Introduction Economic globalization is commonly

Economic globalization is commonly referred to as “the increasing

internationalization of the production, distribution, and marketing of goods and

services” (Harris, 1993, p. 755). Economic globalization in this sense is characterized

by the integration of financial and labor markets via trade, foreign investment and

capital transfers (Shariff, 2003). Increasing exposure to global integration has resulted

in economic dislocations for many countries throughout the world. Much of this

dislocation has come in the form of job losses, poverty, income decrease and

increasing income inequality even among advanced industrial countries (Lawrence,

1996).

Globalization’s negative externalities for the welfare state and questions

regarding what the appropriate policy responses should be have sparked passionate

debate among policy makers and have grabbed the attention of academic researchers.

However, much of the existing academic scholarship is hobbled by theoretical and

empirical limitations that conceal more than they reveal about how global economic

forces are shaping social policies. The motivation of this research is to elucidate the

conditions under which the effects of globalization on social policy are shaped by the

nature of countries’ domestic political institutions and economic structures.

The question that guides this study is: what explains states’ social

expenditures when national economies are increasingly integrated into the global

economy? In answering this question, a dominant theoretical approach in the extant

literature, often referred to as the Efficiency Theory of the welfare state, advances the

proposition that globalization produces a ‘race to the bottom’ effect on social

spending. As global market forces dictate national economic decision-making,

considerations of greater economic efficiency will lead policy makers to sacrifice the

welfare state in order to compete with other states by attracting mobile transnational

capital. As national economies become increasingly integrated into the global

economy, transnational capital will flow to those countries that provide the lowest

levels of social protections for their citizens (Adsera and Boix, 2002, Avelinon et al.,

2005, Cameron, 1978, Garrett, 2001, Garrett, 1998, Hicks and Swank, 1992, Huber,

1999, Iversen and Cusack, 2000b, Katzenstein, 1985, Pierson, 2001, Rodrik, 1998,

Rudra, 2002, Rudra, 2008, Rudra and Haggard, 2001, Swank, 2002).

While the efficiency theory has in some respects become the intellectual

expression of anti-globalization populists on the right and the left (Bhagwati, 2004,

pp.21-25), the theory is limited by its inability to clearly identify the primary causal

mechanisms through which economic globalization produces the ‘race to the bottom’

effect on welfare expenditures.

Notwithstanding the limitations of the efficiency theory, other scholars posit

what they claim to be an alternative theory – the Compensation theory of the welfare

state, which advances the proposition that global economic integration may in fact

produce an expansionary effect on social spending. It is argued that governments will

expand welfare spending to compensate the losers of economic globalization for the

purpose of maintaining their political legitimacy (Miller, 1986, O'Connor, 1971).

However, as it is currently configured, the compensation theory is nothing more than

a statement that global economic integration is correlated with an increase in

governments’ welfare expenditures. As a result, it is not considered a theory in this

analysis but its insights along with those drawn from the efficiency theory are used to

develop an integrated theory of globalization’s effects on the welfare state.

In addition to the literature’s theoretical limitations, the empirical research of

the existing literature has produced evidence that cannot be generalized across a large

sample of countries. Some studies have largely analyzed the relationship between

economic globalization and welfare spending among OECD (Organization for

Economic Cooperation and Development) countries (Cameron, 1978, Garrett, 1998,

Hicks and Swank, 1992, Iversen and Cusack, 2000b), while others have based their

analyses on developing countries (Avelinon, et al., 2005, Kaufman and

SeguraUbiergo, 2001, Rudra, 2002, Rudra, 2008). Studies whose samples rely on

OECD countries suggest that increasing levels of global economic integration increase

government welfare expenditures, consistent with the compensation thesis.

Meanwhile, studies with samples taken from countries in the developing world

suggest that increasing levels of economic globalization significantly reduce

government welfare spending as predicted by efficiency theories (Rudra, 2008).

Given these limitations this study contributes to the existing literature in the

following ways. First, it draws upon efficiency and compensation approaches and

develops an integrated theoretical framework that explains globalization’s effects on

social spending. Second, the study’s theory is systemically tested within the context of

a large-N cross-national pooled time-series analysis of 122 countries during the years

1970-2002. To reinforce the statistical analysis, the study also tests the theory via

comparative case study analyses of South Korea, Chile and Spain. The methodology

that motivated the selection of these countries is discussed in chapter 3. Third, relative

to existing studies, a comprehensive measure of economic globalization is utilized that

adequately captures the theoretical definition of the concept. Fourth, using principal

component analysis, an aggregate indicator that measures the institutional factors of

countries’ domestic politics is constructed. Economic globalization interacts with

these domestic political factors to empirically predict governments’ social spending.

This manuscript is organized in three parts. Part I is comprised of chapters 1,

2 and 3. Chapter 1 reviews the existing literature that features efficiency and

compensation approaches to the welfare state. After discussing the theoretical

limitations of the existing literature an alternative theoretical framework is advanced

that draws upon and integrates efficiency and compensation approaches to explain

social policies under conditions of global economic integration. Chapter 2 presents a

full discussion of the various components of the study’s theoretical argument and

affixes hypotheses at the end of each discussion. Chapter 3 presents the study’s

research design. The outcome variable – welfare spending – and the various

explanatory variables are discussed and operationalized. The chapter concludes with a

discussion of the estimation procedures, which feature various time-series regressions

that are used to analyze the cross-national data as well as a discussion of the

methodology that informed the selection of countries used in the case study analysis.

Part II is comprised of chapter 4, 5 and 6. Chapter 4 presents the empirical

finding of the interactive effect of economic globalization and national capitalist firms

on states’ social spending. Chapter 5 presents the empirical findings of the interactive

effects of economic globalization and domestic political institutions on states’ welfare

spending. A number of interactions are examined. The first analyzes the interaction

between economic globalization and an index of domestic political institutions on

welfare spending. And the analyses that follow examine the interaction between

economic globalization and the disaggregated components of political institutions on

welfare spending; namely, the interaction between economic globalization and regime

type; the interaction between economic globalization and voter’s participation; and the

interaction between economic globalization and electoral competition. Chapter 6

shifts the analysis to a discussion of the empirical findings that feature the interactive

effects of economic globalization and domestic political affiliation on states’ welfare

spending. These include the interaction between economic globalization and labor

unions and the interaction between economic globalization and the ideology of ruling

political parties.

Part III is comprised of Chapters 7, 8, and 9, which presents the case studies

of South Korea, Chile and Spain. The case studies illustrate that the effects of global

economic integration on welfare policy in emerging economies with authoritarian

political histories are conditioned by the variation in the political institutions found in

each country. Chapter 10 concludes with a discussion of the theoretical and policy

implications that emerge from the research.

1.2. The Efficiency Theory of Welfare Spending

The fundamental proposition of the efficiency theory is that high levels of

government social spending undermine economic efficiency and the competitiveness

of domestic firms in international markets (Avelinon, et al., 2005, Garrett, 2001,

Kaufman and Segura-Ubiergo, 2001). It is argued that since social spending is largely

funded from corporate taxes, any increase in social expenditures will be accompanied

by an equivalent increase in the level of taxes (Song and Hong, 2005). Increased taxes

undermine investor confidence and the competitiveness of domestic companies in

both domestic and international markets (Garrett, 2001). Increased social spending

can also result in increased government debt as the state increases its borrowing to

finance its welfare policies. Consequently, increased government borrowing results in

higher interest rates and the devaluation of the currency, both of which increase

production costs and discourage companies from making new investments (Garrett,

2001).

High levels of taxes brought about by increases in government welfare

policies will ultimately facilitate capital flight, as transnational corporations will begin

re-locating their investments to countries that have lower taxes and limited social

protections, hence producing a ‘race to the bottom’ effect on the welfare state (Barnet

and Cavanagh, 1994, Barnet and Muller, 1974, Brecher and Costello, 1994). Since

economic globalization increases the mobility of transnational capital, it is this threat

that forces governments to significantly reduce social expenditures in order to restore

investor confidence. In sum, the efficiency theoretical model posits that economic

globalization and the level of international competition that emerges from it constrain

and limit government welfare spending in order to attract and retain mobile capital.

Recent empirical research seems to confirm the logic of the efficiency theory

of welfare spending. One study assessed the impact of economic globalization on the

growth of government spending in OECD countries and showed that trade and

international financial openness had a negative effect on government spending

(Garrett, 2001). Consistent with this finding, recent research using a sample of Latin

American countries examined the relationship between economic globalization and

welfare spending and found that trade openness had a consistently negative effect on

aggregate social spending and social security transfers (Kaufman and Segura-Ubiergo,

2001).

Razin and Sadka (2005) explain the decline of the welfare state in terms of

the changing demographic patterns and the global integration of national economies.

Under conditions of global economic integration and the growth of an aging

population, governments are caught between a rock and a hard place. On one hand,

increasing taxes on capital runs the risk of driving away mobile transnational capital.

On the other hand, increasing taxes on a young but increasingly small labor force is

both economically and politically unsustainable. Since young people represent an

important element of the median voter, they are likely to effectively resist the

government’s attempt to increase welfare spending. Given this dilemma, government

welfare spending is likely to decline (Razin and Sadka, 2005). To the extent that

economic globalization exerts a downward ‘race to the bottom’ effect on social

spending, the above discussion serves as this study’s theoretical baseline and

generates the following hypotheses:

HE: Increasing levels of global economic integration are associated with a decrease in

welfare spending.

1.3. The Compensation Approach

While recognizing the budgetary constraints of the state under conditions of

increased global economic integration, compensation approaches to welfare spending

emphasize the social demands for welfare allocation and the political incentives of

policy makers to respond to such demands. The welfare system, according to this

approach, is a necessary mechanism for offsetting the costs of global economic

integration (Cameron, 1978, Kaufman and Segura-Ubiergo, 2001, Quinn, 1997).

Scholars in this tradition argue that efficiency theories overlook the political incentive

to increase public programs in response to international economic integration (Garrett,

2001). Since policy makers in democracies are primarily motivated by re-election,

they are more likely to increase welfare spending to offset negative economic

externalities, such as job losses and increased income inequality that emerge from the

competitive nature of the global economy. Hence, knowing that those who are

displaced will blame political incumbents for the negative externalities of economic

globalization, policy makers are more likely to increase welfare spending to pacify

displaced workers. In addition, policy makers will also provide welfare benefits to

insure that the negative externalities of global economic integration do not disrupt

national financial markets (Avelinon, et al., 2005).

David Cameron’s (1978) seminal research provides the first empirical and

historical analysis of the growth of the welfare state among Northern European

countries. The research was the first quantitative analysis of welfare policy that

showed that openness to trade was strongly correlated with what he referred to as the

“scope of the public economy,” which was measured in terms of the change in total

taxes as a percentage of GDP. The research showed that openness to trade was the best

predictor of the growth of government revenues. Large nations that were

economically less open experienced moderate increases in the scope of the public

economy compared to smaller nations with more open economies. While the scope of

the public economy among small Western European countries varied with the

dominance of left parties in Scandinavian countries or the dominance of centrist or

conservative parties in countries like Belgium and Ireland, the best explanation for the

expansion of government expenditures is the degree to which national economies had

been integrated into the global economy (Cameron, 1978).

In his classic, Small States in World Markets, Peter Katzenstein’s (1985)

analysis is consistent with the compensation approach to welfare policy. By

employing a comparative case study analysis of Sweden, Norway, Denmark, the

Netherlands and Belgium, Katzenstein demonstrates that global economic integration

is causally related to welfare expenditures as well as to the various state interventions

that are designed to increase economic growth and productivity. According to

Katzenstein, what distinguishes the small states of Western Europe from larger nations

is the ways in which they have combined liberal policies that are designed to leverage

greater global economic integration with a policy of domestic compensation through

which the country’s national economy is protected from the negative consequences of

liberal openness (Katzenstein, 1985).

In pursuing an effective industrial policy, the mix of international liberalism

and domestic compensation varies widely among small Western European states.

Moreover, the development of an industrial policy is not dependent on size but what

Katzenstein refers to as democratic corporatism, which is the way in which conflicting

economic interests are mediated domestically. Democratic corporatism is

characterized by an “ideology of social partnership expressed at the national level; a

relatively centralized and concentrated system of interest groups; and voluntary and

informal co-ordination of conflicting objectives through continuous political

bargaining between interest groups, state bureaucracies, and political parties”

(Katzenstein, 1985, 32). It is, therefore, the democratic corporatist nature of small

European states that makes it possible to develop an industrial policy that is based on

effectively integrating national economies into the global economy, while at the same

time developing a robust system of domestic compensation (Katzenstein, 1985).

In building on the work of Cameron (1978) and Katzenstein (1985), Rodrik

(1998) developed a cross-national study of the relationship between economic

globalization and the size of government. This study was motivated by a simple

question: is the relationship between trade and government spending negative as

efficiency theory predicts or is the relationship positive as predicted by the

compensation approach? The research shows a positive correlation between countries’

exposure to international trade and the size of government. These results are robust to

most measures of government spending and the inclusion of a wide range of control

indicators as well as various sample selections. According to Rodrik, government

spending reduces societal-risk for countries whose economies are increasingly

vulnerable to global economic integration, and the relationship between trade

openness and the size of government is strongest when the terms-of-trade risk is the

highest (Rodrik, 1998).

Other scholars within this research tradition consider the effects of other

aspects of economic globalization on government welfare spending. Quinn’s (1997)

cross-national study of 38 nations estimated the effects of capital mobility on

government spending and found that greater capital mobility is associated with higher

levels of spending. Other research on the effect of capital mobility on welfare

spending has shown that the integration of capital markets has been associated with

increases in welfare spending as well as higher corporate taxes (Swank, 1998). A

recent empirical treatment of Latin American countries provided additional support

for the compensation thesis. Using a measure of financial openness as well as

measures of trade openness, Avelinon, Brown and Hunter’s research suggests that

trade openness has a positive relationship with education and social expenditures, and

financial openness does not reduce government expenditures for social programs as

predicted by the efficiency theory (Avelinon, et al., 2005).

1.4. Theoretical Limitations of the Existing Literature

A significant limitation of efficiency theory’s ‘race to the bottom’ approach to

social policy is that the central mechanism through which economic globalization is

said to have a reductive effect on welfare spending is not clearly specified. Different

authors in this tradition identify different causal mechanisms. Some identify global

corporations as the central mechanism through which economic globalization

produces a ‘race to the bottom’ effect on welfare spending. These scholars argue that

since economic globalization increases the mobility of transnational capital, it is the

threat of corporate divestment and re-location to other countries that forces

governments to slash welfare expenditures in order to reduce costs and restore

investor confidence (Barnet and Cavanagh, 1994, Barnet and Muller, 1974). Others

point to states as the central mechanism and argue that governments regardless of their

ideological orientation are increasingly willing to sacrifice the interests and rights of

workers and the poor in order to promote an investor friendly environment (Holman,

1993). And still others point to the structural adjustment policies (SAPs) of the

International Monetary Fund (IMF) and the World Bank as the central mechanism and

argue that the conditionalities associated with SAPs force governments to retrench

welfare expenditures in order to improve economic efficiency by reducing public

sector and balance of payments deficits (Bartilow, 1997).

The various mechanisms that scholars identify either directly or indirectly

involve the role of corporate capital. As a result, the theoretical discussion in the next

chapter draws upon Marxist theories of the welfare state to construct the conditions

under which the integration of national capitalist firms into the global production

process establishes the causal mechanism through which globalization exerts a

downward pressure on states’ welfare policies.

A significant limitation of the compensation thesis is that it is based on the

assumption that the welfare state is a necessary mechanism for offsetting the negative

externalities of economic globalization (Cameron, 1978, Kaufman and

SeguraUbiergo, 2001, Quinn, 1997). In fact, the Keynesian welfare state is not a

creature of global capitalism but was created to stabilize the economic contradictions

of capitalism in its national form. In many respects the Keynesian welfare state, as

discussed in chapter 2, is naturally incongruent with the logic of global capitalism

(Teeple, 1995). Therefore, if it is observed that the welfare state offsets globalization’s

negative externalities, then this outcome is not natural to the operations of the

Keynesian welfare state (Miliband, 1969); but is a function of the ways in which

endogenous political institutions, which are absent from the compensation

perspective, averts globalization’s natural tendencies to retrench the welfare state.

The compensation thesis, as it is currently configured, is less theory and

more an observation that the global integration of national economies is correlated

with an increase in governments’ welfare expenditures. The conditions under which

this correlation takes place are never specified. As a result, the compensation thesis is

not considered a theory because it fails to specify the necessary endogenous political

factors that offset globalization’s negative externalities.

On this note, empirical studies in the existing literature have largely overlooked

the importance of how economic globalization’s effect on states’ welfare spending is

conditional on the nature of domestic political institutions. It is only in the past few

years where a handful of scholars have attempted to address this deficit. In their

research, Boix (1998) and Garrette (1998) demonstrate that the impact of global

economic integration on governments’ welfare expenditures is conditional on the

nature of partisan politics. Domestic political variables also feature prominently in the

research of Asera and Boix (2002). They argue that the relationship between the

openness of national economies and the size of the public sector’s welfare spending is

heavily conditional on the nature of the political regime. They contend that

governments strategically provide welfare compensation to build domestic political

coalitions that support free trade, and democratic governments relative to authoritarian

regimes are more likely to use welfare spending to compensate the losers of economic

globalization.

While recent studies have attempted to bring greater theoretical precision by

identifying the mechanisms through which globalization operates in determining

social policies, the mechanisms that are tested in such studies are limited to partisan

politics and political regimes. These studies do not provide a comprehensive analysis

of how the interactions of economic globalization and other domestic political

variables affect states’ welfare spending.

This study argues that government welfare spending is a function of the ways in

which the pressures of economic globalization is conditioned by domestic politics.

Domestic politics consists of political affiliation and political institutional factors that

refer respectively to the willingness and capacity of political systems to initiate

changes in public policy (Glatzer and Rueshemeyer, 2005). Political institutional

factors, which include the characteristics of political regimes and the levels of

electoral competition and political participation, determine the political environment

that shapes the incentives and preferences of government officials who make welfare

policy. Political affiliation factors, which include organized labor and political parties,

determine how government resources - specifically welfare expenditures - are

distributed. This study, therefore, builds on the existing literature by examining the

ways in which economic globalization’s effect on welfare policy is conditional on the

domestic political environment that shapes welfare policy and the political affiliations

of domestic political actors who distribute social benefits.

1.5. Towards an Integrated Theory of the Welfare State

The theoretical limitations of the extant literature present an opportunity to

construct a robust theoretical framework that integrates efficiency and compensation

approaches to the welfare state. While the existing literature treats these approaches as

competing or mutually exclusive ‘theories’ of the welfare state, they are considered

here to be mutually inclusive processes in the development of social policy.

Government welfare policy emerges from the tension of globalization’s proclivity to

retrench social spending and the proclivity of domestic political actors and institutions

to compensate. In essence, the construction of social policy, under conditions of global

economic integration, is a function of the dialectical pressures for greater economic

efficiency and domestic political preferences for greater compensation.

It is postulated that in a world absent of domestic political institutions and

where transnational corporations completely dominate countries’ political economy,

by default, economic globalization will exert a downward pressure on social spending.

However, in the presence of domestic politics, globalization’s natural proclivity for

welfare retrenchment will be averted since its effect on social spending is conditional

on the nature of political institutions and the political preferences of labor unions and

political parties that set a floor against further retrenchment. The dialectical tension

between globalization’s tendency to retrench the welfare state and the tendency of

domestic political institutions to resist retrenchment is fully developed and empirically

tested in the chapters that follow.

Chapter 2

Globalization’s Effect on the Welfare State: Economic Efficiency by

Default but Compensation by Design

In the previous chapter it was argued that, by default, economic globalization

exerts a downward pressure on welfare expenditures through the operations of

transnational corporations. However, since globalization’s effect on social policy is

conditional on the nature of endogenous political forces, its proclivity to retrench

welfare expenditures is averted by the preferences of domestic institutions and

political actors to compensate. In developing this integrated theoretical explanation of

welfare policy, this chapter asks the following questions: under what structural

conditions of corporate capitalism will economic globalization produce a ‘race to the

bottom’ effect on states’ social spending? And under what domestic political

conditions will institutions and political actors avert globalization’s ‘race to the

bottom’ effect on social spending? In answering these questions, the discussion that

follows draws upon Marxist and political democratic theories of the welfare state.

Transnational Corporations and the Race to the Bottom

2.1. Marxist theory of the Welfare State

Marxist scholars have consistently argued that the dynamics of the modern

welfare state cannot be understood apart from the historical development of

capitalism. For Marxists, the productive relations of national capitalism depended on

the welfare state. State intervention was endemic to the birth of capitalism, guided its

early development and has been crucial to the history of capital accumulation, even in

the U.S., a country that prides itself as exceptionally and fiercely committed to rugged

individualism and laissez faire (Manley, 2008, Miliband, 1969, 9). The intervention of

the welfare state was promoted by corporate capital when labor markets were

constrained by national boundaries and capital was largely immobile due to the fact

that national economies were relatively closed before the 1970s. Corporate capital’s

promotion of state intervention emerged from the desire to socialize the costs of the

business cycle whose economic booms and busts created uncertainty, social disruption

and political instability. In addition, corporate capital’s promotion of the welfare state

represented an attempt to diminish the growing interest in socialism that emerged as a

result of the 1930s’ depression. The desire to socialize the costs of capital

accumulation has historically led countries to introduce old age pensions, hospital

insurance and public education. While industrial accident insurance schemes were

partly won by labor unions, they largely emerged from the efforts of corporations to

create a system that would limit corporate liability for industrial accidents as well as

socialize the costs via industry-wide insurance premiums (Teeple, 1995, 13-14).

Marxist scholars argue that the intervention of the welfare state was necessary to

the very survival of national capitalism. State intervention helped to mitigate class

conflict and managed the internal contradictions of capital accumulation, which given

the business cycle produces massive unemployment and economic dislocations for

which capitalism in itself has no mechanism to accommodate. In the attempt to

rationalize capitalism the welfare state subsidizes the costs of capital accumulation by

reproducing the working class, by intervening into labor markets to offset the

dominant leverage that capital has over labor, and by intervening into the production

process. The state’s provision of health care, education, subsidized childcare, child

and family allowances, and food stamps has always been associated with attempts to

propagate the working class and prepare them for the national labor market. The

regulation of minimum wage and the enactment of child labor laws, education and job

training, pensions and unemployment assistance have always been associated with

attempts to bring equity in national labor markets. In addition, for Marxists, the

welfare state’s provision of collective bargaining is nothing more than an institutional

framework that manages class conflict between workers and the owners of capital

(Milward, 2003, 106-110, Teeple, 1995, 15). The scale and scope of the welfare state’s

intervention, for Marxists scholars, underscores the fact that capitalism in its national

form:

“Depends to an ever-greater extent on the bounties and direct support of the state, and

can only preserve its ‘private’ character on the basis of such public help. State

intervention in economic life in fact largely means intervention for the purpose of

helping capitalist enterprises. In no field has the notion of the ‘welfare state’ had a

more precise and apposite meaning than here: there are no more persistent and

successful applications for public assistance than the proud giants of the private

enterprise system” (Miliband, 1969, 78).

While corporate capital in its national form depended on the welfare state,

Marxists scholars argue that under conditions of global economic integration the

welfare state is increasingly at variance with the logic of capital accumulation.

Essentially, the conditions that gave rise to the welfare state have been eroded by the

integration of global markets. As capital became increasingly internationalized and

was no longer limited to the national labor market, it no longer required the welfare

state’s intervention to facilitate political compromise with the national working class.

The growth of the global labor market undermined national labor markets and thereby

undermined the state’s raison d’être to provide welfare benefits and collective

bargaining for the working class. Furthermore, since labor unions are unable to

accompany capital into the global labor market, they have now become anachronistic

relics of an earlier era of capitalism. National jurisdictions have now become less

important in corporate decision-making since transnational corporations can now

secure greater tax concessions from states who increasingly compete for corporate

investment in a never-ending race to the bottom where state revenues dwindle and the

priorities of the welfare state are abandoned (Milward, 2003, 112-115, Teeple, 1995,

69-74). To the extent that Marxist scholars claim that globalization’s ‘race to the

bottom’ effect on social spending is conditional on changes in the structure of

corporate capitalism, Marxist theory, in this respect, is a variant of efficiency theories

and as such the above discussion generates the following hypothesis:

HE1: Increasing levels of global economic integration are associated with a decrease in

welfare spending when the structure of corporate capital is transnational.

Institutions and the Politics of Compensation

2.2. Democratic Regimes and Social Generosity

The authority characteristics of political regimes simultaneously influence

both the pace at which national economies are integrated into the global economy and

the scale of government welfare spending. Relative to authoritarian regimes,

democratic governments who face public pressure have a strong incentive to

compensate economic dislocations that arise from global integration (Garrett, 2001).

