Globalization and Labor in Four Developing Regions: An Institutional Approach
This paper is the introduction to a SCID special issue on “Global Pressures, National Response, and Labor Rights in Developing Countries.” We focus on the potentially conflicting demands that developing countries face from international institutions for better labor standards versus more labor flexibility. This is studied through a comparative analysis of four regions: Eastern Europe, Latin America, East Asia, and the Middle East. The major international institutions we examine are the International Labor Organization, International Monetary Fund, World Bank, bilateral and plurilateral trade agreements, and multinational corporations. This introductory paper presents a review of existing literature on labor standards and labor flexibility with particular focus on the role of international institutions in promoting the two processes and their impact on labor market outcomes. It also describes our project and its contribution to the debates, including a discussion of our main methodological innovation, namely, the construction of new indices for labor standards and flexibility. In empirical terms, it compares the indices across the four regions and provides an analysis of the impact of the indices on labor market outcomes.
KeywordsLabor standards Flexibility International institutions Developing countries
A major critique of the globalization process that characterizes today’s world is that the benefits are not being shared fairly across countries and social groups. Three main sets of actors have protested about unfair treatment: developing country governments (vis-à-vis the industrial world); workers in industrial countries (vis-à-vis global corporations and unfair competition from low-paid workers in developing countries); and workers in developing countries (vis-à-vis their employers, their governments, and various actors in the industrial countries). Most of the attention to this subject has focused on the first two conflicts. This special issue of SCID proposes to contribute to a better understanding of the third.
The impact of globalization on workers in developing (including post-communist) countries is hard to predict at a theoretical level. While theory can help, a satisfactory answer must rely primarily on empirical research. Before turning to theory or our empirical analysis, however, we first need to clarify what we mean by globalization in this volume. When we speak of globalization, we refer to pressures by global institutions rather than the economic aspects of globalization (e.g., trade, capital flows, and migration). With respect to labor, these pressures are of two main kinds. One is the demand for better “labor standards,” while the other is a demand for greater “labor flexibility.” It would appear that the two policy sets are contradictory in intent and impact, although advocates of both claim that they want to improve conditions for workers. Since discussions of the two sets of policies have generally taken place in different arenas, we do not know which has made more inroads in different countries or why. We also have little understanding of how they interact. Nonetheless, the resultant characteristics of labor markets are potentially of great importance to workers’ incomes and well-being, and thus to societies and economies of developing countries as a whole.
The first set of demands—voiced by the US Congress, major labor federations, the International Labor Organization (ILO), and various non-governmental organizations (NGOs)—is for labor standards in developing countries that are equivalent to those in the industrial world. The four core standards are: a ban on forced labor, elimination of discrimination in the workplace, abolition of child labor, and freedom of association and the right to bargain collectively. The fourth leads to the more commonly heard call for decent wages and a safe workplace. In principle, these demands would have a positive impact on income and working conditions for formal-sector labor in developing countries. Opponents claim they would damage competitiveness and thus hinder the creation of new jobs.
The other demand is made by individual companies, trade associations, developed country governments, and the international financial institutions for greater flexibility of labor markets in developing countries. There is less agreement on the content of this demand, but generally, it refers to rule changes that make it easier to hire and fire workers, reassign tasks and hours, establish pay scales and promotion standards based on performance rather than seniority, and limit the role of trade unions. While proponents believe that greater flexibility would raise competitiveness and stimulate job creation, unions fear that it would eliminate jobs, lower wages, and erode other benefits.
This special issue aims to further our understanding of the impact of global institutions on labor markets by juxtaposing the labor standards and labor flexibility debates at the theoretical level and studying their implementation in developing countries. We begin with a chapter by Mark Anner and Teri Caraway which analyzes the changing role of the main international institutions involved in labor issues in developing countries: the ILO, the International Monetary Fund (IMF), and the World Bank. Bilateral and regional policies of the industrial countries also exert significant influence on labor markets, and multinational corporations may have some impact as well. We posit that these external actors play a key role in shaping labor market characteristics, but we also ask how their role compares to that of domestic political actors in individual countries.
Our empirical evidence is obtained through studies of four middle-income regions: Eastern Europe, including former Soviet countries (by Linda Cook), Latin America and the Caribbean (by Katrina Burgess), North and Southeast Asia (by Teri Caraway), and the Middle East and North Africa (by Melani Cammett and Marsha Posusney). This broad range of regions and countries—with different economic structures, political systems, and initial conditions—gives us the basis for studying how context affects attempts at policy change and implementation. At the same time, by restricting the universe to mainly middle-income countries, we avoid cases that are so different as to make comparison overly complicated.1 Within the four regions, of course, we move to the country level since this is where policy on labor is actually made and implemented.
This introductory paper provides essential background for the more detailed analyses in the papers that follow. Following this introduction, the second section presents a brief review of two themes in the existing literature on labor standards and labor flexibility. The third describes our project and its contribution to the debates, including a discussion of our main methodological innovation, namely, the construction of new indices for labor standards and flexibility. The fourth section compares the indices across the four regions. The fifth provides an analysis of the impact of the indices on labor market outcomes. The final section concludes.
