eTextbook, Study Guide, Manual, Slides, Test Bank Please Visit

Managerial Economics and Organizational Architecture

The McGraw-Hill Series in Economics


Brue, McConnell, and Flynn Essentials of Economics Third Edition

Mandel Economics: The Basics Second Edition

Schiller Essentials of Economics Ninth Edition


Colander Economics, Microeconomics, and Macroeconomics Ninth Edition

Frank and Bernanke Principles of Economics, Principles of Microeconomics, and Principles of Macroeconomics Sixth Edition

Frank and Bernanke Brief Editions: Principles of Economics, Principles of Microeconomics, and Principles of Macroeconomics Second Edition

Karlan and Morduch Economics, Microeconomics, and Macroeconomics First Edition

McConnell, Brue, and Flynn Economics, Microeconomics, and Macroeconomics Twentieth Edition

McConnell, Brue, and Flynn Brief Editions: Economics, Microeco- nomics, and Macroeconomics Second Edition

Miller Principles of Microeconomics First Edition

Samuelson and Nordhaus Economics, Microeconomics, and Macroeconomics Nineteenth Edition

Schiller The Economy Today, The Micro Economy Today, and The Macro Economy Today Fourteenth Edition

Slavin Economics, Microeconomics, and Macroeconomics Eleventh Edition


Guell Issues in Economics Today Seventh Edition

Sharp, Register, and Grimes Economics of Social Issues Twentieth Edition


Gujarati and Porter Basic Econometrics Fifth Edition

Gujarati and Porter Essentials of Econometrics Fourth Edition

Hilmer and Hilmer Practical Econometrics First Edition


Baye and Prince Managerial Economics and Business Strategy Eighth Edition

Brickley, Smith, and Zimmerman Managerial Economics and Organizational Architecture Sixth Edition

Thomas and Maurice Managerial Economics Eleventh Edition


Bernheim and Whinston Microeconomics Second Edition

Dornbusch, Fischer, and Startz Macroeconomics Twelfth Edition

Frank Microeconomics and Behavior Ninth Edition


Romer Advanced Macroeconomics Fourth Edition


Cecchetti and Schoenholtz Money, Banking, and Financial Markets Fourth Edition


O’Sullivan Urban Economics Eighth Edition


Borjas Labor Economics Seventh Edition

McConnell, Brue, and Macpherson Contemporary Labor Economics Tenth Edition


Rosen and Gayer Public Finance Tenth Edition

Seidman Public Finance First Edition


Field and Field Environmental Economics: An Introduction Sixth Edition


Appleyard and Field International Economics Eighth Edition

King and King International Economics, Globalization, and Policy: A Reader Fifth Edition

Pugel International Economics Sixteenth Edition

Managerial Economics and Organizational Architecture Sixth Edition




William E. Simon Graduate School of Business Administration

University of Rochester


Published by McGraw-Hill Education, 2 Penn Plaza, New York, NY 10121. Copyright © 2016 by McGraw-Hill Education. All rights reserved. Printed in the United States of America. Previous editions © 2009, 2007, and 2004. No part of this publication may be reproduced or distributed in any form or by any means, or stored in a database or retrieval system, without the prior written consent of McGraw-Hill Education, including, but not limited to, in any network or other electronic storage or transmission, or broadcast for distance learning.

Some ancillaries, including electronic and print components, may not be available to customers outside the United States.

This book is printed on acid-free paper.

1 2 3 4 5 6 7 8 9 0 QVS/QVS 1 0 9 8 7 6 5

ISBN 978-0-07-352314-9 MHID 0-07-352314-3

Senior Vice President, Products & Markets: Kurt L. Strand

Vice President, General Manager, Products & Markets: Marty Lange

Vice President, Content Design & Delivery: Kimberly Meriwether David

Managing Director: James Heine Director, Product Development: Rose Koos Senior Brand Manager: Katie Hoenicke Lead Product Developer: Michele Janicek Senior Product Developer: Christina Kouvelis Director of Digital Content Development:

Doug Ruby

Director, Content Design & Delivery: Linda Avenarius

Executive Manager: Faye M. Herrig Content Project Managers: Mary Jane

Lampe, Sandra Schnee Buyer: Carol A. Bielski Cover Design: Studio Montage Content Licensing Specialist:

Rita Hingtgen Cover Image: © Corbis / Glow Images Compositor: MPS Limited Typeface: 10.75/12 Adobe Garamond Printer: Quad/Graphics

All credits appearing on page or at the end of the book are considered to be an extension of the copyright page.

Library of Congress Cataloging-in-Publication Data

Brickley, James A. Managerial economics and organizational architecture / James A. Brickley, Clifford

W. Smith, Jerold L. Zimmerman, William E. Simon, Graduate School of Business Administration, University of Rochester.—Sixth edition.

pages cm.—(The McGraw-Hill series in economics) ISBN 978-0-07-352314-9 (alk. paper)

1. Managerial economics. 2. Organizational effectiveness. I. Title. HD30.22.B729 2015 658—dc23


The Internet addresses listed in the text were accurate at the time of publication. The inclusion of a website does not indicate an endorsement by the authors or McGraw-Hill Education, and McGraw-Hill Education does not guarantee the accuracy of the information presented at these sites.

Dedicated to our children—

London, Nic, Alexander, Taylor, Morgan, Daneille, and Amy.

PREFACE The past few decades have witnessed spectacular business failures and scandals. In 2001 and 2002, Enron, WorldCom, Arthur Andersen, as well as other prominent com- panies imploded in dramatic fashion. Internationally, scandals emerged at companies such as Parmalat, Royal Dutch Shell, Samsung, and Royal Ahold. In 2007 and 2008, prominent financial institutions around the world shocked financial markets by reporting staggering losses from subprime mortgages. Société Générale, the large French bank, reported over $7 billion in losses due to potentially fraudulent securities trading by one of its traders. JPMorgan Chase bailed out Bear Stearns, a top-tier in- vestment bank, following their massive subprime losses. Washington Mutual and Lehman Brothers were added to the list of “top business failures of all time.”

Due to these cases and others, executives now face a more skeptical investment community, additional government regulations, and stiffer penalties for misleading public disclosures. A common perception is that bad people caused many of these problems. Others argue that the sheer complexity of today’s world has made it virtu- ally impossible to be a “good” manager. These views have raised the cry for in- creased government regulation, which is argued to be a necessary step in averting fu- ture business problems.

We disagree with this view. We suggest that many business problems result from poorly structured organizational architectures. The blueprints for many of these prominent business scandals were designed into the firms’ “organizational DNA.” This book, in addition to covering traditional managerial economic topics, examines how firms can structure organizations that channel managers’ incentives into actions that create, rather than destroy, firm value. This topic is critical to anyone who works in or seeks to manage organizations—whether for-profit or not-for-profit.

New Demands: Relevant Yet Rigorous Education Thirty years ago, teaching managerial economics to business students was truly a “dis- mal science.” Many students dismissed standard economic tools of marginal analysis, production theory, and market structure as too esoteric to have any real relevance to the business problems they anticipated encountering. Few students expected they would be responsible for their prospective employers’ pricing decisions. Most sought positions in large firms, eventually hoping to manage finance, operations, marketing, or information systems staffs. Traditional managerial economics courses offered few insights that obviously were relevant for such careers. But a new generation of economists began applying traditional economic tools to problems involving corporate governance, merg- ers and acquisitions, incentive conflicts, and executive compensation. Their analysis fo- cused on the internal structure of the firm—not on the firm’s external markets. In this book, we draw heavily from this research and apply it to how organizations can create value through improved organizational design. In addition, we present traditional economic topics—such as demand, supply, markets, and strategy—in a manner that emphasizes their managerial relevance within today’s business environment.

Today’s students must understand more than just how markets work and the prin- ciples of supply and demand. They also must understand how self-interested parties within organizations interact, and how corporate governance mechanisms can control these interactions. Consequently, today’s managerial economics course must cover a broader menu of topics that are now more relevant than ever to aspiring managers facing this post-Enron world. Yet, to best serve our students, offering


Preface vii

relevant material must not come at the expense of rigor. Students must learn how to think logically about both markets and organizations. The basic tools of economics offer students the skill set necessary for rigorous analysis of business problems they likely will encounter throughout their careers.

Besides the heightened interest in corporate governance, global competition and rapid technological change are prompting firms to undertake major organizational restructurings as well as to produce fundamental industry realignments. Firms now attack problems with focused, cross-functional teams. Many firms are shifting from functional organizational structures (manufacturing, marketing, and distribution) to flatter, more process-oriented organizations organized around product or region. Moreover, this pace of change shows no sign of slowing. Today’s students recognize these issues; they want to develop skills that will make them effective executives and prepare them to manage organizational change.

Business school programs are evolving in response to these changes. Narrow tech- nical expertise within a single functional area—whether operations, accounting, fi- nance, information systems, or marketing—is no longer sufficient. Effective man- agers within this environment require cross-functional skills. To meet these challenges, business schools are becoming more integrated. Problems faced by man- agers are not just finance problems, operations problems, or marketing problems. Rather, most business problems involve facets that cut across traditional functional areas. For that reason, the curriculum must encourage students to apply concepts they have mastered across a variety of courses.

This book provides a multidisciplinary, cross-functional approach to managerial and organizational economics. We believe that this is its critical strength. Our interests span economics, finance, accounting, information systems, and financial in- stitutions; this allows us to draw examples from a number of functional areas to demonstrate the power of this underlying economic framework to analyze a variety of problems managers face regularly.

We have been extremely gratified by the reception afforded the first five editions of Managerial Economics and Organizational Architecture. Adopters report that the earlier editions helped them transform their courses into one of the most popular courses within their curriculum. This book has been adopted in microeconomics, human resources, and strategy courses in addition to courses that focus specifically on organizational economics. The prior editions were founded on powerful economic tools of analysis that examine how managers can design organizations that motivate self-interested individuals to make choices that increase firm value. Our sixth edition continues to focus on the fundamental importance of markets and organizational de- sign. We use the failures of Enron (Chapter 1), Société Générale (Chapter 1), Arthur Andersen (Chapter 22), and Adelphia (Chapter 10) as case studies to illustrate how poorly designed organizational architectures can be catastrophic. Other books provide little coverage of such managerially critical topics as developing effective organiza- tional architectures, including performance-evaluation systems and compensation plans; assigning decision-making authority among employees; and managing transfer- pricing disputes among divisions. Given the increased importance of corporate gover- nance, this omission has been both significant and problematic. Our primary objective in writing this book is to provide current and aspiring managers with a rigorous, sys- tematic, comprehensive framework for addressing such organizational problems. To that end, we have endeavored to write the underlying theoretical concepts in simple, intuitive terms and illustrate them with numerous examples—most drawn from actual company practice.

viii Preface

The Conceptual Framework Although the popular press and existing literature on organizations are replete with jargon—TQM, reengineering, outsourcing, teaming, venturing, empowerment, and cor- porate culture—they fail to provide managers with a systematic, comprehensive frame- work for examining organizational problems. This book uses economic analysis to develop such a framework and then employs that framework to organize and integrate the important organizational problems, thereby making the topics more accessible.

Throughout the text, readers will gain an understanding of the basic tools of eco- nomics and how to apply them to solve important business problems. While the book covers the standard managerial economics problems of pricing and production, it pays special attention to organizational issues. In particular, the book will help read- ers understand:

• How the business environment (technology, regulation, and competition in input and output markets) drives the firm’s choice of strategy.

• How strategy and the business environment affect the firm’s choice of organi- zational design—what we call organizational architecture.

• How the firm’s organizational architecture is like its DNA; it plays a key role in determining a firm’s ultimate success or failure, since it affects how people in the organization will behave in terms of creating or destroying firm value.

• How corporate policies such as strategy, financing, accounting, marketing, in- formation systems, operations, compensation, and human resources are inter- related and thus why it is critically important that they be coordinated.

• How the three key features of organizational architecture—the assignment of decision-making authority, the reward system, and the performance-evaluation

system—can be structured to help managers to achieve their desired results.

These three components of or- ganizational architecture are like three legs of the accompanying stool. Firms must coordinate each leg with the other two so that the stool remains functional. More- over, each firm’s architecture must match its strategy; a balanced stool in the wrong setting is dysfunc- tional: Although milking stools are quite productive in a barn, tavern owners purchase taller stools.

Reasons for Adopting Our Approach This book focuses on topics that we believe are most relevant to managers. For in- stance, it provides an in-depth treatment of traditional microeconomic topics (demand, supply, pricing, and game theory) in addition to corporate governance topics (assign- ing decision-making authority, centralization versus decentralization, measuring and

The components of organizational architecture are like three legs of a stool. It is important that all three legs be designed so that the stool is balanced. Changing one leg without the careful consideration of the other two is typically a mistake.

Performance Evaluation (What are the key performance measures

used to evaluate managers and employees?)

Rewards (How are people rewarded for meeting performance goals?)

Decision-Rights Assignment (Who gets to make what decisions?)

Preface ix

rewarding performance, outsourcing, and transfer pricing). We believe these topics are more valuable to prospective managers than topics typically covered in economics texts such as public-policy aspects of minimum-wage legislation, antitrust policy, and income redistribution. A number of other important features differentiate this book from others currently available, such as:

• Our book provides a comprehensive, cross-functional framework for analyzing organizational problems. We do this by first describing and integrating important research findings published across several functional areas, then demonstrating how to apply the framework to specific organizational problems.

• This text integrates the topics of strategy and organizational architecture. Students learn how elements of the business environment (technology, compe- tition, and regulation) drive the firm’s choice of strategy as well as the interaction of strategy choice and organizational architecture.

• Reviewers, instructors, and students found the prior editions accessible and engaging. The text uses intuitive descriptions and simple examples; more technical material is provided in appendices for those who wish to pursue it.

• Numerous examples drawn from the business press and our experiences illus- trate the theoretical concepts. For example, the effect of the 9/11 terrorist attacks on demand curves is described in Chapter 4 and how one devastated company located in the World Trade Center responded is discussed in Chapter 14. These illustrations, many highlighted in boxes, reinforce the underlying principles and help the reader visualize the application of more abstract ideas. Each chapter begins with a specific case history that is used throughout the chapter to unify the material and aid the reader in recalling and applying the main constructs.

• Nontraditional economics topics dealing with strategy, outsourcing, leader- ship, organizational form, corporate ethics, and the implementation of man- agement innovations are examined. Business school curricula often are criti- cized for being slow in covering topics of current interest to business, such as corporate governance. The last six chapters examine recent management trends and demonstrate how the book’s framework can be used to analyze and understand topical issues.

• Problems, both within and at the end of chapter, are drawn from real organiza- tional experience—from the business press as well as our contact with execu- tive MBA students and consulting engagements. We have structured exercises that provide readers with a broad array of opportunities to apply the framework to problems like ones they will encounter as managers.

Organization of the Book • Part 1: Basic Concepts lays the groundwork for the book. Chapter 2 summa-

rizes the economic view of behavior, stressing its management implications. Chapter 3 presents an overview of markets, provides a rationale for the exis- tence of organizations, and stresses the critical role of the distribution of knowledge within the organization.

• Part 2: Managerial Economics applies the basic tools of economic theory to the firm. Chapters 4 through 7 cover the traditional managerial-economics top- ics of demand, production and cost, market structure, and pricing. These four chapters provide the reader with a fundamental set of microeconomic tools and

use these tools to analyze basic operational policies such as input, output, and product pricing decisions. Chapters 8 and 9 focus on corporate strategy—the former on creating and capturing values and the latter on employing game the- ory methods to examine the interaction between the firm and its competitors, suppliers, as well as other parties. These chapters also provide important background material for the subsequent chapters on organizations: A robust understanding of the market environment is important for making sound orga- nizational decisions. Chapter 10 examines conflicts of interest that exist within firms and how contracts can be structured to reduce or control these conflicts.

• Part 3: Designing Organizational Architecture develops the core frame- work of the book. Chapter 11 provides a basic overview of the organiza- tional-design problem. Chapters 12 and 13 focus on two aspects of the as- signment of decision rights within the firm—the level of decentralization chosen for various decisions followed by the bundling of various tasks into jobs and then jobs into subunits. Chapters 14 and 15 examine compensation policy. First we focus on the level of compensation necessary to attract and retain an appropriate group of employees. We then discuss the composition of the compensation package, examining how the mix of salary, fringe ben- efits, and incentive compensation affects the value of the firm. In Chapters 16 and 17, we analyze individual and divisional performance evaluation. Part 3 concludes with a capstone case on Arthur Andersen.

• Part 4: Applications of Organizational Architecture uses the framework that we have developed to provide insights into contemporary management is- sues. Chapters 18 through 23 discuss the legal form of organization, outsourc- ing, leadership, regulation, ethics, and management innovations.

Fitting the Text into the Business Curriculum Our book is an effective tool for a variety of classes at the MBA, executive MBA, and undergraduate level. Although this text grew out of an MBA elective course in the eco- nomics of organizations at the University of Rochester, the book’s modular design al- lows its use in a variety of courses. We have been encouraged by the creativity instruc- tors have shown in the diversity of courses adopting this text. Besides the introductory microeconomics course, this book also is used in elective courses on corporate gover- nance, strategy, the economics of organizations, and human resources management. The basic material on managerial economics is presented in the first 10 chapters. The tools necessary for understanding and applying the organizational framework we de- velop within this text have been selected for their managerial relevance. In our experi- ence, these economics tools are invaluable for those students with extensive work experience, and for those who didn’t major in economics as an undergraduate. Those with an economics background may choose to forgo components of this material. We have structured our discussions of demand, production/cost, market structure, pricing, and strategy to be optional. Thus, readers who do not require a review of these tools can skip Chapters 4 through 9 without loss of continuity.

We strongly recommend that all readers cover Chapters 1 through 3 and 10; these chapters introduce the underlying tools and framework for the text. Chapters 4 through 9, as we noted above, cover the basic managerial-economics topics of demand, costs, production, market structure, pricing, and strategy. Chapters 11 through 17 develop the organizational architecture framework; we recommend that these be covered in

x Preface

sequence. Finally, Chapters 18 through 23 cover special managerial topics: outsourc- ing, leadership, regulation, ethics, and the process of management innovation and man- aging organizational change. They are capstone chapters—chapters that apply and il- lustrate the framework. Instructors can assign them based on their specific interests and available time.

Sixth Edition This book is noted for using economics to analyze real-world management problems. The sixth edition maintains and extends this focus. Changes from the fifth edition include:

• Learning objectives have been added to focus on the core concepts of the chap- ter to aid in the assessment of learning outcomes.

• Extended and more in-depth coverage of important managerial economics concepts, including supply and demand analysis, comparative advantage, con- stant versus increasing cost industries, price competition with differentiated products, inter-temporal decisions (Fisher Separation Theorem) and behav- ioral economics.

• Managerial applications, examples, exhibits, and other boxed materials have been updated.

• Key managerial insights from important recent research in organizational economics have been added.

• Data has been updated, where appropriate.

• We have responded in various ways to reader feedback from earlier editions.

Supplements The following ancillaries are available for quick download and convenient access via the Instructor Library material available through McGraw-Hill Connect®.

• PowerPoint Presentations: Fully updated for the sixth edition, each chapter’s PowerPoint slides are closely tied to the book material and are enhanced by animated graphs. You can edit, print, or rearrange the slides to fit the needs of your course.

• Test Bank: The test bank offers hundreds of questions categorized by level of difficulty, AACSB learning categories, Bloom’s taxonomy, and topic.

• Computerized Test Bank: McGraw-Hill’s EZ Test is a flexible and easy-to- use electronic resting program that allows you to create tests from book- specific items. It accommodates a wide range of question types and you can add your own questions. Multiple versions of the test can be created and any test can be exported for use with course management systems. EZ Test Online gives you a place to administer your EZ Test-created exams and quizzes online. Addition- ally, you can access the test bank through McGraw-Hill Connect®.

• Instructor’s Manual: The instructor’s Manual provides chapter overviews, teaching tips, and suggested answers to the end-of-chapter Self-Evaluation Problems and Review Questions.

Preface xi

Digital Solutions McGraw-Hill Connect® Economics

Less Managing. More Teaching. Greater Learning.

McGraw-Hill’s Connect® Economics is an online assessment solution that connects students with the tools and resources they’ll need to achieve success.

McGraw-Hill’s Connect Economics Features

Connect Economics allows faculty to create and deliver exams easily with selectable test bank items. Instructors can also build their own questions into the system for homework or practice. Other features include:

Instructor Library The Connect Economics Instructor Library is your repository for additional resources to improve student engagement in and out of class. You can se- lect and use any asset that enhances your lecture. The Connect Economics Instructor Library includes all of the instructor supplements for this text.

Student Resources Any supplemental resources that align with the text for student use will be available through Connect.

Student Progress Tracking Connect Economics keeps instructors informed about how each student, section, and class is performing, allowing for more productive use of lecture and office hours. The progress-tracking function enables you to:

• View scored work immediately and track individual or group performance with assignment and grade reports.

• Access an instant view of student or class performance relative to learning objectives.

• Collect data and generate reports required by many accreditation organiza- tions, such as AACSB.

• Connect Insight is a powerful data analytics tool that allows instructors to leverage aggregated information about their courses and students to provide a more personalized teaching and learning experience.

Diagnostic and Adaptive Learning of Concepts: LearnSmart and SmartBook offer the first and only adaptive reading experience designed to change the way students read and learn.

Students want to make the best use of their study time. The LearnSmart adaptive self-study technology within Connect Economics provides students with a seamless combination of practice, assessment, and remediation for every concept in the textbook. LearnSmart’s intelligent software adapts to every student response and automatically delivers concepts that advance students’ understanding while reducing time devoted to the concepts already mastered. The result for every student is the fastest path to mastery of the chapter concepts. LearnSmart:

• Applies an intelligent concept engine to identify the relationships between con- cepts and to serve new concepts to each student only when he or she is ready.

xii Preface

• Adapts automatically to each student, so students spend less time on the topics they understand and practice more those they have yet to master.

• Provides continual reinforcement and remediation, but gives only as much guidance as students need.

• Integrates diagnostics as part of the learning experience.

• Enables you to assess which concepts students have efficiently learned on their own, thus freeing class time for more applications and discussion.

Smartbook is an extension of LearnSmart—an adaptive eBook that helps students focus their study time more effectively. As students read, Smartbook assesses comprehension and dynamically highlights where they need to study more.

For more information about Connect, go to, or contact your local McGraw-Hill sales representative.

McGraw-Hill’s Customer Experience Group We understand that getting the most from your new technology can be challenging. That's why our services don't stop after you purchase our products. You can e-mail our Product Specialists 24 hours a day to get product-training online. Or you can search our knowledge bank of Frequently Asked Questions on our support website. For Customer Support, call 800-331-5094, or visit


McGraw-Hill Create™ is a self-service website that allows you to create customized course materials using McGraw-Hill’s comprehensive, cross- disciplinary content and digital products. You can even access third party content such as readings, articles, cases, videos, and more. Arrange the content you’ve se- lected to match the scope and sequence of your course. Personalize your book with a cover design and choose the best format for your students—eBook, color print, or black-and-white print. And, when you are done, you’ll receive a PDF review copy in just minutes!


Go paperless with eTextbooks from CourseSmart and move light years beyond traditional print textbooks. Read online or offline anytime, anywhere. Access your eTextbook on multiple devices with or without an Internet connection. CourseSmart eBooks include convenient, built-in tools that let you search topics quickly, add notes and highlights, copy/paste passages, and print any page.

Preface xiii


ACKNOWLEDGMENTS No textbook springs from virgin soil. This book has its intellectual roots firmly planted in the work of dozens who have toiled to develop, test, and apply organiza- tion theory. As we detailed in the preface to the first edition, the genesis of this book was a course William Meckling and Michael Jensen taught on the economics of or- ganizations at the University of Rochester in the 1970s. Bill’s and Mike’s research and teaching stimulated our interest in the economics of organizations, prompted much of our research focused on organizational issues, and had a profound effect on this text. No amount of citation or acknowledgments can adequately reflect the encouragement and stimulation that they provided, both personally and through their writings.

Bill and Mike emphasized three critical features of organizational design: (1) the assignment of decision rights within the organization, (2) the reward system, and (3) the performance-evaluation system. These three elements, which we call organi- zational architecture, serve as an important organizing device for this book. As read- ers will discover, this structure offers a rich body of knowledge useful for managerial decision making.

Important contributions to the literature on the economics of organizations have been made by a host of scholars. Through the work of these individuals, we have learned a tremendous amount. A number of our colleagues at Rochester also con- tributed to the development of the book. Ray Ball, Rajiv Dewan, Shane Heitzman, Scott Keating, Stacey Kole, Andy Leone, Glenn MacDonald, Larry Matteson, David Mayers, Kevin Murphy, Michael Raith, Mike Ryall, Greg Schaffer, Ronald Schmidt, Larry Van Horn, Karen Van Nuys, Ross Watts, Gerald Wedig, Michael Weisbach, and Ron Yeaple offered thoughtful comments and suggestions that helped to clarify our thinking on key issues. Don Chew, editor of the Journal of Applied Corporate Fi- nance, provided invaluable assistance in publishing a series of articles based on the book; his assistance in writing these articles improved the exposition of this book enormously. Our collaboration with Janice Willett on Designing Organizations to Create Value: From Strategy to Structure (McGraw-Hill, 2003) enriched our under- standing and exposition of many important topics.

This project also has benefited from an extensive development effort. In addition to generations of Simon School students, dozens of colleagues both in the United States and overseas formally reviewed the manuscript and gave us detailed feedback, for which we are very grateful. We offer our sincere thanks to following reviewers, for their thorough and thoughtful suggestions:

Avner Ben-Ner, University of Minnesota Arnab Biswas, University of West Florida Ben Campbell, The Ohio State University Xiujian Chen, Binghampton University Kwang Soo Cheong, John Hopkins University Abbas Grammy, California State University—Bakersfield Charles Gray, University of Saint Thomas Folke Kafka, University of Pittsburgh Brian Kench, University of Tampa Tom Lee, California State University—Northridge Matthew Metzgar, University of North Carolina Ronald Necoechea, Roberts Wesleyan College Harlan Platt, Northeastern University

Acknowledgments xv

Farhad Rassekh, University of Hartford Amit Sen, Xavier University Richard Smith, University of California—Riverside Neil Younkin, Saint Xavier University

We owe special thanks to Henry Butler, Luke Froeb, Mel Gray, and Chris James; each provided insightful comments on the material. In addition, we are grateful for feedback from over 500 individuals who completed various surveys. Their thoughts served to guide our refinement of this work. We appreciate the efforts of Kathleen DeFazio who provided secretarial support. Finally, we wish to thank our colleagues at McGraw-Hill/Irwin—especially Mike Junior—for their encouragement to pursue this project. Through their vision and publishing expertise, they provided us with insights and feedback to help expand our audience while adhering to our mission.

This book represents the current state of the art. Nonetheless, development is on- going as the research evolves and as we continue to learn. Managerial Economics and Organizational Architecture covers an exciting, dynamic area. We hope that a small portion of that excitement is communicated through this text. Reviewers, instructors, and students frequently mention the relevance of material to the business community, the accessibility of the text, and the logical flow within the text’s framework. However, in the final analysis, it is instructors and their students who will determine the true value of our efforts.

We appreciate the extensive feedback we have received from many readers; their generous comments have improved this edition substantially. Although we had a def- inite objective in mind as we wrote this book, it is important to be open to sugges- tions and willing to learn from others who are traveling a similar yet distinct path. Al- though we are unlikely to please everyone, we will continue to evaluate suggestions critically and to be responsive where consistent with our mission. If readers would like to share their thoughts on this work or their classroom experiences, please feel free to contact any of us at the University of Rochester. Many thanks in advance for the assistance.

[email protected] [email protected]

[email protected]


Contents in Brief

Part 1: Basic Concepts

Chapter 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Chapter 2 Economists’ View of Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Chapter 3 Exchange and Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

Part 2: Managerial Economics

Chapter 4 Demand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 Chapter 5 Production and Cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 Chapter 6 Market Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193 Chapter 7 Pricing with Market Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 Chapter 8 Economics of Strategy: Creating and Capturing Value . . . . . . . . . . 257 Chapter 9 Economics of Strategy: Game Theory . . . . . . . . . . . . . . . . . . . . . . . 296 Chapter 10 Incentive Conflicts and Contracts . . . . . . . . . . . . . . . . . . . . . . . . . . 329

Part 3: Designing Organizational Architecture

Chapter 11 Organizational Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355 Chapter 12 Decision Rights: The Level of Empowerment . . . . . . . . . . . . . . . . . 376 Chapter 13 Decision Rights: Bundling Tasks into Jobs and Subunits . . . . . . . . 410 Chapter 14 Attracting and Retaining Qualified Employees . . . . . . . . . . . . . . . . 438 Chapter 15 Incentive Compensation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469 Chapter 16 Individual Performance Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . 502 Chapter 17 Divisional Performance Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . 537

Capstone Case Study on Organizational Architecture: Arthur Andersen LLP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571

Part 4: Applications of Organizational Architecture

Chapter 18 Corporate Governance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578 Chapter 19 Vertical Integration and Outsourcing . . . . . . . . . . . . . . . . . . . . . . . . 615 Chapter 20* Leadership: Motivating Change within Organizations . . . . . . . . . . . 654 Chapter 21 Understanding the Business Environment:

The Economics of Regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655 Chapter 22 Ethics and Organizational Architecture . . . . . . . . . . . . . . . . . . . . . . 684 Chapter 23* Organizational Architecture and the Process

of Management Innovation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 714

Index 715

Glossary* G-1

*These Web chapters and the Glossary can be found online via the Instructor Library material available through

McGraw-Hill Connect®.



Part 1: Basic Concepts Chapter 1: Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Managerial Economics and Organizational Architecture . . . . . . . . . . . . . . . . . . . . . . 3

Organizational Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 Economic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4

Economic Darwinism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Survival of the Fittest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Economic Darwinism and Benchmarking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Purpose of the Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Our Approach to Organizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10

Chapter 2: Economists’ View of Behavior . . . . . . . . . . . . . . . . . . . . .14 Economic Behavior: An Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15

Economic Choice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15 Marginal Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16 Opportunity Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18 Creativity of Individuals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18

Graphical Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20 Individual Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Indifference Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Opportunities and Constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Individual Choice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Changes in Choice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

Motivating Honesty at Merrill Lynch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Managerial Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Alternative Models of Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

Only-Money-Matters Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Happy-Is-Productive Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Good-Citizen Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Product-of-the-Environment Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

Which Model Should Managers Use? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Behavioral Economics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37 Decision Making under Uncertainty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

Expected Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39 Variability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39 Risk Aversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Certainly Equivalent and Risk Premium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Risk Aversion and Compensation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Appendix A: Consumer Choice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Appendix B: Inter-Temporal Decisions and the Fisher Separation Theorem . . . . . . 61

xviii Contents

Chapter 3: Exchange and Markets . . . . . . . . . . . . . . . . . . . . . . . . . 66 Goals of Economic Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Property Rights and Exchange in a Market Economy . . . . . . . . . . . . . . . . . . . . . . . 68

Dimensions of Property Rights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Gains from Trade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

Basics of Supply and Demand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 The Price Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Shifts in Curves versus Movements along Curves . . . . . . . . . . . . . . . . . . . . . .79 Using Supply and Demand Analysis for Qualitative Forecasts . . . . . . . . . . . .79 Linear Supply and Demand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

Supply and Demand—Extended Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .82 Price versus Quantity Adjustments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .82 Short-Run versus Long-Run Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84 Industry Cost Increases and Price Adjustments . . . . . . . . . . . . . . . . . . . . . . . .86

Prices as Social Coordinators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Efficient Exchange and Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90 Measuring the Gains from Trade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Government Intervention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Externalities and the Coase Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

Markets versus Central Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 General versus Specific Knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Knowledge Creation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .100 Specific Knowledge and the Economic System . . . . . . . . . . . . . . . . . . . . . . 102 Incentives in Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

Contracting Costs and Existence of Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Contracting Costs in Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Contracting Costs within Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

Managerial Decisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 Appendix: Shareholder Value and Market Efficiency . . . . . . . . . . . . . . . . . . . . . . . 114

Part 2: Managerial Economics Chapter 4: Demand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 Demand Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 Demand Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

Law of Demand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 Elasticity of Demand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124 Linear Demand Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

Other Factors That Influence Demand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Prices of Related Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 Other Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

Industry versus Firm Demand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Network Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 Product Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 Product Life Cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 Demand Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 Price Experimentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .142

Statistical Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 Appendix: Demand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

Chapter 5: Production and Cost . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 Production Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

Returns to Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .158 Returns to a Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .159

Choice of Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162 Production Isoquants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162 Isocost Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .164 Cost Minimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .165 Changes in Input Prices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .167

Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .168 Cost Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .169 Short Run versus Long Run . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .171 Minimum Efficient Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .175 Learning Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .177 Economies of Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .178

Profit Maximization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .179 Factor Demand Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .180 Cost Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .184 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .185 Appendix: The Factor-Balance Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .191

Chapter 6: Market Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193 Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .195 Competitive Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .195

Firm Supply . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .195 Competitive Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .198

Barriers to Entry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .201 Incumbent Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .202 Incumbent Advantages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .203 Exit Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .204

Monopoly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .204 Monopolistic Competition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .206 Oligopoly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .208

Nash Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .208 Output Competition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .210 Price Competition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .212 Empirical Evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .213 Cooperation and the Prisoners’ Dilemma . . . . . . . . . . . . . . . . . . . . . . . . . . . .214

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .217

Chapter 7: Pricing with Market Power . . . . . . . . . . . . . . . . . . . . . . 223 Pricing Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .224 Benchmark Case: Single Price per Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .225

Profit Maximization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .225 Estimating the Profit-Maximizing Price . . . . . . . . . . . . . . . . . . . . . . . . . . . . .228 Potential for Higher Profits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .231

Contents xix

Homogeneous Consumer Demands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .232 Block Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .232 Two-Part Tariffs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .233

Price Discrimination—Heterogeneous Consumer Demands . . . . . . . . . . . . . . . . . .234 Exploiting Information about Individual Demands . . . . . . . . . . . . . . . . . . . .236 Using Information about the Distribution of Demands . . . . . . . . . . . . . . . . .239

Bundling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .242 Other Concerns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .244

Multiperiod Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .244 Strategic Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .246 Legal Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .247

Implementing a Pricing Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .248 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .250

Chapter 8: Economics of Strategy: Creating and Capturing Value . . . . . . . . . . . . . . . . . . . . .257

Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .258 Value Creation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .259

Production and Producer Transaction Costs . . . . . . . . . . . . . . . . . . . . . . . . . .261 Consumer Transaction Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .261 Other Ways to Increase Demand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .262 New Products and Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .265 Cooperating to Increase Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .265 Converting Organizational Knowledge into Value . . . . . . . . . . . . . . . . . . . . .266 Opportunities to Create Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .267

Capturing Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .269 Market Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .270 Superior Factors of Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .273 A Partial Explanation for Walmart’s Success . . . . . . . . . . . . . . . . . . . . . . . . .278 All Good Things Must End . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .280

Economics of Diversification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .282 Benefits of Diversification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .282 Costs of Diversification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .284 Management Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .284

Strategy Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .286 Understanding Resources and Capabilities . . . . . . . . . . . . . . . . . . . . . . . . . . .286 Understanding the Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .286 Combining Environmental and Internal Analyses . . . . . . . . . . . . . . . . . . . . .287 Strategy and Organizational Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . .288 Can All Firms Capture Value? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .290

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .291

Chapter 9: Economics of Strategy: Game Theory . . . . . . . . . . . . . . 296 Game Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .297 Simultaneous-Move, Nonrepeated Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . .299

Analyzing the Payoffs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .299 Dominant Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .300 Nash Equilibrium Revisited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .301 Competition versus Coordination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .303 Mixed Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .306 Managerial Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .308

xx Contents

Contents xxi

Sequential Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .310 First-Mover Advantage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .312 Strategic Moves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .312 Managerial Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .313

Repeated Strategic Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .314 Strategic Interaction and Organizational Architecture . . . . . . . . . . . . . . . . . . . . . . .316 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .318 Appendix: Repeated Interaction and the Teammates’ Dilemma . . . . . . . . . . . . . . .323

Chapter 10: Incentive Conflicts and Contracts . . . . . . . . . . . . . . . 329 Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .330 Incentive Conflicts within Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .332

Owner-Manager Conflicts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .332 Other Conflicts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .334

Controlling Incentive Problems through Contracts . . . . . . . . . . . . . . . . . . . . . . . . .334 Costless Contracting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .335 Costly Contracting and Asymmetric Information . . . . . . . . . . . . . . . . . . . . . .338 Postcontractual Information Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .340 Precontractual Information Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .343

Implicit Contracts and Reputational Concerns . . . . . . . . . . . . . . . . . . . . . . . . . . . .347 Incentives to Economize on Contracting Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . .349 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .350

Part 3: Designing Organizational Architecture

Chapter 11: Organizational Architecture . . . . . . . . . . . . . . . . . . . . 355 The Fundamental Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .357

Architecture of Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .357 Architecture within Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .358

Architectural Determinants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .360 Changing Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .364 Interdependencies and Complementarities within the Organization . . . . . . .365

Corporate Culture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .366 When Management Chooses an Inappropriate Architecture . . . . . . . . . . . . . . . . . .370 Managerial Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .371

Evaluating Management Advice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .372 Benchmarking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .372

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .373

Chapter 12: Decision Rights: The Level of Empowerment . . . . . . . 376 Assigning Tasks and Decision Rights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .378 Centralization versus Decentralization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .380

Benefits of Decentralization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .380 Costs of Decentralization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .382 Illustrating the Trade-offs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .385 Management Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .389

Lateral Decision-Right Assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .393

xxii Contents

Assigning Decision Rights to Teams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .394 Benefits of Team Decision Making . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .394 Costs of Team Decision Making . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .395 Management Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .395

Decision Management and Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .397 Decision-Right Assignment and Knowledge Creation . . . . . . . . . . . . . . . . . . . . . .399 Influence Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .401 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .403 Appendix: Collective Decision Making . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .407

Chapter 13: Decision Rights: Bundling Tasks into Jobs and Subunits . . . . . . . . . . . . . . . . . . .410

Bundling Tasks into Jobs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .411 Specialized versus Broad Task Assignment . . . . . . . . . . . . . . . . . . . . . . . . . .411 Productive Bundling of Tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .415

Bundling of Jobs into Subunits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .416 Grouping Jobs by Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .417 Grouping Jobs by Product or Geography . . . . . . . . . . . . . . . . . . . . . . . . . . . .419 Trade-offs between Functional and Product or Geographic Subunits . . . . . .420 Environment, Strategy, and Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . .423 Matrix Organizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .424 Mixed Designs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .426 Network Organizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .426 Organizing within Subunits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .426

Recent Trends in Assignments of Decision Rights . . . . . . . . . . . . . . . . . . . . . . . . .427 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .432 Appendix: Battle of the Functional Managers . . . . . . . . . . . . . . . . . . . . . . . . . . . . .436

Chapter 14: Attracting and Retaining Qualified Employees . . . . . . 438 Contracting Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .440 The Level of Pay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .441

The Basic Competitive Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .441 Human Capital . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .442 Compensating Differentials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .444 Costly Information about Market Wage Rates . . . . . . . . . . . . . . . . . . . . . . . .446

Internal Labor Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .447 Reasons for Long-Term Employment Relationships . . . . . . . . . . . . . . . . . . .447 Costs of Internal Labor Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .448

Pay in Internal Labor Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .449 Careers and Lifetime Pay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .449 Influence Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .454

The Salary–Fringe Benefit Mix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .455 Employee Preferences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .455 Employer Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .457 The Salary–Fringe Benefit Choice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .457

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .463

Chapter 15: Incentive Compensation . . . . . . . . . . . . . . . . . . . . . . . 469 The Basic Incentive Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .470

Incentives from Ownership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .473 Optimal Risk Sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .474

Contents xxiii

Effective Incentive Contracts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .476 Principal-Agent Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .476 Informativeness Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .482 Group Incentive Pay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .483 Multitasking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .485 Forms of Incentive Pay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .486 Incentive Compensation and Information Revelation . . . . . . . . . . . . . . . . . . .487 Selection Effects of Incentive Contracts . . . . . . . . . . . . . . . . . . . . . . . . . . . . .488

Does Incentive Pay Work? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .489 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .493 Appendix: Multitasking Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .498

Chapter 16: Individual Performance Evaluation . . . . . . . . . . . . . . . 502 Setting Performance Benchmarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .505

Time and Motion Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .506 Past Performance and the Ratchet Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . .506

Measurement Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .507 Opportunism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .509

Gaming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .510 Horizon Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .511

Relative Performance Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .511 Within-Firm Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .512 Across-Firm Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .513

Subjective Performance Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .513 Multitasking and Unbalanced Effort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .514 Subjective Evaluation Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .515 Problems with Subjective Performance Evaluations . . . . . . . . . . . . . . . . . . .517

Combining Objective and Subjective Performance Measures . . . . . . . . . . . . . . . . .520 Team Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .521

Team Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .522 Evaluating Teams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .524

Government Regulation of Labor Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .525 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .527 Appendix: Optimal Weights in a Relative Performance Contract . . . . . . . . . . . . . .533

Chapter 17: Divisional Performance Evaluation . . . . . . . . . . . . . . . 537 Measuring Divisional Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .539

Cost Centers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .539 Expense Centers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .542 Revenue Centers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .543 Profit Centers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .544 Investment Centers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .544

Transfer Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .549 Economics of Transfer Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .550 Common Transfer-Pricing Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .556 Reorganization: The Solution If All Else Fails . . . . . . . . . . . . . . . . . . . . . . . .560

Internal Accounting System and Performance Evaluation . . . . . . . . . . . . . . . . . . . .560 Uses of the Accounting System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .560 Trade-offs between Decision Management and Decision Control . . . . . . . . .561

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .564 Capstone Case Study on Organizational Architecture: Arthur Andersen LLP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571

Part 4: Applications of Organizational Architecture

Chapter 18: Corporate Governance . . . . . . . . . . . . . . . . . . . . . . . . 578 Publicly Traded Corporations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .580

Corporate Form of Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .580 Stock Exchanges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .581 Stock Ownership Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .581 Governance Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .582

Separation of Ownership and Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .582 Incentive Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .582 Survival of Corporations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .583 Benefits of Publicly Traded Corporations . . . . . . . . . . . . . . . . . . . . . . . . . . . .583

Top-Level Architecture in U.S. Corporations . . . . . . . . . . . . . . . . . . . . . . . . . . . . .584 Sources of Decision Rights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .585 Shareholders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .586 Board of Directors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .591 Top Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .594 External Monitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .598

International Corporate Governance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .601 Market Forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .604 Sarbanes-Oxley Act of 2002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .606 Corporate Governance: An Historical Perspective . . . . . . . . . . . . . . . . . . . . . . . . . .608 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .610 Web Appendix: Choosing among the Legal Forms of Organization . . . . . . . . . . . A-1

Chapter 19: Vertical Integration and Outsourcing . . . . . . . . . . . . . 615 Vertical Chain of Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .617 Benefits of Buying in Competitive Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .620 Reasons for Nonmarket Transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .621

Contracting Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .621 Market Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .624 Taxes and Regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .626 Other Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .627

Vertical Integration versus Long-Term Contracts . . . . . . . . . . . . . . . . . . . . . . . . . .627 Incomplete Contracting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .628 Ownership and Investment Incentives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .628 Specific Assets and Vertical Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . .629 Asset Ownership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .632 Other Reasons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .633 Continuum of Choice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .634

Contract Duration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .635 Contracting with Distributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .636

Free-Rider Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .636 Double Markups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .638 Regulatory Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .641

xxiv Contents

Trends in Outsourcing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .642 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .645 Appendix: Ownership Rights and Investment Incentives . . . . . . . . . . . . . . . . . . . .650

Web Chapter 20: Leadership: Motivating Change within Organizations . . . . . . 654

Leadership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-3 Vision Setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-3 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-4

Decision Making within Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-5 Incentive Problems and Organizational Politics . . . . . . . . . . . . . . . . . . . . . .20-5 Understanding Attitudes toward Change . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-5

Changing Organizational Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-7 Proposal Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-9

Maintaining Flexibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-9 Commitment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-10 Distributional Consequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-10

Marketing a Proposal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-11 Careful Analysis and Groundwork . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-11 Relying on Reputation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-11 Emphasizing a Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-13

Organizational Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-14 Sources of Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-14 Tying the Proposal to Another Initiative . . . . . . . . . . . . . . . . . . . . . . . . . . .20-17 Coalitions and Logrolling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-18 Is Organizational Power Bad? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-19

The Use of Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-20 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-21 Appendix: Strategic Value of Commitment and Crisis . . . . . . . . . . . . . . . . . . . . .20-23

Chapter 21: Understanding the Business Environment: The Economics of Regulation . . . . . . . . . . . . . . . . . . .655

Importance of Regulation to Managers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .656 Economic Motives for Government Intervention . . . . . . . . . . . . . . . . . . . . . . . . . .658

Defining and Enforcing Property Rights . . . . . . . . . . . . . . . . . . . . . . . . . . . .658 Redressing Market Failures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .660 Redistributing Wealth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .666

Economic Theory of Regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .668 Demand for Regulation: Special Interests . . . . . . . . . . . . . . . . . . . . . . . . . . .669 Supply of Regulation: Politicians . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .669 Market for Regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .671 Deadweight Losses, Transaction Costs, and Wealth Transfers . . . . . . . . . . . .674

Managerial Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .675 Restricting Entry and Limiting Substitutes . . . . . . . . . . . . . . . . . . . . . . . . . . .675 Forming Coalitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .678 On Business Participation in the Political Process . . . . . . . . . . . . . . . . . . . . .679

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .681

Chapter 22: Ethics and Organizational Architecture . . . . . . . . . . . 684 Ethics and Choices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .687

Contents xxv

Corporate Mission: Ethics and Policy Setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . .689 Ethics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .689 Value Maximization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .690 Corporate Social Responsibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .692 Economists’ View of Social Responsibility . . . . . . . . . . . . . . . . . . . . . . . . . .693 Corporate Policy Setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .695 Mechanisms for Encouraging Ethical Behavior . . . . . . . . . . . . . . . . . . . . . . .698

Contracting Costs: Ethics and Policy Implementation . . . . . . . . . . . . . . . . . . . . . .702 Codes of Ethics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .704

Altering Preferences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .705 Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .706 Corporate Culture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .709

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .710

Web Chapter 23: Organizational Architecture and the Process of Management Innovation . . . . . . . . . . . . . . . . .714

Management Innovations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-3 The Demand for Management Innovations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-5

The Rise of TQM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-6 Other Innovations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-7

Why Management Innovations Often Fail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-8 Marketing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-8 Underestimating Costs of Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-11 Failure to Consider Other Legs of the Stool . . . . . . . . . . . . . . . . . . . . . . . .23-12

Managing Changes in Organizational Architecture . . . . . . . . . . . . . . . . . . . . . . .23-16 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-19

Index 715

Web Glossary G-1

xxvi Contents


1 C H A P T E R O U T L I N E

Managerial Economics and Organizational Architecture

Organizational Architecture

Economic Analysis

Economic Darwinism

Survival of the Fittest

Economic Darwinism and Benchmarking

Purpose of the Book

Our Approach to Organizations

E nron Corporation was created in 1985 by the merger of two gas pipeline companies. Convinced that impending deregulation of the energy business would create opportunities for firms with the vi- sion to recognize and the willingness to exploit them, Enron moved

aggressively to build and implement an innovative business model. It was a pioneer in the trading of derivative securities tied to assets like natural gas, electricity, and coal. In its transformation from a traditional, capital- intensive gas pipeline company, it established a dramatically smaller re- liance on hard assets, a flatter management structure, and an entrepreneur- ial, risk-taking environment—one that was quite open to creative and unconventional products and practices. It garnered tremendous recognition for these accomplishments; for six years in a row, it was named “Most In- novative” among Fortune’s Most Admired Companies list.

By 2000, Enron operated in several different business segments: transportation and distribution, supplying gas and electric transmission ser- vices; wholesale services, providing energy services and other products to energy suppliers and other firms; retail services, offering business cus- tomers energy products and services; broadband services, providing various service providers with access to a fiber-optic cable network; and other busi- nesses, including water resources and wind energy. In 1990, 80 percent of Enron’s revenues came from its regulated gas pipeline business, but by 2000, over 90 percent of revenues came from its wholesale energy opera- tions and services segment. Enron’s management argued that vertically integrated giants—like ExxonMobil, whose balance sheet was awash with oil reserves, gas stations, refineries, and other hard assets—were dinosaurs. “In the old days, people worked for the assets,” said CEO Jeffrey Skilling. “We’ve turned it around—what we’ve said is the assets work for the people.”

To finance this rapidly expanding array of businesses Enron relied on its bright young CFO, Andrew Fastow. In addition to tapping traditional sources of debt and equity capital, Fastow made extensive use of sophisti- cated partnerships whose financing details were kept off Enron’s balance



1. Define organizational architecture and discuss how economics can be used to

help managers solve organizational problems and structure more effective orga-

nizational architectures.

2. Define Economic Darwinism and discuss its implications related to the bench-

marking of business practices.

P A R T O N E B a s i c C o n c e p t s

2 Part 1 Basic Concepts

sheet.1 For example, to finance its water business, Enron formed Azurix Corporation and raised $695 million by selling one-third of the company to public investors. Enron also formed a partnership called the Atlantic Water Trust in which it held a 50 percent stake. Enron’s partner was Marlin Water Trust, which was marketed to in- stitutional investors. To help attract lenders, Enron guaranteed the debt with its own stock: If Enron’s credit rating fell below investment grade and the stock fell below a stipulated price, Enron itself would be responsible for the partnership’s $915 million debt.

So long as Enron prospered, these guarantees appeared to cost the company little. But several of Enron’s business segments began to experience significant problems. In late summer of 2000, a power shortage in California resulted in blackouts. Enron (along with other energy companies) was blamed by state politicians: California launched an investigation into price gouging by Enron and other power marketers. Enron’s investment in water concessions in Brazil and England ran into political ob- stacles. For instance, British regulators cut the rates that it was allowed to charge its customers. Enron had a 65 percent stake in a $3 billion power project in India. But the power plant became embroiled in a dispute with its largest customer, who refused to pay for electricity. Following the September 11, 2001, terrorist attacks, the pre- cipitous fall in oil prices generated losses for Enron’s trading operations, and tech- nology changes produced a glut of broadband services.

After reaching a peak of nearly $70 billion in August 2000, Enron’s market value collapsed. Its bankruptcy filing in December 2001 is one of the most spectacular business failures ever seen.2 November 2004 saw it emerge from one of the most complex bankruptcies in U.S. history. After 2006 Enron existed as an assetless shell corporation.

What went wrong? According to BusinessWeek,

Enron didn’t fail just because of improper accounting or alleged corruption at the top. . . . The unrelenting emphasis on earnings growth and individual initiative, coupled with a shocking absence of the usual corporate checks and balances, tipped the culture from one that rewarded aggressive strategy to one that increasingly relied on unethical corner cutting. In the end, too much leeway was given to young, inexperienced managers without the necessary controls to minimize failures. This was a company that simply placed a lot of bad bets on businesses that weren’t so promising to begin with.

Thus, BusinessWeek suggests, Enron’s problems were rooted in a fundamentally flawed organizational design. At fault were three key aspects of the company’s cor- porate structure. First, in the course of flattening its management structure, Enron delegated an extraordinary level of decision-making authority to lower-level em- ployees without retaining an appropriate degree of oversight. Second, performance was evaluated largely on near-term earnings growth and success in closing deals. Third, the company offered enormous compensation to its top performers, which en- couraged excessive risk taking. Enron’s internal risk management group was charged with reviewing deals, but the performance appraisals of the 180 employees within the group were based in part on the recommendations of the very people who

1It should be noted that Fastow was recognized by CFO Magazine in October 1999 with their CFO Excellence

Award for Capital Structure Management. 2While the largest U.S. corporate bankruptcy at the time, Enron is now far from the largest. Lehman Brothers

($691 billion in 2008), Washington Mutual ($327 billion in 2008), WorldCom ($103.9 billion in 2002),

General Motors ($91 billion in 2009 and CIT Group ($80.4 billion in 2009) were all greater in size.

Chapter 1 Introduction 3

generated the deals. Enron’s problems appear to stem, at least in part, from its organizational design.

Managerial Economics and Organizational Architecture Standard managerial economics books address a number of questions that are im- portant for organizational success:

• Which markets will the firm enter?

• How differentiated will the firm’s products be?

• What mix of inputs should the firm use in its production?

• How should the firm price its products?

• Who are the firm’s competitors, and how are they likely to respond to the firm’s product offerings?

Addressing these questions is certainly important—and in this book, we do—yet this tale of Enron’s implosion suggests that this list is woefully incomplete. It is also im- portant to address questions about the internal organization of the firm. A poorly de- signed organization can result in lost profits and even in the failure of the institution.

With the benefit of hindsight, it seems easy to identify elements of Enron’s orga- nization that, if changed, might have reduced the likelihood of its collapse. But the critical managerial question is whether before the fact one reasonably could be ex- pected to identify the potential problems and to structure more productive organiza- tions. We believe the answer to this fundamental managerial question is a resound- ing yes. To examine these issues, a rich framework that can be applied consistently is required.

We are not, of course, the first to recognize the importance of corporate organiza- tion or to offer analysis of how to improve it. The business section of any good book- store displays a virtually endless array of prescriptions: benchmarking, empower- ment, total quality management, reengineering, outsourcing, teaming, corporate culture, venturing, matrix organizations, just-in-time production, and downsizing. The authors of all these books would strongly agree that the firm’s organization and the associated policies, adopted by management, can have profound effects on per- formance and firm value; and all buttress their recommendations with selected sto- ries of firms that followed their advice and realized fabulous successes.

The problem with such approaches, however, is that each tends to focus on a par- ticular facet of the organization—whether it be quality control, or worker empower- ment, or the compensation system—to the virtual exclusion of all others. As a con- sequence, the suggestions offered by the business press are regularly myopic. These publications tend to offer little guidance as to which tools are most appropriate in which circumstances. The implicit assumption of most is that their technique can be successfully adopted by all companies. This presumption, however, is invariably wrong. Ultimately, this literature fails to provide managers with a productive frame- work for identifying and resolving organizational problems.

Organizational Architecture

In contrast to the approach of most business best sellers, we seek to provide a sys- tematic framework for analyzing such issues—one that can be applied consistently in addressing organizational problems and structuring more effective organizations.

4 Part 1 Basic Concepts

In this book, we offer a framework that identifies three critical aspects of corporate organization:

• The assignment of decision rights within the company

• The methods of rewarding individuals

• The structure of systems to evaluate the performance of both individuals and business units

Not coincidentally, these are the same three aspects of the organization we identified in the Enron case.

We introduce the term organizational architecture to refer specifically to these three key aspects of the firm. We hesitate to simply use “organization” to refer to these three corporate features because common usage of that term refers only to the organization’s hierarchical structure—that is, decision-right assignments and report- ing relationships—while it generally ignores the performance-evaluation and reward systems. We thus use organizational architecture to help focus specific attention on all three of these critical aspects of the organization.

Stated as briefly as possible, our argument is that successful firms assign decision rights in ways that effectively link decision-making authority with the relevant infor- mation for making good decisions. When assigning decision rights, however, senior leadership—including both management and the company's Board of Directors— must also ensure that the company’s reward and performance-evaluation systems pro- vide decision makers with appropriate incentives to make value-increasing decisions.

Depending on its specific circumstances, the firm will assign decision-making authority differently (some will decentralize particular decisions but centralize oth- ers) and will tailor its reward and performance-evaluation systems. Even though no two firms might adopt precisely the same architecture, successful firms ensure that these three critical aspects of organizational architecture are coordinated.

Our approach is integrative in the sense that it draws on a number of disciplines: accounting, finance, information systems, marketing, management, operations, politi- cal science, and strategy. But what also distinguishes our approach most clearly from that of the best sellers is our central reliance on the basic principles of economics.

Economic Analysis

Economics long has been applied to questions of pricing policy—for example, “how would raising the price of the firm’s products affect sales and firm value?” We ad- dress standard managerial-economics questions involving pricing, advertising, scale, and the choice of inputs to employ in production. In addition, we apply these same tools to examine questions of organizational architecture. For example, “how would changing a division from a cost center to a profit center change incentives, alter em- ployee decisions, and impact firm value?”

In essence, economics provides a theory to explain the way individuals make choices. For example, in designing organizations, it is important to keep in mind that individuals respond to incentives. Managers and employees can be incredibly re- sourceful in devising methods to exploit the opportunities they face. This also means, however, that when their incentives are structured inappropriately, they can act in ways that reduce the firm’s value. In choosing corporate policies, it is critical that managers anticipate potential responses by customers, suppliers, or employees that might produce undesirable outcomes. Neglecting to do so invites individuals to “game” the system and can result in utter failure of well-intentioned policies.

Chapter 1 Introduction 5

We use economics to examine how managers can design organizations that moti- vate individuals to make choices that will increase a firm’s value. For example, the evidence suggests that the problem highlighted in the accompanying box on chief executive officers slashing R&D budgets prior to their retirement is not widespread.3

The research suggests that these perverse incentives can be controlled by basing the CEO’s incentive compensation on stock prices and by managing CEO succession, so that decision rights are gradually transferred to the successor over the years prior to the final departure. Moreover, CEOs’ postretirement opportunities for election to board seats appear linked to performance over the final years of their tenure.4

Standard economic analysis generally characterizes the firm simply as a “black box” that transforms inputs (labor, capital, and raw materials) into outputs. Little consideration traditionally has been given to the internal architecture of the firm.5 In recent years, economists have focused more on questions of organizational architec- ture.6 But little effort has been devoted to synthesizing the material in an accessible form that emphasizes the managerial implications of the analysis. We apply the basic tools of economics to examine the likely effect on a firm’s value of decisions such as centralization versus decentralization, the bundling of tasks into specific jobs and jobs into business units within the firm, the use of objective versus subjective per- formance measures, compensating employees through fixed versus variable (or “in- centive”) compensation, and retaining activities within the firm versus outsourcing. In sum, we examine how managers can structure organizational architecture to mo- tivate individuals to make choices that increase the firm’s value.

R&D and Executive Turnover Suppose a firm links the CEO’s bonus to earnings and the CEO plans to retire in two years. The CEO might reduce the

firm’s research and development budget to boost earnings this year and next. Five years down the road, earnings will

suffer with no new products coming on stream. By then, however, this CEO will be long gone. In fact, research

suggests that this can be a problem for some R&D-intensive firms.

Source: P. Dechow and R. Sloan (1991), “Executive Incentives and the Horizon Problem,” Journal of Accounting and Economics 14,



3K. Murphy and J. Zimmerman (1993), “Financial Performance Surrounding CEO Turnover,” Journal of Accounting and Economics 16, 273–315.

4J. Brickley, J. Linck, and J. Coles (1999), “What Happens to CEOs after They Retire? New Evidence on

Career Concerns, Horizon Problems, and CEO Incentives,” Journal of Financial Economics 52, 341–378. 5Of course, there are several notable exceptions: F. Knight (1921), Risk, Uncertainty, and Profit (London

School of Economics: London); R. Coase (1937), “The Nature of the Firm,” Economica 4, 386–405; and F.

Hayek (1945), “The Use of Knowledge in Society,” American Economic Review 35, 519–530. 6For example, R. Coase (1960), “The Problem of Social Cost,” Journal of Law and Economics 3, 1–44; S.

Cheung (1969), “Transaction Costs, Risk Aversion, and the Choice of Contractual Arrangements,” Journal of Law and Economics 12, 23–42; A. Alchian and H. Demsetz (1972), “Production, Information Costs, and

Economic Organization,” American Economic Review 62, 777–795; K. Arrow (1974), The Limits of Organization (W. W. Norton: New York); M. Jensen and W. Meckling (1976), “Theory of the Firm:

Managerial Behavior, Agency Costs and Ownership Structure,” Journal of Financial Economics 3, 305–360;

Y. Barzel (1982), “Measurement Costs and the Organization of Markets,” Journal of Law and Economics 25,

27–48; O. Williamson (1985), The Economic Institutions of Capitalism: Firms, Markets, Rational Contracting (Free Press: New York); and B. Holmstrom and J. Tirole (1989), “The Theory of the Firm,” in R.

Schmalensee and R. Willig (Eds.), Handbook of Industrial Economics (North-Holland: New York).

6 Part 1 Basic Concepts

In this analysis, ideas of equilibrium—the interplay of supply and demand in product, labor, and capital markets—represent important constraints on manager- ial decisions. Understanding how prices and quantities change in response to changes in costs, product characteristics, or the terms of sale is a critical manage- rial skill. For example, the more than five-fold increase in crude oil prices from below $12 per barrel in 1999 to over $135 in 2008 prompted oil companies to


Economic Incentives and the Subprime Mortgage Crisis “Subprime mortgages” are made to borrowers who do not qualify for standard market interest rates because of

problems with their credit histories or inability to prove that they have enough income to support the monthly

payments. In March 2007, the value of U.S. subprime mortgages was estimated at $1.3 trillion with over 7.5 million

mortgages outstanding. During the second half of 2007, investors in subprime mortgages such as banks, mortgage

lenders, real estate investment trusts, and hedge funds reported losses of close to $100 billion as a result of subprime

mortgage defaults and devaluations. The stock market fell and became quite volatile as more details about the

mortgage crisis were revealed over time.

One important factor that contributed to this crisis was the incentives of the mortgage brokers that originated the

loans. Mortgage brokers, who originated nearly 70 percent of residential mortgages in recent years, don’t lend their

own money. They are paid for originating loans, which are sold to other investors who bear the primary risk. In many

cases, the more loans they originate, the higher their compensation.

The financial incentives for originating mortgages motivated financial companies to offer products that made it

easier for borrowers to qualify for the loans. For example, companies began offering “stated income loans” that

required no proof of income. Consistent with the theory in this book, some borrowers overstated their incomes. In a

recent review of 100 of these so-called liar loans, almost 60 percent of the stated amounts were exaggerated by over

50 percent. For example, in Atlanta a borrower received a $1.8 million loan by stating that he and his wife were top

executives at a marketing firm who earned more than $600,000 per year with personal assets totaling $3 million. In

reality, he was a phone company technician who earned $105,000 per year with savings of only $35,000.

The financial incentives and associated lack of controls produced not only risky loans but also billions of dollars of

fraud. Rings of fraudulent borrowers would (1) recruit people with good credit to apply for very large loans using false

income and asset statements, (2) find home appraisers to significantly inflate the values of the underlying properties,

(3) pay the much lower asking prices to the sellers, and (4) pocket the difference, splitting the proceeds among the

members of the ring. The houses then would go into foreclosure as the loans were not repaid.

Banking executives subsequently testified that they did not foresee this problem—“fraud was not really a

consideration in our world.” The premise of this book is that a careful analysis of the underlying organizational

architecture (incentives and decision-right assignments) can help managers anticipate these types of problems and

develop mechanisms to reduce their severity.

Source: M. Corkery (2007), “Fraud Seen as a Driver in Wave of Foreclosures,” The Wall Street Journal (December 21), A1.


Creative Responses to a Poorly Designed Incentive System A manager at a software company wanted to find and fix software bugs more quickly. He devised an incentive plan that

paid $20 for each bug the Quality Assurance people found and $20 for each bug the programmers fixed. Since the

programmers who created the bugs were also in charge of fixing them, they responded to the plan by creating bugs in

software programs. This action increased their payoffs under the plan—there were more bugs to detect and fix. The

plan was canceled within a single week after one employee netted $1,700 under the new program.

Source: S. Adams (1995), “Manager’s Journal: The Dilbert Principle,” The Wall Street Journal (May 22), A12.

Chapter 1 Introduction 7

increase production, encouraged petrochemical companies to alter their input mix to economize on a now-more-expensive input, made salespeople reevaluate their decisions about contacting potential customers by phone rather than in person, and encouraged auto producers to focus more on gas economy in the design of new models. Yet these incentives to change depend on the structure of the organization. For instance, a salesperson is less likely to switch to greater reliance on telephone and mail when the firm reimburses all selling expenses than when salespeople are responsible for the costs of contacting potential customers.

Economic Darwinism

Survival of the Fittest7

The collapse of Enron, Charles Darwin might have noted, is an example of how com- petition tends to weed out the less fit. As described in The Origin of Species, natural history illustrates the principle of “survival of the fittest.” In industry, we see economic Darwinism in operation as competition weeds out ill-designed organiza- tions that fail to adapt. Competition in the marketplace provides strong pressures for efficient decisions—including organizational decisions. Competition among firms dictates that those firms with low costs are more likely to survive. If firms adopt in- efficient, high-cost policies—including their organizational architecture—competi- tion will place strong pressures on these firms to either adapt or close.

Fama and Jensen suggest that “the form of organization that survives in an activ- ity is the one that delivers the product demanded by customers at the lowest price while covering costs.” This survival criterion helps highlight that while a well- crafted organizational architecture can contribute to a firm’s success, it is not suffi- cient for success. The firm must have a business strategy that includes products for which the prices customers are willing to pay exceed costs. The potential for value creation by a company that manufactures only buggy whips is quite limited no mat- ter how well structured the firm’s organizational architecture.

Nonetheless, given a firm’s business strategy (including its product mix), its choice of organizational architecture can have an important impact on profitability and value. An appropriate architecture can lower costs by promoting efficient pro- duction; it also can boost the prices customers are willing to pay by helping to ensure high-quality production, reliable delivery, and responsive service.

Economic Darwinism and Benchmarking

In the biological systems that Darwin analyzed, the major forces at work were ran- dom mutations in organisms and shocks from the external environment (for instance, from changes in weather). But in the economic systems on which we focus, pur- poseful voluntary changes occur. For instance, in order to compete more effectively with Coke, Pepsi copied many of Coke’s practices. Pepsi spun off its fast-food chains

7This section draws on A. Alchian (1950), “Uncertainty, Evolution, and Economic Theory,” Journal of Political Economy 58, 211–221; G. Stigler (1951), “The Economics of Scale,” Journal of Law and Economics 1, 54–71; and E. Fama and M. Jensen (1983), “Separation of Ownership and Control,” Journal of Law and Economics 26, 301–325.

8 Part 1 Basic Concepts

(Taco Bell, KFC, and Pizza Hut) to focus on its core business—just as Coca Cola had done. Also, Pepsi changed its network of bottlers. One analyst remarked, “Pepsi is starting to look a lot more like Coke.”8 In fact, this practice has been formalized in the process of benchmarking.

Benchmarking generally means looking at those companies that are doing some- thing best and learning how they do it in order to emulate them. But this process also occurs in less formal ways. As Armen Alchian argued, “Whenever successful enter- prises are observed, the elements common to those observed successes will be asso- ciated with success and copied by others in their pursuit of profits or success.”9 For example, if the cover article in the next Fortune reports an innovative inventory con- trol system at Toyota, managers across the country—indeed, around the globe—will read it and ask, Would that work in my company, too? Undoubtedly, the managers with the strongest interest in trying it will be those within firms currently suffering inventory problems.10 Some will achieve success, but others may experience disas- trous results caused by unintended though largely predictable organizational “side effects” (like Fastow’s unchecked incentive for risk taking).

8N. Harris (1997), “If You Can’t Beat ’Em, Copy ’Em,” BusinessWeek (November), 50. 9A. Alchian. “Uncertainty, Evolution, and Economic Theory,” The Journal of Political Economy, Vol. 58,

No. 3 (Jun., 1950), p. 218. 10This raises the question of why any firm with an innovative idea would voluntarily disclose it. Perhaps the

free publicity outweighs the lost competitive advantage.


Economic Darwinism: The Growth in Lead Directors The collapse of Enron in December 2001 and subsequent scandals at Adelphia, Tyco, WorldCom, and other companies

in 2002 shook public confidence in corporate governance. In July 2002, the United States enacted the Sarbanes–Oxley

Act, which mandated substantial changes in corporate accounting and governance practices. Additional scandals and

failures during the 2007–2008 financial crisis raised additional concerns about corporate governance and motivated

additional legislation and regulation.

These events altered the basic business environment for publicly traded corporations. Over the past decade,

investors, regulators, stock exchanges, the media and the general public have placed increased pressure on corporate

boards of directors to become more independent and diligent in their monitoring of CEOs. One important trend in

corporate governance has been the large increase in presiding and lead directors. Presiding directors are independent

directors (a director with no other direct ties to the company or corporate management) who chair executive sessions of

outside directors. Lead directors are more powerful, taking on additional responsibilities (such as serving as the

principal liaison between the independent directors and the CEI and taking the lead role in overseeing formal

evaluations of board members and the CEO). In 2003, only 36 percent of S&P 500 firms had presiding or lead

directors, compared to 90 percent in 2013. Over 60 percent of the S&P 500 firms with presiding or lead directors in

2013 employed the more powerful position of lead director.

If you were to benchmark the current governance practices of large publicly traded corporations, you would find the

appointment of a lead director is a dominant surviving practice in the current business environment. “One size,”

however, is unlikely to fit all firms. Managers should not simply adopt the prevailing organizational practices of other

firms. More careful analysis is required.

Source: Spencer Stuart (2013), “Spencer Stuart Board Index 2013,”

Chapter 1 Introduction 9

Although competition tends to produce efficiently organized firms over the longer run, uncritical experimentation with the organizational innovation du jour can ex- pose the firm to an uncomfortably high risk of failure. Organizational change is ex- pensive. Moreover, successful organizations are not just a collection of “good ideas.” The elements of a successful organization must be carefully coordinated: The differ- ent elements of the firm’s architecture must be structured to work together to achieve the firm’s goals. For this reason, it is important to be able to analyze the likely con- sequences of a contemplated organizational change and forecast its impact on the en- tire firm.

This concept of economic Darwinism thus has important managerial implica- tions. First, existing architectures are not random; there are sound economic ex- planations for the dominant organization of firms in most industries. Second, sur- viving architectures at any point in time are optimal in a relative rather than an absolute sense; that is, they are the best among the competition—not necessarily the best possible. Third, if the environment in which the firm operates changes—if technology, competition, or regulation change—then the appropriate organiza- tional architecture normally changes as well. These three observations together suggest that although improvements in architecture are certainly always possible, a manager should resist condemning prevailing structures without careful analysis. Before undertaking major changes, executives should have a good understanding of how the firm arrived at its existing architecture and, more generally, develop a broader perspective of why specific types of organizations work well in particular settings. Finally, an executive should be particularly skeptical of claimed benefits of proposed organizational changes if the environment has been relatively stable.

Purpose of the Book The primary thrust of this book is to provide a solid conceptual framework for ana- lyzing organizational problems and structuring an effective organizational architec- ture. The book also provides basic material on managerial economics and discusses how it can be used for making operational decisions—for example, input, output, and pricing decisions. This material additionally supplies a set of tools and an un- derstanding of markets, that is, important for making good organizational decisions.


Organizing Xerox Service Center Xerox has developed an expert system to assist employees who answer the company service center’s 800 number to

help callers who have problems with their photocopy machines. The system is designed to lead the employee through a

set of questions to diagnose and fix the problem. If the machine operator cannot fix the problem with the assistance of

the input from the service center employee, a service representative is dispatched to make a service call. This expert

system is designed to evolve more effective prompts as experience accumulates. This will be accomplished by having

service representatives call the service center after a service call. The nature of the problem and the actions taken are to

be entered into the system. Xerox bases pay for the individuals who answer the 800 number on the number of service

calls they handle; it bases compensation for service representatives on the number of service calls they make. Discuss

the incentives these compensation practices create.

10 Part 1 Basic Concepts

Our Approach to Organizations

We begin with two basic notions: People act in their own self-interest, and individu- als do not all share the same information. As we have indicated, this framework sug- gests that the three critical elements of organizational architecture are the assignment of decision rights, the reward system, and the performance-evaluation system. Suc- cessful organizations assign decision rights in a manner that effectively links deci- sion-making authority with the relevant information to make good decisions. Corre- spondingly, successful organizations develop reward and performance-evaluation systems that provide self-interested decision makers with appropriate incentives to make decisions that increase the values of their organizations.

It is also important to note that modern organizations are extremely complex and that developing an understanding of how people within them behave is difficult. As in any book that addresses this set of topics, we face difficult trade-offs between adding more institutional richness to infuse more texture of the actual environment versus omitting details to keep the analysis more focused and manageable. At certain points (especially where little prior formal analysis of the problem exists), we take quite complex problems and discuss them in terms of simplified examples. Nonethe- less, our experience suggests that in these cases, we derive important managerial in- sights to these topics through our admittedly simple examples.

Finally, we believe that a powerful feature of this economic framework is that it can be extended readily to incorporate a broad array of other managerial policies such as finance, accounting, information systems, human relations, operations, and marketing. In this sense, this book can play an important integrating role across the entire business curriculum. Such integration is becoming increasingly important with the expanded use of cross-functional teams within the business community.


Transfers of Organizational Architecture across the Global Economy In 1996, Tianjin Optical & Electrical Communication Group was typical of a Chinese state-owned company. Although

the electronics manufacturer boasted skilled technicians, mismanagement left the company at the brink of bankruptcy.

Motorola, Inc., changed that. It offered to take Tianjin Optical as a supplier, but only if Tianjin adopted the U.S.

telecommunications company’s quality-control and management practices. By 1999, Tianjin Optical was selling a third

of its production to Motorola and reported a small profit. “Now, we think we can survive,” says Zhang Bingjun, Tianjin

Optical’s chairman.

Each Tianjin employee receives an average of two weeks a year in classroom instruction stressing modern

management practices. That effort has paid off: The Tianjin assembly lines produce a slim cellular phone every 2

seconds with virtually the same defect rate as in Motorola’s U.S. plants. Motorola also provides training for more than

100 outside suppliers to boost the quality of their output. Motorola budgets about $2 million annually to “show

[potential suppliers] Western management practices and create a mindset where they understand what we’re doing and

why,” says a training director, Ying Shea.

This assistance in establishing a more effective organizational architecture and internal operating policies provided

by a U.S. multinational corporation to its Chinese partners is but one example of the vital role that foreign businesses

play within the Chinese business sector. Since China opened itself to foreign investment three decades ago, foreign

companies have become an important conduit for economic reform. They have introduced not just modern production

technology but also more efficient organizational architecture to the Chinese business community. Some estimates

suggest that including these collateral benefits, foreign firms and their joint ventures account for as much as a fifth of

China’s trillion-dollar economy.

Source: E. Guyot (1999), “Foreign Companies Bring China More Than Jobs,” The Wall Street Journal (September 15), A26.

1 2

Chapter 1 Introduction 11


Société Générale was founded in the 1860s and in

2013 was France’s third largest bank. Beginning in

the mid-1980s, it pioneered some of the most com-

plex instruments in international finance and became

a global powerhouse in trading derivatives like futures

and options. Through its trading activities, the bank

earned billions of dollars and gained the respect of

bankers throughout the world. In January 2008, Risk, a monthly magazine about risk management, named

Société Générale its “Equity Derivatives House of the


In late January 2008, Société Générale an-

nounced that it had discovered fraudulent securities

trading by one of its low-level traders, Jérôme

Kerviel. The bank reported that it expected the

fraud to cost it a staggering $7.14 billion, making it

one of the largest financial frauds in history. The

announcement shocked the financial markets and

made front-page headlines around the world. Ob-

servers questioned whether the bank could ever re-

gain its former reputation and whether it could con-

tinue to exist without merging with another bank.