Several scholars have examined how political regimes affect social spending and

argue that political regimes play a crucial role when governments decide social

welfare policies under the conditions of increasing economic globalization (Adsera

and Boix, 2002, Avelinon, et al., 2005, Hicks and Swank, 1992, Kaufman and

SeguraUbiergo, 2001). Since policy makers in democracies are subject to pressures

from elections and interest groups, they are more likely to allocate a larger portion of

their budgets for social welfare spending than those in authoritarian regimes. Research

on Latin America demonstrates that in the face of trade expansion, democratic

governments are more likely to provide social welfare programs than non-democratic

regimes (Avelinon, et al., 2005, Kaufman and Segura-Ubiergo, 2001).

Other researchers, however, caution that democracies do not affect all types of

social spending equally. Segura-Ubiergo (2007) argues that lower income groups in

Latin America are likely to pressure governments to increase social spending only to

the extent that they are the direct beneficiaries of such spending. Results from his

research have shown that democracies in Latin America tend to be negatively

associated with social security expenditures but positively associated with health and

education expenditures (Segura-Ubiergo, 2007, 169). These results reflect the fact that

social security beneficiaries in Latin America must be legally employed in the formal

sector, and since lower incomes groups who are largely unemployed have no access to

these benefits, they have no incentive to press their governments to receive them.

Health and education expenditures reach a much larger segment of the population and

lower income groups are more likely to press government to increase such

expenditures (Segura-Ubiergo, 2007). To the extent that economic globalization’s

effect on welfare spending is conditional on the authority characteristics of political

regimes, the above discussion generates the following hypotheses:

H2: Increasing levels of global economic integration are associated with an increase in

welfare spending when political regimes are democratic.

2.3. Competitive Elections and Social Spending

Political democratic theories emphasize the effect that political competition

among political parties has on government welfare policies (Hicks and Swank, 1992,

Kite, 2004). Given the clientelistic nature of competitive electoral politics in many

countries throughout the world, political parties are more likely to propose generous

welfare allotments such as pensions, unemployment insurance, job training, health

care, and social security in order to secure votes. As the global economic integration

of national economies increases, the clientelistic nature of competitive electoral

politics is also expected to increase since parties increasingly seek to provide welfare

benefits for constituent voting districts adversely affected by economic globalization

(Cammack et al., 1988). This discussion generates the following hypothesis:

H3: Increasing levels of global economic integration are associated with an increase in

welfare spending when the level of electoral competition is high.

2.4. Political Participation and Welfare Expansion

Political democratic theories argue that the level of political participation also

affects government welfare expenditures (Hicks and Swank, 1992, Kite, 2004). High

voter turnout is reflective of the political entrance of first time and working class

voters who will most likely to terminate the political careers of incumbents that they

hold responsible for the negative externalities of global economic integration. And

since re-election matters to incumbents, they are more likely to promise increasing

welfare expenditures to appease the wrath of the voters. To the extent that economic

globalization’s effect on welfare spending is conditional on the level of political

participation, the above discussion generates the following hypothesis:

H4: Increasing levels of global economic integration are associated with an increase in

welfare spending when the level of political participation is high.

2.5. Democratization and Welfare Expansion

The above discussions considered the interactive effect of economic

globalization and the institutional factors of domestic politics on states’ welfare

spending. However, because the domestic political system is more than the sum of its

parts, it is also important to consider the interaction between economic globalization

and the aggregate effect of political institutions (which is an index that is comprised of

indicators that measure regime type, the level of electoral competition and political

participation) on states’ welfare spending. When regimes are democratic and the level

of electoral competition and political participation is high, then democratization

defines the nature of political institutions (Vanhanen, 1984, Vanhanen, 1990,

Vanhanen, 1997).

To fully account for the cross-national variation in states’ welfare spending it is

also necessary to estimate the interactive effect of global economic integration and the

aggregate and disaggregate political environmental factors that shape states’ welfare

spending. In estimating the aggregate effect of institutions an index is constructed via

principal components analysis. This discussion generates the following hypotheses:

H5: Increasing levels of global economic integration are associated with an increase in

welfare spending when democratization of political institutions is high.

Political Affiliations and the Politics of Compensation

2.6. Labor Unions and Welfare Expansion

Social democratic corporatist theories emphasize the ways in which organized

labor can directly affect government welfare policies. Although governments in free

market economies tend to be more responsive to the preferences of the business

sector, they will also respond to pressure from labor unions that demand the provision

of social protection benefits to union members who have been adversely affected by

global economic integration. With the threat of strikes, which undermine investor

confidence and economic growth, governments are more likely to provide social

welfare programs such as pensions, unemployment insurance, job training, health

care, and social security under the organized collective power of labor unions

(EspingAnderson, 1990, Garrett, 1998, Hicks, 1999, Kite, 2004, Rodrik, 1998).

Organized labor can also exert indirect pressure on governments to increase

welfare spending. In recent years, the labor movement has effectively lobbied the

World Trade Organization (WTO) to link labor standards to free trade agreements.

The Social Clause, as it is commonly called, would require trading countries to

observe a series of labor standards, which guarantees minimum wages, the right of

labor to engage in collective bargaining, and the prevention of child labor. In addition,

global institutions like the WTO would be given the authority to impose sanctions

against countries that fail to observe these standards (Hughes and Wilkinson, 1998,

Wachtel, 1998, Wet, 1995, Wilkinson and Hughes, 2000).

International agreement on the Social Clause would potentially force

governments to implement labor standards, which would increase the negotiation

power of unions to elicit further concessions from governments to increase welfare

spending. Therefore, to the extent that economic globalization’s effect on welfare

spending is conditional on the strength of organized labor, the above discussion

generates the following hypothesis:

H6: Increasing levels of global economic integration are associated with an increase in

welfare spending when labor unions are strong.

Other scholars argue that globalization has weakened organized labor and

therefore challenge the notion of labor unions’ influence in shaping states’ welfare

policies. Charles Lindblom’s Politics and Markets is a classic treatise on this issue.

Lindblom argues that corporate capital, relative to labor unions, occupies a privileged

position in government. Since economic recession and high inflation can topple

governments, the legitimacy of governments largely depends on private corporations’

providing jobs, generating economic growth and increasing living standards.

Consequently, public policy is always geared towards supporting corporate priorities

and not those of organized labor (Lindblom, 1977, 172-173). Increasing levels of

global economic integration only serve to deepen organized labor’s inferior position in

government and significantly weaken its political influence to extract welfare

allocations from the state.

Case study research about the politics of the labor movement in emerging

economies has corroborated Lindblom’s basic thesis. Since the 1950s the labor

movements in many of these countries have been transformed from being militant

organizations that pressed for the rights of working people, to being increasingly

coopted into the free market priorities of transnational capitalism. Given the fact that

unions in emerging economies increasingly see themselves as a responsible partner to

government and capital they are less likely to press governments to increase welfare

allocations (Gray, 2008, Gray, 2007).

Recent empirical research also shows that with increasing global economic

integration labor unions in the less developed countries (LDCs) have been unable to

prevent the decline of the welfare state. Collective action problems of labor unions in

countries with large pools of low-skilled workers tend to weaken the political

bargaining power of labor unions vis-à-vis the state and transnational corporations.

Therefore, global economic integration is expected to decrease government welfare

spending because labor unions in developing countries are politically too weak to

effectively engage the state to provide societal safety nets against the negative

externalities of economic globalization (Rudra, 2002).

Still other scholars are quick to remind us that the phenomenon of weak labor

unions in LDCs is also taking place in advanced industrial economies, albeit for

different reasons. Teeple (1995) argues that global economic integration has

significantly reduced government welfare spending in advanced industrial economies

largely because governments have consistently enacted legislation intended to curb

the political power that labor unions had progressively won since the late nineteenth

century. Since the 1980s, advanced industrial countries have enacted legislation that

has limited the rights and security of labor unions and consequently has reduced their

legal status, and restricted the ways in which labor unions are financed. All these

factors make it increasingly difficult for unions to organize and lobby the state in

defense of welfare allocations (Teeple, 1995). Therefore, to the extent that increasing

global economic integration has weakened the ability of organized labor to extract

welfare expenditures from the state, the discussion generates the following

hypothesis:

H7: Increasing levels of global economic integration are not associated with an

increase in welfare spending regardless of the strength of labor unions.

2.7. Political Parties and Social Policy

Social democratic corporatist theories also focus on the power of the political

left; namely, leftist parties in shaping the welfare policies of the state (Hicks and

Swank, 1992, Katzenstein, 1985, Kite, 2004). The core theoretical proposition of this

perspective is that the political orientation of leftist parties and their supporters affects

the ways in which the governments respond to economic globalization. It is argued

that the effect of global economic integration on states’ welfare spending is

conditional on the nature of party politics. Governments led by left or centrist political

parties (labor, social democratic, or Christian democratic parties) are more likely to

support robust welfare policies than governments led by parties to the political right

(Huber and Stephens, 2001, Stephens, 2005). Kite (2004) argues that in countries

where social democratic parties are strong the public is less tolerant of economic

inequality and holds government accountable for providing welfare benefits. This

discussion generates the following hypotheses:

H8: Increasing levels of global economic integration are associated with an increase in

welfare spending when ruling political parties are on the left.

2.8. The Null Effect on Social Spending

Some scholars argue that the extent of countries’ integration into the global economy

is exaggerated. It is argued that the great expansion of international trade is grossly

exaggerated since world trade, as a percentage of countries’ gross domestic product, is

roughly the same today as it was before World War 1 (Serrano, 2002). International

trade is not a global phenomenon but largely confined to geographical regions, and

the foreign investments of transnational corporations are not dispersed globally but

largely flow to a few countries. Moreover, since developing countries’ participation in

the global economy is limited by civil war and mounting debt, the extent of their

economic integration and its supposed effect on government welfare spending should

be minuscule (Hirst, 1997, Hirst and Thompson, 1996, Jones, 1995, Scarpf, 1991,

Wade, 1996).

Another reason why economic globalization does not affect the welfare state is

that states still maintain considerable control over their economies and frequently use

government expenditures to provide collective goods (social stability and

infrastructure) that are under-supplied by the market but are nevertheless vital to the

process of capital accumulation (Friden and Rogowski, 1996, Garrett, 1999, Kurzer,

1993, Vernon, 1971). This discussion generates the following hypothesis:

H9: Increasing levels of global economic integration has no effect on states’ welfare

spending.

The previous discussions presented an integrated theoretical explanation of

globalization’s effects on social policy that is informed by efficiency and

compensation perspectives. Drawing on Marxist theory, it was shown that the natural

tendency of economic globalization, as it operates through transnational corporations,

is to exert a downward pressure on social spending. However, given the preferences

and interests of domestic institutions and political actors, globalization’s proclivity to

retrench the welfare state is resisted and averted as it operates through the forces of

domestic politics. The study’s integrated theory treats the efficiency and compensation

perspectives as mutually inclusive processes that determine welfare expenditures

under conditions of economic globalization. The subsequent chapter discusses the

study’s design and data estimation procedures used to empirically test the hypotheses

that emerge from the study’s theory.

Chapter 3

Research Design and Methodology

The Empirical Determinants of Welfare Spending

3.1. Measuring the Dependent Variable

The existing literature has defined and measured governments’ welfare

expenditures in terms of countries’ social spending as a percentage of GDP, countries’

per-capita social spending, or total government spending as a percentage of GDP

(Hicks and Swank, 1992, Rudra, 2002). Much of the research on the relationship

between economic globalization and government welfare spending has used total

central government spending or total government revenues as the dependent variable

(Cameron, 1978, Garrett, 2001, Iversen and Cusack, 2000a, Quinn, 1997, Rodrik,

1998). However, measuring the dependent variable in these ways could produce

measurement error. This is because central government spending or total government

revenues contain numerous items that are not related to welfare spending but include

spending for national security, various government subsidies and government

employment spending. Total central government spending or revenues can increase

without expanding welfare spending. Some governments, in response to

globalization’s negative externalities, may choose to provide subsidies to domestic

corporations (Drunberg, 1998). Others may choose to increase spending on national

security in order to put down public riots and protests against global economic

integration (Looney, 1993). Therefore, operationalizing the dependent variable in

terms of total central government spending or revenues incorrectly conflates welfare

spending with other spending priorities of the state (Rudra, 2002).

To correct this type of measurement error, the dependent variable that is used in this

study is measured in terms of governments’ social spending as a percentage of total

government spending (Kaufman and Segura-Ubiergo, 2001, Nooruddin and Simmons,

2009, Rudra, 2002, Rudra, 2008). The dependent variable is calculated from data

adopted from the United Nations Statistics Division (UNSD) of National Account

Official Country Data. Government welfare spending is calculated via the formulae:

Social Protection Expenditure +Education Expenditure + Health Expenditure

Total Government Expenditure

Social protection expenditure includes social security and welfare affairs

services. Social security is composed of income transfers and in-cash benefits for the

elderly, death survivors, sickness and maternity, work injury, unemployment and

family allowances. Welfare affairs and services are defined as assistance delivered to

clients or groups of clients with special needs, such as the young, the old, and the

disabled (Rudra, 2002). This measurement of the dependent variable eliminates other

types of non-governmental welfare spending that is so often conflated into aggregate

measures of government expenditures. Moreover, this measurement reflects where

governments place their allocative priorities within the national economy (Kaufman

and Segura-Ubiergo, 2001).

3.2. Measuring the Primary Explanatory Variables

For the empirical test of the study’s theory, which integrates efficiency and

compensation approaches to social policy, the central explanatory variables are

operationalized as follows. The first, which captures Marxist theory of globalization’s

‘race to the bottom’ effect on social spending, is an interaction between the structure

of corporate capital within countries’ political and economic systems and the extent to

which they are integrated into global economy. The second, which captures the

tendency of domestic political forces to resist globalization’s downward pressures on

social spending, is an interaction between countries’ domestic politics (which includes

disaggregate measures of political institutions and political affiliations) and their level

of integration into the global economy. In addition, principal component analysis is

used to generate an aggregate index - Domestic Political Institutions - that is

comprised of indicators that measure regime type, the level of electoral competition

and political participation. Domestic Political Institutions also interacts with the level

of countries’ integration into the global economy. Further details regarding the

construction of this index appear in the appendix.

As a constituent element of the interactive term - economic globalization is an

index that measures actual financial flows into countries and government reactions to

such flows (Dreher, 2006). Actual financial flows include trade as a percentage of

GDP, foreign direct investment as a percentage of GDP, portfolio investment as of

percentage of GDP, and income payments to foreign nationals as a percentage of GDP.

Government restrictions of these flows are composed of hidden import barriers, the

mean tariff rate, taxes on international trade as a percentage of current revenue, and

existing theories if globalization is expected to alter the amount of resources available for

welfare or the portion of a nation’s productivity that goes towards such policies.

capital account restrictions. Larger numbers in the index indicate higher levels of

integration into the global economy (Dreher, 2006, Dreher et al., 2008).

The Empirical Determinant of the Race to the Bottom

3.3. Marxist Measures of Corporate Power

In Marxist theory, human society is composed of two parts: the substructure and the

superstructure. In capitalist societies the substructure corresponds to the structure of

corporate capitalism that gives rise to society’s superstructure, which is the political

structure of the state with its corresponding laws that define and regulate the

economics of the market. While the structure of corporate capital (the substructure)

exerts a predominant influence over the state and the nature of the market (the

superstructure), the ideas and policies that are produced by the superstructure also

influence the substructure and are reinforced in its operations (Marx and Stone, 1904).

Since Marxist theory suggests that the structure of corporate capital determines

the political superstructure of the state as well as the economic superstructure of

markets (Miliband, 1982, Miliband, 1983, Miliband, 1969), the power of corporate

capital is calculated in terms of:

The Openness of Political Institutions x The Level of Economic Freedom

The Market Capitalization of Domestic Corporation as a % of GDP

The relative openness of political institutions and the level of economic

freedom, which define the regulatory mechanism of the market, capture the political

and economic superstructure. Measures of the openness of countries’ political

institutions and economic freedom are taken, respectively, from the Polity IV index

and the Economic Freedom of the World index (EFW). The Polity IV index measures

the political characteristics of governments, capturing the nature of a country’s

political superstructure. The EFW index measures the degree of economic freedom

that is present in five major areas of a country’s economic superstructure: the size of

government; legal structures and the security of property rights; access to sound

money; freedom to trade internationally; and the regulation of credit markets, labor,

and business. The denominator represents the substructure of corporate capital. It is

captured by the capitalization or the market value (as a percentage of GDP) of

domestically incorporated corporations listed on countries’ stock exchanges.

The more open countries’ political institutions are and the greater the level of

economic freedom, the more likely it is that the interest and influence of the corporate

substructure will be firmly reflected in the political superstructure of the state as well

as in the economic superstructure of the market (Lindblom, 1977, Mills, 1956).

Countries with open political institutions, as discussed in the previous chapter, and by

extension those with greater economic freedoms are more likely to increase welfare

spending. However, when a larger share of a country’s GDP comes under the control

of corporate capital, open political institutions and economic freedom will affect

welfare spending only through its substructure, which is the level of corporate

capitalization of the domestic economy.

Measuring Political Institutions

3.4. Political Regime Type

The data for countries’ political regime type is adopted from the Polity IV

data set, which measures the institutional authority characteristics of governments.

The measure includes the recruitment of the executive, the constraints on executive

authority, and political competition. The Polity IV data is an index that ranges from 10

(autocratic regimes) to 10 (democratic regimes). The indicator is categorized as

follows: consolidated autocracies range from -10 to -6, anocracies, which are regimes

with weakly constituted political authority range from -5 to +5, and consolidated

democracies that range from +6 to +10 (Marshall and Jaggers, 2000).

3.5. Electoral Competition

The variable that measures the level of countries’ electoral competition is

adopted from Vanhanen’s democratization database. Electoral competition is

measured by the portion of votes received by smaller parties in presidential or

parliamentary elections, or both. The variable is calculated by subtracting the

percentage of votes received by the largest party from 100. If the largest party’s

portion is 45%, the electoral competition value is 55 (= 100-45). Calculating the

distribution of seats in parliament is used to generate the variable when vote

percentages are not available. And in cases where the composition of the government

is not based on popular elections, then electoral competition is zero (Vanhanen, 1990).

3.6. Political Participation

The variable that measures countries political participation is adopted from the

International Institute for Democracy and Electoral Assistance. The level of political

participation is measured by the percentage of the voting age population who actually

voted in Parliamentary and Presidential Elections. Theoretically, the value can range

from 0 (no participation) to 100 (full participation).

Measuring Political Affiliations

3.7. The Strength of Labor Unions

Some studies measure the strength of organized labor in terms of the level of

unionization. However, unionization is not comparable across countries. In some

developing countries, like China, unionization rates are high because labor laws make

union membership compulsory. And yet, unions in these countries have little power to

represent the interest of their members (Chan and Senser, 1997). In this sense,

unionization rates in developing countries “exaggerate labor’s independent political

strength” (Rudra, 2002, 425), which weakens its ability to adequately capture the

strength of unions in the developing world (Banuri and Amadeo, 1991, McGuire,

1997, Valenzuela, 1989, 449). Rudra (2002) resolves this problem by creating the

Potential Labor Power (PLP) indicator that measures unions’ strength as the ratio of

skilled labor to unskilled labor divided by the level of surplus workers as a percentage

of the work force. This study adopts the PLP to measure the strength of labor unions

for all the countries in the data set.

3.8. The Ideology of Political Parties

In measuring the ideology of political parties, the primary focus is on the

ideology of the ruling party in government. And as such, the operational definition of

party ideology follows Castes and Mair’s (1984) typology as leftist, centrist, or

rightist parties. If the ruling party’s ideology is left, then party ideology is coded 1. If

the ruling party’s ideology is centre, then party ideology is coded 0. And if the ruling

party ideology is right, then party ideology is coded -1 (Castles and Mair, 1984).

3.9. Confounding Variables

A survey of the literature shows that a set of important variables affects states’

welfare spending. These variables are summarized in Table 1. They include a lagged

endogenous variable to control for serial correlation (Baltagi, 2005), logged GDP per

capita, GDP growth rate, the number of dependents in countries, the level of

urbanization, logged inflation, logged population and decade dummy variables that

are used to account for important international conditions; namely the oil crisis of the

1970s and the economic recession and debt crisis of the 1980s, that could affect

welfare spending rather than economic globalization.

Following Wagner’s law GDP per capita, which proxies countries’ economic

development, is expected to have a positive effect on welfare spending (Adsera and

Boix, 2002, Avelinon, et al., 2005, Cameron, 1978, Garrett, 2001, Kaufman and

Segura-Ubiergo, 2001, Rodrik, 1998, Rudra, 2002). German economist Adolph

Wagner (1835-1917) postulated that the development of an industrial economy is

accompanied by an increased share of government expenditure in countries’ GDP

(Halicio, 2003, Youseif and Abizadeh, 1992). However, the growth rate of countries’

GDP is expected to have a negative effect on welfare spending as economic expansion

reduces the need for social spending (Avelinon, et al., 2005, Cameron, 1978, Garrett,

2001, Rudra, 2002).

The dependency ratio (sum of young and old in the total population) is

expected to have a positive effect on welfare spending. However, since this variable is

featured in only one study and was reported not to have a statistically significant

effect on welfare spending (Garrett, 2001), it is used in this study to test the robustness

of the empirical results.

Urbanization captures the concentration of the working class and poor and is

expected to have a positive effect on welfare spending. In countries with urban

industrial cities the high concentration of poor and the working class people increases

demand for welfare spending (Avelinon, et al., 2005, Garrett, 2001, Rodrik, 1998,

Rudra, 2002). The level of inflation is expected to have a negative effect on welfare

spending. As inflation increases governments are pressured to cut public spending,

especially social welfare. However, the reported direction of the coefficient is not

consistent in existing research. Some researchers report that inflation has a negative

effect on welfare spending (Avelinon, et al., 2005), while others report a positive

effect (Hicks and Swank, 1992). This study also control for population size although a

survey of the literature has shown that the variable is used in only one study and its

reported effect on welfare spending is zero (Garrett, 2001).

Following Kaufman and Segura-Ubiergo (2001), two-decade dummy variables

are incorporated into the analysis to control for international systemic forces that may

impact states’ welfare expenditures that are independent from economic globalization.

One variable captures the oil crisis years from 1970 to 1981. By the late 1960s world

petroleum production peaked and subsequent production entered into a severe decline.

And as the world’s demand for petroleum remained high, the steep shortfall in

production led to sharp increases in oil prices, which led to extreme price inflation. In

1973, petroleum prices experienced further increases as a result of OPEC’s oil

embargo against the West. And in 1979, oil prices again increased as a result of the

Iranian revolution that severely damaged Iran’s oil fields, which further reduced oil

supplies (Spero, 1981, Chapter 8). The oil crisis of the 1970s, which produced high

levels of inflation, exploded government’s energy expenditures and deteriorated

countries’ balance of payments, is expected to have a negative effect on welfare

spending (Kaufman and Segura-Ubiergo, 2001).

The other decade variable captures the global recession and debt crisis years

from 1982 to 1990, which is often referred to as Latin America’s lost decade. As a

consequence of the oil price shocks and inflation of the 1970s, the world economy, by

the early 1980s, went into recession whereby the GDP growth rates of the developed

and developing world plummeted. For most developing countries the recession

created a breaking point as many experienced a liquidity crisis. Petroleum exporting

countries flush with petrodollars invested their money in international banks, which

'recycled' a major portion of the capital as loans to developing countries, especially to

governments in Latin America. As interest rates increased in the U.S. and in Europe in

1979, debt payments also increased making it harder for borrowing countries to pay

back their debts, which ultimately precipitated the debt crisis of the 1980s (Lomax,

1988, Nunnenkamp, 1986). The global recession and the Third World’s debt crisis of

the 1980s are also expected to have a negative effect on welfare spending.