A Brief Literature Review
Both labor standards and labor flexibility are highly controversial topics, and a good deal has been written about them. Some of it has been for polemical purposes, but much has been serious empirical work—despite problems in obtaining relevant data. In recent years, it should be noted, the relative prominence of the two processes has shifted. In the 1980s and 1990s, attention in the academic and policy literature focused on the need for increased flexibility. In the last decade, by contrast, literature on labor standards is more voluminous, but flexibility continues to be an issue of concern. We address two topics with respect to labor standards and flexibility: the role of international institutions in promoting the two processes and their impact on labor market outcomes.
International Influence on Labor Market Characteristics
In the next paper in this volume, Anner and Caraway analyze the role of the three main global institutions involved with labor issues. In principle, the main organization is the ILO, which is charged by the international political system with taking the lead in promoting rights for workers. Trade unions and socially minded academics began an international campaign for labor rights in the late nineteenth century. The ILO itself, however, came into being as a result of a Labor Commission established as part of the Versailles Peace Conference in 1919 (Berg and Kucera 2008). The current thrust of ILO activities dates from the 1998 “Declaration on Fundamental Principles and Rights at Work,” which established the four core labor rights mentioned earlier—freedom of association and collective bargaining, the elimination of forced labor, the abolition of child labor, and an end to discrimination in employment. These were elaborated in eight conventions that governments are invited to ratify. The vast majority of governments have ratified most conventions; the problem comes with enforcement of the rights after ratification takes place. The ILO itself has the ability to “name and shame” through several committees, but it has turned to the IMF and the World Bank—the so-called international financial institutions (IFIs)—to increase its clout.
Contrary to widespread opinion, the IMF and World Bank have become quasi-partners in promoting labor rights. The process of incorporating them was not a simple one, but by the mid-1990s, the ILO and the IFIs were attending each other’s meetings, and joint activities increased as a result of the Asian financial crisis. Clinton administration officials and the US Congress helped bring them together (Hagen 2003). While the IMF was first to join forces with the ILO, the World Bank has been more active recently. Polaski (2007) reports three positive results that have emerged from the World Bank. First, the Bank’s private finance arm, the International Finance Corporation (IFC), adopted new performance standards that require borrowers to comply with core labor standards and other workplace practices. Second, the World Bank itself has revised its procurement standards for contractors to include core labor standards in infrastructure projects. Both of these changes are still in the process of implementation. Third, the IFC and ILO have launched a “Better Work Program” to improve labor practices in participating countries. The program monitors workplaces to determine whether national labor laws and core labor standards are observed and provides advice and assistance to firms to improve compliance.
At the same time, however, the IFIs have traditionally been strong supporters of labor market flexibility. The World Bank’s annual publication, Doing Business, is perhaps the most public advocate of flexibility. Polaski (2007: 3) calls it the “most glaring example of inconsistency by the World Bank in promoting good workplace practices and respect for workers’ rights.” The report, which is used as an input for other Bank programs and as a guide for potential investors, assigns high scores to countries that promote use of fixed-term contracts, establish a low (or no) minimum wage, roll back restrictions on hours worked, and make it easier to fire workers. As for the IMF, Anner and Caraway report that of the 91 countries that signed “letters of intent” with the IMF in the period 1998–2005, 32% included commitments to create more flexible labor markets. They conclude that the IFIs have, on balance, promoted flexibility over labor standards.2
Attempts have been made to enlist the World Trade Organization (WTO) into the struggle for better labor standards, but with little success.3 Other trade mechanisms, however, have proved more useful. One is the Generalized System of Preferences (GSP) which authorizes industrial countries to provide one-way preferences to certain developing countries. Particularly in the case of the USA, the GSP has been a powerful instrument to promote labor rights.4 Regional trade agreements have also become important instruments for promoting labor rights in developing countries. Since the North American Free Trade Agreement (NAFTA) was negotiated in the early 1990s, US free trade agreements (FTAs) have included labor clauses at the insistence of Congress. In the case of NAFTA itself, labor clauses were incorporated through a side agreement, but in later FTAs, they have been an integral part of the treaties themselves (Elliott 2003; Polaski 2004). Nine FTAs with 14 countries include labor agreements: Australia, Bahrain, Central America, Chile, Jordan, Morocco, Oman, Peru, and Singapore. Three more are pending approval in the US Congress: Colombia, Panama, and South Korea.5 The problem, again, is enforcing these agreements. Critics claim that the labor clauses exist on paper only.6
The European Union has also taken an active role in promoting labor rights. As Cook argues in her paper in this volume, the EU has been perhaps the major force in giving the Eastern European accession countries the best labor standards among the four developing regions we study in this project. In order to become members, countries have to conform to all EU rules (the acquis communautaire). While labor rights are not at the top of the list, they are included, and this has made an important difference.7 For those neighboring countries that are not serious candidate members, however, the EU has been taken a different—and less supportive—role on labor rights. Martín (2007) has studied the agreements between the EU and the Maghreb countries (Euro-Mediterranean or Euromed Partnership) and finds that they have tended to stress flexibility more than labor standards. He attributes this difference to the greater influence of European companies when EU membership is not at stake.8 Others point to European concerns with security and stability on their southern border (Gstöhl 2009).
