A. Some Basic Facts

North-South trade in manufactures is on the whole much more important for the South than the North. Table 1 shows exports and imports of manufactures (as a ratio of GDP) for some newly industrializing economies with the industrial countries. For many of these developing countries manufacturing trade with the industrial countries is now quite large in relation to national output (Brazil and India are exceptions). Exports of manufactures to the North have increased rapidly, but the majority of these developing countries continue to run trade deficits in manufactures, as imports have grown nearly as much.

Table 1.

Developing Country Trade in Manufactured Goods with All Industrial Countries, Selected Years, As a Percent of Developing Country GDP

Sources: UN TARS trade database, IMF International Financial Statistics.

Table 1.

Developing Country Trade in Manufactured Goods with All Industrial Countries, Selected Years, As a Percent of Developing Country GDP

		Exports to Industrial Countries	Imports From Industrial Countries	Trade Balance
Brazil	1970	0.4	3.4	-3.1
	1980	1.3	2.8	-1.5
	1990	2.3	1.8	0.5
	1994	2.1	2.6	-0.5
China	1990	3.1	4.6	-1.5
	1994	9.7	10.2	-0.5
India	1980	1.3	2.1	-0.8
	1990	2.4	2.6	-0.1
	1994	3.8	2.8	1.0
Indonesia	1970	0.1	N/A	N/A
	1980	0.6	7.0	-6.4
	1990	5.1	9.6	-4.5
	1994	6.6	8.2	-1.7
Korea	1970	6.3	10.5	-4.2
	1980	15.6	12.1	3.5
	1990	16.3	12.7	3.6
	1994	11.4	12.7	-1.3
Malaysia	1970	8.2	14.8	-6.6
	1980	10.2	20.9	-10.7
	1990	21.3	33.8	-12.5
	1994	33.2	45.0	-11.8
Mexico	1970	0.9	4.7	-3.8
	1980	0.8	6.2	-5.4
	1990	3.8	7.3	-3.5
	1994	11.4	12.7	-1.3
Thailand	1970	1.3	10.7	-9.4
	1980	4.6	10.2	-5.6
	1990	11.8	18.7	-6.9
	1994	12.4	18.8	-6.4

Sources: UN TARS trade database, IMF International Financial Statistics.

View Table

Table 1.

Developing Country Trade in Manufactured Goods with All Industrial Countries, Selected Years, As a Percent of Developing Country GDP

		Exports to Industrial Countries	Imports From Industrial Countries	Trade Balance
Brazil	1970	0.4	3.4	-3.1
	1980	1.3	2.8	-1.5
	1990	2.3	1.8	0.5
	1994	2.1	2.6	-0.5
China	1990	3.1	4.6	-1.5
	1994	9.7	10.2	-0.5
India	1980	1.3	2.1	-0.8
	1990	2.4	2.6	-0.1
	1994	3.8	2.8	1.0
Indonesia	1970	0.1	N/A	N/A
	1980	0.6	7.0	-6.4
	1990	5.1	9.6	-4.5
	1994	6.6	8.2	-1.7
Korea	1970	6.3	10.5	-4.2
	1980	15.6	12.1	3.5
	1990	16.3	12.7	3.6
	1994	11.4	12.7	-1.3
Malaysia	1970	8.2	14.8	-6.6
	1980	10.2	20.9	-10.7
	1990	21.3	33.8	-12.5
	1994	33.2	45.0	-11.8
Mexico	1970	0.9	4.7	-3.8
	1980	0.8	6.2	-5.4
	1990	3.8	7.3	-3.5
	1994	11.4	12.7	-1.3
Thailand	1970	1.3	10.7	-9.4
	1980	4.6	10.2	-5.6
	1990	11.8	18.7	-6.9
	1994	12.4	18.8	-6.4

Sources: UN TARS trade database, IMF International Financial Statistics.

Table 2 shows analogous trade data from the point of view of the major industrial countries, also as a share of their respective GDPs. Contrary to popular fears about deindustrialization, the major industrial countries continue to have manufacturing trade surpluses with the developing world in the 1990s, except for the United States and Canada, which have shifted to modest deficits. Also, while the rate of increase of import penetration of Southern manufactures in the North is sizeable, the level remains quite low. Overall, imports of manufactures from the developing world currently account for only about 2 percent of industrial country GDP (Krugman and Lawrence, 1994). The decline in manufacturing employment in the developed countries mostly reflects the increasing importance of the service sector in these economies, and the tendency for productivity in manufacturing to increase more rapidly than in services (Rowthorn and Ramaswamy 1997).