3.10. The Data Estimation Procedures

To empirically test the hypotheses of this study, seven regression models are

employed. Model one features the interactive effect of economic globalization and

corporate capital on governments’ social spending. Model two features the interactive

effect of economic globalization and the index of political institutions on

governments’ social spending. Models three through five feature the interactive effects

of economic globalization and regime type, electoral competition and voter

participation on governments’ social spending. The interactive effects of economic

globalization and labor unions and the ideology of the ruling political party on

governments’ social spending are featured in models six through seven. The pooled

time-series cross-sectional regressions are expressed as:

Model 1: Economic Globalization and Corporate Capital

W = α +β1W_1 + β2G + β3CC + β4G*CC + β5CV + …. +ε

Model 2: Economic Globalization and Political Institutions

W = α +β1W_1 + β2G + β3PI + β4G*PI + β5CV + …. +ε

Model 3: Economic Globalization and Political Regime

W = α +β1W_1 + β2G + β3R + β4G*R + β5CV + …. +ε

Model 4: Economic Globalization and Electoral Competition

W = α + β1W_1 + β2G + β3E + β4G*E + β5CV + .... +ε

Model 5: Economic Globalization and Voter Participation

W = α +β1W_1 + β2G + β3V +β4G*V + β5CV + …. +ε

Model 6: Economic Globalization and Labor Unions

W = α +β1W_1 + β2G + β3L+ β4G*L + β5CV + …. +ε

Model 7: Economic Globalization and Political Parties

W = α +β1W_1 + β2G + β3P + β4G*P + β5CV + …. +ε

Where W, represents governments’ social spending; W_1, represents a one year lag of

governments’ social spending; G, represents economic globalization; CC, represents

Corporate Capital; G*CC, represents the interactive term for economic globalization

and corporate capital; PI, represents the index of Political Institutions; G*PI,

represents the interactive term for economic globalization and the index of Political

Institutions; R, represents political regime type; G*R, represents the interactive term

for economic globalization and political regime type; E, represents the level of

electoral competition; G*E, represents the interactive term for economic globalization

and the level of electoral competition; V represents the level of voter participation;

G*V, represents the interactive term for economic globalization and the level of voter

participation; P, represents the ideology of the ruling political party; G*P, represents

the interactive term for economic globalization and the ideology of the ruling political

party; L, represents the strength of labor unions; G*L, represents the interactive term

for economic globalization and labor unions’ strength; and CV, represents the control

variables that are included in the study.

3.11. Case Study Selection Method

A comparative case study design is also utilized to further test the study’s

integrated theory. There are two advantages of case study research designs over large

N-statistical analyses. First, case study designs allow for detailed examination of the

causal mechanisms through which government social policy is generated by the

interactions between economic globalization and the various domestic political and

economic variables. Although statistical analyses allow for greater theoretical

generalization, they are limited in their ability to describe the ways in which variables

of theoretical interest interact under specific political conditions. The case study

component of this research will uncover the specific causal processes through which

the interaction of domestic political factors with economic globalization produces

social welfare policies. Second, case studies allow us to see how the causal

relationships between the independent and dependent variables change over time.

Although large N-statistical analyses allow us to generate theoretical generalizations

about these relationships, they may not explain social welfare outcomes in specific

countries.

The methodology that informs the selection of cases in this research is based on

the most similar systems design. Cases are selected that have similar values for the

confounding variables, while having different values for the central explanatory

variables (King et al., 1994). Following this method, case studies of South Korea,

Chile and Spain are used to test the hypotheses that welfare spending is conditional on

the ways in which countries’ domestic politics and institutions interact with global

economic integration. All three countries transitioned from authoritarian regimes and

therefore share a similar political history. All three countries are similar in terms of

their per-capita wealth. Both Spain and South Korea are high-income OECD member

countries and Chile is classified as an upper-middle income country. The

agedependency ratios in all three countries are also similar. In 2005, the age-

dependence ratios in South Korea, Spain and Chile were respectively 40, 45, and 49

dependents for every 100 working age persons. 12 In addition, all three countries are

non-oil producers and were similarly affected by the oil price shocks of the 1970s.

However, in these countries there are significant variations in the key explanatory

indicators. In all three countries there has been significant variation in the nature of

their external trade policies and regime type. As Table 3.3 shows, these variations can

be sorted into four categories: statist protectionism under military authoritarianism

(South Korea under Park Chung-hee and Spain under Franco); statist protectionism

under democracy (Chile under the socialist government of Salvador Allende);

economic liberalism under military authoritarianism (Chile under the Pinochet

regime); and economic liberalism under democracy (South Korea under Kim

Daejung, Chile under the Concertación coalition government and Spain under the

socialist government of Manuel Chaves González). In addition, case studies of these

countries allow for greater cross-regional variation, which also provides a strong

robust test of the study’s hypotheses. A full discussion of these cases is provided in

chapters seven, eight and nine.

3.12. Tables

Table 3.1: Control Variables of Welfare Spending

Variable Hypothesized

Theoretical Direction

Reported

Direction

Lagged Endogenous Variable positive

Rudra 2002 positive

Avelino Brown, and Hunter 2005 positive

Kaufman and Segura-

Ubiergo 2001

positive

Logged Per Capita GDP positive

Rudra 2002 positive

Cameron 1978 zero

Avelino Brown, and Hunter 2005 zero

Kaufman and Segura-

Ubiergo 2001

zero

Adsera and Boix 2002 positive

Rodrik 1998 zero

Garrett 2001 zero

GDP Growth Rate negative

Avelino Brown, and Hunter 2005 negative

Cameron 1978 negative

Garrett and Mitchell 2001 negative

Rudra 200 zero

Dependents positive

Garrett and Mitchell 2001 positive

Garrett 2001 zero

Urbanization positive

Rudra 2002 zero

Avelino, Brown, and Hunter 2005 positive

Rodrik 1998 negative

Garrett 2001 zero

Logged Inflation negative

Hicks and Swank 1992 positive

Avelino, Brown, and Hunter 2005 negative

Population Not Determined

Garrett 2001 zero

Dummy Decade 1970-1981

(Oil crisis Years)

negative

Kaufman and Segura-

Ubiergo 2001

negative

Dummy Decade 1982-1990

(Economic Recession-Debt

negative

Crisis Years)

Kaufman and Segura-

Ubiergo 2001

negative

Table 3.2: Corporate Market Capitalization and Welfare Spending

Lagged Welfare Spending 0.937*

(0.026)

POLITY 0.003

(0.005)

Index of Economic Freedom 0.002

(0.009)

Market Capitalization % GDP 11334427.769*

(4277302.936)

POLITY*Market Capitalization 0.009

(0.005)

Index of Economic Freedom* -0.002

Market Capitalization % GDP

(0.006)

POLITY*Index of Economic Freedom -0.000

(0.001)

Corporate Power -378279.352**

(156,283.603)

Economic Globalization -0.000

(0.000)

Urbanization -0.000

(0.000)

Dependency 0.006

(0.020)

Growth 0.001

(0.001)

Ln GDP per capita 0.012**

(0.006)

Ln population -0.001

(0.001)

Ln inflation 0.001

(0.002)

Table 3.2: Corporate Market Capitalization and Welfare Spending (continued)

Growth 0.001

(0.001)

Ln GDP per capita 0.012**

(0.006)

Ln population -0.001

(0.001)

Ln inflation 0.001

(0.002)

Oil shock (70’s) 0.000

(0.000)

Debt crisis (80’s) -0.004

(0.004)

Constant -0.087

(0.061)

Observations 296

R-squared 0.96

___________________________________________________________________

Panel correct standard errors in parentheses ** significant at 1%; *significant at 5%

Table 3.3: Case Study Selection

Economic Statist

Liberalization Protectionism

Military

Authoritarian

Regimes

Democratic

Regimes

Part II Empirical

Finding

Park Chung-hee (Korea)

1961-1979

Chun Doo-hwan

1980-1988

Franco (Spain)

1939-1975

Pinochet (Chile) 1973-

1990

Allende (Chile)

1971-1973

Kim Dae-jung (Korea)

1988-2003

Concertación (Chile)

1990-present

Manuel Chaves González

(Spain)

1982-1996

Chapter 4

The Race to the Bottom: Globalization, Transnational Capital and Social

Policy

4.1. Corporate Capital and Social Spending

Marxist theories of the welfare state argue that capitalism in its national form

depends on a robust welfare state to socialize the cost of capital accumulation and

manage class conflict between the owners of capital and the owners of labor.

However, as capital became increasingly internationalized and was no longer limited

to the national labor market, it no longer required the welfare state’s intervention to

facilitate political compromise with the working class. Under conditions of global

economic integration, transnational corporations can now secure greater tax

concessions from states that increasingly compete for corporate investments in a

never-ending race to the bottom, where state revenues dwindle and the priorities of the

welfare state are abandoned (Manley, 2008, Miliband, 1969, Teeple, 1995). To the

extent that Marxist theory provides the causal mechanism through which economic

globalization exerts a downward pressure on social spending, it is a variant of

efficiency theory and as such generated hypothesis HE1, which stated: Increasing

levels of global economic integration are associated with a decrease in welfare

spending when the structure of corporate capital is transnational.

Table 4 presents estimates of the control variables as well as the economic

globalization variable to assess the stability of the empirical platform on which the

study’s welfare expenditure models are built. Model 4a presents estimates for the

main control variables without the variable for economic globalization. Model 4b

presents estimates for economic globalization along with the controls.

The direction of the coefficients for most of the control variables is consistent with

the theoretical predictions and the reported findings of previous empirical research. In

both models, prior levels of welfare spending (the lagged dependent variable) are

positively associated with current levels of welfare spending. And consistent with

Wagner’s Law, higher levels of economic development is positively associated with

welfare spending. Economic growth, urbanization, population (model 4a only),

inflation and the debt crisis of the 1980s are all negatively associated with welfare

spending. The oil shocks of the 1970s and dependency fails to reach statistical

significance. Economic globalization is positively associated with welfare spending in

model 4b.

[Table 4 about here]

Table 4.1 presents the findings of the effect on welfare spending from the interaction

between corporate capital and economic globalization. Model 4.1a presents the results

with the main explanatory variables - corporate capital and economic globalization –

without the interaction term. Model 4.1b presents the main results with the interaction

term.

[Table 4.1 about here]

In models 4.1a and 4.1b, prior levels of welfare spending (the lagged dependent

variable) are positively associated with current levels of welfare spending. While

higher levels of economic development are positively associated with welfare

spending, all other control variables fail to reach statistical significance. In model

4.1a, corporate capital (the constitutive element of the interaction term) is positively

associated with welfare spending. However, since the coefficient on the interaction

term – Corporate Capital*Globalization - is negative, this positive effect diminishes as

corporate capital integrates into the global economy as predicted by Marxist theory. In

figure 4.1, the sloping line indicates the marginal effect of corporate capital on

welfare spending for meaningful changes in countries’ integration into the global

economy. The 95% confidence intervals around the line show the conditions under

which corporate capital has a statistically significant effect on welfare spending – and

it has a statistically significant effect whenever the upper and lower bounds of the

confidence interval are both above (or below) the zero line. It can be observed from

figure 4.1 that corporate capital’s influence on countries’ political economy has a

significant positive effect on welfare spending when its integration into the global

economy is low. However, this positive effect diminishes as its integration into the

global economy increases. Once global economic integration exceeds a score of 50 on

the KOF index, then corporate capital’s influence on countries’ political economy no

longer has a significant positive effect on welfare spending.

[Figure 4.1 about here]

4.2. Robustness Checks: Alternate Measures of Globalization and Social Spending

Alternative measures of social spending and economic globalization, as featured in

the research of other scholars, are used to provide additional empirical test for

hypothesis HE1. These scholars measure welfare spending as a percentage of countries’

GDP and use trade openness and openness to foreign direct investment (FDI) as

proxies for economic globalization (Kaufman and Segura-Ubiergo, 2001, Rudra,

2002, Segura-Ubiergo, 2007). Table 4.2 presents the findings of the effect on welfare

spending from the interaction between corporate capital and openness to trade and

FDI. Model 4.2a presents the results with the main explanatory variables – trade, FDI

and corporate capital – without the interaction terms. Model 4.2b presents the main

results with the interaction terms.

[Table 4.2 about here]

Models 4.2a and 4.2b show that prior levels of welfare spending (the lagged

dependent variable) are again positively associated with current levels of welfare

spending. Urbanization, economic growth, economic development (only in model

4.2a) and the debt crisis of the 1980s are all positively related to welfare spending.

Population and inflation are negatively associated with welfare spending. In Model

4.2a, while corporate capital fails to reach statistical significance, trade openness is

negatively associated with welfare spending. In Model 4.2b, FDI (the constitutive

element of the interaction term) is positively associated with welfare spending.

However, since the coefficient on the interaction term – FDI*Corporate Capital - is

negative, this positive effect diminishes, a finding that is again consistent with

Marxist theory.

Figures 4.2 and 4.3 examine these issues from an alternate vantage point. In Figure

4.2, the sloping line indicates the marginal effect of FDI on welfare spending for

meaningful changes in corporate capital. And 95% confidence intervals around the

line show the conditions under which FDI has a statistically significant effect on

welfare spending. It can be observed that once corporate capital’s influence within

countries’ political economy is greater than 13%, then openness to FDI no longer has

a significant positive effect on welfare spending.

[Figure 4.2 about here]

In Model 4.2b, openness to trade (the constitutive element of the interaction term) is

negatively associated with welfare spending. However, since the coefficient on the

interaction term – Trade*Corporate Capital - is positive, this reductive effect

diminishes, a finding that is not consistent with Marxist theory. In figures 4.3, the

sloping line indicates the marginal effect of trade openness on welfare spending for

meaningful changes in corporate capital. And 95% confidence intervals around the

line show the conditions under which trade openness has a statistically significant

effect on welfare spending. It can be observed that trade openness has a significant

reductive effect on welfare spending at low levels of corporate capital. However, once

corporate capital’s influence within countries’ political economy is greater than 15%,

then openness to international trade no longer has a significant reductive effect on

welfare spending.

[Figure 4.3 about here]

4.3. Summation

The findings presented in this chapter suggest that economic globalization, as

measured by the comprehensive KOF index, will produce a “race to the bottom effect’

on welfare expenditures as it interacts with corporate capital. When the structure of

corporate capital is national, economic globalization has a positive effect on welfare

spending. However, when corporate capital is integrated into the global economy,

economic globalization has a negative effect on welfare spending, as predicted by

Marxist theory. This finding received additional confirmation when FDI was used as a

proxy for economic globalization. When corporate capital is integrated into the global

economy, openness to FDI has a negative effect on welfare spending, which is also

consistent with the Marxist variant of efficiency theory.

However, when corporate capital is integrated into the global trading system,

openness to trade has a positive effect on welfare spending, a finding that is

inconsistent with the Marxist variant of the ‘race to the bottom’ theory. The fact that

trade flows produce the opposite effect suggests that there are other forces at play in

determining social policy and that global economic integration does not have to run

rough shod over the welfare state. It is possible that policy makers could offset

economic globalization’s ‘race to the boom’ effect by carefully choosing how their

countries are integrated into the global economy as well as carefully negotiating the

terms of their countries integration. Negotiating the terms of economic integration

will invariably introduce domestic political forces into global processes that are driven

by market forces. The empirical findings that are presented in the next chapter suggest

that domestic political institutions and political actors play an important role in

averting globalization’s proclivity to retrench the welfare state.

4.4. Tables and Figures

Table 4: Economic Globalization and Welfare Spending

_______________________________________________________________________

Base Model 4a Base Model 4b

Without with

Economic Globalization Economic Globalization

_______________________________________________________________________

Lagged Welfare Spending 0.002* 0.002*

(0.001) (0.001)

Economic Globalization 0.002**

(0.000)

Urbanization -0.001* -0.001*

(0.000) (0.000)

Dependency 0.078 0.095

(0.051) (0.055)

Growth -0.004** -0.005**

(0.002) (0.002)

Ln GDP per capita 0.095** 0.071**

(0.010) (0.013)

Ln population -0.007** -0.001

(0.002) (0.003)

Ln inflation -0.016** -0.009*

(0.004) (0.004)

Oil shock (70’s) 0.000 0.000

(0.000) (0.000)

Debt crisis (80’s) -0.020** -0.006*

(0.003) (0.004)

______________________________________________________________________________

Constant -0316** -0.294*

(0.108) (0.120)

Observations 451 451

R-Squared 0.42 0.46

______________________________________________________________________________

50

Panel-corrected standard errors are in parentheses. ** significant at 1%; *significant at 5%

Table 4.1: Economic Globalization, Corporate Capital and Welfare Spending

_______________________________________________________________________

Model 4.1a Model 4.1b

Without With

Interaction Interaction

_______________________________________________________________________

Lagged Welfare Spending 0.953** 0.958**

(0.023) (0.023)

Economic Globalization -0.000 0.000

(0.000) (0.000)

Corporate Capital 0.001 0.002*

(0.000) (0.021)

Globalization * Corporate Capital -0.00004*

(0.000)

Urbanization -0.000 -0.000

(0.000) (0.000)

Dependency -0.007 0.003

(0.020) (0.019)

Growth 0.001 0.000

(0.001) (0.001)

Ln GDP per capita 0.009* 0.009*

(0.004) (0.004)

Ln population 0.001 0.000

(0.001) (0.001)

Ln inflation -0.001 -0.000

(0.002) (0.002)

Oil shock (70’s) 0.000 0.000

(0.000) (0.000)

51

Debt crisis (80’s) 0.000 0.001

(0.004) (0.004)

______________________________________________________________________________

Constant -0.069 -0.066

(0.036) (0.035) Observations

296 296

R-Squared 0.96 0.96

______________________________________________________________________________

Panel-corrected standard errors are in parentheses. ** significant at 1%; *significant at 5%

Table 4.2: Trade, FDI, Corporate Capital and Welfare Spending

______________________________________________________________________________________

Model 4.2a Model 4.2b

Without with

Interaction Interaction

______________________________________________________________________________________

Lagged Welfare Spending 0.435** 0.468**

(0.109) (0.098)

Trade -0.0002** -0.0004**

52

(0.000) (0.000)

FDI -0.001 0.003**

(0.001) (0.001)

Corporate Capital 0.000 -0.002**

(0.000) (0.000)

Corporate Capital *Trade 0.00008**

(0.000)

Corporate Capital *FDI -0.002**

(0.000)

Urbanization 0.0003** 0.0003**

(0.000) (0.000)

Dependency 0.024 0.030*

(0.016) (0.015)

Growth 0.001** 0.001*

(0.000) (0.000)

Ln GDP per capita 0.008** 0.005

(0.004) (0.004)

Ln population -0.005** -0.005**

(0.001) (0.001)

Ln inflation -0.003** 0.003**

(0.002) (0.001)

Oil shock (70’s) 0.000 0.000

(0.000) (0.000)

Debt crisis (80’s) 0.006** 0.006**

(0.003) (0.003)

______________________________________________________________________________

Constant -0.001 0.021

(0.035) (0.040)

Observations 338 338

R-Squared 0.60 0.63

______________________________________________________________________________

Panel-corrected standard errors are in parentheses. ** significant at 1%; *significant at 5%

53

54

Chapter 5

The Political Environment of Welfare Compensation: Globalization,

Institutions and Social Policy

The evidence that was presented in the previous chapter lends support to the

argument that economic globalization exerts a downward pressure on welfare

expenditures through the operations of transnational corporations. While the use of

alternate measures of social spending and globalization, specifically openness to FDI,

confirms the predictions of efficiency theory, openness to trade does not. And it was

suggested that endogenous political forces might also be at work negotiating the terms of

countries’ integration into the global economy and in the process avert globalization’s

proclivity to retrench the welfare state. This chapter presents the findings that lend

empirical support to this argument.

5.1. The Social Generosity of Democratic Governments

It was argued in chapter 2 that the authority characteristics of political regimes

simultaneously affect the pace at which national economies are integrated into the global

economy and the scale of social spending. Since policy makers in democracies are

55

subject to pressures from elections and interest groups, they are more likely to allocate a

larger portion of their budgets for social welfare spending than those in authoritarian

regimes. And to the extent that economic globalization’s effect on welfare spending is

conditional on the authority characteristics of political regimes, the argument generated

hypothesis 2, which stated: Increasing levels of global economic integration are

associated with an increase in welfare spending when political regimes are democratic.

Table 5.1 presents the findings of the interactive effect of economic globalization and

political regime type on welfare spending. Model 5.1a presents the results with the main

explanatory variables – economic globalization and political regime type – without the

interaction term. And Model 5.1b presents the main results with the interaction term.

[Table 5.1 about here]

The direction of the coefficients for most of the control variables in models 5.1a and 5.1b

are also consistent with the theoretical predictions and the reported findings of previous

empirical research. Prior levels of welfare spending (the lagged dependent variable) are

positively associated with current levels of welfare spending in both models.

Urbanization is negatively associated with welfare spending in both models. Increasing

levels in the ratio of countries’ young and old dependents are positively associated with

welfare spending in both models. Economic growth is negatively associated with welfare

spending in both models. And consistent with Wagner’s Law, higher levels of economic

development is positively associated with welfare spending in both models. Inflation is

negatively associated with welfare spending in both models. While the oil shocks of the

1970s fail to reach statistical significance, the debt crisis of the 1980s is negatively

associated with welfare spending in model 5.1a but fails to reach statistical significance

in model 5.1b. And while population size fails to reach statistical significance in model

5.1a, it is negatively associated with welfare spending in model 5.1b. Economic

globalization and political regime type are both positively associated with welfare

spending in model 5.1a.

Model 5.1b shows that the interactive effect of economic globalization and political

regime type is positively associated with welfare spending. The sloping line in figure 5.1

indicates the marginal effect of economic globalization on welfare spending for

56

meaningful changes in the authority characteristics of political regimes. And 95%

confidence intervals around the line show the conditions under which economic

globalization has a statistically significant effect on welfare spending. It can be observed

from figure 5.1 that economic globalization has a reductive effect on welfare spending

when political regimes are autocratic. However, this reductive effect declines as the

authority characteristics of regimes become more democratic. When the polity score of

political regimes is greater than -3, then economic globalization no longer has a reductive

effect on welfare spending. Hypothesis 2 is supported by the data.

[Figure 5.1 about here]

While this finding shows that the marginal effect of economic globalization on

welfare spending is greater among democracies relative to autocracies and is therefore

lends support to hypothesis 2, additional observations are in order. It is interesting to note

that autocracies whose economies are integrated into the global economy also provide

welfare benefits to their citizens, although at much lower levels than democracies. Table

5.1a, shows the average welfare spending and the average level of economic globalization

from 1971 to 2004 for several regimes with autocratic polities. These regimes have high

levels of welfare spending and they are also all highly integrated into the global economy.

One common characteristic among most of these regimes is that they are all one party

leftist or former Soviet Republic regimes that have traditionally provided generous

welfare benefits to their citizens. Algeria’s National Liberation Front (FNL), the

revolutionary organization that directed its war of independence against France that ended

in 1962 (Fanon, 1965), ruled the country for four decades and during these years

developed Algeria’s welfare state. Azerbaijan, Kazakhstan and Kyrgyzstan are all former

Soviet Republic states with a Marxist tradition for expansive welfare subsidies, which

have been eroded in recent years.

[Table 5.1A about here]

57

Throughout the 1970s, Bolivia was ruled by a military regime headed by left

leaning generals Alfredo Ovando and his successor Juan José Torres. During this period

the Bolivian military regime promoted welfare reforms that were aimed at changing the

deplorable living conditions for the vast majority of Bolivians. The regime also

nationalized the Bolivian operations of the U.S.-based Gulf Oil Corporation (Klein,

2003). Throughout the early 1970s to the late 1980s, Panama was also ruled by populist

military regimes. While the regime led by General Omar Torrijoss was corrupt, the

regime expanded welfare programs and its nationalist foreign policy appealed to the rural

and urban constituencies that were traditionally ignored by the economic elite. After

Torrijoss’s death in 1981 the populist direction of military government was continued

under general Manuel Noriega (Robert C. Harding II, 2001).

While the above regimes were autocratic and presided over economies that were

heavily integrated into the global economy, what motivated their extensive welfare

allocations was not the pressure from democratic elections but a socialist or populist

ideological ethos, which defined and shaped the priorities of public policy.

5.2. Tight Elections and Welfare Expansion

With the attempt to increase countries’ integration into the global economy the

clientelistic nature of competitive electoral politics is expected to increase as political

parties will increasingly seek to provide welfare benefits to constituent voting districts

adversely affected by economic globalization. In chapter 2, this argument generated

hypothesis 3, which stated: Increasing levels of global economic integration are

associated with an increase in welfare spending when the level of electoral competition is

high. Table 5.2 presents the findings of the interactive effect of economic globalization

and electoral competition on welfare spending. Model 5.1a presents the results with the

main explanatory variables - economic globalization and electoral competition – without

the interaction term. And model 5.2b presents the main results with the interaction term.

[Table 5.2 about here]

Similar to Table 5.1, the direction of the coefficients for most of the control variables in

models 5.2a and 5.2b are consistent with the theoretical predictions and the reported

58

findings of previous empirical research. Prior levels of welfare spending (the lagged

dependent variable) are positively associated with current levels of welfare spending in

both models. Urbanization failed to rise to the level of statistical significance in both

models. Economic growth, population size, the ratio of countries’ young and old

dependents, and inflation are all negatively associated with welfare spending. Higher

levels of economic development are positively associated with welfare spending in both

models. While the oil shocks of the 1970s fail to reach statistical significance, the debt

crisis of the 1980s is negatively associated with welfare spending. And while economic

globalization fails to reach statistical significance in model 5.2a, electoral competition is

positively associated with welfare spending.