A more indirect way in which Europe has been important in the debates on labor market characteristics has been through ideas. Specifically, the concept of “flexicurity” has emerged both at the EU and national levels in Europe to mean a combination of labor flexibility and job protection. The idea is that the two can go together rather than being antagonistic as is commonly argued.9 While the concept appears attractive, substantial disagreement exists—especially between unions and employers—on exactly what it means. Moreover, since the most prominent example of flexicurity in practice is in wealthy Denmark, it is also questionable whether it can work in poor developing countries.10 We return to this issue in the concluding section to this paper.
Finally, it should be noted that private corporate actors also play a role in promoting labor rights through corporate social responsibility. Anner (2009) reports that in the face of activist demands for a greater government role in protecting labor rights, corporations proposed self-regulation via codes of conduct and internal monitoring. While many major multinational corporations have introduced them, they vary in scope, design, and effectiveness. Some have been unilateral endeavors, others involve NGOs, and a few have been negotiated with labor unions. All have had difficulties in enforcement and, at best, complement the actions of governments and international agencies.11
Impact of Labor Market Characteristics
Beyond the origin of labor standards and labor flexibility, a vast literature exists on the effects of these labor market characteristics in developing countries. Several recent reviews have tried to summarize lessons learned and questions that remain. Harvard economist Richard Freeman (2009) outlines a number of conclusions derived from his reading of the empirical literature on labor market institutions, which he defines as government regulations and labor union activity. (1) Labor institutions vary greatly across developing countries. (2) Minimum wage regulations have a small negative effect on employment. (3) Employment protection regulations shift output and employment from the formal to the informal sector. (4) Mandated benefits increase labor costs and reduce employment modestly. (5) The impact of labor regulations on growth is inconclusive, and recovery from shocks is not related to the strength of labor institutions. (6) Cooperative labor relations tend to produce better economic outcomes.
Three European economists (Boeri, Helppie, and Macis 2008) offer another review. They focus on four areas: minimum wages, mandated benefits, dismissal costs, and unemployment insurance. On the minimum wage, they say that “the jury is still out.” The one exception is the negative, but small, impact of the minimum wage on low-skill workers. Mandated benefits, they argue, have not yet been systematically analyzed due to the great heterogeneity in policy characteristics and implementation. Employment protection legislation and unemployment insurance can be seen as substitutes for each other. Based on the evidence they review, they say that the latter is more “mobility friendly” and thus helps productivity growth, while the former blocks turnover with negative implications for productivity.
A third review of labor market policies in developing countries, undertaken by Cornell economist Gary Fields (2007), focuses on methodological issues. He stresses the need for good theory and clear definitions. He also discusses the advantages of different kinds of data analysis techniques, including aggregate cross-sectional quantitative data, microdata from comparable cross-sections, panel data, cross-country time series data, experimental data, and qualitative data. He illustrates the six techniques with different studies of labor markets. By way of conclusion, he takes the example of assessing the effects of India’s employment protection laws. Indian law requires that firms employing 100 or more workers must obtain government permission for layoffs or plant closings. A number of researchers have looked at how this affects workers using state-level data on amendments to national labor laws. One set of conclusions is that in the states where the laws were particularly strict, output, employment, investment, and productivity were lower while informality and poverty were higher (Besley and Burgess 2004). In response to criticisms of this study, two World Bank economists tried to distinguish between changes in the laws regarding flexibility in hiring and firing and the de facto use of contract labor. Both were found to hurt workers and impose costs on society, but the authors advocate smarter regulation rather than less regulation (Ahsan and Pagés 2007).
As can be seen, all of the conclusions reported are tentative and/or the effects are found to be small. The authors stress data problems, heterogeneity of policies and context, and the lack of clear theoretical models. Moreover, interaction effects with other policy areas make it difficult to determine causality. All agree that more work needs to be done and better methods used. We hope to make some modest contributions to this endeavor.
Characteristics of the Project
Our project offers several innovations with respect to the study of globalization and labor in developing countries. The first is the focus on institutions rather than on economic flows. In particular, we look at global pressures for labor standards and labor flexibility. As mentioned earlier, the two have generally been studied separately. We bring these discussions together to see which trend is more important and how the two processes have related to each other. The second innovation is the use of a common framework to analyze four regions of the developing world: Eastern Europe, Latin America, East Asia, and the Middle East. Not only are the same questions addressed, but the same statistical measures are employed and the same basic tables and graphs are presented to facilitate the comparative analysis that is presented later in this paper. Third, the statistical measures themselves are innovative. On the one hand, we have developed de jure indices of labor standards and flexibility. In this part of the analysis, we focus on laws and regulations relating to the two and how they are supposed to work in the various countries of the four regions. This is the typical way in which analysts have proceeded, as can be seen by examining the literature. On the other hand, we have gone beyond the laws on the books to try to get a handle on how these laws have (or have not) been implemented. That is, we also present de facto indices of labor standards and flexibility. These are the policies that actually affect workers and the labor market environment.
One of our basic assumptions in the project is that certain characteristics of labor markets are shared within regions.12 These commonalities derive from shared historical experiences. One way in which they come about is through the influence of particular industrial countries in different regions or what some have called colonial legacies.13 For example, the USA has been particularly important for Latin America in the twentieth century, but vestiges of Iberian colonial influence remain in labor codes. Japan has been especially relevant in East Asia, but again, other influences—especially from former European colonial powers—need to be taken into account in the study of labor markets. While Western Europe has played the central role in Eastern Europe in the last two decades, the impact of the period of Soviet dominance continues to linger with respect to labor rights and flexibility. Finally, the Middle East has had more eclectic influences over time than the other three regions; these manifest themselves in the characteristics of labor rights and organization in different ways in different countries.