Société’s CEO Daniel Bouton asserted that the

fraud was the result of one employee’s illegal activ-

ities, did not involve other employees at the bank,

and represented the aberrant and unexplainable ac-

tions of one “rogue trader.” He characterized

Kerviel’s actions as “irrational” since the trades

were made on behalf of the bank “netting the trader

no personal gains.” Bouton emphasized that

Kerviel was a low-level employee who had an an-

nual salary and bonus for 2007 of less than


In principle Kerviel engaged in a quite simple

operation: arbitrage-trading on small differences

between various stock market indexes such as the

CAC in France and the DAX in Germany. Kerviel

should have been able to lock in a virtually riskless

profit by selling a security on the exchange with the

higher price, while simultaneously buying an equiv-

alent instrument on the exchange with the lower

price. And although price differences are typically

small, such arbitrage can produce a substantial profit

if done in sufficient volume. In this arbitrage busi-

ness, although Société Générale might accumulate

large positions on both exchanges, those securities

that it bought and those it sold should balance. The

bank was supposed to face little net exposure to

price changes.

What the bank discovered was that Kerviel had

bought securities on both markets. In effect, he had

made enormous bets that European stock prices

would increase. But they had fallen, and as a result

the bank incurred a substantial loss.

The subsequent investigation revealed that

Kerviel had been placing huge unhedged bets on

European stocks for over a year. Prior to becoming

a trader he had worked in the bank’s trading ac-

counting office. His knowledge of the bank’s risk-

management system allowed him to conceal the

trades and bypass the firm’s control system. He

knew the timing of the nightly reconciliation of the

day’s trades and would delete and then re enter his

unauthorized transactions without being caught.

Bank managers, however, had apparently dismissed

several warning signs about Kerviel’s transactions.

For example, the surveillance office at Eurex, one

of Europe’s biggest exchanges, alerted a compli-

ance officer at the bank that for seven months a

trader named Kerviel had engaged in “several

transactions” that raised red flags. Kerviel’s super-

visors accepted his explanations for these trades

apparently without much investigation.

Various bank officials, investigators, and traders

who worked with Kerviel have concluded that So-

ciété Générale “allowed a culture of risk to flour-

ish, creating major flaws in its operations” that en-

abled Kerviel’s actions to proceed. Several current

and former employees interviewed by the New York Times, indicated that Société Générale traders

were rewarded for making risky investments with

the bank’s money and that it was not uncommon

for traders briefly to exceed limits imposed on their

trading, despite controls meant to prohibit this ac-

tivity. Risk taking apparently was “embraced, as

long as it made money for the bank.” Top execu-

tives and other managers at the bank had received

large bonuses because of the bank’s successful

trading operations.

Kerviel told investigators that all he wanted was to

be respected and to earn a large bonus. He had come

from a modest background and did not have the

12 Part 1 Basic Concepts

A. Alchian (1950), “Uncertainty, Evolution, and Economic Theory,” Journal of Political Economy 58, 211–221.

M. Jensen (1983), “Organization Theory and Methodology,” The Accounting Review 58, 319–339.

M. Jensen and W. Meckling (1992), “Specific and General Knowledge, and Organizational Struc-

ture,” Journal of Applied Corporate Finance 8:2, 4–18.

1–1. Briefly describe Economic Darwinism.

1–2. The Wall Street Journal11 reports that

Franchisees, who pay fees and royalties in exchange for using franchisers’ business formats, have become much more militant in recent years about what they see as mistreatment by franchisers. In general, Ms. Kezios is seeking federal and state laws to give franchisees more power in franchise arrangements. Among her goals: creating legally pro- tected exclusive territories for franchisees.

How would you expect existing franchisees to react to this proposed regulation? How

would you expect a potential new franchisee to react to this proposed regulation?

Solutions to Self-Evaluation Problems 1–1. Economic Darwinism is the economic counterpart of natural selection in biology. Competi-

tion in the marketplace weeds out those organizations that are less efficient and fail to adapt

to the environment. The result is survival of the fittest.

1–2. Reducing the likelihood of encroachment by the franchiser benefits the existing franchisees

to the extent that it shifts future profits from the franchiser to the franchisee. Thus, existing

franchisees are likely to favor the proposed regulation. Potential new franchisees are less

likely to favor the proposal. Presumably, they will have to pay a higher price for a new fran-

chise if the franchiser has to grant the franchisee an exclusive territory. The potential fran-

chisee might prefer to have a nonexclusive territory at a lower price. In any case, the fran-

chiser is unlikely to favor the proposal; if it were efficient to convey exclusive territories, the

original contract could have been structured that way.

1–1. What are the three aspects of organizational architecture?

1–2. In the process of benchmarking, a colleague of yours notes that Lincoln Electric, a producer

of electric arc welders, has much higher productivity than does your company. Unlike your

Review Questions

Self-Evaluation Problems

Suggested Readings

educational pedigree of many of his coworkers who

had advanced degrees in math or engineering from

the prestigious Grandes Ecoles—the MITs of France.

He was noted for working very long hours and had

worked his way up in the bank from being a clerk to

a trader. One of his primary goals was to have his

supervisors recognize his “financial genius.”

1. Do you agree with Société Générale’s CEO

that Kerviel’s actions were “irrational”?

2. Discuss how the bank’s organizational archi-

tecture contributed to the problem.

3. What lessons might you learn from this case if

you were an executive at another bank?

Source: This application is based on a series of articles from the New York Times published in early 2008. In particular see N. D. Schwartz and K. Bennhold, 2008, “A Trader’s Secrets, a Bank’s Missteps,” (February 5).

11J. A. Tannenbaum (1995), “Focus on Franchising: Franchisee Gains,” The Wall Street Journal (June 19), B2.

Chapter 1 Introduction 13

firm, Lincoln has an extensive piece-rate compensation system; much of its employees’

total compensation is simply the number of units produced times the piece rate for that type

unit. Your colleague recommends that your company adopt a piece-rate compensation sys-

tem to boost productivity. What do you advise?

1–3. In the life insurance industry, we see two major ownership structures—common stock in-

surers and mutual insurers. In a common stock company, the owners—its stockholders—are

a separate group from its customers—the policyholders. In a mutual, the policyholders are

also the owners of the company. It has been argued that mutual insurance companies are

dinosaurs—they are large, slow, bureaucratic, and inefficient. How would you respond to

such an argument?


2 C H A P T E R O U T L I N E

Economic Behavior: An Overview

Economic Choice Marginal Analysis Opportunity Costs Creativity of Individuals

Graphical Tools Individual Objectives Indifference Curves Opportunities and Constraints Individual Choice Changes in Choice

Motivating Honesty at Merrill Lynch Managerial Implications Alternative Models of Behavior

Only-Money-Matters Model Happy-Is-Productive Model Good-Citizen Model Product-of-the- Environment Model

Which Model Should Managers Use? Behavioral Economics Decision Making under Uncertainty

Expected Value Variability Risk Aversion Certainty Equivalent and Risk Premium Risk Aversion and Compensation

Summary Appendix A: Consumer Choice Appendix B: Inter-Temporal Decisions and the Fisher Separation Theorem

I n May 2002, Merrill Lynch agreed to pay $100 million to settle charges that its analysts had recommended stocks to clients that they privately thought were poor investments. Internal e-mails provided strong sup- port for this claim leveled by the New York State attorney general. For

example, InfoSpace, an Internet services company, was rated highly in ana- lysts’ reports distributed to clients, yet privately the analysts suggested that it was a “powder keg” and a “piece of junk.” Although InfoSpace’s share price dropped from $261 to $14, Merrill analysts never recommended sell- ing the stock. Merrill analysts rated Excite@Home “accumulate or buy,” while privately the investment team called it a “piece of crap.”

This episode at Merrill sent shock waves through other major investment houses—indeed through the entire investment community. Other investment firms publicly stated that they were taking strong steps to make sure that the situation at Merrill would not be repeated within their organizations. Fortune magazine ran a cover story entitled, “In Search of the Last Honest Analyst.”1 The scandal generated significant concerns throughout the world among both the general public and government regulators. For example, the New York attorney general began a sweeping investigation of analysts at Salomon Smith Barney and other investment firms that had recommended WorldCom to investors. In July 2002, WorldCom became the biggest company ever to file for bankruptcy in U.S. history. In December 2002, the nation’s 10 top investment banks agreed to a $1.2 billion settlement with regulators aimed at “protecting investors from brokerages’ conflicts of interest.”

Economists’ View of Behavior


1. Describe the economic model of behavior.

2. Define and apply marginal analysis in managerial decisions.

3. Define and apply the concept of opportunity costs.

4. Use graphs to explain, predict, and affect behavior in a wide range of


5. Contrast the implications of the economic model with those from other

behavioral models used by some managers.

6. Identify the key concepts that are used to mitigate risk when making decisions

under uncertainty.

1June 10, 2002, issue.

Chapter 2 Economists’ View of Behavior 15

Managers at Merrill, Salomon Smith Barney, and other investment companies had to act quickly to address this potential problem. As a first step, management had to understand what motivated the Merrill analysts to mislead their investment clients. Only then could they choose a policy to redress the situation. If management thought this problem was caused by a few dishonest employees, the appropriate response would have been to try to identify and fire those employees. If, instead, management believed the problem was caused by disgruntled employees taking out their frustra- tions on customers, a potential response would have been to adopt a job-enrichment program to increase employee satisfaction and, it would be hoped, analyst honesty. Alternatively, Merrill Lynch might have created incentives through its compensation plan that caused its analysts to issue misleading investment reports. If so, the appro- priate response would be to restructure its compensation plan. Many other assump- tions and responses are possible.

The example of Merrill Lynch illustrates a general point: Managers’ responses to problems are likely to depend on their understanding of people’s motives and their forecast of people’s reactions—their responses thus depend on their underlying model of behavior. Most managerial actions attempt to change the behavior of indi- viduals, such as employees, customers, union officials, or subcontractors. Managers with different understandings (or models) of what motivates behavior are likely to make different decisions and take different actions.

We begin this chapter by briefly summarizing the general framework economists use to examine individual behavior. Selected graphical tools are introduced to aid our analysis. Next, we use this economic framework to analyze the problem at Merrill Lynch. The managerial implications of this analysis are discussed. We contrast this economic view of behavior with alternative views and explore why the economic framework is particularly useful in managerial decision making. Finally, we analyze decision making under uncertainty. In Appendix A, we analyze the problem of con- sumer choice in more detail and in Appendix B, we illustrate how the graphical frame- work we present in this chapter can be used for analyzing inter-temporal choices.

Economic Behavior: An Overview Individuals have unlimited wants. People generally want greater wealth, more atten- tive service, larger houses, more luxurious cars, and additional personal material items. They want more time for leisure activities. Most also want to improve the plight of others—starving children, the homeless, and disaster victims. People are concerned about vitality, religion, integrity, and gaining the respect and affection of others.

In contrast to wants, resources are limited. Households face limited incomes that preclude all the purchases and expenditures that household members might like to make. The available amount of land, trees, and other natural resources is finite. There are only 24 hours in a day. People become ill; death is inevitable.

Economic Choice

Economic analysis is based on the notion that individuals assign priorities to their wants and choose their most preferred options from the available alternatives. If Kathy Measer is confronted with a choice between a laptop or a desktop computer, she can tell you whether she prefers one over the other or whether she is indifferent

16 Part 1 Basic Concepts

between the two. Depending on the relative prices of the two products, she purchases her preferred alternative. If Kathy has a weekly budget of $1,000, she considers the many ways she might spend the money and then chooses the package of goods and services that will maximize her personal happiness. She cannot make all desired pur- chases on her limited budget. However, this choice is optimal for Kathy, given her limited resources.

Economists do not assert that people are selfish in the sense that they care only about their own personal consumption. Within the economic paradigm, people also care about such things as charity, family, religion, and society. For instance, Kathy will donate $100 to her church, as long as the donation provides greater satisfaction than alternative uses of the money.

Economists, however, often assume for modeling purposes that people care only about their own wealth to simplify the analysis. While wealth is not the only thing that people care about, it is very important to most people. Economic models based on this simplifying assumption often perform quite well relative to more complicated models that add unnecessary complexity to the analysis. Some situations, however, can require models that are based on different assumptions.

Economists do not contend that individuals are supercomputers that make infalli- ble decisions. Individuals are not endowed with perfect knowledge and foresight, nor is additional information costless to acquire and process.2 For example, Kathy might order an item from a restaurant menu only to find that she dislikes what she is served. Within this economic paradigm, she simply does the best she can in the face of her imperfect knowledge. But she learns from her experience and does not repeat the same mistakes in judgment time after time.3

Marginal Analysis

Marginal costs and benefits are the incremental costs and benefits that are associated with making a decision.4 It is the marginal costs and benefits that are important in economic decision making. An action should be taken whenever the incremental benefits of that action exceed its incremental costs. Mary O’Dwyer has a contract to help sell products for an office supply company. She is paid $50 for every sales call that she makes to customers. Thus, Mary’s marginal benefit for making each addi- tional sales call is $50. Mary enjoys playing tennis more than selling. If she places a marginal value of more than $50 on the tennis that she would forgo by making an

2Economists sometimes use the idea of bounded rationality. Under this concept, individuals act in a

purposeful and intendedly rational manner. However, they have cognitive limitations in storing, processing,

and communicating information. It is these limitations which make the question of how to organize economic

activity particularly interesting. H. Simon (1957), Models of Man (John Wiley & Sons: New York). 3At least this learning appears to occur outside the comics. For decades, Charlie Brown from Peanuts continued to try to kick the football held by Lucy van Pelt. Yet Lucy always pulled the ball at the last second.

Few individuals are as incurably optimistic as Charlie Brown—they learn. 4Technical note: Marginal costs and benefits are typically defined as changes in costs and benefits associated

with very small changes in a decision variable. For instance, the marginal costs of production are the

additional costs from producing a small additional amount of the product (for instance, one more unit). Often

decisions involve discrete choices, such as whether or not to build a new plant. In these cases, it is not

possible to define a small change in the decision variable. Incremental costs and benefits are those costs and

benefits which vary with such a decision. For our present discussion, the technical distinction between

marginal and incremental is not important.

Chapter 2 Economists’ View of Behavior 17

extra call, she should not make any more sales calls that day—the marginal costs would have exceeded the marginal benefits. She continues to make additional sales calls as long as the reduction in tennis playing is valued at less than $50.5

Marginal analysis is a cornerstone of modern economic analysis. In economic de- cision making, “bygones are forever bygones.” Costs and benefits that have already been incurred are sunk (assuming they are nonrecoverable) and hence are irrelevant to the current economic decision. Mary paid $5,000 to join a tennis club last month. This fee does not affect her current decision of whether to play tennis or make an extra sales call. That expenditure is ancient history and does not affect Mary’s cur- rent trade-offs.

As another example, consider Ludger Hellweg who owns a company that installs wood floors. He is offered $20,000 to install a new floor. The cost of his labor and other operating expenses (excluding the wood) are $15,000. He has wood for the job in inventory. It originally cost him $2,000. Price increases have raised the market value of the wood to $6,000, and this value is not expected to change in the near fu- ture. Should he accept the contract?

He should compare the incremental costs and benefits from the project. The mar- ginal benefit is $20,000. The marginal cost is $21,000—$15,000 for the labor and operating expenses and $6,000 for the wood. The historic cost for the wood of $2,000 is not relevant to the decision. To replace the wood used on this job costs


Marginal Analysis of Customer Profitability Banks often provide multiple products and services to the same customer (checking and savings accounts, mortgages,

lines of credits, business loans, credit cards, international banking services, insurance, and so on). In the 1980s, most

banks did not consolidate this information, and so it was difficult to determine if serving a given customer was

profitable or not. Today many banks use “profitability software” to consolidate information on each customer. Many

banks have found to their surprise that the incremental costs for serving many of their customers are larger than the

incremental revenues. Fleet Bank, for example, found that as many as one-half of their customers were unprofitable.

Armed with this information, banks work hard to maintain high-profit customers, while they either eliminate or alter

services to unprofitable customers. For example, at many banks profitable customers are given special designations,

such as “Gold Customer Status,” and the banks extend special services to them. Preferred customers are frequently

given special toll free lines; branch managers are furnished with their names and are instructed to meet and greet them

when they visit a branch. They are assigned personal bankers, who call and introduce themselves. Customers are

assigned profitability codes, for example, so employees can know whether they are dealing with a 5, 4, 3, 2, or 1 type

customer (five being most profitable). When the loans for unprofitable customers come up for renewal, they are

renewed at a higher rate, to try to nudge them into profitability, or possibly to get the customers to take their business

elsewhere. In contrast, loan applications by customers in the 4 and 5 categories are quickly processed and given special

attention. Banks provide but one example of how firms are making increased use of information technology to do more

sophisticated marginal analysis—devoting their efforts to customers and products where the incremental revenues are

greater than the incremental costs and eliminating and avoiding unprofitable activities.

Source: A. Hughes (2014), “How Banks Use Profitability Analysis,” Database Marketing Institute,


5To keep this example simple, we abstract from several issues. We ignore any pleasure Mary receives from the

process of selling. Also, selling effort today is likely to have some effect on her future professional progress.

Finally, if Mary values a tennis game at 9 A.M. and one at 7 P.M. equally, she will sell during the business day

and postpone tennis to the evening.

18 Part 1 Basic Concepts

$6,000. Since the marginal costs exceed the marginal benefits, Ludger would be bet- ter off rejecting the contract than accepting it. This example illustrates that in calcu- lating marginal costs, it is important to use the opportunity cost of the incremental resources, not their historic (accounting) cost.

Opportunity Costs

Because resources are constrained, individuals face trade-offs. Using limited re- sources for one purpose precludes their use for something else. For example, if Larry Matteson takes four hours to play golf, he cannot use that same four hours to paint his house. The opportunity cost of using a resource for a given purpose is its value in its best alternative use. The opportunity cost of using four hours to play golf is the value of using the four hours in Larry’s next best alternative use.

Marginal analysis frequently involves a careful consideration of the relevant op- portunity costs. If Larry starts a new pizza parlor and hires a manager at $30,000 per year, the $30,000 is an explicit cost (a direct dollar expenditure). Is he better off man- aging the restaurant himself, since he can avoid the explicit cost of $30,000 by not paying himself a salary? The answer to this question depends (at least in part) on the opportunity cost of his time. If he can earn exactly $30,000 in his best alternative job, the implicit cost of self-management is the same as the explicit cost of hiring an out- side manager: He forgoes $30,000 worth of income if he manages the parlor himself. Both explicit and implicit costs are opportunity costs that should be considered in the analysis. Suppose that Larry’s gross profit from the pizza parlor, before paying the manager a salary, is $35,000 and that he can earn $40,000 in an outside job. Hiring a manager for $30,000 yields a net profit of $5,000 from the pizza parlor. He also earns $40,000 from the outside job, for total earnings of $45,000. If he manages the pizza parlor himself, he earns only $35,000. In this example, it is better for him to work at the outside job and hire a manager to run the restaurant.6

Creativity of Individuals7

Within this economic framework, individuals maximize their personal satisfaction given resource constraints. Indeed, people are quite creative and resourceful in min- imizing the effects of constraints. For instance, when the government adopts new


Opportunity Costs and V-8 The Campbell Soup Company used the idea of an opportunity cost to create a successful ad campaign for its V-8

vegetable juice. Upon finishing a soft drink, the fellow in the ad would look into the camera, slap his forehead, and

exclaim: “Wow—I coulda had a V-8.” Since one is unlikely to drink both a soft drink and a V-8, the opportunity cost of

the soft drink is the forgone V-8—a cost that these commercials sought to convince the viewing audience is quite high.

6Again, to keep the example simple, we assume there is no difference in personal satisfaction between Larry’s

outside job and managing the pizza parlor. We also postpone the discussion of consequences for the success

of the pizza parlor from hiring a manager versus self-management until Chapter 10. 7This section draws on W. Meckling (1976), “Values and the Choice of the Model of the Individual in the

Social Sciences,” Schweizerische Zeitschrift für Volkswirtschaft und Statistik 112, 545–560.

Chapter 2 Economists’ View of Behavior 19

taxes, almost immediately accountants and financial planners begin developing clever ways to reduce their impact. Some self-employed individuals were able to re- duce the impact of recent tax increases by changing their status from a proprietorship to a corporation.

As another example, a 33-year-old Brazilian farm hand recently retired with full social security benefits after he satisfied social security auditors that he had been


You own a business that services trucks. A cus-

tomer would like to rent a truck from you for one

week, while you service his truck. You must decide

whether or not to do this.

You have an extra truck that you will not use for

any other purpose during this week. This truck is

leased for a full year from another company for

$300/week plus $.50 for every mile driven. You

also have paid an annual insurance premium, which

costs $50/week to insure the truck. The truck has a

full 100-gallon fuel tank.

The customer has offered you $600 to rent the

truck for a week. This price includes the 100 gal-

lons of fuel that is in the tank. It also includes up to

500 miles of driving. The customer will pay $.50

for each additional mile that he drives above the

500 miles. You anticipate that the customer will

bring back the truck with an empty fuel tank and

will have driven more than 500 miles. You sell fuel

to truckers at a retail price of $4.00/gallon. Any

fuel you sell or use can be replaced at a wholesale

price of $3.25/gallon.

The customer will rent a truck from another

company if you do not accept the proposed deal. In

either case, you will service his truck. You know

the customer and are confident that he will pay all

charges incurred under the agreement.

1. Should you accept or reject the proposed deal?

2. Would your answer change if your fuel sup-

plier limited the amount of fuel that you could

purchase from him at the wholesale price?



Creative Gaming of the System The U.S. Government in 2009 promoted the sales of presidential and Native American $1 coins by offering free

shipping on any order made to the U.S. Mint (which sold the coins at face value). Enthusiasts of frequent-flyer

mileage programs saw a creative way to “game” the government’s offer. Many credit cards are tied to the frequent-

flyer programs of major airlines—for every dollar charged on the card, a mile is credited to the relevant frequent-flyer

program. Several hundred “mile-junkies” responded by purchasing thousands of dollars worth of coins from the U.S.

Mint using their credit cards. Once they received the coins, they deposited the money in their bank accounts to pay

off their credit card charges before any interest costs were incurred. For example, Patricia Hansen, a San Diego retiree

who loves to travel, ordered $10,000 in coins earning 10,000 miles toward free and upgraded travel. Her husband

took the coins to the bank, as soon as they arrived, so that their credit card bill could be paid. The U.S. Mint

eventually figured out what was going on and stopped the program that had resulted in increased costs for them,

credit card companies, banks, and airlines. This example illustrates an important general point. People often respond to economic incentives in creative ways. Managers and government officials need to craft incentives thoughtfully.

Source: S. McCartney (2009), “Miles for Nothing: How the Government Helped Frequent Fliers Make a Mint,” The Wall Street Jour- nal (December 7), A1.

20 Part 1 Basic Concepts

working since he was three years old. Because Brazil doesn’t specify a minimum re- tirement age, the average Brazilian retires at age 49.8

Similarly, when hackers and corporate spies continue to develop more sophisticated schemes to steal information from Web sites or networks, software tools that detect break-ins also have grown in popularity and sophistication. This intrusion-detection soft- ware was about a $100 million industry in 1999 and is now estimated at over $2 billion.9

Understanding this creative nature of individuals has important managerial impli- cations that we discuss later in this chapter, as well as throughout the book.

Graphical Tools Economists often employ a set of graphical tools to illustrate how individuals make choices. These tools distinguish between the preferences (level of satisfaction) that the individual associates with each potential opportunity and the set of feasible op- portunities that an individual faces. We use these tools throughout this book. They also are used in other courses within the typical business school curriculum, such as in finance, human relations, and marketing courses. Our intent is to introduce these tools so that the reader is comfortable using them in basic business applications. We subsequently apply the tools to analyze the problems at Merrill Lynch. Appendix A provides a more detailed development of the economic theory of individual choice (commonly called the “Theory of Consumer Choice”).

Individual Objectives

Goods are things that people value. Goods include standard products like food and clothing, services like haircuts and education, as well as less tangible emotions such as love of family and charity. The economic model of behavior posits that people ac- quire goods that maximize their personal satisfaction, given their resource con- straints (such as a limited income). Economists traditionally use the term utility in re- ferring to personal satisfaction.

To provide a more detailed analysis of how people make choices, economists repre- sent an individual’s preferences by a utility function. This function expresses the rela- tion between total utility and the level of goods consumed. The individual’s objective is to maximize this function, given the resource constraints.10 This concept can be il- lustrated most conveniently through a simple example where an individual cares about only two goods. The insights from this two-good analysis can be extended readily to the case of additional goods such as food, housing, clothing, respect, and charity.

Suppose that Dominique Lalisse values only food and clothing. In general form, his utility function can be written as follows:

� �

Utility � F(Food, Clothing) (2.1)

8P. Fritsch (1999), “In Brazil Retirement Has Become a Benefit Nearly All Can Enjoy,” The Wall Street Journal (September 9), A1.

9J. D’Allegro (1999), “Intrusion Detection Matures,” National Underwriter (March 8), 9; and Frost and

Sullivan (2007), “World Intrusion Detection and Prevention Markets,” 10Clearly, most individuals do not actually consider maximizing a mathematical function when they make

these choices. However, this formulation can provide useful insights into actual behavior to the extent that it

approximates how individuals make choices. Mathematicians have shown that if an individual’s behavior is

consistent with some basic “axioms of choice” (comparability, transitivity, nonsatiation, and willingness to

substitute), the individual will make choices as if he or she were trying to maximize a mathematical function.

Chapter 2 Economists’ View of Behavior 21

Dom prefers more of each good—thus, his utility rises with both food and clothing. In Dom’s case, his specific utility function is

Utility � Food1�2 � Clothing1�2 (2.2)

For instance, if Dom has 16 units of food and 25 units of clothing, his total utility is 20 (that is, utility � 161�2 � 251�2 � 4 � 5 � 20). Dom is better off with 25 units of both food and clothing. Here, his utility is 25 (utility � 251�2 � 251�2 � 5 � 5 � 25).

Utility functions rank alternative bundles of food and clothing in the order of most preferred to least preferred, but they do not indicate how much one bundle is preferred to another. If the utility index is 100 for one combination of food and clothing and 200 for another, Dom will prefer the second combination. The second bundle does not nec- essarily make him twice as well off as the first bundle.11 Neither does this formulation allow one person’s utility of a bundle to be compared to another person’s utility.

Indifference Curves

Preferences implied by the utility function can be illustrated graphically through indifference curves. An indifference curve pictures all combinations of goods that yield the same utility. Given his utility function in Equation (2.2), Dom is indifferent between either 16 units of food and 25 units of clothing or 25 units of food and 16 units of clothing. Both combinations yield 20 units of utility, and hence are on the same indifference curve. Figure 2.1 shows two of Dom’s indifference curves. For example, if given a choice between any two points on curve 1, Dom would say that he does not care which one is selected—in either case, he obtains 8 units of utility.

Figure 2.1 Indifference Curves

These indifference curves picture all combinations of food and clothing that yield the same amount of utility. The specific utility function in this example is U � F1�2 � C1�2, where F is food and C is clothing. Northeast movements are utility-increasing. Indifference curve 2 represents all combinations of food and clothing that yield 20 units of utility, whereas curve 1 pictures all combinations that yield 8 units of utility. Other indifference curves could be drawn for different levels of utility.

4 16 25 C





Quantity of clothing

Q ua

nt ity

o f

fo od

Increasing utility

1: U = 8

2: U = 20

11This is like rankings on a test—an individual who scores in the 80th percentile is not twice as smart as one

from the 40th.

22 Part 1 Basic Concepts

The slope at any point along one of Dom’s indifference curves indicates how much food he would be willing to give up for a small increase in clothing (his utility remains unchanged by this exchange).12 Standard indifference curves that illustrate trade-offs between two goods have negative slopes. If Dom obtains a smaller amount of one good such as clothing, the only way he can be equally as well off is to obtain more of another good like food. If at a point along an indifference curve the slope is 22, Dom is willing to give up 2 units of food to obtain 1 unit of clothing. Alterna- tively he is willing to give up 1�2 unit of clothing to obtain 1 unit of food. This willingness to substitute has important implications, which we discuss later.

Movements up and to the right in graphs like Figure 2.1 are utility-increasing. Holding the amount of food constant, utility increases by increasing clothing (a rightward movement). Holding the amount of clothing constant, utility increases by increasing the amount of food (an upward movement). Thus, in Figure 2.1, Dom would rather be on indifference curve 2 than on 1. He obtains 20 units of utility rather than 8.

Economists typically picture indifference curves as convex to the origin (they bow in, as in Figure 2.1). Convexity implies that if Dom has a relatively large amount of food, he would willingly exchange a relatively large quantity of food for a small amount of additional clothing. Thus, the indifference curves in Figure 2.1 are steep when the level of food is high relative to the level of clothing. In contrast, if he has a relatively large amount of clothing, he would be willing to substitute only a small amount of food for additional clothing. Correspondingly, the indifference curves in Figure 2.1 flatten as Dom has less food and more clothing. The behavior implied by the convexity of indifference curves is consistent with the observed behavior of many individuals—most people purchase balanced combinations of food and clothing.

Opportunities and Constraints

Dom would like more of both food and clothing. Unfortunately, he faces a budget constraint that limits his purchases. Suppose that he has an income of I and the prices per unit of food and clothing are Pf and Pc, respectively. In this single period analysis, we assume that Dom spends all his income on food and clothing. In a multiperiod context, Dom might want to save part of his income or borrow against future income. We examine these possibilities in Appendix B at the end of this chap- ter. Since he cannot spend more than I, his consumption opportunities are limited by the following constraint:

I � Pf F � PcC (2.3)

where F and C represent the units of food and clothing purchased. This budget con- straint indicates that only combinations of food and clothing that cost no more than I are feasible. Rearranging terms, this constraint can be written as

F � I�Pf � (Pc�Pf)C (2.4)

12Recall that the slope of a line is a measure of steepness, defined as the increase or decrease in height per unit

of distance along the horizontal axis. Slopes of curves are found geometrically by drawing a line tangent to

the curve at the point of interest and determining the slope of this tangent line. The slope at a point along one

of Dom’s indifference curves indicates how the quantity of food changes for small changes in the amount of

clothing in order to hold utility constant. Since by definition Dom is indifferent to this exchange (he remains

on the same indifference curve), he is willing to make the exchange.

Chapter 2 Economists’ View of Behavior 23

Figure 2.2 depicts these consumption opportunities—frequently called a budget line. All combinations of food and clothing on or below the line are attainable. Com- binations above the line are infeasible given an income of I. The F intercept (on the vertical axis) of the line I�Pf indicates how much food Dom can purchase if his en- tire income is spent on food (no clothing is purchased). The C intercept is corre- spondingly I�Pc. The slope of the line �Pc�Pf is �1 times the ratio of the two prices. The ratio Pc�Pf is the relative price of clothing in terms of food. It represents how many units of food he must forgo to acquire a unit of clothing: It is the oppor- tunity cost of clothing. For example, if the price of clothing is $8 and the price of food is $2, the relative price of clothing is 4. To keep total expenditures constant, 4 units of food must be given up for every unit of clothing purchased. The relative price of food is Pf�Pc (in this example, 0.25); 1�4 unit of clothing must be given up for each unit of food purchased.

The constraint changes with changes in Dom’s income and the relative prices of the two goods. As shown in Figure 2.3, changes in income result in parallel shifts of the constraint: Its slope is unaffected. An increase in income shifts the constraint out- ward (up and to the right), while a decrease in income shifts the constraint inward. The slope of the constraint changes with the relative prices of the two goods. As shown in Figure 2.4, if the price of clothing increases relative to the price of food, the constraint becomes steeper. If the price of clothing falls relative to the price of food, the constraint becomes flatter.

Individual Choice

Within this economic framework, Dom’s goal is to maximize utility given his op- portunities. Utility is maximized at the point of tangency between the constraint and

Figure 2.2 Opportunities and Constraints

The constraint reflects the feasible combinations of food and clothing that are attainable given the person’s income (I ). The vertical and horizontal intercepts, respectively, show the amounts of food and clothing that can be purchased if no income is spent on the other good. The slope of the constraint is equal to �1 times the ratio of the prices of the two goods. For instance, if the price of clothing is $8 and the price of food is $2, the slope will be �4. This slope implies that 4 units of food must be given up for 1 unit of clothing. If both goods have the same price, the slope will be �1.

I Pf

I Pc





Combinations above line are unaffordable

Combinations on/below line are affordable

Q ua

nt ity

o f f

oo d

Quantity of clothing

24 Part 1 Basic Concepts

an indifference curve.13 Figure 2.5 portrays the optimal choice. Dom could choose points like b and c on indifference curve 1. However, point a on curve 2 yields greater satisfaction (utility) and thus is preferred. Dom would prefer to be at any point on curve 3. Yet, these points are unattainable given his income.

Figure 2.3 Income Changes

This figure shows that there is a parallel shift in the budget constraint when income changes. The slope of the constraint does not change because there is no change in the prices of the two goods. The slope is �1 times the ratio of the prices.

I Pc

I Pf



Lower income

Original consumption opportunities

Higher income

Quantity of clothing

Q ua

nt ity

o f f

oo d





Figure 2.4 Price Changes

This figure shows how his consumption opportunities change with changes in the price of clothing. The slope of the line is �(Pc�Pf). Thus, an increase in the price of clothing (from Pc to Pc

HI) produces a steeper line, while a decrease (from Pc to Pc

LO) produces a flatter line. Changes in the price of food also affect the slope of the line.


I Pf


Increase in the price of clothing

Original consumption opportunities

Decrease in the price of clothing

Quantity of clothing

Q ua

nt ity

o f f

oo d

I Pc

I Pc



13For simplicity, we ignore the possibility of corner solutions—the points where the budget constraint

intersects the axes. With corner solutions, the individual spends all income on only one good.

Chapter 2 Economists’ View of Behavior 25

This graphical solution to Dom’s choice problem has a simple intuitive interpre- tation. At the point of tangency, the indifference curve and the constraint have equal slopes. Recall that the slope of the indifference curve represents Dom’s willingness to trade food for clothing, whereas the slope of the constraint represents the terms of trade available in the marketplace. At the optimal choice, the willingness and ability to trade are equal. At other feasible combinations of food and clothing, Dom’s util- ity could be increased by making substitutions. For instance, if Dom were at a point where he was willing to trade 5 units of food for 1 unit of clothing and if the relative price of clothing were 4 units (the slope of the indifference curve is steeper than the constraint), Dom would be better off purchasing less food and more clothing. (He is willing to trade 5 units of food for 1 unit of clothing, but only must forgo 4 units of food to obtain 1 unit of clothing in the marketplace.) Alternatively, if Dom were at a point where he was only willing to forgo 1 unit of food for 1 unit of clothing (the slope of the indifference curve is flatter than the constraint), he would be better off purchasing more food and less clothing—since he receives 5 units of food for each unit of clothing forgone.

Earlier in this chapter, we discussed how marginal analysis is the cornerstone of modern economics. It is important to understand that the graphical tools presented in this section depict marginal analysis. In marginal analysis, individuals take actions as long as their incremental benefits are greater than their incremental costs. Our graphical analysis of individual choice corresponds to this decision rule. The relative price ratio, PC�PF, is the marginal cost of a unit of clothing, expressed as units of food—the units of food that are forgone is the opportunity cost of an additional unit of clothing. Similarly, the opportunity cost of an additional unit of food is PF�PC

units of clothing. The slope of the indifference curve reflects the marginal benefit of an additional unit of clothing expressed as units of food. For example, if Dom is will- ing to trade 5 units of food for 1 unit of clothing (slope of the indifference curve � �5), his marginal benefit of one additional unit of clothing must equal the utility

Figure 2.5 Optimal Choice

The individual is best off by choosing point a where the constraint is tangent to indifference curve 2. This optimal combination of food and clothing, F* and C*, yields higher level of satisfaction (utility) than other feasible alternatives (e.g., points b and c). The individual would prefer points on indifference curve 3, but these points are infeasible given his consumption opportunities.

C *


F *



c 1



Quantity of clothing

Q ua

nt ity

o f f

oo d


26 Part 1 Basic Concepts

from 5 units of food. Similarly, his marginal benefit of a unit of food is equivalent to .2 units (1�5) of clothing. If Dom is not at the point of tangency between the indif- ference curve and the budget line, the marginal benefit of trading one good for the other must be greater than the marginal cost. Suppose Dom is willing to trade 5 units of food for 1 unit of clothing, but only has to trade 2 units of food for 1 unit of cloth- ing in the marketplace. In this case, Dom should trade food for clothing since the marginal (incremental) benefit is greater than the marginal (incremental) cost. At the optimum (point of tangency) the marginal benefit of consuming 1 more unit of either good is equal to the marginal cost and there is no reason to make additional trades.

Changes in Choice

Dom’s consumption opportunities will change whenever prices or income change. Consequently, he will make different choices. Recall that changes in relative prices alter the slope of the constraint. When the relative price of a good increases, individ- uals typically choose less of that good.14 Figure 2.6 shows how Dom will purchase less food as its relative price increases—food is more expensive and so less attractive than it was at a lower price. Generally, the amount of clothing purchased can go either up or down; it depends on the location of the new tangency point. (Given the particular utility function assumed in this example, the amount of clothing purchased remains unchanged.) Even though the price of clothing is relatively more attractive, the increase in food prices can limit available income so as to reduce the amount purchased of both goods. Changes in Dom’s income cause parallel shifts in the

Figure 2.6 Optimal Choice and Price Changes

This figure shows how the optimal choice changes with an increase in the price of food. In this example, the individual chooses less food (F1* rather than F0*). This is the typical case—usually, an individual will purchase less of a good when its price increases. Due to the particular utility function used in this example, the amount of clothing purchased remains unchanged (C*). More generally, the amount of clothing purchased can go either up or down. It depends on the location of the new tangency point.

C * C



Consumption opportunities after increase in price of food

Original consumption opportunities

Quantity of clothing

Q ua

nt ity

o f f

oo d



14Although in principle some individuals might purchase more of a good if its price increases, this outcome is

rarely observed.

Chapter 2 Economists’ View of Behavior 27

constraint and will change his optimal choice. In Chapter 4, we examine in more detail how changes in income and prices affect consumption choices. Appendix A contains a more detailed analysis of the effects of price changes on individual choice and illustrates how this basic graphical analysis can be used to study inter-temporal choices, such as the choice between current consumption and savings.

Choices also change if preferences change. Now changes in preferences undoubt- edly occur. (Do you really believe that Toys ‘R’ Us will have any difficulty satisfying the demands for toys that were highly popular in past years, such as Teenage Mutant Ninja Turtle action figures, Tomaguchi virtual pets, Tickle-Me-Elmo dolls, or Poke- mon Cards next Christmas?) Yet, economists rarely focus on such explanations. Eco- nomics has little theory to explain what might cause preferences to change. And since a large premium is placed on operationalism in managerial economics, preference- based explanations generally are appealed to only after other potential explanations are exhausted. In a sense, these preference-based explanations are too easy—they work too well. Virtually any observed behavior could be explained by appealing to prefer- ences: Why did the consumption of frozen yogurt increase relative to that of ice cream? People’s preferences changed so that more frozen yogurt and less ice cream was demanded. But an observed reduction in consumption could have been “explained” just as readily. Without a deeper understanding of why preferences change, one is left “explaining” everything but with an analysis that allows you to predict nothing.

Ultimately, the managerial usefulness of this analysis comes from its power to identify policy instruments that have a predictable impact on the problems at hand. Across a broad array of problems, assuming that underlying preferences are reasonably stable and analyzing the impact of changes in opportunities and constraints regularly will yield important managerial insights and identify productive managerial tools.

ANALYZING MANAGERIAL DECISIONS: Consumer Choice and Graphical Tools

You are a manager for a company that bottles and

sells wine in two different countries. You charge

the same price for a bottle of wine in both coun-

tries. Yet, your wine sales are much higher in one

country than the other. Your boss asks you to de-

velop an explanation for the differences in wine

sales between the two countries and to develop a

plan to sell more wine in the country with low wine


Population sizes and family incomes in the two

countries are very similar. You also know that each

country imposes a per bottle tax on wine.

Begin by providing a plausible economic expla-

nation (focusing on constraints) for the differences

in wine sales in the two countries. Illustrate your

explanation by using indifference curves and bud-

get lines for representative consumers from the two

countries. What data would you want to determine

if your explanation is likely to be correct? Are there

other plausible explanations for the differences in

wine consumption? Are there ways to determine

which of these explanations is most likely to be dri-

ving the differences in consumption?

1. Suppose that your economic explanation is

likely to be correct and that your company will

not allow you to lower the price per liter that

you charge for wine in the two countries.

Discuss at least two potential actions that you

might take to sell more wine in the country

with low demand.

2. Now provide a potential preference-based ex-

planation for the differences in wine sales.

Suppose that this explanation is correct. Dis-

cuss whether there are likely to be feasible

policies that you could use to increase wine

sales in the country with the low demand.

28 Part 1 Basic Concepts

Motivating Honesty at Merrill Lynch Often, economists focus on consumption goods such as food and clothing. This focus is natural given the interests economists have in understanding consumer behavior. Yet this analysis can be extended easily to consider other goods that people care about, such as love and respect.15 Such an extension can be used to analyze the problem at Merrill Lynch.

Suppose that Susan Chen, like other analysts at Merrill Lynch, values two goods—money and integrity. Her utility function is

� � Utility � F (Money, Integrity) (2.5)

Money is meant to symbolize general purchasing power; it allows the purchase of goods such as food, clothing, and housing. Integrity is something Sue values for its own sake—being honest in her dealings with other people provides Sue with satis- faction and she values it for that reason.

Suppose that integrity can be measured on a numerical scale with Sue preferring higher values. For example, 5 units of integrity provide more utility than 4 units of integrity. (In actuality, measuring a good like integrity on a numerical scale might be quite difficult. Yet this complication does not limit the qualitative insights that we can derive from the analysis.)

Merrill paid its stock analysts an annual bonus that was based partly on the ana- lyst’s contribution to the investment banking side of the business (e.g., the firm’s un- derwriting activities). If Sue were completely honest and rated a company as a poor investment, the management of that company might take its investment banking business to another firm. The resulting loss in Merrill’s investment banking revenue would reduce Sue’s annual bonus. This bonus scheme thus confronts Sue with a trade-off. She can be honest and derive satisfaction from maintaining her integrity, or she can be dishonest in her rating of the stock and obtain a higher bonus. (She also might consider the future effects on her income from developing a good or bad rep- utation as an investment analyst. However, the analysis in this chapter is framed in a simple one-period context and does not consider monetary returns from developing a good reputation. In subsequent chapters we extend the analysis and consider such multiperiod effects.)

Figure 2.7 depicts Sue’s implied opportunities. This constraint shows the maxi- mum combinations of income and integrity that are feasible given the compensa- tion plan and conditions at the company.16 If Sue sacrifices all integrity, she earns $max a year. If she is scrupulously honest in her investment recommendations, she earns less (there is a positive floor on her income, $min, since her base salary does depend on the amount of investment banking business and her analysis undoubt- edly will suggest recommending some of Merrill’s clients’ stocks). Intermediate options along the constraint are possible. While Sue would like to earn more than $max, higher income is not feasible in this job.

15G. Becker (1993), “Nobel Lecture: The Economic Way of Looking at Behavior,” Journal of Political Economy 101, 385–409.

16For simplicity, we draw the constraint as linear. Linearity is not necessary for our analysis. Also, we want to

emphasize that we put dollars on the vertical axis only because it is a convenient general indication of value,

not because money is more important than other things. We could illustrate Sue’s willingness to trade

integrity against anything else Sue values, such as Big Macs, pianos, or pairs of jeans.

Chapter 2 Economists’ View of Behavior 29

Sue chooses the combination of integrity and income that places her on the high- est attainable indifference curve. This choice occurs at the point of tangency between her indifference curve and the constraint. Sue ends up selecting relatively low amounts of integrity because the bonus plan adopted by Merrill’s management has made integrity expensive. If Sue were to choose more integrity, she would forfeit a relatively large amount of income.

Management can alter the opportunities Sue and her colleagues face by chang- ing its compensation plan. In the Merrill case, reducing the emphasis of investment banking revenue in determining the annual bonus reduces the gains from dishonest advice and thus flattens the constraint. Changes in the slope of the constraint result in a different tangency point and hence a different choice. Figure 2.8 shows how Sue’s optimal choice changes when the emphasis on investment banking revenue is decreased.17 The result is more honest behavior. In essence, Sue “purchases” more integrity because it now is less expensive. Consistent with this analysis, Merrill, in its settlement with the State of New York, agreed to change the way it evaluated and compensated its analysts. Bonuses now are based on the quality of investment advice—not tied to its investment banking business.

Figure 2.7 Nature of Opportunities Facing an Analyst at Merrill Lynch

The constraint depicts the maximum amounts of money and integrity that are possible for the analyst given the bonus plan and conditions at the company. If the analyst sacrifices all integrity and recommends stocks even if they are poor investments, the employee earns a maximum of $max a year. Investment banking business is lost if the analyst gives objective advice and rates certain stocks as poor investments (selects a higher level of integrity). Income is lower since the analyst is paid a bonus based on investment banking revenues. Ic represents complete honesty. Ic





Quantity of integrity

In co

m e

(in d

ol la

rs )

17We have altered the compensation scheme in a manner that places Sue on the same indifference curve. Our

rationale for doing this is as follows: Merrill Lynch must provide Sue with sufficient job satisfaction

(utility) to retain her at the firm. Below this level of utility, Sue will quit and work elsewhere. Merrill Lynch

is unlikely to want to pay Sue more than this minimum utility because it reduces firm profits. Thus, Merrill

Lynch has an incentive to adjust compensation in a manner that keeps her on the same indifference curve.

Sue’s indifference curve in Figure 2.8 can be viewed as this “reservation” utility. These issues are covered

in more detail in Chapter 14.

30 Part 1 Basic Concepts

Managerial Implications This analysis illustrates how the economic framework can be used to analyze and ad- dress management problems. Managers are interested in affecting the behavior of in- dividuals such as employees, customers, union leaders, or subcontractors. Under- standing what motivates individuals is critical. The economic approach views individual actions as the outcomes of maximizing personal utility. People are willing to make substitutions (e.g., less leisure time for more income) so long as the terms of trade are advantageous. Managers can affect behavior by appropriately designing the opportunities facing individuals. The design of the opportunities affects the trade- offs that individuals face and hence their choices. For example, management can mo- tivate employees through the structure of compensation plans or customers through pricing decisions.

The outcome of individuals making economic choices is a function of both op- portunities and preferences. Individuals try to achieve their highest level of satisfac- tion given the constraints they face. Our discussion of management implications, however, intentionally focuses on opportunities and constraints, not preferences. As a management tool, the usefulness of focusing on personal preferences often is lim- ited. It is difficult to change what a person likes or does not like. Moreover, prefer- ences rarely are observable, and (as we noted earlier) virtually any observed change in choice can be “explained” as simply a matter of a change in personal tastes. For instance, a preference-based explanation as to why employees were dishonest at Merrill Lynch is that these employees gained personal satisfaction from being dishonest (or compared to employees at other firms, Merrill Lynch employees were willing to trade large amounts of personal integrity for small financial rewards). This

Figure 2.8 Optimal Choices of an Analyst at Merrill Lynch under Two Different Compensation Plans

Case 1 reflects the original compensation plan. In this case, compensation includes a high bonus based on investment banking revenues and the constraint is relatively steep. In Case 2, the firm reduces the emphasis on investment banking revenues in compensating analysts. The slope of the constraint is flatter. The result is that the individual chooses a higher level of integrity in Case 2 than in Case 1.



I1* I2*

Case 2

Case 1 I

In co

m e

(in d

ol la

rs )

Quantity of integrity


Chapter 2 Economists’ View of Behavior 31

explanation is not very helpful in giving management guidance on how to address the problem. It suggests that Merrill Lynch might try to fire dishonest employees and replace them with employees who care more about personal integrity. But the diffi- culty of observing personal preferences limits the feasibility of this approach. It would be difficult for Merrill Lynch to know if, as a group, its new hires would be any less dishonest than the old employees. You cannot just ask applicants if they are honest—if they are not, they will have no qualms about claiming that they are.

The fact that individuals are clever and creative in limiting the effects of con- straints greatly complicates management problems. Changing incentives will affect employee behavior, though sometimes in a perverse and unintended manner. Con- sider two of the Soviet Union’s early attempts to adopt incentive compensation to motivate employees. To discourage taxi drivers from simply parking their cabs, they were rewarded for total miles traveled; to encourage additional production, chande- lier manufacturers were rewarded on total volume of production—measured in kilo- grams. In response to these incentive plans, Moscow taxi drivers began driving empty cabs at high speeds on highways outside the city and chandelier manufactur- ers started producing such massive fixtures that they literally would collapse ceil- ings. (It is less costly to make one 100-kilo chandelier than five 20-kilo chandeliers; manufacturers also substituted lead for lighter-weight inputs.) Merrill Lynch initially adopted bonuses to motivate analysts to work harder and cooperate across business units. The deceitful behavior was a side effect that potentially was unanticipated when the plan was adopted.

In summary, the economic approach to behavior has important managerial impli- cations. The framework suggests that a manager can motivate desired actions by es- tablishing appropriate incentives. However, managers must be careful because set- ting improper incentives can motivate perverse behavior.

It is worth noting that economic analysis is limited in its ability to forecast the pre- cise choices of a given individual because individual preferences are largely unob- servable. The focus is on aggregate behavior or on what the typical person tends to do. For example, an economist might not be very good at predicting the responses of individual employees to a new incentive plan. An economist will be successful in


More to Life than Money Should you hire a housekeeper or clean your house yourself? Should you mow your own lawn or hire the kid next door?

People confront these types of questions everyday. Spending money on housekeepers, gardeners, etc. provides extra time

to use for other activities, but it comes at the cost of having less money to purchase other things. People vary in their

attitudes when it comes to time versus money. In a recent survey conducted by LearnVest and Chase Blueprint,

54 percent said they would like more money rather than more time, while 46 percent they want more free time relative to

more money. It is not uncommon to see people in well paid positions retiring early to spend more time with family or to

pursue other personal interests. For example, Kelly Malson was the CFO of World Acceptance Corp. and had an annual

pay package valued at $4.8 million in 2013. During the summer of 2013, one of her college friends died of Lou Gehrig’s

disease at 46. Kelly, who was only 43, decided to quit her job to pursue her dreams while she still could, which included

traveling to all 50 U.S. states and pursuing a graduate degree. Kelly, as well as most other people, clearly value things

other than just money and regularly make choices that trade off money for other things that they value.

Source: C. Oakley (2013), “Would You Rather Have More Time or More Money?” Forbes (12/20) and M. Murphy (2013), “More

CFOs Choose to Retire Early,” The Wall Street Journal (December 10).

32 Part 1 Basic Concepts

predicting that the typical employee will work harder—and thus output for the group will rise—when compensation is tied to output, than when a fixed salary indepen- dent of performance is paid. Managers typically are interested in structuring an or- ganizational architecture that will work well and does not depend on specific people filling particular jobs. Individuals come and go, and the manager wants an organiza- tion that will work well as these changes occur. In this context, the economic frame- work is likely to be useful. To solve management problems where the characteristics of a specific individual are more important, other frameworks may be more valuable. For example, if the board is interviewing a potential new CEO, insights into that in- dividual’s behavior derived from psychology might be extremely useful.

Alternative Models of Behavior18

We have shown how the economic view of behavior can be used in managerial deci- sion making. We now discuss four other models that are commonly used by man- agers (either explicitly or implicitly) to explain behavior. Our discussion of each of these models is simplified. The intent, however, is to capture the essence of a few of the more prominent views that managers have about behavior and to illustrate how managerial decision making is affected by the particular view. We contrast these


Medicare Creates Perverse Incentives for Doctors Doctors do not care about money but are motivated by concerns about providing the best care for patients—right?

Apparently many doctors and the major drug company employees do not think so. Perverse incentives among

physicians arguably have contributed to the problem of spiraling health care costs in the United States.

For years, Medicare (federal health program for the elderly) reimbursement policies allowed individual doctors to

make hundreds of thousands of dollars a year in extra profits from the drugs they administered to patients in their

offices (the doctors would buy the drugs themselves and bill Medicare, rather than having the patients get them directly

from pharmacies). For example, many cancer doctors earned over $1 million per year on drug sales alone. Because the

profits on different drugs varied enormously, doctors had incentives to prescribe medications with the highest profit

margins. Some physicians have acknowledged that they performed treatments that got them the best reimbursements,

“whether or not the treatments benefited patients.”

Drug companies were well aware of the Medicare policies and calculated the profits that doctors received from

prescribing specific drugs “down to the penny.” For example, in 1998 Schering-Plough told its sales representatives

that its drug for the treatment of bladder cancer could produce a profit for a physician of $2,373.84 per patient. Sales

representatives in turn made sure that doctors were well aware when their drugs were in the high-profit category. For

instance, a sales representative for AstraZeneca wrote in a letter to Arizona urologists, “DO THE MATH.”

Medicare changed its reimbursement policies in 2005 and reduced the profits that physicians could make on drug

sales. At least some physicians have responded by shifting from drug intensive treatments to other treatments that have

higher profit margins. To quote one doctor, “People go where the money is, and you’d like to believe it’s different in

medicine, but it’s really no different . . . as long as oncologists continue to be paid by the procedure instead of spending

time with patients, they will find ways to game the system.”

Source: A. Berensen (2007), “Incentives to Limit Any Savings in Treating Cancer,” (June 12).

18This section draws on W. Meckling (1976).

Chapter 2 Economists’ View of Behavior 33

alternative views with the economic view and argue why the economic framework is a particularly useful tool for managers.

Only-Money-Matters Model

Some people believe that the only important component of the job is the level of monetary compensation. But as we have already suggested, people have an incredi- bly broad range of interests, extending substantially beyond money. And these inter- ests are reflected in a diverse array of activities. As examples, much of the work through the Red Cross is undertaken by unpaid volunteers; people frequently choose early retirement, forgoing a regular paycheck to enjoy additional leisure time; riskier occupations command higher pay in order to attract people into those jobs.

Some of this confusion can result from a misinterpretation of standard economic analysis. Central to economics is the study of trade-offs (recall our discussion of in- difference curves illustrating trade-offs between food and clothing). Economists fre- quently use money as one of the goods being considered. But in these cases, money is merely a convenient unit of value: It simply represents general purchasing power. Its use does not suggest that only money matters.

Happy-Is-Productive Model

Managers sometimes assert that happy employees are more productive than un- happy employees. Managers following this happy-is-productive model see as their goal the designing of work environments that satisfy employees. Psycholog- ical theories, such as Maslow’s and Herzberg’s, are frequently used as guides in ef- forts to increase job satisfaction.19


Happy Is Productive versus Economic Incentives—The Affordable Care Act The Affordable Care Act, better known, as “ObamaCare,” has generated significant controversy since it was signed into

law in 2010. The Congressional Budget Office (CBO) conducts economic analysis for Congress and is “widely revered

by both Democrats and Republicans alike as the gold standard for economic analysis.” In 2014, the CBO reported that

it expected the equivalent of 2.5 million Americans, who were otherwise willing and able to work before ObamaCare,

to work less or not at all as a result of the law by 2024. The projection contrasts dramatically with earlier forecasts

made by proponents of ObamaCare, who projected a resulting increase in employment. These positive forecasts were

based in part on an implicit assumption of the Happy is Productive Model of Behavior. The logic is as follows. Most

people want to work because being productive leads to greater happiness. Bad health, however, can prevent people

from working. ObamaCare by promoting a healthier population will increase the number of employed people. This

argument, however, ignores economic incentives. ObamaCare imposes a stiff economic cost on unemployed people

who take jobs—they lose all or part of their government subsidies for purchasing health insurance. Economist Casey

Mulligan sums up the economic incentives created by ObamaCare succinctly, “when you pay people for being low

income, you are going to have more low-income people.” While this is but one effect in evaluating the merits of

ObamaCare, it would appear to be a potentially important one.

Source: J. Rago (2014), “The Weekend Interview with Casey Mulligan,” The Wall Street Journal (February 8), A15.

19F. Herzberg, B. Mausner, and B. Snyderman (1959), The Motivation to Work (John Wiley & Sons: New

York); and A. Maslow (1970), Motivation and Personality (Harper & Row: New York).

34 Part 1 Basic Concepts

A manager adhering to the happy-is-productive model might suggest that the problem at Merrill Lynch was motivated by disgruntled employees who took out their frustrations on customers. This view implies that Merrill Lynch could reduce the problem by promoting employee satisfaction through such actions as designing more interesting jobs, increasing the rates of pay, and improving the work environ- ment. Happier employees would be expected to provide customers with better in- vestment advice.

The economic and happy-is-productive models do not differ based on what peo- ple care about. The economic model allows individuals to value love, esteem, inter- esting work, and pleasant work environments, as well as more standard economic goods such as food, clothing, and shelter. The primary difference in the models is what motivates individual actions. In the happy-is-productive model, employees exert high effort when they are happy. In the economic model, employees exert effort because of the rewards.

To contrast the two models, consider offering an employee guaranteed lifetime employment plus a large salary, which will be paid independent of performance. The happy-is-productive model suggests that the employee will be more productive, because the high salary and job security are likely to increase job satisfaction. The economic model suggests that the employee would exert less effort—since the em- ployee receives no additional rewards for working harder and will not be fired for ex- erting low effort.

Good-Citizen Model

Some managers subscribe to the good-citizen model. The basic assumption is that employees have a strong personal desire to do a good job; they take pride in their work and want to excel. Under this view, managers have three primary roles. First, they need to communicate the goals and objectives of the organization to employees. Second, they must help employees discover how to achieve these goals and objec- tives. Finally, managers should provide feedback on performance so that employees can continue to improve their efforts. There is no reason to have incentive pay, since individuals are interested intrinsically in doing a good job.


Culture and Behavior In Tokyo, lost cell phones, umbrellas, and cash regularly find their way to the Tokyo Metropolitan Police Lost and

Found Center—the Japanese are scrupulous about turning in found articles. In 2002 the center handled $23 million in

cash and 330,000 umbrellas. Scrupulous behavior of this type is far less common in New York City and many other

cities around the world. What accounts for this “cultural difference?” The historic and current behavior of the Japanese

can be explained at least in part by economic incentives.

The system traces its roots to a code written in 718. Lost goods, animals, and servants had to be handed over to a

government official within five days of being found. Not handing over found objects was severely punished. In 1733

two officials who kept a parcel of clothing were led around town and executed. Current law gives the finder seven days

to turn in found goods. If the item is reclaimed, the finder is entitled to a reward (5 to 20 percent). If the item is not

reclaimed within six months, the finder can claim it.

The most commonly reclaimed item is a cell phone—about 75 percent are returned. The least reclaimed are

umbrellas at 0.3 percent.

Source: N. Onishi (2004), “Never Lost, but Found Daily,” New York Times (January 8), A1.

Chapter 2 Economists’ View of Behavior 35

This view suggests that the problems at Merrill Lynch occurred because employ- ees misunderstood what was good for the company. Employees might have thought that increasing investment banking revenues was in the company’s best interests, even if it required a certain amount of dishonesty. Under the good-citizen view, the management of Merrill Lynch could motivate employee honesty by clearly commu- nicating to its analysts that Merrill Lynch would be better off in the long run if they did not deceive their customers. Managers might be instructed to hold a series of an- alyst meetings to stress the value of honesty and objective investment advice.

In the good-citizen model, employees place the interests of the company first. There is never a conflict between an employee’s personal interest and the interest of the company. In contrast, the economic model posits that employees maximize their own utility. Potential conflicts of interest often arise. The economic view predicts that pleas from Merrill Lynch management that analysts be more honest would have little effect on behavior unless they also changed the reward system to make it in the interests of analysts to be more honest.

Product-of-the-Environment Model

The product-of-the-environment model argues that the behaviors of individuals are largely determined by their upbringings. Some cultures and households encourage positive values in individuals, such as industry and integrity, whereas others promote negative traits, such as laziness and dishonesty. This model suggests that Merrill Lynch had dishonest analysts. A response would have been to fire these employees and replace them with honest analysts from better backgrounds.

Which Model Should Managers Use? Behavior is a complex topic. No behavioral model is likely to be useful in all con- texts. For example, the economic model is unlikely to be helpful in predicting whether a given individual will prefer a red shirt to a blue shirt (selling at the same price). But our focus is on managerial decision making. In this context, there are rea- sons to believe that the economic model is particularly useful.

Managers are frequently interested in fostering changes in behavior. For example, managers want consumers to buy more of their products, employees to exert more effort, and labor unions to accept smaller wage increases. In contrast to other mod- els, the economic framework provides managers with concrete guidance on how to alter behavior. Desired behavior can be encouraged by changing the feasible oppor- tunities facing the decision maker. For example, incentive compensation can be used to motivate employees, and price changes can be used to affect consumer behavior.

There is ample evidence to support the hypothesis that this economic framework is useful in explaining changes in behavior. The most common example is that con- sumers tend to buy less of a product at higher prices. The evidence suggests that the model is also useful in explaining aspects of behavior in many other contexts, in- cluding voting; the formation, dissolution, and structure of families; drug addiction; and the incidence of crime.20

20G. Becker (1993).

36 Part 1 Basic Concepts

The good-citizen model appears less successful in predicting behavior in busi- ness settings. Management would be an easy task if employees would work harder and produce higher-quality products simply on request. The happy-is-productive model also has material limitations. Most importantly, the existing evidence sug- gests that there is little relation between job satisfaction and performance (see Scott’s “Criticisms of the Happy-Is-Productive Model” in the accompanying box). Happy employees are not necessarily more productive. Sometimes, managers might want to follow the implications of the product-of-the-environment model and fire employees with undesirable traits. Yet, this approach is of limited use in solving most managerial problems. Also, given laws that limit discrimination, this approach can subject the firm to potentially serious legal sanctions.

The Economic Framework and Criminal Behavior Criminals often are viewed as psychologically disturbed. Evidence, however, suggests that criminal behavior can be

explained, at least in part, by the economic framework. This framework predicts that a criminal will consider the

marginal costs and benefits of a crime and will commit the crime only when the benefits exceed the costs. Under this

view, increasing the likelihood of detection and/or the severity of punishment will reduce crimes. In a pioneering study,

Issac Ehrlich examined whether the incidence of major felonies varied across states with the expected punishment. He

found that the incidence of robberies decreased about 1.3 percent in response to each 1 percent increase in the

proportionate likelihood of punishment. The incidence of crime also decreased with the severity of the punishment.

Since Ehrlich’s study, scholars have conducted extensive research on this topic. In general, the results support the

conclusion that the economic model plays a useful role in explaining criminal activity.

Source: I. Ehrlich (1973), “Participation in Illegitimate Activities: A Theoretical and Empirical Investigation,” Journal of Political Economy 81, 521–565.


Criticisms of the Happy-Is-Productive Model W. Richard Scott summarizes some of the major concerns about the happy-is-productive model (sometimes referred to

as the human-relations movement):

Virtually all of these applications of the human-relations movement have come under severe criticism on both ideological and empirical grounds. Paradoxically, the human-relations movement, ostensibly developed to humanize the cold and calculating rationality of the factory and shop, rapidly came under attack on the grounds that it represented simply a more subtle and refined form of exploitation. Critics charged that workers’ legitimate economic interests were being inappropriately deemphasized; actual conflicts of interest were denied and “therapeutically” managed; and the roles attributed to managers represented a new brand of elitism. The entire movement was branded as “cow sociology” just as contented cows were alleged to produce more milk, satisfied workers were expected to produce more output.

The ideological criticisms were the first to erupt, but reservations raised by researchers on the basis of empirical

evidence may in the long run prove to be more devastating. Several decades of research have documented no clear

relation between worker satisfaction and productivity.

Source: W. R. Scott (1981), Organizations: Rational, Natural and Open Systems (Prentice Hall: Englewood Cliffs), 89–90.


Chapter 2 Economists’ View of Behavior 37

Behavioral Economics The traditional economic model of behavior assumes that individuals are rational, in that they employ marginal analysis in decision making—they balance the incremen- tal costs and benefits to arrive at choices that maximize their personal happiness. One, however, does not have to look far to observe people making choices that by this def- inition appear economically irrational. For example, how many of us have eaten all the food served on a plate, even when we are not hungry and want to lose weight? Do we have some vague notion that we are not wasting food by eating more than we re- ally need or want? How many of us might volunteer to help a distressed person change a flat tire for free, but feel insulted if they offered to pay us for the work?

Simple behavioral observations, as well as more formal evidence from experiments, suggest that people do not always behave rationally. A somewhat new field called Behavioral Economics has emerged that covers a wide range of attempts to extend the standard economics framework to account for potentially relevant features of human behavior that are not included in standard analysis21. While


Interwest Healthcare is a nonprofit organization

that owns 10 hospitals located in three western

states. Cynthia Manzoni is Interwest’s CEO. Vijay

Singh, Interwest’s CFO, and the administrators of

the 10 hospitals report to Manzoni.

Singh is deeply concerned because the hospital

staffs are not being careful when entering data into

the firm’s management information system. This

data involves information on patient intake, treat-

ment, and release. The information system is used

to compile management reports such as those relat-

ing to the costs of various treatments. Also, the sys-

tem is used to compile reports that are required by

the federal government under various grant pro-

grams. Singh reasons that without good informa-

tion, the management and government reports are

less useful and potentially misleading. Moreover,

the federal government periodically audits Inter-

west and might discontinue aid if the reports are

deemed inaccurate. Thus, Singh is worried about

the managerial implications and the potential loss

of federal funds.

Singh has convinced Manzoni that a problem

exists. She also realizes the importance of an accu-

rate system for both management planning and

maintaining federal aid. Six months ago, she invited

the hospital administrators and staff members from

the corporate financial office to a retreat at a resort.

The purpose was to communicate to the hospital ad-

ministrators the problems with the data entry and to

stress the importance of doing a better job. The

meeting was acrimonious. The hospital people ac-

cused Singh of being a bureaucrat who did not care

about patient services. Singh accused the hospital

staffs of not understanding the importance of accu-

rate reporting. By the end of the meeting, Manzoni

thought that she had a commitment by the hospital

administrators to increase the accuracy of data entry

at their hospitals. However, six months later, Singh

claims that the problem is as bad as ever.

Manzoni has hired you as a consultant to ana-

lyze the problem and to make recommendations

that might improve the situation.

1. What are the potential sources of the problem?

2. What information would you want to analyze?

3. What actions might you recommend to in-

crease the accuracy of the data entry?

4. How does your view of behavior affect how

you might address this consulting assignment?

21See Behavioral Economics and Its Application (2007), edited by P. Diamond and H. Vartiainen (Princeton

University Press, Princeton and Oxford) for a collection of papers that summarize behavioral research on a

variety of economic topics.

38 Part 1 Basic Concepts

behavioral economists do not argue that economic incentives are unimportant, they tend to focus more on the cognitive, emotional, and social factors that affect individ- uals in making economic decisions. The ultimate objective of this research is to provide a better understanding of both aggregate market and individual economic behavior than we can get from the standard analysis alone.

The assumptions used in any type of scientific modeling, however, are never en- tirely realistic. Indeed, the purpose of a model is to simplify the analysis to make it tractable. For instance, a physical scientist might assume “unrealistically” that there is a perfect vacuum to derive a scientific formula. The test of her model, however, should not be based on the realism of her assumptions, but on how well the model helps us to predict or understand real-world phenomena. By this criterion, standard economic analysis has stood the test of time. For many decades, it has proven useful in helping managers to understand their competitive environments better and for im- proving their decision making and performance. This book presents the standard economic framework and focuses on the many managerial insights that it provides. A detailed study of Behavioral Economics, while potentially interesting, is beyond the scope of this book.

The need for more “realistic” assumptions in a model depends in part on the pur- pose of the model. For example, suppose that your boss has to go to Chicago for a business meeting and wants you to provide directions on how to drive to the specific location. You have at least two choices: (1) you might download a satellite photo- graph of the Chicago area or (2) you might get a map. Although the satellite photo would certainly include more realism, for this purpose, the map would be more useful. Most of the additional realism in the photograph is extraneous for your par- ticular purpose. The map abstracts from these unimportant aspects of the area to highlight those facets that are important. It is precisely this sense in which within this book, we employ models that abstract from an array of aspects of reality to high- light those things that are important for the particular issue at hand. On a few occa- sions, particularly in the chapters on organizational architecture, we analyze issues for which behavioral economics provides particularly relevant insights for managers. In these cases, we present both the standard economic analysis of the problem, as well as discussion of the implications from the behavior research.

Decision Making under Uncertainty Throughout this chapter, we have considered cases where the decision maker has complete certainty about the items of choice. For instance, Dom Lalisse knew the exact prices of food and clothing, and Sue Chen knew the precise trade-off between integrity and compensation at Merrill Lynch. Decision makers, however, often face uncertainty. For instance, in choosing among risky investment alternatives (such as stocks and bonds), an individual must forecast the likely payoffs. Even so, there can be significant uncertainty about the eventual outcomes. The analysis presented in this chapter can be extended readily to incorporate decision making under uncer- tainty.22 A detailed analysis of decision making under uncertainty is beyond the scope of this book. This section introduces a few key concepts that we will use later in this book.

22For example, E. Fama and M. Miller (1972), The Theory of Finance (Dryden Press: New York), Chapter 5.

Chapter 2 Economists’ View of Behavior 39

Expected Value

Taylor McClure sells real estate for RealCo. He receives a sales commission from his employer. For simplicity, suppose that Taylor has three possible incomes for the year. In a good year, he sells many houses and earns $200,000, whereas in a bad year he earns nothing. In other years, he receives $100,000. Probability refers to the likeli- hood that an outcome will occur. In this example, each outcome is equally likely, and thus has a probability of 1�3 of occurring. The expected value of an uncertain payoff is defined as the weighted average of all possible outcomes, where the probability of each outcome is used as the weights. The expected value is a measure of central tendency—the payoff that will occur on average. In our example, the expected value is23

Expected value � (1�3 � 0) � (1�3 � 100,000) � (1�3 � 200,000) (2.6)

� $100,000


Although Taylor can expect average earnings of $100,000, his income is not certain. The variance is a measure of the variability of the payoff. It is defined as the ex- pected value of the squared difference between each possible payoff and the ex- pected value. In this example, the variance is

Variance � 1�3(0 � 100,000)2 � 1�3(100,000 � 100,000)2 (2.7)

�1�3(200,000 � 100,000)2

� 6.7 billion

The standard deviation is the square root of the variance:

Standard deviation � (6.7 billion)1�2 � $81,650 (2.8)

Variances and standard deviations are used as measures of risk. It does not really mat- ter which we use, since one is a simple transformation of the other (higher standard de- viations correspond to higher variances). In this example, we focus on the standard deviation—in part because the standard deviation is expressed in the same units as the mean, dollars (the units for the variance would be dollars squared). Higher standard deviations reflect more risk. An event with a definite outcome has a standard deviation of zero.

Risk Aversion

Like most people, Taylor is risk-averse: Holding the expected payoff fixed, he prefers a lower standard deviation. He, therefore, gains utility from an increase in expected value, but he experiences a reduction in utility from increases in standard deviation. Figure 2.9 shows three of Taylor’s indifference curves. Each curve shows all combinations of expected value and standard deviation that give Taylor equal

23Note that the expected value need not equal one of the possible outcomes. As a weighted average, it can be

a value between outcomes. In this example, it happens to correspond to one of the possible outcomes,


40 Part 1 Basic Concepts

utility. In contrast to our previous analysis, here one of the objects of choice is a “bad”—Taylor dislikes risk. Thus, in this figure, the indifference curves have posi- tive slopes, and upward and leftward movements are utility-increasing (recall in the standard analysis that the curves have negative slopes, and upward and rightward movements are utility-increasing). The slopes of the indifference curves indicate Taylor’s degree of risk aversion. Steeper slopes translate into higher risk aversion. (If the slopes of the indifference curves are steep, Taylor must receive a relatively large increase in expected value for each additional unit of risk to maintain a con- stant level of utility.) If his indifference curves were totally flat, he would be risk- neutral. A risk-neutral person cares only about expected value and is indifferent to the amount of risk. Indifference curve 3 is associated with the highest level of util- ity, whereas curve 1 is associated with the lowest utility. Taylor is currently on curve 2. Given a choice among compensation plans with different expected payoffs and risk, Taylor will choose the combination that places him on the highest attain- able indifference curve.

Certainty Equivalent and Risk Premium

Figure 2.9 indicates that Taylor is indifferent between the risky commission scheme, which has an expected payoff of $100,000, and a certain income of $80,000. The $80,000 is Taylor’s certainty equivalent for the risky income stream—he is willing to trade the uncertain income of $100,000 for a certain income of $80,000. The dif- ference between the expected value of the risky income stream and the certainty equivalent is called the risk premium. This $20,000 premium, which comes in the form of a higher expected payoff, must be paid to keep Taylor indifferent between the risky income stream and his certainty equivalent.

Figure 2.9 Indifference Curves for Expected Value and Standard Deviation

This figure displays three indifference curves for a risk-averse individual. The individual prefers higher expected value but lower standard deviation. Standard deviation is a measure of risk. Since risk is a “bad,” the indifference curves are positively sloped. Upward and leftward moves are utility-increasing. Currently, the individual has a compensation package that has an expected value of $100,000 and a standard deviation of $81,650. The certainty equivalent of this package is $80,000. The risk premium is $20,000.


E xp

ec te

d va

lu e

(in d

ol la

rs )



1 Risk premium = $20,000

Standard deviation (in dollars)




Increasing utility


Chapter 2 Economists’ View of Behavior 41

Suppose that Taylor receives a job offer from a competing real estate company that would pay him a fixed salary of $90,000 per year. Taylor considers the new job to be the same as his current job in all dimensions other than the compensation plan. Taylor’s current compensation plan will not be sufficient to motivate him to continue to work for RealCo. Even though his current plan has a higher expected payoff, he would prefer the certain $90,000 to RealCo’s risky commission plan. If RealCo wants to retain Taylor, it must offer him a compensation package that provides the same level of utility as the $90,000 for certain.

Risk Aversion and Compensation

Diversified shareholders, who invest in portfolios of companies, own much of the stock of large firms. Managers are often ill-diversified, having much of their human and financial capital invested in one firm. As we will discuss later in this book, this difference in diversification can lead to managers being overly risk averse in their investment decisions relative to those shareholders would prefer. Shareholders can induce managers to undertake more risky investment by adopting compensation plans that reward good outcomes, but that do not penalize bad outcomes heavily. The top management of Enron (see Chapter 1) arguably went too far in this direc- tion, inducing their managers to take too much risk (effectively transforming their behavior from risk averse to “risk loving”).24 We expand on this issue later in the book.

ANALYZING MANAGERIAL DECISIONS: Risk Aversion versus Risk Taking

Lauren Arbittier decides to bet $2,000 on

number 35 of the roulette wheel in a Las Vegas

casino. Almost immediately she starts to question

her decision. Lauren normally is a risk avoider

who hardly ever gambles. But she works at Trilogy

Software where the CEO understands that taking

risks and suffering the consequences are critical to

the firm’s success. The CEO wants to develop peo-

ple who take chances. “You don’t win points . . .

for trying.” Lauren is participating in Trilogy’s

three-month training program for all new recruits.

It educates employees about, among other things,

how to evaluate risky projects, not just to immedi-

ately accept or reject the project because it is risky.

The program also suggests to employees that they

will not be rewarded at Trilogy unless they take

risks. Thus, although Lauren does not like taking

risks, working for Trilogy, she has economic in-

centives to do so.

There are at least three ways in which the Tril-

ogy training program might be effective: (1) It

changes employees’ preferences regarding risk

bearing. (2) It more effectively identifies individu-

als with the risk tolerances that Trilogy desires.

(3) It better communicates the consequences to

Trilogy employees of undertaking risky ventures.

Discuss the likely importance of these three mecha-


SOURCE: E. Ramstad (1998), “High Rollers, How Trilogy Software Trains Its Raw Recruits to Be Risk Takers,” The Wall Street Journal (September 21), A1.