Model 5.2b shows the interactive effect of economic globalization and electoral

competition. Economic globalization (the constitutive element of the interaction term) is

negatively associated with welfare spending. However, since the coefficient on the

interaction term – Economic Globalization*Electoral Competition - is positive, this

reductive effect diminishes. In figure 5.2, the sloping line indicates the marginal effect of

economic globalization on welfare spending for meaningful changes in the level of

countries’ electoral competition. And 95% confidence intervals around the line show the

conditions under which economic globalization has a statistically significant effect on

welfare spending. As predicted, it can be observed from figure 5.2 that economic

globalization has a significant reductive effect on welfare spending when electoral

competition is low. However, this reductive effect diminishes as electoral competition

increases. Once the percentage of votes gained by smaller parties in parliamentary and or

presidential elections is greater than 29%, then economic globalization no longer has a

significant reductive effect on welfare spending. Hypothesis 3 is supported by the data.

[Figure 5.2 about here]

5.3. Voter Turnout and Social Generosity

It was argued that high voter turnout is reflective of the political entrance of first

time and working class voters who hold political incumbents responsible for the negative

externalities of global economic integration. And since incumbent politicians are largely

motivated by re-election, they are more likely to expand welfare expenditures to appease

59

the wrath of voters. In chapter 2 this argument led to hypothesis 4, which stated:

Increasing levels of global economic integration are associated with an increase in

welfare spending when the level of political participation is high.

In testing this hypothesis, table 5.3 presents ordinary least square estimates (OLS), which

assume that political participation - the percentage of the voting age population who

actually voted in parliamentary and presidential elections - is exogenous of welfare

spending. However, some scholars argue that political participation and welfare spending

are endogenous and therefore including the variable in equations that predict welfare

spending will generate biased estimates (Segura-Ubiergo, 2007). It is argued that in the

attempt to prevent conflict and increase social cohesion, governments in some countries

may actually increase welfare spending when political participation is low. In conflict

prone countries like Colombia and Guatemala where voter turnout rarely exceeds 35% to

45%, governments in these countries have a stronger incentive to increase welfare

spending than governments in Uruguay and Costa Rica, where voter turnout is over 80%

of the voting age population (Segura-Ubiergo, 2007, pp. 121-123).

Given this argument, Tables 5.3.1 and 5.3.2 present estimates from a two-stageleast-

squares instrumental variable model (IV), which assumes that political participation and

welfare spending are endogenous. The IV estimates utilize variables that instrument for

political participation. These include indicators that measure the number of legislative

chambers in countries as well as indicators that measure whether elections are governed

by proportional representation. Theoretically, these variables are considered important

determinants of voter turnout. It is argued that an increase in the number of legislative

chambers slows legislation, which in turn reduces turnout because citizens are more

likely to perceive their votes to be less decisive (Jackman, 1987). This indicator is a

dichotomous variable that measures the number of legislative chambers in each country.

The variable is coded 1 if countries’ legislative structure is bicameral and 0 otherwise

(Johnson and Wallack, 2006). Moreover, it is argued that elections that are governed by

proportional representation will increase voter turnout since votes are not wasted and

political parties have a greater incentive to mobilize voters even when their electoral base

is weak (Jackman, 1987, Lijphart, 1994). This variable is coded 0 for majoritarian

electoral systems; 1 for mixed-member majority systems; and 2 for closed-list

proportional representative systems (Gerring and Thacker, 2008, Gerring et al., 2005).

60

Table 5.3 presents OLS estimates of the interactive effect of economic globalization

and voter turnout on welfare spending. Model 5.3a presents the results with the main

explanatory variables - economic globalization and voter turnout – without the interaction

term. And model 5.3b presents the main results with the interaction term.

[Table 5.3 about here]

The direction of the coefficients for most of the control variables in models 5.3a and

5.3b are again consistent with the theoretical predictions and findings that are reported in

previous empirical research. Prior levels of welfare spending (the lagged dependent variable)

are positively associated with current levels of welfare spending in both models. Urbanization

and the ratio of countries’ young and old dependents fail to rise to the level of statistical

significance in both models. Economic growth, population size and inflation are all negatively

associated with welfare spending. Higher levels of economic development are positively

associated with welfare spending in both models. The oil shocks of the 1970s are negatively

associated with welfare spending in model

5.3a, but fail to reach statistical significance in model 5.3b. And the debt crisis of the

1980s is negatively associated with welfare spending in both models. Both economic

globalization and voter turnout are positively associated with welfare spending in model

5.3a. While these variables fail to reach statistical significance in model 5.3b, the

coefficient on the main interaction term – Economic Globalization*Voter Turnout - is

positive.

Since the possible endogeneity between political participation and welfare spending

could bias OLS estimates, it is necessary to estimate the data via an IV model, which

corrects for endogeneity. Table 5.3.2 presents IV estimates of the interactive effect of

economic globalization and voter turnout on welfare spending. Model 5.3.2a presents the

IV results with the main explanatory variables - economic globalization and voter turnout

– without the interaction term. And model 5.3.2b presents the main IV results with the

interaction term. However, before discussing these results, it is necessary to examine the

first stage regression results, which evaluate the relevance, the strength and the validity of

the instruments used in the IV models. More importantly, the discussion of the first stage

61

regression will also provide results from the empirical test for endogeneity and whether

OLS estimates in Table 5.3 are inconsistent.

5.3.1. Voter Turnout: First Stage Regression Results

Table 5.3.1 presents statistics from the first stage equations to assess the quality of the

instrumentation procedure. First, the underidentification test, as reported by the

Kleibergen-Paap statistic, is a test of whether the equations are identified. The test checks

the relevance of the excluded instruments (the exogenous instrumental variables that are

not included in the second stage regression). The statistic is a measure of the instruments’

relevance (i.e. the correlation with the endogenous variable – voter turnout - that cannot

be explained by other control variables), so a rejection of the null indicates that the

equations are identified and that the instruments are relevant (Maddala, 1992: 383-389).

In models 5.3.1a and 5.3.1b the small p-values of the Kleibergen-Paap statistic strongly

indicates that the equations are identified and that the instruments (bicameralism and

proportional representation) are relevant.

[Table. 5.3.1 about here]

Second, the weak identification test, as reported by the Kleibergen-Paap Wald F statistic

and Stock-Yogo critical values test, measures the strength of the instruments’ correlation

with the endogenous variable – voter turnout. Weak identification arises when the

instruments are correlated with the endogenous variable, but only weakly. While different

instrumental variable estimators are more robust to weak instruments, two-stage least

square can perform poorly when instruments are weak (Stock and Yogo, 2002).

Instruments are deemed weak if the first-stage F statistics is less than ten (Staiger and

Stock, 1997). In models 5.3.1a and 5.3.1b the Kleibergen-Paap Wald F statistics suggests

that the models do not suffer from weak instrumentation. This is also confirmed by the

Stock-Yogo critical values test, which compares the F statistic with the critical values for

the nominal bias and size distortion of TSLS. If the F statistic is larger than the critical

values, then the instruments are strong. In models 5.3.1a and 5.3.1b the critical values test

at 10% nominal bias as well as 10% size distortion is less than the F statistic, confirming

that the models do not suffer from weak instruments (Stock and Yogo, 2002).

62

Third, the overidentification test, as reported by the Hansen J statistic, assesses the

validity of the instruments. Specifically, the Hansen J statistic tests whether the

instruments are correlated with the equation’s error term. The null hypothesis is that the

instruments are not correlated with the error term (i.e. that they are correctly excluded

from the equations). Rejecting the null hypothesis indicates that the instruments are not

valid (Davidson and MacKinnon, 2004: 336-338). In models 5.3.1a and 5.3.1b the large

p-values of the Hansen J statistic preclude rejecting the null and suggest that the

instruments are indeed valid.

Fourth, a test for endogeneity in a regression estimated via instrumental variables, as

reported by the Durbin-Wu-Hausman test, evaluates whether voter turnout is endogenous

to welfare spending as suggested by recent research. The null hypothesis states that any

endogeneity among the regressors would not have deleterious effects on OLS estimates.

A rejection of the null indicates that the endogenous regressors' effects on the estimates

are meaningful and instrumental variable techniques are required (Davidson and

MacKinnon, 2004: 338-342). In models 5.3.1a and 5.3.1b the small p-values of the

Durbin-Wu-Hausman test reject the null and clearly indicate that endogeneity exists

between voter turnout and welfare spending and that both models are best estimated via

TSLS instrumental variable procedure.

5.3.2. Voter Turnout: Second Stage Regression Results

Table 5.3.2 presents estimates from the second stage regression. The direction of the

coefficients for most of the control variables in models 5.3.2a and 5.3.2b remain

consistent with the reported findings of previous empirical research. Prior levels of

welfare spending (the lagged dependent variable) are positively associated with current

levels of welfare spending in both models. Economic growth is again negatively

associated with welfare spending in both models. Higher levels of economic development

are positively associated with welfare spending in both models. Urbanization and the

ratio of countries’ young and old dependents fail to rise to the level of statistical

significance in model 5.3.2a but is negatively associated with welfare spending in model

5.3.2b. Population fails to rise to the level of statistical significance in model 5.3.2a but is

negatively associated with welfare spending in model 5.3.2b. Inflation is negatively

63

associated with welfare spending in model 5.3.2a but fails to reach statistical significance

in model 5.3.2b. The oil shocks of the 1970s fail to reach statistical significance in both

models. And the debt crisis of the 1980s is negatively associated with welfare spending in

model 5.3.2a but fails to reach statistical significance in model 5.3.2b. Both economic

globalization and voter turnout are positively associated with welfare spending in model

5.3.2a.

[Table. 5.3.2 about here]

In model 5.3.2b economic globalization (the constitutive element of the interaction term)

is negatively associated with welfare spending. However, since the coefficient on the

interaction term – Economic Globalization*Voter Turnout - is positive, this reductive

effect diminishes. In figure 5.3, the sloping line indicates the marginal effect of economic

globalization on welfare spending for meaningful changes in the level of voter turnout.

And 95% confidence intervals around the line show the conditions under which economic

globalization has a statistically significant effect on welfare spending. It can be observed

from figure 5.3 that economic globalization has a significant reductive effect on welfare

spending when voter turnout is low. However, this reductive effect diminishes as voter

turnout increases. Once voter turnout is greater than 35% of the voting age population,

then economic globalization no longer has a significant reductive effect on welfare

spending. Even after correcting for endogeneity, the results from the IV model support

hypothesis 4.

[Figure 5.3 about here]

5.4. Democratization and Higher Social Spending

Since countries’ domestic political systems, as noted in Chapter 2, is more than the sum

of their parts, it is also important to consider that globalization’s effect on social spending

is conditional on the nature of political institutions (which is an index that is comprised of

indicators that measure regime type, the level of electoral competition and political

participation). When regimes are democratic and the level of electoral competition and

political participation is high, then democratization defines the nature of political

64

institutions. The discussion generated hypothesis 5, which stated: Increasing levels of

global economic integration are associated with an increase in welfare spending when

democratization of political institutions is high.

Table 5.4 presents the findings of the aggregate interactive effect of economic

globalization and political institutions on welfare spending. Model 5.4a presents the

results with the main explanatory variables - economic globalization and political

institution – without the interaction term. And model 5.4b presents the main results with

the interaction term.

[Table 5.4 about here]

In both models the direction of the coefficients for most of the control variables are

consistent with the theoretical predictions and the reported findings of previous empirical

research. Prior levels of welfare spending (the lagged dependent variable) are positively

associated with current levels of welfare spending. Higher levels of economic

development are positively associated with welfare spending. Economic growth, inflation

and urbanization are all negatively associated with welfare spending. While population is

negatively associated with welfare spending in model 5.4a, it is positively associated with

welfare spending in model 5.4b. The ratio of countries’ young and old dependents is

positively associated with welfare spending in both models. The oil shocks of the 1970s

and the debt crisis of the 1980s fail to reach statistical significance in both models.

Economic globalization and political institutions are both positively associated with

welfare spending in model 5.4a. In model 5.4b, the interactive effect of economic

globalization and political institutions is positively associated with welfare spending and

supports hypothesis 5. The solid sloping line in figure 5.4 indicates the marginal effect of

economic globalization on welfare spending for meaningful changes in the

democratization of countries’ political institutions. 95% confidence intervals around the

line allow us to determine the conditions under which economic globalization has a

statistically significant effect on welfare spending. It is easy to see that economic

globalization has a reductive effect on welfare spending when the democratization of

political institutions is low. However, this reductive effect declines as the democratization

of political institutions increases. When the democratization score of countries’ political

65

institutions is greater than 6, then economic globalization no longer has a reductive effect

on welfare spending.

[Figure 5.4 about here]

5.5. Robustness Checks: Alternate Measures of Globalization and Social Spending

As in the previous chapter, alternate measures of welfare spending and economic

globalization are used to estimate the effect of economic globalization on welfare

spending when conditional on the authority characteristics of political regimes. Model

5.5a of Table 5.5 presents results for the interactive effect of the alternate measures of

economic globalization and political regimes on welfare spending. Previous welfare

spending is positively associated with current welfare spending. While all other control

variables fail to rise to the level of statistical significance, urbanization is positively

associated with welfare spending in model 5.5b. And while FDI is positively associated

with welfare spending in model 5.5a, its constitutive element and its interaction with

political regime fail to rise to the level of statistical significance in model 5.5b. Trade

openness (the constitutive element of the interaction term) is negatively associated with

welfare spending in model 5.5b. However, since the coefficient on the interaction term –

Trade*Political Regime - is positive, this reductive effect diminishes. In figure 5.5, the

sloping line indicates the marginal effect of trade openness on welfare spending for

meaningful changes in the authority characteristics of political regimes. And 95%

confidence intervals around the line allow us to determine the conditions under which

openness to trade has a statistically significant effect on welfare spending. It can be

observed from figure 5.5 that openness to trade has a reductive effect on welfare spending

when political regimes are autocratic. However, this reductive effect declines as the

authority characteristics of regimes become more democratic. When the polity score of

political regimes is greater than 4, then openness to trade no longer has a reductive effect

on welfare spending. Hypothesis 2 is also supported by the alternate measures of welfare

spending and economic globalization.

Model 5.6b of Table 5.6 presents results for the interactive effect of the alternate

measures of economic globalization and the level of countries’ electoral competition on

welfare spending. Previous welfare spending is positively associated with current welfare

66

spending. All other control variables fail to rise to the level of statistical significance. FDI

and its interaction with electoral competition have no statistical significant effect on

welfare spending. However, trade openness (the constitutive element of the interaction

term) is again negatively associated with welfare spending. However, since the

coefficient on the interaction term – Trade*Electoral Competition - is positive, this

reductive effect diminishes.

In figure 5.6, the sloping line indicates the marginal effect of trade openness on welfare

spending for meaningful changes in the level of countries’ electoral competition. And

95% confidence intervals around the line show the conditions under which economic

globalization has a statistically significant effect on welfare spending. It can be observed

from figure 5.6 that openness to trade has a significant reductive effect on welfare

spending when electoral competition is low. However, this reductive effect diminishes as

electoral competition increases. Once the percentage of votes gained by smaller parties in

parliamentary and or presidential elections is greater than 34%, then openness to trade no

longer has a significant reductive effect on welfare spending. Hypothesis 3 is supported

by the alternate measures for welfare spending and the economic globalization.

Model 5.7b of Table 5.7 presents results for the interactive effect of the alternate

measures of economic globalization and political participation on welfare spending.

Previous welfare spending is positively associated with current welfare spending. And

again, all other control variables fail to rise to the level of statistical significance. FDI and

its interaction with voter turnout have no statistically significant effect on welfare

spending. However, in model 5.7b, trade openness (the constitutive element of the

interaction term) is negatively associated with welfare spending. However, since the

coefficient on the interaction term – Trade*Voter Turnout - is positive, this reductive

effect diminishes.

In figure 5.7, the sloping line indicates the marginal effect of trade openness on welfare

spending for meaningful changes in the level of voter turnout. And the 95% confidence

intervals around the line show the conditions under which openness to trade has a

statistically significant effect on welfare spending. It can be observed from figure 5.7 that

openness to trade has a significant reductive effect on welfare spending when voter

turnout is low. However, this reductive effect diminishes as voter turnout increases. Once

voter turnout is greater than 15% of the voting age population, then openness to trade no

67

longer has a significant reductive effect on welfare spending. Hypothesis 4 is supported

by the alternate measures for economic globalization and welfare spending. Model 5.8b

of Table 5.8 presents results for the interactive effect of the alternate measures of

economic globalization and the index of political institutions on welfare spending.

Previous welfare spending is positively associated with current welfare spending. In

model 5.8a, both FDI and population are negatively associated with welfare spending,

while all other variables fail to rise to the level of statistical significance. Urbanization

fails to reach statistical significance in model 5.8a but is positively related to welfare

spending in model 5.8b. In model 5.8b, FDI and its interaction with domestic political

institution have no statistical significant effect on welfare spending. Trade openness (the

constitutive element of the interaction term) is negatively associated with welfare

spending. However, since the coefficient on the interaction term –

Trade*Domestic Political Institutions – is positive, this reductive effect diminishes.

In figure 5.8, the sloping line indicates the marginal effect of trade openness on welfare

spending for meaningful changes in the democratization of countries’ political

institutions. 95% confidence intervals around the line allow us to determine the

conditions under which openness to trade has a statistically significant effect on welfare

spending. It is easy to see that openness to trade has a reductive effect on welfare

spending when the democratization of political institutions is low. However, this

reductive effect declines as the democratization of political institutions increases. When

the democratization score of countries’ political institutions is greater than 11, then

openness to trade no longer has a reductive effect on welfare spending. Hypothesis 5 is

supported by the alternate measures of welfare spending and economic globalization.

5.6. Summation

The results presented in this chapter strongly suggest that domestic political forces avert

globalization’s ‘race to the bottom’ effect on social policy. The effect on social spending

from global economic integration is largely conditional on the nature of domestic

institutions that shape the political environment in which welfare policy is made. When

democratization of political institutions is low, economic globalization has a consistently

negative effect on welfare spending, an outcome that is consistent with the predictions by

efficiency theories. However, when democratization of political institutions is high,

68

domestic political forces will most likely avert globalization’s proclivity to retrench the

welfare state by expanding social spending.

5.7. Tables and Figures

Table 5.1: Economic Globalization, Political Regimes and Welfare Spending

Model 5.1a Model 5.1b

Without With

Interaction Interaction

Lagged Welfare Spending 0.003** 0.003**

(0.001) (0.001)

Economic Globalization 0.002** 0.001**

(0.000) (0.000)

Political Regime 0.004** -0.002

(0.001) (0.002)

Globalization *Political Regime

0.00014**

(0.000)

Urbanization -0.001** -0.001**

(0.001) (0.000)

Dependency 0.102** 0.093*

(0.037) (0.038)

Growth -0.004** -0.004**

(0.001) (0.001)

Ln GDP per capita 0.033** 0.033**

(0.007) (0.007)

Ln population -0.005 -0.006*

(0.002) (0.002)

Ln inflation -0.022** -0.022**

(0.004) (0.004)

Oil shock (70’s) -0.003 0.001

(0.005) (0.005)

69

Debt crisis (80’s) -0.009* -0.007

(0.005) (0.005)

Constant 0.091 0.135

(0.079) (0.072)

Observations 801 801

R-Squared 0.32 0.33

Panel-corrected standard errors are in parentheses. **significant at 1%; *significant at 5%

70

Table 5.1A

Country Years Polity Average Welfare Average Level of

Score Spending % Total Economic

Government Spending Globalization

______________________________________________________________________

Algeria 1973-1976 -9 0.45 30.00

Azerbaijan 1991-1994 -3 0.42 45.40

Azerbaijan 1995-1997 -6 0.46 50.00

Azerbaijan 1998-2004 -7 0.58 61.03

Bolivia 1971-1977 -7 0.40 44.02

Bolivia 1978-1979 -4 0.45 44.50

Kazakhstan 1991-1994 -3 0.53 60.10

Kazakhstan 1995-2001 -4 0.50 64.20

Kyrgyzstan 1991-2004 -3 0.43 60.15

Panama 1969-1977 -7 0.51 68.54

Panama 1982-1983 -5 0.45 69.00

Panama 1987-1988 -8 0.53 59.82

________________________________________________________________________

71

Table 5.2: Economic Globalization, Electoral Competition and Welfare Spending

Model 5.2a Model 5.2b

Without With

Interaction Interaction

Lagged Welfare Spending 0.002** 0.002**

(0.001) (0.001)

Economic Globalization 0.001 -0.005**

(0.000) (0.001)

Electoral Competition 0.028** -0.013

(0.008) (0.011)

Globalization*Electoral Competition 0.001**

(0.000)

Urbanization -0.000 -0.000

(0.000) (0.000)

Dependency -0.108** -0.107**

(0.024) (0.024)

Growth -0.003** -0.003**

(0.001) (0.001)

Ln GDP per capita 0.042** 0.045**

(0.009) (0.008)

Ln population -0.010** -0.010** (0.003) (0.003)

Ln inflation -0.021** -0.021** (0.004) (0.004)

Oil shock (70’s) -0.008 -0.007 (0.005) (0.005)

Debt crisis (80’s) -0.018** -0.017**

(0.004) (0.004)

______________________________________________________________________________

Constant 0.137 0.285**

(0.072) (0.075)

Observations 708 708

R-Squared 0.42 0.42

______________________________________________________________________________

Panel-corrected standard errors are in parentheses. **significant at 1%; *significant at 5%

72

73

Table 5.3: Economic Globalization, Voter Turnout and Welfare Spending

OLS Model 5.3a OLS Model 5.3b

Without With

Interaction Interaction

Lagged Welfare Spending 0.002** 0.002**

(0.001) (0.001)

Economic Globalization 0.001* -0.001

(0.000) (0.001)

Voter Turnout 0.160** -0.008

(0.027) (0.071)

Globalization*Voter Turnout

0.003*

(0.001)

Urbanization -0.000 -0.000

(0.000) (0.000)

Dependency -0.015 -0.005

(0.039) (0.041)

Growth -0.004** -0.004**

(0.001) (0.001)

Ln GDP per capita 0.047** 0.047**

(0.008) (0.007)

Ln population -0.007* -0.008**

(0.003) (0.003)

Ln inflation -0.021** -0.021**

(0.004) (0.005)

Oil shock (70’s) -0.015* -0.013

(0.007) (0.007)

Debt crisis (80’s) -0.020** -0.019**

(0.004) (0.004)

74

Constant -0.019 0.122

(0.085) (0.099) Observations

699 699

R-Squared 0.42 0.43

Panel-corrected standard errors are in parentheses. **significant at 1%;*significant at 5%

Table 5.3.1: Economic Globalization, Voter Turnout and Welfare Spending

Results from the First Stage IV Regression

IV Model 5.3.1a IV Model 5.3.1b

Without With

Interaction Interaction

p-value 0.00 0.02

________________________________________________________________________

Table 5.3.2: Economic Globalization, Voter Turnout and Welfare Spending

IV Model 5.3.2a IV Model 5.3.2b

Without With

75

Underidentification test:

Kleibergen-Paap rk LM statistic 67.48 15.01

p-value 0.00 0.00

Weak identification test:

Kleibergen-Paap rk Wald F statistic 40.71 37.23

Stock-Yogi weak ID critical values test:

TSLS size of nominal bias 10% 34.62 36.36

Maximal IV size distortion 10%

Overidentification test of all instruments:

19.93 19.93

Hansen J statistic 1.59 0.31

p-value

0.21 0.58

Test for Endogeneity:

Durbin-Wu-Hausman Test 10.16 5.12

Interaction Interaction Robust

standard errors are in parentheses. **significant at 1%;*significant at 5%

76

Lagged Welfare Spending 0.002** 0.002**

(0.001)

Economic Globalization 0.001** -0.026*

(0.000) (0.013)

Voter Turnout 0.349** -2.113

(0.099) (1.115)

Globalization*Voter Turnout

0.038*

(0.019)

Urbanization -0.000 -0.001**

(0.000) (0.000)

Dependency 0.122 -0.421*

(0.105) (0.218)

Growth

-0.005** -0.004*

(0.002) (0.002)

Ln GDP per capita

0.063** 0.089**

(0.015) (0.023)

Ln population

0.000 -0.021* (0.004) (0.008)

Ln inflation

-0.017** -0.003

(0.006) (0.007)

Oil shock (70’s)

-0.026 -0.004 (0.024) (0.031)

Debt crisis (80’s)

-0.031** -0.018 (0.011)

(0.013)

Constant -0.404 1.589

(0.220) (0.865)

Observations 546 546

R-Squared 0.44 0.24

77

Table 5.4: Economic Globalization, Political Institutions and Welfare Spending

Model 5.4a Model 5.4b

Without With

Interaction Interaction

Lagged Welfare Spending 0.003** 0.003**

(0.001) (0.001)

Economic Globalization 0.002** 0.001**

(0.000) (0.000)

Domestic Political Institution 0.037** -0.005 (0.007)

(0.013)

Globalization *Political Institution 0.001**

(0.000)

Urbanization -0.001** -0.001**

(0.000) (0.000)

Dependency 0.132** 0.112**

(0.041) (0.042)

Growth -0.004** -0.004**

(0.001) (0.001)

Ln GDP per capita 0.034** 0.035**

(0.007) (0.007)

Ln population -0.005* 0.007**

(0.002) (0.002)

Ln inflation -0.022** -0.021**

(0.004) (0.004)

Oil shock (70’s) -0.002 -0.000

(0.006) (0.005)

Debt crisis (80’s) -0.010 -0.008

(0.005) (0.005)

Constant 0.077 0.116

(0.082) (0.074)

Observations 801 801

R-Squared 0.33 0.34

78

Panel-corrected standard errors are in parentheses. **significant at 1%; *significant at 5%

79

Table 5.5: Trade, FDI, Political Regime and Welfare Spending

Model 5.5a Model 5.5b

Without With

Interaction Interaction

Lagged Welfare Spending 0.786** 0.772**

(0.082) (0.081)

Trade -0.000 -3.34e-06*

(0.000) (0.000)

FDI -2.15e-06* -0.000

(0.000) (0.000)

Political Regime 0.000 -0.001

(0.000) (0.001)

Trade*Political Regime 5.40e-07**

(0.000)

FDI*Political Regime 0.000

(0.000)

Urbanization 0.000 0.0002*

(0.000) (0.000)

80

Constant -0.002 -0.010

(0.037) (0.039)

Observations 923 923

R-Squared 0.69 0.70

Dependency 0.011 0.015

(0.012) (0.013)

Growth 0.000 0.000

(0.000) (0.000)

Ln GDP per capita 0.003 0.003

(0.003) (0.003)

Ln population -0.002* -0.001

(0.001) (0.001)

Ln inflation -0.000 -0.001

(0.001) (0.001)

Oil shock (70’s) 0.001 0.003

(0.003) (0.003)

Debt crisis (80’s) 0.001 0.002

(0.002) (0.002)

Panel-corrected standard errors are in parentheses. **significant at 1%; *significant at 5%

The

Authority Characteristics of Political Regimes Dashed lines give 95%

confidence interval.