Beyond external influence, the countries of the four regions themselves have formed organizations that have contributed to shared approaches in various policy arenas, including labor. These are seen especially in the form of political and economic integration schemes. We argue that these historical and organizational factors make it relevant to use region as a basic unit of analysis, but of course we also examine the behavior of individual countries within the four regions. The table and graphs in the next section reveal the extent to which the four regions differ on the variables in which we are interested.
Within the four regions, the authors identify differences across countries with respect to labor market characteristics. To help explain them, they group countries according to the criteria that seem most relevant in the particular regional context. Political regime is important in all four cases, but regime variables are incorporated in different ways. In the case of Latin America, Burgess argues for a historical approach to categorization. Specifically, she creates four “legacy regimes”—labor populism, pluralist welfarism, paternalist dictatorship, and conservative oligarchy—based on the type of political regime that prevailed when the pattern of state–labor relations was established. Cook, looking at Eastern Europe, explains intra-regional differences with respect to labor standards by a combination of democratic political system and relationship to the European Union. She finds less variation on flexibility. In the article on East Asia, Caraway also highlights the role of political regime—authoritarian, semi-democracy, and democracy—though pointing out the overlap between per capita income and political system. Cammett and Posusney remind us that all political systems in the Middle East are authoritarian, but argue that the oil monarchies behave differently than two sets of poor countries—non-oil monarchies and single-party regimes—with respect to labor institutions.
Despite these different sources of categorization, all of the authors use the same measures to analyze the situation with respect to labor standards and flexibility in the countries they study. Four indices were developed—de jure and de facto measures for both labor standards and flexibility; in general, they are calculated for the year 2006. We decided to develop new indices rather than use or improve existing ones. Most cross-national research either combines de jure and de facto indicators (Kucera 2004; Mosley and Uno 2007) or focuses exclusively on the law (Botero et al. 2004; World Bank 2007). While legal rights are important, they are meaningless if they are not enforced. Offering separate indices for de jure and de facto labor rights also expands the set of questions that scholars can address in their work.
The de jure index for labor standards is based on collective labor rights such as freedom of association, collective bargaining, and the right to strike. These are the most controversial of the core labor standards. The de jure scores are based exclusively on each country’s labor regulations and ratification of ILO conventions. The de facto index subtracts points for documented labor violations, weak rule of law, and/or restrictions on associational or organizational rights. Varying from 0 to 100, a higher index indicates better labor standards. Sources include International Confederation of Free Trade Unions (ICFTU) annual reports, US State Department annual human rights reports, and ILO committee on freedom of association reports. We recognize the problems with these sources, e.g., their varying accuracy and consistency across regions, the possible political biases, and their definitional problems.14 Nonetheless, we believe that they are the best available for our purposes.
The de jure flexibility index is based on the ease with which employers can hire, fire, and adjust working hours. It combines the World Bank’s measure of employment rigidity with the legally mandated level of severance pay (firing costs). The de facto scores subtract points for weak rule of law. Again, possible scores vary between 0 and 100. In this case, higher scores indicate more flexible labor regimes. The World Bank’s rule of law index (ROL), together with the other world governance indicators (WGI), are the subject of heated scholarly debate. Critics have called attention to the lack of attention to concept validity, the high correlations between indicators, the absence of clear explanations for why particular indicators are assigned to specific components of the WGI, the alleged pro-business or neoliberal bias in both the data sources and the understanding of “good governance,” and the inclusion of policy-oriented indicators.15 Again, we recognize the shortcomings of the WGI; we use its ROL measure because it provides the best data available for the set of countries analyzed in this volume (for greater detail on the construction of the indices and the sources used, see the Appendix to this paper.)
Patterns of labor standards and labor flexibility vary substantially across the four regions we study; they also vary within regions. In this section, we concentrate mainly on cross-regional variation, while the cross-country differences are the principal topic of the regional papers. Nonetheless, we also provide some comparative data on cross-country variation within regions.
Labor market indices in four developing regions, circa 2006
Labor standards index
Labor flexibility index
Eastern Europe/Former Soviet Union
Middle East/North Africa
The labor flexibility indices differ from labor standard measures in two important ways. First, for labor flexibility, de facto scores are higher than de jure. That is, countries’ labor markets are more flexible in practice—workers have fewer protections—than appears from examining their laws and regulations. On average across the four regions, the de jure flexibility index is 52 of a possible 100, with the de facto index rising to 61. Second, there is less difference across regions with respect to flexibility than to labor standards. The de facto scores vary only between 58 and 64. The most flexible region is the Middle East, while the least flexible is Latin America, but the gap is not great.
How can these differences be explained? We begin with two general explanatory factors and then move to more specific ones when we examine individual regional data below. With respect to the different patterns between the de jure and de facto indices, the apparent difference that the de facto numbers are lower for labor standards but higher for flexibility really reflects the same phenomenon. That is, all kinds of labor rights are greater on paper than in reality because of lack of enforcement. In the case of flexibility, the lack of enforcement shows up as higher de facto indices or more flexible labor markets.