24Most managers have risk-averse preferences (utility functions). Managerial actions, however, are a function

of both preferences and constraints. Thus risk-averse preferences can be offset or reinforced by the design of

the compensation plan. Compensation plans that limit the upside potential but not the downside induce less

risky choices, whereas plans that limit the downside but not the upside induce more risky choices.

42 Part 1 Basic Concepts

Summary In this chapter we summarize the way economists view behavior. In the economic model, individuals are seen as having unlimited wants but limited resources. They rank alternative uses of limited resources in terms of preference and choose the most preferred alternative. Individuals are clever in figuring out ways of maximizing their satisfaction (utility) in the face of resource constraints. Individuals are not necessar- ily selfish in the sense that they care only about their personal wealth: They also care about charity, family, religion, and society. They are not infallible supercomputers.

The opportunity cost of using a resource is the value of the resource in its best al- ternative use. For example, the cost of having a manager use five hours to work on a project is the value of the manager’s time in working on the next best alternative pro- ject. Economic decision making requires careful consideration of the relevant oppor- tunity costs.

Marginal costs and benefits are the incremental costs and benefits that are associ- ated with a decision. In calculating marginal costs, it is important to incorporate the opportunity costs of the incremental resources. For example, in deciding whether to purchase a new laptop computer, the marginal cost is its price and the marginal ben- efit is the value that the person places on the new computer. It is the marginal costs and benefits that are important in economic decision making. Action should be taken when the marginal benefits are greater than the marginal costs. Sunk costs that are not affected by the decision (e.g., unrecoverable funds previously spent on computers) are not relevant.

A utility function is a mathematical function that relates total utility to the amounts that an individual has of whatever items the individual cares about (goods). Preferences implied by a utility function are pictured graphically by indifference curves. Indifference curves depict all combinations of goods that yield the same level of utility. Individual choice involves maximizing utility given resource constraints. Graphically, the constraint depicts all combinations of goods that are feasible to ac- quire; it defines the feasible consumption opportunities. The optimal choice is where the indifference curve is tangent to the constraint. At this point, the individual is at the highest level of utility possible given the feasible opportunities.

Changes in opportunities result in changes in the optimal choice. An important implication is that managers can affect behavior by affecting constraints and opportunities. Managers, however, have to be careful. Individuals are clever at max- imizing their utility, and establishing disfunctional incentives can have perverse consequences.

We contrast the economic model with other models of human behavior that man- agers often use. We argue that the economic model is often more useful than alter- native models in managerial decision making.

The analysis in this chapter can be extended to the case where the decision maker faces uncertainty about the items of choice. An example of decision making under uncertainty is choosing among risky investment alternatives. One concept on which we will rely later in this book is risk aversion. When confronted with both a risky and a certain alternative having the same expected (or average) payoffs, a risk-averse per- son always will choose the certain outcome. A risk premium must be offered to en- tice the person to choose the risky alternative.

Throughout this chapter, we focus primarily on how managers might use this eco- nomic view to analyze and influence the behavior of employees. As we will see, the economic view is quite powerful and useful in explaining behavior in a variety of different contexts.

Chapter 2 Economists’ View of Behavior 43

G. Becker (1993), “Nobel Lecture: The Economic Way of Looking at Behavior,” Journal of Politi- cal Economy 101, 385–409.

M. Jensen and W. Meckling (1994), “The Nature of Man,” Journal of Applied Corporate Finance 7,


2–1. Suppose there are only two goods that Bob cares about—(1) material welfare and

(2) leisure time that he “buys” from the outside world at $40 per unit and $20 per hour, re-

spectively. He currently lives and works in Atlanta, has a budget totaling $1,000 per week for

these two goods, and consumes 11 units of material welfare and 28 hours of leisure time.

a. Show Bob’s consumption choice on a graph using the actual budget line and a hypothet-

ical indifference curve. (Label the axes and show the x and y intercepts.)

b. How will Bob’s optimal consumption choice change if the government imposes a

$10/unit tax on the material welfare good?

c. Suppose Bob has a chance to move (at zero cost) to Saint Louis where material welfare

and leisure time cost $50 and $10, respectively. His budget remains the same as before

and the government has decided not to impose a tax on the material good. Would Bob

move to Saint Louis? Why or why not? Explain.

2–2. Amiko is an investor in the stock market. She cares about both the expected value and stan-

dard deviation of her investment. Currently she is invested in a security that has an expected

value of $25,000 and a standard deviation of $10,000. This places her on an indifference

curve with the following formula: Expected Value � $15,000 � Standard Deviation

a. Is Amiko risk-averse? Explain.

b. What is Amiko’s “certainty equivalent” for her current investment? What does this mean?

c. What is the risk premium on Amiko’s current investment?

2–3. You have won a free ticket to see an Eric Clapton concert (which has no resale value). Bob

Dylan is performing on the same night and is your next best alternative activity. Tickets to

see Dylan cost $40. On any given day, you would be willing to pay up to $50 to see Dylan.

Assume there are no other costs of seeing either performer. Based on this information, what

is the opportunity cost of seeing Eric Clapton? (a) $0, (b) $10, (c) $40, or (d) $5025

Solutions to Self-Evaluation Problems 2–1. Individual Choice


b. The $10 tax increases the price of material welfare from $40 to $50. Refer to Figure 2.6

to see the effects of a price increase on optimal consumption (replace food with material



M at

er ia

l w el

fa re

u ni





Leisure time (hours)

Self-Evaluation Problems

Suggested Readings

25Suggested by R. Frank (2005), “The Opportunity Costs of Economics Education,” New York Times (September 1), C2.

44 Part 1 Basic Concepts

welfare on the vertical axis and clothing with leisure time on the horizontal axis). Bob is

likely to reduce his consumption of material welfare due to the price increase. His con-

sumption of leisure time could either go up, down, or stay the same depending on the

exact nature of his adjustment. The change in relative prices (reflected in the flatter bud-

get line) will work in the direction of motivating Bob to substitute units of material wel-

fare for more leisure time. However, Bob’s new budget line will not allow him to stay on

the same indifference curve as before (it rotates inward). His consumption of leisure time

will depend on the exact location of the new tangency point on a lower indifference curve

in the graph. Bob was spending $560 to purchase 28 hours of leisure time. Whether Bob

consumes more or less leisure time after the price increase depends on how much money

he has left over after buying fewer units of material welfare at the new higher price. For

example, if he only reduces his consumption of material welfare from 11 to 10 units, he

will have less money to spend on leisure time than before ($500 versus $560). Alterna-

tively, if he reduces his consumption of material welfare to 8 or fewer units he will have

more money than he had before to spend on leisure time. Which option he chooses de-

pends on his specific utility function.

c. Bob would want to move to Saint Louis. In Saint Louis, his current consumption bundle

only costs $830. This leaves $170 to spend on other goods. Since more is better than

less, he can always do better in Saint Louis than he is currently doing in Atlanta.

2–2. Decision Making Under Uncertainty

a. Yes, Amiko is risk-averse. She is willing to take on more risk only if it is associated

with a sufficiently higher expected return.

b. Amiko’s certainty equivalent is $15,000. She would be willing to accept a certain return

of $15,000 (the vertical intercept of her indifference curve) in lieu of her current risky

investment that has an expected return of $25,000 and a standard deviation of $10,000.

c. The risk premium on Amiko’s current investment is $10,000. This is the difference in

the expected return of her risky investment and the risk-free investment (certainty equiv-

alent). The $10,000 risk premium is what it takes in expected return to make her indif-

ferent between the risk and risk-free investments.

2–3. Opportunity cost is a subtle concept that requires careful analysis to implement. Even

trained economists can make mistakes if they are not careful to include all relevant costs in

the analysis. Two researchers from Georgia State University (P. Ferraro and L. Taylor)

posed the question to 200 professional economists at an annual meeting.

A careful application of the definition of opportunity costs yields a clear answer—$10.

The next best alternative use of your time, going to the Bob Dylan concert, produces a net

benefit of $10 (the $50 value you place on the Dylan concert minus the $40 to purchase

the ticket). Marginal analysis implies that you should go to the Clapton concert as long as

you obtain at least $10 worth of happiness from the concert. For example, if you value the

Clapton concert at $15, you are $5 better off going to the Clapton concert than the Dylan

concert, which yields only $10 of net value. Interestingly, only 21.6 percent of the profes-

sional economists surveyed chose the correct answer, a smaller percentage than if they had

chosen randomly. Additional surveys revealed that the incorrect answers were driven by

faulty analysis and not by the specific wording of the question. College students who had

taken a course in economics did even worse.

The lesson is that managers, students, and even economists should be careful to include

all of the relevant explicit and implicit opportunity costs in their analyses. Missing a hypo-

thetical question on opportunity costs is inconsequential. Managers can destroy significant

value if they make mistakes in evaluating opportunity costs in their decision making.

2–1. Which costs are pertinent to economic decision making? Which costs are not relevant?

2–2. A noted economist was asked what he did with his “free time.” He responded by saying that

“time is not free.” Explain this response.

Review Questions

Chapter 2 Economists’ View of Behavior 45

2–3. The Solace Company has an inventory of steel that it originally purchased for $20,000. It

currently has an offer to sell the steel for $30,000. Should Solace’s management agree to sell?


2–4. Suppose that you have $900 and want to invest the money for one year. There are three ex-

isting options.

a. The city of Rochester is selling bonds at $90 per unit. The bonds pay $100 at the end of

one year when they mature (no other cash flows).

b. Put the money under your mattress.

c. The one-year interest rate of saving in the Chase Bank is 7 percent.

Which one will you choose? What is the opportunity cost of your choice? Explain.

2–5. Suppose Juan’s utility function is given by U � FC, where F and C are the two goods avail-

able for purchase: food and clothing.

a. Graph Juan’s indifference curves for the following levels of utility: 100, 200, and 300.

b. Are these curves convex or concave to the origin? What does this shape imply about

Juan’s willingness to trade food for clothing?

c. Suppose Juan’s budget is $100 and the prices of F and C are both $5. Graph the budget


d. How many units of food and clothing will Juan purchase at these prices and income?

Show graphically. What is his corresponding level of utility?

e. The Johnson Company is the sole producer of clothing. What can the company do to in-

duce Juan to purchase more clothing? Show graphically. (The graph does not have to be


2–6. Suppose that Bob’s indifference curves are straight lines (as opposed to being convex to the

origin). What does this imply about Bob’s willingness to trade one good for the other? Give

examples of goods where this type of behavior might be expected?

2–7. Suppose that Bob’s indifference curves are perfectly L-shaped with the right angle occurring

when Bob has equal amounts of both goods. What does this imply about Bob’s willingness to

trade one good for the other? Give examples of goods where this type of behavior might be


2–8. a. Briefly describe the five models of behavior presented in this chapter.

b. What are the implications of these models for managers attempting to influence their

employees’ behavior?

2–9. Employees in a plant in Minnesota are observed to be industrious and very productive.

Employees in a similar plant in southern California are observed to be lazy and unproduc-

tive. Discuss how alternative views of human behavior and motivation might suggest differ-

ent explanations for this observed behavior.

2–10. Employees at a department store are observed engaging in the following behavior: (a) they

hide items that are on sale from the customers, and (b) they exert little effort in designing

merchandise displays. They are also uncooperative with one another. What do you think

might be causing this behavior, and what might you do to improve the situation?

2–11. One of the main tenets of economic analysis is that people act in their own narrow self-

interest. Why then do people leave tips in restaurants? If a study were to compare the size

of tips earned by servers in restaurants on interstate highways with those in restaurants near

residential neighborhoods, what would you expect to find? Why?

2–12. Several school districts have attempted to increase teacher productivity by paying teachers

based on the scores their students achieve on standardized tests (administered by outside

testing agencies). The goal is to produce higher-quality classroom instruction. Do you think

that this type of compensation scheme will produce the desired outcome? Explain.

2–13. A company recently raised the pay of employees by 20 percent. Employee productivity

remained the same. The CEO of the company was quoted as saying, “It just goes to show

that money does not motivate people.” Provide a critical evaluation of this statement.

46 Part 1 Basic Concepts

2–14. One physician who worked for a large health maintenance organization was quoted as saying:

One day I was listening to a patient’s heart and realized there was an abnormal rhythm. My first thought was that I hoped that I did not have to refer the patient to a specialist.

Indeed, HMO physicians have been criticized for not making referrals when they are war-

ranted. How do you think the physician was compensated by the HMO? Explain.

2–15. Insurance companies have to generate enough revenue to cover their costs and make a normal

profit—otherwise, they will go out of business. This implies that the premiums charged for

insurance policies must be greater than the expected payouts to the policyholders. Why would

a person ever buy insurance, knowing that the price is greater than the expected payout?

2–16. Critically evaluate the following statement: “Risk-averse people never take gambles.”

2–17. Suppose that an investment can yield three possible cash flows: $5,000; $1,000; or $0. The

probability of each outcome is 1�3.

a. What is the expected value and standard deviation of the investment?

b. How much would a risk-neutral person be willing to pay for the investment?

c. How much would a risk-averse person be willing to pay for the investment?

2–18. In order to spur consumer spending in 1998, the Japanese government considered an $85 bil-

lion voucher system whereby every Japanese consumer would receive a shopping voucher

that could be used to purchase Japanese products. For simplicity, assume the following: Each

consumer has wealth of 1 million yen, consumers must allocate this wealth between con-

sumption now (c1) and consumption later (c2), the interest rate is zero, the voucher is worth

100,000 yen, and it can be spent only in the current period. If it is not spent, it is lost.

a. Plot a budget line for a representative consumer both before and after the voucher pro-

gram (c1 and c2 are on the axes).

b. Do you expect that current consumption of a typical consumer will increase by the full

100,000 yen of the voucher? Explain.

c. How does the impact of this 100,000-yen voucher differ from simply giving the individ-

ual 100,000 yen?

2–19. People give to charity.

a. Is this action consistent with the “economic view of behavior”? Explain.

b. Suppose there is a big drop in charitable giving. At the same time there has been no de-

cline in per capita income or total employment. Using the economic model, what poten-

tial factors might have led to this decline in giving?

c. How might the decline in giving be explained by the product-of-the environment model?

2–20. The Japanese are very good at returning lost property to local police stations. If you lose a

wallet filled with cash in Japan it is likely to be turned into the police. This is true even

though the person finding it could keep it without anyone else knowing. This behavior is not

what you would find in New York City.

a. Does this observation about Japan imply that the economic model does not explain be-

havior in Japan? Explain.

b. Police stations in Japan are filled with lost umbrellas. It used to be that the typical

Japanese would make a trip to the local police station to search for a lost umbrella. Now

they don’t. Explain this behavior using the Economic Model.

c. Do you think that the typical Japanese is more likely to come to a police station to find

a lost cell phone or a lost umbrella? Explain using the Economic Model.

2–21. Some states in the United States allow citizens to carry handguns. Citizens can protect them-

selves in the case of robberies by using these guns. Other states do not allow citizens to carry

handguns. Criminals, however, tend to have handguns in all states. Use economic analysis to

predict the effects of handgun laws on the behavior of the typical criminal. In particular:

(1) Do you think criminals will commit more or fewer robberies in the states with the laws?

(2) How do you think the laws will affect the types of robberies criminals commit? Be sure

to explain your economic reasoning.

2–22. Discuss the following statement: “Sunk costs matter. People who pay $20,000 to join a golf

club play golf more frequently than people who play on public golf courses.”

Chapter 2 Economists’ View of Behavior 47

2–23. Jenny is an investor in the stock market. She cares about both the expected value and stan-

dard deviation of her investment. Currently she is invested in a security that has an expected

value of $15,000 and a standard deviation of $5,000. This places her on an indifference

curve with the following formula: Expected Value � $10,000 � Standard Deviation.

a. Is Jenny risk averse? Explain.

b. What is Jenny’s “certainty equivalent” for her current investment? What does this mean?

c. What is the risk premium on her current investment?

2–24. Accounting problems at Enron ultimately led to the collapse of the large accounting firm

Arthur Andersen. When the Enron scandal first became public, Andersen’s top management

blamed one “rogue partner” in the Houston office who they claimed was less honest than

other partners at the firm. They fired the partner and asked that people not hold the remain-

ing partners accountable for “one bad apple.” What model of behavior was Andersen’s man-

agement using when it analyzed the source of the problem? According to the economic

view of behavior, what was the more likely cause of the problem?

2–25. According to a recent article in The Atlanta Journal-Constitution (January 29, 2004), “ma-

terialism, not necessity, gave birth to dual-income families.” In supporting the argument, the

author cites the following figures from the Department of Commerce: In 1970 the average

wage per job was $6,900, which in 2001 dollars (adjusting for inflation) amounts to

$31,500. In 2001, the average wage per job was $35,500. The main thesis of the article is

that dual-income families are a result of a shift in consumer preferences toward consump-

tion as opposed to leisure time/time spent with the family.

a. Assume the average person worked 250 days during a year both in 1970 and 2001, and

that, as reported in the article, only one person worked in the average family in 1970,

while both parents did in 2001. Provide a graphical analysis of the typical family’s

choice between family income and combined parent leisure time that supports the au-

thor’s argument, relying on the tools presented in class. Be careful in labeling your

graph(s), and provide a clear and concise explanation for your graph(s). Note that there

are 365 days in a year so that the total parent leisure time that is possible is 730 days

(assuming neither spouse works). Assume it is possible for each family member to work

anywhere from 0 to 365 days a year (at the going salary rate) if they choose to do so.

b. Assume that in 1971 the average single person worked 220 days per year, while the

same person worked 260 days per year in 2001. Moreover, suppose the average daily

wage in 2001 dollars was $125 in 1970 and $140 in 2001. Show graphically how the au-

thor’s argument would not necessarily apply to the average single person (i.e., assume

preferences are unchanged). Explain clearly and concisely why the average worker may

be choosing to work more in 2001 and carefully label your graph.

2-26. Russell and Joe have hired Maria to help cook in their restaurant. Maria had previously

owned her own breakfast business. Her speed in cooking was well-known. Russell and Joe

have been surprised that her productivity has fallen significantly since she became their paid

employee. Use the economic view of behavior and marginal analysis to provide a potential

explanation for Maria’s reduced productivity.

2-27. Michael is a fan of the Rhinos—the local professional soccer team. At the beginning of the

season, he purchased nonrefundable season tickets to their 10 home games for a total of

$100. Michael places equal value on each of the home games. His value for any given game

is independent of how many other games he attended during the year. Michael would be

willing to stay at home, which he derives no benefit from, and miss an individual game, if

he could sell the ticket for one game for $20 or more.

Michael has attended three out of the last five home games. The sixth home game of the

season is tomorrow night. Michael’s friend Fred has offered to sell him an extra ticket to a

sold-out concert for $50 that happens to be on the same night as the game. Normally,

Michael would be willing to pay $70 to attend the concert. There is no way that Michael

can attend both the Rhinos’ game and the concert. Looking online, Michael finds that he

can sell his ticket for tomorrow night’s Rhinos’ game for $5.

a. What type/types of cost is the $100 that Michael paid for the season tickets? Explain why.

b. Which event will Michael decide to attend? Explain why.

48 Part 1 Basic Concepts

c. How much would Fred have to charge Michael for the concert ticket in order to make

Michael not care which event he attended? Explain why.

2-28. An entrepreneur quits his job as a banker and invests $100,000 of his savings in a new busi-

ness venture that he will manage. Discuss the two most obvious opportunity costs that he

will incur from this decision.

2-29. In one hour, John can assemble either 20 telephones or 10 answering machines. It takes

Sally two hours to assemble either 20 telephones or 10 answering machines. Does either

person have an absolute advantage in assembling either product? What about a competitive

advantage? Explain (make sure you define absolute and competitive advantage and provide

the opportunity costs for each person). Draw the combined production possibilities curve

for the two people assuming that they work eight hours. Put answering machines on the

horizontal axis.

2-30. You are trying to decide whether to fly or drive from Rochester to Boston during your sum-

mer visit. The trip is approximately 400 miles. You can purchase a round-trip nonstop flight

for $230. The duration of each flight is 3.5 hours (seven hours in total). If you drive, it will

take you approximately seven hours each way (14 hours in total). Your car gets 32 miles to

the gallon and you expect gas will cost $3/gal. Tolls are $15 each way. You value your time

at $15 per hour.

a. Calculate the total cost of driving to Boston and back.

b. Under these conditions, will you drive or fly?

c. What if you value your time at $20, rather than $15, per hour?

d. You are offered a deal for $180 total airfare if you take a flight with a one hour layover

in NYC (total flying time does not change). Do you take the deal? Will this change

your flying versus driving decision? (Use the $15 per hour value of time in the


e. What are some other costs you may want to consider in this analysis?

f. If you were expecting a snow storm, how might this change your analysis? What are

some other costs you may want to take into consideration?

2-31. The school is having a happy hour on Friday. If you go, you will get two free drink tickets

and snacks, for which you would normally pay $15. However, you will have to pay $10 for

the cab fare home. You also have a free student ticket to the local profession team’s soccer

game. There is no resale value, as free tickets are still available.

a. What is the opportunity cost of going to the soccer game?

b. If you would not normally pay to go to the soccer game, which will you choose?

2-32. Bill and Chris produce balls and bats. In one hour, Bill can produce two bats or four balls,

while Chris can produce one bat or three balls. They each work 10 hours a day.

a. Which of the following, if any, is true? Explain why.

1. Chris has an absolute advantage producing balls and a comparative advantage

producing balls.

2. Chris has an absolute advantage producing balls and Bill has a comparative

advantage producing balls.

3. Bill has an absolute advantage producing balls and a comparative advantage

producing balls.

4. Bill has an absolute advantage producing balls and Chris has a comparative

advantage producing balls.

b. If Chris and Bill each split their time evenly between producing balls and producing

bats, what is the total number of bats and balls that will be produced in one day?

c. Can the combined production of balls be increased while holding the number of bats

produced constant? How?

d. Graph the combined production possibilities frontier for Bill and Chris. Place Bats on the

vertical axis and Balls on the horizontal axis. [The graph would show the maximum com-

bined amount of bats that could be produced by Bill and Chris for each feasible combined

amount of balls that could be produced.]

Chapter 2 Economists’ View of Behavior 49

Appendix A: The main text of this chapter provided a simple graphical analysis of individual choice.

Consumer It introduced utility functions, indifference curves, and budget lines. It depicted the op-

Choice timal choice as the tangency between an indifference curve and the budget line where the willingness and ability to trade are equal. This simple analysis explained how the optimal choice changes with changes in relative prices or income. This economic framework has important implications since managers often want to influence and/or predict the behavior of individuals, such as customers and employees.

This appendix extends the economic framework of individual choice (commonly called consumer choice). Its intent is to provide a deeper and richer understanding of this important model of behavior. This appendix also discusses how this model re- lates to an important topic that is covered in more detail later in this book—demand functions.

Marginal Utility A utility function expresses the relation between a person’s total utility and the level of goods consumed. Utility functions can take many forms. For illustration, suppose that Tom Morrell values only food and clothing and that his utility function is

U � FC (2.9)

where F is the units of food and C is the units of clothing that Tom consumes within the period. Notice, this simple utility function is multiplicative in both food and clothing. If Tom has no clothing, then no matter how much food he has, he has utility of zero. Likewise, if he has no food, then no matter has much clothing he has, his utility again is zero. If Tom’s consumption bundle consists of 20 units of both food and clothing his utility is 400, while his utility is only 100 if he has 10 units of both goods. Tom prefers the first bundle, but he is not necessarily four times happier when he has 20 units, rather than 10 units, of each good. The utility function provides an ordinal ranking of consumption bundles—not a cardinal rank- ing where “absolute” comparisons can be made.

Marginal utility measures the additional utility that is obtained by consuming one additional unit of a good, while holding all other goods constant. Marginal utility is an important concept in economic analysis since optimizing individuals focus on the mar- ginal (incremental) benefits and costs in making consumption choices. Figure 2.10 graphs Tom’s utility as a function of food, while holding clothing constant at 10 units. The equation for this graph is

U � 10F (2.10)

The marginal utility of food in this example is 10—for each additional unit of food that Tom consumes he receives 10 additional units of utility. More generally for any given quantity of clothing, C, the marginal utility of food is C (given Tom’s utility function). Similar logic implies that the marginal utility of clothing is F.26 We denote the marginal utilities for food and clothing by MUF and MUC , respectively.

26Note for the mathematically inclined: The marginal utility of any good X is equal to the partial derivative of

the utility function with respect to X. In this example, the partial derivative with respect to F is C, and with

respect to C is F. In this example, the marginal utility of each good is constant. More realistically, the

marginal utility of a good will eventually decline as the consumer continues to receive more of the good. We

employ this utility function only to simplify the presentation.

50 Part 1 Basic Concepts

Slope of an Indifference Curve Now that we have defined marginal utility, we can derive the slope of an indifference curve. Consider Tom’s indifference curve for U � 100. The equation for this indif- ference curve, which is pictured in Figure 2.11, is F � 100�C.27 Because all points on the indifference curve generate 100 units of utility, the total gain in utility


U til










1 3 5 7 9 11

Slope = 100/10 = 10

13 15 17 19 21 23 25 27 29 31 33 35 Food (F )



Figure 2.10 Tom’s Utility as a Function of Food with Clothing Held Constant at 10 Units (U � 10F )

This figure displays Tom’s utility as a function of food with clothing held constant at 10 units. Marginal utility measures the additional utility that is obtained by consuming one additional unit of a good, while holding all other goods constant. The marginal utility of F in this example is 10 (the slope of the line)—for each additional unit of F that Tom consumes he receives 10 additional units of utility.

27The decision to place food on the y-axis is arbitrary. If we had placed clothing on the y-axis the equation for

the indifference curve would be C � 100�F.

5 10 20 Quantity of clothing




Q ua

nt ity

o f

fo od

Indifference curve for 100 units of utility

A (5, 20): MRS = 4

B (10, 10): MRS = 1

C (20, 5): MRS = .25

Figure 2.11 Slope of One of Tom’s Indifference Curves: �(MUC/MUF)

This figure displays an indifference curve for 100 units of utility from Tom’s utility function: U � FC. The equation for the indifference curve is F � 100�C. The slope of an indifference curve at any point is �(MUC�MUF). The slope at a point is defined as the slope of the tangency line at that point. The tangency lines at points A and C are two examples. In this example, the slope at any point is �(F�C). The absolute value of the slope, which is called the Marginal Rate of Substitution (MRS), declines continuously along the curve. This property implies that Tom becomes less willing to trade F for C as C increases relative to F.

Chapter 2 Economists’ View of Behavior 51

associated with an increase in C must be balanced by an offsetting decline in utility from reduced consumption of F (as Tom moves along the curve):

MUC (�C ) � MUF (�F ) � 0 (2.11)

where �F and �C represent the changes in food and clothing. The first term in Equa- tion (2.11) represents the change in utility from changing the amount of clothing, while the second term represents the change in utility from changing the amount of food. These terms are equal in magnitude and of opposite sign along an indifference curve, and so total utility remains unchanged.

The slope of the indifference curve for a small change in C is given by (�F��C ). Rearranging Equation (2.11):

(�F��C ) � �(MUC�MUF) (2.12)

Slope of the indifference curve � �(MUC�MUF)

In the example where U � FC, the slope of the indifference curve is �(F�C ) since MUC � F and MUF � C. For example, at the point where F � 5 and C � 20 the slope is �.25. The intuition for this result is as follows: If a marginal unit of clothing yields five units of utility, while a marginal unit of food yields 20 units of utility, food can be traded for clothing at a rate of .25 units for one unit, and utility will remain the same (for small changes in the two goods).

The absolute value of the slope of an indifference curve is called the Marginal Rate of Substitution (MRS). The MRS reflects the individual’s willingness to trade at a point on an indifference curve (in this example, trading food for clothing). Consider Tom’s willingness to trade when he has a consumption bundle of 10 units of both goods. The slope at this point is �(10�10) � �1. The MRS, which is equal to one, implies that Tom is willing to give up a small quantity of food to receive an equal number of units of clothing. (If Tom were to increase his con- sumption of C by 1 unit he would have to reduce his consumption of F by ap- proximately 1 unit to keep his utility the same.) The MRS declines along the con- vex curve indicating that Tom becomes less willing to trade food for clothing as the amount of clothing increases relative to food.28

Individual Choice Recall that the equation for the budget line is

F � (I�PF) � (PC�PF)C (2.13)

The absolute value of the slope, (PC�PF), reflects the consumer’s ability to trade in the market place. For example, when the price of clothing is $2/unit and the price of food is $1/unit, two units of food must be given up to consume one additional unit of clothing (PC�PF � 2). The intercept (I�PF) indicates how many units of food could be purchased if the entire budget is spent on food.

At the optimal consumption bundle the budget line is tangent to an indifference curve (which is the highest attainable indifference curve given the budget constraint).

28The slope of a curve is defined at a point on a curve and the slope changes along the curve. The slope of �1

at the point [10,10] reflects Tom’s willingness to trade for very small changes in C. For a full unit change in

C, the decline in F is only approximately (as we reduce the size of the exchange—.1 unit of food or .01 unit

of food—the change in utility approaches zero) equal to one. A one unit increase in C and a one unit decrease

in F produce a consumption bundle with 99 units of utility. This value is approximately equal to the starting

point of 100.

52 Part 1 Basic Concepts

This condition implies that the MRS is equal to the ratio of the prices at the optimum. Since the MRS is equal to the ratio of the marginal utilities for the two goods:

MUC�MUF � PC�PF (2.14)

The left side of Equation (2.14) represents the willingness to trade, while the right side reflects the ability to trade. At the point where the consumer is maximizing his utility (the optimum), the two are equal.

We can rearrange Equation (2.14) as follows:

MUC�PC � MUF�PF (2.15)

Equation (2.15) is an important and familiar result in consumer theory. It says that the consumer’s utility is maximized when the budget is allocated among goods so that the marginal utility per dollar of expenditure is the same for each good. At any combination where this condition does not hold, the consumer can be made better off by making feasible changes in the consumption bundle. For example, suppose Tom has an initial bundle where the marginal utility per dollar for clothing is 10 and for food is 20. Since he is getting more utility per dollar from food, he should spend less money on clothing and more on food. As he trades clothing for food, his marginal utility of clothing increases while his marginal utility of food decreases. Tom will eventually reach the optimal consumption bundle where the marginal utility-to-price ratios are equal. Equation (2.15) reflects a condition known as the equal marginal principle—the marginal utility per dollar is the same for all goods at the optimum.29

This principle reappears in various forms in the economic analysis of both consumer and producer behavior.

Solving for the Optimal Consumption Bundle Suppose that Tom has a budget of $100 and the prices of food and clothing are $1 and $2, respectively. How much of each good will he buy? This problem is straightforward since it involves two unknown variables (F and C ) and two independent equations. One equation is the optimality condition Equation (2.14); the second is the budget line Equation (2.13).

At Tom’s optimal choice, the MRS must equal the price ratio (i.e., MUC�MUF � PC�PF). Substituting the values for Tom’s marginal utilities yields

F�C � 2 (2.16)

C � F�2

Tom must also satisfy his budget constraint:

F � 100 � 2C (2.17)

We can solve for the amount of food that Tom will purchase by substituting Equation (2.16) into Equation (2.17):

F � 100 � 2 (F�2)

F* � 50

C* � 25 is found by substituting F* � 50 into Equation (2.16). At the optimal consumption bundle, Tom obtains 1,250 units of utility (25 � 50).

He can increase his consumption of clothing by purchasing 26 units of clothing and

29This condition is also referred to as the “equimarginal principle.”

Chapter 2 Economists’ View of Behavior 53

48 units of food (26 � $2 � 48 � $1 � $100). However, this bundle would yield only 1,248 units of utility (26 � 48). Alternatively, he could decrease his consump- tion of clothing by one unit and increase his consumption of food by two units (C � 24; F � 52). This bundle also would produce only 1,248 units of utility. Indeed any feasible alternative bundle would yield less than 1,250 units of utility.

The equal marginal principle holds at Tom’s optimal consumption bundle—the marginal utility per dollar of expenditure (MUi�Pi) is 25 for both goods. This condition implies that Tom has no incentive to shift expenditures from one good to the other since both goods yield the same marginal utility per dollar of spending.

The marginal utility of income is defined as the additional utility that the con- sumer receives from one additional dollar of income. It can be shown that at the op- timum, the marginal utility-to-price ratio for all goods is equal to the marginal utility of income. For example, if Tom’s income increases by $1, he could increase his util- ity by 25 units by purchasing additional quantities of either good.

Demand Functions A demand function expresses the mathematical relation between the quantity de- manded for a product (how many units consumers will purchase) and the factors that determine consumer choice (such as prices and income). In a more general setting than our simple example, the demand for clothing is likely to be affected by the price of clothing, consumer income, the prices of other products, and other variables (such as advertising expenditures). Managers care about consumer choice since a good under- standing of the demands for their products is important for making productive invest- ment, pricing, advertising, and other decisions. In subsequent chapters, we focus on ag- gregate demand for a product (total demand across all consumers in the market) without directly tying the analysis back to individual consumer behavior as analyzed in this chapter. Nevertheless, it is useful to recognize that aggregate demand for a given product can conceptually be derived from the framework presented in this appendix.

The derivation of Tom’s demand function for either food or clothing is particu- larly easy (given his utility function). Tom’s optimal consumption bundle is where his MRS equals the price ratio. In Tom’s case, this condition is

F�C � PC�PF (2.18)

Tom’s total expenditures on either food or clothing is equal to the quantity purchased of the good times its price. By cross-multiplying Equation (2.18) we see that Tom’s expenditures on food and clothing are always equal:

(F � PF) � (C � PC) (2.19)

Equation (2.19) implies that Tom will always spend half his income on each good (this result is driven by his particular utility function). It follows that Tom’s total ex- penditures on clothing are (C � PC) � I�2. Solving for C produces Tom’s demand function for clothing:

C � I�(2PC) (2.20)

This demand function implies that Tom will purchase more clothing as his income rises and less clothing as his income falls. His clothing purchases vary inversely with the price of clothing.

In our example, Tom had an income of $100 and PC � $2/unit. Consistent with Equation (2.20), we found that he consumed 25 units of clothing. The demand func- tion implies that if Tom’s income were to increase to $200 (holding price constant)

54 Part 1 Basic Concepts

he would purchase 50 units of clothing. In contrast, his clothing purchases would fall to 12.5 units if PC increased to $4 (holding income constant at $100). Tom is only one consumer who purchases clothing. The total (aggregate) quantity demanded for clothing at a given price is equal to the sum of the purchases made by all consumers in the market.

Since Tom always spends half his income on clothing, the amount of clothing that he purchases is not affected by the price of food. This is a special case, which does not hold for many other utility functions. Consider Anne George whose utility func- tion is U � C .5 � F.5. Anne’s demand for clothing is C � I�[(P2

C�PF) � PC].30 This function indicates that Anne’s clothing purchases increase with income and the price of food, but decrease with the price of clothing. For example, at the initial prices and income (PF � $1, PC � $2, and I � $100) Anne purchases 16.67 units of clothing. If the price of food were to increase from $1 to $2, her demand for clothing increases to 25 units. When the prices for food and clothing are equal she spends half her in- come on each good. As relative prices change, Anne spends a higher percentage of her income on the relatively less-expensive good.

Income and Substitution Effects Equation (2.20) indicates that Tom’s demand for clothing decreases with the price of clothing. Figure 2.12 displays the example where the price of clothing increases from $2 to $4 (holding income constant at $100 and the price of food at $1). Remember, Tom’s utility function is U � CF. The price increase causes the budget line to rotate inward. The new budget line, B 2 is steeper than the original line, B1 (slopes of �4 and �2, respectively).

The inward shift of the budget line implies that Tom has less purchasing power than he had prior to the price increase. The area between the two budget lines con- tains consumption bundles that he could have purchased when PC � $2 that he can no longer afford. The reduced consumption possibilities imply that Tom has effec- tively less purchasing power than he had prior to the price increase. Thus, an increase in the price of clothing has two effects. One is to increase the price of clothing rela- tive to the price of food (i.e., PC�PF increases); the other is to reduce Tom’s effective income (purchasing power). As we will see, both effects influence Tom’s response to the price increase.

Tom purchased 25 units of clothing and 50 units of food when PC � $2, PF � 1, and I � $100. This choice, which is at the point of tangency between the original indifference curve, I1, and budget line, B1, is labeled as t1 in Figure 2.12. His optimal consumption bundle following the price increase, consisting of 12.5 units of clothing and 50 units of food, is pictured by the point of tangency, t2, between the indifference curve, I2, and the new budget line, B2. The decline in the quantity demanded for clothing from 25 units to 12.5 units represents the total effect of the price change (a decrease of 12.5 units). The total effect can be decomposed into a substitution effect and an income effect.

The substitution effect is the change in the quantity demanded of a good when its price changes, holding the prices of other goods and utility constant. To hold utility constant, Tom must be compensated for the price increase by receiving enough addi- tional income to maintain his previous level of utility of 1,250 units (U � FC � 25 � 50 � 1,250). Without this increase in income, he could not afford any of the bundles

30For practice, derive the demand function from Anne’s utility function (for this utility function: MUC �

.5C (�.5) and MUF � .5F (�.5)).

Chapter 2 Economists’ View of Behavior 55

on the original indifference curve at the new prices. To focus on the effect of changes in relative prices (versus changes in effective income), we examine what Tom would do if he actually received this hypothetical income increase. This hypothetical situa- tion is depicted by an imaginary budget line, B in Figure 2.12—which is parallel to the new budget line, B2, and tangent to the original indifference curve, I1, at t . The resulting combination of 70.6 units of food and 17.7 units of clothing is the least expensive bundle that Tom can purchase at the new prices that yields 1,250 units of utility.31 Since the cost is $141.40, Tom’s income would have to increase by $41.40 to afford this combination. Thus, if he were to receive enough additional income to compensate for the price increase, he would respond by purchasing 7.3 units less of clothing and 20.6 units more of food than when the price of clothing was only $2. This substitution between clothing and food occurs because clothing is relatively more expensive. Figure 2.12 depicts the 7.3 unit decline in Tom’s clothing purchases with an arrow labeled “substitution effect.” The convexity of the indifference curves implies that the substitution effect is positive.



F oo



12.5 25 50 Clothing


Substitution effect: 25 – 17.7 = 7.3

Income effect: 17.7 – 12.5 = 5.2

Total effect = (7.3 + 5.2) = 12.5

I1 (utility = 1,250 units) I2 (utility = 625 units)







Figure 2.12 Income and Substitution Effects

This figure illustrates income and substitution effects. Tom’s original budget line and indifference curve are denoted by B1 and l1; here he chooses 25 units of clothing and 50 units of food as denoted by t1. An increase in the price of clothing from $2 to $4 causes the budget line to rotate inward as pictured by B2. At the new optimum, t2, Tom purchases 12.5 units of clothing and 50 units of food. The 12.5 unit decline in the demand for clothing is the total effect of the price change, which is the sum of the substitution and income effects. The substitution effect is 7.3 units. It is pictured by Tom’s optimal choice, t , which assumes that Tom has received additional income to keep him on the original indifference curve. The income effect of 5.2 units is the additional decline in demand due to the fact that Tom does not actually receive the hypothetical increase in income. The hypothetical increase is used to isolate the “pure price effect” from the effect of reduced purchasing power due to the price increase.

31This consumption bundle is found by solving two equations simultaneously. One equation is for the

indifference curve containing bundles that yield 1,250 units of utility (F � 1,250�C); the second equation is

that the slope of the indifference curve and the new budget line are equal at the point of tangency (F�C � 4).

56 Part 1 Basic Concepts

The hypothetical $41.40 increase in income is used to isolate the “pure effect” of the change in relative prices. Since Tom does not actually receive this extra income, he will not be able to purchase 70.6 units of food and 17.7 units of clothing (his ac- tual income is still $100). The decline in purchasing power from the price increase has an additional effect on Tom’s demands for clothing and food. The income effect is the change in the quantity demanded of a good because of a change in purchasing power, holding prices constant. The parallel shift in the budget constraint from B to B2 in Figure 2.12 captures Tom’s effective decrease in income. As Tom’s budget de- creases from the hypothetical level of $141.40 to the actual level of $100, he con- sumes 5.2 fewer units of clothing (17.7 � 12.5 � 5.2). The 5.2 unit reduction in quantity demanded is depicted in Figure 2.12 with an arrow labeled “income effect.”

The total effect is that Tom’s quantity demanded for clothing drops by 12.5 units (25 � 12.5 � 12.5) due to the price increase. The total effect, which is the sum of the substitution and income effects, is pictured by an arrow labeled “total effect.” The final result is that Tom purchases 12.5 units of clothing and 50 units of food and ob- tains 625 units of utility (50 � 12.5):

Total Effect � Substitution Effect � Income Effect

12.5 � 7.3 � 5.2

The substitution effect is always positive—changes in relative prices motivate sub- stitutions toward the relatively less-expensive good. The income effect for a normal good is also positive. As income decreases (increases) total consumption must decrease (increase); thus, on average the demand for goods must move in the same di- rection as the income change. Nonetheless, for some goods the income effect is nega- tive. For example, in contrast to a normal good the demand for canned meat products is likely to vary inversely with income (wealthy people are likely to shun canned meat and purchase fresh meat, such as steak). We call goods with demands that vary in- versely with income inferior goods. A positive income effect reinforces the substitu- tion effect and increases the magnitude of the response, while a negative income effect mitigates the substitution effect and reduces the magnitude of the response. In Tom’s case the 7.3 unit substitution effect is reinforced by the 5.2 unit income effect. For most goods, the income effect is small relative to the substitution effect, and thus the total effect usually is in the same direction as the substitution effect.

The income effect in Tom’s case is relatively large (42 percent of the total effect). This is due to the assumption that Tom can only purchase two goods. Since Tom spends half his budget on clothing, he experiences a large drop in purchasing power when the price of clothing doubles. In contrast to this simple example, most con- sumers purchase many goods and spend a relatively small percentage of their bud- gets on any one good (e.g., salt, toothpaste, apples, and so on). Thus, a change in the prices of the typical good does not have an important effect on the purchasing power of the consumer. This observation implies that for many products the substitution ef- fect is much more important than the income effect. For example, suppose that your nearby grocery store raises the price of cucumbers by $1/pound. Conceptually, your income (purchasing power) is lower than it was before since you can no longer pur- chase as many potential consumption bundles. This small decline in effective in- come, however, is not likely to be the driving force behind your response to the price change. The relative increase in the price of cucumbers might motivate you to use more tomatoes and fewer cucumbers in your next salad. However, this decision is driven by the change in relative prices of cucumbers and tomatoes—not by the small change in your purchasing power.

Chapter 2 Economists’ View of Behavior 57

Some goods, such as housing or transportation, constitute a relatively large propor- tion of the typical consumer’s budget. For these goods, income effects can be more important. Since these goods are the exception rather than the rule, we can safely ignore income effects in many applications. Correspondingly, we tend to concentrate on sub- stitution effects in analyzing the effects of changes in relative prices. The reader, how- ever, should be aware that other applications exist where income effects are important.

One prominent case where income effects can be important is the supply of labor. Figure 2.13 depicts Ralph Kramden who is choosing between work and leisure time. Ralph is a bus driver whose employer allows him to choose the number of hours he works each week. Ralph has a total of 100 hours per week that he divides between work and leisure activities (the remaining hours are used for sleeping, etc.). At a wage rate of $10 per hour, Ralph chooses to work 60 hours per week and has a total income of $600; the other 40 hours are used for leisure activities. The budget line ro- tates outward when the wage rate is increased to $20/hr. The new budget line is flat- ter than the original line (slope is 1�20 versus 1�10). The reduced slope captures the increase in the opportunity cost of leisure—leisure now costs Ralph $20/hr. The sub- stitution effect works in the direction of motivating Ralph to reduce his leisure time and to work more hours. The substitution effect in this example, however, is out- weighed by the income effect. At the higher wage rate, Ralph chooses to work only 40 hours per week; his total income is $800, which is $200 more than he made work- ing 60 hours at $10/hr. At an income level of $800, Ralph values an additional hour of leisure time at more than the $20 he could make from using the hour for work. At the lower level of income ($600) he placed a smaller value on an extra hour of leisure time (he had to work more hours to provide basic support for his family).

Magnitude of the Substitution Effect Economists typically assume that indifference curves are convex to the origin. Con- vexity is consistent with the behavioral observation that a person’s willingness to





Le is

ur e

tim e

(h r)

Total income $1,000 $2,000

Budget line for wage = $20/hr.

Budget line for wage = $10/hr.

Figure 2.13 Income Effects in the Supply of Labor

This figure displays Ralph Kramden’s choice between work and leisure. Ralph has a total of 100 hours per week that he divides between work and leisure activities (the remaining hours are used for sleeping, etc.). At a wage rate of $10/hr., Ralph works 60 hours/week and has a total income of $600. At a wage rate of $20/hr., he chooses to work fewer hours (40) and to consume more leisure time (60 hours). While the increase in the wage rate increases the opportunity cost of leisure time, the income effect is larger than the substitution effect. At the higher income level ($800), Ralph places greater value on leisure time and works 20 fewer hours than when the wage rate was $10/hr.

58 Part 1 Basic Concepts

trade one good for the other generally declines as the relative amount of the second good increases (i.e., the MRS declines as the good on the horizontal axis increases). While Tom is willing to trade a large amount of food for a unit of clothing when he has lots of food and little clothing, his willingness to trade food for clothing declines as he reduces his food stock relative to his supply of clothing.

The substitution effect is always positive with convex indifference curves. An in- crease in the relative price of one good motivates substitution away from that good to- ward other goods (holding utility constant). The magnitude of the substitution effect, however, varies depending on the convexity (curvature) of the indifference curve.

Figure 2.14 compares the typical indifference curve with the two extremes. The first extreme is the case of perfect complements where the indifference curve is shaped as a right angle. In this case, the two goods are used in fixed proportions. An individual receives no additional utility from receiving more units of just one of the goods. For utility to increase, the quantity of both goods must increase. An example is shoes. The typical individual requires both a left and right shoe. Utility is not in- creased if the individual receives a right shoe unless it is matched with a left shoe.

The other extreme is perfect substitutes, where the indifference curve is a straight line. In this case, the MRS does not change as the person receives more of one good relative to the other. For example, a person’s willingness to trade $10 bills for $20 bills remains at 2 for 1 regardless of the relative supply of the two goods.

While most goods are neither perfect complements nor perfect substitutes, the con- vexity of indifference curves varies among products. Some indifference curves have sig- nificant curvature (tend to be closer to right angles), while others are relatively straight. The substitution effect is smaller when the indifference curve is more convex (closer to perfect complements). For example, a small increase in the price of left shoes will not motivate consumers to purchase fewer left shoes and more right shoes. In contrast, a small price change can motivate large shifts from one good to another when they are

G oo

d Y

Good X

G oo

d Y

Good X

G oo

d Y

Good X

Normal Case Perfect SubstitutesPerfect Complements

Figure 2.14 Convexity of Indifference Curves

This figure compares the typical indifference curve with the two extremes. The first extreme is the case of perfect complements where the indifference curve is shaped as a right angle. In this case, the two goods are used in fixed proportions. An individual receives no additional utility from receiving more units of just one of the goods. The other extreme case is perfect substitutes, where the indifference curve is a straight line. In this case, the MRS does not change as the person receives more of one good relative to the other. The substitution effect in response to a change in relative prices is larger when the two goods are close substitutes than when they are close complements.

Chapter 2 Economists’ View of Behavior 59

close substitutes. For instance, a small price change can motivate a consumer to switch from one brand of orange juice to another if the consumer is largely indifferent between the two brands (i.e., they are viewed as close substitutes). Geometrically, as the convexity of an indifference curve increases, the consumer does not have to move as far from the initial optimum to reach the new optimum as relative prices change (higher convexity implies that the slope of the indifference curve is changing more rapidly along the curve).

Additional Considerations Our analysis has focused exclusively on interior solutions where the consumer opti- mally purchases positive quantities of both goods. This focus is justified because it is the usual case with convex indifference curves. Nevertheless, there are special cases where it is optimal for the consumer to spend the entire income on only one of the goods. This outcome is known as a corner solution. For example, in the case of perfect substitutes it is optimal for the consumer to purchase only one of the goods unless the budget line and straight-line indifference curve have the same slopes, in which case the consumer is indifferent between purchasing either of the two goods (and thus might purchase positive quantities of both goods). For instance, you might be relatively in- different between holding your cash as $5 bills or $10 bills at an exchange rate of two for one. However, you will hold only one type of bill at other exchange rates. If you have the ability to trade one $10 bill in the marketplace for three $5 bills you should clearly do so. Similarly, while you might be indifferent between purchasing two simi- lar brands of orange juice when they sell for identical prices, you will quickly shift to buying only one brand if that one goes on sale.

For simplicity we have focused on an example based on only two goods. When there are more than two goods in the marketplace, an increase in the price of one good can motivate a reduction in the demand for other complementary goods. For example, an increase in the price for playing golf can reduce the demand for golf equipment. The substitution effect constitutes movement away from golf-related goods to other goods, which are now relatively less expensive. We examine comple- mentarity and substitutability of products in greater detail in Chapter 4.

Calculus Derivation of Equal Marginal Principle The equal marginal principle states that the marginal utility to price ratio is equal for all goods at the consumer’s optimal consumption bundle. At any combination where this condition does not hold, the consumer can be made better off by making feasi- ble changes in the consumption bundle. This section provides a calculus-based de- rivation of this principle.

The utility function for the two-good case takes the following general form:

U � f (xi, xj) (2.21)

To find the slope of an indifference curve, we totally differentiate Equation (2.21). We set this differential equal to zero, since utility does not change along an indifference curve:

dU � [�U��xidxi] � [�U��xjdxj] � 0 (2.22)

The slope of the indifference curve is defined by dxi�dxj (when good i is placed on the y-axis). Thus the

slope of the indifference curve � �(�U��xj)�(�U��xi) (2.23)

� �MUj�MUi (2.24)

60 Part 1 Basic Concepts

This expression has a straightforward interpretation. For illustration, assume that at some fixed combination of xi and xj, the marginal utility of good i is 1 and the mar- ginal utility of good j is 2. This means that 2 units of i can be given up for 1 unit of j and utility will stay the same. This is true by definition, since j has twice the mar- ginal utility of i.

At a consumer’s optimum the slope of the budget line (–Pj�Pi) is equal to the slope of the indifference curve:

�MUj�MUi � �Pj�Pi (2.25)

Rearranging this expression yields the Equal Marginal Principle:

MUj�Pj � MUi�Pi (2.26)

This principle immediately generalizes to utility functions with more than two goods.


1. Define the following terms: marginal utility, ordinal utility, marginal rate of substitution, equal marginal principle, demand function, substitution effect, income effect, normal good, inferior good, perfect complement, and perfect substitute.

2. Susan Pettit’s preferences for coffee (by the

pound) and doughnuts (by the dozen) can be

characterized as follows:

MUcoffee � MUx � y2

MUdoughnuts � MUy � 2xy

a. If the ratio of relative prices is (Px�Py) � 6�3 � 2, and Susan’s income is $90 per

period, what combination of pounds of

coffee and dozens of doughnuts will she


b. Now let the ratio of coffee to doughnut

prices decline to unity (�1), holding the

price of doughnuts constant. How does

Susan respond to the reduction in the

relative price of coffee?

c. Redo parts (a) and (b) for the case of

income of $60 per period.

d. Derive Susan’s demand function for coffee.

e. Is coffee a normal or inferior good for this


f. Does Susan consider coffee and doughnuts

to be either perfect complements or perfect

substitutes? Explain.

3. Susan’s demand function for coffee in the previ-

ous problem includes only the price of coffee

and income. Thus, changes in the price of

doughnuts do not affect the demand for coffee.

Does this imply that there is no substitution effect between the two goods? Explain.

4. (More challenging problem) Mario Casali is a

TV newscaster who gets an annual clothing

allowance to buy suits that he must wear

during his televised forecasts. He allocates the

allowance each year between expensive Italian

suits and cheap American suits. Mario’s utility

function for suits is SA.5 where S is the number

of Italian suits bought and A is the number of

American suits bought. Last year, Mario

bought two Italian suits and four American

suits. [Note: MUS � A.5 and MUA � .5SA (�.5)]

a. If Mario was maximizing his utility last

year, what was the ratio of the price of an

Italian suit to the price of an American suit


b. What was Mario’s clothing allowance last

year if the price of an Italian suit was


c. If Mario has the same allowance this year

as last year, and American suit prices have

not changed, how high would the price of

Italian suits have to rise in order for Mario

to want to buy exactly one Italian suit this


Chapter 2 Economists’ View of Behavior 61

Appendix B: Introductory economics generally concentrates on single-period problems. For ex-

Inter-Temporal ample, how does a consumer choose to spend her income within a single period?

Decisions and How much should a firm produce in a single period to maximize its profits? What

the Fisher prices should it charge for this output?

Separation The frameworks introduced in this book, however, also are used to analyze

Theorem multiperiod problems. Indeed, this is what the modern study of finance is largely about. How does an individual choose between consumption today versus saving for the future? How does a firm choose between paying cash dividends now versus in- vesting to earn higher future profits?

This appendix provides a simple example of how the consumer choice frame- work presented in this chapter can be used to analyze inter-temporal consumption decisions. It also provides a brief introduction to the important problem of a firm deciding on whether to pay higher cash dividends now or to invest the cash to earn higher profits in the future. In so doing, we introduce an important concept that is the starting point of the modern study of finance—the Fisher Separation Theorem.

Simple Example of Inter-temporal Choice Mary Donaldson graduated from college five years ago and has been working as an entry-level employee at a bank. She has taken leave from her job to enter a one-year, full-time MBA program, which begins today. The bank has paid for her tuition, books, and basic room and board. It has also promised to promote her when she com- pletes the program. While Mary earns no salary from the bank while she is in the pro- gram, the bank has promised to pay her $75,000 per year in her new position. Mary currently has $25,000 in savings.

Mary can borrow money from a bank at a 5 percent interest rate if she wants to consume more while she is in school than she can buy with her $25,000. For exam- ple, she might want to rent a better apartment than the one she has been provided. She also might want to travel internationally on her school breaks or to upgrade her music system or automobile. Alternatively, Mary might want to save all or part of her $25,000 so that she can consume more in the future. She knows that she will likely want to buy a house once she starts her new job. She will also have moving and other expenses that will not be covered by her company. She earns 5 percent interest on her savings.

Mary’s problem is to decide how much to consume today versus how much to consume in the future given her current savings, future income, and the market in- terest rate. The key insights that arise from analyzing this problem can be illustrated most conveniently by assuming that Mary cares only about consumption at two points in time: today and one year from today. (This simplification is similar to our earlier focus on only two goods—food and clothing.) To simplify the analysis fur- ther, we assume that there is no uncertainty about Mary’s future income of $75,000 and that she pays no taxes on her interest or job income.

If Mary saves all $25,000 over the next year, she will have (1.05 � $25,000) � $26,250 at the end of the year in her account. Added to her $75,000 salary, the max- imum sum she could have to spend next year is $101,250. Mary, however, might want to increase her current consumption at the cost of not being able to consume as much next year. If Mary wants to spend more than $25,000 today, she will have to borrow to do it. The bank will not loan Mary more money than she is able to pay

62 Part 1 Basic Concepts

back (with interest) given her future income of $75,000. The maximum (B) that she can borrow is

(1.05) B � $75,000

B � $75,000/1.05 � $71,428

Adding the maximum borrowing of $71,428 to the initial $25,000 gives Mary a max- imum of $96,428 to consume today, assuming she consumes nothing next year.

Mary is highly unlikely to want to consume all of her funds in just one period. Rather she is likely to want consume at least something in each period. We refer to her $25,000 in current savings and the $75,000 she will receive next year in income as her endowment. One option is for Mary to consume her endowment at the time it is received. In this way, she neither saves nor borrows. Depending on her prefer- ences, however, Mary might want to borrow something from the bank to consume more today or to save part of her $25,000 to consume more in the future.

The solution to Mary’s decision problem can be pictured using the graphical framework introduced in this chapter. Figure 2.15 displays Mary’s inter-temporal budget line. The variable on the y-axis is Mary’s consumption next year (c2); the

100 20 70 Consumption today (C1) in $000’s

Mary is a net saver

Mary is a net borrower

Mary’s inter-temporal budget line slope = –(1.05)



C on

su m

pt io

n in

o ne

y ea

r (C

2) in

$ 00

0’ s

30 40 50 60 80 90 100 110 120 130














Figure 2.15 Mary’s Inter-Temporal Consumption Choice

Mary’s initial endowment consists of $25,000 in current savings and $75,000 in future income. The interest rate for borrowing and lending is 5 percent. The inter-temporal budget line shows all feasible combinations of consumption today and consumption next year, given Mary’s endowment and the interest rate. Mary’s optimal choice is where the budget line is tangent to an indifference curve. The figure depicts two possible optima that assume different preferences for Mary. The one on the left is the case where she is a net saver. She saves part of her initial $25,000 so that she can increase her consumption in the second period. The one on the right shows the case where she borrows against her future income to increase her current consumption. The actual outcome depends on Mary’s particular preferences.

Chapter 2 Economists’ View of Behavior 63

variable on the x-axis is her consumption today (c1). The budget line goes through her endowment and shows all the feasible alternatives that Mary has, given her en- dowment and the 5 percent interest rate. The y-intercept of $101,250 is the maximum amount she can consume next year if she chooses to consume nothing today. The x-intercept of $96,428 is the maximum she can consume today if she chooses to consume nothing next year. The slope of the budget line is �1.05, which more generally can be expressed as �(1 � r), where r is the interest rate. Note that a change in the interest rate will change the slope of the budget line. An increase would make the line steeper, while a decrease would make it flatter. In all cases the line would include the endowment point.

Mary’s preferences can be pictured with standard indifference curves. All combi- nations of c1 and c2 along a given indifference yield the same level of utility. Mary’s optimal consumption choice occurs at the point of tangency between her budget line and an indifference curve (assuming an interior solution). Mary would like to be on an even higher indifference curve, but this is the best she can do, given her endow- ment and the interest rate of 5 percent. Whether Mary decides to borrow or save de- pends on her preferences. Figure 2.15 depicts two possibilities. The tangency on the left occurs at a point where Mary is a net saver (she saves part of her $25,000 to in- crease next year’s consumption); the tangency on the right shows the case where Mary is a net borrower (she borrows to consume more today at the cost of not being able to consume as much next year). Which of these, if either, that she will choose depends on Mary’s particular preferences for inter-temporal consumption.

Exercise for the Reader. As previously noted, a change in the interest rate will change the slope of the budget line. This, in turn, will change Mary’s optimal choice. Use the standard choice diagram depicted in Figure 2.15 to demonstrate each of the following: If Mary is a borrower when the interest rate is 5 percent, she will also be a borrower if the interest rate falls to 3 percent. However, if Mary is a borrower at 5 percent and interest rate rises to 7 percent, she may still be a borrower, but she might decide to switch to being a lender. Now suppose that Mary is a lender at the initial 5 percent rate. What will she do if interest rates rise? What if they fall?

Terminology. The maximum that Mary can borrow, given her future income of $75,000, is $71,428 at the 5 percent interest rate. In the business world, the $71,428 is referred to as the present value of the $75,000 future cash flow. It is the amount that Mary would need today to obtain $75,000 in one year investing at the market in- terest rate of 5 percent. The ability to earn 5 percent on her savings implies that Mary would be indifferent between receiving $71,428 today or $75,000 one year from now (since she could reproduce the $75,000 future cash flow with $71,428 today). The process of transforming future cash flows into present values is called discounting. The present value in this example was obtained by dividing the $75,000 by 1.05. More generally the present value of a cash flow, C, occurring one year in the future is C/(1 � r), where r is the annual interest rate.32 The $75,000 cash flow in this ex- ample is the future value of the $71,428.

32Calculating present vales is a bit more complicated when expected future cash flows occur at multiple dates

in the future and when they vary in their riskiness. However, the basic concept is the same. A dollar today is

worth more than a dollar in the future (since you could obtain the same dollar in the future by investing a

smaller amount today).

64 Part 1 Basic Concepts

Fisher Separation Theorem Now consider an inter-temporal investment decision by a firm. Deon and Ramona each own 50 percent of the firm. The firm has $100,000 in cash that it could distrib- ute to the owners as cash dividends. This would give Deon and Ramona additional cash today. Alternatively the firm could use the funds to invest in a project that would allow it to pay higher future dividends to its owners.

Our previous example suggests that individuals can differ in their inter-temporal consumption preferences (some might want to save for the future, while others might want to spend more on current consumption). An important question is whether the managers of the firm should consider Deon and Ramona’s time preferences in de- ciding whether to invest in the project. The somewhat surprising answer is that under certain assumptions the answer is no.

Modern finance courses typically start with the assumption of perfect capital mar- kets, which are characterized by zero transaction costs, no taxes, and perfect infor- mation. In a perfect market, Deon and Ramona will unanimously agree that the firm should invest in the project if it increases the present value of the firm’s cash flows, even if they have quite different time preferences. This important result is one part of what is known as the Fisher Separation Theorem—named after the economist Irving Fisher. The other implication, which focuses on how the project is financed, will be addressed later in this appendix.

The basic logic for why the firm’s investment decision can be separated from its owners’ preferences is as follows. If the firm invests the $100,000 today it will re- ceive $110,000 as a cash inflow in one year. For simplicity, assume the firm makes this investment, then (1) the investment is riskless (the firm will receive the $110,000 for sure) and (2) the firm will discontinue operations at the end of the year and dis- tribute $55,000 each to Deon and Ramona. Alternatively, the firm might forgo the investment, discontinue operations and pay $50,000 to each of the owners in cash dividends today. We assume for this analysis that the firm, Deon, and Ramona can all borrow or lend at a market interest rate of 5 percent (subject to having the funds to pay back loans with interest).

If the firm decides to liquidate and pay $100,000 in cash dividends, the present value of the firm’s cash flows is by definition $100,000 (since the cash flows occur at the present time). The present value of $110,000 received in one year is $110,000/1.05 � $104,762. According to the Fisher Separation Theorem, both Deon and Ramona will want the firm to invest regardless of their inter-temporal preferences for consumption since it increases the present value of the firm’s cash flows.

To see why, suppose that Deon wants to consume all he can today, while Ramona wants to save all she can today to consume more in the future. If the firm invests in the project, it will be able to pay $55,000 to each owner in one year. Deon can bor- row $52, 381 from a bank using his future $55,000 as collateral ($52,381 � 1.05 � $55,000). In contrast if the firm does not make the investment, Deon will only have the current dividend of $50,000 to consume today—a loss of $2,381 in current con- sumption. Ramona, in turn, could borrow nothing today and have $55,000 to spend next year if the firm decides to invest. If instead the firm discontinues operations today, she will only have (1.05) � $50,000 � $52,500 to consume next year—a loss of $2,500 in future value. The difference is due to the fact that the firm can earn 10 percent on its investment, compared to the 5 percent that Ramona earns from per- sonal savings. The conclusion is that Deon and Ramona will both want the firm to invest even though they have dramatically different time preferences.

Chapter 2 Economists’ View of Behavior 65

Large corporations often have thousands of shareholders. The basic result illus- trated in this simple two-person example, however, readily extends to many joint owners. In a perfect capital market, all owners will unanimously agree that the firm should invest in any project that will increase the present value of its cash flows. The owners, in turn, can use their own borrowing and lending in capital markets to meet their individual inter-temporal preferences.

Financing Decision. The second part of the Fisher Separation Theorem focuses on the firm’s financing decision. According to the Fisher Separation Theorem, the fi- nancing decision does not affect value in a perfect capital market and is therefore ir- relevant. In other words, the firm’s financing decision can be separated from its in- vestment decisions. The investment decision should be based on present value, while the financing decision is irrelevant in a perfect market. In our current example, this implies that it would not matter whether the firm forgoes its current dividend and uses the $100,000 to finance the investment (as analyzed above) or pays the $100,000 in current dividends and borrows $100,000 to finance the investment.

The logic for this result can be illustrated by comparing how Deon and Ramona would fare if the firm were to borrow to finance the investment in the case analyzed earlier, where the project was financed by forgoing current dividends. If the firm pays the $100,000 it has on hand in current dividends and borrows $100,000 to fund the investment, it will have to pay $105,000 back to the lender in a year. Since its cash flow from the project is $110,000, it will have $5,000 leftover to distribute to Deon and Ramona in one year as dividends. If Ramona invests the initial $50,000 cash dividend at 5 percent, she will have $52,500 at the end of the year in savings plus the $2,500 dividend that she will receive at that time. The total of $55,000 is ex- actly the same as when the firm paid no dividends and used its own cash to fund the project. Thus, Ramona is indifferent as to how the project is financed. Similarly under the second option, Deon could borrow $2,500/(1.05) = $2,381 to finance cur- rent consumption (using his future dividend of $2,500 as collateral). Added to the $50,000 current cash dividend, he will have $52,381 to spend today—the same amount that he would have if the firm financed the investment with cash and paid no current dividends. Deon is also indifferent as to which of the two financing options the firm chooses.

Modern finance theory starts with this perfect capital market analysis. The as- sumptions of zero transaction costs and perfect information are then relaxed to ex- amine under what circumstances the Fisher Separation Theorem breaks down. The analysis focuses on whether there are “real-world” circumstances where the owners of the firm can disagree on a firm’s investment decisions or where the financing de- cision affects firm value? The answer to these questions is yes, but answering them is beyond the scope of this book.


3 C H A P T E R O U T L I N E

Goals of Economic Systems Property Rights and Exchange in a Market Economy

Dimensions of Property Rights Gains from Trade

Basics of Supply and Demand The Price Mechanism Shifts in Curves versus Movements along Curves Using Supply and Demand Analysis for Qualitative Forecasts Linear Supply and Demand

Supply and Demand— Extended Analysis

Price versus Quantity Adjustments Short-Run versus Long- Run Effects Industry Cost Increases and Price Adjustments

Prices as Social Coordinators Efficient Exchange and Production Measuring the Gains from Trade Government Intervention Externalities and the Coase Theorem

Markets versus Central Planning

General versus Specific Knowledge Knowledge Creation Specific Knowledge and the Economic System Incentives in Markets

Contracting Costs and Existence of Firms

Contracting Costs in Markets Contracting Costs within Firms

Managerial Decisions Summary Appendix: Shareholder Value and Market Efficiency

M uch of the world’s economic activity occurs within “free mar- kets” where individual decisions are coordinated through the price mechanism. For example, four of the countries with the largest gross domestic products (GDPs) in 2012 (the United

States, Japan, Germany, and France) all have developed market systems. China with the second largest GDP has made increased use of markets since the 1990s. Prior to that time it had been a centrally planned economy with an extremely low per capita GDP.

On closer inspection, however, it is evident that a substantial amount if not most of the production in modern economies takes place inside firms, where multimillion dollar resource allocation decisions (e.g., on what to produce and how to produce it) are made by managers without the use of market transactions. The monetary size of the world’s largest firms exceeds that of many economies. For instance, the 2012 GDPs of Peru, Kenya, and Portugal were $205 billion, $41 billion, and $212 billion, respectively; the 2012 net sales of ExxonMobil, Walmart, and Chevron were $453 billion, $447 billion, and $246 billion, respectively.

Exchange and Markets


1. Illustrate the concept of Pareto efficiency.

2. Explain the role of alienable private property rights in markets and why

voluntary trade takes place.

3. Define and apply the concept of comparative advantage.

4. Explain the difference between demand and supply functions versus demand

and supply curves.

5. Distinguish between movements along supply and demand curves and shifts in

the curves.

6. Explain the forces that move prices and quantities toward their equilibrium lev-

els in a competitive market.

7. Explain why long-run demand and supply curves are generally more elastic than

short-run curves.

8. Predict (qualitatively) the relative changes in price versus quantity when

demand or supply changes in applied settings.

9. Define and interpret consumer and producer surplus; define and interpret dead-

weight loss in terms of the value of foregone gains from trade.

10. Explain the effects of price controls both within the supply and demand model

and in real-world terms.

Chapter 3 Exchange and Markets 67

To be effective, managers must have a working understanding of both markets and firms. In this chapter, we contribute to this understanding by examining five impor- tant questions: Why do most people actively participate in market exchanges? How do market systems work? What is supply and demand analysis and how can it be used by managers? What are the relevant advantages of using market systems com- pared to central planning in large economies? Why do we observe so much economic activity conducted within firms in market economies?

Answers to these questions are particularly important to managers for two rea- sons. First, an understanding of how markets work helps managers make appropriate strategic and operational decisions (e.g., input, output, and pricing decisions). The supply and demand analysis that we introduce in this chapter is especially useful in many management applications. Second, understanding the relative advantages and disadvantages of markets, central planning, and firms is directly relevant to under- standing firm-level issues such as which decision rights to be decentralized to employees and whether to make or buy each of the firm’s inputs. The basic tools and concepts introduced in this chapter are used to analyze these specific management decisions in more depth in subsequent chapters.

Goals of Economic Systems Every economic entity—be it a national economy, firm, or household—is confronted with three basic issues:

• What to produce

• How to produce it

• How to allocate the final output

Economic entities can be organized in alternative ways to address these issues. For instance, national economies can rely on either central planning or free markets. Similarly, firms and households can use centralized decision making, where the CEO or head of household makes all major decisions. Alternatively, other people in the firm or household can be granted substantial decision-making authority.

Given the alternatives, what is the best way to organize economic activities? To answer this question, we need some criterion for comparing alternative systems. Un- fortunately, uniform agreement over such a criterion is unlikely. For instance, you might argue that an ideal system would produce your preferred mix of products and give them all to you—although your neighbor would certainly disagree. Given these differences in opinion, economists generally focus on a relatively uncontroversial but narrow criterion for comparing the effectiveness of economic systems: Pareto efficiency.1 The production and distribution of goods and services in an economy is said to be Pareto-efficient if there exists no alternative that keeps all individuals at least as well off but makes even one person better off. If an economic system is not producing or distributing goods efficiently, it is conceptually possible to make its members better off by adopting Pareto-improving changes (thus benefiting some members without hurting others).

To illustrate the concept of Pareto efficiency, suppose that an economy can pro- duce two goods: desktop and tablet computers. Currently the economy is producing

1The term is named after Vilfredo Pareto, 1848–1923, an Italian economist and sociologist.

68 Part 1 Basic Concepts

1,000 desktops and 2,000 tablets. The two goods are being produced efficiently only if it is impossible to increase the production of one of the goods without decreasing the production of the other (given existing technology and resources). In contrast, suppose that some of the firms that are currently producing desktops are wasting raw materials due to suboptimal production methods (e.g., the firms could be scrapping more metal than necessary because of the way they cut metal sheets into final parts). In this case, the firms could increase their production of desktop computers by choosing more effective production methods without having any effect on the num- ber of tablets that are being produced. Doing so would be a Pareto improvement. The distribution of desktops and tablets (once they are produced) among consumers in this economy is Pareto efficient if there is no alternative distribution that keeps all individuals at least as well off but makes even one person better off.

As example of inefficient distribution, suppose that John owns a tablet but prefers a desktop, while Gunter owns a desktop and is indifferent between owning a tablet or a desktop. The current allocation is not Pareto efficient since John would be made better off if he and Gunter were to exchange the two products, while Gunter would be no worse off. The trade would be Pareto improving. If a change in the allocation of the two goods in the economy adversely affects even one person, the move would not be Pareto-improving and an economist would have little formal basis to conclude whether the move would be good or bad for society at large.2

Within centrally planned economies, government officials decide what to pro- duce, how to produce it, and who obtains the final output. In free markets, these de- cisions are decentralized to individuals within the economy. At least in concept, a central planner could order any feasible production and distribution of goods. Thus, any allocation of resources that could be achieved by a market economy also could be achieved by a centrally planned economy—at least in principle. We begin by dis- cussing how market systems work and how they can produce a Pareto-efficient allo- cation of resources. We then discuss why in large economies a market is more likely to produce an efficient resource allocation than central planning.

Property Rights and Exchange in a Market Economy A property right is a legally enforced right to select the uses of an economic good. A property right is private when it is assigned to a specific person. Private property rights are alienable in that they can be transferred (sold or given) to another individ- ual. For example, if Valerie Fong owns an automobile, she can use the automobile as she sees fit (within limits set by traffic laws). Valerie can restrict others from using her vehicle. She also can sell the automobile (transfer to another person whatever property rights her ownership confers in the vehicle). The government maintains police and a court system to help enforce these property rights.

An important feature of a market economy is the use of private property rights. Owners of land and other resources have the legal rights to decide how to use these re- sources and frequently trade these rights to other individuals. They are free to start new

2Therefore, economics does not address the question of which of the many possible efficient resource

allocations is best for a society. Producing your preferred set of products and giving them all to you is

efficient (the allocation cannot be changed without making you worse off). However, others will argue that

the allocation is not fair or equitable. Economists have no special training in resolving these fairness or equity

issues. Thus, we focus our attention on efficiency, which most people will agree is a laudable objective—

given limited resources it is good not to waste them.

businesses and to close existing businesses. In contrast, in centrally planned economies, property tends to be owned by the state; government officials decide how to use these resources.

Dimensions of Property Rights

Ownership involves two general dimensions: use rights and alienability rights. These aspects of ownership are not always bundled together. You own your body in the sense that you can decide what activities to pursue. Yet, there are significant legal restrictions on alienability. For instance, you cannot enter a legally enforceable con- tract to sell one of your kidneys, despite the fact that you have two, can live com- fortably with one, and might value your second kidney much less than a wealthy in- dividual who is dying from kidney failure. This restriction eliminates the possibility of a free market in kidneys. In some transactions, it is possible to sell use rights while retaining alienability rights. For instance, in a rental contract, the renter obtains the rights to use an apartment, but does not own or have the right to sell the unit. Con- versely, the landlord has the right to sell the apartment, but does not have the right to use it during the term of the lease. (Rental, lease, and franchise agreements separate alienability and use rights; we examine these contracts in a later chapter.)


Patent for Government-enforced patents better define property rights in new inventions. Patents in the United States are awarded

for processes, machines, manufacturers, or compositions of matter that are considered useful, novel, and not obvious.

Patents protect the intellectual property rights of the inventor and thus protect the common good by providing

incentives to innovate novel and not obvious inventions. received a patent for the world’s first buyer-driven e-commerce system where users can go to the

Internet to name their price for goods and services. challenged whether’s process is really

novel and not obvious. In 2001, the parties settled. Internet businesses where consumers can name their own price have

to pay a royalty. This royalty is a tax on all Internet consumers. Awarding a patent for something that is

obvious lowers incentives for future innovations that use this process.

Source: J. Gurley (1999), “The Trouble with Internet Patents,” Fortune (July 19), 118; L. Flynn (2002), “The Web World Watches

Closely as British Telecommunications Stakes a Patent Claim on a Now-Ubiquitous Function: Hyperlinking,” New York Times (March 11).

Chapter 3 Exchange and Markets 69


Property Rights Insecurity in Colombia Colombia has a continuing stream of impoverished farmers who are leaving the countryside and migrating to cities.

They live in shantytowns that breed crime and violence. Yet Colombia has substantial arable land—an area equivalent

in size to North Dakota. And only about 20 percent is used for agriculture. These seemingly inconsistent facts are both

by-products of Colombia’s more than four decades long conflict between the government and a paramilitary force that

is deeply involved in drugs. The resulting violence induces many to flee. But others are forced off their land or

intimidated into selling at bargain-basement prices.

Source: J. Forero (2004), “Colombia’s Landed Gentry: Coca Lords and Other Bullies,” New York Times (January 21), A4.

70 Part 1 Basic Concepts

Gains from Trade

To understand how a market economy works, we must understand the motives for trading property rights. Why do people buy and sell? The basic answer is to make themselves better off.

