems may make the rich worse off.

While these child labor models are probably reasonable depictions of many developing areas, we do not know enough about conditions in unskilled labor markets to say how significant these types of multiple equilibria and severe credit constraints really are as explanations for child labor. Thus, it would be potentially counterproductive, if even enforceable, to seek an immediate ban on all child labor in all parts of the world today. As a result, an intermediate approach is currently dominant in international policy circles.24

There are four main approaches to child labor policy current in develop- ment policy. The first recognizes child labor as an expression of poverty and recommends an emphasis on eliminating poverty rather than directly address- ing child labor; this position is generally associated with the World Bank (pov- erty policy is discussed further in Chapters 5, 9, and 15).

The second approach emphasizes strategies to get more children into school, including expanded school places, such as new village schools, and condi- tional cash transfer (CCT) incentives to induce parents to send their children to school, such as the Progresa/Oportunidades Program in Mexico, discussed in this chapter’s case study, or the experimental Malawi program discussed in Box 8.5. This strategy has widespread support from many international agen- cies and development bodies. It is probably a more effective approach than mak- ing basic education compulsory, because without complementary policies, the incentives to send children to work would still remain strong and enforcement is likely to be weak, for the same reasons that regulation of the informal sec- tor has proved almost impossible in many other cases. Compulsory schooling is a good idea, but it is not by itself a sufficient solution to the problem of child labor. Improving the quality of basic schooling and increasing accessibility are also very important; the fraction of national income spent on basic education in a majority of low-income countries remains problematic. As the ILO points out,

In sub-Saharan Africa, about half of all low-income countries spend less than 4 per-cent of their national income on education. In South Asia, Bangladesh devotes only 2.6 per-cent of its national income to education and Pakistan, 2.7 per-cent. India invests a smaller proportion of GNP (around 3.3 per-cent) than the median for sub-Saharan Africa, even though average incomes are around one third higher. Even more worrying is that the share of national income devoted to education is stagnating or decreasing in key countries, including Bangladesh, India and Pakistan, which account for over 15 million out-of-school children.25

The third approach considers child labor inevitable, at least in the short run, and stresses palliative measures such as regulating it to prevent abuse and to provide support services for working children. This approach is most com- monly associated with UNICEF, which has prepared a checklist of regulatory and

Conditional cash transfer (CCT) programs Welfare benefits provided condition- ally based on family behavior such as children’s regular school attendance and health clinic visitations.

395CHAPTER 8 Human Capital: Education and Health in Economic Development

social approaches that could meet the “best interest of the child.” The regula- tions included on UNICEF’s checklist include expanding educational oppor- tunities through “time off” for standard or workplace schooling, encouraging stricter law enforcement against illegal child labor trafficking, providing

BOX 8.5 FINDINGS Cash or Condition? Evidence from Malawi

What programs are effective at addressing the nexus of poverty and unmet health and educa- tion needs, especially for girls growing up in extreme poverty? As Sarah Baird, Craig McIntosh, and Berk Ozler note, school enrollment and effective learning, and marriage and fertility outcomes are of “central importance to the long-term prospects of school-age girls” living in poverty. What programs would be most cost-effective?

Findings from a randomized control trial study of a cash transfer program targeted to adolescent girls in Malawi offer important insights. Baird, McIntosh, and Ozler compared families who were randomly assigned to one of three groups: no cash transfer, uncondi- tional cash transfers (UCTs), and cash transfers that were made conditional on the girls’ continued school attendance (CCTs). Given this structure, the research- ers examined education achievements and marriage and childbearing outcomes. They found both transfer programs led to higher rates of continued enrollment (avoiding dropouts); but CCTs had well over double (about 2.3 times) the impact as UCTs. On the other hand, some earlier studies (primarily in Latin Amer- ica) had implied little or no effect of UCTs; perhaps the difference reflects conditions prevailing in low- income Africa. The research found that girls in the CCT program outperformed those in the UCT pro- gram on English reading comprehension (a “modest but significant” difference).

At the same time, the authors found that the CCTs were far more cost-effective in raising enrollment and attendance than the UCTs, even taking into account the extra expenses of running the more administra- tively complex CCT program. The authors examined different transfers and found that even the smallest

amount studied—$4 per month to the parents and $1 per month to the school-age girl—"were sufficient to attain the average schooling impacts observed under the CCT arm.”

On the other hand, the UCT program was found to have a strong impact on “delaying marriage and childbearing—by 44% and 27%, respectively, after 2 years.” And while the CCTs worked better at keep- ing girls in school and learning effectively, they still “had no effect on reducing the likelihood of teenage pregnancies or marriages.” The authors found this was “entirely due to the impact of UCTs on these outcomes among girls who dropped out of school” but whose families continued to receive the transfer benefits (because, after all, the transfer is unconditional). The authors concluded that the “offer of a CCT appears to have been ineffective in dissuading those with a high propensity to drop out of school from getting mar- ried and starting childbearing, especially among girls sixteen or older.” Meanwhile, families living in pov- erty whose daughters did drop out of school ended up receiving nothing, precluding other poverty-reduction benefits.

These findings reflect the difficulties in identifying a single program design to effectively achieve pov- erty reduction, health, education, and social progress goals. As the authors conclude, “This study makes clear that while CCT programs may be more effective than UCTs in obtaining the desired behavior change, they can also undermine the social protection dimen- sion of cash transfer programs.”

Source: Based on Sarah Baird, Craig McIntosh, and Berk Ozler, “Cash or condition? Evidence from a cash transfer experiment,” Quarterly Journal of Economics 126, No. 4 (2011): 1709–1753.