Table 2.

Industrial Country Trade in Manufactured Goods with All Developing Countries, Selected Years, As a Percent of Industrial Country GDP

Sources: UN TARS trade database, IMF International Financial Statistics.

Table 2.

Industrial Country Trade in Manufactured Goods with All Developing Countries, Selected Years, As a Percent of Industrial Country GDP

		Exports to Developing Countries	Imports From Developing Countries	Trade Balance
United States	1970	0.8	0.3	0.5
	1980	1.9	1.1	0.8
	1990	1.5	2.0	-0.5
	1995	2.1	3.2	-1.1
Japan	1970	3.2	0.5	2.7
	1980	5.2	0.7	4.5
	1990	3.1	1.0	2.1
	1994	3.4	1.2	2.1
West Germany	1970	4.1	1.6	2.5
	1980	5.4	2.2	3.2
	1990	4.8	3.2	1.6
France	1970	2.6	0.8	1.9
	1980	3.6	1.1	2.5
	1990	2.9	1.6	1.3
	1995	3.4	2.0	1.4
United Kingdom	1970	4.1	1.3	2.8
	1980	4.8	2.1	2.7
	1990	3.2	2.9	0.3
	1994	3.8	3.4	0.4
Canada	1970	1.0	0.3	0.7
	1980	1.5	0.7	0.9
	1990	0.8	1.4	-0.6
	1995	1.4	2.7	-1.4
Italy	1970	3.3	0.9	2.3
	1980	7.1	1.2	5.8
	1990	5.0	1.5	3.5
	1994	9.8	2.0	7.8

Sources: UN TARS trade database, IMF International Financial Statistics.

View Table

Table 2.

Industrial Country Trade in Manufactured Goods with All Developing Countries, Selected Years, As a Percent of Industrial Country GDP

		Exports to Developing Countries	Imports From Developing Countries	Trade Balance
United States	1970	0.8	0.3	0.5
	1980	1.9	1.1	0.8
	1990	1.5	2.0	-0.5
	1995	2.1	3.2	-1.1
Japan	1970	3.2	0.5	2.7
	1980	5.2	0.7	4.5
	1990	3.1	1.0	2.1
	1994	3.4	1.2	2.1
West Germany	1970	4.1	1.6	2.5
	1980	5.4	2.2	3.2
	1990	4.8	3.2	1.6
France	1970	2.6	0.8	1.9
	1980	3.6	1.1	2.5
	1990	2.9	1.6	1.3
	1995	3.4	2.0	1.4
United Kingdom	1970	4.1	1.3	2.8
	1980	4.8	2.1	2.7
	1990	3.2	2.9	0.3
	1994	3.8	3.4	0.4
Canada	1970	1.0	0.3	0.7
	1980	1.5	0.7	0.9
	1990	0.8	1.4	-0.6
	1995	1.4	2.7	-1.4
Italy	1970	3.3	0.9	2.3
	1980	7.1	1.2	5.8
	1990	5.0	1.5	3.5
	1994	9.8	2.0	7.8

Sources: UN TARS trade database, IMF International Financial Statistics.

Table 3 shows the huge disparities in labor costs across countries, with the example of hourly textile labor costs (wages and fringe benefits).⁸ Wages in the North are much higher than those in the South, although wages in the South themselves vary greatly too. For medium-income countries like Korea and Mexico, labor compensation levels have reached about one quarter to one third of those in the United States, while low-income countries such as India and Bangladesh have wage rates that are under 5-10 percent of the U.S. levels. Many in the North view such low wages as both inhumane and unfair, but economic analysis reveals that these concerns are greatly exaggerated, as discussed next.

Table 3.

International Comparison of Labor Costs in the Textile Industry, Selected Countries, 1993

Source: Textile World, 1994.

Table 3.