Within the economic framework, people order their preferences and take actions that maximize their level of satisfaction (utility). Trade takes place because the buyer places a higher value on the item than the seller. The corresponding gains from trade make both parties better off—voluntary trade is mutually advantageous. For example, if José Coronas is willing to pay up to $26,000 for a particular automobile and Rochester Motors is willing to sell the automobile for as little as $20,000, the poten- tial gains from trade are $6,000 ($26,000 � $20,000). If the automobile trades at $23,000, both parties are $3,000 better off. José gives up $23,000 to buy something that he values at $26,000, while Rochester Motors obtains $23,000 for something it values at only $20,000. At other prices between $20,000 and $26,000, the total gains are still $6,000 but they are not split evenly. For example, at a price of $25,000, José gains $1,000 in value, while Rochester Motors gains $5,000.3

From where do these gains from trade come? One source is differences in prefer- ences. The buyer and seller simply may place different values on the traded item. For

3Sometimes, individuals regret a trade after the fact. For instance, José might be unhappy after he purchased

a particular automobile from Rochester Motors. But given the information he had at the time of the

transaction, he must have expected it to be advantageous to purchase the automobile or else he would not

have done so (at least from Rochester Motors). José’s ability to say no limits the extent to which he can be

exploited in any voluntary trade.


While Animosity between the Governments of Venezuela and the United States Grows, So Does Trade Former Venezuelan President Hugo Chavez predicted that “capitalism will lead to the destruction of humanity.” In turn

he worked hard to redirect his nation’s trade away from the United States to what he considered “more like-minded

nations,” such as China and Iran. Washington has also taken steps to limit trade with Venezuela, such as halting

American weapon sales to Venezuela.

The potential gains from trade between Venezuela and the United States, however, are large. For example,

Venezuela is a leading producer of oil that is in high demand in the United States, while U.S. manufacturers produce

automobiles and other products that are in high demand in Venezuela. Meanwhile many Venezuelans place lower value

on an array of products produced by the like-minded nations, such as Chinese cars.

Despite the acrimony between the Venezuelan and U.S. governments, trade between the two countries continues to

soar. Venezuela is the fourth largest oil supplier to the United States, while non-oil exports to the United States

increased 116 percent during the first three months of 2006. Meanwhile, General Motors and Ford have been striving

to meet soaring demand in Venezuela, with automobile sales up over 28 percent between July 2005 and 2006. General

Motors, as Venezuela’s largest car manufacturer, indicated that it planned to invest $20 million to expand its output in

the country by 30 percent, adding 600 new workers.

Trading partners are made better off through exchange. There are strong incentives to engage in trade. This example

illustrates that these incentives are not easily thwarted by political rhetoric. Trade continues to thrive even though

companies and individuals face potential government actions that could affect their trading relationships and

corresponding investments.

Source: S. Romero (2006), “For Venezuela, as Distaste for U.S. Grows So Does Trade,” (August 16).

example, some people value new automobiles more than other people do. Another im- portant source of gains is that the seller may be able to produce the item more cheaply than the buyer and thus has a comparative advantage in its production. In advanced economies, individuals specialize in producing goods where they have a comparative advantage; they then trade to acquire other goods. Specialization greatly enhances the standard of living of a society. Imagine that you had to be completely self-sufficient, making your own clothing, growing your own food, building your own house, and producing your own vehicles for transportation. Your overall standard of living would be dramatically lower than it is living in a modern, specialized economy.

Table 3.1 presents a numerical example of comparative advantage. Donna Meyers and Mario Santini each produce their own food and drink through hunting and brew- ing beer. Panel A shows how many hours it takes for them to produce 1 pound of meat and 1 quart of beer. Panel B shows their allocation of time and resulting output work- ing independently prior to their meeting and trading. Both work 30 hours per week.

A. Time it takes for Donna and Mario to produce meat and beer

Meat (1 lb) Beer (1 quart)

Donna 1 hour 2 hours Mario 6 hours 3 hours

B. Allocation of time (30 hours per week) and output prior to specialization and trading

Meat Beer

Donna 18 hours; 18 lbs 12 hours; 6 quarts Mario 18 hours; 3 lbs 12 hours; 4 quarts Total production 21 lbs 10 quarts

C. Production with specialization

Meat (lbs) Beer (quarts)

Donna 30 0 Mario 0 10 Total production 30 lbs 10 quarts

D. One possible allocation after trading

Meat (lbs) Beer (quarts)

Donna 23 6 Mario 7 4

Table 3.1 Comparative Advantage

This table provides an example of comparative advantage. Panel A shows how many hours it takes for Donna and Mario to produce 1 pound of meat and 1 quart of beer. Donna and Mario each work 30 hours per week. Panel B shows their allocation of time and resulting output prior to meeting and trading. While Mario is less productive than Donna in an absolute sense for both goods, he has a comparative advantage in making beer (opportunity cost of 1⁄2 pound of meat for 1 quart of beer compared to Donna’s opportunity cost of 2 pounds of meat). Donna has a comparative advantage in producing meat. Panel C illustrates how total production can be increased by having both people specialize in the activity where they have a comparative advantage. Panel D displays a possible final allocation after Donna and Mario trade. Specializing and trading produce real gains for both people.

Chapter 3 Exchange and Markets 71

72 Part 1 Basic Concepts

Donna spends 18 hours per week hunting and 12 hours per week making beer, pro- ducing a total of 18 pounds of meat and 6 quarts of beer. Mario spends 18 hours hunt- ing and 12 hours making beer, producing a total of 3 pounds of meat and 4 quarts of beer. Their total production prior to meeting is 21 pounds of meat (18 � 3) and 10 quarts of beer (6 � 4).

Donna has an absolute advantage over Mario in making both goods—it takes her fewer hours to produce either a pound of meat or a quart of beer. Mario, however, has a comparative advantage (lower opportunity cost) for producing beer. Mario’s op- portunity cost for producing 1 quart of beer is 1⁄2 pound of meat (he could have pro- duced 1⁄2 pound of meat with the 3 hours he uses to produce a quart of beer), while Donna’s opportunity cost is 2 pounds of meat. Conversely, Donna has a comparative advantage in hunting. Donna’s opportunity cost for producing 1 pound of meat is 1⁄2 quart of beer, while Mario forgoes 2 quarts of beer to produce a pound of meat.

Panel C shows how total production can be increased by having each person spe- cialize in producing the product for which they have a comparative advantage. Donna can produce 30 pounds of meat by spending all 30 hours on hunting, while Mario can produce 10 quarts of beer by focusing exclusively on beer production. This specialization maintains total beer production at 10 quarts and increases the production of meat by 9 pounds.4 By specializing and trading, both parties can be made better off—there are gains from trade. The final allocation depends on the spe- cific bargain reached by Donna and Mario. One possible outcome is presented in Panel D, where both parties have the same amount of beer as before but more meat. Specializing and trading results in a Pareto improvement relative to working in isolation.


Strategic Business Planning: Ignoring Economics of Trade During the 1970s, many firms adopted a particular form of strategic business planning. The idea behind this process is

to treat the projects of a firm like stocks in a portfolio. Through systematic analysis, winners are to be kept and losers

sold. Specifically, all projects of the firm are ranked based on growth potential and market share. Projects with high

growth potential and high market share are called stars, while projects with low growth potential and market share are

referred to as dogs. Dogs are sold, while stars are kept. Funding for the stars comes from cash cows, projects with high

market share and low growth potential. Thus, money is invested in the winners to enhance the firm’s competitive


Although the idea might sound intriguing, its underpinnings are inconsistent with the basic economics of trade—

sell if, and only if, you can get a price that exceeds the value of keeping the item yourself. This principle implies that,

contrary to the process, dogs should be kept unless they can be sold at sufficiently high prices. Moreover, stars should

be sold if the price is sufficiently high.

By the 1980s, many firms found that violating the basic economics of trade had led them to accumulate suboptimal

collections of projects. Large increases in stock prices were observed as these firms reshuffled plants, divisions, and

subsidiaries through sell-offs, spin-offs, and divestitures.

Source: (1984) “The New Breed of Strategic Planner,” BusinessWeek (September 17), 62–68.

4In this example, Donna and Mario completely specialize and produce only one product. More generally, at

least one of the two people will specialize in producing one product. The other person might allocate some

time to producing the same product (the one for which he does not have a comparative advantage) if

additional gains are derived from producing more of the product than can be produced by the first person.

Figure 3.1 provides a graphical analysis of Mario and Donna’s gains from spe- cialization and trade. A Production Possibilities Frontier (PPF) shows all combina- tions of the two goods that can be produced with efficient production. Recall, with efficient production it is not possible to increase the production of one good without decreasing the production of the other. The figure pictures Mario and Donna’s indi- vidual PPFs, as well as their combined PPF that assumes appropriate specialization. The individual PPF ’s picture Donna and Mario production choices when each work independently with no trade. As drawn, Donna can produce both more beer and more meat than Mario if there is no trade.

The absolute value of the slope of Mario’s PPF, 1⁄2 lb. meat/qt. beer, is his oppor- tunity cost for producing beer. The absolute value of the slope of Donna’s PPF,

350 5 10 15 20 25 30









Beer (quarts)

M ea

t ( lb

s. )

No trade combined total

Com bined production possibilities

curve (PPF)— efficient production

Mario's production possibilities (PPF)

Donna's production

possibilities (PPF)

Donna with specialization

and trade

Mario with specialization

and trade

Example of complete specialization (Depending on their preferences they may want to locate somewhere else on the combined PPF)

No trade

No trade

G ai


G ai


G ai


Figure 3.1 Comparative Advantage

This figure displays the example of comparative advantage presented in Table 3.1 graphically. Donna and Mario’s Production Possibilities Frontiers (PPFs) show the combinations of the two products that each can produce individually, working 30 hours per week with no trade. The absolute value of the slope of Mario’s PPF, 1⁄2 lb. meat/qt. beer, is his opportunity cost for producing beer. The absolute value of the slope of Donna’s PPF, 2 lb. meat/qt. beer, is her opportunity cost for producing beer. Mario’s lower opportunity cost implies that he has a comparative advantage in beer production. The combined PPF shows combinations of meat and beer production that are possible if they divide the work based on comparative advantage. Any point on the combined PPF is efficient in the sense that the output of one good cannot be increased without decreasing the output of the other good, given their productive capacities. The y-intercept of 35 is the maximum meat that can be produced if they produce no beer. Moving from there, the slope of the PPF is initially �1/2, which assumes Mario will produce the beer because of his comparative advantage. The kink in the PPF occurs where Mario reaches his maximum production of 10 beers. Donna must produce any desired beers beyond that point, and she has a higher opportunity. The combined and individual gains from specialization and trade are also pictured.

Chapter 3 Exchange and Markets 73

74 Part 1 Basic Concepts

2 lb. meat/qt. beer, is her opportunity cost for producing beer. Mario’s lower oppor- tunity cost implies that he has a comparative advantage in beer production. The ini- tial production choices with no trade are pictured on the respective PPFs. The com- bined PPF shows combinations of meat and beer production that are possible if they divide the work based on comparative advantage. The y-intercept of 35 is the maxi- mum amount of meat that can be produced jointly if they produce no beer. Moving from there to positive beer production, the slope of the PPF is initially �1⁄2, which assumes Mario is assigned the task of producing beer because of his comparative advantage.

The kink in the PPF occurs where Mario reaches his maximum production of 10 beers for the week. Donna must produce any additional beers beyond that point, and she has a higher opportunity cost. The slope of the PPF from that point on is �2. Ar- rows showing the increased amount of meat that goes to each person (without re- ducing the beer they receive) picture the combined and individual gains from spe- cialization and trade. These gains are based on the assumption of complete specialization and Donna trading Mario 7 lbs. of meat for six beers. Where the two will actually produce on the PPF depends on their preferences for beer and meat. In simple two good examples of this type with linear PPFs, it is never optimal for both people to produce both goods. It, however, can be optimal for one person to be com- pletely specialized and the other person to produce some of both goods, depending on the demand for the two products.

While it is possible to have an absolute advantage in producing all goods, it is im- possible to have a comparative advantage in all activities.5 Specialization and trading are common features in economies throughout the world. Comparative advantage also arises in many management situations. For example, while a top-level manager might be able to perform many activities more effectively and in less time than a lower- level employee, the manager should not try to do all activities himself (make sales calls, work on the manufacturing line, change lightbulbs, answer phones, and so on). More value will be created if managers concentrate on activities for which they have a com- parative advantage.

5Note for the mathematically inclined: Donna’s opportunity cost for producing 1 pound of meat is 1⁄2 quart of

beer (1 qt. beer/hr.)/(2 lb. meat/hr.) � 1⁄2 qt. beer/lb. meat. The reciprocal of this ratio, 2 lb. meat/qt. beer, is

Donna’s opportunity cost for producing beer expressed in pounds of meat. If Donna’s ratio is smaller than

Mario’s ratio for one product, the reciprocal of Mario’s ratio must be smaller than the reciprocal of Donna’s

ratio. Thus Donna has a comparative advantage in producing the first product, while Mario has a comparative

advantage in producing the second product. It is a mathematical impossibility for one person to have a

comparative advantage in producing all products.

Gains from Trade In 1880, the United States was about to become the world’s most efficient economy. Yet labor productivity varied

substantially among states. North Carolina, the least productive state, was only 18 percent of Nevada, the most

productive. (In 1880, Nevada’s productivity was high because many had migrated there to work in the mines.) In 2002,

New Mexico, the poorest state, had a per capita income that was almost 60 percent of Connecticut, the richest state. As

a giant free-trade zone, incomes in the United States have converged to similar standards. Although there are still

differences, those differences have fallen substantially—and not at the expense of the rich states.

Source: V. Postrel (2004), “A Case Study in Free Trade: American Incomes Converge, but Not at the Bottom,” New York Times (February 24), C2.


A common misconception is that trade takes place because people have too much of some goods—people sell to others what they cannot use themselves. This view, however, does not explain why individuals sell houses, cars, jewelry, land, and other resources that they value highly and have in short supply. The economic explanation for trade argues that trade takes place not because people have too lit- tle or too much of a good. Rather, trade takes place because a person is willing to pay a higher price for a good than it is worth to its current owner. While you might love your new sports car, you would still sell it if someone offered you a high- enough price. And winning bidders of collectibles auctioned on eBay are fre- quently individuals with collections of related items.

It is important to recognize that trade is an important form of value creation. Trading produces value that makes individuals better off. Gains from trade also pro- vide important incentives to move resources to more productive uses. If George Nichols can make the most productive use of a piece of land, he will be willing to pay a higher price for the land than other potential users. The current owner, Jody Crowe, has the incentive to sell the land to George, because she gets to keep the pro- ceeds from the sale. It is these incentives that help to promote a Pareto-efficient al- location of resources in a market economy. After all mutually advantageous trades are completed, it is impossible to change this allocation without making someone worse off.

Basics of Supply and Demand Gains from trade explain why individuals buy and sell. But what coordinates the sep- arate decisions of millions of individuals in a market economy to prevent chaos? Why are there not massive surpluses of some goods and huge shortages of other goods? What restricts the amounts demanded by the public to the amounts supplied? Answers to these questions come from an understanding of the market price system.

ANALYZING MANAGERIAL DECISIONS: Comparative Advantage in the Workplace

You are a manager of a division of a company that

is responsible for the final assembly of two com-

puter products, modems and keyboards. You man-

age two employees, Julio and Chenyu, who each

work 8 hours per day. Currently you have assigned

both Julio and Chenyu to spend the first 7 hours of

the day assembling keyboards and the last hour as-

sembling modems. Julio can assemble 2 modems

per hour and 14 keyboards per hour. Chenyu is

more highly skilled in both activities. She averages

3 modems per hour and 15 keyboards per hour.

1. How many modems and keyboards are being

assembled under the current work assign-


2. What are Julio’s opportunity costs for assem-

bling modems and keyboards? What are

Chenyu’s? Does either employee have a

comparative advantage in assembling one

of the products?

3. Devise a way of reassigning the work activi-

ties between the two employees that keeps the

number of modems being assembled the

same as before but increases the number

of keyboards.

4. What are potential reasons why you might not

want to change the work assignments (assume

that more assembly of either or both products

is desirable)?

Chapter 3 Exchange and Markets 75

76 Part 1 Basic Concepts

The Price Mechanism

The basic economics of a price system can be illustrated through standard supply and demand diagrams. Figure 3.2 displays a supply and demand diagram for a particular model of personal computer—for example, a Pentium dual-core machine with standard quality and features. The vertical axis on the graph shows the price for a PC, and the horizontal axis shows the total quantity of PCs demanded and supplied in the market for the period (e.g., a month).

The market includes all potential buyers and sellers of this type of PC. Suppose that in this market there are many buyers and sellers and that individual transactions are so small in relation to the overall market that the price is unaffected by any single sale or purchase. In this case, no buyer or seller has market power: All trades are made at the going market price. We label this type of market as competitive.

The demand curve depicts how many total PCs consumers are willing to buy at each price (holding all other factors constant). The demand curve slopes downward because consumers typically buy more if the price is lower. For example, consumers are likely to buy more PCs if the price is PLO (say, $900) than if the price is PHI (say, $1,500).

The supply curve depicts how many PCs producers are willing to sell at each price (holding all other factors constant). The curve slopes upward: At higher prices, pro- ducers are able and willing to produce and sell more units. For example, at a price of $900, many potential producers cannot cover their costs, and thus they refrain from entering production. At a price of $1,500, more units are manufactured and brought to market.

P ric

e (in

d ol

la rs








P *




Quantity of PCs

Figure 3.2 Supply and Demand in the PC Industry

The demand curve shows the number of PCs that consumers want to purchase at each price. The supply curve shows the number of PCs that producers want to sell at each price. Equilibrium occurs where the two curves intersect. Here, the quantity supplied equals the quantity demanded. If the price is above the market-clearing price of P*, say at PHI, there is a surplus of PCs. Producers supply more PCs than consumers want to purchase, and inventories build. If the price is below the market-clearing price, say at PLO, there is a shortage. Producers supply fewer PCs than consumers want to purchase and inventories shrink. Surpluses and shortages put pressure on prices and quantities to move to equilibrium levels of P*.

The two curves cross at the market-clearing price P* and quantity Q*. At the mar- ket-clearing price, the quantity of PCs demanded exactly equals the quantity sup- plied. Here, at a price of $1,200, the market is said to be in equilibrium.

There are strong pressures within markets that push prices and quantities toward their equilibrium levels. To see why, suppose that the market price is above the equi- librium price, such as PHI in Figure 3.2. At this higher price, there is a surplus of PCs—suppliers produce more PCs than consumers are willing to purchase. As inven- tories of unsold PCs build, this surplus places downward pressure on prices as suppli- ers compete to try to sell their products. As prices fall, fewer PCs will be produced and more will be demanded, thus reducing the surplus. In contrast, if the price is below the market-clearing price, such as P LO in Figure 3.2, inventories dwindle and back orders accumulate—there is a shortage of computers. Here, consumers will bid up the price of PCs as they compete for the limited supply. As prices rise, producers increase their output and consumers demand fewer PCs, thus reducing the shortage. When the mar- ket is in equilibrium, there is no pressure on prices and quantities—the quantity de- manded exactly equals the quantity supplied. Inventories are stable at their desired levels, and the market price is stable at this point.

Supply and demand diagrams like that in Figure 3.2 are snapshots at a point in time. As time passes, both the supply and the demand curves are likely to change. Figure 3.3 shows the effects of a shift in the demand curve in the PC market. The left panel depicts an increase in demand. Here, there is a shift in the demand curve to the right, since at each price, consumers demand more PCs. Demand for PCs might increase for a variety of reasons, including an increase in the purchasing power of consumers or a decline in the prices of supporting software. These types of changes motivate consumers to purchase more PCs at any given price. After the demand shift at the old equilibrium price, inven- tories shrink and there is a shortage of PCs. This shortage places upward pressure on prices; higher prices in turn stimulate more production. The end result is a higher equi- librium price and quantity. The right panel shows that the opposite effect occurs with a reduction in demand. This shift to the left in the demand curve also can be caused by a variety of factors (e.g, a recession that causes businesses to reduce their purchases of PCs or an increase in personal tax rates that reduces consumers’ purchasing power).

Chapter 3 Exchange and Markets 77


Initial demand

Initial demand


P ric

e (in

d ol

la rs



Increase in demand Decrease in demand


S0 S0




Q0 Q0Q1 Q2


Quantity of PCs Quantity of PCs


P1 *

P0 *

* **





Figure 3.3 The Effects of a Shift in Demand on the Equilibrium Price and Quantity of PCs

The initial equilibrium is where the demand curve, labeled D0, intersects the supply curve, labeled S0. The left panel shows the effects of an increase in demand. The result is a higher equilibrium price and quantity. The right panel shows the effects of a decrease in demand. The result is a lower equilibrium price and quantity.

78 Part 1 Basic Concepts

Figure 3.4 depicts the effects of a shift in supply in the PC market. The left panel displays a shift in the supply curve to the right. A rightward shift implies an increase in supply, because at each price producers make and offer more PCs. Many factors might cause an increase in supply. For example, a decline in the prices of labor and other inputs used for manufacturing PCs will make PC production more profitable and increase supply. Supply also might increase because of changes in technology that allow for less expensive, more efficient production. After the supply shift at the old equilibrium price, inventories accumulate and there is a surplus of PCs. This sur- plus places downward pressure on prices, which in turn increases the quantity of PCs demanded. The end result is a lower equilibrium price and higher equilibrium quantity. The right panel shows that the opposite effect occurs when supply shifts to the left.

Initial supply

P ric

e (in

d ol

la rs


Initial supply


P *1


S0 S2

S1 S0


P *0

Increase in supply Decrease in supply

P *2

P *0


$ $

Quantity of PCs Quantity of PCs

Q0 Q0Q1 Q2 * * **

Figure 3.4 The Effects of a Shift in Supply on the Equilibrium Price and Quantity of PCs

The initial equilibrium is where the demand curve, labeled D0, intersects the supply curve, labeled S0. The left panel shows the effects of an increase in supply. The result is a lower equilibrium price and an increase in equilibrium quantity. The right panel shows the effects of a decrease in supply. The result is a higher equilibrium price and a lower equilibrium quantity.


Shifts in Demand, Quantity, and Price at the PGA Tournament The PGA Tournament features competition among the world’s top golfers. In 2013, the PGA was held at Oak Hill

Country Club in Rochester, New York. The event attracted over 30,000 spectators a day. Many of these spectators were

from outside the Rochester area.

A significant number of these visitors were avid golfers who wanted to play while they were in Rochester.

Rochester has several courses that are open to the public. However, many courses in the area are private (only members

and their guests can play). Facing this dramatic temporary increase in the demand for public golf courses, several of

the private courses decided to become public during the week of the PGA. These courses charged high fees ranging

from $150 to $350 per round (their normal guest fees were approximately $75). This example highlights that shifts in

demand motivate increases in the quantity supplied and the price of a product (in this case, golf times).

Chapter 3 Exchange and Markets 79

Shifts in Curves versus Movements along Curves

Demand and supply curves depict the quantities that will be demanded and supplied at each possible price, holding all other relevant variables constant. The price observed in the market determines the specific quantity demanded and quantity supplied in the marketplace (i.e., the price determines the relevant points on the two curves).

We have provided examples of how a change in an important non price variable can cause either the demand and/or supply curve to shift. It is common to refer to a rightward shift in the demand curve as an increase in demand and a leftward shift as a decrease in demand. Similarly, a rightward shift in the supply curve is referred to as an increase in supply (higher quantity is supplied at any given price) and a left- ward shift as a decrease in supply.

A movement along a given demand or supply curve is caused by a change in price (holding other variables constant). The change in price is said to result in a change in the quantity demanded or quantity supplied.

It is important to be able to distinguish between shifts in demand and supply curves (changes in demand or supply) and movements along a given curve (changes in quantity demanded or quantity supplied). For example, reconsider the increase in supply pictured on the left in Figure 3.4. As we have discussed, this increase would have been motivated by a change in some relevant variable other than the price of the product, such as a decline in the hourly wage paid to labor. The rightward shift in supply indicates that producers will supply more of the product at any given price. The demand curve has not changed, so there is no change in demand. However, there is a change in the quantity demanded when supply increases—at the lower equilib- rium price consumers purchase more of the product.

Note that a change in supply or demand is motivated by a change in a relevant variable other than the price of the good. On the other hand, a change in the quantity demanded or quantity supplied is caused by a change in the price of the good, in- duced by a shift in the other curve, holding all other variables that affect the position of the curve constant.

Using Supply and Demand Analysis for Qualitative Forecasts

Consulting firms, large companies, and governmental agencies use formal statistical analysis to develop quantitative estimates of demand and supply to use in analyzing specific markets. Managers, analysts, the media, and others, however, often use sup- ply and demand analysis on a much less formal basis simply to forecast the direction of changes in prices and quantities in the marketplace.

For example, suppose that Mr. Fan owns a restaurant chain in the United States that features a menu of specialty beef dishes. Fan has just read in his morning news- paper that there have been newly reported cases of “Mad Cow Disease” in the United States. Several people who ate contaminated beef died recently from the disease. Fan recalls that Japan, South Korea, and other countries immediately stopped importing American beef when the first case of Mad Cow disease was reported in the United States a number of years ago. Among other things, Fan wants to know what effect the scare will have on the price that he will have to pay for beef over the upcoming months.

Fan can use supply and demand analysis to forecast the directional effect that the Mad Cow scare will have on beef prices. First, he needs to consider whether it will

80 Part 1 Basic Concepts

affect the demand for beef, the supply of beef or both. He is fairly confident that the scare will cause demand to decrease. As pictured in Figure 3.3, a decrease in demand is expected to result in a lower price for beef. However, Fan should not forget supply. For example, what if the government immediately ordered the slaughter and disposal of 90 percent of all cows in the United States to protect consumers from the disease? He needs to take both effects into account, unless he has good reason to assume that one of the effects is not likely to be important.

For practice, suppose that you are Fan and use supply and demand analysis to forecast the directional change in the price beef. Start by drawing a standard supply and demand diagram, as pictured in Figure 3.2 to depict the beef market prior to the scare. Draw the projected shifts in supply and demand caused by the scare. Note that the shifts in both curves have the same directional effect in reducing the equilibrium quantity. The effects on price, however, are in opposite directions. In such cases, you need to ask yourself, which effect is likely to be larger? Sometimes it is hard to fore- cast. Suppose in this case, Fan is relatively confident that the short-term supply of beef will not change very much. The government is unlikely to order the mass de- struction of cows due to a few reported cases of the disease. If so, he might assume that the demand effect is likely to dominate and the near-term effect is likely to be a decline in beef prices. In the longer run, the cost of cattle ranching and beef process- ing could increase due to new government regulations, additional testing for Mad Cow disease, and so on. Thus the longer run effects are harder to forecast. For addi- tional practice, what affect do you think the scare will have on the price of chicken products?

Fan is also likely to be interested in how the scare will affect customer demand for his beef dishes. Demand for products is the subject of the next chapter.

Linear Supply and Demand

Throughout this book we use linear demand and supply curves (as pictured in Fig- ures 3.2 to 3.4). Linearity simplifies the analysis and is often a reasonable approxi- mation in actual applications (at least over the range of actions being considered). This section provides a numerical example of supply and demand analysis using lin- ear supply and demand curves.

Suppose that the supply function for apples is

Qs � 30 � 0.2Pa � 3W (3.1)

where Qs is the quantity supplied of apples in millions of pounds, Pa is the market price for apples in cents, and W is the hourly wage rate paid to agricultural workers. The supply function indicates that farmers will produce more apples as either the market price of apples rises or the wage rate for workers falls. Farmers produce more apples when the wage rate falls since production costs are lower. Currently the wage rate is $10. Substituting this value in Equation (3.1) and solving for Pa produces the current supply curve6:

Pa � 5Qs (3.2)

6Recall that when graphing the supply and demand curves, the convention is to place price on the vertical axis.

Chapter 3 Exchange and Markets 81

Supply curves show the relation between price and quantity supplied holding all other factors constant (in this case the wage rate for agricultural workers).

Suppose that the demand function for apples is

Qd � 20 � 1�3Pa � 0.002I (3.3)

where I � per capita income. The demand function indicates that consumers will purchase more apples as the price falls and/or as income increases. Currently income is $10,000. Substituting this value into the demand function and solving for Pa pro- duces the current demand curve:

Pa � 120 � 3Qd (3.4)

In equilibrium, the quantity supplied equals the quantity demanded: Qs � Qd � Q* where Q* denotes the equilibrium quantity. Substituting Q* into the supply and demand curves (Equations [3.2] and [3.4]) and setting them equal (since there is one equilibrium price) allows us to find the equilibrium quantity, Q*:

5Q* � 120 � 3Q* (3.5)

Q* � 15

The equilibrium price of 75 cents is found by substituting the equilibrium quantity of 15 into either the demand or supply curve Equations ([3.2] or [3.4]).7


Over the past decade, the federal government has

taken significant steps to encourage the develop-

ment of ethanol and other fuels made from plants

as a partial replacement for gasoline. These actions

have been undertaken by politicians in the midst of

public concerns about the dependence on foreign

oil, war in the Middle East, and global warming.

The primary input for ethanol production is corn.

In 2011, the 13.9 billion gallons of ethanol pro-

duced in the United States consumed over 20 per-

cent of the domestic corn supply.

Suppose that the government has just passed

new legislation mandating increased annual pro-

duction of corn ethanol. You manage the Hog

Heaven restaurant chain. Your restaurant chain,

which has about 300 outlets throughout the United

States, specializes in barbecue pork dishes but also

offers chicken, beef, and vegetarian meals. Cur-

rently about 80 percent of your revenue comes

from your pork dishes. The price of pork has a

major impact on your costs. You are concerned

that the federal promotion of ethanol might have

an impact on pork prices and the profitability of

your restaurant chain. Feed cost is typically about

50 to 60 percent of the total cost of production of

pork producers. About 80 percent of the feed that

hogs consume is corn.

1. Use basic supply and demand analysis to illus-

trate the likely effect of the government’s man-

dated increase of ethanol production on (1)

corn prices and (2) pork prices.

2. What actions might you consider given the

results of your analysis?

7We could have found the equilibrium price by setting the demand and supply functions (Equations [3.1] and

[3.3]) equal after substituting for the current values of W and I. The equilibrium quantity then could be found

by substituting the equilibrium price into either Equation (3.1) or (3.3). We took the extra steps of solving for

the demand and supply curves to illustrate how they are derived from the underlying demand and supply

functions. We elaborate on this derivation in the case of the demand curve in the next chapter.

82 Part 1 Basic Concepts

Note that changes in the wage rate shift the supply curve, while changes in income shift the demand curve. See if you can find the new equilibrium price and quantity if income increases to $20,000. Answer the problem before looking in this footnote for the answer.8

Determining the equilibrium in simple numerical supply and demand problems can be summarized as follows. Begin by inserting the current values for variables other than price into the supply and demand functions and solve the functions for P to get the demand and supply curves. Equate the supply and demand curves and solve for the equilibrium quantity, Q*. Put Q* into either the supply or demand curve equations and solve for the equilibrium price, P*.

To consider the effect of a change in a non price variable, replace the original value with the new value and repeat the above steps to obtain the new equilibrium. It is a good idea to graph these kinds of problems on a standard supply and demand diagram, which shows the original equilibrium, as well as the shift in the relevant curve and the new equilibrium. Graphical analysis can often provide a more intuitive understanding of a problem. It also can be helpful as a check for arithmetic errors that can arise in purely algebraic solutions to the problem.

Supply and Demand—Extended Analysis This section uses supply-and-demand framework to analyze other issues of manage- rial interest. It begins by considering whether a change in supply or demand is likely to have a greater impact on the equilibrium price or quantity. This analysis is followed by a related discussion of short-run versus long-run responses to changes in the marketplace. The section ends by considering under what circumstances a per- unit cost increase in an industry can be passed on to consumers through higher prices.

Price versus Quantity Adjustments

We have seen that the equilibrium price and quantity typically change when either the demand or supply curves shifts. Forecasting the direction of price and quantity changes in a market can be very useful to managers. However, it is even more useful to be able to forecast whether most of the impact of the change will be on price or quantity.

To analyze this question, we need to introduce the concept of demand and supply elasticities. Elasticities, which are defined more precisely later in this book, measure the sensitivity of quantity demanded and supplied to price changes.9 The left panel of Figure 3.5 depicts two extreme demand curves—one is vertical and the

8An increase in income in this example shifts the demand curve to the right, resulting in both a higher

equilibrium price and quantity (see Figure 3.2). More specifically, shifting the income from $10,000

to $20,000 results in an equilibrium quantity of 22.5 million pounds and an equilibrium price of $1.125

(112.5 cents). 9The responsiveness of consumption and production decisions to price changes varies across products. For

example, consumers tend to be relatively responsive to price changes when it comes to restaurant meals but

pay little attention to changes in the price of toothpaste. Similarly, the supply decisions of producers can be

greatly affected by price in some cases and vary little in others (in the latter case consider a farmer who has

grown a fixed quantity of a highly perishable commodity that must be sold before rots).

other is horizontal. The right panel displays similarly sloped supply curves. A ver- tical curve indicates that the quantity demanded or quantity supplied is the same regardless of price. Vertical supply and demand curves are referred to as perfectly inelastic. The horizontal curves depict the extreme opposite case where no quantity is demanded at prices above P*, and no quantity is supplied at lower prices. Horizontal supply and demand curves are referred to as perfectly elastic. Typically industry demand curves and supply curves are less extreme—industry demand curves typically slope downward, while industry supply curves typically slope up- ward. At a given price and quantity, demand and supply curves with more vertical slopes are referred to as relatively more inelastic (since the relative change in quantity to a price change is small), while flatter curves are termed relatively more elastic.

In extreme cases, a change in demand or supply will result in only a price or quantity change—the other variable will remain unchanged. Figure 3.6 depicts these cases for an increase in demand. When supply is completely inelastic, an in- crease in demand increases price, but has no effect on quantity. For example, con- sider land in central New York City. Its supply is inelastic and the price for the land is determined by demand. The higher the demand, the higher will be the price. In contrast, when supply is perfectly elastic the increase in demand will cause an increase in quantity but no increase in price. Later in this book, we will discuss why horizontal long-run supply curves are reasonably common in certain types of industries.

P ric

e (in

d ol

la rs


0 0

10987654321 Quantity











P ric

e (in

d ol

la rs


0 0

10987654321 Quantity











Perfectly inelastic supply

Perfectly elastic supply

Perfectly inelastic demand

Perfectly elastic demand

Figure 3.5 Perfectly Inelastic and Elastic Demand and Supply

This figure displays perfectly inelastic and elastic demand and supply curves. When demand (supply) is perfectly inelastic the quantity demanded (supplied) does not change with price. With perfectly elastic demand (supply), a small increase (decrease) in price relative to $5 in this figure will reduce the quantity demanded (supplied) to zero.

Chapter 3 Exchange and Markets 83

84 Part 1 Basic Concepts

Figure 3.7 depicts the two extreme cases for an increase in supply. The effects are similar to what we saw for the increase in demand. An increase in supply will be fully reflected in price when demand is perfectly elastic and fully in quantity when demand is perfectly elastic. In the next chapter, we will examine in more detail the determinants of demand elasticities.

We have focused on increases in demand and supply. The effects are the same but in the opposite direction for decreases in supply and demand. While we have focused our attention on the extremes, the results generalize to more common in-between cases. If the supply curve is relatively inelastic, a change in demand will primarily affect the price of the product. In contrast, if the supply curve is relatively elastic a change in demand will primarily affect the quantity. Similarly, if the demand curve is relatively inelastic a shift in supply will be reflected primarily in price; if the demand curve is relatively elastic it will primarily affect the quantity.

Short-Run versus Long-Run Effects

Supply and demand curves tend to be relatively more inelastic in the short run than the long run. To see why, consider how consumers might respond to a large increase in the price of gasoline. Consumption of gasoline might not change very much in the

100 1 2 3 4 5 6 7 8 9













P ric

e (in

d ol

la rs


Perfectly inelastic supply

Perfectly elastic supply



Increase in demand

Figure 3.6 Increase in Demand with Perfectly Inelastic or Elastic Supply

This figure displays the effect of a demand increase on the equilibrium price and quantity when supply is either perfectly inelastic or elastic. The original equilibrium price and quantity are $5 and 4 units respectively. When demand increases, only the equilibrium price changes when supply is perfectly inelastic (from $5 to $7). With perfectly elastic supply, only the equilibrium quantity changes (from 4 to 6). If demand were to decrease back to D1, the same general effects would happen in reverse.

first weeks after the price increase. People have to get to work, school, and other places. The locations of people’s homes and the types of automobiles they drive are largely fixed in the short run. All of these factors make the short-run demand rela- tively inelastic. With more time to adjust, however, consumers can reduce their con- sumption of gasoline by forming car pools with others to commute to work, pur- chasing vehicles that get better gas mileage, and so on. In the even longer run, they can relocate moving closer to their jobs, change jobs to work at locations closer to home, and so on. A similar analysis holds on the supply side of the market. Supply curves tend to be relatively more inelastic in the short run than the long run because suppliers have more flexibility to make changes over the longer run. Figure 3.8 depicts supply curves and demand curves for the so-called short run, medium run, and long run. Focus on the intersection point of the three curves as the initial starting price and quantity. Notice how a change in price is met with a greater change in quantity in the longer run, since the curves are more elastic.

We previously discussed how a shift in demand or supply will have a greater ef- fect on price when the other curve is relatively inelastic and a greater effect on quan- tity when it is relatively elastic. Because demand and supply curves are likely to be more inelastic in the short run than the long run, shifts in demand and supply will tend to be reflected in price changes in the short run and in quantity changes in the long run.

100 1 2 3 4 5 6 7 8 9 Quantity

Perfectly inelastic demand

Perfectly elastic demand



Increase in supply












P ric

e (in

d ol

la rs


Figure 3.7 Increase in Supply with Perfectly Inelastic or Elastic Demand

This figure displays the effect of a supply increase on the equilibrium price and quantity when demand is either perfectly inelastic or elastic. The original equilibrium price and quantity are $5 and 4 units respectively. When supply increases, only the equilibrium price changes when demand is perfectly inelastic (from $5 to $3). With perfectly elastic demand, only the equilibrium quantity changes (from 4 to 6). If supply were to decrease back to S1, the same general effects would happen in reverse.

Chapter 3 Exchange and Markets 85

86 Part 1 Basic Concepts

Prices communicate important information to consumers and suppliers. For exam- ple, price increases signal to consumers to reduce their consumption of a product and to producers to figure out how to supply more of it. Ultimately, these responses trans- late into less extreme price changes and greater quantity changes in the longer run. In a sense, the large price changes in the short run help to motivate the output and con- sumption changes in the long run.

Industry Cost Increases and Price Adjustments

John MacDonald manages a company in a competitive industry that bottles and sells healthy juices to consumers. The current market price for juices in his industry is $5 per bottle. The government has just announced a new $2 per bottle tax that it is going to impose on suppliers in this industry. John wants to know whether he will be able to pass this cost increase on to consumers, for example, by charging $7 per bot- tle. Since his firm operates in a competitive industry he has no power to set the price. His hope is that the market price will increase to $7 to offset the cost increase. The question is will it? It turns out that the answer depends on the relative elasticities of supply and demand curves in the industry.

Figure 3.9 displays a graphical analysis of this example where the absolute values of the slopes of the demand and supply curves are roughly the same. The $2 per unit


Decrease in Supply of Pilots Results in Reduction of Flights and Small Increases in Pilot Wages In August 2013, the Federal Government increased the minimum experience required for commercial airline pilots

from 250 hours to 1,500 hours. This reduced the supply of entry-level pilots materially. The primary employer of entry-

level commercial pilots are the smaller regional airlines. Before this change, the starting salary paid at 14 U.S regional

carriers averaged about $22,400 per year.

Supply and demand analysis suggests that this decrease in the supply of pilots would either decrease the number of

pilots employed by the airlines, increase pilot wages, or some combination of both. Many of the routes flown by the

regional airlines were only marginally profitable, and the airlines had only limited power to increase ticket prices to

cover the increased costs for hiring pilots. These conditions imply that their demand for pilots would have been

relatively elastic. The supply and demand framework predicts that the primary effect of a decrease in supply with

relatively elastic demand would be a decline in the number of flights by these smaller regional airlines. Consistent

with this forecast, many of the regional airlines cut their number of flights and hired fewer pilots, rather than

raising wages of entry-level pilots. In the first quarter of 2014. Silver Airways announced that it was cutting its flights

by 13 percent. Republic Airways, one of the nation’s largest regional carriers announced due to the limited lumber of

qualified commercial pilots, it was removing 27 of their 243 aircraft from operation. Great Lakes Aviation Ltd. stated

that it was reducing the number of pilots from 300 in 2013 to about 100. However this reduction in the supply of

qualified pilots, also resulted in small increases in pilot wages by 2014. For example, Silver Airways, a Florida-based

airline with 35 planes, offered its current pilots salary increases of 5 to 10 percent and promised a $6,000 bonus if they

continued to work for the company for one year.

Consistent with the economic view of behavior, some regional airlines responded “creatively.” Because the new

federal rules only required 250 hours of experience for commercial pilots flying planes with fewer than 10 seats they

could hire pilots with this lower level of experience by removing 10 seats from a 19-seat airplane. Since these pilots

were more plentiful, they were less expensive to hire.

Source: J. Nicas and S. Carey (2014), “What Can New Pilots Make? Near Minimum Wage” The Wall Street Journal (February 12).

cost increase causes the supply curve to shift upward by $2. Suppliers have to pay $2 per unit to the government and thus require $2 more per unit to induce them to bring any specific quantity to the market, relative to the original supply curve. For example, suppliers were willing to bring 6,000 units to market at a price of $5. Now they require $7 to induce them to produce the same quantity. Note that the decrease in supply (upward/leftward shift in the curve) causes the price to increase to $6. Thus, suppliers are able to pass on half the cost to consumers. John and the other producers collect $6 for each unit sold. However, they only net $4 after paying the tax to the government. Each side of the market is $1 per unit worse off.10

More generally the sharing of a per-unit cost increase depends on the relative elas- ticities of the demand and supply curves. In the previous example, the supply and de- mand elasticities were about the same, so each side of the market bore about one-half the cost. When the two elasticities are not the same, the side of the market with the less elastic curve bears the larger share of the cost increase. Figure 3.10 depicts the two extreme cases for the demand curve. If the demand curve is perfectly inelastic, quantity remains unchanged and the price increases by the full amount of the cost

100 1 2 3 4 5 6 7 8 9 Quantity

Long-run supply

Medium-run supply

Short-run supply

Medium-run demand

Short-run demand

Long-run demand












P ric

e (in

d ol

la rs


Figure 3.8 Short, Medium, and Long Run

This figure displays the typical pattern observed for supply and demand curves in the short, medium, and long run. Both demand and supply tend to be more elastic in the longer run than the short run because consumers and producers have more time to make adjustments in quantities when price changes.

Chapter 3 Exchange and Markets 87

10Aside from paying the increase costs on units transacted, consumers and producers also experience lost

gains from trade due to the reduction in quantity transacted. Their combined loss due to the reduction in

trade is pictured by the deadweight loss (DWL) triangle in the graph. The tax does not only transfer money

from consumers and producers to the government, but also causes a reduction in quantity and gains from

trade. These losses are not offset by gains to others and so are commonly referred to as “deadweight losses.”

88 Part 1 Basic Concepts

increase. Here the entire cost increase is passed on to consumers. In contrast, when demand is perfectly elastic the price remains the same and the quantity adjusts. In this case, none of the cost increase is passed on to consumers. Similarly, it can be shown that if supply is perfectly inelastic, producers bear all the costs. If supply is perfectly elastic all the cost is passed on to consumers. When neither side of the

10 11 12 130 1 2 3 4 5 6 7 8 9 Quantity (000’s)

$2 per unit tax




Borne by consumers Borne by producers













P ric

e (in

d ol

la rs


Figure 3.9 Passing on Part of a Cost Increase to Consumers

This figure displays an example where suppliers face a new $2 per unit tax. The original equilibrium price and quantity were $5 and 6,000, respectively. The tax causes the supply curve to shift upward by $2. The new equilibrium price and quantity are $6 and 4,000 units, respectively. Consumers pay $6, but producers only net $4 after tax. Each side of the market is $1per unit worse off than before the tax. The government collects $8,000 in taxes. The shaded rectangles reflect the sharing of this cost. Each side of the market bears $4,000, so the cost increase is evenly split. The even split is due to the fact that the absolute values of the slopes of the supply and demand curves in this example are equal. More generally, the side of the market with the relatively less elastic curve will bear a greater share of a per unit cost increase. The deadweight loss triangle represents the lost gains from trade due the quantity reduction due to the tax.


Supply of Online Résumés Bogs Down Employers The Internet has reduced significantly the cost of submitting résumés to would-be employers. Job seekers no longer

must print their résumés on high-quality paper, address, stamp, and mail an envelope. A click of the mouse and the

résumé is gone. Some companies have thousands of résumés dumped into their e-mail boxes each day. During 1999

there were almost 5 million résumés on the Internet—200 times more than in 1994. When the cost of producing a good

(like submitting a résumé) falls, its supply increases.

Source: S. Armour (1999), “Online Résumés Bogging Down Employers,” Democrat and Chronicle (July 19), 1F.

10 11 12 130 1 2 3 4 5 6 7 8 9 Quantity (000’s)

Perfectly elastic demand

Perfectly inelastic demand



$2 per unit tax

New equilibrium with elastic


New equilibrium with inelastic


Tax borne by consumers with inelastic demand












P ric

e (in

d ol

la rs


Tax borne by suppliers with Elastic demand

Figure 3.10 Sharing of a Cost Increases at the Extremes

This figure shows the effects of a $2 per unit cost increase with perfectly inelastic and perfectly elastic demand. When demand is perfectly inelastic, the entire cost increase is passed on to consumers through a higher price and quantity does not change. With perfectly elastic demand the entire cost is borne by suppliers (none can be passed on). Price does not change and quantity falls.


Elastic Demand Limits Steelmakers’ Abilities to Pass through Cost Increases Higher prices for inputs, such as iron ore and coking coal, were expected to force up steel prices during 2011. Analysts,

however, predicted that steelmakers would be unlikely to be able to raise steel prices enough to pass on the full increase in

costs. Theory suggests that the ability of producers to pass through cost increases depends on the elasticity of demand.

Analysts argued that weaker demand from the construction industry and the expectation that China might tighten its

monetary policy due to concerns about rising inflation implied that steel makers would have difficulty in maintaining higher

prices throughout the year. A UBS analyst summarized the basic economics of the situation as follows, “While mills can

easily pass through higher input costs when demand is strong, unfortunately the current environment is less supportive.”

Source. D. Maylie (2010), “Steelmakers Grapple with Price-Increase Pressures,” The Wall Street Journal (December 28)

Chapter 3 Exchange and Markets 89

market is at one of the extremes, the cost is split. The side of the market with the rel- atively more elastic curve bears a smaller share of the cost.

Sometimes a cost increase is imposed on consumers, rather than on suppliers. For ex- ample, some states impose taxes on consumers when they purchase bottles of liquor at a store. While we do not do so here, it is relatively easy to show that the sharing of a cost increase does not depend on which side of the market initially incurs it. When a cost in- crease is imposed on consumers, the demand curve shifts downward, while the supply curve stays the same. The resulting price change results in the same net outcome for both parties as when the tax is imposed on suppliers. What matters in both cases are the relative elasticities of the supply and demand curves.

90 Part 1 Basic Concepts

Prices as Social Coordinators The equilibrium of supply and demand highlights the crucial role that prices play in coordinating the consumption and production decisions of individuals. For example, if too few PCs are being produced, inventories will shrink and dealers will raise prices. High-prices signal would-be producers to shift from producing lower-valued products to producing computers. Because property rights are private, individuals reap the reward from redirecting their efforts and therefore have strong incentives to shift production. Higher prices also motivate consumers to reduce the quantity of PCs demanded. The end result is that the quantity demanded equals the quantity supplied. This is what Adam Smith referred to as “the invisible hand.”

Efficient Exchange and Production

If everyone trades in a competitive marketplace and all mutually advantageous trades are completed, the price system results in a Pareto-efficient resource allocation.11

No government intervention or central planning is required. Rather, consumers and producers, acting in their own self-interest, react to price signals in a manner that produces an efficient resource allocation. Prices act to control and coordinate the many individual decisions made in the economy. After trading is completed, the output mix and final distribution of products cannot be changed without making someone worse off.

The basic logic for efficiency in a competitive economy is straightforward. At equilibrium prices, the quantity supplied equals the quantity demanded for all goods and there are no shortages or surpluses. Everyone who wants to make trades has done so, and all gains from trade have been exhausted. In making supply decisions, firms have strong private incentives to adopt the most efficient production methods and the value-maximizing output mix (these production choices maximize their profits). No changes in either production or distribution can be made without making someone worse off.

Measuring the Gains from Trade

In some applications it is useful to have measures of the gains from trade that are in units, such as dollars, that are independent of individuals’ subjective utilities. Con- sumer surplus and producer surplus display this property and are commonly used to measure the gains from trade and the effects of specific actions and events on con- sumers, producers, and society as a whole.12

Figure 3.11 displays supply and demand curves for a market in which the equilib- rium price and quantity are $10 and 10 units. The demand curve indicates that some

11These conditions will be met in a competitive market when trading costs are sufficiently low. Later, we will

discuss factors that can motivate inefficiency in a market economy. 12Technically, consumers must have a specific type of utility function for consumer surplus to be an exact

measure of their gains from trade in a market. However, it is generally a good approximation whenever

consumers allocate their expenditures across many goods, which most do. The measure is widely used and

accepted by most economists.

consumer would be willing to pay $19 to obtain the first unit of the product; how- ever, the consumer only has to pay the market price of $10. The $9 “surplus” is a measure of the consumer’s gains from trade from purchasing the first unit. The sum of the surplus at all points along the demand curve up to the equilibrium quantity of 10 units represents the aggregate difference between what consumers would be will- ing to pay for the product and what they have to pay given the market price. We call this difference, consumer surplus and display it graphically in Figure 3.11 by Triangle A. In this example, the consumer surplus is $50 (1�2 � 10 � 10). Note that in the aggregate, consumers would be willing to pay up to $150 dollars to obtain the 10 units (Triangle A � Triangle B � Triangle C), but only have to pay $100 (Trian- gle B � Triangle C). Their gains from trade from participating in this market are $50.

The same idea is used for measuring the net gains to producers. Later in this book, we show that the area under the supply curve represents the incremental costs that producers incur to produce the output. In this example, the incremental costs are $50 (Triangle C). Producers are willing to supply 10 units as long as the incremental rev- enue is at least equal to the incremental costs of $50. Producers, however, receive $100 (Triangle B � Triangle C). The extra $50 (Triangle B) represents the gains to trade for producers and is called producer surplus.

The total gains from trade produced in the market are measured by the sum of consumer and producer surplus. In this example, the total gains from trade (surplus) are $100 (Triangle A � Triangle B). We use the concepts of producer, consumer, and total surplus in several places in this book to measure the effects of various actions on consumers, producers, and “social welfare.”

Government Intervention

We have discussed how a well-functioning price system can produce an efficient al- location of resources without government intervention or central planning. Nonetheless, governments sometimes intervene to establish caps or floors on




SupplyConsumer surplus (Triangle A)

Producer surplus (Triangle B)

Incremental production costs (Triangle C)

Q 10





Figure 3.11 Consumer and Producer Surplus

In this figure, consumers pay $100 to obtain 10 units of the product ($10 � 10 units). They would have been willing to pay an additional $50 to obtain the 10 units. This $50 (Triangle A) represents the gains from trade to consumers and is called consumer surplus. Producers receive $100 of revenue, but would have been willing to supply the product if they had covered their incremental production costs of $50 (Triangle C). The extra $50 (Triangle B) is the gains from trade to producers and is called producer surplus.

Chapter 3 Exchange and Markets 91

92 Part 1 Basic Concepts

prices. This section examines the economic effects of these actions (in an otherwise well-functioning market).

Price Controls The average retail price for gasoline in the United States was about $1.15 per gallon in 2003. Gasoline prices increased dramatically over the period 2004–2008 due to factors such as the increased demand for oil in China and India, the war in Iraq, and Hurricane Katrina. Gasoline prices, which averaged $1.60 per gallon in 2004, broke the $2.00 per gallon barrier in July of 2005. In August 2005, gasoline supply was dis- rupted by Katrina, the devastating storm that hit the Gulf Coast. Gasoline prices surged to over $3.00 per gallon. In subsequent months gasoline prices were volatile, rising from a low of $2.10 per gallon in November 2005 to a high of $3.20 per gallon in May 2007. In 2008 the price was above $4.00 per gallon. Consumers expressed fear and outrage over the high gasoline prices. Some groups asked the U.S. government to implement price controls to protect consumers from “unfair” gasoline prices.

Figure 3.12 displays the economic effects of a cap on the price of gasoline. The free market equilibrium price is $3.00 per gallon. However, suppose that the government passes a law that does not allow stations to charge more than $2.00 per gallon. At the $2.00 price, the quantity demanded, QD, is greater than the quantity supplied, QS— there is a shortage of gasoline. The excess demand implies that gasoline must be al- located through nonprice mechanisms. One mechanism is to serve customers in the order that they arrive until the supply is exhausted. This mechanism is likely to





Excess demand (shortage) for gasoline

$2 price




Lost gains from trade


Figure 3.12 Economic Effects of a Government Price Cap on Gasoline

The free market equilibrium price in this figure is $3 per gallon. The government does not allow stations to charge more than $2 per gallon. At the $2 price, the quantity demanded is greater than the quantity supplied— there is a shortage of gasoline. The excess demand implies that gasoline must be allocated through nonprice mechanisms, such as waiting in line. Triangles A and B represent the lost gains from trade to consumers and producers (consumer and producer surpluses), respectively, induced by the price cap. Rectangle C represents a transfer of surplus from producers to consumers (ignoring other costs imposed on consumers). The price cap can also distort incentives of consumers to reduce consumption of gasoline, for example, by moving closer to work or buying smaller automobiles.

produce long lines and waits for gasoline. Also, customers who place the highest value on the gasoline (e.g., due to the importance of their travel) do not necessarily re- ceive the product. The quantity supplied falls as a result of the price cap, resulting in lost gains from trade (total surplus). The reduction in consumer surplus and producer surplus is pictured by Triangles A and B, respectively. The consumers who obtain the gasoline for $2.00 potentially benefit from the cap at the expense of the gasoline deal- ers, the gasoline distributors, the gasoline refiners, and the individuals who own the mineral rights and receive a lower price for the quantity sold (unless the gains are off- set by costs such as having to wait in line). Rectangle C pictures the transfer from pro- ducers to consumers (ignoring other costs).

During the 1970s, there was a severe shortage of gasoline in the United States due to an oil embargo and price controls. People still remember the long gas lines and substantial inconveniences experienced during that period. Some gas stations served customers based on their license plate numbers—odd numbers one day and even numbers the next. Customers also were limited in the number of gallons that they could purchase. These inconveniences have rarely been observed since the elimina- tion of price controls in the early 1980s.

Consumers responded to higher gasoline prices during the 2004–2007 period in a variety of ways. Some reduced their travel plans, while others shifted to less expen- sive forms of transportation (e.g., carpooling, buses, and bikes). Some moved closer to work, while others purchased more fuel-efficient automobiles. Price controls not only produce shortages, but also distort incentives to take actions that reduce the consumption of the product.

Price Floors A prominent example of a price floor is a minimum wage law. In the United States, the Fair Labor Standards Act (FLSA) requires employers to pay employees at least a minimum wage (in 2014, $7.25 per hour) for all hours they work. If a state has a min- imum wage that is higher than the federal minimum, employers are obligated to pay the higher rate to employees working in that state. For example, the minimum wage in the State of Washington was $9.32 per hour in early 2014.

Figure 3.13 displays the economic effects of a minimum wage law. The market clearing price for unskilled labor in this example is $6.00 per hour. However, em- ployers are not allowed to pay wages below $7.25 per hour. At $7.25, the quantity supplied of labor, QS, is greater than the quantity demanded, QD —there is unem- ployment. Firms would hire more workers at $6.00 and fewer people would enter the labor market; at $6.00, the quantity supplied equals the quantity demanded, Q*, and there is no unemployment.13

The QD people who are employed at the minimum wage of $7.25 benefit from the law at the expense of their employers who have to pay higher wages (unless the employees incur offsetting costs to obtain and keep their jobs). Rectangle C pictures the transfer of surplus from employers to employed workers. In contrast, the people between QD and Q* are hurt by the law. These people are willing to work for as lit- tle as $6.00 per hour and would obtain jobs in a free market. However, they are un- employed due to the minimum wage law. Overall, there is a reduction in the total gains from trade in the labor market. The lost surpluses for firms and labor are

Chapter 3 Exchange and Markets 93

13In reality, measured unemployment would not be zero absent minimum wage regulation. There are always

going to be individuals changing jobs or searching for better jobs. We abstract from these considerations in

this example.

94 Part 1 Basic Concepts

pictured by Triangles A and B, respectively. These triangles depict the lost surplus from not allowing companies and unemployed workers to enter into mutually beneficial employment relations below the minimum wage. In January 2014, the unemployment rate among teenagers in the United States was 22.6 percent com- pared to 6.8 percent for adults. One potential reason for the high unemployment rate among teenagers is the minimum wage.14

Minimum wage laws can also affect how people are paid. For example, suppose that prior to a new minimum wage law unskilled workers are paid $6.00 per hour plus health benefits. One likely effect of forcing employers to pay higher cash wages is a reduction in health benefits. Employees who prefer the health benefits to higher cash wages are worse off after the law.

14The U.S. government has tried to reduce the effects of the minimum wage law on teenagers by exempting

them from the law for their first 90 days of employment. During this period, teenagers must be paid at least

$4.25 per hour. Note that the amount of unemployment caused by the minimum wage laws depends on the

slopes of the supply and demand curves (see Figure 3.13). Although most economists would agree with the

direction of the effect, there is disagreement as to its magnitude. Many economists believe this effect is large,

but some disagree. For example, D. Card and A. Krueger (1995), Myth and Measurement: The New Economics of the Minimum Wage (Princeton University Press: New Jersey).



Labor demand

Quantity of labor

Labor supply

Minimum wage


W ag

e ($

/h r.





Lost gains from trade (surplus)

Unemployment (excess supply

of labor)



Figure 3.13 Economic Effects of Minimum Wage Laws

This figure displays the supply and demand for unskilled labor. The free market equilibrium is a $6.00 wage rate and Q* people being hired. The government has imposed a minimum wage of $7.25 that results in an excess supply of labor (unemployment). The QD

people who are employed at the minimum wage of $7.25 benefit from the law (unless they incur offsetting costs to obtain and keep their jobs). Rectangle C represents the transfer of surplus from employers to employed workers. In contrast, the people between QD and Q* are hurt by the law. These people are willing to work for as little as $6.00 per hour and would obtain jobs in a free market. However, they are unemployed due to the minimum wage law. Overall there is a reduction in the total gains from trade in the labor market. The lost surpluses from reduced trade for firms and labor are pictured by Triangles A and B, respectively.

Externalities and the Coase Theorem15

Externalities exist when the actions of one party affect the well-being or production possibilities of another party outside an exchange relationship. Externalities can pre- vent a free market from being efficient. If a firm emits pollution into the air, it can adversely affect the welfare of the firm’s neighbors. If the firm does not bear these

15This section draws on R. Coase (1960), “The Problem of Social Cost,” Journal of Law and Economics 3,


The Coase Theorem and the “Fable of the Bees” A prominently discussed case of externalities is the so-called “Fable of the Bees.” Beekeepers provide pollination

services for the surrounding fruit growers, and the growers, in turn, provide nectar for the bees. Many economists

would consider this to be a classic case of externalities. If beekeepers and growers do not receive compensation for the

benefits they bestow on other parties, they will underinvest in their activities (from a social standpoint).

The Coase Theorem suggests that beekeepers and growers can privately negotiate to overcome this externality

problem. This is exactly what is done. Beekeepers and growers often enter into contracts. Fruit growers hire beekeepers

to supply hives of bees for pollination of those trees that give little suitable nectar, while the beekeepers pay growers for

the privilege of “grazing” their bees on high nectar-producing trees. Given these payments, beekeepers and growers

have incentives to consider the effects on the other party when they make their investment decisions. Through this

process, beekeepers and growers can reach efficient levels of investment without help from the government.

Source: S. Cheung (1973), “The Fable of the Bees: An Economic Investigation,” Journal of Law and Economics 16, 11–34.


Chapter 3 Exchange and Markets 95

CBO Estimates Effects of Proposed Increase in the Federal Minimum Wage The Federal Minimum wage was $7.25 per hour in 2014. The President and various members of Congress supported

proposed legislation that would have increased it to $10.10 per hour. They argued that $7.25 was not a “living wage”

and that raising the minimum wage would lift many out of poverty. Opponents argued that by decreasing the number of

jobs and thus increasing unemployment, such an increase would hurt many of the people who it was designed to help.

Economic theory argues that a binding minimum wage will decrease the number of jobs and increase

unemployment. It also predicts that the people who find jobs at the higher rate will potentially be better off (it depends

in part on how costly it was for them to find employment). The Congressional Budget Office (CBO) is a highly

respected agency that conducts bipartisan economic analysis for Congress. In February 2014, the CBO issued a report

projecting the impact of the proposed increase in the minimum wage. Consistent with economic theory, they predicted

that the proposed hike would cause a loss of about 1�2 million jobs. They further predicted that 16.5 million workers

would experience wage increases. About 900,000 of them would be lifted out of poverty.

The higher wages paid to workers will have to be paid by someone—there is no “free lunch.” As discussed in this

chapter, the ultimate payers are not necessarily the companies that are required to pay higher wages. At least some of

this cost increase is likely to be passed to consumers in the form of higher prices. Moreover, many employees currently

hired at the minimum wage are entry-level workers with limited work experience and education. The increase in the

quantity of labor supplied at the higher wage rate is likely to include suburban middle-class teenagers and others who

have some more prior work experience. Given the limited number of jobs openings relative to people seeking work at

the higher wage rate, the prospects of a minority, high-school dropout finding work could be even more bleak.

Source: E. Morath, D. Paletta, and C. Lee (2014), “Wage-Rise Report Sees Fewer Jobs, Less Poverty” The Wall Street Journal (February 20).


96 Part 1 Basic Concepts

costs, it is likely to select an inefficient level of pollution (i.e., to overpollute). In choosing how much to invest in pollution control equipment, the firm’s managers will consider only its own costs and benefits. Efficient investment would require them also to consider costs and benefits imposed on neighbors (the efficient level of investment is where the total marginal costs of additional investment equal the total marginal benefits—not just those incurred privately by the firm).

Economists used to think that externalities surely would prevent a market system from producing an efficient allocation of resources. Government intervention seemed to be required to enhance efficiency. For example, the traditional recom- mendation was to tax firms based on their levels of pollution. Such a tax would give firms incentives to reduce pollution.

In 1960, Nobel Prize winner Ronald Coase presented a compelling argument that exchange in a free market is more powerful in producing efficient results than had been thought previously. As long as property rights can be traded, there is an incen- tive to rearrange these rights to enhance economic efficiency. The often-recom- mended government intervention might be unnecessary and in many cases undesir- able. Suppose that a firm has the legal right to pollute as much as it wants. Its neighbors always can offer to pay the firm to reduce its pollution level. Thus, the firm faces a cost for polluting (if the firm pollutes, there is an opportunity cost of not receiving compensation from its neighbors). The firm will pollute only if the pollution is more valuable to the firm than the costs it imposes on its neighbors. This efficient solution is obtained without a pollution tax. The same level of pollu- tion can occur even if the neighbors have the legal right to stop the firm from emit- ting any pollution as opposed to the firm’s having the legal right to pollute as much as it wants. In this case, the firm can pay its neighbors for the right to pollute. Re- gardless of whether the firm or the neighbors have the legal right, the gains from trade are exhausted when the marginal benefit to the firm of polluting is equal to the sum of the marginal costs imposed on its neighbors plus those that the firm bears.

Coase’s argument convinced most economists that externalities were less of a problem than previously thought. It also implied that the distribution of property (legal) rights might have less of an effect on the ultimate use of resources than it has on the distribution of income—as long as these rights can be exchanged. In our ex- ample, the firm might emit the same amount of pollution regardless of who initially is assigned the property right. However, the party with the property right obtains more wealth (since it is the one receiving or avoiding payments).

Nonetheless, as Coase points out, market exchange will not always solve the problem of externalities. The transactions that are necessary to overcome this

problem are not free: There are contracting costs. These costs in- clude search and information costs, bargaining and decision costs, and drafting, policing, and enforcement costs.16 These costs can pre- vent a preferred outcome from occurring. In our example, the firm might limit its pollution for a payment that is far lower than the col- lective damage imposed on its neighbors. Nonetheless, the costs of bargaining with the firm and the costs of reaching agreement on how the neighbors should split the payment can prevent this mutually ben- eficial agreement from being reached. Generally, the costs of reaching an agreement increase with the number of bargainers. In our example,

16C. Dahlman (1979), “The Problem of Externality,” The Journal of Law and Economics 22, 148–162.

The Coase Theorem The ultimate resource allocation will be efficient, regardless of the initial assignment of property rights, as long as contracting costs are sufficiently low and the property rights are assigned clearly, are well enforced, and can be exchanged readily.

the likelihood of reaching an efficient agreement is highest if the firm only has to bar- gain with a single neighbor who owns all the surrounding property.

It also is important that property rights be clearly assigned, enforced, and ex- changeable. Suppose there were no legal system to enforce property rights. Neigh- bors would be reluctant to pay a firm not to pollute—they do not obtain an enforce- able property right to prevent the firm from polluting. After collecting the payment, the firm could renege on its promise to reduce pollution and the neighbors would have no recourse.

This discussion suggests that market economies will tend to produce an efficient resource allocation whenever property rights are clearly assigned and contracting costs of exchanging them are sufficiently low. When these conditions are met, effi- ciency will occur regardless of the initial distribution of property rights. This general principle is often referred to as the Coase Theorem.

The driving force behind the Coase Theorem is gains from trade: Individuals have incentives to search out and undertake mutually advantageous trades. This principle has important managerial implications. Even if a manager does not have all the prop- erty rights necessary to undertake a particular project, it does not mean that the pro- ject cannot be undertaken. If the proposed project creates enough value, the manager often can acquire the necessary property rights from their current owners. Suppose the Watts Construction Company can create substantial value by developing a shop- ping center on a site that currently is zoned for residential housing. Surrounding property owners might support a change in the zoning requirement, as long as they share in the value creation. Watts might be able to increase this support by offering to develop a new neighborhood park near the shopping mall.

Chapter 3 Exchange and Markets 97


Property Rights Help Make Niger Greener Niger in northern Africa historically has been known for being a barren, dust-choked country with many starving

citizens. Most observers saw no hope in Niger’s continued battle against desertification. The end result would be

increased poverty and starvation.

Recent studies of vegetation patterns, however, show that Niger recently has added millions of new trees and is far

greener than it was 30 years ago. Interestingly some of the vegetation is densest in some of the most densely populated

regions of the country. This finding runs counter to the conventional claim that population growth leads to the loss of

trees and accelerates land degradation.

One important factor for why Niger has become greener is a change in property rights. From colonial times, all

trees in Niger were property of the state. State ownership gave farmers little incentive to grow and protect trees. Trees

were cut down by residents for construction and chopped for firewood with little regard for the environmental costs.

Government foresters were supposed to foster the growth of trees and to protect them from illegal destruction, but there

were not enough foresters for an area nearly twice the size of Texas. Now the government allows individuals to own

trees—there is private rather than public ownership. Farmers make money from their trees by selling branches, pods,

fruits, and bark. Because these sales are more lucrative over time than cutting trees for firewood, the farmers preserve

and protect them.

Niger remains in a fragile position since it is located in a drought-prone area. Recently acquired private property

rights, however, have helped to foster the growth and protection of trees. According to experts, “more trees will help

Niger’s people to withstand whatever changes the climate might bring.” Observers note that the improved situation in

Niger was accomplished without having to spend a “lot of money.”

Source: L. Polgreen (2007), “In Niger, Trees and Crops Turn Back the Desert,” (February 11).

98 Part 1 Basic Concepts

The Coase Theorem also suggests that contracting costs are central to the study of organizations. In the absence of contracting costs, efficient outcomes will occur in- dependent of the way decision rights are assigned. From an efficiency standpoint, it does not matter whether decision rights are centralized or decentralized. It is con- tracting costs that make these organizational considerations important. We elaborate on this issue in the section that follows.

Markets versus Central Planning History suggests that the price system is more efficient at controlling and coordinat- ing production and consumption decisions in large economies than is central plan- ning. Without the aid of government planners, market economies have produced products that are highly valued by consumers while avoiding large shortages or sur- pluses. In planned economies such as the former Soviet Union, shortages, surpluses, and other production mistakes are common.

There are at least two reasons why markets have been more successful than central planning in large economies. First, the price system motivates better use of knowledge and information in economic decisions. Second, it provides stronger incentives for in- dividuals to make productive decisions. As we will discuss later, an understanding of these advantages can be useful for managers in making firm-level decisions, such as what decision-making authority to delegate to employees and whether to make or buy a firm’s inputs.

General versus Specific Knowledge17

Figure 3.14 shows how the costs of transferring knowledge can be displayed on a continuum. At one end of this continuum is general knowledge. General knowledge is virtually free to transfer. Examples of general knowledge are prices and quanti- ties—a storekeeper easily can tell you that the price of sugar is $1 per pound. As the costs of information transfer increase, the information is said to become more specific. We use the term specific knowledge to denote knowledge that is relatively high on this scale: It is expensive to transfer.

Information transfer costs

General knowledge Specific knowledge

Figure 3.14 The Cost of Transferring Knowledge

The costs of transferring knowledge can be displayed on a continuum. At one end is general knowledge, which essentially is free to transfer. As the costs of information transfer increase, the information is said to become more specific. We use the term specific knowledge to denote knowledge that is relatively expensive to transfer.

17This section draws on M. Jensen and W. Meckling (1995), “Specific and General Knowledge, and

Organizational Structure,” Journal of Applied Corporate Finance 8:2, 4–18.

At least three factors influence the costs of transferring information. First are the characteristics of the sender and receiver. Generally, it is less expensive for people of similar training, language, and culture to communicate than for people from differ- ent backgrounds. Second is the technology available for communication. For exam- ple, the development of electronic mail (e-mail) has lowered the costs of transferring information. Third is the nature of the knowledge itself. Some knowledge is difficult to summarize, comprehend, or transfer in a timely fashion. Depending on the exact setting, the following types of knowledge often are specific in nature:

• Idiosyncratic knowledge of particular circumstances. The employee on the spot is most likely to know if a particular truck has room for additional cargo or if a certain customer wants to purchase a specific product. If this information is not used immediately, it may become useless. For example, by the time the in- formation about the truck is transferred to another person (such as a central scheduler), the opportunity to load the truck with additional cargo can be lost (for instance, if the truck has departed).

• Scientific knowledge. Knowledge of how recombinant DNA works is not eas- ily transferred to nonscientists.

• Assembled knowledge. An accountant who has completed a client’s tax re- turns for several years is likely to have assembled important knowledge about the relevant parts of the tax code and the idiosyncrasies of the individual’s


The Dynamic Nature of Specific Knowledge Historically, economies of scale have motivated firms in retailing to concentrate on standardized production and

distribution. Knowledge about the idiosyncratic demands of people in particular neighborhoods tended to be ignored in

stocking individual stores within a large retail chain: The information simply was too expensive to collect and process.

This limited their ability to compete with small local stores that catered to the specific demands of local customers.

But the development of computers and electronic scanners has made information about idiosyncratic demands of

individuals less specific. As a result, retail companies have begun to engage in more micromarketing. For instance, the

Sears outlet in the North Hollywood section of Los Angeles is tailor-made to suit the neighborhood’s Hispanic

population. Signs are in Spanish. The store is stocked with ethnic items, such as a broad selection of compact discs and

tapes by Latin American artists. A few hundred miles to the north, the Sears store in San Jose offers a large number of

clothing items in extra-small sizes to attract the area’s Asian population. On the other hand, Sears stores in Florida

carry large, roomy clothes that appeal to the large population of elderly residents.

Source: (1995) “Customers on Target,” Financial Times (August 18).

Chapter 3 Exchange and Markets 99


Topic-Specific Search Engines is competing with Yahoo and Google—and winning. Its search engine for engineers has 3.5 million

users and adds 20,000 more each week. “They own that market,” says Charlene Li of Forrester. GlobalSpec has a well-

defined customer base and detailed understanding of its users; this sets it apart from the generalist search engines.

These features allow its “vertical site” to provide search results from a select group of topic-specific Web sites and

precisely target advertising at particular audiences.

Source: (2007) “Specific Knowledge about Your Customer,” The Economist (July 14), 75.

100 Part 1 Basic Concepts

income and deductions. Another example is learning to operate a complex ma- chine. In neither case is this information easily transferred to others.

Specific knowledge is critical in properly allocating resources. Many economic opportunities are short-lived and must be acted on quickly by the person on the spot (who has the specific information of the opportunity) or lost. Not incorporating the proper scientific or assembled knowledge into economic decisions can have costly implications. For an economic system to be successful, it must promote the use of relevant specific knowledge in economic decisions.

Knowledge Creation

It is important to realize that knowledge is dynamic. There are at least two factors that can motivate changes in the costs of transferring knowledge. The first is tech- nology: Improved communications and computer technology have greatly lowered the costs of transferring certain types of information, making it more general. Sec- ond, individuals can take actions to convert specific knowledge to more general knowledge, for example, by drafting an operating manual.

Nonaka and Takeuchi argue that converting hunches, perceptions, mental mod- els, beliefs, experiences, and other types of specific knowledge into a form that can be communicated and transmitted in formal and systematic language is a key aspect of successful new product innovation.18 As one example, consider Matsushita’s development of an automated fresh bread maker in the 1980s. Specific knowledge of how to knead dough to produce tasty bread was held by master bakers. This knowledge was not easily transferred to others, and past attempts to produce fully automated bread makers had failed because they produced poor-quality bread. Yet specific knowledge about how to manufacture automated bread machines was held by engineers. To produce a successful bread machine, relevant specific knowledge had to be transferred between bakers and engineers. To accomplish this transfer, managers from Matsushita took bread-making lessons from a master baker at an Osaka hotel. Eventually, the managers discovered that the key to good bread mak- ing is to twist and stretch the dough during the kneading process. This concept was general knowledge that could be passed along to design engineers. Matsushita’s “Home Baker” was the first fully automatic bread-making machine for home use and has become a quite successful product.

In 1778, economist–clergyman Thomas Malthus predicted that population would grow more rapidly than the food supply resulting in mass starvation.19 His argument was straightforward: Because land and other natural resources are finite, the growth in population (fueled by the “passion of the sexes”) would eventually exceed the available food supply. But this prediction—which prompted economics to be labeled the “dismal science”—has not come to pass. For instance, population in the United States increased from 76 million in 1900 to 296 million in 2005 (an increase of nearly 300 percent), yet the amount of land and the number of workers devoted to

18I. Nonaka and H. Takeuchi (1995), The Knowledge-Creating Company (Oxford University Press: New

York). 19T. Malthus (1778), “An Essay on the Principle of Population” (printed for J. Johnson, in St. Paul’s Church

Yard, London).