Even though the underlying trends for both indices are similar, the difference among countries is much larger for labor standards than for flexibility. This difference could be accounted for in two ways.17 On the one hand, it could derive from greater international pressures on all regions for flexibility in comparison to labor standards. On the other hand, it could mean that there is greater internal consensus within the various countries and regions with respect to the advantages of flexible labor markets. These two could come together insofar as external pressures are reflected in the opinions of important domestic policy makers.18 In practice, the relative mix may well vary across regions as we will now see.19
The nations of Eastern Europe have a high average level of de facto labor standards, but a broad range from 93 in Slovenia to 46 in Russia (see Fig. 1a). On flexibility, the range is much narrower: from 68 in Belarus and Romania to 49 in Estonia and Slovenia. The relationship between the two indices is negative with better labor standards accompanying lower flexibility. This is especially the case if the outliers circled (and explained below) are excluded. As seen in Table 1, the gap between de jure and de facto measures is relatively small for both indices in comparison to the other three regions.
According to Cook’s analysis, this pattern can be explained by three main factors. First, and perhaps most important, is the influence of the European Union. For countries wanting to accede to the EU, a variety of policy requirements are set out in the rules of the regional community. One indicator of the EU’s importance is that the countries in Eastern Europe that have become members are all located at the top of the Fig. 1a, indicating better labor standards, while the three who have not (Russia, Ukraine, and Belarus) are shown in the circle at the bottom of the graph. Second, in addition to the EU but hard to disentangle, is the influence of internal democratic politics. Those countries that have become members of the EU are all democracies, while non-members have more authoritarian systems even if they hold elections. Greater democracy is likely to mean a larger role for unions and other advocacy groups that push for better labor standards. Finally, initial conditions are also relevant. All of these countries are former communist systems where workers had greater protection than in many other parts of the developing world. In terms of flexibility, both external influence and initial conditions are important explanatory factors. Despite the fact that the EU pushes for better labor conditions, it also stresses flexible labor markets. At the same time, the lack of flexibility in the Soviet era still has an influence on the region.
Latin America shows a different pattern, as seen in Fig. 1b. The average level of de facto labor standards is lower than in Eastern Europe and the range is somewhat narrower: from 84 in Uruguay to 50 in Guatemala and Honduras. The flexibility index range is about the same as in Eastern Europe: the highest is 68 in Nicaragua, while the lowest level is 51 in Bolivia. With all countries included, the two indices do not seen to be related. Excluding the circled outliers, there is again a moderately strong negative relationship. These outliers are defined according to deviations from Burgess’s legacy regimes. In particular, five countries (Bolivia, Chile, Costa Rica, Nicaragua, and Uruguay) underwent profound periods of social change that disrupted the historical patterns.
Latin America also has different causal patterns at play in promoting labor standards and flexibility than those present in Eastern Europe. Burgess identifies a number of important external influences. In particular, she points to the pressures for better labor standards embodied in US trade policy, including the US GSP in Central America, which was more important for de jure than de facto changes. She also highlights the role of free trade agreements with Mexico, Chile, Central America, Peru (and pending FTAs with Colombia and Panama). Other influences in favor of improving labor standards include conditionality of the World Bank and IMF as well as codes of conduct established by multinational corporations in certain firms and/or sectors. Burgess argues that US FTAs and World Bank and IMF agreements have raised the profile of labor rights in general but are weak on implementation and ambivalent about collective rights. The FTAs may also have had a strong, indirect impact on flexibility, but this was not a stated goal. Nonetheless, she says that all of these external pressures have been heavily mediated through domestic political processes, especially in terms of enforcement, such that it would be misleading to argue that the main influence on labor market characteristics has come from outside.
East Asia has a lower level of de facto labor standards than Eastern Europe or even Latin America, as seen in Fig. 1c. The range is large, from Indonesia at 67 to China at 35 out of a possible 100. Flexibility among East Asian countries also differs greatly when compared to counterpart regions: from 87 in Singapore to 45 in Taiwan. The relationship between labor standards and flexibility is again negative, especially if outlying Singapore is eliminated from the graph. Singapore has far more flexible labor markets than any country in our four regions (labor standards, by contrast, are quite typical). A city state whose economy is based heavily on multinational corporate investment, Singapore is very open in all economic dimensions, and labor policy is tied into this general picture. Its history as a former British colony may also help to account for its openness.
Caraway argues that international influence has not been particularly important in labor market policies in East Asia with respect to either labor standards or flexibility. In part, this is because these countries—until the financial crisis of 1997–1998—were such economic success stories that they did not have to deal extensively with the IFIs. Nor did they negotiate trade agreements with the USA or the European Union until recently. Even after the crisis, South Korea was the main case of successful external pressure on labor markets. Otherwise political characteristics in the countries themselves, and initial conditions including fairly open economies and political authoritarianism, have been more relevant.
Finally, labor standards in the Middle East and North Africa are the lowest among our four regions, and the differences among countries are great, as seen in Fig. 1d. Indeed, two of the oil monarchies in the region (Saudi Arabia and the UAE) have no formal labor rights at all. The highest de facto labor standards are found in Morocco (58). Flexibility scores also vary substantially from Bahrain (77) to Morocco (48). Unlike the other three regions, little relationship is found between the two indices even when the two outliers (Saudi Arabia and the UAE) are excluded.