International Comparison of Labor Costs in the Textile Industry, Selected Countries, 1993

	Hourly Cost (U.S.$)	Ratio (U.S. = 100)
Japan	23.65	204
West Germany	20.50	177
Canada	13.44	116
United States	11.61	100
Australia	10.84	93
South Korea	3.66	32
Mexico	2.93	25
Malaysia	1.18	10
Thailand	1.04	9
Philippines	0.78	7
India	0.56	5
China	0.36	3
Bangladesh	0.23	2

Source: Textile World, 1994.

View Table

Table 3.

International Comparison of Labor Costs in the Textile Industry, Selected Countries, 1993

	Hourly Cost (U.S.$)	Ratio (U.S. = 100)
Japan	23.65	204
West Germany	20.50	177
Canada	13.44	116
United States	11.61	100
Australia	10.84	93
South Korea	3.66	32
Mexico	2.93	25
Malaysia	1.18	10
Thailand	1.04	9
Philippines	0.78	7
India	0.56	5
China	0.36	3
Bangladesh	0.23	2

Source: Textile World, 1994.

B. Gains and Losses from North-South Trade

Comparative and Absolute Advantage. Contrary to popular fears, international differences in wage rates do not preclude mutually beneficial trade. On the contrary, international trade theory suggests that the benefits of trade increase with economic diversity between partners. The fear that high-wage countries are unable to compete with low-wage countries (the so-called “sweatshop labor” argument) confuses the fundamental distinction between comparative and absolute advantage. Economic theory teaches us that overall differences in productivity (absolute advantage) determine wages, while sector-specific variations in productivity and costs determine trade patterns.⁹

To the extent that low wages reflect low labor productivity, any advantage to employing low-wage labor is offset. The “sweatshop labor” argument for protection ignores the connection between wages and productivity. Even if nominal wages are rigid (or artificially repressed, as is often alleged), exchange-rate adjustments can eventually be expected to offset international differences in costs. Indeed, as noted above, most developing countries continue to run trade deficits in manufactures with the industrial countries, which would be unlikely if their unit labor costs in manufacturing were as low, relative to the industrial countries, as their wages.

A related concern is that the acquisition of foreign technology and capital by the South will equalize productivity in low- and high-wage countries. It is true that capital mobility allows developing countries to import capital and technology more easily and thus converge more rapidly to industrial country productivity levels. But in practice capital flows seem to play a limited role in equalizing productivity across countries. Even with the increasing ease of technology transfer and capital mobility, other factors can hold down productivity in poor countries, such as low levels of human capital and poor public infrastructure and transportation services. In addition, international capital flows to poor countries are generally not large enough to rapidly alter factor proportions in these countries. Despite the increasing degree of financial market integration across the world, the bulk of domestic investment is still financed by domestic saving.¹⁰ In any case, if productivity growth did accelerate in some low-wage countries, the advantage would most likely be reduced by compensating wage gains or exchange-rate appreciation.

Recent research (Golub 1995, 1996) confirms that aggregate manufacturing labor costs and labor productivity are related, although exchange rate movements and other macroeconomic shocks can sometimes induce substantial short-run deviations.¹¹ Figure 1 shows, relative to the United States, manufacturing wages,¹² labor productivity, and unit labor costs (the ratio of wages to productivity) for several emerging market economies in 1990. Productivity and wage levels are closely related in most cases, i.e., international differences in unit labor costs are invariably much smaller than differences in wage rates suggest. In fact, calculations indicate that unit labor costs in several emerging market economies in 1990 were actually higher than those of the United States, i.e., the productivity gap is even bigger than the wage gap. There is also evidence that wages and productivity generally move together over time. For example, Korea had the most rapid productivity growth, and it also had the highest wage growth. One major exception is Mexico, where real wages dropped dramatically and the peso crashed in the aftermath of the 1982 debt crisis. Overall, however, the evidence in Golub (1995) shows a close connection between wages and productivity at the level of aggregate manufacturing.

View Full Size

Developing Country Labor Productivity, Compensation, and Unit Labor Costs, Relative to the United States, 1990

(United States = 1.0)

Citation: IMF Working Papers 1997, 037; 10.5089/9781451845532.001.A001

Source: Golub (1995).