Chapter 3 Exchange and Markets 101

agricultural production over the same period fell dramatically. Today less than 3 per- cent of the U.S. population works in agriculture, while the per capita food supply is at an all-time high. And this increase in the food supply is not just a U.S. phenomenon. From 1951 to 1992, world food production per capita increased 34 percent.20 So why was Malthus wrong?

Malthus, as well as more recent pessimists, has underestimated the importance of improvements in technology—figuring out better ways to use existing re- sources.21 For example, today’s computers are far more powerful than those of a decade ago, yet they take fewer resources to produce. Moreover, computer de- signers have discovered ways to make the materials used in computers more re- cyclable. This process of discovering better ways to use existing resources occurs not only in manufacturing but also in service-related industries as well. For ex- ample, consider the implications of McDonald’s innovations in the 1950s in the delivery of “fast food.”

Economist Paul Romer argues that the opportunities for this type of discovery and growth essentially are unlimited.22 People are constantly taking ideas and knowledge that are in their “wetware” (brains) and converting them into “software” (recipes and formulas) that can be employed to produce new products and services. Matsushita’s conversion of the specific knowledge held by the master baker into more general knowledge that could be used by engineers is a good example. As we will discuss in Chapter 8, good managers understand this mechanism for creating knowledge and value and foster it within their firms.

20J. Perloff (2001), Microeconomics (Addison Wesley: Boston), 154. 21For a more recent example of concern about the implications of finite resources, see D. Meadows New York

(1977), Limits to Growth: A Report for the Club of Rome’s Project on the Predicament of Mankind (New

York American Library). 22For nontechnical discussions of Romer’s theory of economic growth, see P. Romer (1993), “Economic

Growth,” Fortune Encyclopedia of Economics (Time Warner Books: New York); P. Romer (1995), “Beyond

the Knowledge Worker,” World Link, (January/February), 55–60.


Converting IT Wetware into Software Charles Belford, president of the Canadian-based management consulting firm Managements Smarts, Inc., advises

clients to get more value out of their existing information technology (IT) without buying new software or hardware.

He argues that firms should revise their current delegation of authority, governance, and planning and management

practices to identify low-cost, low-risk IT enhancements. Managers should be given incentives to ensure that successful

pilot tests are properly identified and then disseminated throughout the organization. He encourages clients to

reorganize so that a senior manager has enterprise-wide oversight to exploit IT packages to ensure their full potential

for the company. For an example, Mr. Belford says, “By revising the current delegation of authority for managing Web

site content in your company, you may be able to turn your obese Web site or your bulimic internal intranet network

into healthy assets that actually serve their respective constituencies.”

Mr. Belford maintains that the only way to convert IT wetware (successful local IT applications) into recipes or

software and then leverage this newly created software throughout the firm is by changing the company’s

organizational architecture.

Source: C. Belford, “Add Value to Tech Assets without Breaking the Bank,” The Globe and Mail (March 28, 2002), B16.

102 Part 1 Basic Concepts

Specific Knowledge and the Economic System23

Nobel Prize–winner Friedrich Hayek offered a compelling argument that market economies are more likely than centrally planned economies to incorporate relevant specific knowledge in economic decision making. He argued that the relevant spe- cific knowledge for economic decision making is not given to any one individual; in- stead, it is distributed among many people in the economy. This knowledge, by def- inition, does not lend itself to statistical aggregation; it is costly to transfer. A central planner invariably lacks the mental or computing ability to process large volumes of this sort of information. Hayek thus concluded that central planners often ignore im- portant specific knowledge in economic decisions.

In contrast, economic decisions in a market system are decentralized to individu- als who are likely to have the relevant specific knowledge. Technical and marketing geniuses, like William Gates at Microsoft and Michael Dell at Dell Computer, are free to start new businesses and to market products of their choosing. The information that motivates these decisions does not have to be transferred to some central office in Washington where centralized production decisions are made. Thus, the information is more likely to be used effectively.

The activities of decentralized decision makers are coordinated by prices. For in- stance, an increase in market-determined wage rates (the price of labor) signals to pro- ducers that labor is in short supply and should be conserved. Higher wages, in turn, motivate producers to conserve labor. An important advantage of the price system that is stressed by Hayek is that prices economize on the costs of transferring information to coordinate decisions. Companies normally do not have to know all the details of why labor costs have increased. The simple fact that wages have increased tells them most of the things they need to know to make value-maximizing decisions.24

Incentives in Markets

Private property rights are critical for making a market economy work because they provide strong incentives for decentralized decision makers to act on their specific information—the wealth effects of economic decisions are borne directly by the re- source owners. If Alice Chan owns a piece of property, she has incentives to use the land productively because she gets to keep any profits. If Jamal Hammoud can make more productive use of the land, Alice will sell the land to Jamal (there are gains from trade). Property rights are rearranged so that decision rights over re- sources are linked with the relevant specific knowledge.

In contrast, decision makers in centrally planned economies have limited incentives to make productive use of information (even if they have it) since they do not own the resources under their control. Further, lower-level bureaucrats have lim- ited incentives to carry out decisions made by the central authority. The best use of a particular automobile might be to transport tourists from a local airport. A central planner, however, might give the car to his brother because he is more concerned about making his brother happy than about making the economy more productive. After all, he does not keep the profits from transporting tourists—they go to the state.

23This section draws on F. Hayek (1945), “The Use of Knowledge in Society,” American Economic Review 35,

519–530. 24Producers also might want to know the expected future prices of labor. For instance, if the price increase is

expected to be transitory, the company might want to avoid making layoffs.

Contracting Costs and Existence of Firms Hayek’s argument suggests that markets are better than central planning. Why, then, is so much activity conducted within firms, where resource allocation decisions are made by managers in a manner that is closely akin to central planning?25

Conceptually, firms do not have to exist. All production and exchange could be car- ried out through market transactions. In the case of the PC, each consumer could buy all the parts that make up the PC in separate market transactions and then pay some- one to assemble them. In reality, of course, most computers are made by firms and only the final products are sold to the consumer.

Ronald Coase provides an answer to the question as to why resources are allocated by both markets and firms.26 His basic argument is that economic transac- tions involve contracting costs, including search and information costs, bargaining and decision costs, and policing and enforcement costs. There is also an opportunity

ANALYZING MANAGERIAL DECISIONS: Nobel Prize–Winner F. A. Hayek on the “Miracle” of the Price System

It is worth contemplating for a moment a very sim-

ple and commonplace instance of the action of the

price system to see what precisely it accomplishes.

Assume that somewhere in the world a new oppor-

tunity for the use of some raw material, say, tin, has

arisen, or that one of the sources of supply of tin

has been eliminated. It does not matter for our pur-

pose—and it is significant that it does not matter—

which of these two causes has made tin more

scarce. All that the users of tin need to know is that

some of the tin they used to consume is now more

profitably employed elsewhere and that, in conse-

quence, they must economize tin. There is no need

for the great majority of them even to know where

the more urgent need has arisen, or in favor of what

other needs they ought to husband the supply. If

only some of them know directly of the new de-

mand and switch resources over to it, and if the

people who are aware of the new gap thus created

in turn fill it from still other sources, the effect will

rapidly spread throughout the entire economic sys-

tem. This influences not only all the uses of tin but

also those of its substitutes and the substitutes of

these substitutes, the supply of all things made of

tin, and their substitutes, and so on. All this takes

place without the great majority of those instru-

mental in bringing about these substitutions know-

ing anything at all about the original cause of these

changes. The whole acts as one market, not be-

cause any of its members surveys the whole field,

but because their limited individual fields of vision

sufficiently overlap so that through many interme-

diaries the relevant information is communicated to

all. The mere fact that there is one price for any

commodity—or rather that local prices are con-

nected in a manner determined by the cost of trans-

port, etc.—brings about the solution which (if con-

ceptually possible) might have been arrived at by

one single mind possessing all the information

which is in fact dispersed among all the people in-

volved in the process.

Some people (e.g., Hayek) argue that decentral-

ization of economic decisions in the economy leads

to an efficient resource allocation. What differences

exist within the firm that make the link between de-

centralization and efficiency less clear?

Source: F. Hayek (1945), “The Use of Knowledge in Society,” American Economic Review 35, 519–530.

25Within a firm, resources often are transferred from one division to another by an administrative order from

management. For example, managers often are transferred among divisions by administrative decisions.

Prices are not used to make these decisions—the divisions typically do not bid for the managers. 26R. Coase (1937), Economica, “The Nature of the Firm,” New Series, IV, 386–405.

Chapter 3 Exchange and Markets 103

104 Part 1 Basic Concepts

cost if the transaction results in an inefficient resource allocation (we discuss this in detail in Chapter 10). The optimal method of organizing a given economic transac- tion is the one that minimizes contracting costs.27 In some cases, the method will be market exchange. In other cases, the method will involve firms.

Contracting Costs in Markets

A primary set of costs of using markets for exchange involves the discovery and ne- gotiation of prices.28 For example, firms have the following two potential advan- tages:

• Fewer transactions. If there are N customers and M factors of production, a firm can hire the M factors and sell to the N customers. The total transactions are N � M. In contrast, if each customer contracts separately with each factor of production, there are N � M transactions. For example, 10 workers might be required to assemble a computer. If there are 1,000 customers and each cus- tomer negotiates with each worker, there are a total of 10,000 transactions. If a firm hires the 10 workers and sells computers to the 1,000 customers, there are 1,010 transactions.

• Informational specialization. Think of buying a PC. How much do you know about buying each separate part? PC producers, on the other hand, specialize in this knowledge. The consumer buying from a firm only has to be concerned with the quality of the end product.

In Chapter 19, we shall elaborate on one particularly important set of contract- ing costs that motivates the existence of firms, those associated with specific assets. Assets are specific when they are worth more in their current use than in alternative uses. An extreme example is a machine that is used to produce parts that can be used only by one particular producer. The machine is valuable in producing parts for the particular buyer but is essentially worthless in alternative uses. In this case, independent suppliers are reluctant to purchase the machine since they do not want to be at the mercy of a single buyer. For instance, suppliers might worry that the buyer will try to force a reduction in future prices, make unreasonable quality or quantity demands, or curtail purchases. It is these concerns that make simple mar- ket transactions between buyers and sellers unlikely when the relevant assets are highly specific. A potential response to this problem is for the producer to own the machine and make the input parts within a single larger firm.

Another potential advantage of firms is that in some cases they can reduce contracting costs through established reputations. Individuals are likely to have con- fidence in trading with parties who are expected to continue to participate in the marketplace over a long time. They understand that these parties have incentives to be honorable in order to enhance their reputation and future business opportunities.

27It is not always possible to separate contracting costs from the basic costs of production. The optimal method

of production can depend on the way the transaction is organized. Therefore, it is more precise to say that

the optimal method of organization is the one that minimizes total costs (production and contracting costs).

The basic arguments are easier to explain if we focus on contracting costs. 28Economists generally agree that contracting costs motivate the existence of firms. There is disagreement

concerning which contracting costs are most important. Our intent in this chapter is to give the reader a

general sampling of the kinds of costs that can be important.

Organizations tend to have longer lives than individuals and thus might be expected to be more likely to honor agreements than unknown individuals (some major cor- porations date back to the nineteenth century). This increased trust can motivate lower expenditures on negotiating and policing agreements.

Government regulation also helps explain the existence of some firms. Sometimes firms can produce more cheaply because they avoid taxes at intermediate stages of production compared to market transactions.

Contracting Costs within Firms

We have discussed several contracting costs that can motivate the existence of firms. Given these costs, why isn’t the economy just one big firm? The answer is that re- source allocation by firms also involves contracting costs. For example, as firms be- come larger, it becomes increasingly difficult for managers to make efficient and timely decisions. They are more likely to make errors and to be less responsive to changing circumstances. As a firm grows, important decisions must be delegated to employees who are not owners of the firm, thereby generating costs to motivate these nonowners to work in the interests of the owners.

Herbert Simon on Organizations and Markets The United States often is referred to as a market economy. In reality, much of the economic activity in the United

States, as well as in other market economies, is conducted within firms. To quote Herbert Simon, a former Nobel Prize


Suppose a visitor from Mars approaches the earth from space, equipped with a telescope that reveals social structures. The firms reveal themselves, say, as solid green areas with faint interior contours marking out divisions and departments. Market contracting costs show as red lines connecting firms, forming a network in the spaces between them. Within the firms the approaching visitor also sees pale blue lines, the lines of authority connecting bosses with various levels of workers. . . . No matter whether the visitor approached the United States or the Soviet Union, urban China or the European Community, the greater part of the space below would be within the green areas, for almost all the inhabitants would be employees, within firm boundaries. Organizations would be the dominant feature on the landscape. A message sent back home, describing the scene, would speak of “large green areas interconnected by red lines.” It would not likely speak of a “network of red lines connecting green spots.”

Source: H. Simon (1991), “Organizations and Markets,” Journal of Economic Perspectives 5, 25–44.


Chapter 3 Exchange and Markets 105

Corporate Focus and Stock Returns Ronald Coase argues that the use of markets involves contracting costs and that sometimes these costs can be reduced

by including transactions within firms. However, firms also involve contracting costs. In the 1990s, many companies

concluded that they had become too large and diversified. These companies, in turn, decided to refocus on their core

businesses and to shed unrelated activities (e.g., through asset sales). Evidence suggests that on average, these firms

increased their stock market values by increasing their focus on core activities.

Source: R. Comment and G. Jarrell (1995), “Corporate Focus and Stock Returns,” Journal of Financial Economics 37, 67–87.


106 Part 1 Basic Concepts

Efficient Organization Individuals involved in trade and production have incentives to implement cost-re- ducing methods of organization because there are greater gains to be shared. For ex- ample, at a given price, more profits can be generated if costs are reduced.29 In com- petitive markets, individuals will constantly search for new and better ways to reduce costs to improve their competitive advantage and profits. The bottom line is that firms will be used to organize economic activities whenever their cost is lower than that of using markets, and vice versa. Also, as we will see, this same process has im- portant implications for the internal design of organizations.

Managerial Objectives Our discussion to this point has treated decision makers within firms as owners. Owners have a strong interest in increasing the profits of the firm, since they get to keep the proceeds. In public corporations managers are rarely major owners of the firm. Nonetheless, in Chapters 4 to 9 the book, we assume that managers strive to maximize firm profits—or more precisely firm value, the present value of the firm’s profit stream: They make input, output, and pricing decisions with value maximization as their sole objective.30 This perspective is a reasonable starting point because if firms fail to make profits over time, they cease to exist. Most managers are under constant pressure to create value. There also are other mechanisms, such as incentive compensation, that can be structured to align the interests of managers and owners. These mechanisms help make profit maxi- mization a reasonable first approximation of the managers’ objective function. Profit maximization is the basic premise used in most economics textbooks. Start- ing in Chapter 10, however, we shall present a richer characterization of the firm and analyze management/owner conflicts in greater detail. The appendix to this chapter provides a more detailed discussion of managerial objectives, focusing on the topics of shareholder value and stock-market efficiency.

Managerial Decisions We began this chapter with an overview of how market economies operate. An under- standing of this topic is critical if managers are to make productive economic decisions. It is important to understand how a shift in either supply or demand affects

29A firm’s profit (II) is the difference between its total revenues (TR) and total costs (TC): II � TR � TC. If

a company has sales of $1 million and costs of $750,000, it earns a profit of $250,000. 30Much of our basic analysis focuses on maximizing profit in a single period. This approach yields useful

managerial insights without overly complicating the analysis.

CEO Turnover and Firm Profits A standard assumption in microeconomics is that managers strive to maximize profits. One reason that managers are

likely to be concerned about profits is that poor profits and stock price performance increase the likelihood that they

will be fired. For instance, research suggests that the worst performing firms are about 1.5 times as likely to have a

management change as the best performers.

Source: For a review of this evidence, see J. Brickley (2003), “Empirical Research on CEO Turnover and Firm Performance: A Dis-

cussion,” Journal of Accounting and Economics 36, 227–233.


product prices. (In Chapters 4 to 9, we shall extend this analysis and examine in more detail how managers might make optimal input, output, and pricing decisions.)

We also discussed the role of knowledge and incentives in determining the ef- fectiveness of alternative economic systems and the importance of contracting costs in determining whether or not economic transactions are conducted within

Firms versus Markets: When Markets Ruled Economic theory argues that activities are organized within firms when the cost is lower than using markets, and vice

versa. Today, much of the economic activity in the world is conducted within firms. It is hard to envision a world where

large firms do not play an important role in the production and distribution of products. The importance of firms,

however, is a relatively recent phenomenon. Prior to the middle of the nineteenth century, there were virtually no large

firms. Most production was conducted by small, owner-managed operations. The activities of these operations were

coordinated almost entirely through market transactions and prices. To quote Alfred Chandler in describing business

organization before 1850,

The traditional American business was a single-unit business enterprise. In such an enterprise an individual or a small number of owners operated a shop, factory, bank, or transportation line out of a single office. Normally this type of firm handled only a single economic function, dealt in a single product line, and operated in one geographic area. Before the rise of the modern firm, the activities of one of these small, personally owned and managed enterprises were coordinated and monitored by market and price mechanisms.

The large firm became feasible only with the development of improved energy sources, transportation, communications,

and legal/court systems. Coal-fired steam power generators provided a source of energy that made it possible for the

factory to replace artisans and small mill owners, and railroads enabled firms to ship production in large quantities to

newly emerging urban centers. The telegraph allowed firms to coordinate activities of workers over larger geographic

areas. These developments tended to make it less expensive to coordinate production and distribution using

administrative controls, rather than to rely on numerous market transactions among all the intermediaries in the system.

Source: A. Chandler (1977), The Visible Hand: The Managerial Revolution in American Business (Harvard University Press:

Cambridge, MA).


ANALYZING MANAGERIAL DECISIONS: Labor Unions and a Proposed Tax on Luxury Goods

You work as a stock analyst. You specialize in ana-

lyzing markets for “high-end” manufactured prod-

ucts. The President has recommended that Con-

gress impose a tax of $20,000 on every person who

purchases a new yacht. He argues that rich people

who can afford to purchase yachts (they cost

around $2,000,000 each) should pay higher taxes.

The money raised would be used to provide addi-

tional benefits to returning war veterans.

You are writing a research report on the firms

that manufacture yachts. You want to include a

statement in the report about whether you think the

proposal is likely to pass Congress and become

law. You have analyzed the proposal and the voting

records of members of Congress. You have decided

that the proposal will not obtain enough votes to

pass Congress, unless the major labor unions rep-

resenting employees in this industry support it. You

know that many of the yacht-producing firms hire

veterans who could benefit from the programs

funded by the tax.

a. Draw a standard supply and demand diagram

depicting the market for yachts prior to the tax.

b. Analyze the effects of imposing a $25,000 per

yacht tax on consumers (on a qualitative basis).

c. Based on this analysis, do you think that the

labor unions will support the proposed tax?

Explain. What does it depend upon?

Chapter 3 Exchange and Markets 107

108 Part 1 Basic Concepts

markets or organizations. Although we have focused our discussion at the economic-system level, these issues are directly relevant to understanding firm- level decisions on organizational architecture. If firms are to be productive, they must be structured in ways that promote the use of the relevant specific knowledge and economize on the costs of organization. They also must establish appropriate incentives, so that their employees act in a productive manner. Starting in Chapter 10, we shall extend the concepts introduced in this chapter to questions of organi- zational architecture.

Summary There are many different ways of organizing economic activities. Economists focus on Pareto efficiency in evaluating the effectiveness of alternative economic systems. An allocation is Pareto-efficient if there is no alternative that keeps all individuals at least as well off but makes at least one person better off. Pareto-improving changes in a resource allocation are viewed as welfare-increasing.

An important feature of a market economy is the use of private property rights. A property right is a legally enforced right to select the uses of an economic good. A property right is private when it is assigned to a specific person. Private property rights are alienable in that they can be transferred (sold or gifted) to other individuals.

In free markets, property rights frequently are exchanged. Trade occurs because it is mutually advantageous. The buyer values the good more than the seller, and there are gains from trade. Trade is an important form of value creation. Trading produces value that makes individuals better off. Gains from trade also motivate the movement of resources to more productive users. Total output and standards of living often in- crease when individuals specialize in production activities for which they have a comparative advantage (lower opportunity costs).

Prices coordinate the individual actions in a market economy. If too little of a good is being produced, inventories will shrink, prices will rise, and producers will have incentives to increase output to exploit the profit opportunity. If too much of a good is being produced, prices will fall, inventories will build, and producers will have incentives to cut production. The market is in equilibrium when the quantity supplied of a product equals the quantity demanded. There are strong pressures in competitive economies that move the market toward equilibrium. In equilibrium, there are no shortages or surpluses and inventories are stable at their desired levels. Equilibrium prices and quantities change with changes in the supply and demand for products.

Consumer surplus and producer surplus are measures of the gains from trade to consumers and producers from participating in a market. Government-imposed price caps or floors result in market imbalances and lost surplus (in an otherwise well- functioning market).

Externalities exist when the actions of one party affect the consumption or pro- duction possibilities of another party outside an exchange relationship. Externalities can cause markets to fail to produce an efficient resource allocation. Competitive markets will produce a Pareto-efficient allocation of resources if the costs of making mutually advantageous trades are sufficiently low. The Coase Theorem indicates that the ultimate resource allocation will be efficient, regardless of the initial assignment of property rights, as long as contracting costs are sufficiently low and property rights are clearly assigned, well enforced, and readily exchangeable.

General knowledge is inexpensive to transfer, whereas specific knowledge is ex- pensive to transfer. Specific knowledge is quite important in economic decisions.

Chapter 3 Exchange and Markets 109

Central planning often fails because important specific knowledge is not incorpo- rated in the planning process. Within market systems, economic decisions are decentralized to individuals with the relevant specific knowledge. Prices convey gen- eral knowledge that coordinates the decisions of individuals. Private property rights provide important incentives to individuals to act productively, since they bear the wealth effects of their decisions.

In principle, all economic activity could be conducted through market transac- tions. However, even in market economies, much economic activity occurs within firms, where administrative decisions rather than market prices are used to allocate resources. Firms exist because of the contracting costs of using markets. However, organizing transactions within firms also involves costs. Individuals have incentives to organize transactions in the most efficient manner—to increase the gains from trade. Economic activities tend to be organized within firms when the cost is lower than that of using markets, and vice versa.

This chapter provides important background information on both markets and organizations. In Chapters 4 to 9, we shall extend the analysis of markets and study important managerial decisions such as output, inputs, pricing, and strategy. In these next six chapters, we assume that managers strive to maximize firm profits. In the remainder of the book, we shall extend the analysis of organizations and cover a variety of important topics about organizational design. A reader interested primarily in organizational design can move directly to Chapter 10 without loss of continuity.

R. Coase (1988), The Firm, the Market, and the Law (The University of Chicago Press: Chicago).

J. Earwell, M. Milgate, and P. Newman (1989a), Allocation, Information, and Markets (W.W.

Norton: New York).

——— (1989b), The Invisible Hand (W.W. Norton: New York).

F. Hayek (1945), “The Use of Knowledge in Society,” American Economic Review 35, 519–530.

M. Jensen and W. Meckling (1995), “Specific and General Knowledge, and Organizational Struc-

ture,” Journal of Applied Corporate Finance 8:2, 4–18.

O. Williamson (1985), The Economic Institutions of Capitalism (Free Press: New York).

3–1. Two men, Tom Hanks and Forest Gump, have been marooned separately on the same de-

serted island. There are two activities each man can undertake to obtain food: fishing and

gathering coconuts. Tom can catch 40 fish per hour or gather 10 coconuts per hour. Forest

can catch 10 fish per hour or gather 8 coconuts per hour. Answer the following questions:

a. Does Tom have a comparative advantage in producing both products? Explain.

b. Tom and Forest have not yet met. Tom is working 2 hours a day and producing (and

consuming) 48 fish and 8 coconuts (Note: the fish are very small). Forest is also work-

ing 2 hours a day, but he is producing and consuming 15 fish and 4 coconuts. Now

assume that Tom and Forest meet and develop a trading relationship. Come up with a

production and trading scheme such that they can each work the same amount per day

as before, but each is better off than before. Provide specific numbers to show how

they are better off.

3–2. a. Suppose sugar has the demand curve P � 50 � 5Q and the supply curve P � 5Q.

Compute the equilibrium price and quantity and show graphically. Calculate the con-

sumer surplus and producer surplus associated with this outcome.

b. What factors might cause the equilibrium price and quantity of sugar to change?

Self-Evaluation Problems

Suggested Readings

110 Part 1 Basic Concepts

Solutions to Self-Evaluation Problems 3–1. a. No, while Tom has an absolute advantage in producing both products, he only has a

comparative advantage in fishing. Forest has a comparative advantage in gathering co-

conuts. His marginal cost for gathering one coconut is 1.25 fish (10�8), while Tom has a

marginal cost of 4 fish (40�10).

b. One specific example is as follows: Say Tom produces only fish and Forest produces

only coconuts. There will be a total of 80 fish and 16 coconuts. Now suppose they

set a price of 2 fish per coconut and Tom buys 10 coconuts (for a price equal to

20 fish). Then Tom consumes 60 fish and 10 coconuts and Forest consumes 20 fish

and 6 coconuts. They each consume more of both commodities, so they are each bet-

ter off even though they are working the same amount as before. Many other exam-

ples could be constructed.

3–2. a. Set demand equal to supply: 50 � 5Q � 5Q and solve for the equilibrium quantity,

Q* � 5. Place Q* into either the supply or demand equation and solve for the equilib-

rium price, P* � $25. Graphically, the picture is

Triangle C pictures the consumer surplus. The area of a triangle is .5(Base � Height).

Therefore the consumer surplus is .5(5 � 25) � $62.5. The producer surplus pictured

by Triangle P is also $62.5.

b. Changes in the equilibrium price and quantity are induced by shifts in either the supply

or demand curves. Factors that affect demand include such things as consumer income

and the prices of other goods. For example, if an increase in consumer income caused

the demand curve for sugar to shift to the right (an increase in demand) both the equilib-

rium price and quantity would increase. Factors that affect supply of sugar include such

things as the prices of inputs used in the production process (e.g., land, labor, and fuel

prices). If an increase in the price of labor caused the supply curve to shift to the left

(a decrease in supply), the equilibrium price of sugar would increase and the equilibrium

quantity would decrease.

3–1. What is Pareto efficiency? Why do economists use this criterion for comparing alternative

economic systems?

3–2. What is a property right? What role do property rights play in a market economy?

3–3. Twin brothers, Tom and Bill, constantly fight over toys. For instance, Tom will argue it is his

turn to play with a toy, while Bill argues it is his turn. Their parents frequently have to inter-

vene in these disputes. Their mom has conceived an idea that might reduce these conflicts. In

particular, every toy in the house would be “owned” by one of the boys. The owner would

have complete authority over the use of the toy. The mom reasons that ownership would cut

Review Questions

5 Q



$50 Supply




Chapter 3 Exchange and Markets 111

down on disputes. Any time there is an argument over a toy, the owner gets the final and im-

mediate say. The boys’ dad is concerned that this idea will prevent the boys from learning to

“share.” He envisions that under the new system, Tom will not allow Bill to play with his

toys and Bill will not allow Tom to play with his toys. The old system forces them to figure

out a way to share the toys. Do you think that their dad’s concerns are valid? Explain.

3–4. Many economists favor free trade between nations. They argue that free trade will increase

total world output and make people of trading nations better off. Discuss how this argument

relates to concepts presented in the chapter.

3–5. What do you think will happen to the price and quantity of DVD players if

a. The availability of good movies to play on DVD players increases?

b. Personal income increases?

c. The price of inputs used to produce DVD players decreases?

d. Ticket prices at local movie theaters decline substantially?

3–6. Suppose that the U.S. government caps the price of milk at $1 per gallon. Prior to the cap

milk sold for $1 per gallon. Picture the effects of the price cap using a supply and demand

graph. Explain how the cap affects consumers and producers.

3–7. a. What is an externality?

b. Why might externalities lead a firm to discharge too much pollution into a river?

c. Congress has passed a law that limits the level of cotton dust within textile factories.

Why might a textile firm allow too much cotton dust within its workplace?

3–8. What is the difference between general and specific knowledge? How can specific

knowledge motivate the use of decentralized decision making?

3–9. Evaluate the following statement:

Using free markets and the price system always results in a more efficient resource allocation than central planning. Just look at what happened in Eastern Europe.

3–10. a. What are contracting costs?

b. Give a few examples of contracting costs.

c. What effect does the existence of contracting costs have on market economies?

3–11. If markets are so wonderful, why do firms exist?

3–12. In certain professional sports, team owners “own” the players. Owners can sell or trade

players to another team. However, players are not free to negotiate with other team own-

ers on their own behalf. The team owners initially obtain the rights to players through an

annual draft that is used to allocate new players among the teams in the league. They

also can obtain the rights to players by purchasing them from another team. Players do

not like this process and often argue that they should be free to negotiate with all teams

in the sporting league. In this case, they would be free to play for the team that offers the

most desirable contract. Owners argue that this change in rights would have a negative

effect on the distribution of talent across teams. In particular, they argue that all the good

players would end up on rich, media-center teams such as New York or Los Angeles

(because these teams could afford to pay higher salaries). The inequity of players across

teams would make the sport less interesting to fans and thus destroy the league. Do you

think the owners’ argument is correct? Explain.

3–13. The guide at the Washington Monument tells your 10-year-old nephew, “Enjoy the monument.

As a citizen, you are one of its owners.” Your nephew asks you if that is true. What do you say?

3–14. Locust Hill Golf Club is a private country club. It charges an initiation fee of $23,000. When

members quit the club, they receive no refund on their initiation fees. They simply lose their

membership. Salt Lake Country Club is also a private golf course. At this club, members join

by buying a membership certificate from a member who is leaving the club. The price of the

membership is determined by supply and demand. Suppose that both clubs are considering

installing a watering system. In each case, the watering system is expected to enhance the

quality of the golf course significantly. To finance these systems, members would pay a spe-

cial assessment of $2,000 per year for the next three years. The proposals will be voted on by

112 Part 1 Basic Concepts

the memberships. Do you think that the membership is more likely to vote in favor of the pro-

posal at Locust Hill or for the one at Salt Lake Country Club? Explain.

3–15. Critically evaluate the advice of the Providence Consulting Group, which recommended to

your company,

That you analyze all the business divisions in your company. Rank them on growth po- tential. Sell all the low-growth units and invest the money in the high-growth units. Make sure not to sell the high-growth units.

3–16. Suppose that the U.S. government begins charging a $1 sales tax to all consumers for each

dress shirt they buy.

a. What is likely to happen to the price (not including the tax) and quantity demanded of

dress shirts? Show using supply and demand graphs.

b. What is likely to happen to the demand for sport shirts (not taxed) and undershirts

(which are worn primarily with dress shirts)? Explain.

3–17. Title-loan firms offer high-interest loans (the interest rate can exceed 200 percent per year)

to high-risk customers. The title of a car often is used as collateral. If the borrower defaults

on the loan, the company can repossess the car. Recently, the financial press has reported

stories of poor people who have had their cars repossessed by title lending companies. Leg-

islation is being proposed in some states to make this lending practice illegal. A proponent

of the law made the following argument. “The market for loans is very competitive given all

of the banks, savings and loans, and finance companies. Outlawing title lending will make

poor people better off. It will motivate the lending companies to provide loans with less

onerous terms. Thus, low income people and people with bad credit histories will be able to

obtain credit on more favorable terms.” Do you agree with this argument? Explain.

3–18. Suppose that annual demand in the U.S. market for ice cream cones can be expressed as QD

� 800 � 0.2I � 100P, where QD is the number of cones demanded in millions of cones, I equals average monthly income in dollars, and P is price in dollars per cone. Supply can be

expressed as QS � 200 � 150P (with the same units for quantity and price).

a. Graph the demand and supply curves for ice cream cones, assuming that average

monthly income is $2,000, and solve for the equilibrium price and quantity.

b. Now assume that average monthly income drops to $750 and supply is unchanged.

Draw the new demand curve on the same graph as used in (a) above and solve for the

new equilibrium price and quantity. How would you describe the shift in demand


3–19. The rent control agency of Rochester has found that aggregate demand is P � 500 � 5QD.

Quantity, QD, is measured in thousands of apartments. Price, P, equals the monthly rental

rate in dollars. The city’s board of realtors acknowledges that this is a good demand esti-

mate and has shown that supply can be expressed as P � 5QS.

a. If the agency and the board are right about demand and supply, respectively, what is the

free-market price? How many apartments are rented?

b. If we assume an average of three persons per apartment, what is the expected change in

city population if the agency sets a maximum average monthly rent of $100 and all

those who cannot find an apartment leave the city?

3–20. Assume that before the ice storm of 2003, the weekly demand and supply for ice in the

Rochester Metro Area was given by the following equations:

Dpre: P � 100 � Q

Spre: S � 5 � 0.5Q

a. Draw a graph representing the Rochester ice market before the storm and label it care-

fully. What was the equilibrium price for the Rochester ice market before the storm?

And the total quantity of ice traded?

b. As a result of the ice storm, electricity went out in the Rochester area. The demand for

ice increased due to the lack of electricity to power refrigerators. The lack of power also

Chapter 3 Exchange and Markets 113

caused the supply to decrease. Ice producers were still able to produce some ice using

electric generators. Other ice had to be imported from other areas with power. The rele-

vant post-storm equations are the following:

Dpost: P � 110 � Q

Spost: P � 10 � 2Q

Draw a graph representing the Rochester ice market after the storm and label it care-

fully. What is the new equilibrium price? What is the quantity?

c. An open-ad in a local newspaper, commenting on the dramatic increase in price of ice

following the storm, stated:

Obviously, avarice and greed won out over decency and morality as ice-vendors took advantage of the ice storm to increase prices and gouge their loyal customers.

Do you agree with this statement? Explain.

3–21. Suppose the supply and demand for wheat is given by:

Supply: Qs � 1,800 � 240P

Demand: Qd � 2,550 � 10I � 266P

Where P � the price per bushel of wheat and I � income. The current value of I is 100.

a. Find the current equilibrium price and quantity of wheat sold in the marketplace.

b. Find the equilibrium price and quantity if income increases to 150.

c. Show the change in equilibrium using a standard supply and demand graph. Make sure

to label the axes and the curves. The graph does not need to be to scale. Just illustrate in

a general way what is going on.

3–22. Assume that the demand curve for sporting guns is described by QD � 100 � 2p and the

supply is described by Q S � �20 � p (Q D and Q S are in millions, p is in $).

a. Compute the competitive equilibrium price and quantity. Draw a graph of a supply and

demand curve and label it correctly. Compute the total value created in the market for

sporting guns (Hint: total value � consumer surplus � producer surplus).

b. Suppose that the government views sporting guns as a luxury product and taxes the con-

sumers $6 for each sporting gun they buy. Solve the new competitive equilibrium. What

losses do consumers of sporting guns incur as a result of the tax? What losses, if any, do

the producers of sporting guns incur?

3–23. Suppose there has been a storm in Nebraska that has destroyed part of the corn crop in the

field. The demand curve for corn has not changed. As a result, the market clearing prices

and quantities before and after the storm are: Pb � 50, Qb � 2,000; Pa � 100, Qa � 1,500.

(The subscripts a and b refer to “after the storm” and “before the storm.”)

a. Assume a linear demand curve for corn; that is P � � � �Q. Calculate �, � with the

provided information, and draw the demand curve with P on the y-axis and Q on the x-

axis. Label the intercept and the slope on the graph.

b. The supply curve for the period after the storm is P � (1�15)Q, and it is parallel to the

supply curve before the storm. Is the supply curve before the storm above or below that

after the storm? Calculate the slope and the intercept of the supply curve before the

storm. Draw both supply curves on a new graph with P on the y-axis and Q on the x-

axis. Add the demand curve (calculated in part a) to the graph.

c. Suppose consumers care only about corn consumption and apple consumption (they live

in a two-good world). How would the change in the price of corn affect the budget con-

straint of the typical consumer? Show graphically. How would the change in relative

prices affect the typical consumer’s consumption of corn versus apples? Is this result

consistent with your observation from the demand and supply framework (i.e., an in-

crease in the price of corn is associated with a decrease in the equilibrium quantity)?


114 Part 1 Basic Concepts

3-24. Suppose that the demand function for wheat is QD � 1.1 � .25P � .1I, where P � the price

of wheat and I � per capita income. The supply function for wheat is QS � 4.3 � .5P � .04Po, where Po � the price of oil. Currently, I � 20 and Po � 50.

a. Plot the supply curve and demand curve on the standard supply and demand graph.

b. Find the equilibrium price and quantity and depict it on your graph.

c. Calculate the resulting consumer surplus and producer surplus.

d. What would happen to the equilibrium price and quantity if income were to increase

(holding other factors constant)? Give a brief economic explanation for why this would


e. What would happen to the equilibrium price and quantity if the price of oil were to in-

crease (holding other factors constant)? Give a brief economic explanation for why this

would occur.

f. What would happen to the equilibrium price and quantity if income and the price of oil

were both to increase at the same time? Give a brief economic explanation for why this

would occur.

3-25. Input cost increases have caused the industry supply curve for golf balls to shift. The equi-

librium quantity changed, but the price did not. Can you say anything about the elasticity of

either the demand or supply curves from observing these effects? Will consumer expendi-

tures on golf balls increase or decrease? Explain.

3-26. Suppose the demand function for a product is P � 100 � (1/6)Q � 2I, where I is income

and the supply function is P � (1/3)Q. Currently I � 25.

a. Find the current equilibrium price and quantity.

b. Suppose the government imposes a $5 per unit excise tax on the product (charged to the

supplier). Find the new equilibrium price and quantity.

c. Draw a graph that shows the equilibriums before and after the tax. Show the areas of the

graph that represent the tax borne by consumers and the tax borne by suppliers (together

they add up to the total tax collected). Also depict the deadweight loss imposed by the


d. Explain in words what the deadweight loss represents.

e. What happens if the $5 per unit tax is imposed on consumers rather than suppliers?

3-27. In 1990, Congress adopted a new tax on luxury goods, such as yachts, private airplanes, and

jewelry. Assume this was charged as a per unit tax to the consumer. Was this a good way to

accomplish the objective of taxing rich people? Explain.

3-28. Your firm operates in a competitive industry. All firms in the industry have experienced an

increase in the cost of raw materials. Do you think that the firms in your industry will be

able to pass the cost increase on to consumers through higher prices? Explain.

3-29. Suppose that the Swiss government wants to encourage good grooming by Thun citizens. To

“encourage” better grooming it places a 15 CHF price cap on haircuts (no company can

charge more than 15 CHF for a haircut). The government official who sponsored the new

regulation argued (1) more people will want to get haircuts at a lower price and (2) thus more

people will get haircuts. Do you agree with the economic reasoning of this official? Explain.

3-30. Suppose that the demand curve for wheat is P � 100 � Q and the supply curve is P � Q.

Find the equilibrium price and quantity. Calculate consumer and producer surplus. Suppose

that the government imposes a price cap of 30. Show the effect graphically and calculate the

resulting consumer surplus, producer surplus and deadweight loss.

In the United States, top managers (officers and directors) have a fiduciary duty to act in the interests of the corporation and its shareholders. Consistent with this legal obligation, executives constantly profess to the media, stock analysts, and other constituencies a fundamental allegiance to increasing shareholder value, as reflected by the price of the common stock. Some critics, however, contend that managers ignore their legal duty and make decisions that benefit themselves at the

Appendix: Shareholder Value and Market Efficiency

expense of shareholders. We have little doubt that abstract legal principles, like fiduciary duty, are insufficient to induce corporate leaders to maximize shareholder value. However, theory and evidence suggests that there are a variety of internal and external control mechanisms that provide additional incentives to managers to be concerned about shareholder value. We examine these mechanisms beginning in Chapter 10.

Analyzing managerial decisions based on the assumption of shareholder wealth maximization has two benefits. First, it suggests what managers should do to meet their fiduciary responsibilities. Second, it describes what good managers actually do when they have sufficient incentives to focus on shareholder value.

Finance courses analyze investment and financing decisions where it is crucial to consider intertemporal tradeoffs among cash flows. For example, should a manager invest $1,000,000 to build a plant today that has the potential to yield $100,000 per year profit in the future? In contrast, managerial economics largely focuses on opera- tional decisions where intertemporal tradeoffs in cash flows are relatively less impor- tant. For example, what current price should the manager charge to maximize profits? These decisions can be analyzed under the simplifying assumption that managers seek to maximize single-period profits. Little is gained from the added complexity of as- suming that managers seek to maximize shareholder wealth, which involves the valu- ation of multiperiod cash flows.31

In Chapters 4 to 7, we present an economic analysis of demand, production, cost, market structure, and pricing. Here we follow the standard approach in managerial eco- nomics and assume that managers strive to maximize single-period profits. In subse- quent chapters, where intertemporal considerations are more important, we assume that mangers seek to maximize shareholder wealth.

The main text does not require a detailed understanding of stock market valuation. It is sufficient simply to understand that share price incorporates the effects of man- agerial actions on both current and future profits—appropriately adjusted for the timing and risk of the cash flows. In this appendix we go beyond what is necessary for the main text by providing a more detailed analysis of shareholder value. We begin by introducing the concept of present value and deriving an expression for the current value of a share of common stock. We then discuss the concept of stock mar- ket efficiency and its resulting managerial implications.

Present Value Is it better to receive a dollar today or a dollar a year from now? The obvious answer is that it is better to receive the dollar today, since it can be invested to yield more than a dollar in the future. Suppose that you invest a dollar in a risk-free asset (such as a U.S. Treasury security) with a 5 percent interest rate. At the end of one year, your investment will be worth $1.05. The investment, which promises $1.05 for certain in one year, and a dollar today have equivalent value since the dollar can be invested to produce the same future cash flow.32 The future value of a dollar invested in the risk- free asset for one year is $1 � (1.05) � $1.05. Conversely, the present value of $1.05 received for certain in one year is $1.05�1.05 � $1.

31In stationary settings where the same action is optimal in each period, the two objective functions are

equivalent. 32This equivalence of the risk-free investment and the dollar today holds even if the investor wants to consume

a dollar today as long the investment can be sold in the marketplace for its present value.

Chapter 3 Exchange and Markets 115

116 Part 1 Basic Concepts

The concept of present value extends to multiple periods. One dollar invested at 5 percent for n periods is worth $(1.05)n at the end of the nth period. The present value of this future value is $(1.05)n�(1.05)n � $1.00. More generally the present value of W dollars received for certain at the end of n periods is W�(1 � r)n, where r is the risk- free interest rate. For example, suppose that a risk-free investment promises a cash flow of $1.50 in five years and the annual interest rate is 5 percent. The present value of the investment is $1.50�(1.05)5 � $1.175 (investing $1.175 for five years at 5 percent yields a terminal value of $1.50, assuming any intermediate cash flows are reinvested).

Investments often generate multiple cash flows over time. For example, a 10-year U.S. Treasury bond pays a fixed rate of interest every 6 months until the security ma- tures at which time a final principal payment is made along with the final interest payment (20 total payments). The present value of a stream of cash flows is equal to the sum of the present values of the cash flows for each period. The general formula for the present value of a risk-free investment is

Present Value � �CFt�(1 � r)t (3.6)

where CFt is the cash flow that occurs in period t (t � 1 to n, the terminal period). For example, the present value of a 3-year bond that pays $100 at the end of each year and a principal payment of $1,000 at maturity is: 100�(1.05) � 100�(1.05)2 � 1,100�(1.05)3

� $1,136.16 (assuming a 5 percent risk-free rate). The concept of present value is extremely useful when comparing alternative in-

vestments with different time-series patterns of cash flows. Present value allows an “apples-to-apples” comparison of the alternatives since all are expressed in a common dimension—their present values. The best investment is the one with the highest net present value—the difference between the present value of its benefits and costs.

Share Value Our analysis of present value has focused on risk-free investments, where promised fu- ture cash flows were known with certainty and correspondingly discounted by the risk- free rate to obtain present values. Common stocks, however, are risky investments— the ultimate payouts to shareholders are not certain, but depend on the fortunes of the firm. While a firm might be expected to pay a given stream of dividends over time, in- vestors might receive higher or lower payments depending on the fortunes of the firm.

The values of stocks and other risky investments are determined by discounting expected cash flows by risk-adjusted discount rates. The discount rate used for valuing an investment increases with the risk of the expected cash flows. (Finance courses teach that the relevant risk from the perspective of a shareholder who holds this firm’s stock in a well-diversified portfolio is its systematic risk, that which cannot be elimi- nated through holding a diversified portfolio of investments.)

The intuition of stock valuation can be illustrated using a simple example. Suppose that a stock has the potential to pay a liquidating dividend to shareholders of either $50 or $150 at the end of the year. Each outcome has a probability of .5, and thus the expected cash flow is (.5 � $50) � (.5 � $150) � $100. If we discount the $100 expected cash flow at the risk-free rate of 5 percent we obtain $100�(1.05) � $95.24. Purchasing the stock at a price of $95.24 by definition yields an expected return of 5 percent ((100 � 95.24)�95.24). Risk averse investors, however, will not purchase the stock at this price since they can earn the same return for certain by in- vesting the $95.24 in a risk-free asset. For the stock to appeal to a risk-averse investor, it must sell for a lower price and thus offering a higher expected return.

Suppose investors would purchase this stock if it yielded an expected rate of return of 10 percent. The resulting price of $90.91 is found by discounting the expected cash flow of $100 by the risk-adjusted discount rate of 10 percent ($100�1.10 � $90.91).

More generally, companies are expected to last more than one period and may pay dividends in multiple periods. The value of a stock is equal to the sum of the present values of each of the expected future dividends. Expressed in equation form, the cur- rent value of a share of stock, P0, is

P0 � D1�(1 � k) � D2�(1 � k)2 . . . � D q

�(1 � k)q (3.7)

where Dt is the expected dividend paid to the investor at each time t, and k is the risk- adjusted discount rate (expected return).33 Cash flows that are expected to occur fur- ther into the future have less impact on the valuation due to this discounting process. Thus, analysts generally exert most of their effort predicting cash flows over the first 5 to 10 years of the investment. Simplifying assumptions typically are used for esti- mating the present value of the remaining cash flows.

In special cases, Equation (3.7) reduces to simple expressions. One prominent ex- ample is the constant growth model. This model assumes that investors expect the firm’s dividends will grow each period at a constant rate g � k, and that the firm will last forever. It is easy to show that Equation (3.7) reduces in this case to

P � D1�(k � g) (3.8)

Suppose that investors expect that the HG Corporation will pay a $5 dividend at the end of the year and that the dividend will grow at 5 percent annually forever. The price of the stock at the beginning of the year, assuming a risk-adjusted discount rate of .10 percent, is P0 � $5�(.10 � .05) � $100.

Equation (3.8) illustrates that the value of the stock is not determined by the cur- rent dividend alone. Growth firms often pay few dividends in their early years so that internally generated cash can be reinvested in the business. Nonetheless, they can sell at high prices if the market anticipates that they will make large payouts after their growth slows. Consider Microsoft Corporation as its growth rate slowed in 2004. In July 2004, Microsoft announced that it would double its annual dividend to share- holders by $3.5 billion per year, pay a one-time special dividend of $32 billion, and repurchase $30 billion of company stock over the following four years. The special dividend payment was the largest in S&P 500 history and quickly turned Microsoft from a small dividend payer to the tenth highest on the S&P 500 Index.

Stock Market Efficiency Investments with identical cash flows and risk sell for the same price in a well-func- tioning stock market.34 Since rational investors always will purchase the lower- priced of two identical securities, identical securities must sell for the same price for the market to clear (quantity supplied equal to quantity demanded for each security). In equilibrium the expected returns on identical securities are equal.

33Most investors receive part of their returns in the form of capital losses or gains when they sell the stock.

This observation does not invalidate Equation (3.7) since the price of the stock at the time of the sale reflects

the discounted value of the remaining expected cash flows. Also, the equation holds even when some

shareholders receive payouts in the form of share repurchases. 34Other characteristics such as liquidity and tax implications can be valued by the market. For simplicity we

concentrate on risk. The same principle holds in the more general case; investments with identical

characteristics that are valued by the market should sell at the same price.

Chapter 3 Exchange and Markets 117

ANALYZING MANAGERIAL DECISIONS: Shareholder Value and Market Efficiency

1. Suppose that you purchase a newly issued 10-

year U.S. Treasury bond for $10,000. The bond

has a promised interest rate of 5 percent ($250

every six months). The stated interest rate of 5

percent (annual payment of $500 divided by

the initial face value of $10,000) does not

change over the life of the bond. Do you ex-

pect that the market value of the bond will be

constant or variable over the life of the bond?


2. Calculate the present value of an investment

with the following expected cash flows at a

discount rate of 10 percent: year 1 � $500,

year 2 � $600, and year 3 � $650. Recalculate

the present value at discount rates of 15 per-

cent and 5 percent.

3. Is the discount rate used by investors to value

a given stock necessarily constant over time?


4. Find an event reported in today’s business press

that is likely to have an important effect on the

cash flows for a given firm. Use Yahoo’s fi-

nance Web site to produce a chart of the com-

pany’s stock price around the time or day of

the announcement of the event (http://www Explain why the market

reacted the way it did.

118 Part 1 Basic Concepts

A stock market is efficient if it responds quickly and rationally to new information (i.e., share prices fully reflect available information; each stock is priced to yield a competitive return given its risk; stocks are not systematically under- or overvalued). In an efficient capital market, the market values of securities reflect the present values of the expected future net cash flows to shareholders, including expected cash flows from future investments. If an event occurs that changes expected cash flows or risk of a firm, the share price will adjust quickly to reflect the new information.As a result, in- vestors should expect to receive competitive returns from purchasing stocks at current prices (i.e., the market-determined expected return for securities with similar charac- teristics). Depending on the fortunes of the firm, investors may end up earning more or less than the expected return. However, they should not expect to “beat the market” on a systematic basis.

The efficient markets hypothesis is perhaps the most extensively tested hypothe- sis in all the social sciences. The evidence is consistent with the view that stock mar- kets are at least reasonably efficient with respect to public information. Major stock markets react quickly to new information, and investors generally can expect that they will not earn abnormal returns from trading stocks based on publicly available information.

The research on efficient markets has important implications for corporate man- agers. First, there is no ambiguity about the firm’s objective function—managers should maximize the current market value of the firm. Hence, management does not have to choose between maximizing the firm’s current value or its future value, and there is no reason for management to have a time horizon that is too short. Second, management decisions that increase reported earnings, but do not affect current or fu- ture cash flows, represent wasted effort. Third, if new securities are issued at market prices, which reflect an unbiased assessment of future payoffs, then concern about di- lution or the sharing of positive net present value projects with new security holders is eliminated. Fourth, security returns are meaningful measures of firm performance. This allows scholars, management, and analysts to use security returns to estimate the effects of various corporate policies and events on the market value of the corporation.

For example, soon after Hurricane Katrina wreaked havoc on the Gulf Coast in sum- mer 2005, the stocks of property insurance companies fell dramatically, while the stocks of companies such as Home Depot (which sells lumber and other building sup- plies) increased. According to the efficient markets hypothesis these reactions reflect the stock market’s unbiased estimate of the valuation effects of the storm. The evi- dence that security returns are a meaningful measure of firm performance also provides support for using equity-based compensation to provide incentives to top management.

Chapter 3 Exchange and Markets 119


4 C H A P T E R O U T L I N E

Demand Functions

Demand Curves

Law of Demand

Elasticity of Demand

Linear Demand Curves

Other Factors That Influence Demand

Prices of Related Products


Other Variables

Industry versus Firm Demand

Network Effects

Product Attributes

Product Life Cycles

Demand Estimation


Price Experimentation

Statistical Analysis



Appendix: Demand

T he Players Theater Company is a regional repertory theater in the Midwest. Each year, it produces six plays, ranging from Shake- speare to contemporary musicals. PTC has priced its tickets at $30. On a typical night, approximately 200 of the theater’s 500 seats are

filled. The PTC board met recently to discuss a possible price decrease to $25 for next season. Advocates of the proposal argued that the decrease in ticket prices would increase the theater’s customer base, the number of tick- ets sold, and revenues for the company.

At the meeting, the PTC board engaged in a heated debate over the pro- posal. It soon became evident that the board had insufficient information to make a sound decision. For instance, nearby restaurants, which serve PTC customers, have indicated that they are planning to implement substantial price increases before the beginning of the next season. Would this increase affect the demand faced by PTC and thus the appropriate ticket price? Al- though customers might buy more tickets at lower prices, would total rev- enue or profits necessarily increase? Would it be better to attract additional customers by lowering price or by improving the quality of PTC plays? After a lively discussion, the proposed decrease in price was tabled for fur- ther study.

This discussion at the PTC board meeting highlights the fact that man- agers require a detailed understanding of product demand to make sound pricing decisions. Understanding product demand also is important for deci- sions on advertising, production levels, new product development, and capi- tal investment projects.


P A R T T W O M a n a g e r i a l E c o n o m i c s


1. Define and mathematically express a demand function.

2. Explain the Law of Demand and the connection between declining marginal

value and downward sloping demand curves.

3. Explain price elasticity and its relation to total revenue.

4. Define marginal revenue and discuss its relation to the demand curve, total rev-

enue, and price elasticity.

5. Interpret and explain income and cross elasticities.

6. Identify differences between industry and firm demand.

7. Discuss network effects.

8. Derive and graph demand curve, marginal revenue, total revenue, and price

elasticity when given a linear demand function.

Chapter 4 Demand 121

Chapter 3 offered an introduction to supply and demand analysis. In that chapter, we introduced the notion of a demand curve and discussed some of the factors that can cause a demand curve to shift. In this analysis, we focused on the aggregate de- mand curve for all the firms in the industry. In this chapter we provide a more exten- sive analysis of demand at the firm level. Important topics include demand functions, demand curves, factors affecting demand, industry versus firm-level demand, net- work effects, demand for product attributes, product life cycles, and demand estima- tion. In the technical appendix to this chapter, we derive point elasticities, analyze marginal revenue for a linear demand curve, and examine a special (log-linear) demand function. Most of the analysis in this chapter (e.g., the relation between price and revenue changes) applies to both industry-level and firm-level demand curves. We focus on firm level demand because of its particular importance to managers.1

Demand Functions Managers require a fundamental knowledge of the factors that affect the demand for their product. Only by understanding these factors can they make sound decisions on pricing, output, capital expenditures, and other strategic issues. In fact, poor pricing decisions can destroy firm value and damage executive careers. Kraft Foods demoted the head of its NorthAmerican business unit, Betsy Holden, after she raised prices and Kraft lost market share. Over this period Kraft’s stock price fell 20 percent while other food companies’ stock prices rose 9 percent.2

A demand function is a mathematical representation of the relation between the quantity demanded of a product and all factors that influence this demand. In its most general form, a demand function can be written as

Q � f (X1, X2, . . . Xn) (4.1)

where the Xis are those factors that affect the demand for this product. The quantity demanded Q is the dependent variable in the demand function, since

its value depends on the variables on the right-hand side of the equation. The Xis are the independent variables. In this chapter, we focus on three particularly important independent variables: the price of the product, the prices of related products, and the incomes of potential customers. This analysis can be extended to include other vari- ables, such as advertising expenditures, tastes and preferences, and consumer expec- tations (e.g., about future prices).

For concreteness, we continue to focus on PTC as an example. We assume that PTC faces a demand function for tickets on any given night that can be expressed by the following function:3

Q � 117 � 6.6P � 1.66Ps � 3.3Pr � 0.0066I (4.2)

1The supply and demand analysis in Chapter 3 implicitly assumes that the market is “perfectly competitive.”

As explained later in this book, firm-level demand curves are horizontal in perfectly competitive industries.

The firm-level demand curves in this chapter are downward sloping, which implies that the firm has at least

some “market power.” Most firms have at least a small degree of market power, for example, due to locational

or other advantages in serving a subset of customers. 2S. Ellison and V. O’Connell (2003), “Kraft Removes Holden as Co-Chief,” The Wall Street Journal (December

17), A3 and A8; E. Herman (2005), “Former Co-CEO Holden Leaves Kraft,” Chicago Sun-Times (June 25). 3Note that this function assumes that PTC can sell fractional tickets. This assumption does not have a material effect

on our analysis. However, it allows us to draw continuous demand curves. One way to think of quantity in this

example is as the average number of tickets sold for a performance. In this case, fractional tickets are possible. Note

also that we assume that demand is constant for each performance by PTC. PTC performances are all scheduled for

Friday and Saturday nights. If they expand their schedule to include weeknights or matinees, it is likely that demand

conditions for these performances will differ and hence so should prices. These issues are discussed in Chapter 7.

where P is the price of PTC tickets, Ps is the ticket price at a nearby symphony hall, Pr is the average meal price at nearby restaurants, and I is the average household in- come of area residents.

As our starting point, we assume PTC tickets are currently priced at $30; symphony tickets and meals are priced at $50 and $40, respectively; income is $50,000. Given these values, the demand function implies that PTC sells 200 tickets per night. We now examine each of the independent variables in this demand function in more detail.

Demand Curves The price of the product is particularly important in demand analysis for two rea- sons: First, prices are among the most important variables that customers consider in making purchasing decisions. Second, managers choose the price of their products; variables such as the prices of other products and income levels largely are beyond their control. Given its special importance, economic analysis traditionally singles out the effects of price from other independent variables in the demand function.

A demand curve for a product displays for a particular period of time how many units will be purchased at each possible price, holding all other factors fixed.4 The left panel of Figure 4.1 depicts the demand curve for PTC tickets. By convention,

122 Part 2 Managerial Economics

406.6400 400

$ $



60 Income = $50,000

Income = $51,000

T ic

ke t p

ric e

(in d

ol la

rs )


D D0



Quantity of PTC tickets Quantity of PTC tickets

Figure 4.1 Demand Curves

The left panel shows the demand curve for the Players Theater Company tickets. By convention, price is placed on the vertical axis, while quantity is placed on the horizontal axis. The equation for PTC’s demand curve is: P � 60 � 0.15Q. The curve indicates that, for example, 200 tickets are purchased at $30 and 133 tickets are purchased at $40. The right panel indicates that the demand curve shifts to the right as income increases from $50,000 to $51,000—at each price, consumers buy more tickets. Movements along a demand curve are motivated by changes in price and are called changes in the quantity demanded. Movements of the entire demand curve are motivated by other factors, such as changes in income, and are referred to as changes in demand.

4Technical note: We derive an individual’s demand curve from the indifference curve/budget line analysis in

the appendix to Chapter 2. The price of one good—say, food—is varied, holding the price of other goods and

income fixed. The person’s optimal choices are recorded. The individual’s demand curve simply plots the

optimal choices of the good (in this case, food) against the associated prices. The firm-level demand curve,

in turn, is the sum of the demands of all individuals at each price.

Chapter 4 Demand 123

price is placed on the vertical axis, while quantity is placed on the horizontal axis.5

The equation for PTC’s demand curve is

P � 60 � 0.15Q (4.3)

This expression is obtained by substituting the current values of the other variables into Equation (4.2) and solving for P. The equation indicates that, for example, 200 tickets are purchased at $30 and 133 tickets are purchased at $40.6

Demand curves hold other factors fixed. Changes in income or the prices of sym- phony tickets or restaurant meals will cause shifts in the position of the demand curve (the intercept changes). For instance, the right panel of Figure 4.1 indicates that the demand curve shifts to the right as income increases from $50,000 to $51,000—at each price, patrons purchase 6.6 more tickets. Movements along a de- mand curve reflect changes in price and conventionally are called changes in the quantity demanded. Movements of the entire demand curve are caused by other factors (such as this change in income) and are referred to as changes in demand.

Law of Demand

Demand curves generally slope downward—individuals purchase less (or certainly no more) of a product as the price increases. PTC’s demand curve has a slope of �0.15. Although it is conceptually possible that individuals might purchase more of a product as the price rises, as a practical consideration, managers are quite safe in assuming that the quantity demanded for their products varies inversely with price.7 It would be fool- ish for PTC board members to think that they would sell more tickets if they raised the price. The negative slope of demand curves has become known as the law of demand.

5In subsequent chapters, we consider costs which are a function of quantity produced. Placing P on the vertical

axis allows us to display both demand (revenue) and costs on the same graph in a convenient fashion. 6Rounded to the nearest dollar. 7Goods for which the income effect swamps the substitution effect so that consumers purchase more at higher

prices are called Giffin Goods. We ignore this possibility throughout the rest of this book.


Learning the Law of Demand the Hard Way Trillions of dollars of products are now sold online. It is not uncommon for companies to make mistakes in posting

online prices. For example, in 2012 United Airlines posted a price of $43 for first-class airline tickets from New York

to Hong Kong—the normal price is around $11,000. In 2013, Walmart accidently posted an $8.85 price for computer

monitors, digital projectors, and other products that normally sell for hundreds of dollars. Consistent with the law of

demand these types of errors typically motivate a rapid and large increase in the number of orders placed by customers.

Online pricing errors are particularly problematic, since potential customers throughout the country or around the

globe view the posted prices. Placing a wrong price tag on an item in a store is typically much less consequential. In

extreme cases, such as the United Airline and Walmart errors, companies often tell customers the price was a mistake

and cancel the orders. However, this action understandably upsets potential customers, who might be less likely to

purchase from the company in the future. Also the legality of canceling orders can be unclear. Companies frequently

honor small pricing errors to avoid upsetting customers.

Source: S. Kim (2013), “Why Walmart Canceled Mispriced Item Orders,” ABC News (November 7); and T. Hume (2012) “Too Good

to be True: New York to Hong Kong for $43,” CNN Travel (July 23).

Elasticity of Demand

Demand curves vary in their sensitivity of quantity demanded to price. In some cases, a small change in price leads to a big change in quantity demanded, whereas in other cases a big price change leads to only a small change in quantity demanded. Information on this sensitivity is critically important for managerial decision mak- ing. For instance, the board would not want to lower ticket prices to $25 if it could fill the theater by reducing prices only to $28.

One measure of the responsiveness of quantity demanded to price is simply the slope of the demand curve. But this measure is of limited usefulness, in part be- cause it depends on the particular dimensions in which quantities are quoted. For instance, if the slope of a demand curve is �2 when the quantity is expressed in tons, it is only �0.001 when the quantity is stated in pounds. Using the magnitude of the slope coefficient to derive insights into the sensitivity of quantity demanded to price requires additional computation. Economists more frequently use a di- mensionless measure of this sensitivity known as the price elasticity of demand, �. (Frequently, this elasticity is simply referred to as the elasticity of demand.)

Demand elasticity measures the percentage change in quantity demanded given a percentage change in its price. The law of demand indicates that price elasticities are negative; convention, however, dictates that we state this elasticity as a positive number. Higher price elasticities mean greater price sensitivity. The elasticity of demand, �, thus is given by

� � �(% change in Q )�(% change in P ) (4.4)

124 Part 2 Managerial Economics

ANALYZING MANAGERIAL DECISIONS: Setting Tuition and Financial Aid

The Board of Ursinus College in Pennsylvania

raised its tuition and fees 17.6 percent to $23,460

in 2000. It subsequently received 200 more appli-

cations than the year before. The president of the

college surmised that “applicants had apparently

concluded that if the college cost more, it must be

better.” Other colleges that raised tuition to match

rival colleges in recent years include University of

Notre Dame, Bryn Mawr College, Rice University,

and the University of Richmond. They also experi-

enced an increase in applications. In contrast,

North Carolina Wesleyan College lowered their tu-

ition and fees about 10 years ago by 22 percent and

attracted fewer students. The college president con-

cluded that “it didn’t work out the way it had been

hoped. People don’t want cheap.”

You are hired as a consultant to a President of a

liberal arts college in the East. You are asked to

evaluate a recommendation by the college’s

Admissions Director, Susan Hansen, to increase

tuition and to reduce financial aid to students.

Susan argues that the data from competing colleges

suggest that the demand curves for colleges slope

upward—the quantity demanded increases with

price. Susan projects that the increase in tuition and

reduction in financial aid will solve the school’s fi-

nancial problems. Last year, the college enrolled

400 new students who each paid an effective tuition

of $15,000 (after financial aid), totaling

$6,000,000. She projects that with the increased

demand from charging an effective tuition of

$25,000, the college will be able to enroll 600 new

students (of equal or better quality), totaling

$15,000,000. Evaluate Susan’s analysis and


Source: J. D. Glater and A. Finder (2006), “In Twist on Tuition Game, Popularity Rises with Price,” (December 12).

Chapter 4 Demand 125

Calculating Price Elasticities This elasticity can be approximated between any two points using the concept of arc elasticity.8 The formula for an arc elasticity is

� � �[�Q �(Q1 � Q2)�2] � [�P�(P1 � P2)�2] (4.5)

where � represents the change between the two points.9 Figure 4.2 displays two points on PTC’s demand curve for theater tickets. As shown in the figure, the arc elas- ticity between these two points is 1.4. Hence, over this region, for every 1 percent increase in price, patrons reduce the quantity of tickets purchased by approximately 1.4 percent.

Price elasticities lie between zero and infinity. If the price elasticity is zero, quan- tity demanded is unaffected by price. In this case, as depicted in the left panel of Figure 4.3, the demand curve is vertical. If the price elasticity is infinite, a small in- crease in price will lead customers to purchase none of the product. In this case, as displayed in the right panel of Figure 4.3, the demand curve is a horizontal line. For instance, a small farmer might not be able to sell any soybeans if they were priced above the prevailing market price. Demand is elastic if the price elasticity is greater than one, unitary if equal to one, and inelastic if less than one.

Elasticity varies along most demand curves. For instance, with a linear demand curve, elasticity will be high when quantities are low and approach zero as the


Walmart Supercenters and Demand Elasticities for Grocery Products Price elasticities usually increase with available substitutes. Over the last 25 years, the price elasticity for grocery products

has increased substantially in many markets where Walmart built supercenters—a substitute for traditional supermarkets

and other grocery stores. Unlike traditional supermarkets, supercenters offer a wide range of products at substantially

discounted prices (e.g., groceries, clothing, home furnishings and electronic equipment). Supercenters are very large

relative to most grocery stores, averaging about 179,000 square feet in 2014. Walmart began establishing supercenters in

the South and Southwest regions of the United States in the late 1980s. It is now the largest grocery market chain in the

country with supercenters in 48 of the 50 states. Researchers have found that the prices for groceries at supercenters are

more than 25 percent lower than at traditional supermarkets prior a Walmart entry. As a result, Walmart entry into a

market motivates competing stores to lower their prices. Walmart supercenters benefit some consumers directly by selling

groceries to them at much lower prices than they paid historically. Consumers, who continue to shop at traditional

supermarkets, benefit indirectly from lower prices at these stores. Not surprisingly, traditional supermarkets oppose

Walmart supercenters. Labor unions also oppose them because the increased competition makes it difficult for traditional

grocery stores to continue to pay the historic level of wages to organized labor. Zoning restrictions and other regulations,

which are endorsed by traditional supermarkets and labor, have blocked Walmart entry in some markets. Consumers in

these markets continue to pay relatively high prices for their grocery products.

Source: J.Hausman and E. Leibtag (2007), “Consumer Benefits from Increased Competition in Shopping Outlets: Measuring the Ef-

fect of Walmart,” Journal of Applied Econometrics and Form 10-K Walmart Stores, Inc., for fiscal year ended January 31, 2014.

8Price elasticity can be measured at a point on the demand curve. The concept of point elasticity requires

elementary knowledge of calculus and, more importantly, a smooth mathematical demand curve. While our

example assumes such a curve, data on demand often is available for only a few price-quantity combinations.

We show how to calculate point elasticities in the appendix to this chapter. 9Equation (4.4) can be expressed as � � ��Q �Q � �P�P. When calculating the elasticity between two

points, the question arises as to which Q and P to use in this expression, the starting or ending values.

Equation (4.5) uses the average of these two values—the initial plus the ending values divided by 2.

126 Part 2 Managerial Economics


(133, $40)

� = −[ΔQ/(Q1 + Q2)/2] � [ΔP/(P1 + P2)/2]

(200, $30)


60 = −[−67/(200 + 133)/2] � [10/(30 + 40)/2]

= 1.4



Quantity of PTC tickets

T ic

ke t p

ric e

(in d

ol la

rs )



Q1 Q2

Figure 4.2 Arc Elasticity

This figure displays two points on PTC’s demand curve for theater tickets. As displayed in the figure, the arc elasticity between these two points is 1.4. Thus, over this range, for every 1 percent increase in price there is approximately a 1.4 percent reduction in the quantity of tickets purchased.

D (� = 0)

D (� = �)




Quantity Quantity

P ric

e (in

d ol

la rs


P ric

e (in

d ol

la rs


Figure 4.3 Range of Price Elasticities

Price elasticities lie between zero and infinity. If the price elasticity is zero, quantity demanded is unaffected by price. In this case, as depicted in the left panel of the figure, the demand curve is vertical. If the price elasticity is infinite, as in the right panel, a small increase in price will cause people to purchase none of the product, and the demand curve is a horizontal line.

quantities become large. (Try calculating some arc elasticities along PTC’s demand curve.) We discuss this topic in greater detail later. (In the appendix to this chapter, we present a special demand curve that has constant elasticity—it does not vary along the curve.)

Chapter 4 Demand 127

Price Changes and Total Revenue One of the board’s concerns is how total revenue changes if it lowers ticket prices. We now demonstrate that the relation between revenue and price depends on the de- mand elasticity. Total revenue is calculated by multiplying the quantity purchased times the price (i.e., P � Q ). If price elasticity is inelastic (less than one), then the quantity demanded is less responsive to a change in price; a 1 percent increase in price results in less than a 1 percent decrease in quantity. Thus total revenue in- creases. Conversely, a price decrease results in a decrease in revenue. In contrast, if demand elasticity is unitary (equal to one), a 1 percent change in price results in an offsetting 1 percent change in quantity and hence total revenue is unchanged. Finally, if demand is elastic (value greater than one), a small increase in price results in a de- cline in revenue, whereas a small decrease in price results in an increase in revenue. These relations are summarized in Figure 4.4. We discuss these relations in greater detail below.

Determinants of Price Elasticities The elasticity of demand tends to be high when there are good substitutes for the product. For instance, if a flight is overbooked airlines have little trouble finding

Inelastic demand (� < 1) ↑ P ⇒ ↑ Total revenue ↓ P ⇒ ↓ Total revenue

Unitary elasticity (� = 1) ΔP ⇒ No change in total revenue

Elastic demand (� > 1) ↑ P ⇒ ↓ Total revenue ↓ P ⇒ ↑ Total revenue

Figure 4.4 Price Elasticities, Price Changes, and Total Revenue

How total expenditures on a product change with price depends directly on the price elasticity. This figure displays the relation between small price changes, total revenue, and price elasticities.


Price Elasticities Economists have estimated the price elasticities of various products, such as

Sugar � 0.31

Potatoes � 0.31

Tires � 1.20

Electricity � 1.20

Haddock � 2.20

Movies � 3.70

These estimates indicate that sugar and potatoes have relatively low-price elasticities. This might be expected given that

these products represent a small portion of most people’s budgets. Also, sugar has few close substitutes. Haddock and

movies have high elasticities. Haddock is a narrowly defined product (as opposed to fish) and has many close substitutes.

Movies are a luxury item for many people; higher prices cause individuals to consume other forms of entertainment.

Source: E. Mansfield and G. Yohe (2004), Microeconomics (W.W. Norton: New York), 135.

volunteers to surrender their seats when alternate flights are available that involve no material delay in arrival. Conversely, if the overbooked flight is the last of the day, to elicit volunteers might require several free tickets.

The elasticity of demand for PTC tickets is likely to increase with the number of competing events in the city. With many entertainment options, a small increase in the price of PTC tickets might be sufficient to induce a substantial number of potential consumers to attend other events. When alternatives are more limited, additional customers will decide to pay the higher price for PTC tickets rather than just stay at home.

Demand elasticities also can depend on the importance of the goods within con- sumers’ budgets. Goods such as salt and pepper, which consume a relatively small proportion of a person’s income, tend to be relatively price-insensitive, or inelastic. On the other hand, goods such as major appliances and automobiles represent more substantial purchases. Customers are more likely to comparison-shop to collect product information and thus are more likely to be price-sensitive.

A third determinant of price elasticity is the length of the period to which the demand curve pertains. Demand tends to be more elastic or responsive to price changes over a longer period than within a shorter period. An increase in PTC ticket prices is likely to result in an immediate decline in tickets sold. Long-run effects will be even larger as consumers identify other entertainment options or fail to renew season tickets (these effects will cause the demand curve to shift to the left). Similarly, a large increase in the price of oil will result in a near-term de- cline in the quantity of oil demanded—people will set their thermostats to a lower temperature and drive less. Over time, the effect will be larger as consumers in- sulate their homes better, buy smaller, more fuel-efficient cars, and shift to alter- native energy sources.

Other factors that can affect price elasticities include the degree of brand loyalty and whether consumers view the product as a “necessity” or “luxury good.” Branded products with high customer loyalty and products that are viewed as necessities will generally be less price elastic than unbranded products and luxury goods. Consumers vary in what they consider to be necessities versus luxuries. Many people consider sail boats to be a luxury good, but for people who race them for a living, they are more likely to be viewed as a necessity. Differences in the demands for products across potential customers will be explored in more detail in a later chapter.

128 Part 2 Managerial Economics


Demand Elasticities and Airline Pricing Round-trip airfares are substantially lower if the traveler stays over a Saturday night. Airline companies offer this

discount to increase revenues (and profits). The typical traveler who stays over a Saturday night is a tourist. Tourists

have relatively high price elasticities for air travel. Lowering the price from the standard fare correspondingly increases

revenue: The price decrease is more than offset by the increase in tickets sold. Airline companies do not offer

comparable discounts to travelers who complete the round-trip midweek. These customers are primarily business

travelers who have relatively inelastic demands. Lowering price would decrease revenue because the decrease in price

would not be offset by an increase in tickets sold. Airline companies also offer fewer discounts during peak periods,

such as the period around the Thanksgiving holiday. During these periods, demand is relatively inelastic and they can

fill the planes without offering substantial discounts.

Chapter 4 Demand 129

Linear Demand Curves

The PTC board’s decision on whether or not to lower prices depends on the relation between price and total revenue and thus its demand elasticity: It would make little sense for the board to lower prices if a price reduction would lower total revenue. We now provide a more in-depth analysis of the relation between price and revenue and discuss the PTC board’s optimal pricing policy. Through this analysis, we illustrate the properties of linear demand curves. Knowing these properties is useful for un- derstanding the subsequent analysis in this book.10

Total Revenue PTC’s total revenue (TR) for any given performance is equal to the quantity of tick- ets sold times the price. Price is given by the demand curve in Equation (4.3). Thus, total revenue can be expressed as

TR � P �Q (4.6)

� (60 � 0.15Q )Q � 60Q � 0.15Q 2

Figure 4.5 displays PTC’s demand and total revenue curves. Total revenue in- creases as price decreases up to the midpoint of the demand curve. Over this range, demand is elastic: The percentage decline in price is smaller than the percentage in- crease in quantity demanded. The elasticity is unitary at the midpoint. Past the mid- point, price declines result in reduced total revenue; thus, demand is inelastic over this range. These are general properties of linear demand curves.