396 PART Two Problems and Policies: Domestic

support services for parents and for children working on the streets, and working to develop social norms against the economic exploitation of children.

The fourth approach, most often associated with the ILO, favors banning child labor. If this is not possible, however, and recognizing that child labor may not always result from multiple equilibria problems, this approach favors banning child labor in its most abusive forms. The latter approach has received much attention in recent years; the ILO’s “Worst Forms of Child Labor Con- vention” was adopted in 1999. The worst forms covered under the convention include “all forms of slavery or practices similar to slavery, such as the sale and trafficking of children, debt bondage and serfdom and forced or compul- sory labor"; child prostitution and pornography; other illicit activities, such as drug trafficking; and work that “by its nature or the circumstances...is likely to harm the health, safety or morals of children.” The ILO set a working target to completely eliminate the worst forms of child labor by 2016; significant prog- ress has been made, but as of 2011, the ILO reported that progress was not fast enough to meet this goal.

A 2003 study by the ILO estimated that eliminating child labor and extend- ing quality schooling for all children up to age 14 over a 20-year period would result in the baseline case of $5 trillion of economic gains (in present dis- counted value), after accounting for opportunity costs. Even when changing the assumptions of the study to be very conservative about the likely income gains, the result is an enormously productive economic investment with a 44% internal rate of return in the baseline case and 23% in a conservative case.26

Finally, many activists in developed countries have proposed the imposition of trade sanctions against countries that permit child labor or at least banning the goods on which children work. This approach is well intentioned, but if the objec- tive is the welfare of children, it needs to be considered carefully, because if chil- dren cannot work in the export sector, they will almost certainly be forced to work in the informal sector, where wages and other working conditions are generally worse. Export restrictions may also make it more difficult for poor countries to grow their way out of poverty. Of course, the worst forms of child labor can never be tolerated. It seems clear that if efforts at banning imports from developing countries were channeled instead into working to secure more public and private development assistance for nongovernmental organizations that work with child laborers, much more would be accomplished to help these children.

8.4 The Gender Gap: Discrimination in Education and Health

Education and Gender

Young females receive less education than young males in most low-income developing countries. While youth literacy is now much higher than it was as recently as 1990, Figure 8.4 shows that in most regions, girls still lag behind boys. Large majorities of illiterate people and those who have been unable to attend school around the developing world are female. The educational gender gap is especially great in the least developed countries in Africa, where female literacy rates can be less than half that of men in countries such as Niger, Mali,

Educational gender gap Male-female differences in school access and completion.

397CHAPTER 8 Human Capital: Education and Health in Economic Development

Guinea, and Benin. The gap is also relatively large in South Asia; in India, the adult female literacy rate is just 47.8%, which is just 65% of the male rate (the female youth literacy rate is 67.7%, 80% of the male youth literacy rate). In Pakistan, the adult female literacy rate is just 36%, only 57% of the male rate (in this case, the female youth literacy rate is 54.7%, some 72% of the male rate). While globally, 123 million youth (aged 15 to 24) lack basic reading and writing skills; 61 per cent of them are young women.27 Recall that the target for Millennium Development Goal 3 (“promote gender equality and empower women”) is to “eliminate gender disparity in primary and secondary educa- tion preferably by 2005, and at all levels by 2015.” Although the 2005 date was missed in many countries, progress has been dramatic in many others. In most low-income countries and many middle-income countries, women make up a minority—sometimes a small minority—of college students. But the long- term trend in higher-income countries for a significantly higher and growing share of female than male enrollment in tertiary (university) education has been extending recently to many upper-middle-income countries in the Mid- dle East, Latin America, and elsewhere.

School completion is also subject to gender inequalities, and the gap is often particularly large in rural areas. For example, in rural Pakistan, 42% of males complete their primary education, while only 17% of females do. In the cities, the gender gap is smaller though still substantial, as 64% of males com- plete primary education versus 50% of females in urban areas.28

Empirical evidence shows that educational discrimination against women hinders economic development in addition to reinforcing social inequality.

FIGuRE 8.4 Youth Literacy Rate, 2008

Source: International Bank for Reconstruction and Development/The World Bank, World Development Indicators, 2010. Reprinted with permission.

East Asia and Pacific

Europe and Central Asia

Latin America and Caribbean

Middle East and North Africa

South Asia

Sub-Saharan Africa

0 25 50 75 100 Percent

Female

Male

398 PART Two Problems and Policies: Domestic

Closing the educational gender gap by expanding educational opportunities for women, a key plank of the Millennium Development Goals, is economi- cally desirable for at least three reasons:29

1. The rate of return on women’s education is higher than that on men’s in most developing countries. [This may partly reflect that, with fewer girls enrolled, the next (marginal) girl to enroll is likely to be more talented on average than the marginal boy.]

2. Increasing women’s education not only increases their productivity (and hence also earnings) in the workplace but also results in greater labor force participation, later marriage, lower fertility, and greatly improved child health and nutrition, thus benefiting the next generation as well. The latter is because a mother ’s education directly increases knowledge that can help child survival, nutrition, education, and indirectly by making possible higher earnings for the family—noting in particular that moth- ers generally spend a somewhat larger fraction of an additional dollar on their children than do fathers.

3. Because women carry a disproportionate burden of poverty, any signifi- cant improvements in their role and status via education can have an important impact on breaking the vicious circles of poverty and inade- quate schooling.

Health and Gender