• Download Figure
• Download figure as PowerPoint slide

International differences in wage rates are entirely compatible with mutual gains from trade, which is based on comparative advantage--sectoral differences in productivity and costs. For example, Malaysia’s wages and productivity in aggregate manufacturing were both about 10 per cent of the U.S. level in 1990, but it seems likely that the Malaysia/U.S. productivity ratio would be greater than 10 per cent in simple products such as textiles and less in sophisticated products like machinery. If skill-adjusted wages are about the same in all industries in Malaysia, as they should be in a competitive labor market, Malaysian relative unit labor costs will be low in textiles and high in machinery. If so, Malaysia will export textiles and import machinery. Golub (1995) provides empirical evidence supporting this approach to comparative advantage. Of course, there are other determinants of comparative advantage besides sectoral labor productivity and labor costs. The important point is that both developed and developing countries benefit from specializing in what each produces relatively efficiently.

Distributional Effects. Economic theory also suggests that, even though the North as a whole gains from trading with low wage-countries, imports of labor-intensive products are likely to be harmful to unskilled workers in the North. The adverse effect on unskilled workers derives from two sources. First, the Stolper-Samuelson theorem shows that even with perfectly competitive labor markets, unskilled labor will lose from trade as the price of labor-intensive goods declines.¹³ Second, international competition may reduce wages in highly-unionized and imperfectly-competitive industries where economic rents were previously shared by workers and firms (Borjas and Ramey 1995).¹⁴ But, as Krugman (1995) argues, the small size of the imports of manufactures from developing countries shown in Table 2 seems hard to square with a dominant role for such trade.¹⁵ In any event, almost all empirical estimates suggest that trade accounts for at most one quarter of the decline in unskilled workers’ real wages in the United States.¹⁶

If North-South trade has played a relatively small part in the deterioration of the position of unskilled labor in the North, why then has it been receiving so much attention? Perhaps the answer is that it is politically appealing and convenient, and on the surface plausible, to blame imports from low-wage countries, so such trade becomes a lightning rod for the frustrations of workers, even if its importance is swamped by domestic factors. Also, there is the prospect of expansion of North-South trade, especially as China and India become major exporters. It is within the context of these controversies that the debate about international labor standards is best understood.

A. Economic Arguments

There are several frequently-advanced economic arguments for international harmonization of labor standards: (1) There are international market failures requiring joint action by governments. (2) Countries with low standards have an unfair competitive edge in international trade. (3) Partly as a result of the latter, there will be a “race to the bottom” as countries attempt to attract capital inflows by competitive lowering of standards. (4) Developing countries’ standard of living will benefit from external pressure to raise labor standards.

International market failures. In the case of some types of environmental damage, e.g., depletion of the ozone layer, there is a clear need for international coordination to deal with a global public good. Some argue that adherence to labor standards is also a public good insofar as individuals find some practices abhorrent but are unable to curtail these practices collectively due to the free rider problem—the incentive for each individual to avoid bearing the cost and to “free-ride” on the efforts of others. Others argue that international labor standards are not public goods, since individuals are free not to purchase goods made under conditions they disapprove of.²⁹ According to the view that labor standards are private goods, the main problem is how to disseminate the appropriate information about production conditions. Also, trade sanctions are not a desirable means of achieving this end, as they may simply lower the gains from trade without directly addressing the underlying market failure. These issues are considered further in Section V.

Unfair competition. Weak labor standards are alleged to artificially hold down labor costs for two reasons. First, weak core labor standards reduce the bargaining power of labor. Second, low labor standards directly lower non-wage components of labor costs. As discussed above, several studies show a high correlation between wages and productivity in developing countries, casting doubt on the first proposition. It is mainly the second proposition that is considered here.

Many economists are skeptical of the popular fear that low labor standards provide an unfair competitive advantage in international trade.³⁰ Indeed, this is just a slightly modified version of the “sweatshop labor” argument discussed above. Instead of wages, the unfair standards argument often focuses on fringe benefits and safety conditions, but it reflects the same confusion of comparative and absolute advantage discussed in Section II. Moreover, wages are the most important component of labor costs and compensation even in developed countries, so public outrage about working hours or health and safety conditions is misdirected. For a given level of labor costs, the division between wages and fringe benefits is largely irrelevant for international competitiveness, although it may not be a matter of indifference to workers.