81

Table 5.6: Trade, FDI, Electoral Competition and Welfare Spending

Model 5.6a Model 5.6b

Without With

Interaction Interaction

FDI*Electoral Competition 0.000

(0.000)

Urbanization 0.000 0.000

(0.000) (0.000)

Dependency 0.011 0.014

82

Lagged Welfare Spending 0.812** 0.804**

(0.075) (0.074)

Trade -0.000 -5.93e-06**

(0.000) (0.000)

FDI -0.000 -0.000

(0.000) (0.000)

Electoral Competition 0.000 -0.000

(0.000) (0.000)

Trade*Electoral Competition 1.26e-07**

(0.000)

Constant -0.018 -0.015

(0.036) (0.035)

Observations 946 946

R-Squared 0.70 0.70

(0.013) (0.014)

Growth 0.000 0.000

(0.000) (0.000)

Ln GDP per capita 0.003 0.003

(0.003) (0.003)

Ln population -0.001 -0.000

(0.001) (0.001)

Ln inflation -0.001 -0.001

(0.001) (0.001)

Oil shock (70’s) 0.002 0.003

(0.003) (0.003)

Debt crisis (80’s) 0.002 0.002

(0.002) (0.002)

Panel-corrected standard errors are in parentheses. **significant at 1%; *significant at 5%

83

Table 5.7: Trade, FDI, Voter Turnout and Welfare Spending

Model 5.7a Model 5.7b

Without With

Interaction Interaction

Lagged Welfare Spending 0.809** 0.802**

(0.076) (0.076)

Trade -0.000 -4.13e-06**

(0.000) (0.000)

FDI -0.000 -0.000

(0.000) (0.000)

Voter Turnout 0.000 -0.000

(0.000) (0.000)

Trade*Voter Turnout 1.04e-07*

(0.000)

FDI*Voter Turnout 0.000

(0.000)

Urbanization 0.000 0.000

(0.000) (0.000)

Dependency 0.015 0.017

(0.014) (0.015)

Growth 0.000 0.000

(0.000) (0.000)

Ln GDP per capita 0.003 0.003

(0.003) (0.003)

Ln population -0.001 -0.000

(0.001) (0.001)

Ln inflation -0.000 -0.001

(0.001) (0.001)

Oil shock (70’s) 0.002 0.003

(0.003) (0.003)

Debt crisis (80’s) 0.002 0.002

(0.002) (0.002)

Constant -0.022 -0.021

84

(0.036) (0.036)

Observations 946 946

R-Squared 0.70 0.70

Panel-corrected standard errors are in parentheses. **significant at 1%; *significant at 5%

85

Table 5.8: Trade, FDI, Domestic Political Institutions and Welfare Spending

Model 5.8a Model 5.8b

Without With

Interaction Interaction

Lagged Welfare Spending 0.786** 0.775**

(0.082) (0.135)

Trade -0.000 -2.08e-06*

(0.000) (0.000)

FDI -2.14e-06* -0.000

(0.000) (0.000)

Domestic Political Institution 0.001 -0.006

(0.002)

(0.004)

Trade*Domestic 3.20e-06**

Political Institution (0.000)

FDI*Domestic -0.000

Political Institution

(0.000)

Urbanization 0.000 0.00014*

(0.000)

Dependency 0.013 0.015

(0.012) (0.013)

Growth 0.000 0.000

(0.000)

Ln GDP per capita 0.003 0.003

(0.003) (0.003)

Ln population -0.002** -0.000

(0.001)

Ln inflation -0.001 -0.001

(0.001) (0.001)

Oil shock (70’s) 0.001 0.002

(0.003) (0.003)

Debt crisis (80’s) 0.001 0.001

86

(0.002) (0.002)

Constant -0.002 -0.012

(0.037) (0.040)

Observations 923 923

R-Squared 0.70 0.70

Panel-corrected standard errors are in parentheses. **significant at 1%; *significant at 5%

The Democratization of Domestic Political Institutions

Dashed lines give 95% confidence interval.

87

Chapter 6

The Distribution of Welfare Compensation: Globalization, Political

Affiliation and Social Policy

The previous chapter presented evidence that the political institutional environment that

shapes the incentives and preferences of policy makers who make social policy averts

globalization’s natural proclivity to retrench the welfare state. Domestic political actors

such as labor unions and political parties shape the distribution of states’ social policy. Do

the political and ideological preferences of these actors also apply the brakes to

globalization’s ‘race to the bottom’ effect on social spending? The discussion that follows

addresses this question.

6.1. Labor Unions and Social Generosity

Recent scholarship contends that globalization’s effect on states’ social spending is

conditional on the relative strength of organized labor. It is argued that the power of

organized labor is displayed when it exerts pressure on governments to increase welfare

spending by effectively lobbying the WTO to link labor standards to free trade

agreements (Hughes and Wilkinson, 1998, Wachtel, 1998, Wet, 1995, Wilkinson and

Hughes, 2000). The discussion in chapter 2 generated hypothesis 6, which stated:

Increasing levels of global economic integration are associated with an increase in

welfare spending when labor unions are strong.

Yet, as discussed in chapter 2, other scholars challenge this notion arguing instead that

global economic integration has weakened the power of organized labor to influence

states’ welfare policies. Organized labor’s weakened position is a function of the fact that

the labor movements in many countries have been transformed from being militant

organizations that once pressed for the rights of the working class, to being co-opted into

88

the free market priorities of transnational capitalism and increasingly see themselves as

partners to corporate capital no longer willing to press the state for welfare benefits (Gray,

2008, Gray, 2007). These counter-arguments generated hypothesis 7, which stated:

Increasing levels of global economic integration is not associated with an increase in

welfare spending regardless of the strength of labor unions.

Table 6.1 presents the findings of the interactive effect of economic globalization and

labor unions on welfare spending. Model 6.1a presents the results with the main

explanatory variables - economic globalization and labor unions – without the interaction

term. And model 6.1b presents the main results with the interaction term.

[Table 6.1 about here]

The direction of the coefficients for most of the control variables in models 6.1a and 6.1b

are consistent with theoretical predictions that are reported in the findings of previous

research. Prior levels of welfare spending (the lagged dependent variable) are positively

associated with current levels of welfare spending. Urbanization, economic growth,

population, inflation and the debt crisis of the 1980s are all negatively associated with

welfare spending. Higher levels of economic development are positively associated with

welfare spending. The ratio of countries’ young and old dependents and the oil shocks of

the 1970s fail to reach statistical significance.

In model 6.1a, economic globalization and organized labor are positively associated with

welfare spending. While model 6.1b shows that both economic globalization and

organized labor (the constitutive elements of the interaction term) are positively

associated with welfare spending, the coefficient on the interaction term – Economic

Globalization*Labor Power – has no statistically significant effect on welfare spending,

which support hypothesis 7. In other words, this suggest that organized labor’s effect on

states’ welfare spending is not through the forces of global economic integration. This

finding lends support to the argument that economic globalization may have weakened

and co-opted labor to support the priorities of transnational capital.

89

6.2. Party Ideology and Social Generosity

Social democratic theories contend that economic globalization’s effect on social

spending is conditional on the nature of party politics. Governments led by leftist or

centrist political parties are more likely to support robust welfare policies than

governments led by parties to the political right (Huber and Stephens, 2001, Stephens,

2005). In chapter 2, these arguments generated hypothesis 8, which stated: Increasing

levels of global economic integration are associated with an increase in welfare spending

when ruling political parties are on the left.

Table 6.2 presents the findings of the interactive effect of economic globalization and the

ideology of ruling parties on welfare spending. Model 6.2a presents the results with the

main explanatory variables - economic globalization and the ideology of the ruling party

– without the interaction term. And model 6.2b presents the main results with the

interaction term.

[Table 6.2 about here]

The direction of the coefficients for most of the control variables in models 6.2a and 6.2b

are consistent with the theoretical predictions that are reported in the findings of previous

research. Prior levels of welfare spending (the lagged dependent variable) are positively

associated with current levels of welfare spending. Urbanization, economic growth,

population, inflation, the oil shock of the 1970s (in model 6.2a only) and the debt crisis of

the 1980s are all negatively associated with welfare spending. Higher levels of economic

development are positively associated with welfare spending. The ratio of countries’

young and old dependents and the oil shocks of the 1970s (in model 6.2b only) fail to

reach statistical significance.

Economic globalization is positively associated with welfare spending in models 6.2a

and 6.2b. The ideology of the ruling political party is positively associated with welfare

spending in model 6.2a, but fails to reach statistical significance in model 6.2b. The

coefficient on the interaction term – Economic Globalization*Ruling Party Ideology - is

positive. In figure 6.2, the sloping line indicates the marginal effect of economic

globalization on welfare spending for meaningful changes in the ideology of ruling

political parties. And 95% confidence intervals around the line show the conditions under

90

which economic globalization has a statistically significant effect on welfare spending.

Figure 6.2 shows that global economic integration increases welfare spending regardless

of political parties. While welfare expenditures are not significantly reduced under right

parties, left parties tend to be more generous in the provision of welfare benefits. From

this perspective, since social democratic theory predicted that right parties are more likely

to slash states’ welfare spending, while left parties are expected to significantly increase

spending, hypotheses 8 is partially confirmed by the finding. A clearer picture emerges

when considering this issue from a different vantage point. In figure 6.2a, the sloping line

indicates the marginal effect of the ideology of ruling parties on welfare spending for

meaningful changes in countries’ integration into the global economy. From this

perspective, right parties have a reductive effect on welfare spending at low levels of

global economic integration. However, once left of center parties govern and countries’

integration into the global economy exceeds a score of 45 on the KOF index, then an

ideological movement from the right to the left increases welfare spending.

6.3. Robustness Checks: Alternate Measures of Globalization and Social Spending. As

was done in the previous empirical chapters, alternate measures of social spending and

economic globalization are used to provide additional empirical test for the study’s

hypotheses. Table 6.3 presents the findings of the interactive effect of labor unions and

openness to trade and FDI on welfare spending. Model 6.3a presents the results with the

main explanatory variables - labor unions and openness to trade and FDI – without the

interaction terms. And model 6.3b presents the main results with the interaction terms.

[Table 6.3 about here]

The direction of the coefficients for most of the control variables in models 6.3a and 6.3b

are also consistent with theoretical predictions that have been reported in previous

research. Prior levels of welfare spending (the lagged dependent variable) are positively

associated with current levels of welfare spending. Urbanization, economic growth,

population, inflation, the oil shock of the 1970s and the debt crisis of the 1980s are all

negatively associated with welfare spending. Consistent with Wagner’s Law, higher levels

91

of economic development is positively associated with welfare spending. And the ratio of

countries’ young and old dependents fails to reach statistical significance.

The power of organized labor has a positive effect on welfare spending in model

6.3a, but interestingly becomes negative in model 6.3b. Openness to trade is negatively

associated with welfare spending, in models 6.3a and 6.3b. However, since the coefficient on

the interaction term – Labor Power*Trade – is positive, this reductive effect diminishes, a

finding that is consistent with hypothesis 6. In figures 6.3, the sloping line indicates the

marginal effect of trade openness on welfare spending for meaningful changes in the Potential

Labor Power index (PLP). And 95% confidence intervals around the line show the conditions

under which trade openness has a statistically significant effect on welfare spending. It can be

observed from figure 6.3 that trade openness has a significant reductive effect on welfare

spending when the power of organized labor on the PLP index low. However, this reductive

effect diminishes as the power of organized labor increases. Once labor power is greater than 10

percentage points on the PLP index then openness to international trade no longer has a

significant reductive effect on welfare spending.

While this analysis has shown that the effects on the welfare state that come from global

economic integration, measured in terms of the KOF globalization index - are not

conditional on the strength of organized labor, the effects on welfare spending that come

from openness to international trade are conditional on the power of the labor movement.

This finding seems to support the argument that organized labor can affect welfare policy

by attaching labor standards to free trade agreements, which bind governments into

making concessions for increased welfare benefits.

Table 6.4 presents the findings of the interactive effect of party ideology and openness to

trade and FDI on welfare spending. Model 6.4a presents the results with the main

explanatory variables – party ideology and openness to trade and FDI – without the

interaction terms. And model 6.4b presents the main results with the interaction terms.

[Table. 6.4 about here]

The directions of the coefficients for most of the control variables in models 6.4a and

6.4b are consistent with the theoretical predictions that are reported in the findings of

previous research. Again, prior levels of welfare spending (the lagged dependent variable)

92

are positively associated with current levels of welfare spending. Urbanization, economic

growth, population, inflation, the oil shock of the 1970s and the debt crisis of the 1980s

are all negatively associated with welfare spending. Higher levels of economic

development and the level of urbanization are positively associated with welfare

spending. The ratio of countries’ young and old dependents and FDI all fail to reach

statistical significance. While openness to trade is negatively associated with welfare

spending in models 6.4a and 6.4b, party ideology fails to reach statistical significance.

And the interactive terms Party Ideology*Trade and Party Ideology*FDI also fail to reach

statistical significance, suggesting that these alternative measures of economic

globalization do not affect states’ welfare spending via the ideology of political parties.

However, the finding presented in Table 6.2 does show that economic globalization,

broadly defined and measured by the KOF index, positively affect welfare spending

through party ideology.

6.4. Summation

The results presented in this chapter suggest that political and ideological preferences of

labor unions and political parties avert globalization’s downward pressure on social

policy. When organized labor is politically strong, economic globalization, measured in

terms of openness to trade, will have a positive effect on states’ welfare spending.

Economic globalization will increase welfare spending regardless of the ideology of

ruling political parties, although relative to right parties, left parties provide more

generous welfare benefits. The study’s cross-national empirical findings have shown that

globalization’s effect on the welfare state is conditional on the nature of domestic

political institutions and political actors. In the chapters that follow these findings are

confirmed via the comparative case studies of globalization’s effects on the welfare state

in South Korea, Chile and Spain.

6.5. Tables and Figures

Table 6.1: Economic Globalization, Organized Labor and Welfare Spending

Model 6.1a Model 6.1b

Without With

Interaction Interaction

Lagged Welfare Spending 0.003* 0.003**

(0.001) (0.001)

93

Economic Globalization 0.001** 0.002**

(0.000) (0.000)

Labor Power 0.001** 0.002**

(0.000) (0.000)

Economic Globalization* Labor Power

-0.000

(0.000)

Urbanization

-0.001** -0.001**

(0.000) (0.000)

Dependency -0.002 0.001

(0.024) (0.024)

Growth -0.005** -0.005**

(0.001) (0.001)

Ln GDP per capita 0.030** 0.029**

(0.006) (0.006)

Ln population -0.006** -0.006**

(0.002) (0.002)

Ln inflation -0.018** -0.018**

(0.005) (0.003)

Oil shock (70’s) -0.013 -0.014

(0.007) (0.007)

Debt crisis (80’s) -0.013* -0.014*

(0.006) (0.007)

Constant 0.249** 0.244**

(0.080) (0.079)

Observations 737 737

R-Squared 0.31 0.31

Panel-corrected standard errors are in parentheses. **significant at 1%; *significant at 5%

Table 6.2: Economic Globalization, Party Ideology and Welfare Spending

Model 6.2a Model 6.2b

Without With

94

Interaction Interaction

Panel-corrected standard errors are in parentheses. **significant at 1%; *significant at 5%

95

Lagged Welfare Spending 0.002** 0.002**

(0.001) (0.001)

Economic Globalization 0.002** 0.002**

(0.000) (0.000)

Ruling Party Ideology 0.012** -0.024

(0.004) (0.013)

Globalization*Ruling Party Ideology 0.001**

(0.000)

Urbanization -0.0005* -0.0005*

(0.000) (0.000)

Dependency -0.017 -0.013

(0.031) (0.032)

Growth -0.004** -0.004**

(0.001) (0.001)

Ln GDP per capita 0.037** 0.033**

(0.008) (0.008)

Ln population -0.008** -0.007**

(0.002) (0.002)

Ln inflation -0.015** -0.015**

(0.004) (0.004)

Oil shock (70’s) -0.013* -0.011

(0.006) (0.006)

Debt crisis (80’s) -0.017** -0.014**

(0.004) (0.004)

Constant 0.173 0.194**

(0.089) (0.090)

Observations 717 717

R-Squared 0.34 0.35

96

97

Table 6.3: Trade, FDI, Organized Labor and Welfare Spending

Model 10.a Model 10.b

Without With

Interaction Interaction

Lagged Welfare Spending 0.003** 0.003**

(0.001) (0.001)

Trade -0.0005** -0.001**

(0.000) (0.000)

FDI 0.001 0.002

(0.002) (0.002)

Labor Power 0.001** -0.003** (0.000)

(0.000)

Labor Power*Trade 0.00009**

(0.000)

Labor Power*FDI -0.000

(0.000)

Urbanization -0.001** -0.001**

(0.000) (0.000)

Dependency -0.005 0.014

(0.028) (0.029)

Growth -0.004** -0.005**

(0.001) (0.001)

Ln GDP per capita 0.052** 0.063**

(0.007) (0.008)

Ln population -0.014** -0.009**

(0.003) (0.003)

Ln inflation -0.023** -0.025**

(0.005) (0.005)

Oil shock (70’s) -0.022** -0.022*

(0.009) (0.009)

Debt crisis (80’s) -0.022* -0.023**

(0.098) (0.111)

Constant 0.244** 0.129

98

(0.098) (0.111)

Observations 709 709

R-Squared 0.31 0.33

Panel-corrected standard errors are in parentheses. **significant at 1%; *significant at 5%

99

Table 6.4: Party Ideology, Trade, FDI and Welfare Spending

Model 6.4a Model 6.4b

Without With

Interaction Interaction

Lagged Welfare Spending 0.003* 0.003*

(0.010) (0.001)

Trade -0.0005** -0.0005**

(0.000) (0.000)

FDI 0.000 0.001

(0.002) (0.002)

Ruling Party Ideology 0.008 0.007

(0.004) (0.008)

Party Ideology*Trade 0.000

(0.000)

Party Ideology*FDI -0.005

(0.000)

Urbanization -0.0005* -0.0004*

(0.000) (0.000)

Dependency -0.022 0.026

(0.030) (0.030)

Growth -0.004** -0.004**

(0.001) (0.001)

Ln GDP per capita 0.053** 0.051**

(0.007) (0.007)

Ln population -0.018** -0.018**

(0.001) (0.002)

Ln inflation -0.024** -0.024**

(0.004) (0.004)

Oil shock (70’s) -0.023** -0.022**

(0.005) (0.005)

Debt crisis (80’s) -0.029** -0.029**

(0.004) (0.005)

100

Constant 0.266** 0.281**

(0.087) (0.087)

Observations 757 757

R-Squared 0.31 0.31

Panel-corrected standard errors are in parentheses. **significant at 5%; *significant at 1%

Part III

Economic Globalization and the Welfare

State in Emerging Economies with Authoritarian Political

Histories

101

Chapter 7

Globalization and the Welfare State in South Korea

7.1. Introduction

The South Korean welfare state has grown over the past forty years from a system

with limited benefits into a system that is relatively comprehensive. What explains this

dynamic is the interplay between the variation of South Korea’s global economic

integration and variation in the nature of its political institutions, which was characterized

by a transition from authoritarian military rule to democratic governance.

South Korea’s military regimes used social welfare as an instrument to promote

economic growth and industrialization. The welfare state under South Korea’s military

regimes was an ideal portrayal of the developmental use of social policy. The essential

components of the corporatist consensus of South Korea’s developmental welfare state

included: trade policies that protected corporate capital from foreign competition;

corporate capital’s provision of full employment and private welfare benefits to industry

workers; and labor unions acceptance of limited welfare benefits from the state in

exchange for full employment of its members (Goodman and White, 1998, p. 17, Wade,

1990, White, 1988, Woo-Cumings, 1999). This chapter presents a discussion of how the

pressures of global economic integration and political democratization unraveled this

consensus making social policy in South Korea increasingly consistent with

compensatory approaches of the welfare state.

Many intricacies are associated with the dynamics of South Korea’s welfare state

making it desirable to present an historical overview of its evolution from

authoritarianism to democracy. This will comprise the content of the next section, to be

followed by a discussion of the corporatist consensus of South Korea’s developmental

welfare state. This is followed by a discussion of the various welfare programs under the

military regimes and the ways in which these programs were limited in coverage and

were subordinated to the priorities of industrial development. The next sections discuss

how the process of economic globalization and democratic governance unraveled the

consensus of the developmental welfare state and significantly transformed welfare

policy in South Korea. The chapter concludes with a discussion of the future challenges

102

and the limitations of South Korea’s welfare state in an increasingly integrated global

economy.

7.2. Historical Overview

In 1961 a military coup d'état led by Major General Park Chung-hee effectively

ended the democratic government in South Korea’s Second Republic. After the

assassination of Park Chung-hee in 1979, popular protest against the military regime

emerged. These protests reached a climax with the coup d'état led by Chun Doo-hwan and

the establishment of martial law. In 1980, millions of people protested the military

regime’s massacre of Gwangju, which killed 207 democratic protesters. As a result of

these protests, the regime made political concessions for the transition to democracy.

When the first democratic presidential election was held in 1987, Kim Young-sam and

Kim Dae-jung, former critics of South Korea’s military regimes, ran against each other

and split the opposition vote, enabling former general Roh Tae-woo to win the election. In

1991, Kim Young-sam’s Unification Democracy Party merged with Roh's ruling

Democratic Justice Party. And as the candidate of the ruling conservative political party

he defeated Kim Dae-jung in the 1992 presidential election. The Asian financial crisis in

1997 and the subsequent collapse of the South Korean economy eroded the electoral

credibility of the ruling conservative government and in that year ushered into office the

liberal government of Kim Dae-jung.

7.3. The Corporatist Consensus of the Developmental Welfare State

South Korea’s military regimes induced agreements on economic and social

policy among labor unions and corporate capital. There was consensus on the importance

of protecting the domestic market from international competition. Trade policies

advantaged domestic firms by encouraging exports and restricting imports. In return,

domestic firms provided full employment and private welfare benefits to workers in the

industrial sector (Robinson, 2002, 257, Wiarda, 1997). Labor unions accepted limited

welfare benefits from the state in exchange for economic growth and stable employment

(Park, 2008, Song and Hong, 2005).

Since an expansive welfare state was seen as an obstacle to economic development, social

spending was restricted and subordinated to the priorities of industrial development. The

103

state’s expenditure on education and job training were considered as investments in

industrial development rather than social welfare (Goodman and Peng, 1996, Holliday,

2000, Jones, 1993, Kwon, 2007, Kwon and Holliday, 2007). Consequently, expenditures

for education and job training accounted for 85-90% of all social spending, while

expenditures on health and social protection remained at 1% of total government

spending throughout the 1970s and 1980s (United Nations, 2007). Expenditure on

education was seen as providing a foundation for transforming South Korea from a low-

skilled labor-intensive economy to one that produced high value-added technologies for

export markets. In this regard, welfare benefits were limited to workers in the public and

industrial sectors that were critical to economic development. The state provided

generous tax-cuts and investment subsidies to encourage private firms to provide health

care and social protection benefits to their workers. With the exception of public pensions

for workers in the public sector, private firms and the contributions of their workers

funded all the social insurance and pension programs (Kwon, 2007, Kwon, 1999, Kwon

et al., 2009). In short, since welfare policy was subordinated to the priorities of South

Korea’s industrial development, the essential logic of the developmental welfare state

was largely consistent with efficiency theoretical explanations of states’ welfare policies.