Marginal Revenue An important concept in economics is marginal revenue, which is defined as the change in total revenue given a one-unit change in quantity. Intuitively, marginal rev- enue for the first unit is just its price. Thus, the intercepts of the demand and marginal revenue curves are the same. As quantity increases, marginal revenue is below price—to sell an extra unit, the price charged for all units must decrease. Marginal revenue is positive up to the midpoint of the demand curve (total revenue is increas- ing over this interval). At the midpoint, demand elasticity is unitary and marginal revenue is zero. Beyond the midpoint, marginal revenue is negative—the increase in revenue from selling another unit is less than the decline in revenue from lowering price (see the appendix). Hence, marginal revenue (MR) for a linear demand curve is a line with the same intercept as the demand curve but with twice the negative slope (see Figure 4.5). The equation for PTC’s marginal revenue is

MR � 60 � 0.3Q (4.7)

Profit Maximization All of PTC’s costs are fixed and do not depend on the quantity of tickets sold on a given evening—actors and utilities have to be paid regardless of how many people attend the performance. Thus, the PTC board’s objective is to maximize total revenue (for PTC, with costs fixed, maximizing total revenue is equivalent to

10Technical note: This result also can be applied to nonlinear demand curves. At a specific point, construct the

tangent to the demand curve. Now, if the tangency point is at the midpoint of the tangent, the elasticity is

unitary, and so forth.

130 Part 2 Managerial Economics

Inelastic demand (� < 1)

200 Q








� = 1

Elastic demand (� > 1)

T ot

al r

ev en

ue (

in d

ol la

rs )

T ic

ke t

pr ic

e (i

n do

lla rs


Quantity of PTC tickets

Figure 4.5 Demand, Total Revenue, and Marginal Revenue for Linear Demand Curves

This figure displays PTC’s demand and total revenue curves in the upper and lower panels. Total revenue increases as price decreases up to the midpoint of the demand curve. Thus, over this range, demand is elastic: The percentage decline in price is smaller than the percentage increase in quantity demanded. The elasticity is unitary at the midpoint. Past the midpoint, price declines result in reduced total revenue; and thus, demand is inelastic over this range. An important concept in economics is marginal revenue, which is defined as the change in total revenue given a unitary change in quantity. In the appendix, we show that marginal revenue (MR) for a linear demand curve is a line with the same intercept as the demand curve but with twice the negative slope. The marginal revenue curve for PTC is pictured in the figure.

ANALYZING MANAGERIAL DECISIONS: Demand Curve for an Electronics Product

You work for a company in India that manufac-

tures and exports batteries and other charge stor-

age devices. You are the sales manager for a DC-

DC converter that is used to step up or step down

the voltage in various industrial applications. You

currently price the product at 4,000 Indian Rupees

(INR) and sell 100,000 units. You estimate that if

you priced the product at 3,000 INR you would

sell 150,000 units.

You think it is reasonable to assume that your

demand curve is linear.

1. Derive the equation for your demand curve

from the two price and sales points.

2. Are you currently operating in the elastic or

inelastic portion of your demand curve?

3. You are paid a sales commission based on your

total sales. What price would you charge to

maximize your bonus?

4. Is this price likely to be optimal from your

firm’s standpoint, which has profit maximiza-

tion as a goal?

Chapter 4 Demand 131

maximizing total profit). Figure 4.5 indicates that revenues are maximized at a price of $30. Hence under current conditions, the PTC board should not lower the ticket price to $25. Currently, the company is collecting $30 � 200 � $6,000 in revenue per night. If the price is decreased to $25, total revenue would be $25 � 233 � $5,825 per night.11 (The upcoming increase in restaurant prices will change the optimal pricing policy. A practice problem at the end of this chapter explores this policy change.)

Note that, in contrast to this example, most firms do not want to maximize total revenue. PTC, with only fixed costs, is a special case. In most firms, both costs and revenues vary with output. A profit-maximizing firm must consider both effects. We discuss these considerations in greater detail in Chapters 5 through 7.

Other Factors That Influence Demand In addition to a product’s own price, the prices of related products and incomes of potential customers are among the more important factors that influence product demand.

Prices of Related Products

Complements versus Substitutes The demand for a product can be affected by the prices of related products. For instance, if the local symphony raises its ticket prices, arts patrons will be less likely to attend the symphony and more likely to attend the PTC. Thus, there is a positive relation in Equation (4.3) between the demand for PTC tickets and the price of sym- phony tickets. Goods that compete with each other in this manner are referred to as substitutes. In contrast, if local restaurants raise their prices, the demand for PTC tickets falls (note the negative sign in the demand function). For instance, some po- tential PTC customers will choose to stay home because the total cost of an evening on the town has increased. Products like theater tickets and meals at restaurants, which tend to be consumed together, are complements. Another example of com- plements is digital video disk players and DVD movies. Between 1997 and 1999, the price of DVD players fell from $600 to $299. Sales of DVD players rose from under 50,000 to 600,000 over these two years. And as consumers experience the better sound quality and video, they also are buying bigger TVs and better sound systems. Big-screen TVs were up 12 percent and audio sales 11 percent. “The

11From the demand curve: $25 � 60 � 0.15Q. Therefore, Q � 233.


Complementarity between Computer Hardware and Software Over the past 25 years there has been a dramatic decrease in the price of personal computers. Not only has the price of

PCs decreased, but their quality and computing power have improved substantially as well. This decrease in the price

of personal computers has increased the quantity of PCs demanded enormously. In addition, it also has increased the

demand for software products. Today, some of the largest companies in the world (e.g., the Microsoft Corporation)

specialize in writing of software for PCs. Computer hardware and software are complements and thus have negative

cross elasticities.

132 Part 2 Managerial Economics

biggest market driver is DVD,” said Terry Shimek, owner of Shimek’s Audio Video in Anchorage, Alaska. Moreover, the demand for DVD movies increased as well. Initially caught flat-footed, Hollywood started making more movies available on DVD. The number jumped from 1,800 in 1988 to 5,000 in 1999. (That is compared to 18,000 on VHS tape.)12

Cross Elasticities One frequently used measure of substitution between two products is the cross elas- ticity of demand. Cross elasticity is defined as the percentage change in the quantity demanded of a good, given a percentage change in the price of some other good. Cross elasticities between any two goods, X and Y, can be calculated using a formula that is analogous to Equation (4.5):

�xy � [�Qx�(Qx1 � Qx2)�2] � [�Py�(Py1 � Py2)�2] (4.8)

Unlike price elasticities, which are invariably positive (at least when you multiply them by �1), cross elasticities can be either positive or negative—substitutes have positive cross elasticities whereas complements have negative cross elasticities. Whether a commodity has strong substitutes or complements depends, in part, on how finely the commodity is defined. Pepsi and Coke might have relatively large cross elasticities. The cross elasticities between colas, more broadly defined, and other soft drinks are likely to be smaller.13

Cross elasticities are useful because managers frequently want to forecast what will happen to their own sales as other companies change their prices. The PTC board is concerned about the effects that a forthcoming increase in restaurant prices would have on its ticket demand. If meals in local restaurants and theater tickets are strong complements, a substantial increase in restaurant prices would cause a serious decline in the demand for PTC tickets. In this case, the PTC board might want to off- set this shift in demand by lowering ticket prices or advertising more heavily. In con- trast, if meals and tickets are weak complements, the increase in meal prices would have little effect on ticket demand. In our example, a $10 increase in meal prices will result in 33 fewer ticket sales per night. Using the formula in Equation (4.8), the cor- responding cross elasticity between these two points [(200, $40); (167, $50)] is


Derived Demand Some products are demanded, not because individuals receive pleasure from consuming them, but rather because they

are useful in the consumption of other products. Demands for these products are derived from the demands from other

products. Take motor oil, for example. Few people derive satisfaction from purchasing oil for their automobiles.

Rather, it is a derived demand from consuming transportation services provided by their cars. Procter & Gamble

(P&G) discovered that spraying a bit of their Clean Shower bathroom cleaning product on a razor each day can extend

the razor’s life three or four times. They are formulating a product targeted to this use. Thus, this new product’s

demand is derived from the demand for razor blades.

12E. Ramsted (1999), “As Prices Tumble, Sales of DVD Players Explode for the Holidays,” The Wall Street Journal (December 9), 31.

13Next, we extend this discussion to show how cross elasticities can be used by managers to define a firm’s


Chapter 4 Demand 133

�0.81: For every 1 percent increase in meal prices over this range there is, on average, a 0.81 percent decline in ticket sales. This elasticity suggests that PTC tickets and restaurant meals are rather strong complements.


Normal versus Inferior Goods Another factor that frequently affects the demand for a product is the income of po- tential buyers. As a person’s income increases, more products are purchased, and the combined expenditures across all products rise. The demand for specific products,


Estimates of Cross Elasticities Economists have estimated the cross elasticities for various commodities. Below are a few of these estimates:

Electricity and natural gas � 0.20

Beef and pork � 0.28

Natural gas and fuel oil � 0.44

Margarine and butter � 0.81

All the pairs of commodities listed above are substitutes. Complements such as DVD players and DVD movies have

negative cross elasticities. Natural gas apparently is not a very strong substitute for electricity. Although people can use

either gas or electricity for heating, natural gas is not generally used for lighting. On the other hand, natural gas and fuel

oil are closer substitutes (both tend to be used for heating). Margarine and butter are strong substitutes. Note that the cross

elasticities for some goods depend on how much time consumers have to adjust to changes in relative prices. For example,

the cross elasticity between natural gas and fuel oil is likely to be much higher in the long run, since it takes time to

replace old furnaces and other equipment with newer equipment powered by the less-expensive fuel source. In contrast,

customers can adjust to a relative change in the prices of margarine and butter on their next trips to the grocery store.

Source: E. Mansfield and G. Yohe (2004), Microeconomics (W. W. Norton: New York), 135.


Cross Elasticity and Pfizer’s Lost Profits in Japan Pfizer, the world’s largest pharmaceutical company, produces the leading drug to help people quit smoking. It is

marketed under the brand name of Chantix in the United States and Champix in Japan. According to the Wall Street Journal, Pfizer lost millions of dollars of potential sales by underestimating the cross elasticity of demand between

Champix and cigarettes. On October 1, 2011 the price of cigarettes increased significantly in Japan due to a

government tax. Tens of thousands of people, who wanted to quit smoking, went to their doctors for prescriptions of

Champix. Unfortunately, Pfizer had produced far too little of the drug to meet the demand. Clinic directors were upset

and argued, “they should have predicted something like this.” After all, Pfizer had known about the planned tax

increase for over a year. A Pfizer spokesman responded to these complaints by stating, “we expected more demand, but

not to this extent.” Whether or not Pfizer could have predicted the very large cross elasticity between Champix and

cigarettes with available data and information remains an open question. But had they done so, they would have

increased their revenues and profits by potentially millions of Yen.

Source: H. Tabuchi (2011), “In Japan, Pfizer is Short of Drug to Help Smokers, The Wall Street Journal (January 3).

however, can either rise or fall as income increases. The demand for goods such as gourmet foods or jewelry would be expected to increase with income, whereas the demand for other goods like canned processed meat or cabbage might decline. Goods for which demand increases with income are called normal goods. PTC tick- ets are normal goods. Goods for which demand declines with income are called inferior goods.

Income Elasticities The sensitivity of demand to income is measured by their income elasticity. The in- come elasticity is defined as the percentage change in the demand for a good, given a percentage change in income (I). Income elasticities can be calculated using the following formula.

�I � [�Q �(Q1 � Q2)�2] � [�I�(I1 � I2)�2] (4.9)

The income elasticity is positive for normal goods and negative for inferior goods. The income elasticity of a firm’s product has important implications. Firms pro-

ducing products with high income elasticities are more affected by cyclical fluctua- tions; they tend to grow more rapidly in expanding economies but contract more sharply in depressed economies. Managers must anticipate these fluctuations in managing cash flows and hiring decisions. Demands for products with small income elasticities are more stable over economic cycles. Studies indicate that goods like do- mestic servants, medical care, education for children, and restaurant meals tend to have relatively large income elasticities, whereas goods such as most food products, gasoline, oil, and liquor have relatively small (in absolute value) income elasticities.

Income elasticities also can influence location decisions. For instance, PTC has a relatively high-income elasticity (above 1.6). This elasticity was one of the factors that motivated the founders to locate their theater in a community with a high per capita income. They anticipated that they would have fewer customers if they located in a less affluent area.

134 Part 2 Managerial Economics


Estimates of Income Elasticities Economists have estimated the income elasticities for various products. Below are a few of these estimates:

Flour � �0.36

Margarine � �0.20

Milk � 0.07

Meat � 0.35

Dentist services � 1.41

Restaurant consumption � 1.48

According to these estimates, flour and margarine are inferior goods. People spend less on these goods as their

incomes rise. The other goods are normal goods (expenditures on the products rise with income). Dentist services and

restaurant consumption are particularly sensitive to income changes.

Source: E. Mansfield and G. Yohe (2004), Microeconomics (W. W. Norton: New York), 135.

Chapter 4 Demand 135

Other Variables

We have concentrated on three of the more important independent variables in most demand functions—the product’s own price, prices of related products, and income. Other variables, such as advertising expenditures, also can be important. In all cases, the analysis is similar. Demand responds to a change in some other variable. Sensi- tivity can be measured by the appropriate elasticity—for instance, an advertising elasticity. Obviously, managers do not have the time to consider all the conceivable variables that might have trivial impacts on the demand for their products. Good de- cision making requires that managers understand the effects of the more important factors, which usually include the product’s own price, the prices of close substitutes and important complements, and incomes.

Industry versus Firm Demand Industry Demand Curves Although we have concentrated our analysis on firm-level demand, demand func- tions and demand curves can be defined for entire industries. For instance, a demand function could be specified for the entertainment industry in PTC’s market area. Such a function would relate the total ticket sales for all entertainment events to fac- tors that affect this demand. Managers often are interested in total industry demand because it provides important information on the size of their potential markets and trends that affect them. For instance, a company’s executives might judge the perfor- mance of a store manager that reports flat sales quite differently if market demand is shrinking and the store is increasing market share versus a case where market de- mand is increasing but the store is losing market share. Moreover, estimates of in- dustry demand sometimes can be obtained at modest cost from outside analysts or business publications.

Firms within an industry compete directly and their products are likely to be rel- atively strong substitutes. The overall industry, on the other hand, is less likely to have strong substitutes. A person wanting to go to an entertainment event might


A Pampered Dog Loses His Stylist Even wealthy consumers are affected by changes in income. Consider Betsy Illium, a marketing consultant to medical

practices, who owns four Manhattan apartments (three are investment properties). In early 2008, Illium became

concerned that her income would decline given forecasts of a looming economic recession. “It’s frightening,” she said,

with much of her wealth being tied up in real estate. In response she decided to replace her dog Dobbin’s regular

groomer, who charged $130, with one from Petco, which charged only $65. She also decided to send her dirty sheets

and towels to a laundry service rather than to a higher priced dry cleaner. In her case, high-priced dog groomers and

dry cleaning are normal goods (her demands for these products increase with income), while laundry services and low-

priced dog groomers are inferior goods (her demands for these products decreased with income). Illium became

“appalled when she calculated that Dobbin’s grooming, her own weekly hair, nail and massage appointments, gourmet

groceries, restaurant meals and Starbuck’s coffee cost nearly $2,000 a month.” She realized that she would have to

“tighten her alligator belt” due to the threat of recession.

Source: S. Rosenbloom (2008), “Tightening the Alligator Belt,” (January 27).

choose among several options based on price. Entertainment events more broadly defined have fewer alternatives. Thus, demands facing individual firms within an industry tend to be more price-elastic than those for the entire industry.

Defining Industry and Market Area We have indicated that managers can gain important insights by analyzing industry- level demand. One problem that managers face in conducting this type of analysis is defining the relevant industry and market area. Is PTC competing in the live theater

136 Part 2 Managerial Economics


9/11 Causes Massive Shifts in Demand Curves The 9/11 terrorist attacks on the World Trade Center and Pentagon caused the demand curves for numerous goods and

services to shift. Consider the following examples of industries whose demand curves shifted left:

• Hotels lowered room rates by 30 to 40 percent. Marriott Hotels put a third of its hourly employees on part-time


• Disney trimmed the hours at its theme parks and reduced the hours of 40,000 part-time workers.

• The U.S. airline industry estimated that 9/11 caused it to lose about $7 billion and saw its revenues drop by

40 percent. Following 9/11 many planes flew at less than 50 percent capacity and airlines laid off 20 percent of

their workforce. The U.S. government provided $15 billion of loan guarantees to the airline industry.

• Spot prices for crude oil plummeted to a two-year low of $20 per barrel.

But not all firms saw the demand for their goods and services adversely affected by 9/11. The following firms saw their

demand curves shift to the right:

• Sales of American flags skyrocketed. Annin & Co., a 675-employee flag maker and market leader, increased

output of its most popular 3-by-5-foot U.S. flags from 30,000 to 100,000 a week by October 2001. Its five

plants added shifts, but dealers still had to wait 15 weeks for delivery of popular-size flags.

• The demand for biometric systems increased. Biometric systems identify people using digital face, iris, or

fingerprint scans. The biometric trade association doubled its worldwide revenue growth forecasts following

9/11. The three largest publicly traded biometric firms’ stock prices rose on average 70 percent when the stock

markets reopened on 9/17.

• InVision Technologies, which makes bomb-detection systems for airports that cost between $750,000 and

$1.5 million each, saw its stock price increase 165 percent on the first trading session after the markets


• The U.S. government increased the defense department’s budget by $10 to $15 billion. A significant percentage

of those funds were expected to be spent on intelligence technologies, such as those produced by software

companies like Narus, which track and sort data and e-mail on the Internet to identify potential terrorists.

On average, publicly traded companies suffered significant reductions in demand. For instance, the stock market

suffered its worst one-week loss since the Great Depression; following 9�11 stocks lost $1.2 trillion in market value.

Nonetheless, there were winners scattered among the larger population of losers.

Source: E. Brown (2001), “Heartbreak Hotel?” Fortune (November 26), 161–163; M. Gunther (2001), “The Wary World of Disney,”

Fortune (October 15), 104; K. Marron (2002), “Systems That Use Physical Traits to Control Access to Sensitive Data Are Catching On

in Post-Sept. 11 Era,” The Globe and Mail (March 28), B16; B. O’Keefe (2001), “Securing the Air, One Bag at a Time,” Fortune (October 15), 244; B. Powell (2001), “The Economy under Siege,” Fortune (October 15), 87–108; J. Simons (2001), “Greed Meets

Terror,” Fortune (October 29), 145–146; S. Tully (2001), “From Bad to Worse,” Fortune (October 15), 118–128; D. Voreacos (2001),

“As Country Wraps Itself in Flags, Company Strains to Make Them,” Houston Chronicle (October 7), 10; M. Warner (2001), “Web

Warriors,” Fortune (October 15), 148.

Chapter 4 Demand 137

14Cross elasticities also are used as evidence in antitrust cases. Antitrust cases generally focus on whether or

not a company has significant market power within an industry. Thus, the definition of the industry is quite

important. A company might have a significant market share (and thus apparent power) in a narrowly

defined industry, but a small market share in a more broadly defined market. For instance, the government

suggested that the ReaLemon Company had monopolized the reconstituted concentrated lemon juice

market—supplying over 90 percent of that market. The company responded that the appropriate market

definition was broader: They faced vigorous competition from reconstituted natural-strength lemon juice,

fresh lemons, frozen lemonade, lime juice, and so on.


Store Layout Affects Demand Paco Underhill calls himself a “retail anthropologist.” His consulting firm videotapes consumers as they shop at his

clients’ stores such as Sears, The Gap, and McDonald’s. He then offers recommendations for store layout. For

example, most North Americans turn right after entering a store while most British and Australian customers turn left.

Consumers tend to avoid narrow aisles; they apparently dislike being jostled from behind (what he calls the “butt-brush

factor”). Junk food should be placed on low or middle shelves so kids can reach them. After finding that women spend

only half the time in the store when accompanied by a man, he recommends placing numerous chairs around stores so

men can sit comfortably while the women shop.

Source: K. Labich (1999), “Attention Shoppers: This Man Is Watching You,” Fortune (July 19), 131–133.

industry or in a more broadly defined entertainment industry? Cross elasticities pro- vide important information to answer these types of questions. The cross elasticity be- tween PTC tickets and symphony tickets is 0.4. This relatively high value (see the box titled Estimates of Cross Elasticities presented earlier in this chapter) suggests that PTC competes against companies in a broader entertainment industry than just live theater.14 The managers at PTC also must define the relevant geographic area of their marketplace. If PTC raises its prices, will its customers shift to theaters in other nearby cities? If so, these cities should be included in the definition of PTC’s market area.

Network Effects For some products, demand increases with the number of users. For example, fax machines and telephones are not particularly useful unless there is a network of users. This consideration is quite important for many of today’s communication and information products. For instance, consumers were reluctant to buy new products, such as 3-D television sets, Blu-ray players, and new word processing programs, until they became convinced that the products had the potential for widespread use. Consumers understand that if a product does not garner sufficient demand, important complementary products, such as Blu-ray movies, will not be produced in high volumes or at attractive prices. Also they worry that it will be dif- ficult to acquire reliable, inexpensive service for the new product. They may learn to use a new technology only to find that it is discontinued because of insufficient demand. For example, despite vigorous efforts by Sony to promote its Betamax technology for video recording, the technology was displaced completely by the VHS format. VHS suffered a similar fate from DVDs and Blu-ray.

Products where these network concerns are important often have relatively elastic demands. When price is lowered, there are two effects. One is the standard price

effect: Consumers purchase more of the product because it is being sold at a more at- tractive price. The second is the network effect: Demand for the product increases even more because more people are using the product. When a new consumer purchases the product, there is an externality for other users; because there is an additional user of the product, the product becomes more attractive for other current and potential users.

Network effects have important implications not only on product pricing, but on product design as well. For instance, when software manufacturers are designing software upgrades, they have to decide whether to make the new product compatible with prior versions of the software and with competing products on the market. Mak- ing a new product compatible with competing products can reduce the uniqueness of the product. However, the net effect can be to increase overall demand for the prod- uct because of network effects. For example, a consumer might be more willing to buy an Apple computer if it is compatible with Windows computers because of the enhanced ability to interact with Windows users.

Product Attributes Thus far, we have taken the attributes of the product as given. Our analysis of the de- mand for PTC tickets is based on the existing quality and selection of plays, their starting times, the quality of seating, and so on. Given these characteristics, we ex- amined how price and other factors affect the demand for PTC tickets.

Understanding consumer demand also plays an important role in the design of the product. For instance, do local patrons prefer Shakespeare or more contemporary

138 Part 2 Managerial Economics


Demand Elasticity for Gasoline The industry-level demand for gasoline is relatively inelastic: The price of gasoline can change substantially and have

little effect on the overall quantity demanded. The demand elasticities facing individual gas stations are much larger. If

several gas stations are located at the same intersection, an individual station can suffer a remarkable loss of business to

its local competitors by raising its price.


eBay and Network Effects The largest online auction site, eBay, is one of the few thriving dot-coms. Recently it has been growing at 72 percent

per year. It had reported profits of $90 million in 2001 on revenues of $749 million and attracted 37 million customers

per quarter. Over a million sellers offer products on eBay. Many sellers work full-time selling products over eBay; for

example, Angie Cash sells $10,000 a month of items that cost no more than $20 each. eBay does lots of things right

such as clever e-mails, message boards, and a self-monitoring system whereby buyers and sellers rate each other’s

performance. eBay is a trading platform that exploits network effects. The demand for eBay increases as the number of

buyers and sellers increases. More sellers increase the chance that buyers will find what they want, and more buyers

increase the chance sellers will sell their products at higher prices.

Source: E. Brown(2002), “How Can a Dot-Com Be This Hot?” Fortune (January 21), 78–84.

Chapter 4 Demand 139

plays? Do they prefer mysteries or musicals? Do they value comfortable seating with additional leg room or seating that is closer to the stage? Can the anticipated de- crease in demand from increased restaurant prices be offset by changing the starting time of the plays? (Delaying the starting time by 30 minutes might give people more time to eat at home before going to the play.)

Answers to these types of questions are important in managerial decision making and establishing corporate strategy. Indeed, when managers speak of the importance of understanding consumer demand, they often are referring to understanding the specific product attributes that are important to customers. Marketing managers are responsible for understanding the broad range of product attributes that affect de- mand. These include price, product design, packaging, promotion and advertising, and distribution channels.15 This broad focus on demand has played an especially important role in management innovations like total quality management programs (see Web Chapter 23, available via Connect).

An important problem facing most firms is how to incorporate information that may be held by many people throughout a firm—for example, about such matters as consumer demand—into the decision making process for product design. We defer discussions of this problem until the last half of this book. These sections provide in- sights into how to design the firm’s organizational architecture to help ensure that relevant information is incorporated in the decision making process.

Product Life Cycles Our discussion of product attributes suggests that managers constantly seek to de- velop new and better ways to identify and respond to consumer demands. This ac- tivity leads to the introduction of new products. Managers generally recognize that market demand for a new product is unlikely to remain stable over time. Often, the

15For a more formal economic analysis of the demand for product attributes, see K. Lancaster (1966), “A New

Approach to Consumer Theory,” Journal of Political Economy 74, 132–157.


Understanding What Consumers Want Innovative companies strive to develop new products that will be demanded by customers. Managers, however, vary in

their beliefs about the best way to discover consumer demand. Steve Jobs, former CEO of Apple Inc., thought that

consumer surveys and other market research were relatively useless, since consumers often don’t know whether they

will purchase a new product until they see it. Many industry “experts” were highly skeptical about the potential demand

for iPads when the product was first announced. They criticized Jobs for not conducing market research prior to the

launch. Jobs, however, strongly believed that consumers would want to purchase the product once they experienced it,

and he was right. Henry Ford, the founder of Ford Motor expressed this same belief about market research when he

famously said, “If I had asked people what they wanted, the would have said faster horses.” Jim Goodnight, one of the

founders and long–time CEO of SAS, strongly believed that it is critical for his company to obtain ideas and suggestions

from customers about potential new software products. SAS has successfully developed an expanding range of new and

upgraded software products for nearly four decades. In 2013, it employed 13,764 people and had sales of over $3 billion.

While these CEOs differ in their beliefs about the value of market research, all three were highly successful in managing

their companies and in introducing new products that were highly demanded by customers.

Source: W. Isaacson (2011), Steve Jobs (Simon & Schuster) and SAS Overview and Annual Report 2013.

industry demand curve for a new product shifts outward as the product becomes more widely known. Eventually, however, the demand is likely to shift inward as consumers shift toward other new and improved products. This pattern in the de- mand for new products is known as the product life cycle.

As depicted in Figure 4.6, the product-life-cycle hypothesis suggests that the de- mand for a product can be categorized into four main phases: introduction, growth, maturity, and decline. In the growth phase, the industry-level demand increases rapidly. In the maturity phase, the demand continues to increase and then begins to decrease. In the decline phase, the demand continues to fall. Eventually, the product is withdrawn from the market. Managers should recognize these trends in new- product planning, as well as in entry, exit, and pricing decisions for given products.

The increase in demand during the growth phase encourages new firms to enter the industry. For instance, the growth in the demand for personal computers (PCs) during the 1990s prompted many firms to begin production. Given the entry of new firms, original firms typically lose market share. If industry demand grows at a faster rate than the number of firms, existing firms realize sales growth, even though their market

140 Part 2 Managerial Economics

In du

st ry

q ua

nt ity

o f o

ut pu


Time T

Introduction Growth Maturity Decline

Product life cycle

QFigure 4.6 Product Life Cycle

The product-life-cycle hypothesis suggests that the industry demand for a new product goes through four main phases: introduction, growth, maturity, and decline.


Demand for Prostitutes Declines in the United States Over Time The demand for goods and services frequently changes over time. Popular media opinion is that the demand for

prostitutes, “the world’s oldest profession,” is booming in the United States. The evidence, however, suggests

otherwise. The sexual mores are much more liberal today than they were a century ago. It is generally much easier for

men to find premarital “hookups.” The increased supply of noncommercial sex appears to have reduced the demand for

prostitutes. In 1948, 69 percent of men surveyed indicated that they had paid for sex at least once, while in 2006 the

percentage was only 15. Consistent with basic supply and demand analysis, the decline in demand has been

accompanied by a decrease in the price paid to street prostitutes. The average annual income for prostitutes is

estimated to have declined from $25,000 per year to $18,000 per year (both in 2007 dollars) over the period


Source: (2014) “Sex, Lies and Statistics,” The Economist (March 22).

Chapter 4 Demand 141

share falls. If the number of new firms grows faster than industry demand, existing firms will experience a reduction in demand, depressing prices and firm profits.

This discussion suggests that the first firms to introduce a successful product sometimes can have “first-mover advantages.” In this case, they enjoy high profits until competitive entry occurs. They also can develop a customer base and have a longer time period to learn how to produce the product efficiently. These advantages explain why firms frequently strive to be the first to develop and launch new products. However, in attempting to exploit an innovative product, managers must anticipate the impact that their policies are likely to have on entry decisions by potential rivals.

The analysis also suggests that managers should be careful in evaluating whether to enter an industry during its growth phase. Competition during this phase can be intense; moreover, the demand they face is expected to decline at some point in the future. To prosper in such an environment, a new firm must have some type of competitive advantage over its rivals (e.g., being a low-cost producer).

Demand Estimation16

In our PTC example, we know the demand function. Most managers are not so lucky: They must estimate their demand functions. Sometimes it is easy to estimate demand, at least for the very near term. Other times it is quite difficult. Some com- panies employ statistical techniques to provide numerical estimates of demand func- tions. Other companies use more qualitative approaches.

Demand estimation is a complex topic that is largely beyond the scope of this book. Here, we simply provide a brief discussion of three general techniques used in estimating demand: interviews, price experimentation, and statistical analysis. Our intent is to provide insights into the basic costs and benefits of each approach. These insights make managers more informed consumers of demand estimates and offer guidance as to the type of demand analysis to employ in a given situation. Although each approach has its limitations, the approaches are not mutually exclusive. Be- cause the limitations differ, many managers employ several methods and aggregate the estimates to increase their understanding of demand.


Interview approaches attempt to estimate demand through customer surveys, ques- tionnaires, and focus groups. Perhaps the most naive version of this approach is simply to ask consumers what they would purchase if faced with different prices. The answers to these questions can be remarkably unreliable. First, people have in- centives to be less than completely truthful since customers would like the firm to offer lower prices. Second, even if they try to be truthful, they might have difficulty forecasting what they would actually purchase given the array of alternatives avail- able in the marketplace.

More sophisticated approaches to customer interviews are possible. For example, an individual might be asked about the difference in price between two competing products. Now if you found that individuals had purchased one of the products but

16This section draws on W. Baumol (1977), Economic Theory and Operations Analysis (Prentice Hall:

Englewood Cliffs, NJ), 234–236.

did not know the price of the other, you might conclude that customers were rela- tively insensitive to price.

Sometimes companies use a simulated market where people are given play money and asked to simulate purchase decisions. These experiments can yield useful in- sights. Again, however, the decisions people make with play money need not mirror the decisions they would make with their own money.

Consumer surveys play a particularly important role in providing information about the attributes that are valued by customers. Many businesses request that buy- ers fill out customer-service and complaint forms. Businesses often follow up sales or service with telephone calls to customers to ask about product and service quality and customer satisfaction. Among the most important sources of information about customer preferences are the direct contacts that salespeople and other company rep- resentatives have with their customers.

All the interview approaches, however, can produce remarkably inaccurate in- formation if the sample is not representative of the population of the firm’s cus- tomers. For instance, if you are interested in estimating demand for a good with a negative income elasticity, distributing surveys at an upscale mall might be a poor way to proceed. More subtle problems with eliciting interview information also can arise. One team of researchers cautions, “The curious, the exhibitionistic, and the succorant are likely to overpopulate any sample of volunteers. How secure a base can volunteers be with such groups over-represented and the shy, suspicious, and inhibited under-represented?”17

Price Experimentation

A second approach is to undertake price experiments. For instance, the board might decrease PTC’s ticket price to $25 and carefully track changes in ticket sales. However as part of the company’s marketing strategy, PTC prints brochures that detail the sea- son’s plays, costs, dates, and ticket prices. Thus, experimenting with their ticket prices would require reprinting their brochures. This raises the cost to PTC of this type of price experimentation. Some other types of firms incur few costs in changing prices; for instance, it is particularly easy for companies that market through the Internet to experiment with their prices.

Many firms are unlike PTC in that they operate at multiple locations. If a firm has the flexibility to vary prices across different geographic markets, it has the potential to gain more information than if it is limited to experimenting at a single location. But care must be exercised. Ideally, the local markets are separated geographically and have their own media outlets. Thus advertising lower prices in one market will not shift demand from the firm’s other locations.

There are at least three limitations in the use of price experimentation. First, de- mand can differ, depending on whether customers anticipate that a price change will be permanent or temporary. It can be difficult to identify customers’ expectations about future prices. Second, direct-market tests are not controlled experiments; sev- eral changes might be occurring simultaneously. For instance, the board might lower PTC’s ticket prices at the same time that the symphony changes its prices. The observed change in demand would reflect both effects. Third, some managers worry

142 Part 2 Managerial Economics

17E. Webb, D. Campbell, R. Schwartz, and L. Schrest (1966), Unobtrusive Measures (Rand McNally College

Publishing Company: Chicago).

Chapter 4 Demand 143

On Estimating Demand Curves for Common Stocks There has been a long-running debate over the demand elasticities of common stocks of individual firms. Many

economists argue that these demand curves are perfectly elastic, since there are numerous stocks with similar risk-

return characteristics available in the market. In this case, the demand curves for individual stocks are horizontal.

Others argue that each stock is unique and has very few substitutes. Here, the individual demand curves would be


Managers care about the slopes of the demand curves for their common stock since these slopes affect the price

at which they can sell new securities. If demand curves slope downward, price must be decreased below the current market

price to sell new securities. If demand curves are horizontal, new securities can be issued at the current market price.

Managers, of course, want to sell new stock at the highest possible price.

The existing empirical evidence suggests that stock prices decline by about 3 percent when firms announce new

issues of common stock. This finding seems to suggest that the demand curves for common stocks are downward-

sloping. This finding, however, is subject to alternative interpretations. If investors think that managers tend to issue

new stock when they believe it is overvalued, an announcement of a new issue will cause the entire demand curve to

shift down and price will decline (since investors infer from the announcement that the firm is overvalued). The

observation that price declines when new stock is issued is not sufficient to allow us to identify the price elasticity of a

firm’s common stock—the price decrease might be due to either a shift in demand or a shift in quantity demanded.

This example illustrates that it is not always easy to estimate demand curves, even when extensive data on prices

and quantities are readily available. Indeed, the data on prices and volumes for publicly traded securities are among the

best available in the world.

Source: C. Smith (1986), “Investment Banking and the Capital Acquisition Process,” Journal of Financial Economics 15, 3–29.


that price experimentation is risky. They are concerned that customers lost as a result of a price increase might be difficult to regain even if subsequently the price were lowered. Alternatively, it might be difficult to raise the price once a firm had lowered it (customers might be annoyed and purchase from rivals).

Statistical Analysis

Often, companies use statistical techniques such as regression analysis to estimate demand functions. Computers and large databases on sales, prices, and other relevant factors have increased the usefulness of this approach materially. By using statistical techniques, the effects of specific factors often can be isolated. It is possible to ana- lyze large samples of actual market data to obtain more reliable results.

Even though statistical approaches can provide managers with important infor- mation on demand, they must recognize that there are potential problems. Just be- cause a researcher can produce reams of computer output formatted into tables and multicolored graphs implies neither that the analysis is well done nor that the results are reliable. Below, we briefly discuss three types of problems that managers en- counter regularly in statistical approaches to estimating demand.

Omission of Important Variables The problem of omitted variables can be illustrated by an example. Assume that the actual demand function for a company is

Sales � 120 � 2P � 8A � 0.04I (4.10)

where P is the price of the product A is advertising expenditures I is income

Table 4.1 presents the data for 2006, 2007, and 2008. While this data is potentially available to the marketing manager, Brendis Isaccsdottir, who wants to estimate de- mand, does not necessarily know that both advertising and income are important de- terminants of demand. Suppose that Brendis ignores income and uses statistical techniques to estimate a relation between sales and advertising.18 Standard regres- sion techniques would yield the following equation:

Sales � 140 � 48A (4.11)

The model appears to predict sales perfectly (based on the data in the table). The equation, however, materially overstates the true influence of advertising and can lead to spectacular mistakes in decision making. Based on this analysis, Brendis might budget far too much for advertising. This omitted-variables problem is present when- ever important excluded variables are correlated with explanatory variables that are included in the statistical analysis.19 Including unimportant variables does not bias es- timated coefficients for the other variables (however, including irrelevant variables re- duces the precision of the various estimates).

Multicollinearity If the factors that affect demand are highly correlated (tend to move together), it might be impossible to estimate their individual effects with much precision. For in- stance, two important variables in the demand function might be income and educa- tion. If in the data set to be analyzed high income is always associated with high ed- ucation, it might be impossible to separate the two effects.

144 Part 2 Managerial Economics

2006 2007 2008

Income (I) $3,000 $4,000 $3,500 Advertising (A) 2 3 2.5 Price (P) 10 10 10 Sales (S) 236 284 260 True demand S � 120 � 2P � 8A � 0.04I Estimated demand S � 140 � 48A

Table 4.1 An Example of the Omitted-Variables Problem

The true demand curve of the company in this example is Sales � 120 � 2P � 8A � 0.04I. The data for 2006 to 2008 are presented in the table. If the analyst omits income and uses statistical techniques to estimate a relation between advertising and sales, the analyst will obtain the following equation: Sales � 140 � 48A. The model predicts sales perfectly (based on the data in the table). The estimated equation, however, significantly overstates the influence of advertising. The omitted-variables problem is present whenever important variables are left out of the analysis that are correlated with the explanatory variables that are included in the analysis.

18The manager does not have to worry about controlling for price, since it was constant over the period ($10). 19The problem does not always result in overstated coefficients on the explanatory variables. Depending on the

nature of the correlation among the explanatory variables, the coefficients can be either overstated or

understated. The estimated coefficient in this example is overstated because advertising and income are

positively correlated.

Chapter 4 Demand 145

Identification Problem Another potentially important problem that can confront Brendis is the identification problem. This problem also can be illustrated by example. Suppose the marketing manager has collected data on past prices and sales for a given industry with the aim of estimating an industry demand curve. In the past three years, the following sales and price combinations have been observed: (10, $10), (12, $8), and (14, $6). Is it valid for Brendis to connect these three points as an estimate of the demand curve? Because of the identification problem, the answer is generally no.

Each data combination reflects the intersection of the demand curve and supply curve for the industry for each year. If the demand curve has shifted over the three years due to changes in factors such as personal income, the points come from three different demand curves. Connecting the points does not provide an estimate of the current demand curve. In fact, if supply considerations have been stable while de- mand has shifted, it will trace out the industry’s supply curve. Suppose in our exam- ple that both the demand and supply curves have shifted in each year. As shown in Figure 4.7, the resulting combinations of price and quantity are observed equilibrium points, given the conditions during the relevant time periods. Connecting these points provides a poor estimate of the current demand curve D3.

Sometimes, Brendis will not have enough information to solve the identification problem and is better off using consumer interviews or market experiments to esti- mate demand. Other times, she has enough information to identify the demand func- tion (she needs to be able to specify factors that influence demand, but not supply, and vice versa). One special case in which Brendis does not have to worry about the identification problem is when the demand curve is stable. Suppose the demand curve did not shift over the three years and all the different sales-price combinations were caused by changes in supply. In this case, she can obtain a reasonable estimate of the demand curve simply by connecting the observed sales-price combinations.










Q Q2 Q3



Estimated demand

P ric

e (in

d ol

la rs



Figure 4.7 An Example of the Identification Problem

An analyst has collected data on past prices and sales for her firm’s industry. The demand and supply curves have shifted over the three years. Connecting the three price-quantity points provides a poor estimate of the current industry demand curve (labeled D3 in the graph). The three points are equilibrium points, given all conditions that affect the demand and supply of the product at each point in time. They are not three points along the same demand (or supply) curve.

146 Part 2 Managerial Economics


We have discussed some of the difficulties that managers face in trying to estimate the demand for their product. These problems can be difficult to solve. Nonetheless, estimates of demand play a critical role in decision making—especially the pricing decision. Successful managers address these problems the best they can, given im- perfect knowledge and limited resources.


Personal video recorders (PVRs) are digital video

recorders used to record and replay television

programs received from cable, satellite, or local

broadcasts. But unlike VCRs, which they replace,

PVRs offer many more functions, notably the abil-

ity to record up to 3,000 hours of programs and

easy programming. A PVR consists of an internal

hard disk and microprocessor. After the owner in-

stalls the hardware, the PVR downloads all upcom-

ing TV schedules to the hardware via a phone or

cable connection. Users merely enter the name of

the show(s) they want recorded and the system

finds the time and channel of the show and

automatically records it. Users must subscribe to a

cable or satellite system if they wish to record

programs off these channels.

Besides ease of programming and much larger

recording capacity than videotape, PVRs allow the

user to watch a prerecorded show while the unit is

recording up to five new programs, pause watching

live programs (e.g., if the phone rings) and then re-

sume watching the rest of the live broadcast, view

instant replays and slow motion of live programs,

and skip commercials. In effect, PVRs, like older

VCRs, allow viewers to control when they watch

broadcast programs (called “time shifting”). How-

ever, PVRs provide much sharper pictures and are

much simpler to operate than VCRs, and PVRs

allow the user to download the television schedule

for the next week.

Two companies that begin selling PVRs and

subscription services were: TiVo and ReplayTV.

Both firms started in 1997. As of 2013 TiVo had

about one million subscribers and ReplayTV had

been purchased by DirectTV. Companies are devel-

oping new technologies that make it even easier for

users to “snip” commercials.

Cable companies now offer a combined cable

box and PVR in one unit for a small additional

monthly charge. This further simplifies setup and

operation, and the user gets a single bill.

1. Discuss how PVRs will affect the demand

from advertisers?

2. Suppose you are in charge of setting the price

for commercial advertisements shown during

Enemies, a top network television show. There

is a 60-minute slot for the show. However, the

running time for the show itself is only 30 min-

utes. The rest of the time can be sold to other

companies to advertise their products or do-

nated for public service announcements. De-

mand for advertising is given by:

Qd � 30 � 0.0002P � 26V

where Qd � quantity demanded for advertising

on the show (minutes), P � the price per

minute that you charge for advertising, and V is the number of viewers expected to watch the

advertisements (in millions).

a. All your costs are fixed and your goal is to

maximize the total revenue received from

selling advertising. Suppose that the ex-

pected number of viewers is one million

people. What price should you charge?

How many minutes of advertising will you

sell? What is total revenue?

b. Suppose price is held constant at the value

from part (a). What will happen to the

quantity demanded if due to PVRs the num-

ber of expected viewers falls to 0.5 million?

Calculate the “viewer elasticity” based on

the two points. Explain in words what this

value means.

Chapter 4 Demand 147

3. As more viewers begin using PVRs, what hap-

pens to the revenues of the major networks

(CBS, NBC, ABC, and FOX)?

4. Discuss the long-run effects if a significant

proportion of the viewers begin adopting these

“advertising snipping” systems.

5. What advice would you give the major

commercial networks and producers of

programming for these networks as more

consumers adopt PVRs?

Source: J. Gudmundsen (2002), “Video Gizmos Change the Rules,” Democrat and Chronicle (August), 5E and 8E; B. Fisher (2003), “TiVo and Replay TV Have Features to Satisfy Any TV Junkie,” Detroit News (June 24); R. Reilly (2003), “Great Invention Period,” Sports Illustrated (December 22).

Summary Understanding product demand is critical for many managerial decisions such as pricing, setting production levels, undertaking capital investment, and establishing an advertising budget. This chapter provides a basic analysis of demand.

A demand function is a mathematical representation of the relations among the quantity demanded of a product over a specified time period and the various factors that influence this quantity. We focus on three independent variables in the demand function: the price of the product, the prices of related products, and customers’ incomes.

A demand curve for a product displays how many units will be purchased over a given period at each price holding all other factors fixed. Movements along a de- mand curve reflect changes in price and are called changes in the quantity de- manded. Movements of the entire demand curve are caused by other factors, such as changes in income, and are referred to as changes in demand.

Demand curves generally slope downward to the right: Quantity demanded varies inversely with price. This relation often is referred to as the law of demand. Demand curves vary in their sensitivities of the quantity demanded to price. Price elasticity is defined as the percentage change in quantity demanded from a percent- age change in price (expressed as a positive number). The price elasticity tends to be high when there are close substitutes for the product and when the good repre- sents a significant expenditure for the consumer. Demand tends to be more elastic over the long run than over the short run. How total revenue from a product changes with price depends on the price elasticity. A small price increase results in an in- crease in expenditures when demand is inelastic and a decrease in expenditures when demand is elastic. Total expenditures remain unchanged when the demand elasticity is unitary.

An important concept in economics is marginal revenue, which is defined as the change in total revenue given a one-unit change in quantity. Marginal revenue for a linear demand curve is given by the line with the same intercept as the demand curve but with twice the negative slope. Total revenue increases with quantity when marginal revenue is positive and decreases with quantity when marginal revenue is negative.

The price of related products can affect the demand for a product. Goods that compete with each other are referred to as substitutes. Products that tend to be con- sumed together are complements. One frequently used measure of substitution be- tween two products is the cross elasticity of demand. The cross elasticity is positive for substitutes and negative for complements.

148 Part 2 Managerial Economics

Another factor that can affect the demand for a product is the income of potential buyers. The sensitivity of demand to income is measured by the income elasticity. The income elasticity is positive for normal goods, and negative for inferior goods.

Demand curves can be defined for individual firms or entire industries. The price elasticities for individual firms within an industry are generally higher than for the industry as a whole. Cross elasticities can be helpful in defining the appropriate industry.

For some products, demand increases with the number of users. For example, telephones are not very useful unless there is a network of users. Products where these network concerns are important often have relatively elastic demands. When price is lowered, there is both a standard price effect and a network effect.

The standard economic analysis of demand takes the attributes of the product as given. Information about consumer demand, however, is also important in the initial design of products. In Chapters 11–23 of this book, we provide important insights into how to design the firm’s organizational architecture to help ensure that this type of information is incorporated in the decision making process.

Managers use three basic approaches to estimate demand: interviews, price exper- imentation, and statistical analysis. All three approaches can suffer from potentially serious problems. Managers have to do the best they can given imperfect information and limited resources. Knowledge of the potential pitfalls can make managers more intelligent producers and users of demand estimates.

G. Stigler (1987), The Theory of Price (Macmillan: New York), Chapter 3.

4–1. Suppose Product A has the demand function QA � 10 � 5PA � 2PB � 0.01I. The initial

values of the variables are QA � 15, PA � $4, PB � $2.5 and I � $2,000.

a. When PA moves to $3.4, keeping other variables at their initial values, QA becomes 18.

What is the corresponding own-price arc elasticity of demand?

b. If income, I, increases to $2,250 per period with all other variables held at their ini-

tial values, QA becomes 17.5. What is the corresponding income arc elasticity of de-


c. If PB increases to $3 with all other variables held at their initial values, QA becomes 16.

What is the corresponding cross-price arc elasticity of demand?

d. Is Product A an inferior or normal good? Are Product A and Product B substitutes or

complements? Explain.

e. Is the firm charging the revenue maximizing price for Product A at the initial values?


f. Compute the MR at the initial values.

4–2. Suppose your firm faces a demand curve of P � 90 � .30Q. Find the revenue maximizing

output and price. Calculate the total revenue. Is this outcome on the elastic, inelastic, or uni-

tary elastic part of the demand curve? Is this price the optimal price for your firm to charge?

Display this choice graphically (showing the demand and marginal revenue curves).

4–3. The BJC Company has the following demand function:

Q � 300 � 30(price) � 0.01(income)

Currently, price is $5 and income is $20,000.

a. Calculate the point elasticities for price and income.

b. Is the product a normal or an inferior good?

Self-Evaluation Problems

Suggested Reading

Chapter 4 Demand 149

4–4. Last year, Americans bought 5,000 Ferraris. The average retail price of a Ferrari was

$100,000. Statistical studies have shown that the price elasticity of demand is 0.4. Assume

the demand curve is linear. Estimate it using the above information.

a. Is demand elastic or inelastic?

b. What will happen to revenue if the company raises its price?

Solutions to Self-Evaluation Problems

4–1. a.



d. Product A is normal because an increase in income leads to an increase in demand for the

product. Products A and B are substitutes because the cross-price elasticity is positive.

Consumers substitute away from B and purchase more A when the price of B increases.

e. No, the firm is not charging the revenue maximizing price for Product A. Revenue maxi-

mization occurs at the price/quantity where MR � 0 and the corresponding own-price

elasticity is one. But here we see that � 1.12 � 1. Therefore, the current price is

higher than the revenue maximizing price, while the current quantity is lower than the

revenue maximizing quantity.

f. Start with the original demand function, QA � 10 � 5PA � 2PB � 0.01I and insert the

initial values for PB and I. Solve for PA. This produces the demand curve for Product A:

PA � 7 � .2Q. The MR function has the same intercept at the demand curve, but twice

the negative slope: MR � 7 � .4Q. At the initial quantity of 15, MR � $1. Producing an

additional unit increases revenue by $1.

4–2. The demand curve is given by P � 90 � .30Q. The marginal revenue curve is MR � 90 � .60Q. Revenue maximization occurs where MR � 0. Thus the revenue maximizing quan-

tity, Q*, is

90 � .60Q* � 0

Q* � 150.

The revenue maximizing price is found by inserting Q* into the demand function:

P* � 90 � (.3 � 150) � $45.

Revenue maximization occurs at the midpoint of the demand curve, where the price elastic-

ity is equal to one (unitary elasticity). The goal of the firm is to maximize profits, not rev-

enue. Thus the firm is not at the optimal price unless the marginal cost of production � $0.

Graphically the analysis is pictured as follows:








Unitary price elasticity




QA =

(16 - 15)

(3 - 2.5) *

2.75 15.5


1 0.5


2.75 15.5

= 0.354



(17.5 - 15)

(2250 - 2000) *

2125 16.25


2.5 250


2125 16.25

= 1.307


QA = Absolute Value

- (18 - 15)

(3.4 - 4) *

3.7 16.5


3 0.6


3.7 16.5

= 1.12

150 Part 2 Managerial Economics

4–3. a. The current quantity demanded is 350. Thus, the price elasticity is 30(5�350) � .429.

The income elasticity is .01($20,000�350) � .571.

b. The product is a normal good (income elasticity is positive).

4–4. If the demand curve is linear, it must take the form P � a � bQ. To estimate a and b we use

the information provided. The elasticity of demand can be expressed as:

� � abs [(�1�b) � (P�Q )]

We are told that at P � 100,000 and Q � 5,000, the elasticity of demand equals 0.4. Substi-

tuting the known values, solve for b :

0.4 � (1�b) � (100,000�5,000)

b � 50

100,000 � a � 50 � 5,000

a � 350,000

4–1. What is the difference between a demand function and a demand curve?

4–2. How will each of the following affect the position of the demand curve for DVD players?

a. An increase in the price of DVD movies.

b. A decrease in the price of DVD players.

c. An increase in per capita income.

d. A decrease in the price of movie tickets.

4–3. If the demand for a product is inelastic, what will happen to total revenue if price is increased?


4–4. What are the signs of cross elasticities for substitute products? Explain.

4–5. Distinguish between normal and inferior goods.

4–6. Is it true that a normal good must have an income elasticity that is more than one? Explain.

4–7. Suppose that the price of Product A falls from $20 to $15. In response, the quantity demanded

of A increases from 100 to 120 units. The quantity demanded for Product B increases from

200 to 300. Calculate the arc cross elasticity between Product B and Product A. Is B a substi-

tute or complement for A? Explain. Does Product A follow the “law of demand”? Explain.

4–8. How can cross elasticities be used to help define the relevant firms in an industry?

4–9. Suppose the price of heating oil increases significantly. Discuss the likely short-run and

long-run effects.

4–10. The Alexander Machine Tool Company faces a linear demand curve. Currently, it is selling

at a price and quantity where its demand elasticity is 1.5. Consultants have suggested that

the company expand output because it is facing an elastic demand curve. Do you agree with

this recommendation?

4–11. For three years in a row, income among consumers has increased. Alexander Machine

Tool has had sales increases in each of these three years. Does Alexander Machine Tool

produce inferior or normal goods? Forecasts predict that income will continue to rise in

the future. Should Alexander Machine Tool anticipate that demand for its products will

continue to rise? Explain.

4–12. The cross elasticity between Product A and Product B is 10. Do you think that Product A is

likely to face an elastic or inelastic demand curve? Explain.

4–13. Vijay Bhattacharya is interested in estimating the industry demand curve for a particular

product. He has gathered data on historical prices and quantities sold in the industry. He

knows that the industry supply curve has been stable over the entire period. He is consider-

ing estimating a regression between price and quantity and using the result as an estimate of

the demand curve. Do you think this technique will result in a good estimate of the demand

curve? Explain.

Review Questions

Chapter 4 Demand 151

4–14. Maria Tejada, a civil engineer, uses data on population trends to forecast the use of a partic-

ular highway. Her forecasts indicate severe road congestion by the year 2010. She suggests

building a new road. Comment on this approach.

4–15. Alexander Machine Tool faces the demand curve P � $70 � 0.001Q. What price and quan-

tity maximize total revenue? What is the price elasticity at this point?

4–16. Studies indicate that the income elasticity of demand for servants in the United States ex-

ceeds 1. Nevertheless, the number of servants has been decreasing during the last 75 years,

while incomes have risen significantly. How can these facts be reconciled?

4–17. Prior to a price increase, the price and quantity demanded for a product were $10 and 100,

respectively. After the price increase, they were $12 and 90.

a. Calculate the arc elasticity of demand.

b. Is the demand elastic or inelastic over this region?

c. What happened to total revenue?

4–18. Define marginal revenue. Explain why marginal revenue is less than price when demand

curves slope downward.

4–19. In 1991, Rochester, New York, had a serious ice storm. Electric power was out in houses for

days. The demand for power generators increased dramatically. Yet the local merchants did

not increase their prices, even though they could have sold the units for substantially higher

prices. Why do you think the merchants adopted this policy?

4–20. Seven teenagers, four boys and three girls, were given $200 each to go on a shopping spree.

An advertising agency, which specializes in youth markets, gave the teens the money. An

account executive accompanied the teens while they were shopping. The agency wanted to

learn not only what they bought, but also what they talked about to see what was on their

minds. “It’s not so much to stay in tune with trends, because trends are elusive. It’s more

what’s really happening with teens and what’s important to them.”20

a. Discuss the trade-offs between sample size (7 teens), cost, and reliability of what is

learned from this experiment.

b. An agent accompanied the teens while they were shopping. Why didn’t the ad agency

avoid this expense and just look at what the teens bought?

4–21. Southwest Airlines estimates the short-run price elasticity of business air travel to be 2 and

the long-run elasticity to be 5. Is ticket demand more elastic in the short run or long run?

Does this seem reasonable? Explain.

4–22. Gasoline prices increased substantially in 2004 and 2005. What adjustments did people

make to minimize the long-term effects of this price increase?

4–23. Assume that demand for product A can be expressed as QA � 500 � 5PA � 3PB and de-

mand for product B can be expressed as QB � 300 � 2PB � PA. Currently, market prices

and quantities for these goods are PA � 5, PB � 2,QA � 481, and QB � 301.

a. Suppose the price of product B increases to 3. What happens to the quantity demanded

of both products?

b. Calculate the arc cross-elasticity between product A and product B using prices for

product B of 2 and 3.

c. Are these goods substitutes or complements?

4–24. The Zenvox Television Company faces a demand function for its products that can be

expressed as Q � 4,000 � P � 0.5I, where Q is the number of televisions, P is the price

per television, and I is average monthly income. Average monthly income is currently equal

to $2,000. Answer the following questions.

a. Graph the demand curve (sometimes called the “inverse” demand curve) faced by Zenvox

at the current income level. Be sure to label this and all graphs you draw carefully. On the

same graph, depict marginal revenue. At what price and quantity is Zenvox’s total revenue

maximized? What is the marginal revenue at this point? Show the calculation.

20“Teens Track Retail Trends for Ad Agency (1999),” Democrat and Chronicle (September 5), 1E.

152 Part 2 Managerial Economics

b. What is the price elasticity of Zenvox’s demand function at the price and quantity de-

rived in part (a)? Explain what this value means in words.

c. Why might Zenvox choose to produce at a price and quantity different than that derived

in part (a)?

4–25. According to an article in Forbes (March 2001) teen cigarette smoking declined signifi-

cantly between 1975 and 2000. The most dramatic decline occurred in the years

1975–1981. Since then teen smoking has increased in some years and declined in others.

Between 1975 and 1981 there was a slight decrease in the price of cigarettes. Thus the dra-

matic decline in smoking is not attributable to an increase in cigarette prices. One theory is

that the significant increase in gasoline prices over this period motivated many teens not to


a. Discuss how a rise in gasoline prices might affect the demand for cigarettes among


b. Suppose there are two goods in the world, cigarettes and gasoline. Draw a figure that

shows how an increase in gasoline prices can result in a decline in both gasoline and

cigarette consumption. Use the standard consumer behavior graph with budget lines

and indifference curves. Be sure to label your figure appropriately.

c. In the late 1990s the price of cigarettes increased from $2.50 per pack to $3.25 per pack.

In one community during this time period, the number of packs of cigarettes consumed

by teenagers fell from 10,000 to 9,000. Assume that everything except cigarette prices

remained the same. Calculate the arc price elasticity among teens between these price


d. Calculate the total expenditures on cigarettes by the teens in part (c) both before and

after the price increase. Did total revenue increase or fall? Discuss how this answer is

implied by the arc elasticity that you calculated in part (c).

4–26. In an article appearing in the Dow Jones News Service on February 5, 2004, the agency

cites Saudi Arabia’s concern about the overproduction of oil by the OPEC cartel.

Assume the current daily demand for OPEC’s oil is given by the following equation:

P � 50 � 0.001Q

where P is the price per barrel (ppb) and Q is the quantity of barrels sold daily (in thou-

sands). Moreover, suppose the marginal cost of producing a barrel is constant at zero. a. Would it surprise you to learn that OPEC’s declared objective is to sell 25 million bar-

rels a day for an average price of $25 per barrel? Why or why not? Explain. You may

use a graph to support your argument.

b. Assume that after OPEC’s meeting this week, the new demand for OPEC oil will be

given by P � 40 � .001Q. Would OPEC’s stated objective (25 million barrels at an

overall price of $25) be attainable after this change? Explain. Assume OPEC ignores the

demand shift. What’s the maximum price per barrel they can charge if they decide to

keep producing 25 million barrels per day? What is the profit in this case?

c. Now suppose that OPEC recognizes that demand has changed (as in [b]) and wants to

maximize profits. What is the daily quantity they should supply? At what price? What

is the profit in this case? What is the price elasticity of demand at this price/quantity

combination? Explain.

4–27. As a result of strikes in Canada the world price of nickel rose by 20 percent in December.

Over the same period, the quantity demanded of nickel decreased from 10,000,000 to

8,500,000 pounds worldwide. The world price of nickel was 70 cents per pound before

the strikes.

a. Show graphically the effect of Canadian strikes on the market for nickel.

b. Given the information above, what’s the price elasticity of the world demand for nickel

over the relevant price range?

c. Did the total expenditure for nickel increase, decrease, or remain constant after the

strikes? How is this consistent with your answers to parts (a) and (b)? Explain clearly

and concisely.

Chapter 4 Demand 153

4–28. Assume the demand curve for gasoline is given by the equation

P � 10 � 0.0005Q,

where P is the price per gallon and Q is the quantity of gasoline in gallons. Assume that the

only supplier of gasoline in the region is General Gasoline Co. and that the marginal cost of

production is constant at zero.

a. If the company is currently charging $4 a gallon, is it maximizing profit? If so, prove

it. If not, find out the price that maximizes its profit, and compare the profits at the

two prices.

b. Discuss the likely effect of the introduction of a fuel-efficient car in the region; that

is, what would happen to the equilibrium quantity. Show the changes on a graph that

displays (you don’t need to show actual numbers) General Gasoline’s pricing solution

and explain.

4–29. The accompanying chart presents data on the price of fuel oil, the quantity demanded of

fuel oil, and the quantity demanded for insulation.

Fuel Oil Insulation

Quantity Demanded Quantity Demanded Price per Gallon (millions of gallons) (millions of tons)

$3 100 30 $5 90 35 $7 60 40

a. Calculate the price elasticity (arc elasticity) of demand for fuel oil as its price rises

from 30 to 50 cents; from 50 to 70 cents. Calculate the change in total revenue in

the two cases. Explain how the changes in revenue relate to your estimated elasticities.

b. Calculate the arc cross elasticity of demand for insulation as the price of fuel oil

rises from 50 to 70 cents. Are fuel oil and insulation substitutes or complements?


4–30. Japan has 4,350 miles of expressway—all toll roads. In fact, the tolls are so high that

many drivers avoid using expressways. A typical 3-hour expressway trip can cost $47. A

new $12 billion bridge over Tokyo Bay that takes 10 minutes and costs $25 rarely is busy.

One driver prefers snaking along Tokyo’s city streets for hours to save $32 in tolls.21

Assume that the daily demand curve for a particular stretch of expressway is:

P � 800 yen � .16Q

a. At what price-quantity point does this demand curve have a price elasticity of one?

b. Assume the government wishes to maximize its revenues from the expressway, what

price should it set? And how much revenue does it generate at this price?

c. Suppose that traffic engineers have determined that the efficient utilization of this par-

ticular toll road is 4,000 cars per day. This traffic level represents an optimum tradeoff

between congestion (with its associated reduction in speeds and increase in accidents)

between expressways and surface roads. If 4,000 cars per day is the socially efficient

utilization of the toll road, what price should be set on the toll road? And how much

revenue is collected by the government?

d. Which price, the one in part (b), or the one in part (a) would you expect the government

to set?

21J. Singer (2003), “Lonesome Highways: In Japan, Big Tolls Drive Cars Away,” The Wall Street Journal (September 15), A1 and A15.

154 Part 2 Managerial Economics

Appendix: In the chapter, we presented formulas for arc elasticities that estimate elasticities

Demand22 between two points on the demand curve. This appendix shows how to calculate elas- ticities at single points on the demand curve. It also derives the equation for marginal revenue for a linear demand curve and discusses a special (log-linear) demand function.

Point Elasticities Elasticities measure the percentage change in quantity demanded for a percentage change in some other variable. There are several ways to express the formula for an elasticity. One way, using price elasticity as an example, follows:

� � �(�Q�Q )�(�P�P ) (4.12)

� �(�Q��P ) � (P�Q )

By definition, as the change in P goes to zero, the limit of the first term (�Q��P ) is the partial derivative of Q with respect to P. At a particular point on the demand curve, the elasticity of demand for small changes in P is given by

� � �( Q� P ) � (P�Q ) (4.13)

As an example, consider the demand function for PTC theater tickets:

Q � 117 � 6.6P � 1.66PS � 3.3PR � 0.0066I (4.14)

The point elasticity at the current price-quantity combination of $30 and 200 tickets is

� � �(�6.6) � (30�200) � 1 (4.15)

Recall that this is the value that we derived graphically in the text (see Figure 4.5). Other point elasticities—for example, point cross elasticities—can be calculated

in a similar fashion. Simply substitute the appropriate variable (e.g., the price of another product) for P in Equation (4.13).

Marginal Revenue for Linear Demand Curves Marginal Revenue (MR) is the change in total revenue for an additional unit of quan- tity. As the change in quantity becomes very small, the limit of this definition is the partial derivative of total revenue with respect to Q.

Linear demand curves take the following form:

P � a � bQ (4.16)

Thus, total revenue, P � Q can be written as:

TR � (a � bQ ) � Q (4.17)

� aQ � bQ 2

Marginal revenue is

MR � TR� Q � a � 2bQ (4.18)

22This appendix requires elementary knowledge of calculus.

Chapter 4 Demand 155

This formula is a line that has the same intercept as the demand curve, but with twice the negative slope.

Marginal Revenue and Demand Elasticity In this section, we derive the relation between marginal revenue and demand elastic- ity. This relation is useful in a number of contexts. For example, it underlies a for- mula that we use in Chapter 7 to analyze how a firm’s optimal price markup over cost relates to the product’s demand elasticity.

By definition:

TR � PQ (4.19)

MR � TR� Q � ( � Q � Q ) � P

Multiply the quantity in parentheses by P�P :

MR � ( P� Q � Q ) � (P�P ) � P (4.20)

� ( P� Q � Q�P ) � P � P � P � [1 � 1��]

Equation (4.20) indicates that marginal revenue is equal to price when demand is perfectly elastic. In this case, the firm can sell one more unit at the market price with- out having to lower the price. MR falls as the elasticity decreases and is negative when demand is inelastic (� 1).

Log-Linear Demand Functions The following demand function is frequently used in empirical demand estimation:

Q � �P �I � (4.21)

where Q is the quantity demanded P is price I is income

(Other variables such as advertising and the price of other goods are commonly in- cluded as other explanatory variables.) An important property of this demand func- tion is that the price and income elasticities are constant (they do not vary along the demand function) and are equal to �� and �, respectively. In particular:

� � �( Q� P ) � (P�Q ) (4.22)

� �(��P ��1I� ) � (P��P �I�)

� ��

Similarly, � is the income elasticity. Taking the natural logarithm of the demand function in Equation (4.21) yields

lnQ � ln� � �lnP � �lnI (4.23)

This equation is linear in the logarithms; it thus can be estimated by standard regres- sion analysis using data on Q, P, and I. The estimated coefficients � and � are estimates of the price and income elasticities. Other types of elasticities—for example, cross elasticities—can be estimated by including other variables in the demand equation.


5 C H A P T E R O U T L I N E

Production Functions

Returns to Scale

Returns to a Factor

Choice of Inputs

Production Isoquants

Isocost Lines

Cost Minimization

Changes in Input Prices


Cost Curves

Short Run versus Long Run

Minimum Efficient Scale

Learning Curves

Economies of Scope

Profit Maximization

Factor Demand Curves

Cost Estimation


Appendix: The Factor- Balance Equation

T he global demand for steel has increased in recent years due in part to major construction projects in countries such as China and India. Construction crews were particularly busy in China on projects re- lated to the 2008 Beijing Olympics, the 2010 Shanghai World Ex-

position, and new housing. While the world production of steel also has in- creased, the net result has been an increase in steel prices from their historic levels. In the early 1990s, steel sold for under $100 per ton. In July 2014, the price was $740 per ton. This more than sevenfold increase in the price of steel far exceeds the rate of inflation, which was about 70 percent over the same period. Steel prices not only increased over this period but also dis- played high volatility. In spring 2002 the price was as low as $222, but by fall 2004 it had increased to above $700. It subsequently retreated to under $500 in late 2005, but by mid-2008 exceeded $1,200.1

Automobile manufacturers are a major user of steel. High steel prices provide strong incentives to these companies to find ways to mitigate the price increase, for example, by substituting away from steel toward rela- tively less expensive production materials. Consider their actions during the 1990s when steel prices began their recent assent. In 1994, domestic steel

Production and Cost

1Steel prices are from


1. Explain what is meant by a production function.

2. Distinguish between returns to scale and returns to a factor.

3. Create a graphical analysis of the cost-minimizing input mix and explain how it

is affected by changes in relative prices of inputs.

4. Define and describe the relationship among total, marginal, and average costs.

5. Describe the connection between production and cost functions.

6. Distinguish between short- and long-run costs curves.

7. Define fixed and variable costs and their role in decision making.

8. Explain long-run costs, sources of economies and diseconomies of scale and

scope, and the notion of minimum efficient scale.

9. Explain why MR � MC at the profit maximizing output.

10. Understand how many units of a factor (such as labor) a firm should purchase

at different factor prices.

Chapter 5 Production and Cost 157

prices increased as the U.S. economy recovered from a recession.2 Indeed, the steel market was the strongest it had been in 20 years: Specifically, steel prices had risen from below $90 to over $135 per ton between 1992 and 1994. After significant price increases earlier in the year, domestic steel companies were planning to increase sheet-steel prices by another 10 percent at year’s end. In the tight electrogalvanized markets, price increases as high as 20 percent were expected. (In fact, in 1995, prices exceeded $142 per ton.)

To counter the effects of the increase in domestic steel prices, U.S. auto compa- nies actively pursued new overseas suppliers. For instance, in July 1994, General Motors invited bids for sheet steel from foreign companies such as Sidmar, Solldac, Thyssen, and Klockner. The increases in steel prices affected both companies’ pric- ing and output decisions. The increases in steel prices also placed pressure on U.S. automakers to use other raw materials in the production process. For example, auto companies increased their use of aluminum in engines, transmissions, body com- ponents, heating and cooling systems, and suspension systems in 1995.3

(Aluminum prices had been relatively stable; they were $.534 per pound at the be- ginning of 1992 and $.533 in January 1994.) Potential applications focused on re- placing cast iron or steel with aluminum. In addition, auto companies increased re- search on new ways to use plastics, magnesium, and recyclable materials in their production process.

This example raises a number of questions that are of interest to managers. First, how do firms choose among substitutable inputs in the production process? How does the optimal input mix change with changes in the input prices? How do changes in input prices affect the ultimate cost of production and the output choices of firms? In this chapter we address these and related questions. Major topics in- clude production functions, choice of inputs, costs, profit maximization, cost esti- mation, and factor demand curves. In the appendix, we derive the factor-balance equation.

Production Functions A production function is a descriptive relation that links inputs with output. It speci- fies the maximum feasible output that can be produced for given amounts of inputs. Production functions are determined by the available technology. Production func- tions can be expressed mathematically. For instance, given current technology, an automobile supplier is able to transform inputs like steel, aluminum, plastics, and labor into finished auto parts. In its most general form, the production function is ex- pressed as

Q � f (x1, x2, . . . xn) (5.1)

where Q is the quantity produced and x1, x2, . . . xn are the various inputs used in the production process.

2Details of this example are from “General Motors Eyes Imports to Counter Price Increases,” Metal Bulletin (July 11, 1994), 21.

3A. Wrigley (1994), “Automotive Aluminum Use Climbing in 1995’s Models: Automotive Applications Will

Use Some 120 Million Lbs. in 1995,” American Metal Market (August 9), 1.

158 Part 2 Managerial Economics

To simplify the exposition, suppose that the auto part in this example is pro- duced from just two inputs—steel and aluminum. An example of a specific produc- tion function4 in this context is

Q � S1�2A1�2 (5.2)

where S is pounds of steel, A is pounds of aluminum, and Q is the number of auto parts produced.

With this production function, 100 pounds of steel and 100 pounds of aluminum will produce 100 auto parts over the relevant time period, 400 pounds of steel and 100 pounds of aluminum will produce 200 auto parts, and so on.5

Returns to Scale

The term returns to scale refers to the relation between output and the proportional variation of all inputs taken together. With constant returns to scale, a 1 percent change in all inputs results in a 1 percent change in output. For example, Equation (5.2) presents a production function with constant returns to scale. If the firm uses 100 pounds of each input, it produces 100 auto parts. If the firm increases both in- puts by 1 percent to 101 pounds, it produces 101 auto parts.6

With increasing returns to scale, a 1 percent change in all inputs results in a greater than 1 percent change in output. An example of such a production function is

Q � S A (5.3)

Here, 100 pounds of steel and 100 pounds of aluminum produce 10,000 auto parts, while 101 pounds of steel and aluminum produce 10,201 auto parts (a 2 percent in- crease in output). Firms often experience increasing returns to scale over at least some range of output. One major reason is that a firm operating on a larger scale can engage in more extensive specialization. For instance, if an automobile company has only three employees and three machines, each employee and each machine has to


Increasing Returns to Scale Motivates Amazon to Invest in Large Warehouses The Law of Increased Dimensions (the “Container Principle”) implies that the doubling of the height and width of a

container leads to more than a proportionate increase in cubic capacity (volume). The corresponding increasing returns

to scale have led to invest in huge warehouses that are capable of storing hundreds of thousands of items.

As of 2014, Amazon owned or leased over 84 million square feet of space for housing its fulfillment and data centers.

Its largest fulfillment centers are over 1.2 million square feet in size. These large centers are located near major

airports, which are used in the shipping of orders. The resulting cost efficiencies deriving from the size and location of

these centers have contributed to Amazon’s performance.

Source: (2014), Form 10-K for Fiscal Year Ended December 31, 2013 (January 31)

4This production function is an example of a Cobb-Douglas production function, which takes the general

form Q � �S �A�. Cobb-Douglas production functions are used frequently in empirical estimation. Not all

firms, however, have production processes that are well described by this particular type of production

function. 51001�2 � 1001�2 � 10 � 10 � 100, and 4001�2 � 1001�2 � 20 � 10 � 200. 6[(100 � 1.01)1�2] � [(100 � 1.01)1�2] � 101.

perform a myriad of tasks for the company to produce automobiles. Given the broad array of tasks that each worker and machine has to perform, efficiency is likely to be low. In contrast, a large firm employing thousands of workers and machines can en- gage in much greater specialization. (As noted in Chapter 3, specialization often pro- duces efficiency gains.)

With decreasing returns to scale, a 1 percent change in all inputs results in a less than 1 percent change in output. An example is

Q � S1�3A1�3 (5.4)

The likelihood that a firm will choose to operate where it experiences decreasing re- turns to scale is open to debate. Some economists argue that firms should seldom dis- play decreasing returns to scale. If a facility of a given size can produce a given out- put, why can’t the firm simply replicate that facility and produce twice the output with twice the inputs? Indeed, most empirical studies on the subject suggest that the typical firm initially experiences increasing returns to scale, followed by constant re- turns to scale over a quite broad range of output. On the other hand, several empiri- cal studies indicate that some firms probably do experience decreasing returns to scale.7 Also, casual observation suggests that some larger firms suffer from ineffi- ciencies to a greater extent than do smaller firms—for example, coordination and control problems become more severe as a firm becomes larger.

In our examples, the returns to scale are the same over all ranges of output. For instance, Equation (5.2) always displays constant returns to scale, while Equation (5.4) always displays decreasing returns. Most production functions vary in returns to scale over the range of output. Most frequently, production functions have in- creasing returns to scale when output is relatively low, followed by constant returns to scale as output continues to increase, and possibly decreasing returns to scale when output is high. Other combinations are possible.

Returns to a Factor

Returns to a factor refers to the relation between output and the variation in a single input, holding other inputs fixed. Returns to a factor can be expressed as total, mar- ginal, or average quantities. The total product of an input is the schedule of output obtained as that input increases, holding other inputs fixed. The marginal product of an input is the change in total output associated with a one-unit change in the input, holding other inputs fixed. Finally, the average product is the total product divided by the number of units of the input employed.

To illustrate these concepts, consider the production function in Equation (5.2): Q � S1�2A1�2. Table 5.1 presents the total, marginal, and average product of S, holding A fixed at 9.8 For this production function, total product increases as S increases; marginal product, however, declines. This means that although total product increases with S, it does so at a decreasing rate. Average product also decreases over the entire range.

7For example, E. Berndt, A. Friedlaender, and J. Chiang (1990), “Interdependent Pricing and Markup

Behavior: An Empirical Analysis of GM, Ford, and Chrysler,” working paper, National Bureau of Economic

Research, Cambridge, MA. 8The production function assumes that production does not have to take place in discrete units. For instance,

output might be expressed in tons; clearly production in fractions of tons is possible.