Cammett and Posusney stress intra-regional differences in terms of external versus internal influence on labor market characteristics. They indentify FTAs with the USA as the most important type of external influence in some of the oil monarchies, leading to improved labor standards, especially freedom of association. The poorer countries, by contrast, typically suffered from budget deficits that led them to seek support from the IFIs. This relationship promoted greater flexibility due to IFI conditionality. Like their colleagues, however, Cammett and Posusney point to internal structures and processes as being at least as important as external influence. In the case of the oil monarchies, the reliance until recently on imported labor together with the lack of an industrial sector meant they had long had flexible labor markets. Indeed, they have become less flexible recently because of the trend toward indigenization of the work force and some move toward industrial investment. In the non-oil monarchies and the single-party regimes, the presence of an important domestic industrial sector enabled both labor and local businesses to push for their own preferred type of labor markets—although the authoritarian political systems meant that private sector pressures were less important than in other regions we study.
The Impact on Labor Market Outcomes
Up until now, we have concentrated on trying to explain the impact of globalization on labor market policies—specifically, the implementation (or not) of better labor standards and greater labor flexibility—in developing countries. That is the main focus of the project. In this introductory paper that compares across regions, however, we attempt an additional task and ask if de facto labor standards and labor flexibility have had an impact on labor market outcomes. In particular, we are interested in wages, unemployment, and informality. The initial idea was to do some econometric modeling, but the data problems proved too great due to the small size of the sample and a single point in time for the indices. Moreover, labor statistics are notoriously poor even in comparison with other data from developing countries.
Given the problems with the econometrics, we present below some simple bivariate scatterplots that can serve as an entry point into a more extensive analysis in the future if and when better data become available. Four labor market outcomes are analyzed: the legally mandated annual minimum wage (for the latest available year), the average monthly wage across sectors (in 2006), the average unemployment rate (for the period 2002–2006), and the average rate of informality (for the period 2002–2006). Informality is defined as the self-employed and unpaid family members as a share of the workforce.20
As far as minimum and average wages are concerned, the expectation is that labor standards would be positively related to wages, while flexibility would be negatively related. Better labor standards should increase the power of unions, making it possible to push for higher wages; just the opposite holds for flexibility. Key components of the World Bank index we use—the easing of rules about hiring and firing, the ability of employers to set pay scales based on performance, and especially the limits on labor union activity per se—reduce union power. For unemployment, by contrast, the relationship with labor standards would be positive insofar as they put a damper on hiring. In the opposite direction, if flexibility makes it easier to fire workers, then a standard argument is that employers would be more willing to take a risk with new workers, and unemployment would show a negative relationship. Finally, informality should also be positively related to labor standards as the latter place more barriers to small businesses entering the formal market. Since flexibility correspondingly places fewer barriers, it should thus be negatively related to informality.
The impact of globalization on labor in developing countries is important for economic, social, and political reasons. This topic has often been a polemical one, but we bring empirical evidence to bear with respect to the influence of global institutions on labor market characteristics. Specifically, we study the relevance of demands by international actors for better labor standards and more labor flexibility. To do so, we create innovative measures for both de jure and de facto labor standards and flexibility.
Our results suggest that global institutions have been important, but in different ways and to different degrees in the four middle-income regions we study. For Eastern Europe, the EU was very influential through its requirements for accession countries (though less so for others). In Latin America, US trade policy and multinational corporations’ codes of conduct appear to have played a role in strengthening labor standards, while the IFIs have been influential in promoting labor market flexibility. Similar factors were important in the Middle East, especially IFI loans and FTAs with the USA. Interestingly, trade agreements with the EU were not relevant despite the Euromed program between Europe and various Middle Eastern countries. East Asia, by contrast, has experienced less external influence than the other regions precisely because the external factors that have been important elsewhere—IFI loans or trade agreements with the United States or EU—have been absent in North and Southeast Asia until recently. Economic success until the 1997–1998 crisis meant that there was little reason to deal with the IMF and World Bank, and no trade agreements with the West were negotiated until the last few years.
Despite the importance of international institutions, not all the variation in labor policies was explained by them. Of course, we never expected that would be the case. The era of attributing overwhelming dominance to the industrial countries or IFIs or multinational corporations is long over. In all four regions, the authors stress that domestic political structures and processes have been at least as important as external influences in determining labor market characteristics. A distinction needs to be drawn between direct and indirect effects. Directly, external pressures have perhaps been most influential on de jure outcomes, but even these are heavily mediated by domestic factors. Indirectly, external pressures combined with internal actors to shape de facto outcomes, overwhelming de jure rights in many cases. The challenge is to see how external and internal forces interrelate. We believe we have made some progress in this area.
We are also interested in the relationship between labor standards and flexibility. We expected to find a negative correlation, and that expectation held when pooling the data from the four regions. It also held within regions in the cases of Eastern Europe and East Asia. When certain outliers were eliminated in the Latin American sample, a negative link also appeared there. In the Middle East, the relationship was extremely weak, probably because sub-regional groups with very different structural characteristics canceled each other out. Nonetheless, there is a strand of thought and practice—the European notion of flexicurity—which argues that the relationship need not be negative. Some Latin American countries, especially Chile and Costa Rica, also have tried to combine flexibility and good labor standards. But it remains to be seen if poorer developing countries can imitate their European counterparts since substantial resources are necessary for the active labor market policies involved. This will be an important topic to investigate in the future (see Weller 2009 on flexicurity in Latin America.)