7.4. The Social Policies of the Military Regimes

The welfare policies of South Korea’s developmental welfare state can be

categorized into three programs. These include social insurance, which included

industrial accident insurance, national health insurance, and employment insurance;

public pensions; and public assistance (Kim, 2007, Kwon, 2007).

In 1963, the Park Chung-hee regime introduced the industrial accident insurance

program to promote economic development by facilitating corporate investments in hard

industry like construction, shipbuilding, and the machine industry. The coverage of the

program was funded by the corporate sector and was restricted to firms that had more

than 500 employees (Kwon, 2007). Two years later, the government introduced health

insurance that only provided coverage for workers in the public sector, private schools

and those employed in the corporate sector of the economy. The state’s financial

contribution to health insurance was negligible and limited to funding the cost of

administering the program. However, since the funding for health insurance came from

104

the financial contributions of workers and their employers, coverage was unaffordable for

the self-employed, small businesses, and rural farmers – the most vulnerable in South

Korean society (Kim, 2007, Kwon, 2007). Similarly, the government’s unemployment

insurance program, which provided benefits to workers displaced by temporary

unemployment was only available to workers in the private sector who were critical to the

country’s economic development (Kwon, 2007, Kwon, 1999).

While the social insurance program was used to promote economic development,

public pensions were designed to consolidate the political support of the military regime

and consequently was only available to government and military personnel (Song and

Hong, 2005). By 1988 the military government introduced a national pension program

(Kim, 2007, Kwon, 2007). And similar to the national health insurance, the national

pension program was only available to those workers that were critical to South Korea’s

industrial development (Kwon, 2007, Kwon, 1999). The national pension program was

also financed by the contributions of industrial workers and private firms (Kim, 2007).

Public assistance is the only social protection program that was fully funded by

the military regime. The program provided small amounts of food, condiments, clothing,

monetary allowance or “consolation money” and subsidized heating costs for the aged.

However, the program was not universal and was restricted to the elderly whose family

members were not of a working age (Song and Hong, 2005).

7.5. Political Democratization and Economic Liberalization

The national protest against the Chun Doo-hwan military regime during the 1980s

forced the government to make several concessions for the transition to democracy. These

included the amendment of the constitution, the promotion of basic human rights and

greater freedoms to political parties, the restoration of presidential elections, the

introduction of local autonomy, the promotion of free speech and the extension of

amnesty to pro-democracy leader, Kim Dae-jung (Han-Gyo-Rae, 1987). In December

1987, presidential election was restored, and with the election of the conservative

government led by Kim Young Sam, the first civilian regime in 1993, the government

embarked on a number of political reforms. The political power of the state was

decentralized by giving regional jurisdictions more autonomy. The electoral system was

restructured to increase voter participation in national elections.

105

Political democratization was followed by international demands to liberalize

South Korea’s economy. Major trading partners, like the U.S., pressed the government to

liberalize South Korea’s trade and investment policies that protected domestic firms from

foreign competition. In response, the government liberalized its trade policy, which

effectively ended the protectionist practices of the previous military regimes. Policies that

protected the domestic market from foreign capital flows were also abandoned, while the

government’s membership to the World Trade Organization (WTO) in 1993 and the

Organization for Economic Cooperation and Development (OECD) in 1996 strengthened

domestic economic reform and deepened South Korea’s integration into the global

economy (Yang and Moon, 2005). As a result, trade expansion accounted for over 50% of

South Korea’s GDP throughout the 1990s. Capital flows into South Korea accounted for

US$538 million in 1986 and increased to US$1,473 million in 1987. These increases

continued throughout the 1990s, exceeding US$10,200 million in 1997 (Song and Hong,

2005).

7.6. The Crisis of the Developmental Welfare State

The liberalization of the South Korean economy was an abrogation of the

corporatist consensus of the developmental welfare state that was based on trade

protection. Domestic firms could no longer rely on the state to protect them from foreign

competition. In addition, facing increasing competition from foreign multinationals, they

could no longer afford to provide full employment or private welfare for the industrial

work force. Foreign competition, especially in labor-intensive industries, increased

production costs and forced many firms into bankruptcy, which increased unemployment.

With rising unemployment, labor unions would increasingly look to the state to provide a

social safety-net (Song and Hong, 2005, Yang and Moon, 2005).

The response of the Kim Young Sam government was to reform the existing

developmental welfare system by incrementally expanding the government’s role in the

system. The government provided funding for job training, job security grants and

unemployment compensation through the Employment Insurance Program. Unemployed

workers from the industrial sector were eligible for unemployment benefits and job

security grants were given to domestic firms to discourage them from cutting their

106

payrolls. The Employment Insurance Program covered unemployed workers in the

private sectors and private school teachers. And similar to the welfare policies of the

previous military regimes, the program was financed by the contributions of workers and

their employers (Kwon, 2007). The government also expanded the National Pension

Program to the self-employed and established several task force committees to study

reforming social insurance, pensions, health, housing, education, public protection and

employment. The committees included the Welfare Policy Committee for the Disabled

and the Medical Care Reform Committee, and the Committee for Social Security Policy,

which was responsible for drafting the Five-Year Welfare Development Plan (Song and

Hong, 2005).

While the previous military regimes viewed welfare policy as an instrument for

industrial development, Kim Young Sam’s conservative democratic government viewed it

as a mechanism to improve the quality of life as well as an instrument to facilitate

development (Song and Hong, 2005). However, welfare reform by the Kim Young Sam

government was never comprehensive or universal in scope. The role of the state in

welfare reform was restricted to being an organizer of the system, still passing the

financial burden of welfare to private companies, social and religious organizations and

individual workers (Koh, 1998). In this respect, the welfare policies of Kim Young Sam’s

conservative government were fundamentally on the same line with those of the previous

military regimes.

7.7. The Asian Financial Crisis, the Democratic Left and the IMF

Attempts at welfare reform collapsed in face of the Asian financial crisis in 1997.

The crisis started in Thailand with the financial collapse of the Thai baht, which was

driven in part by the financial overextension in the real estate market and the country’s

rising foreign debt. As the crisis spread, most of Southeast Asia saw slumping currencies,

devalued stock markets and a steep rise in private debt. The financial contagion also

spread to Northeast Asian countries like South Korea. Indonesia, Thailand and South

Korea were the countries the most affected by the crisis. The South Korean economy

experienced negative growth rates of 6.7 percent, unemployment rose to 7 percent in

1998, then climbing to 8.6 percent in early 1999, and real income plummeted to 9.3%

(Song and Hong, 2005, Yang and Moon, 2005).

107

South Korea’s electorate blamed the conservative government of Kim Young Sam

for the financial crisis and in the presidential elections in 1997 replaced the government

with the liberal government of Kim Dae-jung. The new government inherited an economy

with high levels of debt and depleted foreign exchange reserves. To restore the financial

stability of the economy the government accepted a short-term stabilization loan from the

International Monetary Fund (IMF). The economic reforms demanded by the IMF

included the further liberalization of trade, the privatization of public enterprises, the

tightening of monetary policy to arrest the depreciation of the currency, increases in the

interest rate to control inflation, and banking reforms to promote greater transparency, all

of which furthered the process of liberalizing the South Korean economy (Yang and

Moon, 2005). Higher interest rates, in the short-term, coupled with the further

liberalization of the economy undermined domestic investment, increased corporate

bankruptcies, unemployment, and poverty. The urban poverty rate in South Korea surged

from 7 percent in 1997 to 21 percent in 1998 (National Statistics Office, 1999, Song and

Hong, 2005).

7.8. Social Democratic Corporatism and Comprehensive Welfare Reform

The management of the economic crisis provided the liberal government with the

opportunity to fundamentally transform welfare policy from one that was largely selective

and based on industrial development to one that was more universal. In pursuing welfare

reform, the government attempted to navigate the implementation of two conflicting

policies: neoliberal economic reform to resolve the financial crisis and social reforms to

provide adequate safety-nets to those displaced by the financial crisis. Neoliberal

economic reform threatened to impede the development of social reform, the policy

preference of labor unions and the lower classes – the main supporters of the liberal

government (Song and Hong, 2005). On the other hand, the neoliberal economic policies

of the IMF were important to stabilize the market to facilitate economic growth.

In resolving this dilemma, the government forged a social democratic corporatist

consensus among labor, capital and government regarding the importance of continuing

the process of integrating South Korea into the global economy, while developing a

comprehensive welfare system (Katzenstein, 1984, Katzenstein, 1985). The creation of

the Tripartite Commission in 1998, which included representatives from the government,

108

labor unions and private industry were charged with the task developing a viable

compromise to economic and social reforms (Song and Hong, 2005, Yang and Moon,

2005). Members of the commission agreed to introduce economic policies that would

facilitate the process of economic liberalization as well as introduce policies that would

establish a comprehensive social safety-net to compensate those who are displaced by the

process of global economic integration (See Table 7.1). Coupled with the political

preference of the liberal government, the financial crisis expedited the pace of economic

liberalization and facilitated the transformation of South Korea’s welfare system (Song

and Hong, 2005). The liberal government, relative to its predecessors, not only

structurally reformed South Korea’s social policy but also expanded the country’s social

welfare programs, as indicated in Table 7.2.

The liberal government extended the coverage of the Employment Insurance

Program. The program was expanded in four stages. First, to companies with ten or more

workers in January 1998; then to companies with more than five workers in march 1998;

and then to all the companies in October 1998; and finally to all the employees including

temporary workers, part-time workers, and day workers in September 2000 (Gazier and

Herrera, 2000, Yang and Moon, 2005). The duration of unemployment benefit was also

extended from 60 to 120 days (Kwon, 2007).

In 1999, the government significantly expanded the National Pension Program to

cover the entire population. Farmers, the self-employed and temporary workers were

excluded from the program in the past because it was difficult to ascertain their income.

The reform of the National Pension Program made it possible to determine their income

by calculating the amount that they would contribute to the program based on the value of

their homes and cars. (Song and Hong, 2005, Yang and Moon, 2005).

Coverage under the Industrial Accident Insurance Program was also expanded. In

the early 1960s, the injury insurance program covered companies with more than 500

employees. Before the reform, the insurance covered 7.5 million employees in companies

with more than five workers. Small companies were excluded from the program.

However, as a result of the reform, the program covered an additional 1.6 million

employees in small companies with four or fewer workers in 2000. At present, enrollment

in the program is compulsory for any company with at least one employee. (Yang and

Moon, 2005).

109

A comprehensive National Health Insurance program did not exist in South Korea.

Different insurance programs administered health insurance. The Industrial Workers’

Health Insurance covered 38 percent of the population, the Public Employee Health

Insurance covered 11 percent of the population, and the Regional Health Insurance

Program covered 51 percent of the population (Song and Hong, 2005). However, since

having separate programs administer health insurance proved to be financially inefficient,

the government successfully integrated these programs into a National Health Insurance

program with universal coverage (Yang and Moon, 2005).

In addition to reforming the various social insurance programs, the government

redesigned the public assistance program and expanded its coverage. With the passage of

the National Basic Livelihood Security Law in 1998, the government introduced a

generous and comprehensive public assistance program because the existing program

only covered people who did not have family members of working age. Consequently,

many people in poverty were not eligible to receive the benefits. The new program

allowed for benefits to be extended to the poor without regard to any of the eligibility

requirements that governed the old program. In addition, the new public assistance

program provided cash benefits to the poor and the disabled (Kim, 2007, Kwon, 2007).

7.9. Limitations and Future Challenges of the Welfare State

The South Korean welfare system was fundamentally transformed from a system

with limited benefits to one where benefits have increasingly become universal. For

example, during 1997 to 2002, social welfare spending increased by 78 percent, health

care expenditures increased by 120 percent, education expenditures increased by 60

percent, and social protection expenditures increased by 49 percent (United Nations,

2007). Thus, the South Korean case lends support to the argument that states’ welfare

policy is the outcome of the conditional relationship between economic globalization and

the process of political democratization.

However, despite its transformation, South Korea’s welfare policy is not without

limitations. For example, the unemployment insurance program does not cover school

and college graduates because it is limited to the workers who have been employed for a

minimum of six months. The National Pension Program is a voluntary rather than a

mandatory system. In addition, by making the program voluntary the government

110

intended to appease the self-employed who saw their contributions to the program as a

type of quasi-tax. To date, approximately half of these workers have not registered for the

program (Kim, 2000, pp. 10-11, Lee, 2001, p. 6, Yang and Moon, 2005).

South Korea’s aging population is a major challenge to the future financial

sustainability of the welfare state. Demographic aging will increase the demand for social

protection benefits. While demographic aging is also occurring in other OECD countries,

the speed and scale of aging is faster in South Korea than other OECD countries. The

speed of demographic aging in South Korea is also faster than in Japan where for the last

two decades welfare policies have been adjusted to serve the social needs of the elderly.

South Korea’s demographic aging is also faster than in France and the United States (see

Table 7.3). Given this phenomenon, the number of pensioners in the future will increase

significantly and over burden the pension fund. Similarly, government health care

expenditures will increase as the number of elderly utilizing health care services increases

and the quality of medical care that they demand will also become increasingly expensive

(Kwon, 2007, pp. 8-10).

The welfare reforms that were initiated by the liberal government during the late 1990s were

based on the principle that South Korea’s welfare policies should be socially inclusive,

economically developmental and democratically shaped. However, the demographic aging of

the South Korean society poses the greatest challenge to the future financial sustainability of a

welfare system that is based on these principles. Will the pressures of demographic aging

coupled with the pressures of economic globalization unravel a welfare system that is socially

inclusive and once again place greater emphasis on the developmental importance of social

policy? This question highlights the continuing complexity in studying the dynamics of the

welfare state in South Korea.

111

7.10. Tables

Table 7.1: Key Contents of the Tripartite Commission’s “Social Compromise to

Management transparency

and corporate restructuring

Enhancing labor market

flexibility

Policies to promote

employment stability and

combat unemployment

Enhancing labor rights

Extension and consolidation of

the social security system

Overcome the Economic Crisis”

Improvement of the corporate financial structure

More responsible and more transparent corporate governance

Promotion of business competitiveness

Permission for employers to dismiss workers in cases of managerial

need

Permission for the establishment of temporary work agencies

Expansion and improvement of employment insurance

Livelihood support for the unemployed

Expansion and improvement of the public employment service

Expansion of vocational training

Job creation through public works and business start-up subsidies

Consultation and re-hiring requirements in case of redundancy

dismissals

Permission for public servants to form workplace associations

Permission for teachers to join trade unions

Permission for trade unions to engage in political activities Right of

dismissed and unemployed workers to join trade unions

Integration of social partners in social security steering

committees

Wage guarantee in bankruptcy cases

Extension of social insurance coverage to non-regular workers

Source: Organization for Economic Cooperation and Development (2000:49) in Yang and Moon (2005:

81)

Table 7.2: Globalization and Social Spending

Year

Welfare Expenditure (% of GDP)

Social

Insurance

Welfare

Service

Social

Expenditure

112

Chun Doo-hwan Military Government

1980

0.50

--

1981 0.50 -- --

1982 0.50 -- --

1983 0.50 -- --

1984 0.80 0.1 --

1985 0.20 0.1 --

1986 0.20 0.1 2.6

1987 0.20 0.1 2.6

Roh Tae-woo Transitional Government

1988

0.20

0.1

3.2

1989 0.30 0.1 3.6

1990 0.80 0.1 3.9

1991 0.90 0.1 3.8

1992 1.00 0.1 4.2

Kim Young-sam Democratic

Government

1993

0.90

0.2

4.4

1994 0.90 0.2 4.7

1995 0.80 0.2 5.1

1996 0.90 0.2 5.3

1997 0.90 0.2 6.8

Kim Dae-jung Democratic Government

1998

1.02

0.2

6.9

1999 1.19 0.4 7.2

2000 1.60 0.5 10.2

2001 2.00 0.7 n/a

2002 2.00 0.6 n/a

Source: National Statistical Office in Song and Hong (2005:183)

Table 7.3: The Speed of Demographic Aging

Years that the rate of

elderly population reach to

Number of years

when elderly

Number of years when

elderly increase from

113

increase from 7% to

14% of the

population

14% to 20% of the

population

7% 14% 20%

Korea

2000

2019

2026

19

7

Japan

1970

1994

2006

24

12

France 1864 1979 2020 115 41

USA 1942 2013 2028 71 15

Source: National Statistical Office (2001) in H.J. Kwon (2007)

Chapter 8

114

Globalization and the Welfare State in Chile

8.1. Introduction

The welfare state in Chile was transformed from a system with comprehensive benefits

that was heavily funded by government revenues into one where benefits were significantly

limited and regulated by market forces. Welfare policy has recently evolved into a quasi-

comprehensive system in which state subsidies and benefits have been expanded but guided by

the logic of the private market. Like South Korea, these dynamics are largely the result of the

interplay between the variation of Chile’s integration into the global economy and variation in

the nature of its domestic political system, which witnessed a transition from democratic

governance to authoritarianism and then the resumption of democratic practices after sixteen

years of military rule.

This chapter begins with an historical overview and then highlights the ways in which

Chile’s democratic regimes, while pursuing an industrial strategy that featured import-

substitution industry (ISI), trade protection and the nationalization of key industries, developed

a comprehensive welfare state. The chapter then discusses how the economic and political

crisis of the ISI induced welfare state led to the military coup d'état in 1973. The chapter then

examines how the economic stabilization and trade liberalization policies of the military junta

globally integrated the economy and in the process replaced Chile’s traditional welfare state

with one that featured significantly limited benefits that were regulated by the free market. This

is followed by a discussion of the restoration of democracy in Chile and the creation of a quasi-

comprehensive welfare state under conditions of greater global economic integration and the

political dominance of the Concertación regime, which is composed of a ruling coalition of

centreleft political parties. The chapter concludes with a discussion of the future challenges to

Chile’s welfare state.

8.2. Historical Overview

Chile achieved independence from Spain in 1811 and by 1932 established an electoral

democracy that ended in 1973. Throughout much of this period, various governments

attempted to reform Chile’s social and economic system by pursuing ISI, trade protectionism,

expanding the welfare state and statist policies that sought to nationalize key industries. These

115

statist policies were instituted following the election of Eduardo Frei of the Christian

Democratic Party (PDC) in 1964. The government acquired majority ownership of the copper

industry, redistributed land, and expanded access to education. Despite these changes Chile’s

political left pressed for more radical reforms, which in 1970 culminated with the election of

Salvador Allende of the Popular Unity party. The Allende government accelerated the reforms

of the Frei administration by fully nationalizing the copper and telecommunication industries

and expanded land reform and the welfare state. The PDC allied with Chile’s parties on the

right to block the legislative initiatives of Allende’s Popular Unity government. The ideological

gridlock prevented the government from addressing the economic depression. Unemployment

and inflation increased, while international capital flows to Chile plummeted. And as the

economy continued to deteriorate along with the indecisive outcome of the 1973 Legislative

elections, the military intervened on September 11 (Collier and Sater, 1996).

The Chilean military, led by General Augusto Pinochet, deposed the Allende

government in a violent coup and terminated democratic practices and civil liberties and

regarded the organized left as an internal enemy of the state. In 1978, General Pinochet won a

tightly controlled referendum, which institutionalized the junta’s rule. The military regime

implemented a series of neo-liberal economic reforms that liberalized trade and investment,

privatized state holdings in the economy, and dismantled the comprehensive welfare state. In

1980, General Pinochet won another referendum that approved the new Constitution, which

called for a plebiscite in 1988. Chileans were given the opportunity to reelect Pinochet to

another 8-year term or reject him in favor of contested democratic elections. The collapse of

the economy in 1982 sparked a nationwide protest against the military junta, which helped to

galvanize opposition to Pinochet’s reelection among Chile’s political parties. In the ensuing

plebiscite, 55% of the Chilean people rejected 8 more years of military rule and called for

democratic elections in 1989 (Constable and Valenzuela, 1993, Falcoff, 1989).

Two major coalitions of parties emerged to contest the 1989 elections. These included

the center-left Coalition of Parties for Democracy, or Concertación, and the center-right

Democracy and Progress coalition. Patricio Aylwin, a Christian Democrat and the candidate of

the Concertación, won the presidency with 55% of the vote and the Concertación won

majorities in the Chamber of Deputies and among the elected members of the Senate. The

Concertación coalition has governed Chile continuously since the transition to democracy.

Eduardo Frei Ruiz-Tagle was elected president in 1993, followed by Ricardo Lagos in 1999,

116

and recently Michelle Bachelet in 2005. While the Concertación coalition governments

maintained the neo-liberal economic policies of the Pinochet regime, they have also

implemented social programs, although at much reduced levels than the Allende era, to reduce

poverty and expand access to education and health care (Rector, 2005).

8.3. Import-Substitution and the Welfare State

From the 1930’s to the mid-1970s, Chile’s import substitution model of industrial

development, which was promoted by the United Nations Economic Commission for Latin

America (ECLA) and at that time was adopted by governments throughout Latin America,

involved the use of discriminatory tariffs, exchange rate controls and tax policies to help

establish national industries and protect them from overseas competition. During this period,

Chile’s ISI strategy was intimately connected with its Bismarkian social interventions to

provide a comprehensive welfare state.

The logic of ISI was that since developing countries faced a declining terms of trade

with advanced industrial countries and the infancy of their industries placed them at a

competitive disadvantage vis-à-vis Western transnational firms, Latin American countries

should reduce their dependence on the global economy by encouraging the local production of

industrial goods. Protectionist trade policies were, therefore, used to help strengthen national

firms to the point where they could compete with foreign producers (Prebisch, 1959, Prebisch,

1950).

Chile’s import-competing industries supported trade protection since such policies

limited foreign competition and produce economic rents for national firms. Since 1838, the

Sociedad National de Agricultura (SNA), which represented the interests of Chile’s landowning

agricultural producers, secured legislation that imposed tariffs on agricultural imports. And by

1897, the political clout of the Sociedad de Fomento Fabril (SFF), that represented the interests

of Chile’s industrial manufacturers was instrumental to the passage of Law 980, which raised

tariffs on imported textiles and other manufacturing goods (Edwards and Lederman, 1998, pp.

31-32). Labor unions supported trade protection since it stabilized the national labor market by

allowing domestic firms to provide full employment, which in turn strengthens the labor

union’s collective bargaining power (Barrera and Valenzuela, 1986). At the end of the Allende

administration, Chile’s import tariffs averaged 105 percent and were highly dispersed across a

range of imported products where some products were subjected to nominal tariffs of more

117

than 700 percent. Moreover, trade was severely constrained by a battery of quantitative

restrictions. Some of these restrictions included import prohibitions of certain products and the

maintenance of a multiple exchange rate system, which further discouraged imports (Edwards

and Lederman, 1998, p. 3).

The ISI strategy also involved the nationalization of industries, especially foreign

companies that the state considered crucial to national development. Protectionist trade policies

were re-enforced by the successive nationalization policies of the Christian Democratic and

Socialist governments of Edurdo Frei and Salvador Allende. In his stateof-the-nation address in

May 21, 1969, President Frei unveiled plans for the “Chileanization” or part ownership of

Chile’s copper industry, which culminated in an agreement with Anaconda – the American

copper company – for the “negotiated and progressive” nationalization of the company’s major

mines throughout the country. A similar agreement was also negotiated with the American

owned Kennecott Copper Corporation that allowed the government’s share of the companies’

profits to increase from 72.6 percent to 91.8 percent (Sobel, 1974, p. 22-24). In explaining his

economic policies in the months prior to his Presidential inauguration, Allende noted:

“We must recover our basic resources that are in the hands of foreign capital, especially

American – copper, iron ore, nitrates, which are in your hands, the hands of American

monopolies. Then we must nationalize the monopolies that influence the social and economic

development of the country. To this we must add a serious wide-ranging profound agrarian

reform, the nationalization of banking and state control over foreign trade” (Sobel, 1974, p.

33).

And in the first year of Allende’s presidency, the Chilean Congress unanimously approved a

constitutional amendment that authorized the president to fully nationalize the copper

companies of Anaconda, Kennecott and Cerro. All 158 senators and deputies who represented

every major political party in Chile voted in favor of the reform. Although 60 percent of

Chile’s private banking system had already been nationalized by previous administrations, the

Allende government accelerated this trend by nationalizing the subsidiaries of the Bank of

America, the Bank of London and Citibank of New York. The government also nationalized

the U.S-owned International Telephone & Telegraph Corporation (ITT), RCA Victor - the U.S-

owned electronic corporation and the assembly plants of General Motors Corporation (Sobel,

1974, pp. 50-51; pp. 56-57; p. 79). In 1973, at the end of the Allende administration, the state

118

controlled 80 percent of Chile’s industrial production, 400 corporate enterprises and

approximately 60 percent of the country’s GNP (Collier and Sater, 1996, p. 342).