Chapter 5 Production and Cost 159

160 Part 2 Managerial Economics

More generally, marginal and average products do not have to decline over the en- tire range of output. Indeed, many production functions display increasing marginal and average products over some ranges. However, most production functions reach a point after which the marginal product of an input declines. This observation often is called the law of diminishing returns (or law of diminishing marginal product), which states that the marginal product of a variable factor eventually will decline as its use is increased. To illustrate this principle, consider the classic example of farm- ing a plot of land. Land is fixed at 1 acre, and no output can be harvested without any workers. If 10 bushels of grain can be produced by one worker, the marginal product of the first unit of labor is 10 bushels. The change in output might be even greater as the firm moves from one to two workers. For instance, two workers might be able to produce 25 bushels of grain by working together and specializing in various tasks. The marginal product of labor is 15 bushels and thus, over this range, marginal

Units Units Total Product Marginal Product Average Product of S of A of S of S of S

1 9 3.00 3.00 3.00 2 9 4.24 1.24 2.12 3 9 5.20 0.96 1.73 4 9 6.00 0.80 1.50 5 9 6.70 0.70 1.34

Table 5.1 Returns to a Factor

This table shows the total, marginal, and average products of steel for the production function Q � S1�2 A1�2. Aluminum is held fixed at 9 units. The total product of S is the total output for each level of S; the marginal product of S is the incremental output from one additional unit of S; and the average product of S is output divided by the total units of S.


Studying for an Exam—the Law of Diminishing Returns Your performance on the CPA exam depends on both your effort and aptitude in the subject. Your aptitude is largely a

fixed input (e.g., it is hard to increase your basic IQ). Effort on the other hand is something you can control and vary.

For example, if you are preparing for the CPA exam, you could spend many or few hours reading books and working

problems. The choice is yours. If you exert no effort, you are unlikely to do well on the exam and easily could fail. If

you study, your performance is likely to improve. Initially, as you begin to allocate additional hours to studying for the

exam, your rate of improvement might be quite large. For instance, you might expect to increase your exam score by

20 points if you study one versus no hours. As you continue to spend more time on studying for the exam, your rate of

improvement (marginal product of effort) is likely to decline—you will reach a point of diminishing returns. Indeed at

some point your exam score could decline with additional effort as you become too tired to take the exam—the

marginal product of effort becomes negative. Basic economics says that you should study for an exam up to the point

where the incremental benefits of studying additional time are equal to the incremental costs. The law of diminishing

returns implies that the incremental benefits will eventually become smaller and possibly negative. The incremental

costs of studying an additional hour depend on your opportunity cost of time—for what else could you use the time? It

is unlikely to be optimal for you to spend additional time studying for the CPA exam if that causes you to neglect your

other professional responsibilities.

product is increasing. Eventually, as the firm continues to add more workers, while holding land fixed, output will grow at a slower rate. At some point, total output might actually decline with additional workers because of coordination or conges- tion problems. In this case, the marginal product is negative.

Figure 5.1 illustrates returns to a factor in this common case. The upper panel dis- plays total product, and the lower panel displays marginal and average products. As the use of input S goes from zero to S1, marginal product rises. Over this range total product is convex—total product increases at an increasing rate.9 At S1, diminishing returns set in and the marginal product begins to fall. Between S1 and S2, marginal product is positive and so total product continues to increase. However, it does so at a decreasing rate (the curve is concave). Beyond S2, marginal product is negative, hence total output falls with increases in S. Average product is rising where marginal product is above average product and is falling where marginal product is below

Marginal product

Average product





Total product


Q ua

nt ity

o f a

ut o

pa rt

s Q

ua nt

ity o

f a ut

o pa

rt s

(p er

u ni

t o f s

te el


Quantity of steel


Figure 5.1 Returns to a Factor: A Common Case

This figure illustrates a common pattern for total product, marginal product, and returns to a factor. In the lower panel, marginal product rises, then falls, and eventually becomes negative. When marginal product is rising (between zero and S1), total product increases at an increasing rate (the curve is convex) in the lower panel. When marginal product is falling but positive (between S1 and S2), total product continues to increase but does so at a decreasing rate. Beyond S2, marginal product is negative and total product falls with additional output. Average product is rising where it is below marginal product and is falling where it is above marginal product. Average and marginal products are equal where average product is at a maximum.

9Technical note: The marginal product at a point is equal to the slope of the total product curve at that point

(MP � �TP/�S). Thus, marginal product is decreasing when the total cost curve is concave and increasing

when it is convex.

Chapter 5 Production and Cost 161

162 Part 2 Managerial Economics

average product. Marginal product and average product are equal where average product is at its maximum.10 This relation is a general rule.11 The accompanying box on baseball batting averages illustrates the intuition behind this relation.

Choice of Inputs

Production Isoquants

Most production functions allow some substitution among inputs. For example, sup- pose that Alexi Dyachenko is chief operating officer, managing a firm with the pro- duction function Q � S1�2 A1�2, and he wants to produce 100 auto parts. In this case, there are many different combinations of steel and aluminum that will yield 100 auto parts. For instance, 100 auto parts can be produced using 100 pounds of steel and 100 pounds of aluminum, 25 pounds of steel and 400 pounds of aluminum, or 400 pounds of steel and 25 pounds of aluminum. Figure 5.2 displays all the possible combinations of inputs that can be used to produce exactly 100 auto parts. Obvi- ously, 100 auto parts also could be produced with more inputs—points above or to the right of a point on this isoquant—but those points represent inefficient production methods. This curve is called an isoquant (iso, meaning the same, and quant from quantity). An isoquant shows all input combinations that produce the same quantity assuming efficient production. There is a different isoquant for each possible level of production. Figure 5.2 shows the isoquants for 100, 200, and 300 auto parts.

Production functions vary in terms of how easily inputs can be substituted one for another. In some cases, no substitution is possible. Suppose that in order to produce 100 auto parts you must have 100 pounds of aluminum and 100 pounds of steel, to produce 200 auto parts you must have 200 pounds of aluminum and 200 pounds of steel, and so on. Having extra steel or aluminum without the other metal yields no ad- ditional output—they must be used in fixed proportions. As shown in Figure 5.3, iso- quants from fixed-proportion production functions are shaped as right angles. At the other extreme are perfect substitutes: The inputs can be substituted freely one for an- other. Suppose that one auto part always can be produced using either 2 pounds of steel or 2 pounds of aluminum. In this case, the firm can produce 100 auto parts by using either 200 pounds of aluminum or 200 pounds of steel, or any combination in


Baseball Batting Averages Marginal product is above average product when average product is rising and below average product when average

product is falling. This relation is a general property of marginals and averages. A useful illustration is a baseball

player’s batting average. The batting average is defined as the number of hits divided by the number of times at bat.

Suppose a player starts a game with an average of .300. If the player gets two hits with four at bats, the marginal

batting average for the day is .500 and the player’s batting average must rise. If the player gets one hit with four at bats,

the marginal is .250 and the overall average must drop.

10Graphically, marginal product is the slope of a line drawn tangent to the total product curve of that level of

output; average product is the slope of the line connecting a point on the total product curve with the origin. 11Averages and marginals also are equal when the average is at a minimum.

between. As shown in Figure 5.3, the corresponding isoquant is a straight line. Most production technologies imply isoquants that are between these two extremes. As depicted in Figure 5.3, typical isoquants have curvature, but are not right angles. The degree of substitutability of the inputs is reflected in the curvature: The closer the isoquant is to a right angle (the more convex), the lower the degree of substitutability.

Generally, isoquants are convex to the origin (as pictured in the center panel in Figure 5.3—the typical case). Convexity implies that the substitutability of one input

Q ua

nt ity

o f a

lu m

in um

Quantity of steel





AFigure 5.2 Isoquants

An isoquant displays all possible ways to produce a given quantity. There is a different isoquant for each possible level of production. This figure shows the isoquants for 100, 200, and 300 auto parts for the production function Q � S1�2A1�2.

Q ua

nt ity

o f a

lu m

in um

Fixed proportions



Quantity of steel



Perfect substitutes

Quantity of steel



Normal case

Quantity of steel

Figure 5.3 Isoquants for Fixed Proportion Production Functions, Perfect Substitutes, and the Normal Case

Production functions vary in terms of how easily inputs can be substituted for one another. In some cases, inputs must be used in fixed proportions and no substitution is possible. Here, isoquants take the shape of right angles. At the other extreme are perfect substitutes, where the inputs can be freely substituted for one another. Here, isoquants are straight lines. Most production functions have isoquants that are between the two extremes. The isoquants in the normal case have curvature but are not right angles.

Chapter 5 Production and Cost 163

164 Part 2 Managerial Economics

for another declines as less of the first input is used. In our example, if the firm is using a large quantity of steel and little aluminum, it can eliminate a relatively large quantity of steel with the addition of only a small quantity of aluminum while keep- ing output the same (see Figure 5.2). In this case, aluminum would be much better suited than steel to construct some components of the auto part. But as the firm uses higher proportions of aluminum, its ability to substitute aluminum for steel declines: Steel is better suited for other components. Most production processes display this property.

Isocost Lines

Given that there are many ways to produce a given level of output, how does Alexi choose the most efficient input mix? The answer depends on the costs of the inputs. Suppose that the firm faces competitive input markets and can buy as much of each input as it wants at prevailing market prices. The price of steel is denoted Ps, whereas the price of aluminum is denoted Pa. Total cost (TC) is equal to the sum of the quan- tities of each input used in the production process times their respective prices. Thus,

TC � PsS � Pa A (5.5)

Isocost lines display all combinations of S and A with the same cost. Suppose Ps � $.50 per pound and Pa � $1 per pound, and the given cost level is $100. In this case,

$100 � $.50S � $A (5.6)

or equivalently,

A � 100 � 0.5S (5.7)

Figure 5.4 graphs this isocost line. Note that the intercept, 100, indicates how many pounds of aluminum could be purchased if the entire $100 were spent on aluminum. The slope of �0.5 is �1 times the ratio of the two prices (Ps�Pa): Since aluminum is twice as expensive as steel, 0.5 pounds of aluminum can be given up for 1 pound of steel and costs remain the same.

Holding the prices of the inputs constant, isocost lines for different cost levels are parallel. Figure 5.4 illustrates this property using the isocost lines for $100 and $200. Note that the further away the line is from the origin, the higher the total cost. Thus, holding output constant, the firm would like to be on the lowest possible isocost line.


Substitution of Inputs in Home Building Builders in the Pacific Northwest use large quantities of wood in the construction of residential houses. For instance,

wood is used for framing, siding, floors, roofs, and so on. Home builders in the Southwest (e.g., Arizona) use much

more stucco and tile in home construction. An important reason for this difference is that, in contrast to the Pacific

Northwest, the Southwest does not have large nearby forests. This example suggests that home builders are able to

substitute among inputs in building a home. Home builders in the Southwest, however, still use wood to frame the

house: The substitution of other inputs for wood is not complete.

The slope of an isocost line changes with changes in the ratio of the input prices. As depicted in Figure 5.5, if the price of steel increases to $1, the line becomes steeper (slope of �1). Here, the firm must give up 1 pound of aluminum to obtain 1 pound of steel. Alternatively, if the price of steel falls to $.25 (not depicted in the figure), the line becomes flatter (slope of �0.25). In this case, the firm has to give up only 0.25 pound of aluminum for every pound of steel. Similarly, the slope of the line also changes with changes in the price of aluminum. What determines the slope of the line are the relative prices (recall the slope is �Ps�Pa).

Cost Minimization

For any given level of output, Q*, Alexi will want to choose the input mix that min- imizes total costs. As shown in Figure 5.6, the cost-minimizing mix (S*, A*) occurs at the point of tangency between the isoquant for Q* with the isocost line. Alexi




$100 line $200 line


Quantity of steel

Q ua

nt ity

o f a

lu m

in um



Figure 5.4 Isocost Curves

Isocost lines display all combinations of inputs that cost the same. In this example, Ps � $.50 per pound and Pa � $1 per pound. The figure shows isocost lines for $100 and $200 of expenditures. The slope of an isocost line is �1 times the ratio of the input prices—in this example, �0.5. Isocost lines for different expenditure levels are parallel.



Ps = $1.00 per pound

Ps = $.50 per pound


Quantity of steel

Q ua

nt ity

o f a

lu m

in um



Figure 5.5 Isocost Lines and Changes in Input Prices

This figure depicts the effect of changes in input prices on the slopes of isocost lines. The solid line shows the isocost line when the price of aluminum is $1 and the price of steel is $.50. The dotted line shows the isocost line where the prices of both inputs are $1. Total cost in each case is $100.

Chapter 5 Production and Cost 165

166 Part 2 Managerial Economics

would like to produce the output less expensively (using an isocost line closer to the origin). However, lower-cost production is not feasible. Alexi could select other input mixes to produce Q*.12 But any other input mix would place the firm on a higher isocost line. Consider the combination (S�, A�) in Figure 5.6. This combina- tion of inputs also produces Q* units of output. Yet this output can be produced at a lower cost by using less aluminum and more steel.

In the appendix to this chapter, we show that at the optimal input mix, the follow- ing condition holds

MPs�Ps � MPa�Pa (5.8)

where MPi is the marginal product of input i. (Recall that the marginal product of an input is described in Table 5.1.) Condition (5.8) has a straightforward interpretation. The ratio of the marginal product to price indicates how much additional output can be obtained by spending an extra dollar on the input. At the optimal output mix this quantity must be the same across all inputs. Otherwise, it would be possible to in- crease output without increasing costs by reducing the use of inputs with low ratios and increasing the use of inputs with high ratios. For instance, if the ratio is 10 units per dollar for aluminum and 20 units per dollar for steel, the firm could hold costs constant but increase output by 10 units by spending one less dollar on aluminum and one more dollar on steel. Alexi has not chosen an optimal input mix when such sub- stitution is possible.



Q *

S �

A �


S *

Quantity of steel Q

ua nt

ity o

f a lu

m in


Figure 5.6 Cost Minimization

The input mix that minimizes the cost of producing any given output, Q*, occurs where an isocost line is tangent to the relevant isoquant. In this example, the tangency occurs at (S*, A*). The firm would prefer to be on an isocost line closer to the origin. However, the firm would not have sufficient resources to produce Q*. The firm could produce Q* using other input mixes, such as (S�, A�). However, the cost of production would increase.

12Note the similarity between this cost minimization problem and the consumer’s utility maximization

problem introduced in Chapter 2. The mathematics are the same—both are constrained optimization

problems. The consumer maximizes utility for a given budget. Cost minimization is equivalent to

maximizing output for a given budget (where the budget is that associated with the lowest-cost method of

producing the output).

Changes in Input Prices

An increase in the relative price of an input will motivate Alexi to use less of that input and more of other inputs. Figure 5.7 illustrates how the optimal input mix for produc- ing Q* changes as the price of steel increases: Alexi chooses less steel and more aluminum to produce the output. This effect is called the substitution effect. The strength of the substitution effect depends on the curvature of the isoquant. The


$15 per hour Minimum Wage Increases the Relative Cost of Labor at SeaTac Airport The minimum wage for hotel and car services employees working near SeaTac Airport (Seattle Washington) increased

to $15 per hour at the beginning of 2014, making it the highest in the country. The new minimum was more than twice

the Federal minimum of $7.25 per hour.

Economic theory predicts that SeaTac hotels and car rental agencies will respond to the higher relative price of

labor by substituting capital for labor. For example, a restaurant owner would have increased incentives to convert to

having customers order food with tablet computers located at each table (an increasingly common practice), while a

car rental agency would have increased incentives to use kiosks and other technologies for serving customers.

It is too early to assess all the effects this increase in the minimum wage will actually have on employment in the

area. It depends in part on the ability of companies to substitute capital for labor. Some local hotel managers have

stated that they are in a service industry and are highly unlikely to substitute capital for labor due to the higher wage

rate. The Clarion Hotel, however, within six weeks of the new law, had closed their full-service restaurant and was

considering replacing it with a less labor-intensive café.

Many people have been watching this “experiment” in increasing the minimum wage to such a high level. They

will have to wait to observe the long-run effects, since it takes time for businesses to convert to more capital-intensive

processes, such as using tablet computers for food ordering.

Source: A. Martinez (2014), “$15 Wage Floor Slowly Takes Hold in SeaTac,” Seattle Times (February 13).



Q *

High steel price

Low steel price


A* 1

S * 2 S *


Quantity of steel Q

ua nt

ity o

f a lu

m in


Figure 5.7 Optimal Input Mix and Changes in Input Prices

This figure illustrates how the optimal input mix for producing a given output, Q*, changes as the price of an input increases. In this example, the price of steel increases and the firm uses less steel and more aluminum to produce the output. This effect is called the substitution effect. The strength of the substitution effect depends on the curvature of the isoquant. The greater the curvature, the less the firm will substitute between the two inputs.

Chapter 5 Production and Cost 167

168 Part 2 Managerial Economics

greater the curvature, the less Alexi will substitute between the two inputs for any given change in prices. The substitution effect helps explain the reactions of automo- bile companies to the 1994 increases in domestic steel prices. These companies in- creased their use of foreign steel. They also searched for additional ways to replace steel with other inputs such as aluminum.

Costs We have analyzed how firms should choose their input mix to minimize costs of pro- duction. We now extend this analysis to focus more specifically on costs of produc- ing different levels of output. Analysis of these costs plays an important role in out- put and pricing decisions.

ANALYZING MANAGERIAL DECISIONS: Choosing the Mix of People and Machines in a Retail Supercenter

You manage a large retail supercenter that sells gro-

ceries and other products to 30,000 customers per

week. Currently, you employ 80 check-out clerks

and 10 automated check-out machines (customers

scan and pay for their purchases without a clerk’s

assistance). Each clerk is paid wages and fringe

benefits of $800 per week. It also costs you $800 per

week to lease each machine (price includes installa-

tion, software support, and servicing). A vendor

has offered to lease you additional machines at this

price. You estimate that by leasing 10 more ma-

chines you can meet your service requirements with

30 fewer clerks. Should you lease the additional ma-

chines or continue to service your customers with

your current input mix?

1. You conduct additional analysis and estimate

that you can service the 30,000 customers with

the following combinations of clerks and ma-

chines. Calculate the total costs for each of these

combinations. What combination of inputs

serves the customers at the lowest possible cost?

Clerks Machines

80 10 50 20 30 30 22 40 15 50 12 60

2. Plot the input combinations in the table on a

graph that contains clerks on the vertical axis

and machines on the horizontal axis. Connect

the points by lines to approximate an isoquant

as pictured in Figure 5.6. Add the cost mini-

mizing isocost curve to the graph (you can

derive this line from the input prices and the

total cost of the low-cost input combination).

How do the slopes of the isocost curve and

isoquant compare at the optimal input


3. Suppose that the marginal product of clerks

at the optimal input combination is 500.

Explain in words what this means. What is

the marginal product of machines at this

point? Explain why.

4. Suppose that the cost of leasing a machine

declines to $500 per week. What is your new

optimal input mix? How does this affect

your graph?

5. Are there any other factors that should be

considered in making this decision on the

optimal mix of machines and clerks

Discuss briefly.

Cost Curves

The total cost curve depicts the relation between total costs and output. Conceptu- ally, the total cost curve can be derived from the isoquant/isocost analysis discussed above. For each feasible level of output, there is a least-cost method of produc- tion—as depicted by the tangency between the isoquant and the isocost line. The total cost curve simply displays the cost of production associated with the isocost line and the corresponding output. For instance, if the least-cost method of produc- ing 100 auto parts is $1,000, one point on the total cost curve is (100, $1,000). If the least-cost method of producing 200 parts is $1,500, another point is (200, $1,500). Marginal cost is the change in total costs associated with a one-unit change in out- put. Average cost is total cost divided by total output. Managers sometimes refer to marginal cost as incremental cost, whereas they use the term unit cost to refer to average cost.

Figure 5.8 displays the total, marginal, and average cost curves for a hypothetical firm. (This figure illustrates a common pattern for cost curves, although not all firms have cost curves with this same shape.) The upper panel indicates that total cost increases with output. Between zero and Q1, total cost increases but at a decreasing rate (the curve is concave). As shown in the lower panel, over this range, marginal cost

Marginal cost





Average cost

Total cost

C os

t p er

u ni

t o f o

ut pu

t (in

d ol

la rs

) T

ot al

c os

ts (

in d

ol la

rs )

Quantity of output

Q1 Q2

Figure 5.8 Cost Curves

This figure displays the total, marginal, and average cost curves of a hypothetical firm. The upper panel pictures total cost. Total cost increases with output. Between zero and Q1, total cost increases but at a decreasing rate (the curve is concave). As shown in the lower panel, over this range, marginal cost decreases. Past Q1, total cost increases at an increasing rate (the curve is convex) and marginal cost increases. Average cost declines where marginal cost is below average cost and rises where marginal cost is above average cost. This relation is a general rule.

Chapter 5 Production and Cost 169

170 Part 2 Managerial Economics

decreases.13 Past Q1, total cost increases at an increasing rate (the total cost curve is convex) and marginal cost increases. Average cost is declining where marginal cost is below average cost and is rising where marginal cost is above average cost. Aver- age cost equals marginal cost where average cost is at its minimum point. As previ- ously discussed, these relations are general rules: They apply to average costs and average products, as well as batting averages and GPAs.

Production Functions and Cost Curves With constant input prices, the shapes of cost curves are determined by the underlying production function. For instance, if the production function displays increasing returns to scale over some range of output, long-run average cost must decline over that range. With increasing returns to scale, a 1 percent increase in input expenditures re- sults in a greater than 1 percent increase in output and average cost must fall. In con- trast, with decreasing returns to scale, a 1 percent increase in input expenditures results in a less than 1 percent increase in output and average cost must rise. Finally, constant returns to scale imply constant average cost. U-shaped curves (as pictured in Figure 5.8) normally are used to illustrate average costs. This slope suggests an initial region of increasing returns to scale, followed by decreasing returns to scale.14

There is also a direct link between the marginal cost curve and the underlying pro- duction function. Recall from Equation (5.8) that cost minimization requires the ratio of the marginal product to price to be equal across all inputs. For illustration, suppose at the optimal input mix to produce 100 auto parts, the ratio of the marginal product to price for both steel and aluminum is 2. By expending $1 more on either input, output increases by 2 units. The reciprocal of this ratio (1�2) is their marginal cost of producing one additional unit of output—if 2 units are produced with $1 of additional expenditure on inputs, the marginal cost of producing one extra unit is $.50. This example indicates that, holding input prices constant, marginal cost is de- termined by the marginal productivity of the inputs: The higher their marginal pro- ductivity, the lower the marginal cost. If the marginal productivities in our example were doubled, the ratio of the marginal product to price would be 4 and the marginal cost would be $.25. The inverse relation between marginal productivity and marginal cost makes intuitive sense. If with a given increase in inputs more output can be pro- duced, the marginal cost of producing that output is lower.

Input prices also can affect the shapes of the cost curves. For instance, a declining average cost can be motivated by discounts on large volume purchases. Similarly, a machine that produces 20,000 units might not be twice as expensive as a machine that produces only 10,000 units. Alternatively, if the firm bids up the price of inputs with large purchases, average cost can rise with increased output. Thus, the long-run average cost curve can slope upward even if the firm does not experience decreasing returns to scale.

13Technical note: The marginal cost at a point is the derivative of total cost (MC � �TC��Q). Graphically, it

is equal to the slope of the total cost curve at that point. Thus, marginal cost decreases when the total cost

curve is concave and increases when it is convex. The average cost curve is the slope of the line connecting

a point on the total cost curve with the origin. 14Some economists argue that the typical long-run average cost curve is flat to the right of its minimum

efficient scale. Once that output is reached, additional output can be produced at a constant average cost by

simply replicating the process (the production function does not experience decreasing returns to scale). But

this argument presumes that organizational costs do not increase disproportionally with firm size. See P.

McAfee and J. McMillan (1995), “Organizational Diseconomies of Scale,” Journal of Economics and Management Strategy 4:3, 399–426.

Opportunity Costs Managers must be careful to use the correct set of input prices in constructing cost curves. In Chapter 2, we defined opportunity cost as the value of a resource in its next best alternative use. Current market prices for inputs more accurately reflect opportu- nity costs than historical costs. For instance, if an auto supplier purchases 1,000 pounds of aluminum for $600 and subsequently the market price increases to $900, the oppor- tunity cost of using the aluminum is $900. If the company uses the aluminum, its re- placement cost is $900. Alternatively, the current inventory could be sold to another firm for $900. In either case, the firm forgoes $900 if it uses the aluminum in its pro- duction process.

The relevant costs for managerial decision making are opportunity costs. It is im- portant to include the opportunity costs of all inputs whether or not they have actually been purchased in the marketplace. For instance, if an owner spends time working in the firm, the opportunity cost is the value of the owner’s time in its next best alterna- tive use.

Short Run versus Long Run

Cost curves can be depicted for both the short run and the long run. The short run is the operating period during which at least one input (typically capital) is fixed in sup- ply. For instance, in the short run it might be infeasible to change plant size or change the number of machines. In the long run, the firm has complete flexibility—no inputs are fixed.

The definitions of short run and long run are not based on calendar time. The length of each period depends on how long it takes the firm to vary all inputs. For a cleaning- services firm operating out of rented office space, the short run is a relatively brief period—perhaps only a few days. For a large manufacturing firm with heavy invest- ments in long-lived specialized plant and equipment, the short run might be a rela- tively long time period—it might be a matter of years.

Short-run cost curves sometimes are called operating curves because they are used in making near-term production and pricing decisions. For these decisions, it often is appropriate to take the plant size and certain other factors as given (since these factors are beyond the control of the managers in the short term). Long-run cost curves frequently are referred to as planning curves, since they play a key role in longer-run planning decisions relating to plant size and equipment acquisitions.

Chapter 5 Production and Cost 171


Industry Responds to Higher Metals Prices Metals prices rose substantially between 2003 and 2004; for instance, the price of hot rolled steel increased by more

than 80 percent. Manufacturing firms limited the impact of these higher raw materials prices through improved

productivity and switching to less expensive substitutes. For example, some stainless steel makers began to use more

chromium and manganese and less nickel. Craig Yarde of Yarde Metals in Southington, CT, said that the run-up in

prices had benefited his company by increasing the market value of the 40 million pounds of metals in its inventory,

mostly aluminum and stainless steel.

Source: B. Simmon (2004), “Surge in Cost of Metal Squeezes Pricing and Profits,” New York Times (February 26), C1.

172 Part 2 Managerial Economics

Fixed and Variable Costs In the short run, some costs are fixed and do not vary with output. These fixed costs are incurred even if the firm produces no output. For instance, the firm has to pay managers’ salaries, interest on borrowed capital, lease payments, insurance premi- ums, and property taxes whether or not it produces any output. Variable costs change with the level of output. These costs include items like raw material, fuel, and certain labor costs. In the long run, all costs are variable.

Short-Run Cost Curves Figure 5.9 displays the short-run cost curves for the TAM Corporation. For this firm, suppose that the basic plant size is fixed and that all other inputs are variable. The upper panel depicts total cost. Total cost is the sum of the fixed cost (FC) and total variable cost (TVC). The shape of the total cost curve is completely determined by the shape of the total variable cost curve. Fixed costs simply shift up the location of the curve. Between 0 and Q1, the total cost curve is concave. Over this range, the marginal productivity of variable factors increases (assuming fixed input prices). Past Q1, the total cost curve is convex and the marginal productivity of variable fac- tors decreases. This type of pattern is expected given the law of diminishing returns. At low output levels, fixed inputs are not efficiently utilized. Increasing the variable

Marginal cost




$ Average total cost

Average variable cost

Average fixed cost

Total variable cost

Total cost

C os

t p er

u ni

t o f o

ut pu

t (in

d ol

la rs

) T

ot al

c os

ts (

in d

ol la

rs )

Quantity of output

Q1 Q2 Q3

Figure 5.9 Short-Run Cost Curves

This figure displays the short-run cost curves of a hypothetical firm. The upper panel depicts total cost (TC) and total variable cost (TVC). Fixed costs simply shift the position of the variable cost curve. The lower panel depicts marginal and average costs. Average fixed cost declines with output since the fixed cost is being spread over more units. Marginal cost (MC) declines to Q1 and then increases beyond that point due to diminishing returns. Marginal cost depends only on the variable input factors and is completely independent of the fixed cost. Average total cost (ATC) and AVC decline as long as marginal cost is lower than the average cost and increase beyond that point. Marginal cost is equal to both ATC and AVC at their respective minimum points. Average total cost is always larger than AVC, since ATC � AFC � AVC. However, this difference becomes smaller as output increases and AFC declines.

inputs increases output materially. Over this range, total cost increases—but does so at a decreasing rate. Eventually, the marginal productivity of the variable inputs de- clines and it becomes increasingly expensive to produce extra units of output.

The lower panel depicts marginal and average costs. Average fixed cost (AFC) is total fixed cost divided by output. Average fixed cost declines with output since the fixed cost is spread over more units. Marginal cost (MC) declines up to Q1 and then


Drugstore Chains “Play Doctor” Due to Increased Demand and Low Variable Costs Large chains, such as CVS, Walgreens, Target and Kroger, operate thousands of retail pharmacies at locations

throughout the United States. In recent years, a growing number of these pharmacies have begun to offer walk-in

clinics for basic medical services, such as vaccinations, diabetes screening, earwax removal, and treatments for sore

throats and the flu. The number of retail pharmacies with medical clinics (as well as the services they provide) is

expected to increase substantially due to expanded insurance coverage under the Affordable Care Act. CVS, currently

the market leader with 800 locations, has plans to open another 150 facilities in 2014.

Many of the costs for providing medical services at these retail locations are fixed. For example, the costs for

leasing the building, electricity, security guards, and much of the staffing would largely be the same whether or not

additional medical services are provided at the given retail location. To be profitable, the chains only have to cover

their variable costs for providing the new service, such as hiring additional nurse practitioners and the direct or

opportunity costs for providing space and seating for waiting patients. CVS charges prices between $79 and $99 for

most of its services. The relatively low prices, relatively quick service, convenient hours, and opportunities to shop

while waiting have proven attractive to many customers.

Not surprisingly, some physician groups have objected to these clinics which compete with them for patients. They

argue that the quality of medical services is inferior at the clinics. Some researchers, however, have found that the

quality of care at these retail clinics is on par with traditional medical offices for certain medical services.

Source: S. Reddy (2014), “Drug Stores Play Doctor: Physicals, Flu Diagnosis, and More,” (April 7).


Small Airport in “Big Trouble” The difficulties of competing with plant sizes significantly below minimum efficient scale are highlighted by the

problems facing small U.S. airports. Commercial airline companies have cut many flights in recent years at these

airports due to the lack of profitability. It is estimated that U.S. airlines reduced the number of flights by 14 percent

from 2007 through 2012 with midsize and small airports being the hardest hit. As examples, Des Moines, Iowa and

Burlington, Vermont lost 22 and 24 percent of their flights, respectively over this period.

The low volume of passengers and limited number of flights increase the average costs of providing services to

customers at these smaller airports. Airlines that continue to offer flights at these airports have increased their fares

substantially. For example, the small airport at Huntsville, Alabama had the highest average domestic fares among the

tracked airports in the fall of 2013, averaging $559 for a round trip fare compared to the national average of $390. Yet

the economic viability of continuing to offer flights at these many of these smaller airports remains questionable.

Local officials argue that the airports in their smaller communities are good for business and hope that the

convenience to local customers more than offsets the higher ticket prices. Many local fliers, however, opt to drive to

larger airports where the ticket prices are less expensive, the planes larger and the flight options more extensive. Some

local communities have opted to provide financial incentives and subsidies to airline companies to offset their lack of

profitability at their smaller airports. The economic and political viability of these programs, however, remains unclear.

Source: S. Carey (2014), “Why Small Airports are in Big Trouble,” (April 7).

Chapter 5 Production and Cost 173

174 Part 2 Managerial Economics

increases beyond that point due to diminishing returns. Note that marginal cost de- pends only on the variable input factors and is completely independent of the fixed cost. Average variable costs (AVCs) are total variable costs divided by output. Both average total cost (ATC) and average variable cost decline as long as marginal cost is lower than average cost; they increase beyond that point. Marginal cost is equal to both average total cost and average variable cost at their respective minimum points. Average total cost is always larger than average variable cost, since ATC � AFC � AVC. However, this difference becomes smaller as average fixed cost declines with higher output.

Long-Run Cost Curves In the short run, firms are unable to adjust their plant sizes. In the long run, however, if a firm wants to produce more output, it can build a larger, more efficient plant. In the long run, the average cost (LRAC) of production is less than or equal to the short- run average cost of production. Indeed, the LRAC curve can be thought of as an envelope of the short-run average cost curves. Figure 5.10 illustrates this concept. The figure shows four potential plant sizes. Each of the four plants provides the low- cost method of production over some range of output, assuming that only these four plant sizes are feasible. For instance, the smallest plant provides the lowest-cost method of producing any output from zero to Q1, while the next largest plant pro- vides the low-cost method of producing outputs from Q1 to Q2, and so on. The heavy portion of each curve indicates the minimum long-run average cost for producing each level of output.




Q2 Q3



C os

t p er

u ni

t o f o

ut pu

t (in

d ol

la rs


Quantity of output

Figure 5.10 Long-Run Average Costs as an Envelope of Short-Run Average Cost Curves

In the long run, the average cost (LRAC) of production is less than or equal to the short-run average cost (SRAC) of production. The LRAC curve can be thought of as an envelope of the short- run average cost curves. The figure shows four potential plant sizes. Each of the four plants provides the low-cost method of production over some range of output. For instance, the smallest plant provides the lowest-cost method of producing any output from zero to Q1, while the next largest plant provides the low-cost method of producing output from Q1 to Q2, and so on. The heavy portion of each curve indicates the minimum long-run average cost for producing each level of output, assuming that there are only these four possible plant sizes.

If we extend this analysis by assuming there are many different feasible plant sizes that vary only slightly in size, the resulting LRAC curve will be relatively smooth, as pictured in Figure 5.11. This figure also pictures the long-run marginal cost curve (LRMC). As we have discussed, the marginal cost is below average cost where average cost is falling and above average cost where it is rising. The two are equal at the minimum average cost.

Minimum Efficient Scale

Minimum efficient scale is defined as that plant size at which long-run average cost first reaches its minimum point. In Figure 5.11, this minimum occurs at Q*. The minimum efficient scale affects both the optimal plant size and the level of potential competition.

Average production cost is minimized at the minimum efficient scale. As we dis- cuss in the next chapter, competition provides incentives for firms to adopt this plant size. If firms build plants that depart materially from minimum efficient scale, they will be at a competitive disadvantage and could be forced out of business. One com- plicating factor is transportation costs. If transportation costs are high, cost


Public Utilities The production of electric power typically is associated with large economies of scale: The average cost of producing

electricity decreases with the quantity produced. This production characteristic implies that it is generally more

efficient to have one large plant that produces power for an area than several smaller plants. A problem with having one

producer of electric power in an area, however, is that the firm has the potential to overcharge consumers for electricity

since there are limited alternative sources of supply. Concerns about this problem provide one motivation for the

formation of public utility commissions that regulate the prices that utility companies can charge consumers.

Q * Q

Long-run marginal


Long-run average cost


C os

t p er

u ni

t o f o

ut pu

t (in

d ol

la rs


Quantity of output

Figure 5.11 Long-Run Average and Marginal Cost Curves

If there are many different plant sizes that vary only slightly in size, the resulting long-run average cost (LRAC) curve is relatively smooth, as pictured in this figure. The long-run marginal cost (LRMC) is below average cost where average cost is falling and above average cost where it is rising. The two are equal at the minimum average cost. The minimum efficient scale is defined as the plant size at which LRACs are first minimized (Q* in this example).

Chapter 5 Production and Cost 175

176 Part 2 Managerial Economics

disadvantages of smaller regional plants can be more than offset by cost savings in transporting the product to customers. In this case, when total production and distri- bution costs are considered, firms with plants that are smaller than the minimum ef- ficient scale can survive in a competitive marketplace.

Generally, the number of competitors will be large and competition more vigorous when the minimum efficient scale is small relative to total industry demand. For instance, suppose that Kate Polk is evaluating the possibility of entering an industry where she sees established firms reporting substantial profits. If her firm would have to produce 10 percent of the market’s output to be cost-efficient, Kate should be con- cerned that her entry is likely to drive the price down and thus she would be less likely to enter the market than if she needed to produce only 1 percent of the market’s output for efficient production.


Size Doesn’t Always Matter Regis Corporation has 10,000 beauty salons, buys shampoo by the train load, spends millions on advertising, and uses

sophisticated technology to track performance at each salon. Nonetheless, the 300,000 independent salons in the

United States still compete quite effectively with Regis chains like Supercuts. Paul Finkelstein, Regis CEO says. “We

don’t run a big business. We run ten thousand $300,000 businesses.” He says it’s salon location first, and quality of

stylists second that make the business go. Those are factors that an independent salon also can offer.

Source: F. Bailey (2003), “In Some Businesses, Size Is Irrelevant to Success,” The Wall Street Journal (November 11), B13.

ANALYZING MANAGERIAL DECISIONS: Developing Economies of Scale for Malaysia’s Proton Holdings

Proton Holdings Bhd is a national carmaker in

Malaysia. In late 2007, the Malaysian government

owned 43 percent of the company. The remaining

stock of the company traded on public stock ex-

changes. Proton was among Malaysia’s worst per-

forming companies in 2005, after competition from

foreign carmakers and a lack of new models cost

the firm significant market share and profits. It has

since hired a new chief executive, sold its loss-

making MV Agusta motorbike firm, and pledged to

find a new technology partner. The company has

been under substantial pressure, with its share of

domestic sales falling from 75 percent to 44 per-

cent over the past decade.

Analysts polled in late 2007 noted that the com-

pany’s new management had made several moves

to revamp the company and that these efforts were

bearing fruit in terms of increased sales volume and

market share. New models such as the Persona, a

sport edition of Savvy, and Satria Neo were relative

successes. The management also implemented

stringent cost controls. Nonetheless, analysts con-

cluded that Proton’s long-run ability to survive de-

pends on whether it can achieve increased produc-

tion volume and economies of scale. Without

sufficient scale it is unlikely that the company will

survive the intense local and worldwide competi-

tion. The analysts assert that by itself, Proton

would find it hard to achieve economies of scale

and to develop new technologies.

Suppose that you are hired as a consultant to

advise Proton’s management. What do the analysts

mean when they say that Proton needs to achieve

economies of scale to be competitive? Discuss at a

general level the types of actions that the company

might want to consider to achieve the necessary


Source: K. Fong (2007), “No Economies of Scale for Proton with- out Global Partner,” StarBiz, thestar online (November 21).

Industries where average cost declines over a broad range of output are charac- terized as having economies of scale. Significant economies of scale limit the num- ber of firms in the industry. For instance, if the minimum efficient scale is 25 per- cent of total industry sales, there is room for only four firms to produce at that volume. The level of competition among existing firms can vary significantly, even if there are only a few firms in the industry. However, threat of entry is less press- ing than in industries where scale economies are low. The threat of potential new competitors is often an important consideration in a firm’s strategic planning. In subsequent chapters, we examine how a firm’s market structure affects managerial decision making.

Learning Curves

For some firms, the long-run average cost of producing a given level of output de- clines as the firm gains production experience. For example, with more output, em- ployees might gain important information on how to improve production processes. They also become more proficient as they gain experience on the job.15 A learning curve displays the relation between average cost and cumulative production vol- ume. Cumulative production is the total amount of the product produced by the firm across all previous production periods. Figure 5.12 presents an example where there are significant learning effects in the early stages of production. Eventually, how- ever, these effects frequently become minimal as the firm continues to produce the product.

15A. Alchian (1959), “Costs and Outputs,” in The Allocation of Economic Resources, by M. Abramovitz and

others (Stanford University Press: Palo Alto, CA), 23–40.

Average cost of producing Q * units

Cumulative quantity of output produced

Learning curve


C os

t p er

u ni

t o f o

ut pu

t (in

d ol

la rs



Figure 5.12 Learning Curve

A learning curve displays the relation between average cost for a given output period, Q*, and cumulative past production. In this example, there are significant learning effects in the early stages of production. These effects become minimal as the firm continues to produce the product.

Chapter 5 Production and Cost 177

178 Part 2 Managerial Economics

Figure 5.13 illustrates the difference between economies of scale and learning ef- fects. Economies of scale imply reductions in average cost as the quantity being pro- duced within the production period increases. Learning effects imply a shift in the entire average cost curve: The average cost for producing a given quantity in a pro- duction period decreases with cumulative volume. Learning effects sometimes can provide existing firms in an industry a competitive advantage over potential entrants; it depends on the nature of the information (Chapter 3) and the distribution of that in- formation across employees. We discuss this issue in more detail in Chapter 8.

Economies of Scope

Thus far, we have focused on the production of a single product. Most firms, however, produce multiple products. Economies of scope exist when the cost of producing a set of products jointly within one firm is less than the cost of producing the products sep- arately across independent firms. Joint production can produce cost savings for a vari-


Average cost with low cumulative volume

Average cost with high cumulative volume


Learning effect

C os

t p er

u ni

t o f o

ut pu

t (in

d ol

la rs

) Quantity of output

Figure 5.13 Economies of Scale versus Learning Effects

This figure shows the average cost curves for a firm when it has experienced high and low cumulative volume. In both cases, there are economies of scale (average cost declines with output). The average cost for each level of output, however, is lower where the firm has experienced high cumulative volume because of learning effects.

Economies of Scale and Learning Effects in the Chemical Processing Industry Marvin Lieberman studied economies of scale and learning effects in the chemical processing industry. He found that

for each doubling in plant size, average production costs fell by about 11 percent. For each doubling of cumulative

volume, the average cost of production fell by about 27 percent. Thus, there is evidence of both economies of scale and

learning effects in the chemical processing industry. The size of the estimates suggests that learning effects are more

important than economies of scale in explaining the observed decline in costs within the industry from the 1950s to

the 1970s.

Source: M. Lieberman (1984), “The Learning Curve and Pricing in the Chemical Processing Industries,” Rand Journal 15, 213–288.


Chapter 5 Production and Cost 179

ety of reasons. Efficiencies can result from common use of production facilities, coor- dinated marketing programs, and sharing management systems. Also, the production of some products provides unavoidable by-products that are valuable to the firm. For instance, a sheep rancher jointly produces both mutton and wool.

Economies of scope help explain why firms produce multiple products. For in- stance, PepsiCo is a major producer of soft drinks; yet it also produces a wide range of snack foods (e.g., corn chips and cookies). These multiple products allow PepsiCo to leverage its product development, distribution, and marketing systems.

Economies of scope and economies of scale are different concepts. Economies of scope involve cost savings that result from joint production, whereas economies of scale involve efficiencies from producing higher volumes of a given product. It is possible to have economies of scope without having economies of scale and vice versa.

Profit Maximization Thus far, we have focused on the costs of producing different levels of output. How- ever, what output level should a manager choose to maximize firm profits? To an- swer this question, we return to the concept of marginal analysis that we initially in- troduced in Chapter 2.

Marginal costs and benefits are the incremental costs and benefits that are asso- ciated with a particular decision. It is these incremental costs and benefits that are important in economic decision making. An action should be taken whenever the incremental benefits of that action exceed its incremental costs. In deciding whether or not to produce one more unit of a product, the incremental benefit is marginal revenue (see Chapter 4), while the incremental cost is equal to marginal production cost (including any distribution costs)—fixed costs do not affect the de- cision. Therefore, the firm should produce extra units so long as marginal revenue exceeds marginal cost; the firm should not produce extra units if marginal revenue


Jimmy Beans Wool—Economies of Scope Fail to Materialize Jimmy Beans Wool is an online yarn and fabric seller with a store in Reno, Nevada. The company was founded in 2002

and by 2013 had sales of over $7 million dollars. The company initially focused exclusively on selling yarn. It grew

“organically” at an annual rate of around 50 percent over the period 2007–2012 and earned a national reputation for its

high rate of growth. The owners envisioned that Jimmy Beans could be a $100 million business within a few years. The

owners invested to expand their product line by selling fabrics to grow sales. They made significant investments in

inventory, building remodeling, and marketing. Unfortunately, sales did not grow, and the company suffered a

significant financial setback.

The owners say that they “fell into the trap of thinking if we can sell yarn, we can sell anything.” This is an all too

common managerial misconception that has led to other unsuccessful expansions and mergers. Good managers avoid

this kind of hubris and are careful not to overestimate economies of scope. Economies of scope are most likely to exist

in closely related products where there are true synergies in either production or distribution. The assumption that

economies are produced by a manager who can “manage anything” quite often proves false.

Source: A. Gardella (2014), “Seeking Even Faster Growth, An E-Commerce Company Stumbles,” New York Times (April 2).

180 Part 2 Managerial Economics

is less than marginal cost. At the profit-maximizing level of production, the follow- ing condition holds:16

MR � MC (5.9)

As we saw in Chapter 4, marginal revenue depends on the demand curve for the product. The effective demand curve that the firm faces will be affected by the degree of competition in the product market. In Chapter 6, we examine how the output de- cisions of firms vary across different market settings.

The changes in metal prices throughout the 1990s changed the total cost of auto- mobile manufacturing. Typically, such changes are accompanied by changes in the marginal cost of production. For example, a reduction in steel prices would mean not only a substitution toward steel from other inputs but also an increase in output. Figure 5.14 illustrates this effect. Note that this analysis holds other factors constant. If the demand for automobiles is falling at the same time (thus shifting marginal rev- enue downward), the net effect could be an increase in output. However, the increase in output would be less than if steel prices were constant.

Factor Demand Curves In discussing the optimal input mix, we noted that the following condition must hold for efficient production:

MPi�Pi � MPj�Pj (5.10)

16Technical note: Since profits equal total revenues minus total costs, Equation (5.9) is the first-order

condition for profit maximization. This condition holds at both minimum and maximum profits. At the

maximum, the marginal cost curve cuts the marginal revenue curve from below—the second-order






Q* Q * Q

Quantity of output

C os

t/r ev

en ue

p er

u ni

t (in

d ol

la rs


0 1

Figure 5.14 Changes in Marginal Cost and Optimal Output

This figure illustrates that a decrease in marginal cost (from MC0 to MC1) raises the optimal output of the firm (from Q*0 to Q*1).

Chapter 5 Production and Cost 181

for all inputs i and j. The ratios of marginal product to price reflect the incremental out- put from an input associated with an additional dollar expenditure on that input. The reciprocals of these ratios reflect the dollar cost for incremental output or the marginal cost:

Pi�MPi � Pj�MPj � MC (5.11)

At the profit-maximizing output level, MR � MC. Therefore, at the optimal output level the following condition must hold

Pi�MPi � MR (5.12)


China Becomes the “World’s Smokestack” This chapter focuses on how firms choose among alternative input mixes. Another important production decision is where

to locate the plant. Costs can vary dramatically across locations due to differences in the prices for labor, land,

transportation and other inputs, taxes, environmental and safety regulations, threat of terrorism, and political risk. The steel

industry provides a good example of how changes in costs can motivate significant changes in the location of production.

During the 1950s the Ruhr Valley in Germany had the world’s highest growth rate in steel production. In its heyday,

Germany produced 10 percent of the world’s steel supply. In 2006, Germany produced only 3.8 percent of the world’s

supply, ranking seventh behind China, Japan, the United States, Russia, India, and South Korea. China, which has

displayed meteoric growth in steel production, supplied 34 percent.

German steel production slowed in the 1960s as miners had to dig deeper for coal and taxes and labor costs

continued to increase. Another important factor was new environmental regulation. The emissions from steel plants had

made the Ruhr Valley one of the most polluted places in the world. The air was dark and grimy. The residents suffered

from an inordinate incidence of lung and other pollution-related diseases. The white shirts that men wore to church on

Sunday turned to grey by the time they came home. In an effort to “green the country,” the government imposed costly

pollution control requirements on the steel companies.

Differences in labor costs and environmental standards motivated a shift in steel production from Germany to China

as “smoke-spewing plants” were disassembled in the Ruhr Valley and moved 5,000 miles away to China. The Phoenix

steel mills in Dortmund had been among Germany’s largest since before World War II. In the late 1990s, they were

slated for closure and were likely headed for the scrap heap. The Chinese realized that they could buy a relatively

sophisticated German blast furnace for a small fraction of what a new one would cost. A Chinese company sent

workers to Dortmund who labeled every part of the seven-story blast furnace, disassembled it, and packed it into

wooden crates for the voyage to China. They worked day and night to accomplish this task in a much shorter time than

it would have taken German workers, who were governed by strict union and government work rules.

The Hebei Province is “China’s new Ruhr Valley.” Its air is heavily polluted and its citizens suffer from a variety of

associated health problems. Meanwhile, the Ruhr Valley’s pollution level has substantially improved. This

improvement, however, has come at a cost. Dortmund, which in 1960 had 30,000 residents working in the steel

industry, now has less than 3,000. While Dortmund continues to have high unemployment, the decline in steel jobs has

been offset to some extent by new jobs in other less-polluting industries.

The willingness to pay for clean air generally increases with a country’s income. In March 2014, Chinese Premier

Li Keqiang stated in a speech before the Chinese legislature, “we will declare a war on pollution and fight with the

same determination we battled poverty.” Government officials have pledged a series of pollution reforms targeted at

energy-intensive industries, such as steel, aluminum, cement, and coal. Economics, however, teaches us “there is no

free lunch.” Analysts predict that the proposed reforms will increase production cost and potentially force some of the

factories to close.

Source: J. Kahn and M. Landler (2007), “China Grabs West’s Smoke-Spewing Factories,” (December 21); and W. Ma

and C.-W. Yap (2014), “China Needs Industry to Enlist a War on Poverty,” Dow Jones Reprints (March 6).

182 Part 2 Managerial Economics

or equivalently,

Pi � MR � MPi (5.13)

Equation (5.13) is the firm’s demand curve for input i.17 It has a straightforward in- terpretation. The right-hand side of the equation represents the incremental revenue that the firm obtains from employing one more unit of the input (the incremental out- put times the incremental revenue). We call this incremental revenue the marginal revenue product (MRPi) of input i. Figure 5.15 illustrates the demand curve for an input.18 At the current input price of P*i, the firm optimally uses Q*i units of the input. The firm optimally employs additional units of the input up to the point where the marginal cost of the input (its price with constant input prices) is equal to the mar- ginal revenue product of the input. Intuitively, if the marginal revenue product is greater than the input price, the firm increases its profitability by using more of the input. If the marginal revenue product is less than the price of the input, the firm in- creases profitability by reducing the use of the input. Profits are maximized when the two are equal.

Our discussion of the profit-maximizing output level and the optimal use of an input might appear to suggest that these decisions are two distinct choices. The two de- cisions, however, are linked directly. Once the firm chooses the quantities of inputs,

17Technical note: The marginal product of input i can depend on the levels of other inputs used in the

production process. Thus, the demand curve for an input must allow other inputs to adjust to their optimal

levels as the price of input i changes. This adjustment is not important if the marginal product of input i is

not affected by the levels of the other inputs. 18Technical note: The second-order condition for maximum profits ensures that the demand curve for the input

is the downward-sloping portion of the marginal revenue product curve. Thus, Figure 5.15 displays only the

downward-sloping portion of the curve.




Q * Qi

C os

t/r ev

en ue

p er

u ni

t o f i

np ut

(in d

ol la

rs )


Quantity of input i

Figure 5.15 Factor Demand Curve

The demand curve for a factor of production is the marginal revenue product curve (MRP) for the input. The marginal revenue product is defined as the marginal product of the input times the marginal revenue. It represents the additional revenue that comes from using one more unit of input. The firm maximizes profits where it purchases inputs up to the point where the price of the input equals its marginal revenue product.

Chapter 5 Production and Cost 183

output is determined by the production function. Thus, profit-maximizing firms choose the output where marginal revenue equals marginal cost and produce that out- put so that the price of each input is equal to its marginal revenue product. In our auto example, an increase in steel prices would be expected to motivate simultaneous ad- justments in both the number of automobiles produced and the methods used to pro- duce them.


Hog Producers React to Increase in Corn Prices Pork is the most widely eaten meat in the world, providing about 38 percent of daily meat protein worldwide (despite

the fact that it is not consumed by some people due to religious restrictions). The United States Department of

Agriculture reports that in 2006 the per capita consumption of pork was 43.9 kg, 40 kg, and 29 kg in China, Europe,

and the United States, respectively. Ironically in China’s “Year of the Pig” (2007), there was a crisis that increased pork

prices by 50 percent. The dramatic price increase caused both social unrest and government intervention to “help keep

pork affordable.”

Feed cost is typically about 50–60 percent of the total cost of producing pork. In the United States, corn accounts

for about 80 percent of the typical hog feed. Rising fuel prices and government policies promoting the use of corn-

based ethanol as an alternative to gasoline caused U.S. corn prices nearly to double in the summer and fall of 2006.

This price change significantly increased the cost of feeding hogs. U.S. hog producers were able to lessen the effects of

the increase in corn prices by switching to feed mixes that use less corn and more dried distilled grain and solubles

(DDGS). DDGS is a by-product of the corn-based ethanol production process. Scientists found that DDGS could be

substituted for corn at a 10 percent inclusion rate without having a significant effect on the efficiency, growth, or

carcass traits of the hogs. Corn and DDGS, however, are not perfect substitutes in the hog production process. Feed

mixes with higher DDGS inclusion rates (e.g., 20 percent and 30 percent) produced smaller pigs that offset the

advantages of lower feed costs.

This example highlights a general point about production costs. Increases in input prices increase production costs. The

effect of the price increase often can be mitigated by shifting the input mix toward relatively less expensive inputs. The

ability to reduce costs in this manner depends on the degree of substitutability among the inputs.

Source: J. Lawrence (2006), “Impact on Hog Feed Cost of Corn and DDGS Prices,” Iowa Farm Outlook (November 15), 1–4.


Demand for Labor Falls Following 9/11 Terrorist Attacks Organized labor was particularly hard hit by the terrorist attacks following 9/11. Of the 760,000 job cuts in the three

months following the attack, roughly 50 percent were union members—nearly four times organized labor’s

13.5 percent of the U.S. workforce. The reason for this higher-than-average job loss is because unions are

disproportionately represented in the hard-hit travel and tourism industries. Unions representing public employees also

are seeing large job cuts as state and local budgets are cut. September 11 significantly shifted consumers’ demand for

travel and tourism services to the left. When demand shifts to the left, so does the firm’s marginal revenue curve.

Hence, labor’s marginal revenue product—the product of marginal revenue and the factor’s marginal product (holding

constant labor’s marginal product)—also shifts to the left.

Source: A. Bernstein (2001), “A Sock in the Eye for Labor,” BusinessWeek (December 17), 44.

184 Part 2 Managerial Economics

Cost Estimation Our discussion indicates that a detailed knowledge of costs is important for manage- rial decision making. Short-run costs play an extremely important role in operating decisions. For instance, when the marginal revenue from increased output is above the short-run marginal cost of production, profits increase by expanding production. Alternatively, if marginal revenue is below short-run marginal cost, reducing output increases profits. Long-run costs, in turn, provide important information for deci- sions on optimal plant size and location. For instance, if economies of scale are im- portant, one large plant is more likely optimal with the product transported to re- gional markets. Alternatively, if scale economies are small, smaller regional plants, which reduce transportation costs, are more likely optimal.

If managers are to incorporate costs in their analyses in this manner, they must have accurate estimates of how short-run and long-run costs are related to various factors both within and beyond the control of the firm.19 Among the most com- monly used statistical techniques for estimating cost curves is regression analysis. A regression estimates the relation between costs and output (possibly controlling for other factors, such as the product mix or the weather, which affect costs). The data for this analysis can be either time-series data on costs, output, and other vari- ables, or cross-sectional data, which includes observations on variables across firms or plants at a point in time. For instance, in many applications, it is assumed that short-run total costs are approximately linear20

VC � a � bQ (5.14)

where VC is total variable costs for the period and Q is the quantity of output produced.

A detailed discussion of cost estimation is beyond the scope of this book. Suffice it to say that similar problems arise in cost estimation as arise in the case of demand estimation (e.g., omitted-variables problems). Among the most common problems in cost estimation are difficulties in obtaining data on relevant costs. Cost estimates often are based on accounting reports, which record historical costs. As we have in- dicated, these historical costs do not necessarily reflect the opportunity costs of using resources. Moreover, there is the issue of choosing the appropriate functional form. Equation (5.14) presumes a linear model. However, cost curves need not be linear. For instance, it might be appropriate to use a quadratic model, which would include an additional Q 2 term.

One of the more serious problems complicating cost estimation is the fact that most plants produce multiple products. Multiple products are produced in the same plant because there are economies of scope. Rather than produce two different types of cereals in two separate plants, it typically is cheaper to produce them in one plant; fixed resources can be used more efficiently. If a plant produces multiple products, total and average costs for each product can be calculated only by allocating fixed

19In addition, some firms estimate cost curves to obtain insights into their underlying production functions.

Recall that the shapes of cost curves depend on the underlying production functions. Thus, it often is

possible to infer the characteristics of a production function from the shape of the corresponding cost curves.

Typically, the data for estimating cost curves is more readily available than the necessary data for estimating

production functions. 20Variable costs are normally estimated with an intercept. Although variable costs undoubtedly are zero when

output is zero, most cost curves are nonlinear. Forcing the intercept to be zero yields a less precise estimate

of this slope—the change in costs associated with a change in output.

Chapter 5 Production and Cost 185

costs across the products. This allocation often is arbitrary and complicated further by the existence of joint costs. Cost accountants use accounting records to track costs of individual products. Fixed and variable resources used by each product are recorded. These product costs, calculated by the cost accountants, typically are used to estimate short-run and long-run average and marginal costs.

Despite these estimation problems, cost curves play an important role in manage- rial decision making. Nonetheless, it is important that managers maintain a healthy skepticism when using these estimates. For instance, in making major decisions, it generally is instructive for managers to examine whether a proposed decision is still attractive with reasonable variation in the estimated parameters of the cost func- tion—that is, to conduct sensitivity analysis.

Summary A production function is a descriptive relation that connects inputs with outputs. It specifies the maximum possible output that can be produced for given amounts of in- puts. Returns to scale refers to the relation between output and a proportional varia- tion in all inputs taken together. A production function displays constant returns to scale when a 1 percent change in all inputs results in a 1 percent change in output. With increasing returns to scale, a 1 percent change in all inputs results in a greater than 1 percent change in output. Finally, with decreasing returns to scale, a 1 percent change in all inputs results in a less than 1 percent change in output. Returns to a fac- tor refers to the relation between output and the variation in only one input, holding other inputs fixed. Returns to a factor can be expressed as total, marginal, or average quantities. The law of diminishing returns states that the marginal product of a vari- able factor will eventually decline as the use of the input is increased.

Most production functions allow some substitution of inputs. An isoquant displays all combinations of inputs that produce the same quantity of output. The


Gina Picaretto is production manager at the Rich

Manufacturing Company. Each year her unit buys

up to 100,000 machine parts from Bhagat Incorpo-

rated. The contract specifies that Rich will pay

Bhagat its production costs plus a $5 markup (cost- plus pricing). Currently, Bhagat’s costs per part are

$10 for labor and $10 for other costs. Thus the

current price is $25 per part. The contract provides

an option to Rich to buy up to 100,000 parts at this

price. It must purchase a minimum volume of

50,000 parts.

Bhagat’s workforce is heavily unionized. During

recent contract negotiations, Bhagat agreed to a

30 percent raise for workers. In this labor contract,

wages and benefits are specified. However, Bhagat

is free to choose the quantity of labor it employs.

Bhagat has announced a $3 price increase for its

machine parts. This figure represents the projected

$3 increase in labor costs due to its new union

contract. It is Gina’s responsibility to evaluate this


1. Why do many firms use cost-plus pricing for

supply contracts?

2. What potential problems do you envision with

cost-plus pricing?

3. Should Gina contest the price increase?


4. Is the increase more likely to be justified in the

short run or the long run? Explain.

5. How will a $3 increase in the price of machine

parts affect Gina’s own production decisions?

186 Part 2 Managerial Economics

optimal input mix to produce any given output depends on the costs of the inputs. An isocost line displays all combinations of inputs that cost the same. Cost minimization for a given output occurs where the isoquant is tangent to the isocost line. Changes in input prices change the slope of the isocost line and the point of tangency. When the price of an input increases, the firm will reduce its use of this input and increase its use of other inputs (substitution effect).

Cost curves can be derived from the isoquant/isocost analysis. The total cost curve depicts the relation between total costs and output. Marginal cost is the change in total cost associated with a one-unit change in output. Average cost is total cost di- vided by total output. Average cost falls when marginal cost is below average cost; average cost rises when marginal cost is above average cost. Average and marginal costs are equal when average cost is at a minimum. There is a direct link between the production function and cost curves. Holding input prices constant, the slopes of cost curves are determined by the underlying production technology.

Opportunity cost is the value of a resource in its next best alternative use. Current market prices more closely reflect the opportunity costs of inputs than historical costs. The relevant costs for managerial decision making are the opportunity costs.

Cost curves can be depicted for both the short run and the long run. The short run is the operating period during which at least one input (typically capital) is fixed in supply. During this period, fixed costs can be incurred even if the firm produces no output. In the long run, there are no fixed costs—all inputs and costs are variable. Short-run cost curves are sometimes called operating curves because they are used in making near-term production and pricing decisions. Fixed costs are irrelevant for these decisions. Long-run cost curves are referred to as planning curves, since they play a key role in longer-run planning decisions relating to plant size and equipment acquisitions.

The minimum efficient scale is defined as that plant size at which long-run aver- age cost is first minimized. The minimum efficient scale affects both the optimal plant size and the level of potential competition. Industries where the average cost declines over a broad range of output are characterized as having economies of scale.

A learning curve displays the relation between average cost and the cumulative volume of production. For some firms, the long-run average cost for producing a given level of output declines as the firm gains experience from producing the output (i.e., there are significant learning effects).

Economies of scope exist when the cost of producing a joint set of products in one firm is less than the cost of producing the products separately across independent firms. Economies of scope help explain why firms often produce multiple products.

The profit-maximizing output level occurs at the point where marginal revenue equals marginal cost. At this point, the marginal benefits of increasing output are offset exactly by the marginal costs.

The marginal revenue product of input i (MRPi) equals the marginal product of the input times marginal revenue. Profit-maximizing firms use an input up to the point where the MRP of the input equals the input price. At this point, the marginal benefit of employing more of the input is offset exactly by its marginal cost.

Managers often use estimates of cost curves in decision making. A common statis- tical tool for estimating these curves is regression analysis. One common problem in statistical estimation is the difficulty of obtaining good information on the opportunity costs of resources. Another problem with estimating cost curves involves allocating fixed costs in a multiproduct plant. Cost accountants track the costs and estimate product costs.

Chapter 5 Production and Cost 187

G. Stigler (1987), The Theory of Price (Macmillan: New York), Chapters 6–10.

5–1. The Zimmerman Company digs ditches. It faces the production function, Q � L1�2K, where

Q is the number of ditches dug, L is hours of labor, and K is the number of digging tools.

a. Complete the following table:

K � 0 K � 1 K � 2 K � 3

L � 0 L � 1 L � 2 L � 3

b. Does the production function display increasing, decreasing, or constant returns to

scale? Explain.

c. Are the marginal products of K and L increasing, decreasing, or constant? Explain.

d. Assume constant input prices. Draw the general shapes of the following: (1) long-run

average cost; (2) short-run marginal cost, assuming L is fixed; (3) short-run marginal

cost, assuming K is fixed.

5–2.21 A product is produced using two inputs x1 and x2 costing w1 � $10 and w2 � $5 per unit,

respectively. The production function is y � 5x1 1.5x2

2 where y is the quantity of output, and

x1, x2 are the quantities of the two inputs. The marginal products of inputs 1 and 2 for this

production function are:

MP1 � 7.5 � x1 0.5 � x2


MP2 � 10 � x1 1.5 � x2

a. What input quantities (x1*, x2*) minimize the cost of producing 10,000 units of output?

b. What is the total cost of producing the 10,000 units?

Solutions to Self-Evaluation Problems 5–1. a.

K � 0 K � 1 K � 2 K � 3

L � 0 0 0 0 0 L � 1 0 1 2 3 L � 2 0 1.41 2.83 4.24 L � 3 0 1.73 3.46 5.19

b. The production function shows increasing returns to scale. As you increase both inputs

by the same proportion, output goes up by a higher proportion (e.g., if you double both

inputs output goes up by more than double). This can be seen along the diagonal of the


c. The marginal product of K is constant. If you hold L fixed and increase K, the marginal

increase in output is constant as you add more and more K. You can see this along the

Self-Evaluation Problems

Suggested Reading

21This problem requires slightly higher skills in algebra than most of the other problems in the book. Readers

who cannot work the problem on their own should study the general approach used in the solution to obtain

a better understanding of the conditions for cost minimization.

188 Part 2 Managerial Economics

rows of the table. The marginal product of L is decreasing. If you hold K fixed and in-

crease L, the marginal increase in output declines as you add more and more L. This can

be seen along the columns of the table.

d. Increasing returns to scale implies that LRAC declines as output is increased (there are

economies of scale). The marginal product of K is constant. Thus SRMC is constant when

K is the variable input. For example, when L � 1, each additional unit of K produces one

extra unit of output (see the second row of the table). In this case, SRMC (the cost of pro-

ducing an additional unit of output) is simply the price of K. The marginal product of L is

decreasing. Thus SRMC is increasing when L is the variable input. The general shapes of

the graphs are:

5–2. a. The cost-minimizing combination of inputs is found by equating their marginal product

to price ratios: MP1�P1 � MP2�P2. This condition can be expressed as:

So, we get


x1 � (3�8) x2 (i)

Use the relation in equation (i) to express the production function for y � 10,000 units as

a function of only x2. You now have a solvable equation with one unknown variable, x2:

Solving for x2 and obtaining x1 from equation (i):

x2* � 13.357

x1* � 5.01

b. The total cost of producing the 10,000 units is simply the sum of the expenditures made

to acquire each of the inputs:

TC � w1x1 � w2x2 � 10 � 5.01 � 5 � 13.357 � $116.87

5–1. Distinguish between returns to scale and returns to a factor.

5–2. Your company currently uses steel and aluminum in a production process. Steel costs $.50

per pound, and aluminum costs $1.00 per pound. Suppose the government imposes a tax of

Review Questions

y � 5x1 1.5x2

2 1 10,000 � 5a3x2

8 b 1.5

(x2) 2 � 5a3

8 b 1.5

x2 3.5

1 3

4 x�1

1 x2 � 2 1 x2


� 8




� w1

w2 1

7.5 � x1 0.5 � x2


10 � x1 1.5 � x2

� 10




� w1


$/unit $/unit $/unit




SRMC (K fixed)


SRMC (L fixed)

Chapter 5 Production and Cost 189

$.25 per pound on all metals. What affect will this have on your optimal input mix? Show

using isoquants and isocost lines.

5–3. Your company currently uses steel and aluminum in a production process. Steel costs $.50

per pound, and aluminum costs $1.00 per pound. Suppose that inflation doubles the price of

both inputs. What affect will this have on your optimal input mix? Show using isoquants

and isocost lines.

5–4. Is the “long run” the same calendar time for all firms? Explain.

5–5. You want to estimate the cost of materials used to produce a particular product. According

to accounting reports, you initially paid $50 for the materials that are necessary to produce

each unit. Is $50 a good estimate of your current production costs? Explain.

5–6. Suppose that average cost is minimized at 50 units and equals $1. What is marginal cost at

this output level?

5–7. What is the difference between economies of scale and economies of scope?

5–8. What is the difference between economies of scale and learning effects?

5–9. Suppose that you can sell as much of a product as you want at $100 per unit. Your marginal

cost is: MC � 2Q. Your fixed cost is $50. What is the optimal output level? What is the

optimal output, if your fixed cost is $60?

5–10. Discuss two problems that arise in estimating cost curves.

5–11. Suppose that the marginal product of labor is: MP � 100 � L, where L is the number of

workers hired. You can sell the product in the marketplace for $50 per unit, and the wage

rate for labor is $100. How many workers should you hire?

5–12. Textbook authors typically receive a simple percentage of total revenue generated from

book sales. The publisher bears all the production costs and chooses the output level. Sup-

pose the retail price of a book is fixed at $50. The author receives $10 per copy, and the firm

receives $40 per copy. The firm is interested in maximizing its own profits. Will the author

be happy with the book company’s output choice? Does the selected output maximize the

joint profits (for both the author and company) from the book?

5–13. Suppose your company produces one product and that you are currently at an output level

where your price elasticity is 0.5. Are you at the optimal output level for profit maximiza-

tion? How can you tell?

5–14. Semiconductor chips are used to store information in electronic products, such as personal

computers. One of the early leaders in the production of these chips was Texas Instruments

(TI). During the early period in the development of this industry, TI made the decision to price

its semiconductors substantially below its production costs. This decision increased sales, but

resulted in near-term reductions in profits. Explain why TI might have made this decision.

5–15. The AFL-CIO has been a steadfast proponent of increasing the minimum wage. Offer at

least two reasons why they might lobby for such increases.

5–16. Mountain Springs Water Company produces bottled water. Internal consultants estimate the

company’s production function to be Q � 300L2K, where Q is the number of bottles of

water produced each week, L is the hours of labor per week, and K is the number of ma-

chine hours per week. Each machine can operate 100 hours a week. Labor costs $20 per

hour, and each machine costs $1,000 per week.

a. Suppose the firm has 20 machines and is producing its current output using an optimal

K�L ratio. How many people does Mountain Springs employ? Assume each person

works 40 hours a week.

b. Recent technological advancements have caused machine prices to drop. Mountain

Springs can now lease each machine for $800 a week. How will this affect the optimal

K�L ratio (i.e., will the optimal K�L ratio be smaller or larger)? Show why.

5–17. The Workerbee Company employs 100 high school graduates and 50 college graduates at

respective wages of $10 and $20. The total product for high school graduates is 1,000 �

190 Part 2 Managerial Economics

100Q H, whereas the total product for college graduates is 5,000 � 50QC. Q H � the number

of high school graduates, while QC � the number of college graduates. Is the company hir-

ing the optimal amount of each type of worker? If not, has it hired too many high school or

too many college graduates? Explain.


0 1 500 80 2 60 3 50 4 60 5 75 6 95 7 120 8 150 9 185

a. Complete the above table.

b. Graph TC, TFC, TVC, MC, AC, AFC, and AVC against Q.

5–19. Suppose the Jones Manufacturing Company produces a single product. At its current input

mix the marginal product of labor is 10 and the marginal product of capital is 20. The per

unit price of labor and capital are $5 and $10, respectively. Is the Jones Company using an

optimal mix of labor and capital to produce its current output? If not, should it use more

capital or labor? Explain.

5–20. Suppose the production function of PowerGuns Co. is given by

Q � 25LK

where Q is the quantity of guns produced in the month, L is the number of workers em-

ployed, and K is the number of machines used in the production. The monthly wage rate is

$3,000 per worker and the monthly rental rate for a machine is $6,000. Currently

PowerGuns Co. employs 25 workers and 40 machines. Assume perfect divisibility of labor

and machines.

a. What is the current average product of labor for PowerGuns Co.? What is the current

marginal product of machines? (Assume 1 unit increase in machines.)

b. Does PowerGuns’ production function display increasing, decreasing, or constant

returns to scale? Explain.

c. What is the total cost of the current production of PowerGuns in a month? What is the

average cost to produce a shooting gun? Assuming the number of machines does not

change, what is the marginal cost of producing one additional gun?

d. What is the law of diminishing returns? Does this production display this characteristic?


5–21. Assume Canon’s production function for digital cameras is given by Q � 100(L 0.7K 0.3),

where L and K are the number of workers and machines employed in a month, respectively,

and Q is the monthly output. Moreover, assume the monthly wage per worker is $3,000 and

the monthly rental rate per machine is $2,000. (Note: Given the production function, the

marginal product functions are MPL � 70(L�0.3K 0.3 ) and MPK � 30(L 0.7K�0.7 )).

a. If Canon needs to supply 60,000 units of cameras per month, how many workers and

machines should it optimally employ?

b. What are the total cost and average cost of producing the quantity given in (a)?

5–22. For simplicity, throughout this problem, assume labor (L), capital (K ), and quantity pro-

duced (Q) can be infinitely divided—that is, it is fine to hire 3.3 workers, rent 4.7 machines,

and/or produce 134.2 units. Answer the following questions, assuming the production

Chapter 5 Production and Cost 191

function for DurableTires Corp. is Q � L1�3K 1�2, where Q is the quantity of tires produced,

L is the number of workers employed, and K is the number of machines rented.

a. What is the quantity of tires produced when the company employs 64 workers and

36 machines?

b. What are the average product of labor (L) and the average product of machines (K)

when the input mix is the one given above? Clearly and concisely, please explain how

you would interpret these numbers.

c. Continue to assume the input mix given above: What is the marginal product of labor

(L), if the number of workers is increased by 1 unit? What is the marginal product of cap-

ital (K), if the number of machines is increased by 1 unit, instead? Clearly and con-

cisely, explain how you would interpret these numbers.

d. Does DurableTires’ production function display increasing, decreasing, or constant re-

turns to scale? Explain. Would your answer change, if the production function were

Q � L1�2K 1�2? How? Explain.

e. Does DurableTires’ production function display increasing, decreasing, or constant

returns to labor? Explain. Would your answer change, if the production function were

Q � L1�2K 1�2? How? Explain.

5–23. Answer the following questions, continuing to assume the production function for

DurableTires Corp. is Q � L1�3K 1�2, where Q is the quantity of tires produced, L is the

number of workers employed, and K is the number of machines rented. Moreover, assume

the wage per unit of labor (WL) is $50 and the rental price per machine is $200 (WK ).

a. What is the total cost of producing the quantity of tires you found in your answer to

question 5–23(a)? And the average cost? Assuming the number of machines rented does

not change, what is the marginal cost of producing one additional tire?

b. Given the production function above, the marginal product of labor and the marginal

product of capital are MPL � 1�3(L�2�3K 1�2) and MPK � 1�2(L1�3K �1�2), respectively.

Given the wage and rental rate above, is DurableTires Corp. adopting an optimal input

mix to produce the quantity of tires found in question 5–23(a)? If yes, why? If not, why

not, and how could DurableTires Corp. save money producing that same quantity of

tires? Explain.

c. What happens to the optimal input mix you found in question 5–23, if the government

introduces a tax that raises the cost of labor to $150 per worker? Explain.

5–24. Assume DurableTires Corp. faces the following demand curve, P � 250 � 0.1Q. If Durable-

Tires’ marginal cost is constant at $35, how many tires should it produce in order to maxi-

mize its profits? What’s DurableTires’ profit in this case? Should the elasticity of demand

be greater, equal, or less than 1 at the profit-maximizing price and quantity? Explain (Hint: you may use a graph to support your argument).

Appendix: The This appendix derives the factor-balance equation—Equation (5.9) in the text:

Factor-Balance MPi�Pi � MPj�Pj (5.15) Equation22

This condition must hold if the firm is producing output in a manner that minimizes costs (assuming an interior solution).

Recall that at the cost-minimizing method of production, the isoquant curve and isocost line are tangent. Thus, they must have equal slopes. The factor-balance equa- tion is found by setting the slope of the isoquant equal to the slope of the isocost line and rearranging the expression. In the text, we showed that the slope of the isocost line is �Pj�Pi. We now derive the slope of an isoquant.

22This appendix requires a basic knowledge of calculus.

192 Part 2 Managerial Economics

Slope of an Isoquant The production function in the two-input case takes the following general form:

Q � f (xi, xj) (5.16)

To find the slope of an isoquant, we totally differentiate Equation (5.16). We set this differential equal to zero, since quantity does not change along an isoquant:

dQ � [�Q��xidxi] � [�Q��xjdxj] � 0 (5.17)

The slope of the isoquant is defined by dxi�dxj. Thus,

Slope of an isoquant � �(�Q��xj)�(�Q��xi) (5.18)

� �MPj�MPi (5.19)

This expression has a straightforward interpretation. For illustration, assume that at some fixed combination of xi and xj, the marginal product of i is 1 and the marginal product of j is 2. At this point, the slope of the isoquant is �2. This means that 2 units of i can be given up for 1 unit of j and output will stay the same. This is true by defin- ition since j has twice the marginal product of i.