In addition, as Ehrenberg (1994) and Lawrence (1996) point out, mandating increases in fringe benefits such as vacation and maternity leave may simply alter the composition of labor compensation and not its total size. This is the case, for example, if fringe benefits are paid for through payroll taxes, and the supply of labor is inelastic, so workers bear the entire burden of the payroll taxes. According to Ehrenberg (1994) the empirical evidence suggests that the incidence of payroll falls mostly on labor in the long run.³¹ If so, the level of fringe benefits has no effect on labor costs.

In contrast, some labor market standards, e.g., child labor laws, are likely to alter effective labor supplies, and therefore affect equilibrium labor costs. Even where they do affect labor costs, weak labor standards, like low wages, are likely to be a consequence of low productivity and poverty, not an independent source of international competitiveness and wage determination. As Bhagwati and Srinivasan (1996) emphasize, economic development, cultural differences, or other national preferences may give rise to different choices about labor standards. These national characteristics, along with factor endowments and other more traditional aspects, determine national comparative advantage. Low labor standards and low wages both reflect the abundance of unskilled labor and low productivity in developing countries. Some workers, even children, may understandably choose to work long hours under unsafe working conditions relative to Northern standards, given the alternatives available to them. Forcing poor countries to adopt and enforce tougher labor standards may raise their labor costs if the implementation of such standards in effect lowers the labor-intensity of production. Therefore, high labor standards imposed on a poor country are like a tax on labor-intensive production and reduce the comparative advantage of the country in this sector.³² In this sense, imposition of high labor standards on conditions of work is tantamount to protectionism, as many developing countries argue.

Several recent studies provide some empirical evidence on the effects of labor standards on wages and trade by developing quantitative measures of adherence to labor standards and relating the latter to other variables. Rodrik (1996) uses a variety of measures of labor standards, including number of ILO Conventions ratified, use of child labor, statutory hours of work, annual days of leave, and percentage of the labor force that is unionized. Focusing on manufacturing, he finds that his measures of labor standards are positively associated with wages but have little effect on trade patterns. He fails to note, however, the seeming contradiction between these two findings. If labor standards have large effects on wages relative to productivity, why don’t they have any effect on trade patterns? Rodrik may be obtaining large wage effects because he does not fully control for productivity as the underlying determinant of wage differences.³³ In any case, he finds that most of his labor standard variables are insignificant determinants of comparative advantage in labor intensive goods. OECD (1996) found little evidence of weak labor standards on either wages or trade patterns, using indexes of freedom of association as their measure of labor standards. Examining bilateral U.S. trade with ten major developing countries, Aggarwal (1995) finds no correlation between growth rates of imports in the United States and adherence to core labor standards in the exporting country. She also fails to find strong evidence that the sectors most exposed to competition from imports allegedly produced under substandard labor conditions such as textiles, carpets, footwear, and toys, have fared unusually poorly in terms of changes in wages and employment in the U.S.³⁴ In short, the available empirical evidence, although rather limited, does not provide support for the view that weak labor standards have much effect on trade patterns.

Race to the Bottom. Perhaps the deepest fear of many in the North, is that low Southern labor standards act as a magnet for capital inflows, and lead to a “prisoner’s dilemma” situation where the non-cooperative outcome is a competitive lowering of labor standards in all countries. For example, Barnet and Cavenagh (1994, p. 49) write that “nations that observe basic worker rights are losing out to regimented societies where employees do not have the right to organize or bargain collectively, much less to strike.” Langille (1996, p. 253) argues that “our [Canadian] ability to establish our own domestic labor standards has been eroded.”

It is in principle possible that such a prisoner’s dilemma situation could emerge, but it is far less plausible than the popular fears suggest. Again, if low labor compensation reflects low productivity, there is no unfair competitive advantage.

Wilson (1996), using the literature on local public finance, provides an illuminating perspective on the alleged benefits of using lax standards to attract footloose inputs like capital. First, if there are no market failures, and capital earns its true social marginal product, there is no reason to attract capital by depressing labor standards. Doing so would introduce a distortion which entails net losses for the home country. Second, if market imperfections exist and give rise to an excess of the social return to capital over the private return, direct capital subsidies are a more effective and less distortionary way of attracting capital. Why relax labor standards that may be valued for other reasons when a more direct method is available? Third, market imperfections are as likely to give rise to a race to the top as to the bottom, since the social rate of return to capital can fall short of the private return. Wilson (1996) labels this NIMBY (not in my back yard). In short, there is little reason for a country to deliberately depress labor standards as a way of attracting increased foreign investment.