ISI was also an integral component of Chile’s welfare state. The Allende government’s

expanded control over the country’s industrial output was correlated with a massive increase in

social spending. During the period of Chile’s largest nationalization initiatives, social spending

on health care, education, housing and social assistance, for each year form 1970-1973, more

than doubled the annual average of the previous four years (See Table. 8.1).

In the 1940’s, White Collar workers from the financial and manufacturing industries

established their own health care system – the National Medical Services for Employees

(SERMENA) that established a private provider system with private doctors.

A limited public health care system was established in 1952, with the creation of the National

Health Care Service (SNS), which only provided prenatal care and family planning. However,

by 1971, the Allende government unified the health care systems into the Servicio Unico de

Salud – a national health care system. The state built a network of hospitals and clinics in

working class neighborhoods on the outskirts of major cities and provided care for every

44,000 inhabitants. The government also established the National Milk Plan that provided

3,470,000 persons with a ½ liter of milk a day to reduce malnutrition, which affected 50% of

Chile’ children in 1970. Spending on education increased primary and secondary school

enrollments. Primary school enrollment increased from an annual average of 34% during the

period 1960-70 to 65% in 1971 to 1972. Secondary school enrollment also grew at a rate of

18.2% during 1971-72 (Isreal,

1989, pp. 35-36). While Chile’s state-run Social Security system – the Servicio de Seguro

Social (SSS) was designed to serve the needs of miners, urban blue-collar and service workers,

the Allende government extended coverage to informal workers and peasants and effectively

transformed the social security system into a universal entitlement system (Hudson, 1994,

Illanes and Riesco, 2007).

The nationalization initiatives of the state were also used to provide the working classes

with employment and salary increases as well as empower the labor unions. In 1970, the

Allende government signed an agreement with one of the country’s largest unions, the Central

Workers Union, which provided for the participation of workers in the planning and the

administration of state-owned and mixed corporate enterprises, the reduction of unemployment

119

by the provision of 180, 000 jobs and the creation of a Central Committee on Wages and

Salaries to formulate new wage and salary policies. In negotiations with the government in

1971 for a new contract, Workers at the Chuquicamata copper mines were given a 21.8% pay

increase plus a commitment by the Allende government to spend $3.7 million dollars during

1972 on social programs for the town of Chuquicamata (Sobel, 1974, p. 38; p. 71).

8.4. The Economic and Political Crisis of the Welfare State

By the early 1970’s, there was growing evidence that Chile’s ISI strategy and the

welfare state that it supported was financially unsustainable. Social spending per person,

between 1920 and 1970, increased by 38%, while GNP per capita increased by only 2.3%

(Arellano, 1985, p. 414). Trade protectionism provided no incentive for domestic firms to

become efficient, but merely encouraged rent seeking behavior. In addition, the small size of

the domestic market never allowed domestic firms to reach economies of scale to be able to

compete in export markets. Consequently, overall industrial production fell even among the

nationalized industries. And the state’s social policies that increased employment and the

purchasing power of the working class coupled with the decline in industrial production gave

rise to black market activities, shortages and inflationary pressures, which surpassed 200

percent by 1972. In addition, the falling prices for copper (Chile’s main export) and the tripling

price for imported oil plunged the balance of payments into deficit, which forced the

devaluation of the official exchange rate from 48 Chilean escudos to the U.S. dollar to 85

escudos to the dollar. And on the black market the currency fell from 150 to 300 escudos to the

dollar (Collier and Sater, 1996, pp. 345346, Hudson, 1994, Isreal, 1989, Sobel, 1974, p. 85).

The crisis in the macro economy led to a significant decline in tax revenues, which

severely eroded the state’s ability to finance its increasingly costly social programs. As a result

of the Allende government’s expansion of the social security system, there were 35 different

pension funds and more than 150 social security programs for different occupational groups.

Since the newly incorporated participants of the program secured by law new benefits that were

denied to original participants (even when the programs of the new participants were added to

existing pension funds), the social security system became inherently unequal and was in

financial crisis by 1972. The state’s support for the national health care program also became

financially unsustainable given the fact that it was tied to the social security system (Hudson,

1994, Mesa-Lago, 1989).

120

Chile’s economic crisis triggered a political crisis within Allende’s coalition

government. The Socialist being the largest party within the coalition continued to press

Allende to accelerate the economic and social policies in Chile’s transition to socialism. The

PDC members of the coalition gradually shifted to the right and became the most important

opposition to the Allende government and they actively promoted legislation that attempted to

reverse the expansive growth of the state’s acquisition of private capital. Given the fact that the

Allende government expropriated the holding of American corporations, opposition also came

from the U.S. government. The Nixon administration pursued a two-track policy towards

Allende’s government. Overtly, the administration cut U.S. aid to Chile and blocked Chile’s

request for financial assistance from the World Bank at a time when Allende’s government

needed it the most. Covertly, the Nixon administration worked to help destabilize the

government by increasing aid to Chile’s military and opposition political groups. With the

growing economic crisis and political polarization within the Allende government, the military

junta staged a coup d’état on September 11, 1973 (Isreal, 1989, pp. 263-269, Sobel, 1974, pp.

90-91).

8.5. Economic Liberalization and the Retrenched Welfare State

In formulating an economic strategy to arrest the economic crisis the junta relied

heavily on the advise of the Chicago Boys, a group of Chilean neo-liberal economists who

were trained at the University of Chicago under Milton Friedman and Arnold Harberger. The

economic policies that were advocated by the Chicago Boys ended decades of ISI development

in Chile and the comprehensive welfare state that it induced. The junta’s Minister of Finance,

in October 1973, stated that Chile’s “best prospects for growth are in the opening to

international competition” (Méndez, 1979, p. 63-64). The junta’s economic policy was based

on three objectives: the stabilization of the economy; the liberalization of trade; and the

privatization of state holdings (Laban and Larrain, 1995, p. 116, Silva, 1993).

The stabilization policy involved a two-part strategy. Often referred to as ‘shock

treatment’, the first part of the stabilization policy sought to eliminate inflationary pressures by

cutting the fiscal deficit by 25% within the first six months of the military dictatorship. The

reduction of the fiscal deficit involved the retrenchment of public sector jobs and across the

board cuts in the social programs of the welfare state (Friedman, 1975). Second, the

121

stabilization policy also concentrated on arresting the balance of payments crisis by securing

external financing from international creditors. By January 1974, the International Monetary

Fund (IMF) approved a stand-by arrangement that allowed Chile to borrow U.S$94.8 million

over 12 months to overcome the foreign exchange deficit of its balance of payments. The Inter-

American Development Bank, in April of that year, approved a U.S$73.3 million loan to Chile,

which was denied to the Allende government in 1972. In addition, a syndicate of international

commercial banks that included Bankers Trust, Irving Trust, and First National City Bank and

several Canadian banks opened a U.S$170 million line of credit to Chile. In addition, just two

months after the coup d’état, Manufactures Hanover Trust Co. of New York extended a U.S$44

million loan to Chile’s Central Bank and a combination of other American and Canadian

commercial banks offered Chile additional loans that totaled U.S$150 million. For its part, the

U.S. government also extended loans totaling U.S$52 million to finance Chile’s imports of

American corn and wheat (Sobel, 1974, p. 161; p. 173).

The junta’s liberalization policy dismantled the Allende government’s differentiated

tariff structure of rates between 10% and 35%. The objective was to reduce tariffs to a uniform

rate of 10% by 1979 (Edwards and Lederman, 1998, p. 4). In fact, as shown in Table 8.2, the

junta successfully implemented its trade reforms by significantly reducing tariffs from their

1973 levels to 10% by 1979. However, this pattern was temporarily suspended in 1982-83,

when Chile experienced its worst economic crisis since the 1930s. Table 8.5 shows that during

these years the economy recorded negative growth rates of -10.3% and -3.8% respectively. In

addition to tariff reductions, the junta also gradually reduced and eventually eliminated import

prohibitions and import licenses, which served as non-tariff barriers to trade (see Table 8.3).

While trade liberalization was incongruent with the rent seeking interests of Chile’s import

competing firms, trade reforms found strong political support among export-oriented industries

that were unable to realize economies of scale under ISI. Exporting firms in Chile’s mining and

wine industries gained from trade liberalization and the export promotion policies of the junta.

Since trade liberalization lowered the price of imported inputs of production, the reform also

benefited firms in the construction and transportation industry which participated in several

industrial strikes against the Allende government that temporarily paralyzed the domestic

economy during 1970-73 (Campero, 1991, Edwards and Lederman, 1998, p.

33).

122

Three months after the coup d’état the junta implemented its privatization policy by

announcing that 115 nationalized companies, including 12 that were owned by foreign capital

would be returned to their former owners. The first of these companies were 4 U.S-owned

motion pictures distributors, the U.S-owned General Tire International and Dow Chemical.

Chile’s State Development Corporation announced that another 88 business and industries

would also be returned to their former owners because they were illegally expropriated by

Allende’s Popular Unity government. The junta also signed an agreement with General Motor

Corporation that effectively returned ownership of the nationalized assembly plants back to the

company (Sobel, 1974, p. 162; p. 173). On December 1974, the junta passed a decree that

prohibited the state ownership of commercial banks (Edwards and Lederman, 1998, p. 38).

The junta’s privatization policies, which also liberalized the capital markets, were

reinforced by the repression of labor unions that aligned themselves with the Allende

government and implemented measures that severely weakened the labor movement. On

September 18, 1973, the junta issued a decree that banned the presentation of union demands

and suspended the right of union leaders to use paid working hours to address union issues.

Other decrees were issued that made it easier for private firms to fire workers, including the

firing of workers who lead, in what the junta considered an illegal strike. The junta

implemented measures that suspended previous agreements with unions regarding salary

increases, benefits and other remunerations to their members. The junta’s anti-labor diktat also

suspended unions’ rights for collective bargaining and the automatic adjustment of pensions to

compensate for inflation (Arellano, 1985, p. 415, Barrera and Valenzuela, 1986, p. 235-236).

Consequently, union membership drastically declined from 65% of Chile’s total wage earners

in 1973, the last year of the Allende government, to less than 20% on average for the entire

1980s (Cortázar, 1997, p. 240).

As shown in Table 8.4, the neo-liberal stabilization policies that slashed the fiscal deficit

achieved their objective by drastically reducing inflation from a high of 605.9% in 1973 to

21% in 1989, the last year of military rule in Chile. Controlling inflation and the stabilization

of prices allowed Chile’s industry to achieve greater economies of scale as they benefited from

the privatization of capital markets and trade liberalization, which effectively integrated Chile

into the global economy. As a result, Chile’s annual growth rates during the years of military

rule, with the exception of 1982-83, exceeded the growth rates of the previous democratic

123

governments as well as its Latin American neighbors and established the economic conditions

that have sustained growth into the transitional years under democratic rule (see Table 8.5).

The success of the junta’s neo-liberal policies in resurrecting the economy came at the expense

of Chile’s welfare state and ended five decades of continuous social spending that was financed

by public revenues. As a consequence of the junta’s stabilization policies to reduce inflation,

social spending as a proportion of GNP was reduced from 20% in the second half of the 1960s

to 14% by the start of the 1980s (See Table 8.6). Combined with the junta’s privatization

policies, the allocation of education, housing and social security was largely determined by

market forces with increased participation by the private sector.

In terms of education reform, the junta drastically reduced expenditures on public

schools and placed the burden of administering and supporting the education system on local

municipalities and parents. Between 1980 to1990, government spending on education was cut

by 27%. Schools that were located in affluent municipalities were able to compensate for the

shortfall in government funding but schools in poor districts had no alternative source of

funding (Delannoy, 2000). Instead, in 1981, the junta introduced a nationwide school voucher

program that gave parents the choice between sending their children to private or public

schools. The program created a dynamic market for education with more than a thousand

private schools entering the market for profit, which increased private enrollment rate from

20% to 40% by 1988 and exceeded 50% in Chile’s urban areas. However, the market forces

that were unleashed by the voucher program did not improve education outcomes. Results from

the international test scores in math and science (TIMSS) in which Chile participated in 1970

and 1999 showed that Chile’s ranking, relative to the other 12 countries that participated in

those years, worsened (Hsieh and Urquiola, 2002). Higher education was also market driven.

Under the Allende government university tuition had traditionally been free. However, with the

reform of higher education students were required to pay an enrollment fee. In addition, the

level of state’s funding to the universities was based on the proportion of students who entered

universities with the highest scores on the national aptitude test. Universities were forced to

compete for state funding in their effort to recruit the most qualified students. The reform of

higher education significantly reduced access to students from working class families who

could not afford university fees or whose test scores were not competitive to be recruited by

Chile’s financially strapped universities (Arellano, 1985, pp. 114-116).

124

The Allende government established, Servicio Unico de Salud, a national health care

system by merging SERMENA – the private health care system for white-collar workers with

SNS – the public system for blue-collar workers. The junta’s reform of health care gave

participants the option of switching to private health care institutions. Eight percent of the

participants that switched took with them 40% of the revenues of the public health care. While

privatization increased the quality of health care for the affluent that switched, the access to

quality health care service among the remaining low-income groups suffered from a sharp

decline in state funding (Collier and Sater, 1996, pp. 373374, Laban and Larrain, 1995, 123).

From the 1950s, the state played a major role in Chile’s low-cost housing development

and built 60% of the houses between 1960 and 1972. The junta drastically slashed public

spending on housing to less than half of its 1970 levels, which increased Chile’s housing

deficit. In addition, the junta also reduced subsidies on housing loans and increased the

participation of the private sector in the development of new homes and municipal buildings.

Housing was also allocated to income groups that met certain savings goals, which effectively

reduced poor families’ access to housing since they could not meet the junta’s savings criteria

(Hudson, 1994).

The military junta closed the previously unfunded pay-as-you-go state run pension

system in which benefits were paid directly from taxes and social security contributions. The

pay-as-you-go system was replaced by funded personal retirement accounts that were

administered by private Pension Fund Managing Corporations, often referred to as

Administradores de Fondos de Pensiones (AFPs) (Soto, 2007). Initially, the state supervised the

pension system and required that assets were deposited in the government debt or bank

deposits. However, by the mid to late 1980s, pension fund assets have increasingly been

deposited in private domestic securities and mutual funds (Borzutzky, 2003, DuPont, 1996,

Kritzer, 1996, Mitchell, 1998). The reform of the pension system removed the share of the

employer’s contribution to the pension system by terminating the social security tax and

established a mandatory contribution rate of 10% of employee’s monthly wages (Borzutzky,

2001, Borzutzky, 2003, Castiglioni, 2001). The privatization of the pension system also

provided for two separate poverty safety nets. First, workers who contributed to their personal

accounts for at least 20 years and who did not accumulate sufficient funds in their accounts to

maintain a minimum standard of living upon retirement were entitled to a government-financed

125

“top up” benefit called the minimum pension guarantee. Second, elders who either fail to

contribute for 20 years or were never in the system at all were entitled to a PASIS, which was a

small noncontributory means-tested benefit that was worth half the minimum pension

(Borzutzky, 2003, Kritzer, 1996, Mitchell, 1998).

While Chile’s privatized pension system has been hailed as a success and a model for

pension reform, the new system was limited in its coverage. Before the junta’s privatization,

the coverage of the public pension system accounted for 72 percent of the population, while the

private pension system covered just 60 percent by 2000. The problem of coverage is rooted in

the structure of Chile’s labor market, which is characterized by high levels of self-employment

and an informal labor force. Among the self-employed, who account for 28 percent of Chile’s

labor force, only 4 percent were engaged in the new pension system in 2000. Approximately,

1.5 million self-employed workers were not enrolled in the system (Borzutzky, 2003,

Mideplan, 2000). The 10% mandatory employee contribution coupled with the nature of

Chile’s labor market reduces the incentive for low-wage workers to contribute to the privatized

pension system. As a result, only a minority from Chile’s labor force makes regular

contributions to the pension system, which reduces the likelihood that participants will

accumulate sufficient funds in their personal accounts to maintain a minimum standard of

living upon retirement. In 2006, projections that were based on the history of worker

contributions demonstrated that a large share of the pension system’s participants would indeed

face financial hardship upon retirement. And 45% of participants would also have pensions

below the minimum pension guarantee threshold and would not have met the level of

contribution required to qualify for the subsidized government benefit (Solange et al., 2006).

An additional problem with the privatized pension reforms is that the high administrative costs,

which reduce retirement benefits, also had a negative effect on the level of participation in the

system. Administrative cost that is paid to pension fund managers reduce the rate of benefits

from 12.7 percent before the reforms to 7.4 percent after their implementation (Kay, 1997).

8.6. Democratic Transition and the Welfare State: Continuity and Change

In the 1988 plebiscite the Chilean people voted to reject 8 more years of military rule in favor

of democratic elections in 1989 that brought the Concertación coalition government, led by

Patricio Aylwin, to office. The new government continued the neoliberalization economic

126

policies of the junta and went even further in the process of integrating Chile into the global

economy. Immediately upon taking office, the Aylwin government reduced the uniform tariff

from 15% to 11% in 1991. Unilaterally reduced, even further, the uniform tariff from 11% to

6% from 1999 to 2003 (Bravo-Ortega, 2006,

p. 11; p. 13). And from 1997 to 2009, various Concertación administrations have also extended

trade liberalization by negotiating a series of bilateral and multilateral free trade agreements.

Currently, Chile has signed 13 Free Trade Agreements, two Free Trade association agreements

with MERCOSUR and the European Union and five Economic Complementation Agreements

(See Table 8.7).

However, while continuing and in some respects deepening the free market policies

inherited from the military junta, the Aylwin government also implemented changes in Chile’s

social welfare policies. In terms of education the government expanded spending for

elementary, secondary and post-secondary schools. Table 8.8 shows that public expenditure on

education increased from U.S $940.3 million in 1990 to U.S $3017.7 million in 2001.

Moreover, pubic spending per student in primary, secondary and post-secondary schools almost

tripled the amounts spent from 1990 to 2002. Consequently, relative to the years under military

rule, working class children have greater access to public education. The dropout rate among

children from low income families have been reduced from 4%, for the first half of the 1990s,

to 2% in 1997 (Cox, 2004, p. 5). There was also a marked improvement in the learning

performance among schools with different systems (municipal schools or government

subsidized private schools) and these improvements were not biased in favor of private

schools. The improvement in learning performance, access and retention rates among low-

income children were also a result of the expansion and improved forms of social assistance.

The expansion in government funded social assistance included food, health care,

schoolmaterials supply and grant programs. The main support for primary education came in

the form of school meals and health care. In 2003, 867,589 primary students received school

lunches, while another 100,415 received eyeglasses and hearing aids (Cox, 2004, p. 7; pp. 12-

17).

In terms of housing, the government continued the practices of the junta by allowing

private sector participation in the construction of new homes but increased public spending on

housing by 50%. The government also changed the eligibility requirements for public housing

programs to benefit low-income families and provided subsidies to poor neighborhoods to fund

127

utilities. And while the government maintained the structure of privatized heath care, it

increased funding for the public health care portions of the system that largely served the poor,

especially primary care services. The government increased the salaries of health workers in

the public sector and gave more authority to local and regional governments over the

distribution of equipment and health-care resources and provisions (Hudson, 1994).

The privatized structure of Chile’s pension system was maintained by various

Concertación administrations. The Aylwin government, however, increased the minimum

pension that was paid of what remained of the state-run pay-as-you-go system by 30%

(Hudson, 1994). In 2006, the Concertación government led by Michelle Bachelet moved to

reform the private pension system to address the problems of coverage and the low

participation among low income and the self-employed members of the labor force. Referred to

as the “reform of the reform” became law in 2008 and sought to strengthen Chile’s private

pension system. The latest reforms created a new and more generous noncontributory

“solidarity pension” that replaces both the means-tested PASIS benefit and the minimum

pension guarantee that was established by the military junta. In 2012 when the reform is fully

phased in, elders with family incomes of less than 60% of the national average will be eligible

for a full solidarity pension provided that they have no contributory pension benefit. Under the

pension system that was created by the junta, low-income workers had little incentive to

contribute once they qualified for the minimum pension guarantee. Under the new system,

every additional income that is contributed to the personal accounts will earn an extra return.

The reform includes other measures that are designed to increase participation in the system.

The reforms make participation by the self-employed mandatory. This requirement will be

phased in over a seven-year period. The reforms also seek to boost the participation of young

low-income workers by paying subsidies to their employers who offer them formal-sector jobs.

The personal retirement accounts for women will also be supplemented to compensate for time

spent as non-contributors, while providing child care at home. In addition, the new reforms

include measures that will reduce the administrative fees that are charged by pension fund

managers plus measures that will improve competition among AFPs

(Jackson et al., 2009, pp. 32-33) .

8.7. Future Challenges of the Welfare State

There is tension between the Concertación government’s social welfare initiatives and its

commitment to deepen Chile’s integration into the global economy. This tension will become

128

more pronounced as Chile’s aging population increasingly demand greater outlays in social

assistance. In addition, since Chile’s economy is deeply integrated into the global market, price

fluctuations for copper and a prolonged recession in global capital markets could undermine

the financial viability of public expenditures in education, housing and health care as well as

threaten pension funds that are increasingly invested in overseas capital markets (OECD, 1998,

pp. 17-18). While the Concertación government, in the context of maintaining its domestic

economic neo-liberal agenda, has benefited from the sustained growth of the global economy

over the last two decades to fund Chile’s social programs, the recent collapse of global

financial system and the subsequent recession in major industrial economies will most likely

force the government to retrench the quasi nature of the welfare state.

129

8.8. Tables

Table 8.1: Social Spending During the Allende Government (millions U.S. dollar)

1965-69* 1970 1971 1972 1973

Health 139.4 154.2 211.6 247.8 237.2

Education 281.9 362.0 473.2 524.2 354.9 Housing

133.7 108.6 229.0 228.3 220.9 Child Assistance 0.3

0.7 0.6 0.8 0.7

Social Assistance 6.9 7.8 8.4 10.6 5.3 Social

Subsistence 0.8 1.9 1.5 0.8 0.3 TOTAL

562.8 635.2 924.2 1012.6 828.5 % of total Expenditure+

32.2 28.9 33.5 34.3 21.6

*A

verage

+Excluding debt service expenses

Source: World Bank, Chile, An Economy in Transition (Washington, D.C., 1980), p. 165.

130

Table 8.2: Itinerary of Import Tariff Reductions, 1973-1991

Date Maximum % of Items Tariff % of Average

Tariff Subject to Tariff Mode Items Tariff

December 31, 1973 220 8.0 90 12.4 94.0

March 1, 1974 200 8.2 80 12.3 90.0

March 27, 1974 160 17.1 70 13.0 80.0

June 5, 1974 149 14.4 60 13.0 67.0

January 16, 1975 120 8.2 55 13.0 52.0

August 13, 1975 90 1.6 40 20.3 44.0

February 9, 1976 80 0.5 35 24.0 38.0

June 7, 1976 65 0.5 30 21.2 33.0

December 23, 1976 65 0.5 20 26.2 27.0

January 8, 1977 55 0.5 20 24.7 24.0

May 2, 1977 45 0.6 20 25.8 22.4

August 29, 1977 35 1.6 20 26.3 19.8

December 3, 1977 25 22.9 15 37.0 15.7

June 1, 1978 25 21.6 10 51.6 13.9

June 1, 1979 10 99.5 10 99.5 10.1

March 23, 1983 20 99.5 20 99.5 20.0

February 9, 1984 35 99.5 35 99.5 35.0

March 1, 1985 30 99.5 30 99.5 30.0

June 29, 1985 20 99.5 20 99.5 22.0

January 5, 1988 15 99.5 15 99.5 15.0

June 1, 1991 11 99.5 11 99.5 11.0

Source: Ffrench-Davis, Ricardo. (1987) Import Liberalization: The Chilean Experience, 1973-

1982. In Military Rule in Chile: Dictatorship and Oppositions, edited by J.S Valenzuela and A.

Valenzuela. Baltimore and London: The Johns Hopkins University Press.

Table 8.3: The Elimination of Non-Tariff Barriers (NTB's)

Instruments The Situation in 1973: Dates when Relaxed

The Number of Product and Eliminated

Categories Affected

Import Prohibitions 187 August 1976: Down to 6 products

April 1978: Down to 5 products

August 1981: All eliminated

Prior Deposits* 2,872 January 1974: Wavers granted

August 1976: Eliminated

Import Licenses** 2,278 January 1974: Eliminated Source:

Sebastian Edwards, and Daniel Lederman. (1998) The Political Economy of Unilateral Trade

131

Liberalization: The Case of Chile. In NBER Working Paper Series. Cambridge, MA: National

Bureau of Economic Research, p. 60.