This introductory paper goes one step beyond the shaping of labor markets by policies on labor standards and flexibility. It also begins to examine the impact of standards and flexibility on labor market outcomes. Our simple scatterplots suggest several possible impacts. De facto labor standards are positively related to minimum and average wages, while flexibility is negatively correlated. Unemployment is positively related to labor standards, while flexibility has no correlation. For informality, we find a nonlinear relationship: higher labor standards are correlated with greater informality among countries with poor standards, while the relationship changes sign in countries with better standards. All of these correlations are weak, but they provide some hypotheses that can be explored if better data become available to permit the use of more sophisticated methods. In the meantime, they suggest that the labor market characteristics we try to explain in the project are themselves likely to affect workers’ well-being.
Although 70% of the 55 individual countries in our four regional samples are classified by the World Bank as middle income, 13 countries are defined as high income and four as low income.
The continued promotion of flexibility is not limited to the IFIs. For example, a recent publication by the Carnegie Endowment for International Peace, authored by a progressive former minister from Chile who analyzes policies needed to deal with the current economic crisis, says that “new legislation will be required that reduces the costs of hiring and firing by firms, thus increasing market flexibility.” See Foxley (2009: 16).
On the US and EU GSP processes, and reasons that the former has placed stricter limits on partner countries, see Grynberg and Qalo (2008).
A recent report for the Congressional Research Service describes four models for the inclusion of labor rights in US FTAs; see Bolle (2008).
See, for example, Wise (2003) for a discussion of this issue with respect to Mexico.
Aita (2008) provides an extensive study of the Euromed program; Martín was the scientific advisor of the project.
Flexicurity is being discussed in Latin America. See, for example, Weller (2009).
See Stallings (1995) for an argument about the relevance of regions for socioeconomic analysis.
See www.worldbank.org/wbi/governance for information on one widely used set of institutional indicators. Criticisms of these indicators were discussed above.
It may also have to do with different methods for calculating de facto scores for labor standards and flexibility.
On the difference between external influence as leverage versus linkage, see Stallings (1992). Linkage refers to the internalization of external influence.
This puzzle—why greater consensus exists on flexibility than on labor standards—suggests a weakness with our approach in this project. We focus on the institutional elements of globalization at the cost of ignoring the economic aspects. But the greater convergence on flexibility may well result from these economic characteristics. For example, the ability to export successfully and to obtain access to foreign capital may force countries to follow more flexible policies. Why this would not also hold for labor standards is beyond the scope of the project because of our design.
Sources for these variables are: minimum wage and average wage (ILO, LABORSTA database); unemployment and informality (ILO, key indicators of the labor market database). I thank Bapu Vaitla for help with this analysis.
The Singaporean outlier is a good illustration of the problem with bivariate analysis. Since Singapore has much higher per capita GDP than most countries in our sample, its wages will obviously be high, demonstrating the need to control for income level-in addition to other variables.
The Associational and Organizational Rights component of the Freedom House Civil Rights measure includes freedom of assembly, demonstration and open public discussion; freedom for nongovernmental organizations; free trade unions and peasant organization or equivalents; effective collective bargaining; and free professional and other private organizations. Higher scores reflect stronger rights.
I am grateful for the assistance of Bapu Vaitla, a graduate student at the Fletcher School at Tufts University, who helped with the data analysis on labor market outcomes. I also benefitted from comments by Katrina Burgess, Melani Cammett, Teri Caraway, Linda Cook, and two anonymous reviewers. Funding for the project was provided by the Ford Foundation and the Watson Institute for International Studies at Brown University. I completed this paper and the editing of the special issue while at the Institute for Development and Human Security of Ewha Womans University in Seoul, Korea, which is funded by the WCU (World Class University) program through the National Research Foundation of Korea funded by the Ministry of Education, Science and Technology of the Republic of Korea (grant no. R32-20077).
- Adelman J, editor. Colonial legacies: the problem of persistence in Latin America. New York: Routledge; 1999.Google Scholar
- Ahsan A, Pagés C. Are all labor regulations equal? Assessing the effects of job security, labor dispute, and contract labor laws in India. Policy Research Working Paper No. 4259. Washington: World Bank; 2007.Google Scholar
- Aita S. Employment and labor law in the Arab Mediterranean countries and the Euromediterranean Partnership. Madrid: Fundación Paz y Solidaridad Serafín and Agencia Española de Cooperación Internacional; 2008.Google Scholar
- Anner M. The corporate social responsibility movement: private sphere mechanisms to enforce labor standards in the global economy. Pittsburgh: University of Pittsburg, Department of Labor Studies and Employment Relations, unpublished; 2009.Google Scholar
- Boeri T, Helppie B, Macis M. Labor regulations in developing countries: a review of the evidence and directions for future research. SP Discussion Paper No. 833. Washington: World Bank; 2008.Google Scholar
- Bolle MJ. Overview of labor enforcement issues in free trade agreements. Report for Congress, Order No. RS22823. Washington: Congressional Research Service; 2008.Google Scholar
- Booth AE. Colonial legacies: economic and social development in East and Southeast Asia. Manoa: University of Hawaii Press; 2007.Google Scholar
- Burtless G. Workers’ rights: labor standards and global trade. Brookings Review; 2001.Google Scholar
- Compa L. Corporate social responsibility and workers’ rights. Comp Labor Law J. 2008;30(1):1–10.Google Scholar
- Elliott KA. Labor standards and the Free Trade Area of the Americas. Working Paper No. 03-7. Washington: Institute for International Economics; 2003.Google Scholar
- Emmenegger P. The long road to flexicurity: the development of job security regulations in Denmark and Sweden. Scandinavian Political Studies; 2010 (in press).Google Scholar
- ETUC (European Trade Union Confederation). Flexicurity; 2007. Available at: www.etuc.org/a/4288?var_recherche=%20Flexicurity.