The available empirical evidence suggests that foreign investment is not highly responsive to weak labor standards. Rodrik (1996) finds that foreign investment is actually negatively correlated with weak labor standards. OECD (1996) notes that the scarcity of evidence requires caution, but that labor standards do not seem to play a significant role in investment decisions. OECD (1996) does point out that in a few countries, export processing zones have lower labor standards than in other manufacturing. Against this, however, Aggarwal (1995) observes that labor standards in most developing countries, especially for child labor, are higher in the formal sector, where foreign investment is confined, than in subsistence agriculture and small-scale services. Also, based on U.S. Department of State reports, Aggarwal (1995) points out that working conditions in U.S. owned plants in developing countries are often far better than in domestically-owned plants, because multinationals apply internationally-recognized labor standards. In this context, limitations on unions in export-processing zones may be an effort to keep the price of labor close to its true social opportunity cost rather than at the inflated level in manufacturing. Relatedly, there is little evidence that countries use weak environmental standards to attract investment (Levinson 1996). Finally, as mentioned in Section II, the volume of capital inflows to developing countries remains small relative to domestic saving and investment, so even if capital flows are responsive to international differences in labor standards, the overall effect on capital accumulation is likely to be minor.

Welfare of Southern workers. Human-rights activists are often genuinely motivated by concern over the welfare of workers in the South. Regardless of their motivation, Northern pressures to impose higher labor standards on the South may sometimes be detrimental to Southern workers. The following points apply especially to labor standards regarding conditions of work such as minimum wages, health and safety benefits, hours, and possibly minimum age of employment. They are less relevant to core labor standards such as bans on forced labor and on discrimination, which are founded more on moral rather than economic concerns, as discussed below.

Everyone agrees that improvements in the standard of living in the South are desirable. The question is how to achieve this. Fields (1990) points to the differences between “labor market processes” and “labor market outcomes”. Attempting to mandate certain labor market processes, such as minimum wages, under conditions where they are inappropriate, may sometimes be counterproductive in terms of the ultimate objective of improving labor market outcomes. Consider again the difficult problem of child labor. As Freeman (1994, p. 89) observes, “Better that they work and eat than starve.” Fields (1995, p. 12) and others make the same basic point: most developed country labor standards are simply “unattainable for most of the world’s people.”

An alternative view, argued by Piore (1990) for example, is that there are multiple equilibria in economic development, and that labor standards can push an economy out of an undesirable “sweating” strategy where firms maintain low labor costs and do not seek productivity improvements. Forcing labor costs up induces technological change and growth, according to Piore. In commenting on Piore, Srinivasan (1990) concedes that economic models can generate such multiple equilibria, but he doubts their relevance for understanding economic development. It seems far more plausible that higher labor standards are primarily a consequence rather than a cause of economic growth.³⁵ Krueger (1996), for example, shows that the use of child labor is strongly inversely related to economic development.

As Srinivasan (1990), Portes (1994) and Fields (1995) observe, many developing countries do in fact have detailed labor standards, but these may cause more harm than good. In the NAFTA debate, it was widely recognized that Mexico’s legislated labor standards are in fact stronger than those of the United States and Canada, although the enforcement is often weak (Weintraub and Gilbreath, 1993). Labor market regulations such as minimum wages are often cited as a major contributor to the problem of economic dualism in developing countries, i.e., excessively high wages and high unemployment in the urban areas (Portes 1994). As Srinivasan (1990) puts it, “premature enforcement through public policy of unrealistic standards… creates an island of labor aristocracy in the organized sectors…” Fields (1990, 1995) argues that the high growth rates of the successful East Asian countries entailed high wage growth, despite restrictions on labor rights and relatively weak labor standards. As Lim notes, ³⁶

In Malaysia, labor organization has been restricted in the export-oriented electronics industry for more than twenty years. Singapore has virtually outlawed strikes for thirty years, and its unions are controlled by the government…Neither country has ever had a minimum wage….. Yet …wages, hours of work, working conditions, fringe benefits and social protections are superior by developing-country standards….In Thailand and the Philippines, on the other hand, labor has been free to organize since the late 1980s, and both have a minimum wage. But labor outcomes leave a great deal to be desired…Child labor (including child prostitution), serious workplace health and safety violations…are persistent problems in both countries.