*Required a 90-day non-interest bearing deposit at the Central Bank, which was equivalent to

10% of the value of the imported item. **Government approval was required prior to

importation.

Table 8.4: Chile's Inflation, 1973-89 (% change)

1973 605.9

1974 369.2

1975 343.2

1976 197.9 1977 84.2

1978 37.2

1979 38.0

1980 31.2 1981 9.9

1982 20.7

1983 23.1

1984 23.0

1985 26.4

1986 17.4

1987 21.5

1988 12.7

1989 21.4

Source: Sebastain Edwards and Alejandra Cox Edwards, Monetarism and Liberalism:

The Chilean Experiment (University of Chicago Press, 1991), pp. 28; p. 213)

132

Table 8.5: Chile vs. Latin America: Average Annual Growth Rates (percentages), 1965 – 2008

(Alessandri, Frei, Allende, Pinochet and Concertación Administrations)

Year Chile Latin America

The Alessandri Years

1961 4.8 4.3

1962 4.7 3.5

1963 6.3 7.3

1964 2.2 5.1

The Frei Years

1965 0.8 4.7

1966 11.2 4.5

1967 3.2 6.0 1968

3.6 7.1 1969 3.7

6.8

1970 2.1 6.7

The Allende Years

1971 9.0 6.8

1972 -1.2 7.0

1973 -5.6 8.3

The Pinochet Years

1974 1.0 7.0 1975 -12.9

3.8 1976 3.5 5.4

1977 9.9 4.8 1978

8.2 5.1

1979 8.3 6.5

1980 7.5 5.9 1981

5.3 1.7

1982 -10.3 -1.4

1983 -3.8 -2.5

1984 8.0 3.7

1985 7.1 2.6

1986 5.6 4.1

1987 6.6 3.4

1988 7.3 0.5

1989 10.6 0.9 The

Concertación Years Patricio Aylwin Govt.

1990 3.7 0.3

1991 8.0 4.6 1992 12.3

3.9 1993 7.0 3.7

_______________________________________________________________________ Table

8.5: Chile vs. Latin America: Average Annual Growth Rates (percentages), 1965 –

2008 (Alessandri, Frei, Allende, Pinochet and Concertación Administrations), (continued)

133

Chile Latin America

The Concertación Years

Eduardo Frei Ruiz-Tagle Govt.

Year

1994 5.7 4.7

1995 10.6 0.6 1996

7.4 3.5

1997 6.6 5.5

1998 3.2 2.4 1999

-0.8 0.2

Ricardo Lagos Govt.

2000 4.5 3.9

2001 3.4 0.3

2002 2.2 -0.5

2003 3.9 2.2

2004 6.0 6.1

2005 5.6 4.9

Michelle Bachelet Govt.

2006 4.6 5.6

2007 4.7 5.8

2008 3.2 4.4

________________________________________________________________________

______

Source: Data for the years 1961-1981 was adapted from Jorge Rodriguez Grossi, ed,

Perspectivas Economicas Para La Democracia: Balance y lecciones Para La Experiencia

Chilena (Santiago: Instituto Chileno de Estudios Humanisticos, 1984), p. 32. Data for the years

1982-2008 was adapted from the World Bank’s World Development Indicators, 2009.

Table 8.6: Government Social Spending, 1961-1981 (% GDP)

Alessandri Years

1961 17.0

1963 16.3

Frei Years

1965 20.0

1967 20.1

1969 18.7

Allende Years

1970 19.9

1971 25.2

1972 25.8

Pinochet Years

1974 17.6

1975 18.3

1977 17.4

1979 15.4

134

1981 14.3

Source: Adopted from, Arellano, José-Pablo. (1985) Social Policies in Chile: An Historic al

Review. Journal of Latin American Studies 17:397-418.

135

Table 8.7: Chile: Trade Agreements of the Concertación Government

Bilateral Free Free Trade Economic Complementation+

Trade Agreements Associations Agreements

1997 - Canada 1996 - MERCOSUR 1993 - Bolivia

1998 - Mexico 2003 - European Union 1993 - Venezuela

1999 - Central America 1994 - Colombia

2002 - Costa Rica 1995 - Ecuador

2002 - El Salvador 1995 - Peru

2004 - EFTA#

2004 - U.S

2004 - South Korea

2006 - Panama

2006 - China 2007 -

India* 2007 - Japan

2009 - Australia

* Trade agreement comes with safe guards.

#EFTA countries include: Iceland, Liechtenstein, Norway and Switzerland.

Source: Adopted from McGill University’s Preferential Trade Agreement Database, which can

be accessed at http://ptas.mcgill.ca

+Economic Complementation is an agreement to establish a framework for negotiations leading

to the creation of free trade. Information on these agreements were adopted from, Tim Martyn,

Complete Guide to Regional Trade Agreements of the Asia-Pacific. The Australian APEC

Study Center, 2001.

Table 8.8: Public Expenditure on Education by various Concertación Administrations;

Spending per Student, by Level, 1990-2002

Public Expenditure Ministry of education,

on Education spending per student ($US

Million 2001) Each year, by educational level

(US$ Million 2001)

Year Primary Secondary Post-secondary

1990 940.3 231.8 213.8 823.3

1991 1035.5 244.8 216.5 1109.3

1992 1176.4 270.1 270.7 1111.9

1993 1328.5 302.8 296.3 1097.4

1994 1461.3 325.5 324.5 1148.0

1995 1620.2 371.0 396.3 1180.0

1996 1840.6 402.3 441.1 1240.1

1997 2017.8 443.4 494.3 1319.5

1998 2214.7 480.5 546.0 1333.0

136

1999 2412.3 518.7 550.0 1417.1

2000 2617.8 539.5 609.6 1374.0

2001 2788.8 582.8 623.5 1360.5

2002 3017.7 N/A N/A N/A

%Growth in per student spending 151.4 191.7 63.5

Source: Cristián Cox. 2004. Innovation and reform to improve the quality of primary education:

Chile (EFA Education for All Global Monitoring Report 2005, The Quality Imperative), p. 34.

Chapter 9

Globalization and the Welfare State in Spain

9.1. Introduction

From the autocratic regime of Francisco Franco, which pursued an ISI industrial

strategy that featured protectionist trade policies and where social benefits were developmental

in nature and limited to workers in the industrial sector, to the Socialist government (PSOE) of

Manuel Chaves González where social benefits were universal in nature but commensurate

with the government’s neo-liberal economic strategy, the changing nature of Spain’s welfare

state is largely correlated with variations in the interaction between the country’s integration

into the global economy and changes within its domestic political system.

This chapter begins with a brief overview of Spain’s political history and then

highlights the ways in which the ISI industrial strategy of Franco’s dictatorship, unlike Chile’s

137

military regime but similar to South Korea’s military rulers, established a developmental

welfare state with limited social benefits. The chapter then discusses how Franco’s ISI

developmental strategy undermined economic growth and precipitated the crisis of Spain’s

developmental welfare state. The chapter then examines how the Stabilization Plan that was

adopted by Franco’s regime and continued by the transitional government to arrest the

country’s economic crisis gradually liberalized the economy, while facilitating reforms in

Spain’s social welfare policies. This is followed by a discussion of how the restoration of

democracy, especially during the leadership of the PSOE government, deepened Spain’s

integration into the global economy, significantly expanded the country’s social welfare

policies but increasingly tied this expansion to the growth of the market economy. The chapter

concludes with a discussion of how Spain’s aging population challenges the future

sustainability of the country’s social policies.

9.2. Historical Overview

Franco and the military participated in a failed coup d'état against the Popular Front

government, which led to the Spanish civil war from 1936 to 1939. During this conflict, Franco

emerged as the leader of the Nationalists against the Popular Front government. After winning

the civil war with assistance from Hitler’s Germany and Mussolini’s Italy, Franco dissolved the

Spanish Parliament and established an autocratic regime from 1939 until his death in 1975.

And during World War II Franco permitted Spanish volunteers – the Blue Division – to join the

Nazis in the fight against the Soviet Bolsheviks on the Eastern Front as re-payment for

Germany’s support during the civil war. In 1946, the United Nations (UN) imposed diplomatic

sanctions against the Franco regime for its support of the Axis powers and considered using

economic sanctions to promote democratization in Spain (Baklanoff, 1978, Harrison, 1985).

The post-war diplomatic isolation and the threat of economic sanctions strengthened economic

nationalism as the Franco regime’s control of foreign trade and the adoption of an ISI

development strategy induced the establishment of a developmental welfare state (Anderson,

1970).

Before his death, Franco declared that Juan Carlos would be his successor as King. With

this declaration a constitutional monarchy was established in 1977, which oversaw the

formation of a political confederation by several centrist political parties that led to the creation

of the Union of the Democratic Centre (UCD). In the elections of June 15, 1977 the UCD

138

received 34.4% of the vote taking 166 of the 350 seats in the Congress of Deputies. After 38

years of dictatorship, the transitional UCD led government governed in a coalition with rightist

and leftist parties in the Congress. The UCD ruled the country until its defeat to the PSOE in the

1982 general election in which it received only 6.7% of the vote and 11 seats. The party’s

conservative electoral base defected and gave their support to the newly created alliance

between the Democratic Popular Party (PDP) and the Popular Alliance (AP). During its tenure

in office the PSOE implemented a number of neo-liberal economic reforms that set the stage for

Spain’s entrance into the European Economic Community (EEC). The alliance between the PDP

and AP eventually led to the establishment of the Peoples Party (PP) - the principal conservative

political opposition that defeated the PSOE in the 1996 general election.

9.3. Import-Substitution and the Developmental Welfare State

Spain’s post-war diplomatic isolation and the threat of economic sanctions against the

Franco regime reinforced the protectionist tradition that characterized Spain’s trade policy

since the late 1800s (Oliver-Alonso and Valles, 2005, World Bank, 1963). The threat of

economic sanctions gave impetus to economic nationalism, which created the conditions in

which the ISI developmental strategy was adopted (Pittsburgh Post-Gazette, 1947). The Franco

regime established the National Institute of Industry (INI) to give incentives to domestic firms

to manufacture products that would substitute for imports (Anderson, 1970). In addition to high

tariffs, import licenses and quotas were frequently used to strengthen national firms by

restricting competition from foreign importers (Anderson, 1970, Baklanoff, 1978).

Franco’s trade policy was consistent with the logic of ISI development. Technological

inputs used in domestic industry were imported from behind high trade barriers. The regime’s

trade policies consisted of a system of multiple exchange rates, the importation of industrial

inputs, quotas and state-monopolized trade agreements. The system of multiple exchange rates

not only restricted trade but also severely discouraged foreign investment (World Bank, 1963).

Imports were limited to technological inputs such as raw materials, semi-finished products,

machinery and various chemical inputs that were critical for domestic industrial production.

Thirty-eight percent of Spain’s imports between 1960 to 1961 consisted of industrial inputs and

sixty-two percent consisted of quotas and state-monopolized trade (Baklanoff, 1978). The

composition of trade from the 1950s to the 1970s, as shown in Table 9.1, largely reflected the

139

economic autarchy of the Franco regime, which would prove unsustainable with the onset of

the near collapse of the economy by the early to mid-1950s.

Given the developmental nature of Franco’s welfare state, social benefits were limited

to workers in the industrial sectors (Cousins, 1995, Gibbons, 1999). To encourage the worker

productivity, the regime introduced compulsory insurance for retirement and health care for

workers in the industrial sector (See, Table 9.2). In 1942, the regime provided Health Insurance

for low-income industrial workers and their dependants. In 1947, the Old Age and Invalid

Insurance Program was introduced, which provided old age subsidies for disabled and low-

income industrial workers who were respectively over the age of 60 and 65 years. Private

sector firms’ contribution to the program was based on a pay-as-you-go system, while public

sector firms made contributions via the Regimen de Clases Pasivas (RCP). In 1946, health care

benefits covered just 30% of population (Cousins, 1995).

9.4. Economic Crisis, Liberalism and the Reform of the Developmental Welfare State

As a result of the adverse residual effects of the civil war and the failure of Franco’s ISI

development strategy, Spain was far more economically retarded in 1940s than in the previous

decade. By the early 1950s, the ISI model proved inadequate to generate economic growth.

Spain’s per capita GDP was barely 40 percent of the average for Western Europe countries. The

country experienced high levels of inflation, the tripling of consumer prices and negative

growth rates in some years. The ISI-induced economic crisis, like Allende’s Chile, led to the

growth of the black market and rationing (Solsten and Meditz, 1988). Moreover, the

protectionist trade policies of ISI undermined Spain’s technological development and depleted

the country’s foreign exchange reserves.

Such reserves, which were US$58 million in 1958, dropped to US$6 million by mid1959. And

given the growing demands of the emerging middle class for imported food and luxury items

coupled with the regime’s restraints on foreign trade, exports collapsed and the value of the

Spanish peseta fell on the black market, all of which increased the country’s balance of

payments deficits (Oliver-Alonso and Valles, 2005, 184, Solsten and Meditz, 1988).

To pull Spain form the economic crisis, the Franco regime adopted the

recommendations for liberal economic reforms that were advocated by a group of

developmental technocrats that included bankers, industrial executives, some academic

140

economists and members of the semi-secret Roman Catholic lay organization - Opus Dei

(Giner, 1986, Moreno, 2001). An important aspect of these reforms called for an overhaul of

the tax collection system, which increased tax revenues, and coupled with a program of

monetary and fiscal restraints the public sector, by 1958, were no longer experiencing deficits

but reported a surplus. In addition, Spain joined the International Monetary Fund (IMF), the

World Bank, the Organization for Economic Development and Cooperation (OECD) and the

General Agreement on Tariffs and Trade (GATT). These institutions brought about significant

reforms to Spain’s trade policies through a series of initiatives, which included the introduction

of a single exchange rate regime and the liberalization of price controls and trade restrictions

(Solsten and Meditz, 1988). In implementing these reforms, tariff rates (both nominal and

effective) were significantly reduced. The average level of tariffs declined by 21 percent from

18.7 percent in 1961 to 15.5 percent in 1970 (Baklanoff, 1978, Heston et al., 2006).

The liberal economic reforms allowed the Franco regime to avoid the possibility of

suspending its debt payments to foreign banks that held Spanish currency, and by the end of

1959 Spain’s balance payments showed a US$100-million surplus. The reforms, moreover,

facilitated a steep increase in foreign direct investment between 1960 and 1974 that amounted

to US$7.6 billion, which later contributed to the growth of the economy by an impressive 7

percent from 1962 to 1974 (Oliver-Alonso and Valles, 2005, Solsten and Meditz, 1988).

However, the initial effect of these reforms was deflationary and recessionary as they

significantly deteriorated real incomes and raised unemployment in the first years of their

implementation. The rise in unemployment and reduced wages led to the emigration of

approximately 500,000 Spanish workers to other West European countries in search for

employment. In a country where social policies were limited the negative effects of the reforms

on the Spanish working class induced a crisis in the developmental welfare state. In addition,

although the Franco regime outlawed collective bargaining and organized labor unions, the

frequency of strikes among industrial sector workers drastically increased, which in turn

triggered the growth of various social movements that called for greater democratization and an

expansion of the welfare state (Hipsher, 1996, Hooper, 1986, Mangen, 2001, Ramon, 1985,

Solsten and Meditz, 1988).

In response to the growing dissatisfaction among the industrial working class, the

Franco regime sought to modernize Spain’s social policies in an attempt to appease striking

workers as well as to increase the efficiency of existing social programs. There were no

141

attempts to transform Spain’s limited developmental welfare state into one that provided

universal access to social benefits, which was the standard for most countries in Western

Europe (Rodriguez, 1993). As shown in Table 9.2, the regime introduced unemployment and

illness insurance and social security programs that largely targeted workers in Spain’s industrial

sectors. The provision of social security benefits were based on the occupational category of

workers and the level of contributions provided by the affiliated private and public sector firms

(Cousins, 1995, Lieberman, 1982, Rodriguez, 1993). In addition, the regime created temporary

work programs that were designed to provide employment relief from the recessionary effects

of the liberal economic reforms (Lieberman, 1982, Mangen, 2001). Even with the attempts to

modernize social policy, welfare spending in Spain remained at 4 percent of GDP throughout

the 1960s, which was far below the standard of Western European countries (Mangen, 2001).

9.5. The Oil Crisis, the Transitional Centrist and the Developmental Welfare State

The return to democracy in Spain correlated with the quadrupling of imported oil

prices. And given the fact the Spain imported 70% of its energy, the oil price shocks of the

1970s produced an inflationary effect on economic output. However, since the political capital

of the centrist government of Adolfo Suarez Gonzalez was largely spent on drafting the new

constitution and negotiating the terms for the transition to democracy, the government failed to

implement the necessary economic policies that would help the country adjust to the

inflationary pressures of the global economy. Consequently, industrial productivity plummeted,

which was followed by a sharp increase in wages and consumer prices. As the economy

contracted, Spain experienced a reversal in its migration trends due to the economic recession

throughout Western Europe as well as an increased outflow of labor from the rural areas that

sought diminishing job prospects in the cities. Together, these factors contributed to the sharp

increase in the country’s unemployment. The recessionary effects of the oil price shocks and

government inaction resulted in job losses for 1.5 million workers and by 1982 – the last year

of transitional government - unemployment stood at 17 percent (Harrison, 1985, Mangen,

2001, Solsten and Meditz, 1988).

The negotiations between the Suarez government and the political opposition produced the -

Pactos de la Moncloa (The Moncloa Pacts) in October 1977, which was a series of agreements

that defined the process of democratization. These agreements guaranteed a free press, the right

of labor unions to organize, the legalization of political parties, the submission of the military

142

to civilian authority, the establishment of local government through open elections, and

immunity for members of Franco’s regime for actions taken during the dictatorship. The

creation of a new constitution via a referendum in December 1978 formalized Spain’s

transition to democracy (Oliver-Alonso and Valles, 2005).

The Moncloa Pacts also established the basic structure of Spain’s limited welfare state

during the transitional period. In terms of welfare reform, the Suarez government was similar

to the Franco regime in that it sought to improve the efficiency and the quality of social

services under the existing welfare system by restructuring the fragmentary and duplicative

ways in which agencies delivered welfare benefits (Mangen, 2001). But more importantly, the

Moncloa Pacts created a framework that guided the government’s use of social policy to

compensate the industrial working class during the economic crisis. In this regard, the

government expanded unemployment benefits for industrial workers by extending

unemployment entitlements from 12 to 18 months (McMillion, 1981, OECD, 1984). Health

care coverage was also extended to selfemployed professionals and government subsidies were

given to private firms as an incentive to create industrial jobs (Baklanoff, 1978). However, the

Moncloa Pacts between the Suarez government and the political opposition made no provision

for a comprehensive reform of Franco’s developmental welfare state. Such reforms would later

emerge under conditions of greater political democratization and the deepening of Spain’s

economic integration into the European community.

9.6. Neo-Liberal Socialists, European Integration and Comprehensive Welfare Reform

When the PSOE took office in October 1982, the socialist government inherited an

economy where inflation was running at an annual rate of 16%, the trade deficit stood at US$4

billion, the public sector experienced large deficits, and Spain’s foreign exchange reserves were

nearly depleted. But having secured an absolute majority in both houses of the Spanish

parliament, the socialist government combined its agenda for comprehensive social reform

with a neoliberal economic strategy and was able to implement unpopular economic

adjustment measures that brought the economy back into balance (Magone, 2004, Solsten and

Meditz, 1988).

To reduce the public sector deficit, the government brought the debt-ridden social

security system into better balance. The government passed its pension reform legislation in

1985, which involved substantial cuts in benefits and saved US$600 million (Beremo and

143

Gacia-Duran, 1994, p.109). And through a series of privatization initiatives the socialist

government sought to expand the role of the private sector in the economy. These initiatives

included the closing of state-run corporations that were unprofitable and privatizing others that

belonged to the state holding company INT (National Industry Institute). In other instances, the

government downsized state-owned companies in the coal, iron and steel industries. PSOE also

passed legislation that ended the state’s monopoly of the telecommunications industry and

liberalized the energy sector by allowing the pipeline network of Repsol – the state owned Oil

Company – to be used by private competitors. In addition, PSOE slashed corporate taxes by 8

percent to improve the competitiveness and profitability of Spanish firms (Polavieja and

Richards, 2001, Richards, 1999, pp. 167-168). The government also moved quickly to

deregulate the labor market in order to encourage private investment and make the economy

more competitive. The uniform structure of wages in an inflationary economy undermined

corporate investment since the existing rigidity of the labor market restricted the ability of

private firms to reduce their indebtedness (Beremo and Gacia-Duran, 1994, p. 108).

The neoliberal economic strategy of the socialist government also included the

liberalization of trade, which was largely driven by Spain’s entry into the European Economic

Community (EEC) (Heston, et al., 2006). The PSOE signed the Treaty of Accession to the EEC

in 1986, which reduced tariffs to zero for imports coming from member countries of the EEC

(Ferrera, 2005). The “Europeanization” of Spain was finalized when the socialist government

signed the Maastricht Treaty, which established Spain’s membership in the European Monetary

Union (Oliver-Alonso and Valles, 2005).

The neoliberal adjustment measures of the PSOE reduced the budget deficit to 5 percent

in 1985. Inflation was reduced to 4.5% in 1987 and the deregulation of the labor market

contained the growth of real wages below the rate of inflation. Spain's industrial output grew at

a rate of 5.2% in 1987 and its real GDP registered a growth rate of 5.5 percent, which was the

largest rate of expansion among OECD countries during that year (Solsten and Meditz, 1988).

While the stabilization and growth of the economy provided the PSOE with the

financial basis to pursue welfare reform, European integration, which required the

comprehensive expansion of Spain’s social policies to meet European standards, provided the

government with an external political motive to extend universal health, education and social

protection benefits (Guillen and Alvarez, 2001). The passage of the General Health Act

universalized health care services for all citizens and foreign residents in Spain. Health care

144

coverage increased to 99.8 percent of the population in 1991(Almeda and Sarasa, 1996). The

government also expanded unemployment benefits to 50 percent of the unemployed, and social

benefits to primary and secondary public education was expanded to cover students between 4

and 15 years old. At the university level, the provision of government funded scholarships

allowed low-income families to have greater access to higher education (Moreno and Arriba,

1998).

The ‘Europeanization’ of Spain brought the country’s social policies closer to the

European standard (Oliver-Alonso and Valles, 2005). During its tenure in office (19801996),

the socialist government’s welfare reform expanded overall social spending from 25 percent of

GDP in 1975 to 40 percent in 1996. Social protection spending increased from 15.6 percent

GDP in 1980 to 20.7 percent in 1990 (Guillen and Alvarez, 2001). The government’s spending

on pensions also increased from 8.5 percent of GDP in 1982 to 10.5 percent in 1995 (OECD,

1999), and as Table 9.3 shows, aggregate per capita social spending on education, social

security and health care increased from US$1,311 in 1975 to $US2,625 in 1996.

The Spanish case shows that the transformation of the country’s social policies was a

function of the interactive and mutually reinforcing process of the political transformation of

the state and the economic integration into the European community. In response to the

economic crisis of the 1950s and 1970s, Spain abandoned economic nationalism as well as the

developmental nature of its social policies and embraced economic liberalization and

democracy, while constructing a universal welfare state. Notwithstanding the attempts by the

Franco regime to liberalize the economy during the 1960s and the 1970s, further liberalization

was largely a product of Spain’s membership in the EEC, which required not only the

consolidation of democracy but also the comprehensive reform of the welfare state.

9.7. The Future Challenge to the Welfare State

Like most countries in the European Union, Spain’s aging population poses the greatest threat

to the future financial viability of the welfare state, especially the pension system. Given the

fact that the country’s fertility rate is 1.2, which is the lowest in the European community, and

that 20 percent of the population in 2020 will be over sixtyfive years old, Spain faces a future

of extreme demographic aging and soaring retirement costs (European Commission, 1998,

Moreno, 2001). It is estimated that in countries like Spain, Italy, Germany and France the total

burden of pension expenditures (if the current unfunded pay-as-you-go pension system is left

145

unchanged) is expected to rise to approximately 90% of GDP from 2000 to 2050 (Rother et al.,

2004). However, with fewer workers to cover the rising costs of pensions, Spain’s future

pension debt is estimated to be 244 percent of GDP by 2050. This is significantly larger than its

current debt, which is 49.7 percent of GDP that is owed to creditors (Thomas, 2010, p.1). A

major overall of the pension system is vital to the financial sustainability of Spain’s welfare

state. Such reform may entail a combination of various options such as establishing private

pension accounts, increasing the retirement age or increasing taxes. These options, however,

will be politically costly for Spain’s policy makers. However, if governments lack the political

courage to reform how the current pension system is funded, then Spain’s mounting pension

deficits will decrease future governments’ ability to finance social policies because the high

interest payments that are accrued from such debt will crowd out spending priorities on health

care, education and social protection programs.

146

Students also viewed