- Feinberg R. Corporate social responsibility: contributions to the APEC integration agenda. Paper presented at APEC Study Center Consortium Annual Meeting, Singapore; 2009.Google Scholar
- Foxley A. The global financial crisis and middle-income countries. Washington: Carnegie Endowment for International Peace; 2009.Google Scholar
- Freeman R. Labor regulations, unions, and social protection in developing countries: market distortions or efficient institutions? Working Paper 14789. Cambridge: National Bureau of Economics Research; 2009.Google Scholar
- Fung A, O’Rourke D, Sabel C, editors. Can we put an end to sweatshops? Boston: Beacon; 2001.Google Scholar
- Grynberg R, Qalo V. Development and enforcement of labour standards in US and EU preferential trading arrangements. In: Qalo V, editor. Bilateralism and development: emerging trade patterns. London: Cameron May; 2008. p. 117–215.Google Scholar
- Gstöhl S. The social dimension of EU neighborhood policies. In: Orbie J, Tortell L, editors. The European Union and the social dimension of globalization. London: Routledge; 2009. p. 62–80.Google Scholar
- Hagen K. Policy dialogue between the International Labor Organization and the international financial institutions: the search for convergence. Occasional paper No. 9. Geneva: Friedrich Ebert Stiftung; 2003.Google Scholar
- Jenkins R, Pearson R, Seyfang G, editors. Corporate responsibility and labour rights: codes of conduct in the global economy. London: Earthscan; 2002.Google Scholar
- Keune M. EU enlargement and social standards: exporting the European social model? In: Orbie J, Tortell L, editors. The European Union and the social dimension of globalization. London: Routledge; 2009. p. 45–61.Google Scholar
- Khor M. The WTO and the battle over labor standards. Third World Network Features; 1997. Available at: www.globalpolicy.org.
- Kucera D. Measuring trade union rights: a country-level indicator constructed from coding violations recorded in textual sources. Geneva: International Labor Organization; 2004.Google Scholar
- Kurtz M, Schrank A. Growth and governance: models, measures, and mechanisms. J Polit. 2007;69(2):538–54.Google Scholar
- Madsen PK. How can it possibly fly? The paradox of a dynamic labour market in a Scandanavian welfare state. In: Campbell JL, Hall JA, Pedersen OK, editors. National identity and a variety of capitalism: the case of Denmark. Montreal: McGill University Press; 2006.Google Scholar
- Martín I. In search of development along the southern border: the economic models underlying the Euro-Mediterranean Partnership and the European Neighbourhood Policy. In: Ferragina AM, editor. Bridging the gap: the role of trade and FDI in the Mediterranean. Naples: Conssiglio Nazionale delle Ricerche-Istituto di Studi sulle Società de Mediterraneo; 2007. p. 115–41.Google Scholar
- Mosley L, Uno S. Racing to the bottom or climbing to the top? Economic globalization and collective labor rights. Comp Polit Stud. 2007;40(8):923–48.Google Scholar
- Orbie J, Tortell L. The European Union and the social dimension of globalization. London: Routledge; 2009.Google Scholar
- Polaski S. Trade and labor standards: a strategy for developing countries. Washington: Carnegie Endowment for International Peace; 2004.Google Scholar
- Polaski S. The World Bank’s approach to core labor standards and employment creation: recent developments. Testimony Submitted to the United States House of Representatives, Committee on Financial Services; 2007.Google Scholar
- Seidman G. Beyond the boycott: labor rights, human rights, and transnational activism. New York: Russell Sage Foundation; 2007.Google Scholar
- Stallings B. International influence on economic policy: debt, stabilization, and structural reform. In: Haggard S, Kaufman R, editors. The politics of economic adjustment. Princeton: Princeton University Press; 1992. p. 41–88.Google Scholar
- Stallings B, editor. Global change, regional response: the new international context of development. New York: Cambridge University Press; 1995.Google Scholar
- Stern RM, Terrell K. Labor standards and the World Trade Organization: a position paper. Policy Brief No. 2. Ann Arbor: University of Michigan William, Davidson Institute; 2003.Google Scholar
- Weller J, editor. Regulation, worker protection and active labour-market policies in Latin America. Santiago: Economic Commission for Latin America and the Caribbean; 2009.Google Scholar
- World Bank. Doing business: measuring business regulation. Online database. Washington, DC: World Bank and International Finance Corporation; 2007. Available at: www.doingbusiness.org/CustomQuery