According to this view, pursuing trade and labor-market polices which are conducive to high growth rates will be far more effective in raising labor incomes than mandating levels of wages and benefits. Such growth-oriented labor-market polices need not be laissez-faire, as Lim (1990) illustrates in the case of Singapore. Government investments in education and training in particular are likely to be beneficial. Loss of access to markets in the developed world will hamper developing-country growth prospects and thereby retard the upgrading of labor standards.

National labor market policies may well sometimes enhance both efficiency and equity by remedying market failures (World Bank 1995). But even if it is in the interest of developing countries to upgrade their labor standards, why should this problem not be left to each country to decide for itself, given that the case for adverse international spillovers is not strong? One frequent argument (e.g., Barnet and Cavenagh 1994, Langille 1996) is that the non-democratic political regimes in much of the South prevent the expression of the public’s true preferences. Bhagwati and Srinivasan (1996) object that even in democratic developed countries, such “political market” failures exist. Even if is accepted that a developing country is not democratic, it does not follow that developed countries can legitimately impose higher standards, and that doing so will raise welfare in the developing country. Instead, a more direct and effective method of expanding the public’s voice in the setting of standards is to encourage the spread of democracy in the developing world.³⁷

Finally, if the goal is to enhance the welfare of the South, trade sanctions hardly seem an appropriate and effective approach. Srinivasan (1995) points out that the most effective way of raising standards for workers in developing countries would be to allow free international migration from low to high standard countries, an argument rarely made by proponents of harmonizing labor standards. Even if immigration is judged infeasible, direct income transfers, which could be used for upgrading living standards in the South, would be more consistent with a humanitarian motivation as well as more economically efficient, as Srinivasan (1995) shows formally.³⁸

In summary, there is considerable danger that workers in the South would be harmed rather than helped by international harmonization of standards related to conditions of work, especially if they are enforced through trade sanctions. The surest way to improve labor standards in the South is economic growth, which international trade facilitates.

B. Social and Political Concerns

The economic arguments for harmonizing labor standards are weak. It seems clear that the popular fears regarding “social dumping” are greatly exaggerated and that much of the force of the arguments derive from misunderstanding of the relationships between trade and wages, as well as from interest groups in the North who have obvious reasons for wishing to restrain Southern exports of labor intensive goods. The basic reality is that the developing countries’ comparative advantage in unskilled-labor intensive goods and their low labor standards are both manifestations of their level of economic development and large supplies of unskilled labor. Low labor standards are not the primary source of developing countries’ comparative advantage.

In some cases, however, there are genuine moral concerns. Some labor practices, most obviously slavery, are virtually universally regarded as unacceptable. As noted above, this has led the ILO and OECD to focus on core labor standards which uphold basic labor rights. As famously argued by Okun (1975, Chapter 1) there is a tension in market-based democracies between markets and universal social rights.³⁹ Democracies strike a balance, limiting the domain of markets, e.g., prohibitions on buying and selling votes or children, despite the economic inefficiencies such restrictions entail. Of course, there is a fuzzy and shifting boundary on the acceptability of markets, but according to Okun, the “market has a place, and the market must be kept in its place” to maintain the legitimacy of the society. This point applies with particular force to international trade. Popular support for open trade, especially with developing countries, is likely to be more fragile than for many other market relationships. Rodrik (1994), Lawrence (1996), and Freeman (1994) justify some international harmonization of labor standards with arguments of this type. Freeman notes that all countries have at least some labor standards, and thereby reveal a collective willingness to pay for the costs such standards impose. Likewise, consumers of imported products often care at least a little about the working conditions under which the goods are produced.