Factor-Balance Equation When employing the cost-minimizing production method, the slope of the isoquant is the same as the slope of the isocost line:

�MPj�MPi � �Pj�Pi (5.20)

Rearranging this expression gives us the factor-balance equation:

MPi�Pi � MPj�Pj (5.21)

This expression immediately generalizes to production functions with more than two inputs.


1 Market Structure


6 C H A P T E R O U T L I N E


Competitive Markets

Firm Supply

Competitive Equilibrium

Barriers to Entry

Incumbent Reactions

Incumbent Advantages

Exit Costs


Monopolistic Competition


Nash Equilibrium

Output Competition

Price Competition

Empirical Evidence

Cooperation and the Prisoners’ Dilemma

Summary T he market for cable television has grown tremendously since Home Box Office (a subsidiary of Time Inc.) began broadcasting in 1975.1 Today millions of subscribers purchase multichannel pack- ages from cable companies. Historically consumers in most local

markets had but one choice—purchase cable TV from the one local provider or watch the locally broadcast “free” channels. Subject to regulatory con- straints, local cable companies could set their prices without fear that they would be undercut by the competition. Correspondingly, annual price in- creases for cable TV often exceeded the rate of inflation. According to the Federal Communications Commission, the average monthly price for cable TV rose by more than 90 percent between 1995 and 2005.

In 1994, DirecTV began providing an alternative to cable TV—satellite TV. Initially, customers had to shell out up to $850 for installation and dish equipment, and pay ongoing fees to acquire satellite TV. Many consumers did not consider satellite TV to be a viable alternative because of the price. Fledgling satellite companies were able to attract some customers because


1. List the basic characteristics of market structure.

2. Explain why perfect competition is a useful benchmark model.

3. Explain a firm’s short-run supply decision and firm and industry supply curves.

4. Explain shutdown and exit decisions.

5. Explain graphically and intuitively long-run equilibrium and how changes in

a market affect the equilibrium in the short and long run.

6. Explain the difference between a constant cost and increasing cost industry.

7. List potential barriers to entry in an industry.

8. Contrast the monopolistic and competitive market outcomes.

9. Define oligopoly and Nash equilibrium.

10. Describe standard economic models of oligopolies focusing on output and

price competition, respectively.

11. Describe the “prisoner’s dilemma” and discuss how it relates to cartel


1Some of the details for this example are from P. Grant (2002), “The Cable Guy Cuts His Rates,”

The Wall Street Journal (September 25); Reuters Limited (2002), “FCC: Cable Prices Rose 7.5%

over 12 Months” (April 4); C. Wexler (2006), “Ask Yourself Why . . . Cable Rates Got So High,”

Common Cause (October).

194 Part 2 Managerial Economics

they offered more channels and a clearer digital picture than the typical cable com- pany; they were most successful in locations not served by cable companies.

By 2002, the market had changed significantly. Local cable companies were of- fering more channels and higher quality reception, while satellite companies had lowered their prices and hookup fees substantially. To many consumers, satellite TV and cable TV had become relatively close substitutes. Thus the decision between these two services began to depend more on their relative prices. One consumer who switched from paying $80 per month for Mediacom’s cable service to DirecTV’s satellite service priced at $50 per month summed it up when he said, “I feel like I got everything I had with digital cable but at a lot cheaper price.”

In response to the increased competition, cable company managers initiated several major policy changes. First, they became more competitive in the pricing of cable TV services. Charter Communications, with over 4 million residential subscribers in 2014, is one example of a cable company that altered its pricing policies in response to competitive pressures. In St. Louis, for instance, its basic service of 125� channels for $59.99 per month competed with Dish Network’s service of 120� channels for $29.99 per month and DirectTV’s basic package of 150� channels for $29.99 per month. In addition, Charter faced increased compe- tition from Internet video providers, such as Netflix and Amazon Instant Video. In explaining an earlier price freeze in that market, Charter’s CEO stated, “We’ve got to think twice about rate increases.” Second, cable companies began offering new price/channel packages to cater to various consumer groups with different price/channel sensitivities. For example, AT&T introduced a premium package of 150 channels for $50 per month and an economy package of 100 channels for $40 per month. Previously they only had offered a 125-channel package for $43 month. Third, increased competition also affected the companies’ advertising strategies. For example, companies began promoting the relative benefits of cable TV (such as access to local channels, reduced “rain fade,” not having satellite equipment detract from the appearance of the home, and so on).

The example of cable TV illustrates how policy choices—such as pricing, product design, and advertising—are influenced critically by the market environment. Policies that work within a protected market environment often have to be amended materially when facing a more competitive environment. It is important that managers understand the firm’s market environment and how this set of market circumstances affects deci- sion making. Our purpose in this chapter is to enhance that understanding by exploring the implications of alternative market structures. Our primary focus is on output and basic pricing decisions within different market structures. In subsequent chapters, we examine more complex pricing policies and how other policies, such as aspects of the firm’s strategy and organizational architecture, depend on the market environment.

We begin by discussing markets and market structure in greater detail. We then provide an analysis of competitive industries. Perfect competition is at one end of a continuum based on the environment in which prices are determined within the in- dustry. Competitive markets provide important managerial implications for firms op- erating within a broad class of market settings. Next, we discuss barriers to entry that can limit competition within an industry. This section is followed by an analysis of the market structure at the other end of the continuum: monopoly. In a monopolistic industry, there is but one firm. In contrast to firms in competitive industries, a mo- nopolist has substantial discretion in setting prices. After a brief discussion of a rel- atively common hybrid structure—monopolistic competition—we consider the case of oligopoly, where a small number of rival firms constitute the industry.

Chapter 6 Market Structure 195

Markets A market consists of all firms and individuals who are willing and able to buy or sell a particular product.2 These parties include those currently engaged in buying and selling the product, as well as potential entrants. Potential entrants are all individu- als and firms that pose a sufficiently credible threat of market entry to affect the pric- ing and output decisions of incumbent firms.

Market structure refers to the basic characteristics of the market environment, in- cluding (1) the number and size of buyers, sellers, and potential entrants; (2) the de- gree of product differentiation; (3) the amount and cost of information about product price and quality; and (4) the conditions for entry and exit. We begin our analysis of alternative market structures by examining competitive markets.

Competitive Markets Economists generally characterize competitive markets by four basic conditions:

• A large number of potential buyers and sellers

• Product homogeneity

• Rapid dissemination of accurate information at low cost

• Free entry into and exit from the market

Although few markets are perfectly competitive, many markets closely approxi- mate this description. Moreover, competition establishes a benchmark that yields use- ful insights into other market settings. An example of a market that comfortably sat- isfies the conditions for a competitive market is the market for soybeans. In this market, a relatively large number of farmers grow soybeans, and a large number of firms and individuals purchase soybeans. Soybeans are a relatively homogeneous commodity; the product varies little across producers. There are limited informa- tional disparities, and entry as well as exit are essentially costless.

In competitive markets, individual buyers and sellers take the market price for the product as given—no single participant has any real control over price. If a seller charges more than the market price, buyers simply will purchase the product from other suppliers. And firms always can sell their output at the market price; thus they have no reason to offer discounts to attract buyers. In this setting, firms view their de- mand curves as horizontal—a firm can sell any feasible output at the market price, P*—but sells no output at a price above P*. Figure 6.1 illustrates a horizontal demand curve. With a horizontal demand curve, both marginal revenue (MR) and average revenue (AR) equal price.

Firm Supply

Short-Run Supply Decisions In the last chapter, we saw that a firm’s profit is maximized at the output where mar- ginal revenue equals marginal cost. The intuition of this result is straightforward—it makes sense to expand output as long as incremental revenue is greater than

2The specific characteristics of a product often vary across firms. Knowing which firms and individuals to

group together as a market, therefore, is not always straightforward. As discussed in Chapter 4, cross

elasticities are helpful in defining markets. Products with high cross elasticities can be considered in the same

market because they are “close substitutes.”

196 Part 2 Managerial Economics

incremental cost. Past this point, profits decline with additional output since incre- mental revenue is less than incremental cost. In a competitive market, marginal rev- enue is equal to price (P ). In the short run, the firm takes its plant size (and possibly other inputs) as given. The relevant cost is short-run marginal cost (SRMC). The condition for short-run profit maximization in a competitive industry is

P* � SRMC (6.1)

This condition—one of the more important propositions in economics—indicates that at any price, a competitive firm should produce the output where price equals short-run marginal cost. The firm, however, has the additional option of producing no output at all. When the price of the product is insufficient to cover its average variable cost (AVC), the firm is better off if it ceases production. With no output, the firm loses money since it generates no revenue to cover its fixed costs. However, this loss is smaller than the one it would incur if the firm produced any other level of out- put (since revenue from sales would be lower than its variable production costs). Hence the shutdown condition for the short run is

P* � AVC (6.2)

A firm’s supply curve depicts the quantity that the firm will produce at each price. Therefore the firm’s short-run supply curve is that portion of its short-run marginal cost curve above average variable cost. Figure 6.2 highlights this supply curve.

Long-Run Supply Decisions Firms can lose money in the short run yet still find it optimal to stay in business. In the long run, however, a firm must be profitable or it is better to exit this market. Price must equal or exceed long-run average cost (LRAC). Thus, the shutdown condition for the long run is

P* � LRAC (6.3)


P *


P ric

e (in

d ol

la rs


Quantity (market)

Di = MRi = ARi




P *


Quantity (firm i )

Figure 6.1 Firm Demand Curve in Perfect Competition

In competitive markets, firms take the market price of the product as given. The demand curve is horizontal. Both marginal revenue and average revenue are equal to the market price.

Chapter 6 Market Structure 197

In the long run, a firm can adjust its plant size. The long-run supply decision of a firm is based on long-run marginal costs (LRMC). The long-run supply curve of a firm is that portion of its long-run marginal cost curve above long-run average cost. This supply curve is depicted in Figure 6.3.3





C os

ts p

er u

ni t o

f o ut

pu t

(in d

ol la

rs )

Quantity (firm i )

Figure 6.2 The Firm’s Short-Run Supply Curve

The firm’s short-run supply curve is the portion of the short-run marginal cost (SRMC) curve that is above average variable cost (AVC). At prices below average variable cost, the firm is better off not producing any output.

3There is no inconsistency between short-run and long-run profit maximization. The LRMC at any given output

is equal to the SRMC, given that the firm has the optimal plant size for the output. Hence, the firm

simultaneously can choose an output where P* � SRMC � LRMC.




C os

t p er

u ni

t o f o

ut pu

t (in

d ol

la rs


Quantity (firm i )

Figure 6.3 The Firm’s Long-Run Supply Curve

The long-run supply curve for firm i is the portion of the long-run marginal cost (LRMC) curve that is above long-run average cost (LRAC). If price is below LRAC, the firm should go out of business.

198 Part 2 Managerial Economics



$ $

Qi Q


P *1

P *0P *0

P *1



Q* i1 Q*

i 0 Q* 0 Q*


P ric

e an

d co

st p

er u

ni t o

f o ut

pu t

(in d

ol la

rs )

Quantity (firm i ) Quantity

Figure 6.4 Competitive Equilibrium

The left panel illustrates the long-run supply decision of firm i, a representative firm in the industry. In the right panel, supply and demand curves (labeled S0 and D0) determine the market price, P*0. At the price, P*0, the firm produces Q*i0. At the price P*0, the firm is earning an economic profit. This economic profit is the profit per unit (P*0 � LRACi) times the total output Q*i0 and is depicted by the shaded rectangle. Economic profits will motivate other firms to enter the industry. This entry will shift the supply curve to the right and lower the price. Additional entry will occur up to the point where there are no economic profits. This condition occurs at a price of P*1. Here, there are no incentives for firms to enter or leave the industry (incumbents are earning a normal rate of profit and inventories are stable at their desired levels), and the market is in equilibrium. In a competitive equilibrium, firms produce output at the low point on their average cost curves (P*1 � LRMCi � LRACi). Thus, the equilibrium is associated with efficient production.

Competitive Equilibrium

In Chapter 3, we explained that the market price in a competitive market is determined by the intersection of the industry demand and supply curves. The industry demand curve depicts total quantities demanded aggregated across all buyers in the marketplace at each price. Similarly, the industry supply curve is the sum of all individual supply decisions (discussed earlier). For example, if there are 100 firms in the industry and each produces 20 units at a price of $10, the industry supply at that price will be 2,000 units.

For a graphical illustration of a competitive equilibrium, consider the supply and demand curves, labeled S0 and D0 in the right panel of Figure 6.4. Here, the market price is P*

0. The left panel depicts the long-run supply decision of a typical firm in the industry, firm i. At the price P*

0, firm i produces the quantity of output Q*i0. Cost curves are defined to include a normal rate of profit (a normal return on capital is one component of LRAC). Thus, at the price P*

0, firm i is earning an economic profit (above normal profit). This economic profit is the profit per unit (P*

0 � LRAC) times the output Q*i0 and is depicted by the shaded rectangle. The existence of economic profits will motivate other firms to enter the industry.4 This entry will shift the supply

4Profits reported by firms are based on the accounting definition: Sales revenue minus the explicit costs of doing

business. The calculation of accounting profits, therefore, does not include the opportunity cost of the owner’s

entrepreneurial effort or equity capital. Economic profits include these costs. Positive economic profits attract

entry because the returns are higher than the returns in the alternative activities. Positive accounting profits do

not always invite entry—the returns do not always cover the opportunity costs of the owners.

Chapter 6 Market Structure 199

curve to the right; inventories will build above their desired levels because of the in- creased production; hence firms will lower price. Additional entry will occur up to the point where there are no longer economic profits. This condition is pictured in Figure 6.4 at a price of P*

1. Here, there are no incentives for firms to enter or leave the industry (incumbents are earning a normal rate of profit); inventories are stable at their desired levels and the market is in equilibrium. In a competitive equilibrium, firms produce output at the minimum point on their average cost curves (P* � LRMC � LRAC). Thus, this equilibrium is associated with efficient production.

Constant versus Increasing Cost Industries Standard supply and demand graphs, such as Figures 6.1 and 6.4, usually picture in- dustry supply curves as upward sloping. While short-run industry supply curves typ- ically slope upward, long-run industry curves in many industries are essentially hor- izontal (flat)—supply is perfectly elastic. Whether the long-run industry supply curve is horizontal or upward sloping depends on the incremental costs of expanding industry output.

In the short run, firms have some fixed inputs and entry or exit of firms have not had time to occur. The Law of Diminishing Marginal Returns states that the marginal products of variable inputs will eventually decline as output increases. A firm’s short-run marginal costs will increase once this output level is reached (since the variable inputs are less productive, but cost the same). Upward sloping short-run marginal cost curves imply that both firm and industry supply curves will generally slope upward in the short run (as pictured in the standard graphs). An increase in de- mand for the product will cause the short-run equilibrium price to increase, and ex- isting firms will earn economic profits in the near term. Economic profits, in turn, will induce entry and a resulting long-run fall in price.

Whether the long-run industry supply curve is flat or upward sloping depends on whether the firm is a constant-cost or increasing cost industry. In a constant-cost in- dustry, all existing firms and potential entrants face the same average costs of produc- tion independent of the number of firms in the industry. For instance, they might all have identical cost curves to those pictured for the representative firm in the left panel of Figure 6.4. As an example, consider fast-food restaurants in a large market area. A new restaurant entering this market will likely incur nearly identical costs to existing firms for retail-building space, labor, wholesale food products, and other inputs re- quired for operation. In addition, wage rates and the prices of other inputs are unlikely to be affected with more or fewer restaurants. Thus, the average cost of producing meals is likely to be independent of the number of restaurants in the market area. With


Entry in Low-Carb Food An estimated 32 million Americans are following the Atkins Diet and spending $2.5 billion a year on low-carb foods.

Atkins Nutritionals sells 120 products and has licensed its name to dozens of companies. “There’s not much growth in

the food industry and Atkins is getting most of it,” says John McMillin of Prudential Securities. But food giants from

Kraft to General Mills are beginning to offer competing products. For instance, Heinz is introducing its One Carb

ketchup. “Competition is inevitable,” says Atkins President, Scott Kabak.

Source: B. Grow (2003), “The Low-Carb Food Fight Ahead,” BusinessWeek (December 22), 48.

200 Part 2 Managerial Economics

constant costs, the long-run industry supply curve is horizontal at a price equal to the common minimum long-run average cost for firms in the industry (P*

1 in Figure 6.4). At higher prices, existing firms earn economic profits, which promotes entry and downward pressure on price. At lower prices, firms do not cover average costs and thus are unprofitable. Some will exit the industry putting upward pressure on price.

With a horizontal (perfectly elastic) long-run supply curve, the long-run market price does not change as demand for the product increases or decreases (price is de- termined by costs, not demand). Rather a change in demand affects only the quantity produced and the number of firms in the industry. For example, if the minimum av- erage cost for firms in a constant-cost industry is $50, the long-run equilibrium price will be $50. This is the only price where no firms have incentives either to exit or to enter the industry. If demand increases or decreases, the price will stay at $50 but the equilibrium quantity will adjust (draw a horizontal supply curve at $50 and view what happens as the demand curve shifts). Our analysis in Chapter 3 describes the ability of suppliers to pass on a per-unit cost increase depends on the relative elastic- ities of supply and demand curves. Perfectly elastic long-run supply curves imply that a cost increase, such as a government excise tax, will ultimately be borne by consumers in the form of higher prices.

In increasing cost industries, the costs of production increase as the number of firms in the industry increases. For example, in many agricultural markets additional output must be produced on less productive farmland, which implies higher per-unit costs. In such an industry, price must increase to induce additional production (since the current price is below the incremental cost of producing additional output). In this case, the long-run supply curve is upward sloping. Here a change in demand will affect both the equilibrium quantity and price.

As we will discuss in more detail in Chapter 8, existing firms in an increasing cost industry often do not earn economic profits when the product price increases due to an increase in demand. Rather, the prices of the inputs that made the initial low-cost production possible (e.g., a particularly productive piece of farmland will sell for a higher price than less-productive farmland) are likely to increase, so that in equilibrium no firm has a cost advantage or economic profits. If the product and

Phantom Freight Most plywood in the United States is produced in the Pacific Northwest. Due to this dominance, plywood prices

throughout the country are essentially the Northwest price plus shipping. If this condition did not hold, Northwest

suppliers would curtail shipping plywood to cities with low prices and increase shipping to cities with high prices. The

changes in supply would affect the prices in the cities until, in equilibrium, the prices across cities would differ only by

transportation costs.

In a U.S. court case, Southeast timber producers were sued for charging customers Northwest’s price plus shipping

and then delivering locally produced plywood. It was ruled that these companies were making unjust profits because they

did not actually incur the shipping costs. The jury awarded billions of dollars to the customers. Were these companies

really making economic profits? The answer is probably not. The local production in the Southeast had a shipping

advantage to Southeast customers. The factor that made this advantage possible was scarce timber land in the Southeast.

Presumably, the price of this scarce timber land was bid up to the point where plywood producers were making only a

normal profit given the prevailing price for plywood in the Southeast (which was the Northwest price plus shipping).

Source: A. Alchian and W. Allen (1983), Exchange and Production: Competition, Coordination and Control (Wadsworth Publishing:

Belmont), 228–231.


Chapter 6 Market Structure 201

all the relevant input markets are perfectly competitive, generally no firm in the in- dustry will be able to make economic profits over the long run. It is important for managers to consider this likely reality when they make investment and other strate- gic decisions.

Strategic Considerations Although few markets exactly match economists’ idealized conditions for perfect competition, many markets approximate this structure. In most industries, there are strong competitive forces that reduce economic profits over time. These forces imply that many strategic advantages (e.g., being the first in a new market) are likely to be short-lived. If the conditions in the market resemble the competitive model, it is im- portant to move quickly to take advantage of transitory opportunities. In addition, potential entrants should realize that observed economic profits in an industry are likely to be bid away as time passes. This consideration can affect both long-range capital spending and entry decisions. For instance, given the increased competition, cable TV companies increased the level of scrutiny they applied to internal invest- ment proposals. In a competitive market, firms must strive for efficiency and cost control; inefficient firms lose money and are forced out of the market.

Barriers to Entry5

Although the competitive model is a reasonable approximation in many markets, there are other industries where firms have notable market power—output decisions of individual firms have a noticeable impact on prices. A necessary condition for market power to exist is that there are effective barriers to entry into the industry.

To understand what constitutes an effective entry barrier, it is useful to consider the decisions of individual firms to enter an industry. Firms consider entering a new market when they observe extant firms reporting large profits. For instance, if Wen Ho observes a firm such as a cable TV company reporting large profits, his firm (like a number of other firms) is likely to consider entering the industry. Entry decisions depend on three important factors: First, Wen will be concerned about whether his entry will affect product prices. This depends, at least in part, on how existing firms are likely to respond to a new entrant. For example, are they likely to cut prices? Sec- ond, Wen will be concerned about incumbent advantages. Do existing firms have


The WSJ recently presented data suggesting that

United Airlines was not covering its costs on flights

from San Francisco to Washington D.C. The article

quoted analysts saying that United should discon-

tinue this service. The costs per flight (presented in

the article) included the costs of fuel, pilots, flight

attendants, food, etc. used on the flight. They also

included a share of the costs associated with running

the hubs at the two airports, such as ticket agents,

building charges, baggage handlers, gate charges,

etc. Suppose that the revenue collected on the typi-

cal United flight from San Francisco to Washington

does not cover these costs. Does this fact imply that

United should discontinue these flights? Explain.

5This section provides a brief summary of the literature in economics on barriers to entry; it draws on S. M.

Oster (1994), Modern Competitive Analysis (Oxford Press: New York).

202 Part 2 Managerial Economics

advantages that an entering firm will have difficulty duplicating—ones that make it unlikely that the new firm will enjoy similar profits? Third, Wen will be concerned about costs of exit. How much will it cost to leave the industry if this incursion fails? We discuss each of these factors in turn.

Incumbent Reactions

Specific Assets Specific assets are assets that have more value in their current use than in their next best alternative use. Consider the case of the Alaskan Pipeline. It has a high value in its current use. Yet it is completely specialized for transporting oil from the North Slope to Prudoe Bay—it has virtually no other use. Moreover, it could be moved only at enormous expense. If existing firms in an industry have invested heavily in assets quite specialized to that market, they are likely to fight harder to maintain their positions than if their assets are less specific and can be shifted at low cost to alter- native activities.

Scale Economies Industries with significant economies of scale have minimum efficient scales that occur at high output levels. In such industries, a new entrant must produce at high volume to be cost-effective. Large-scale production is more likely to have a material effect on price. For example, if the minimum efficient scale is 30 percent of total mar- ket demand, price certainly will decline if a new entrant tries to capture such a large share of the market—its entry undoubtedly would trigger vigorous price competition from incumbents. Note that the absolute size of the minimum efficient scale is not as important as is this scale relative to the size of the total market. Minimum efficient scale varies enormously across industries. In one study, estimates of minimum efficient scale, as a percentage of industry capacity, ranged from 0.5 percent (fruit/vegetable canning) to 33 percent (gypsum products).6 Globalization of markets increases effective market size, thereby reducing this entry barrier—for example, consider the size of American versus global automobile markets.

Reputation Effects Potential entrants can be influenced by the reputations of existing firms in the indus- try for reactions to new entrants. In certain circumstances, it can pay for an existing firm to react more aggressively than would be implied by considering only its immediate interests. For example, facing a new rival, the firm might engage in extensive price cutting to establish a reputation as a formidable competitor. Note, however, that threats by firms to cut prices if entry occurs sometimes lack credibil- ity. If new firms actually enter, existing firms might not follow through with their threats because they would be harmed by their own price cuts. Thus, it can be rea- sonable for a potential entrant to ignore threats—if the entrant believes that incum- bents are bluffing. We examine these considerations in greater detail in Chapter 9.

Excess Capacity If firms with excess capacity cut production, they can be confronted with much higher average costs (depending on the slopes of their average cost curves). Also, firms with excess capacity are better able to satisfy the demands of new customers

6K. Lancaster and R. Dulaney (1979), Modern Economics: Principles and Policy (Rand McNally: New York),


Excess Capacity at Alcoa In 1940, Alcoa Aluminum lost an important antitrust case involving its production strategy of maintaining excess

capacity. The judge ruled that he could think of no better “effective” deterrent to entry.


Chapter 6 Market Structure 203

should they lower price and force a rival out of business. Potential competitors, therefore, may be less likely to enter when there is excess capacity in the industry be- cause they anticipate more aggressive reactions on the part of incumbents.7 Excess capacity frequently exists for completely innocuous reasons. For example, a firm facing cyclical production or anticipating growth has excess capacity over some time spans because it has invested in additional capacity to satisfy peak demands better. In other cases, excess capacity may be chosen specifically to deter entry.

Incumbent Advantages

Precommitment Contracts Existing firms often have long-term contracts for raw materials, distribution outlets, shelf space, and delivery of the final product. These contracts can serve as a deterrent to entry, since they limit the opportunities for customers and suppliers to switch from incumbent firms to new entrants.

Licenses and Patents Sometimes, entry is limited through government restrictions such as licensing require- ments and patents. For instance, the number of doctors is limited effectively by state medical licensing requirements. This restriction allows doctors to charge higher prices than if entry were unrestricted. Regulators and licensed physicians justify such restric- tions with arguments based on consumer protection. Yet, whether or not consumers benefit from stringent licensing is debatable—given that they pay higher prices.

Normal patent life is 17 years. Over this period, other firms are prohibited from copying the innovation; thus a patent provides a firm with potential market power. Patents also provide important incentives to innovate. From a practical standpoint, the effectiveness of a patent in blocking entry varies dramatically (some patents can be circumvented by clever design, e.g.).

Learning-Curve Effects In Chapter 5, we discussed how average costs are reduced in some industries through production experience. As production experience accumulates, the firm learns how to lower unit costs. Learning-curve effects can result in new rivals having a cost disad- vantage relative to existing firms. Whether these effects are important depends on whether the new entrants simply can copy the techniques learned by existing firms through their experience.

Pioneering Brand Advantages Sometimes, a firm benefits from being first in an industry. In some industries—over- the-counter drugs, for example—a satisfied customer might be reluctant to switch

7Excess capacity can occur because of significant declines in industry demand. In this case, profits are likely

to be low and entry will not be attractive. Our current discussion focuses on cases where incumbents are

making economic profits and have excess capacity. These economic profits might not induce entry because

of the fear of price cutting by incumbents.

204 Part 2 Managerial Economics

brands even if the price of a competing product is substantially lower. This tendency is likely to be strongest in experience goods, which have to be tried by the customer to ascertain quality. For instance, customers might hesitate to try a new pain reliever because they fear that it might not be as effective as their regular brand. Where qual- ity can be judged by inspection prior to purchase, this advantage of incumbents is lower. Sometimes the incumbent’s advantage with an experienced good can be over- come by a new entrant through free samples, endorsements, or government certifica- tion. Each of these methods entails additional costs—these costs of overcoming in- cumbent advantages deter entry.

Exit Costs

Another important entry consideration centers on the costs of exit. In some indus- tries, it is possible to “hit and run.” For instance, forming a new company to seal as- phalt driveways requires little investment in specialized equipment or training. A new firm can enter quickly when the profit potential is high and exit at low cost if profits decline. In other industries, especially those with specific assets, exit costs can be high. In such industries firms bear significant costs, such as moving employ- ees to new locations and liquidating plants and other assets when they decide to exit. High exit costs deter initial entry.

Monopoly Effective barriers to entry limit the threat of competition and give incumbent firms market power. Although competitive markets are at one end of the spectrum, at the other end is monopoly—where there is but a single firm in the industry. Here, indus- try and firm demand curves are one and the same.


Government Restrictions on Exit Some regulators want to restrict companies from closing plants. These regulators appear motivated by concerns over

people who lose their jobs when a company closes a plant. Restrictions on plant closings, however, are likely to reduce

the desirability of entry into an industry—firms will be reluctant to enter an industry if they cannot exit easily when

they are losing money. Thus, potential effects of government restrictions on exit are less vigorous competition in the

affected industries, higher consumer prices, and lower levels of employment.


Entry in Consumer Electronics Since its founding two decades ago, Xoceco Inc. has evolved from producing low-cost color TVs for the Chinese

market to producing flat-screen TVs. They plan to market in the United States by supplying companies like Dell or

Hewlett-Packard. Rather than spending lavishly in chips and software to power their products, they buy the

components, assemble the gadgets, and undercut the industry leaders’ prices. Thus, consumer electronics leaders like

Sony and Matsushita are threatened the way IBM was by the rise of the PC clone.

Source: E. Ramstad and P. Dvorak (2003), “Off-the-Shelf Parts Create New Order in TVs, Electronics,” The Wall Street Journal (December 16), A1.

Chapter 6 Market Structure 205

Profit Maximization Suppose that a monopolist charges the same price to all customers. (As discussed shortly, such a pricing policy might be motivated by either government regulation or the inability to prevent resale among customers; in Chapter 7, we relax this restric- tion.) The firm’s objective is to choose the price–quantity combination along the de- mand curve that maximizes profits. This combination occurs where marginal revenue equals marginal cost.

For purposes of illustration, consider the following linear demand curve:

P � 200 � Q (6.4)

(Assume that marginal cost is constant at $10.) Recall that the marginal revenue curve for a linear demand curve is a line with the same intercept and twice the negative slope. Figure 6.5 displays the demand curve, marginal revenue curve, and marginal cost curve in this example. Optimal output occurs at 95 units—where MR � MC. To sell this output, the firm charges a price of $105. The firm makes $95 profit per unit ($105 � $10) for a total profit of $9,025 ($95 � $95); this is indicated by the shaded rectangle abcd.

As opposed to pure competition, monopolistic suppliers charge customers more than the marginal and average costs of production and distribution; the firm thus earns an economic profit. Monopolies restrict output compared to competitive in- dustries. In our example, if the industry were competitive, the market price would be $10 (marginal cost) and total quantity sold would be 190 units.

Unexploited Gains from Trade Given the monopolist’s output and pricing choices, some consumers are willing to pay more than the marginal cost of production and distribution, yet do not purchase the product. Thus, not all gains from trade are exhausted. The associated loss in




d MC = AC







P ric

e an

d co

st p

er u

ni t o

f o ut

pu t




Figure 6.5 Monopoly

This figure illustrates the price and output decisions of a monopolist. In the example, demand is P � 200 � Q. Marginal costs are $10. The profit- maximizing output occurs at 95 units, where MR � MC. To sell this output, the firm charges a price of $105. The firm makes $95 per unit profit ($105 � $10) for a total profit of $9,025 ($ 95 � 95), as indicated by the shaded rectangle abcd. Some consumers are willing to pay more than the marginal cost of production, yet do not receive the product. The associated loss in potential gains from trade is pictured by the shaded triangle cde. The firm does not lower the price to sell to these consumers because it does not want to lower the price for other customers.

206 Part 2 Managerial Economics

potential gains from trade is pictured by the shaded triangle cde in Figure 6.5. Consumers along this segment of the demand curve value the product at more than $10 but less than $105. The firm does not lower the price to sell to these consumers because it does not want to lower the price for other customers (recall that in this chapter we presume the firm charges the same price to all customers). From the firm’s standpoint, any gain from selling to additional customers would be more than offset by the loss from lowering its price to all its customers.8

Monopolistic Competition As the name implies, monopolistic competition is a market structure that is a hybrid between competition and monopoly. In this market structure, there are multiple firms that produce similar products. There is free exit from and entry into the industry. Yet competition does not eliminate market power because the firms sell differentiated products. Examples include retail shops, books, movies, gasoline stations (differen- tiated by location and brand), and business schools. It is the most common market structure. For instance, although many shoe stores compete for customers at a large mall, they are not generally viewed as perfect substitutes. Some position themselves as discount stores, while others target customers who are willing to pay high prices for upscale, branded products. Some stores focus on selling athletic shoes, while oth- ers target conservative business customers. Some focus on women, while others focus on men or children. If any of the existing stores are highly profitable, new stores can enter to compete. These companies thus have some market power. New

ANALYZING MANAGERIAL DECISIONS: Pricing and Investment Decisions

You work for a drug manufacturing company that

holds a patent on Hair Grow, the world’s most

effective drug for restoring hair. Your job is to

analyze the pricing and investment decisions facing

the firm. Your marketing group estimates that Hair

Grow has the following demand curve:

P � 101 � .00002Q

1. Your marginal cost for producing a Hair Grow

pill is $1. What is the profit-maximizing price

and quantity? What is your profit?

2. Suppose that your production facility can

only produce 1,000,000 pills. What is

your optimal price and quantity given the

production constraint? What are your


3. Suppose that you could increase the capacity

of your plant to 3,000,000 pills within a two-

year period for a cost of $30,000,000. Should

you undertake the investment (for simplicity,

assume you can borrow the funds for the ex-

pansion at a 0 percent interest rate)?

8Economists frequently refer to these lost gains from trade as a deadweight loss. This inefficiency (or social cost) is one reason why governments might pass regulations like antitrust laws to restrict the formation of

monopolies. But these regulations also can be motivated by concerns about the higher prices that consumers

pay when they face monopolistic suppliers. Although government regulation has the potential to reduce

inefficiencies and wealth transfers from consumers to firms, it is important to keep in mind that government

regulation is not costless. There are salaries for regulators and court costs, for instance. From a societal

viewpoint, the costs of government regulation should be weighed against the benefits. These issues are

discussed in greater detail in Chapter 21.

Chapter 6 Market Structure 207

stores are likely to enter the industry if the existing firms report large profits—there are no significant barriers to entry.

Monopolistic competition is similar to monopoly in that firms under both market structures face downward-sloping demand curves: A toothpaste company can raise its price without losing all sales. Given that the firms face downward-sloping de- mand curves, each strives to select the price–quantity combination that maximizes its profits. The output decision is based on the same analysis as for the pure monop- olist—choose that output where MC � MR.

The difference between monopoly and monopolistic competition is that in mo- nopolistic competition, economic profits invite entry and imitation. If a shoe store earns high profit from carrying a new product line, others will have an incentive to compete by offering that product line or a very similar line. Entry and imitation will shift the original firm’s demand curve to the left and reduce profits. Zero economic profits exist when the demand curve is shifted to the point where average cost equals price. Figure 6.6 shows this condition.


Monopolistic Competition in Golf Balls There are many brands of golf balls. Some golfers view the balls as perfect substitutes and simply purchase the lowest-

priced brand. Other golfers prefer one brand to another. For instance, they might believe one brand of ball flies farther

or provides greater control than competing brands. These golfers are willing to pay a higher price for their favorite ball

than for competing balls. However, they often will substitute if the price difference is more than a few dollars a dozen.

Also, if a company develops some popular feature, like a larger number of dimples on the ball, the feature is typically

copied by other companies within a short time period. Since a golf equipment company has a monopoly in producing

its own brand, it has some market power. However, this power is limited given the competition in the industry.





MRi Q*

P *


P ric

e an

d co

st p

er u

ni t o

f o ut

pu t

(in d

ol la

rs )

Quantity (firm i )



Figure 6.6 Monopolistic Competition

In monopolistic competition, firms sell differentiated products. This figure shows the demand curve for firm i in such an industry. The curve is downward-sloping. Similar to monopoly pricing, the firm selects the output where marginal revenue equals marginal cost. Monopolistic competition differs from monopoly in that abnormal profits will invite entry. Entry shifts the demand curve for the firm to the left (as some of the customers buy from the new firms). The firm makes no economic profits when price is equal to average cost. This condition occurs at price P*i and quantity Q*i.

208 Part 2 Managerial Economics

These competitive responses will tend to force profits to zero. Yet some brands con- tinue to be more distinctive than others. Also, costs can vary because of differences in production techniques and inputs. It is possible for some firms to earn economic prof- its in monopolistic competition.9

Oligopoly Within oligopolistic markets, only a few firms produce most of the output. Examples of oligopolistic industries include automobiles and steel during the 1950s. These in- dustries had important scale economies and other substantial entry barriers. In 1995, the top four cereal makers in the United States produced about 90 percent of industry output, while the top eight accounted for virtually all production. Products may or may not be differentiated. Firms can earn substantial profits. These profits are not reduced through new entry because of effective entry barriers. Yet as we shall see, economic profits sometimes can be eliminated in oligopolistic industries through competition among the existing firms.

In our analysis of other market structures, we assume that firms take the prices of their competitors as given. A firm was not expected to respond to announcements of changes in prices by rival firms. This assumption certainly is reasonable in the case of competitive markets with many small firms, as well as in the case of monopoly with only one large firm. But this assumption rarely is valid within oligopolistic in- dustries. For instance, when American Airlines considers lowering its prices on par- ticular routes, it obviously must be concerned about whether United Airlines and its other competitors will follow suit. In fact, firms in oligopolistic industries ordinarily will be quite concerned about how their rivals will react to most major policy deci- sions, be they advertising campaigns or product design decisions. Decision making within these industries requires strategic thinking. Decision makers must realize that competitors are rational parties operating in their own self-interest. Thus, it is impor- tant for decision makers to place themselves in their rivals’ positions and consider how they might react. (This basic principle, which we now examine briefly, is devel- oped more completely in Chapter 9.)10

Nash Equilibrium

To analyze oligopolies, we need an underlying principle to define an equilibrium when rival firms make decisions that explicitly take each other’s behavior into ac- count. Previously, we used the concept that a market is in equilibrium when firms are doing the best they can, given their circumstances and have no reason to change price

9Monopolistic competition does not exhaust all gains from trade for two reasons. First, as in monopoly, the

firms do not sell to all consumers who value the product at above marginal cost. Second, firms do not operate

at the bottom of their average cost curves (see Figure 6.6). Lower average cost would be obtained with fewer

firms, each producing more output. Nonetheless, regulation to address these inefficiencies is unlikely to be

effective. Consumers value product differentiation and are arguably better off with more variety at slightly

higher average cost than with lower variety produced at lower average cost. Second, with few entry barriers,

the market power of firms is unlikely to be great. 10This chapter presents a basic introduction to game theory. The material provides sufficient background for

the game theory applications found in subsequent parts of the book (these are in the appendices of several

chapters). Chapter 9 extends this introduction of game theory and discusses in more detail how managers

might use this theory as a tool in decision making. Readers interested in a more detailed treatment of game

theory should read Chapter 9.

Chapter 6 Market Structure 209

or output. For example, in a competitive equilibrium, there is no reason for entry or exit (existing firms are making “normal” profits). No existing firm has any reason to change its output level (all are producing where MC � MR � P* and inventories are stable at their desired levels).

We can apply this same basic idea to oligopolistic markets with minor modifica- tion. In the following analysis, a firm does the best it can, given what its rivals are doing. In doing so, the firm anticipates that other firms will respond to any action it takes by doing the best they can as well. Actions are noncooperative in that each firm makes decisions that maximize its profits, given the actions of the other firms. The firms do not collude to maximize joint profits. An equilibrium exists when each firm is doing the best it can, given the actions of its rivals. Economists call this a Nash equi- librium for Nobel laureate John Nash who first developed these general concepts.

To illustrate this approach, assume a simple setting: There are two firms in an indus- try—a duopoly. Each independently chooses a price for an identical product. The firms choose either a high price or a low price. The payoffs are given in Figure 6.7. (The entry on the upper left in each cell is for WonCo, while the entry on the lower right is for TuInc.) For example, if both firms charge a high price, WonCo’s profits are $400 and TuInc’s profits are $200.11

The equilibrium is for WonCo to charge a high price and TuInc to charge a low price. Any other combination is unstable: Given the action of one of the firms, the other firm has the incentive to change its price. For instance, if both firms charge a high price, it is in the interests of TuInc to lower its price—its profits go from $200 to $250. The other combinations of WonCo charging a low price and TuInc a high price and both firms charging a low price are similarly unstable: Each firm has an incentive to alter its price given the other firm’s choice. A Nash equilibrium is self-enforcing. If WonCo charges a high price, it is optimal for TuInc to charge a low price. Similarly, if TuInc charges a low price, it is optimal for WonCo to charge a high price. Given the choice of one firm, there is no reason for the other to alter its strategy.



TuInc—Low price

W on

C o—

Lo w

p ri




TuInc—High price

W on

C o—

Lo w

p ri




TuInc—Low price

W on

C o—

H ig

h pr

ic e



TuInc—High price

W on

C o—

H ig

h pr

ic e

Figure 6.7 Nash Equilibrium

In this example, there are two firms in an industry—WonCo and TuInc. Each independently chooses a price for an identical product. The firms choose either a high price or a low price. The payoffs are given in the table (the upper-left entry in a cell displays the profits for WonCo, the lower right shows the profits for TuInc). The equilibrium is for WonCo to charge a high price and TuInc to charge a low price—the shaded cell. Any other combination is unstable: That is, given the action of one of the firms, the other firm has the incentive to deviate. This equilibrium is called a Nash equilibrium.

11The profits differ due to differences in the underlying production costs.

210 Part 2 Managerial Economics

In this example, the Nash equilibrium is not the outcome that maximizes the joint profits of the two companies. Combined profits would be higher if both firms charged a high price. Conceptually, the combined profits under this pricing policy could be split in a manner that would make both firms better off than in the Nash equilibrium. For instance, the combined profits of $600 could be split, with each firm receiving $300. As this example illustrates, noncooperative equilibria do not neces- sarily maximize the joint value of the firms. (Since the potential gains from trade are not exhausted, it is often the case that one or more firms can be made better off, with- out making other firms worse off, by changing the joint decisions.)

Output Competition

The first major analysis of oligopoly was published by Augustine Cournot in 1838. To illustrate his model, suppose again that there are only two firms in the industry and that they produce identical products. In the Cournot model, each firm treats the output level of its competitor as fixed and then decides how much to produce. In equilibrium, neither firm has an incentive to change its output level, given the other firm’s choice. (Thus, this is a Nash equilibrium.)

Suppose the duopolists face the following total industry demand:

P � 100 � Q (6.5)

where Q � QA � QB. For simplicity, assume that both firms have marginal costs of zero: MCA � MCB � 0. Each firm takes the other firm’s output as fixed. Thus, the anticipated demand curve for firm i (i � A or B) is

Pi � (100 � j) � Q i (6.6)

where j � expected output of the other firm. The marginal revenue for firm i is12

MRi � (100 � j) � 2Q i (6.7)

Firm i ’s profits are maximized by setting marginal revenue equal to marginal cost (in this case, zero). Doing so, and rearranging the expression, yields the following reaction curve:

Q i � 50 � 0.5Q j (6.8)

The reaction curve indicates firm i ’s optimal output given the output choice of firm j. Both firms have the same reaction curve in this example, except that the subscripts are reversed.

Equilibrium occurs where the two curves cross. At these output levels, each firm is maximizing profit given the other firm’s output choice. Neither firm has an incentive to alter its output. The equilibrium is pictured in Figure 6.8. In equilibrium, each firm produces 33 units for a total output of 66 units; the price is $33.34. This output level is lower than in a competitive market. With competition, total output would be 100 units and the price would be zero (where P* � MC). In the Cournot equilibrium, firms make economic profits: Price is $33.34; average costs are zero. Each firm thus reports profits of $1,110.89. This profit is lower than the two firms could obtain if they directly col- luded and jointly produced the monopolistic output of 50 units (e.g., 25 units per firm). With effective collusion, joint profits would be $2,500 rather than $2,221.78. Figure 6.9




12Recall that marginal revenue for a linear demand curve is a line with the same intercept, but twice the

negative slope.

Chapter 6 Market Structure 211



Competition MC = 0

66.66 100 Q


50 0






P ri

ce (

in d

ol la

rs )

Figure 6.9 Comparison of Prices and Outputs among Collusive, Cournot, and Competitive Equilibria

In this example, the total industry demand curve is P � 100 � Q. Marginal cost is zero. The figure shows the price–quantity outputs for the industry under collusive, Cournot, and competitive equilibria. The output is smallest and the price is highest for the collusive equilibrium. The output is largest and the price is smallest for the competitive equilibrium.

Firm A’s reaction curve

Firm B’s reaction curve

a = Competitive equilibrium b = Cournot equilibrium c = Collusive (monopoly) equilibrium









25 50 100



Q ua

nt ity

o f

ou tp

ut b

y F

ir m


Quantity of output by Firm A

Figure 6.8 Cournot Equilibrium

The duopolists in this example face the total industry demand curve, P � 100 � Q, where Q is the sum of the two outputs. Both firms face a marginal cost of zero. The figure shows the reaction curves for each firm. The reaction curve indicates firm i’s optimal output given the output choice of firm j ( i, j � A or B). The Cournot equilibrium occurs where the two reaction curves cross. Each firm produces 33.33 units. The market price is $33.34. The output for the firms is lower and the profits are greater than in the competitive equilibrium. The output for the firms is greater and the profits are lower than in the collusive (monopoly) equilibrium.


In the simple Cournot model, firms make their

output choices simultaneously. In practice, firms

sometimes make these kinds of decisions


Suppose that you manage one of the firms

discussed in the Output Competition example in

the text. The industry demand in this example is

P � 100 � Q and the MC of each firm is zero.

1. Suppose that each firm must make an upfront

investment of $1,000 to enter the market and

that your competition has already paid this

investment and chosen to produce 50 units.

This investment is nonrecoverable (sunk).

Should you make the $1,000 investment and

enter the market? If so, how much should you

produce and what are your profits? Continue to

assume that your firm will survive for only one

production period.

2. How do your profits and those of your

competitor compare to the case of simultane-

ous decisions discussed in the text? Would you

say that this example of output competition has

a first mover advantage or disadvantage?

212 Part 2 Managerial Economics

displays the three price–quantity outcomes using the original industry demand curve. (This model can be extended readily to more than two firms. The same general results hold: As the number of firms grows, outcomes approach those of a competitive market.)

Price Competition

In the Cournot model, firms focus on choosing output levels. An alternative possi- bility is that firms might focus on choosing product price.13 Here, the Nash equilib- rium is for both firms to choose a price equal to marginal cost—the competitive out- come. To see why, suppose one of the firms chooses a price, P� � MC. In this case, it is optimal for the rival firm to charge a price just below P� to capture all industry

13This situation often is referred to as the Bertrand model. Bertrand was a French economist who wrote a short

note almost 50 years after Cournot’s work was published arguing that in some markets, producers set prices

rather than quantities.


Price Wars Firms in oligopolistic industries sometimes engage in intense price competition to their mutual detriment. Examples

include the airline and wireless industries. The major airlines, including American, United, Delta, and Northwest,

periodically enter fare wars that lower the price of air travel for consumers and lower the combined profits of the

airline industry. For instance, in 2005 Delta lowered its fares in the hope of gaining new passengers. This price

reduction was matched by other airlines within hours. In 2002 wireless carriers such as Sprint PCS, AT&T Wireless,

and Verizon stole each other’s customers offering look-alike calling plans and rock-bottom prices. As soon as one

company offered a new calling plan, the rest followed suit. Outside analysts generally agree that the firms lose profits

through these price cuts. Financial analysts following the stocks of wireless service companies recommended to their

clients that they reduce the amount invested in this sector because intense price competition was lowering profits.

Source: S. Mehta (2001), “First That Old Sinking Feeling,” Fortune (December 10), 34–35; and M. Rollins (2002), “Wireless

Services Second Quarter Preview: Less Than a Zero Sum Game,” Solomon Smith Barney Industry Note (July 8); J. Horwich (2005),

“Northwest Matches Delta’s Fare War . . . Reluctantly,” Minnesota Public Radio (January 5).

Chapter 6 Market Structure 213

sales. (Since we assume the firms’ products are identical, customers buy the product from the firm that offers the lower price.) Given that the second firm charges a price just below P�, it is now optimal for the first firm to charge a slightly lower price. This process continues; only when price equals marginal cost does neither firm have an incentive to lower price. (Lowering price further would result in selling below cost, thus generating a loss.) Of course, both firms would like to devise a way of avoiding competition and capturing higher profits. Yet as we discuss below, fostering cooper- ation can be difficult—and in certain cases, illegal.

Empirical Evidence

There are various economic models of oligopoly. We have presented but two of them simply to illustrate that economic theory does not make unambiguous predictions about what to expect within such industries. Some models yield outcomes close to pure competition—firms sell at marginal cost and make no economic profits. Other models yield outcomes closer to pure monopoly. What actually occurs in oligopolis- tic markets depends on the specific market and competing firms (there is no one model that fits all situations). It is ultimately an empirical issue. Available evidence

Price Competition with Differentiated Products We demonstrated in the main text how price competition could lead to zero economic profits in an oligopolistic

industry. The analysis focused on two firms producing homogenous (identical) products. Often competing firms

produce differentiated products. Products are horizontally differentiated when at the same price some consumers buy

one and some buy the other. For example, Ford and General Motors (GM) engage in price competition in most

segments of the car market, but consumers do not view their products as perfect substitutes. At the same price, some

would buy a Ford, while others would buy a Chevy (produced by GM). A given customer, however, might be induced to

purchase her less-preferred product with a sufficient price discount. Ford and GM are likely to consider this possibility

when they price their products. Economists refer to price competition in this setting as Bertrand Competition with Differentiated Products.

The basic economic model of this competition considers two firms, producing horizontally differentiated products.

The degree of differentiation ranges from D � 0 (all customers view the products as homogenous and thus perfect

substitutes) to D � 1 (complete differentiation, where a given customer does not substitute from one product to the

other based on price). Depending on the degree of differentiation, it is possible for both firms to earn long-run

economic profits in this setting, even if they compete on price. Both firms earn zero economic profits when they are

viewed as perfect substitutes (D � 0)—the case analyzed in the text. Profit increases with the degree of differentiation.

With complete differentiation (D � 1), each firm can simply price as a monopolist to the set of customers who prefer

its product (since customers do not view the products as substitutes). Profits fall with the degree of substitution since

the firms have increased incentives to lower price to attract additional customers from the other firm. In equilibrium,

price competition makes each firm worse off than if they had been able to collude on price.

Managers invest in marketing, advertising, packaging, and new product design to differentiate their products from

competitors. Interestingly, successful differentiation can benefit not only the investing firm, but also the competing

firms (there is a positive externality). The investing firm gains by having increased pricing power to extract profits from

customers who prefer its products. However, by definition the other firm is also more differentiated, which allows it to

extract additional profits from the customers who prefer its products. For example, suppose a firm is successful in

promoting its clothing line among teenagers. While teenagers would be more likely to buy the product, more

conservative adults might be more likely to buy the other.


214 Part 2 Managerial Economics

suggests that oligopolies typically result in less output than competitive markets and that firms earn economic profits—at least in some industries.14 Firms sometimes compete on price, to each other’s detriment, and normally earn less in aggregate than a monopolist could.

Cooperation and the Prisoners’ Dilemma

As we have discussed, in oligopolistic industries it is in the private interests of firms to find ways to cooperate and capture more profits than through competition. In prin- ciple, firms are most profitable if they effectively collude and act as a monopolist in jointly setting price and output for the industry. Collusion maximizes joint profits, which then can be divided among the firms in the industry. Many governments un- derstand these incentives and have passed a variety of antitrust laws to limit firms’ ability to engage in fixing prices. These laws are designed to lower the prices con- sumers pay for products. Some of the more restrictive of these laws have been adopted in the United States. Internationally, firms tend to have more latitude in forming cooperative agreements in attempting to increase profits—for example, con- sider the OPEC cartel.15

Prisoners’ Dilemma Even when free to cooperate, firms find that cooperation is not always easy to achieve. Individual firms have incentives to “cheat” and not adhere to output and price agreements. This incentive can be illustrated by the well-known prisoners’ dilemma. In the original prisoners’ dilemma, there are two suspects; hence, suppose the SEC has been investigating an insider trading scheme and their investigation is focused on two securities brokers, Avi Wasserman and Bea Haefner, who are arrested and charged with a crime. Police have insufficient evidence to convict them for in- sider trading violations unless one of them confesses. The police place Avi and Bea in separate rooms and try to get them to confess. If neither confesses, they are con- victed of less serious crimes associated with their trading activities and are sentenced to only 2 months in jail. If both confess, they spend 12 months in jail. However, if one confesses but the other does not, the confessor is released under a plea bargain in return for testifying while the other is sentenced to 18 months in jail—12 for the crime and 6 for obstructing justice. The payoffs in terms of jail time faced by each individual are displayed in Figure 6.10.

The Nash equilibrium is for both suspects to confess. Given these payoffs, it is al- ways in the individual interests of each suspect to confess, taking the action of the other party as given. If Avi does not confess, Bea is set free by confessing. Alterna- tively, if Avi confesses, Bea reduces her jail sentence from 18 to 12 months by also confessing. Either way, it is in Bea’s interests to confess—confessing is a dominant strategy. Since the payoffs are symmetrical, it also is optimal for Avi to confess. Although it is in the individual interests of each party to confess, it is clearly in their joint interests not to confess. By not confessing, each only serves 2 months in jail,

14D. Carlton and J. Perloff (1990), Modern Industrial Organization (HarperCollins: New York), Chapter 10,

discusses some of the relevant empirical literature. 15In smaller countries, much of the local production of key products is exported. In this case, it can be in the

countries’ interests to allow the formation of cartels. Ultimately, consumers pay higher prices and there are

inefficiencies. However, many of these costs are imposed on people in other countries.

Chapter 6 Market Structure 215

compared to 12 months if both confess. The prisoners’ dilemma suggests that any agreement for neither to confess is likely to break down when they make their indi- vidual choices unless there is some mechanism to enforce their joint commitment not to confess. (For particular crimes one such mechanism might be the Mob: Both suspects have incentives not to confess if they expect to be executed for providing evidence to the police.)

Cartels Cartels consist of formal agreements to cooperate in setting price and output levels. (These activities are generally illegal in the United States.) Firms trying to maintain cartels can face a problem like the prisoners’ dilemma—we might call it the cartel’s dilemma. Members can agree to restrict output to increase joint profits. However, in- dividual firms have incentives to cheat. If all other firms restrict output, prices will not be affected significantly by the extra output of one firm. However, that firm’s profits will increase from selling additional output at the cartel-maintained high price. But if all firms react to these incentives by increasing output, the cartel breaks down. Actual cartels often unravel because of such incentives. This outcome is

2 months

2 months

Bea—No confession Bea—Confession

A vi

— N

o co

nf es

si on

A vi

— N

o co

nf es

si on

0 months

18 months

18 months

0 months

Bea—No confession Bea—Confession

A vi

— C

on fe

ss io


12 months

12 months

A vi

— C

on fe

ss io


Figure 6.10 Prisoners’ Dilemma

In the prisoners’ dilemma, there are two suspects: Suppose the SEC has been investigating an insider trading scheme and their investigation is focused on two securities brokers, Avi Wasserman and Bea Haefner, who are arrested and charged with a crime. The police do not have sufficient evidence to convict them for insider trading unless one of them confesses. The police place the suspects in separate rooms and ask them to confess. If neither confesses, they are convicted of minor violations uncovered by the investigation of their trading activities and are sentenced to 2 months. If both confess, they spend 12 months in jail. However, if one confesses and the other does not, the confessor is released under a plea bargain in return for testifying but the other is sentenced to 18 months in jail—12 for the crime and 6 for obstructing justice. The payoffs in terms of jail time faced by each individual are displayed. Each entry in the table lists the jail sentences for Avi and Bea, respectively. The Nash equilibrium is for both suspects to confess—the shaded cell. Given the payoffs, it is always in the individual interests of each suspect to confess (taking the action of the other party as given).

216 Part 2 Managerial Economics

displayed in Figure 6.11, which displays the payoffs for two firms attempting to form a cartel. It is in their joint interests to restrict output. Yet, as in the prisoners’ dilemma, both firms have individual incentives to renege and increase output. The Nash equi- librium is for each to increase output.

A cartel can persist if it can impose sufficient penalties on cheaters. But for these penalties to be effective, cartel members must be able to observe (or reliably infer) that a firm has cheated. To the extent that cartel members expect to interact on a repeated basis, there are greater incentives to cooperate. Repeated interaction also increases incentives to invest in developing effective enforcement mechanisms to limit cheating. Potentially, these incentives can be strong enough to resolve the car- tel’s dilemma. In general, cooperation is easier to enforce if the number of firms in the industry is small: It is easier to identify and punish cheaters.

Even when firms are not permitted to form cartels, there may be ways of cooperat- ing to increase profits. For example, over time, a firm might become known as a price leader. Such a firm changes prices in the face of new demand or cost conditions in a way that approximates what a cartel would do. Other firms follow the price changes, thus acting like members of a cartel. Individually, firms still can have short-run in- centives to cheat (e.g., reducing price to get more sales). However, firms might resist this short-run temptation to foster cooperation in the long run and hence obtain higher long-run profits.16

Another potential mechanism to foster cooperation is the structure of contracts with buyers employed by firms in the industry. Most-favored-nation clauses provide buyers with guarantees that the seller will not sell to another buyer at a lower price. These clauses reduce incentives of sellers to lower the price for one buyer because that same price concession would have to be offered to other buyers as well. Meeting-the-




In c—

Lo w

o ut

pu t


In c—

Lo w

o ut

pu t





BeaCo—Low output A

V In

c— H

ig h

ou tp

ut BeaCo—High output

BeaCo—Low output BeaCo—High output


In c—

H ig

h ou

tp ut



Figure 6.11 Cartel’s Dilemma

Two firms, AVInc and BeaCo, attempt to form a cartel. If both firms restrict output, prices are high and each firm’s profit is $500. If both cheat on the cartel and increase output, price will be low and each firm’s profit is $200. If one firm expands output while the other restricts output, the market price will be at an intermediate level; the firm with the high output will make $600 (because of the increased sales), but the other firm will only make $150 (because of the lower price). These payoffs are displayed. The Nash equilibrium is for both firms to increase output—the shaded cell. Given the payoffs, it is always in the interest of each firm to increase output (taking the output of the other firm as given).

16Indeed, economists have shown that within any long-term relationship, with no specified ending date,

cooperation is a possible equilibrium—the parties need not succumb to the cartel’s dilemma. We discuss this

issue in more detail in Chapter 9 and in the appendix to Chapter 10.

Chapter 6 Market Structure 217

competition clauses guarantee that a seller will meet the price of a competitor. Such a clause makes it difficult for firms to cheat on an agreement not to lower price since price concessions are more likely to be brought to each other’s attention by customers.

Our discussion of these strategic interactions among rivals in output markets is meant only to provide a basic introduction to these issues.

Summary A market consists of all firms and individuals who are willing and able to buy or sell a particular product. These parties include those currently engaged in buying and selling the product, as well as potential entrants. Market structure refers to the basic characteristics of the market environment, including (1) the number and size of buyers, sellers, and potential entrants; (2) the degree of product differentiation; (3) the amount and cost of information about product price and quality; and (4) the conditions for entry and exit.

Competitive markets are characterized by four basic conditions: A large number of potential buyers and sellers; product homogeneity; rapid, low-cost dissemination of information; and free entry into and exit from the market. In competitive markets, individual buyers and sellers take the market price of the product as given: They have no control over price. Firms thus view their demand curves as horizontal. The firm’s short-run supply curve is that portion of its short-run marginal cost curve above short-run average variable cost. The long-run supply curve is that portion of its long- run marginal cost curve above long-run average cost. In a competitive equilibrium, firms make no economic profits. Production is efficient in that firms produce at their minimum long-run average cost. Firms in competitive industries must move rapidly to take advantage of transitory opportunities. They also must strive for efficient production in order to survive. Some firms in the industry can employ resources that give them a competitive advantage (e.g., an extremely talented manager). Yet in such


Collusion in the Lysine Industry In 2009, Warner Brothers produced a feature film named the “Informant!” that focused on price fixing (based on the

book by K. Eichenwald). Matt Damon portrayed Mark Whitacre, a high-ranking executive at Archer Daniels Midland

Company (ADM). He reported to FBI agents in the 1990s that his employer was engaged in illegal price fixing and

helped them collect evidence by secretly recording company executives negotiating pricing agreements with Japanese

competitors. In 1996, ADM pleaded guilty to price fixing and paid a $100 million fine.

Lysine is an amino acid derived from corn used in swine and poultry feed to promote growth. ADM entered the

lysine market in 1991. Prior to that time, the market had been dominated by two Japanese companies. ADM quickly

gained market share. However, with the competition, the price of lysine fell from about $1.30 per pound to $.60 per

pound. According to Whitacre, ADM executives began discussions in 1992 with their Japanese competitors about how

it would be in their mutual self-interest to collude and fix prices. Collectively, the competitors were forgoing millions

of dollars of profit per month because of the competition among the three companies. Whitacre indicates that a favorite

saying at ADM was (a prominent line in the movie): “The competitor is our friend, and the customer is our enemy.” Whitacre lost his whistle-blower immunity when the FBI discovered that he had committed a multimillion dollar

fraud against ADM. He was sentenced to 10 years in jail and was released from prison in December 2006 after serving

8.5 years. In recent years, he had been employed as the Chief Operating Officer and President of Operations of the

biotech company, Cypress Systems.

Source: Warner Brothers (2009), “The Informant!” (movie).

218 Part 2 Managerial Economics

cases, any excess returns often go to the factor of production responsible for the par- ticular advantage, rather than to the firm’s owners.

Although the competitive model provides a useful description of the interaction between buyers and sellers for many industries, there are others where firms have substantial market power—prices are affected materially by the output decisions of individual firms. Market power can exist when there are substantial barriers to entry into the industry. Expectations about incumbent reactions, incumbent advantages, and exit costs all can serve as entry barriers.

The extreme case of a firm with market power is monopoly, where the industry consists of only one firm. Here, industry and firm demand curves are one and the same. In contrast to competitive markets, consumers pay more than marginal cost and the firm earns economic profits. Output is restricted from competitive levels. With a monopoly, not all the potential gains from trade are exhausted.

Monopolistic competition is a hybrid between competition and monopoly. It is like monopoly in that firms under both market structures face downward-sloping demand curves. Market power comes from differentiated products. Examples include retail shops at malls, gas stations (that differ in location and brand), books, movies, and busi- ness schools. The analyses of output and pricing policies are similar in the two cases. The difference between monopoly and monopolistic competition is that in monopo- listic competition, economic profits invite entry that limits profits.

In oligopolistic markets, only a few firms account for most production. Prod- ucts may or may not be differentiated. Firms can earn substantial profits. How- ever, these profits can be eliminated through competition among existing firms in the industry. To analyze output and pricing decisions in oligopolistic industries, we use the concept of a Nash equilibrium: A Nash equilibrium exists when each firm is doing the best it can given the actions of its rivals. In the Cournot model, each firm treats the output level of its competitor as fixed and then decides how much to produce. In equilibrium, firms make economic profits. However, these profits are not as large as would be made if the firms effectively colluded and posted the monopoly price. Other models of oligopoly yield different equilibria. For instance, one model based on price competition yields the competitive solu- tion in the case of homogeneous products: Price equals marginal cost with no eco- nomic profits. Economic theory makes no clear-cut prediction about the behavior of firms in oligopolistic industries. Economic models must be crafted to fit the specific details of the industry and the nature of the competition. Available evi- dence suggests that in some oligopolistic industries, firms restrict output from competitive levels and hence capture some economic profits.

It is in the economic interests of firms in oligopolistic industries to find ways to co- operate, thereby capturing higher profits. An ostensible motive for antitrust laws is to limit the dead-weight social costs that occur when firms collude on price. Even when firms are free to cooperate, effective cooperation is not always easy to achieve. Indi- vidual firms have incentives to deviate from agreed-on outputs and prices. This in- centive is illustrated by the prisoners’ dilemma. This model highlights incentives that can cause cartels to be unstable. However, firms sometimes can cooperate success- fully when they can impose penalties on noncooperative firms. Cooperation also can be sustained through the incentives provided by long-run, repeated relationships.

While short-run economic profits are possible in all market structures, the overall analysis suggests that most firms will not earn economic profits over the longer run. Firms do not earn long-run economic profits in perfect competition or in the more common structure of monopolistic competition. Absent government fiat, competition,

Chapter 6 Market Structure 219

and technological change makes it hard to maintain monopolies and oligopolistic market power over time. Increases in international trade also can reduce the market power of domestic firms. An important bottom line is that managers should consider potential competition in making decisions in all market structures.

A. Dixit and B. Nalebuff (1991), Thinking Strategically (Norton: New York).

G. Stigler (1987), The Theory of Price (Macmillan: New York), Chapter 3.

R. Pindyck and D. Rubinfeld (1992), Microeconomics (Macmillan: New York), Chapters 8–13.

6–1. The total and marginal cost of producing Product A are

TC � $1,000 � 2Q2

MC � 4Q

The $1,000 is a fixed cost in the short run, but can be avoided in the long run by shutting

down (going out of business). There is only one possible plant size for this operation; thus

SRMC � LRMC � 4Q in this problem. Derive and graph the firm’s short-run and long-run

supply curves (on separate graphs).

6–2. Suppose your firm faces a demand curve of P � 90 � .30Q and the marginal cost of produc-

tion is $10 per unit. Find the profit-maximizing output and price. Display this choice graphi-

cally (showing the demand, marginal revenue, and marginal cost curves). Is this outcome on

the elastic, inelastic, or unitary elastic part of the demand curve? What are your profits?

6–3. Genesee and Natural Light are the two sole competitors in the ultra low-end beer market in

the Rochester metro area. Both firms have marginal costs of 0 and fixed costs of 200. The

industry demand curve is P � 100 � 0.1Q, where Q � Q1 � Q2

a. Assume the firms compete on quantity (Cournot competition). What is the equilibrium

price and total production in the market? How much profit will each firm make?

b. If firms compete on price (rather than quantity), is the answer in Part A a Nash

equilibrium? Explain.

c. Now assume that Genesee acquires Natural Light’s Rochester operations and has a total

regional monopoly on ultra low-end beer. What price will they charge? How much

profit will the combined firm make? Is this more or less than in Part A? Explain briefly

why you found what you did.

d. Now suppose that neither firm has entered the Rochester market and that it costs

$10,000 to build a plant. Suppose that Genesee spends $10,000 to build a plant and

starts producing 400 units. Will Natural Light want to spend $10,000 to enter the mar-

ket? Show why. Assume that there is only one production period to consider.

Solutions to Self-Evaluation Problems 6–1. The firm’s short-run supply curve is the portion of its short-run marginal cost curve that lies

above average variable cost (AVC). AVC is VC�Q � 2Q2�Q � 2Q. SRMC, which is 4Q, is

everywhere above 2Q. Therefore the firm’s short-run supply curve is P � 4Q:


Q 1


SR supply

Self-Evaluation Problems

Suggested Readings

220 Part 2 Managerial Economics

The firm’s long-run supply curve is the portion of its long-run marginal cost curve that lies

above long-run average total cost (LRAC). LRAC in this problem is

LRAC � TC�Q � $1,000�Q � 2Q

Recall that LRAC falls when LRMC is below it and rises when LRMC is above it.

LRAC is minimized when it equals marginal cost. This is the point where positive supply

begins (the firm stays in business), since its LRMC is above LRAC at every quantity

beyond this point. The point can be found by setting LRAC � LRMC and solving for

Q and LRMC:

$1,000�Q � 2Q � 4Q

$500 � Q2

Q � 22.36 units

At this quantity LRMC � LRAC � $89.44. At this price the firm breaks even

(TR � TC). At lower prices the firm loses money and in the long run it is optimal to shut

down. Thus the firm’s long-run supply is 0 up to a price of $89.44 and is P � 4Q at

higher prices:

6–2. Profit maximization occurs where MR � MC. The MR function has the same intercept as

the linear demand curve but twice the negative slope. Therefore the profit-maximizing con-

dition is

90 � .60Q � 10.

Solving this equation produces the profit-maximizing quantity, Q* � 133.33 units. The op-

timal price (from the demand curve) is $50. The firm makes a profit of $40 per unit for a

total profit of $5,333.20. The outcome is on the elastic portion of the demand curve (above

the midpoint on the linear curve). It is never optimal for a firm to produce on the inelastic

portion of its demand curve since revenue is increased and costs reduced by lowering out-

put. The graph for this problem is


133.33 Q









22.36 Q

LR supply


Chapter 6 Market Structure 221

6–3. a. In Cournot competition, each firm maximizes its own profits taking the other firm’s out-

put as given. In the equilibrium, both of the firm’s expectations about the other firm’s

output are realized. The profit-maximizing conditions for the two firms are

MR1 � (100 � .1Q2 *) � .2Q1 � 0 (optimal condition firm 1)

MR2 � (100 � .1Q1 *) � .2Q2 � 0 (optimal condition firm 2)

The equilibrium occurs at quantities that jointly satisfy these two equations. In this prob-

lem, both firms are identical. Thus in equilibrium Q1 * � Q2

*. Substitute this relation into

either of the two equations to produce a single equation with one unknown variable.

Using the first equation:

(100 � .1Q1 *) � .2Q1

* � 0

Q1 * � 333.33 units � Q2


Total production (Q1 * � Q2

*) � 666.66 units. The equilibrium price from the industry

demand curve is $33.34. Each firm makes profits of ($33.34 � 333.33) � $200 � $10,913.22.

b. No, the solution in Part A is not a Nash equilibrium if the firms compete on price. In a

Nash equilibrium, each firm is choosing optimally given the choice of the other firm. If

one firm chooses a price of $33.34, the other firm would do better by undercutting the

price and capturing the entire market. For example, if firm 1 cuts its price by $1 to

$32.34 it will sell 676.6 units (assuming the other firm does not match the price de-

crease) and make a profit of $21,674.48 (greater than the $10,913.22 in Part A). The

Nash equilibrium with price competition and homogenous products is for both firms to

price at MC ($0 in this problem).

c. After the merger, Genesee is a monopolist. Its demand curve is the industry demand

curve. It will maximize profits by setting its MR, 100 � 0.2Q � 0. Q* � 500 units and

P* from the demand curve is $50. The firm makes a profit of ($50 � 500) � $200 � $24,800. This profit is higher than the combined profit in Part A ($21,826.44). As a

monopoly, the firm can charge the price (and produce the corresponding quantity) that

maximizes total profits for the industry. Firms make economic profits in the Cournot

equilibrium but their competition (noncollusion) produces an outcome with lower

profits than monopoly.

d. No, Natural Light will not enter. If Genesee is producing 400 units, Natural Light’s de-

mand curve is P � (100 � .1 � 400) � .1Q2. Its optimal production (setting its MR � 0) is 300 units and the resulting price will be $30 (given 700 total units are being pro-

duced). Natural Light’s profits are (300 � $30) � 200 � $8,800, which are less than

the $10,000 entry fee.

6–1. What four basic conditions characterize a competitive market?

6–2. The short-run marginal cost of the Ohio Bag Company is 2Q. Price is $100. The company

operates in a competitive industry. Currently, the company is producing 40 units per period.

What is the optimal short-run output? Calculate the profits that Ohio Bag is losing through

suboptimal output.

6–3. Should a company ever produce an output if the managers know it will lose money over the

period? Explain.

6–4. What are economic profits? Does a firm in a competitive industry earn long-run economic

profits? Explain.

6–5. The Johnson Oil Company has just hired the best manager in the industry. Should the own-

ers of the company anticipate economic profits? Explain.

6–6. A Michigan court ruled in the 1990s that General Motors did not have the right to close a

particular Michigan plant and lay people off. Do you think this ruling benefited the people

of Michigan? Explain.

Review Questions

222 Part 2 Managerial Economics

6–7. The Suji Corporation has a monopoly in a particular chemical market. The industry demand

curve is P � 1,000 � 5Q. Marginal cost is 3Q. What is Suji’s profit-maximizing output and

price? Calculate the corresponding profits.

6–8. Assume the industry demand for a product is P � 1,000 � 20Q. Assume that the marginal

cost of product is $10 per unit.

a. What price and output will occur under pure competition? What price and output will

occur under pure monopoly (assume one price is charged to all customers)?

b. Draw a graph that shows the lost gains from trade that result from having a monopoly.

6–9. In 1981, the United States negotiated an agreement with the Japanese. The agreement called

for Japanese auto firms to limit exports to the United States. The Japanese government was

charged with helping make sure the agreement was met by Japanese firms. Were the Japan-

ese firms necessarily hurt by this limited ability to export? Explain.

6–10. Compare the industry output and price in a Cournot versus a competitive equilibrium. Do

firms earn economic profits in the Cournot model? Does economic theory predict that firms

always earn economic profits in oligopolistic industries? Explain. What does the empirical

evidence indicate?

6–11. What is a Nash equilibrium? Explain why a joint confession is the Nash equilibrium in the

prisoners’ dilemma.

6–12. Candak Corporation produces professional quality digital cameras. The market for profes-

sional digital cameras is monopolistically competitive. Assume that the inverse demand

curve faced by Candak (given its competitors’ prices) can be expressed as P � 5,000 � 0.2Q and Candak’s total costs can be expressed as TC � 20,000,000 � 0.05Q2. Answer the

following questions.

a. What price and quantity will Candak choose?

b. Is this likely to be a long-run equilibrium for Candak Corporation? Why or why not? If

not, what is likely to happen in the market for professional digital cameras, and how

will it affect Candak?

6–13. Will a monopolist ever choose to produce on the inelastic portion of its demand curve?



7 C H A P T E R O U T L I N E

Pricing Objective

Benchmark Case: Single Price per Unit

Profit Maximization

Estimating the Profit- Maximizing Price

Potential for Higher Profits

Homogeneous Consumer Demands

Block Pricing

Two-Part Tariffs

Price Discrimination— Heterogeneous Consumer Demands

Exploiting Information about Individual Demands

Using Information about the Distribution of Demands


Other Concerns

Multiperiod Considerations

Strategic Interaction

Legal Issues

Implementing a Pricing Strategy


I ntuit began as a small software company in 1983 with its new program Quicken—a personal finance program that addressed the common household problem of balancing the family’s checkbook. By 2013, In- tuit had grown to a company with over $4.1 billion in annual revenue

and publicly traded stock on the Nasdaq Stock Market. Its flagship products in 2014 included Quicken, QuickBooks, and TurboTax.

A visit to Intuit’s website in 2014 would have revealed many interesting pricing decisions. Consider the following examples: (1) Intuit offered five Windows versions of Quicken including its Starter, Deluxe, Premier, Home & Business and Rental Property Manager editions. The prices for these prod- ucts ranged from the Starter Edition selling at $39.99 to the Rental Property Manager Edition selling at $164.99. (2) The four higher priced editions could read computer files produced by older versions of Quicken, but the low- priced Starter Edition could not. Thus existing users who wanted to update their software, but use existing data files, had to purchase one of the higher priced alternatives. (3) A Mac version of Quicken was priced at $49.99. (4) Customers who purchased Quicken were offered free shipping and a free mobile app. (5) Online customers could enter a “Special Offer Code” that gave discounts to customers with the code. (6) Intuit provided a

Pricing with Market Power


1. Explain a profit-maximizing firm’s basic pricing objective.

2. Explain how block pricing and two-part tariffs can sometimes be used to increase

profits when facing homogenous consumers.

3. Define price discrimination and explain how it can sometimes be used to increase

profits relative to the benchmark case of a single per-unit price.

4. Develop both cost and valuation-related explanations for real-world examples of

apparently similar goods sold at different prices.

5. Explain “personalized pricing” (first-degree price discrimination) and provide

examples; discuss the social cost implications.

6. Explain “group pricing” (third-degree price discrimination) and provide


7. Explain “menu pricing” (second-degree priced discrimination) and provide


8. Understand how coupons and rebates are sometimes used to price discriminate.

9. Discuss how product bundling might increase profits and provide examples.