Bhagwati (1995, p. 754), however, cautions that “universally-condemned practices such as slavery are rare indeed.” Even the case for bans on imports made by foreign prison labor are debatable, according to Fields (1995): why not let prisoners defray the costs of their crimes by working? In this regard, some view forced labor by political prisoners as much more objectionable than for common criminals, but the more fundamental problem is the existence of political prisoners rather than the fact that they are forced to work. As noted earlier, other core labor rights are also less universal than sometimes claimed. If “freedom of association” means unions that exacerbate wage inequalities, this may be undesirable. Disparities in income levels as well as cultural factors give rise to wide differences in acceptable behavior in labor markets, e.g., for child labor. Discrimination between men and women is acceptable in some Islamic societies. Without a clear basis for international agreement, according to Bhagwati (1995), labor standards will degenerate into protectionism.

Another argument against harmonization of basic labor rights is the arbitrariness of the focus on labor rights to the exclusion of other human rights problems, such as lack of freedom of expression. As Klevorick (1996, p. 463) puts it, “… if the citizens of one country are concerned about the well-being of populations elsewhere—if the condition of distant populations enters the utility functions of citizens of a given country—why should that concern manifest itself with respect to specific aspects of the distant folks’ lives and not with their overall well-being, the utility or welfare they perceive themselves as achieving?” The focus on labor rights rather than human rights in general leads one to suspect that economic considerations more than ethical concerns underly some of the calls for harmonization of labor standards.

Summary. The economic and moral dimensions of harmonization of standards cannot be completely separated. While lack of adherence to core labor standards is surely not the main source of developing countries’ comparative advantage in labor intensive goods, and trade is not the main source of unskilled workers’ difficulties, violation of such standards in the South undermines support for free trade in the North, at a time when unskilled labor is faring poorly and import penetration of Southern manufactures is rising rapidly. Increased international compliance with a few core labor standards can therefore be defended as a way of increasing the legitimacy of a liberal international trading system, as well as being desirable in itself. The following section assesses alternative approaches.

A. A Limited Social Clause

A number of authors (e.g., Leary 1996, Rodrik 1994, 1996) favor a social clause which would allow trade sanctions for violations of fundamental labor rights, but would avoid mandating harmonization of employment conditions such as minimum wages and working hours.⁴⁰ Proponents of a limited social clause generally agree that such a clause should be negotiated multilaterally either at the ILO or the WTO rather than adopted unilaterally.

Rodrik (1994, 1996) proposes a “social safeguards” clause for the WTO which focuses on basic labor rights. To avoid protectionist capture of this provision, Rodrik would establish a demanding set of conditions which would have to be met before trade sanctions are imposed. First, the social concern at issue would have to be broadly shared by export and consumer interests in the importing country. Second, import restrictions would have to be judged effective in furthering the objective. Third, if the exporting country is democratic, the importing country would have to offer compensation for any trade sanctions. Unfortunately, it would not be easy to establish any of these criteria, and abuse seems likely, as Anderson (1995) points out. International agreement on such a subject appears to be a daunting task. As discussed above, differences of views and difficulties of implementation explain why the ILO has eschewed sanctions. It is hard to see how the new proposals would fare better, given the strong opposition by developing countries and genuine disagreements about what constitute basic human rights. Myrdal (1994) argues that this divisive issue could spell the demise of the ILO. ILO officials themselves have expressed concerns that the use of trade sanctions could disrupt the consensus-oriented approach of the ILO (Maier 1994). The ILO took up this question at it International Labor Conference in 1994 which was marked by “vigorous re-statement of conflicting views.” (Hansenne 1996). The ILO now feels that greater mutual understanding has been reached, however, and is “reasonably optimistic that it will be possible to agree on a number of common rules, even though these common rules may be a far cry from the kind of social clause originally proposed.” (Hansenne 1996, p. 4). New or enhanced roles for the ILO are considered in the next section.

According to Bhagwati and Srinivasan (1996), application of trade sanctions is almost always undesirable, but they can be consistent with existing WTO procedures, as long as compensation is offered. Bhagwati (1995, p. 758) puts it as follows: “Thus, unless one resents having to pay for one’s virtue…, this is a perfectly sensible solution even to politically-unavoidable unilateralism: do not import glass bangles made with child labor…, but make some other compensatory trade concession.” Bhagwati also argues that existing international fora are adequate when there is an international consensus for stronger sanctions to respond to truly egregious cases.

B. Voluntary Initiatives

Enhancing the role of the ILO. As discussed earlier, the ILO has a long history of dealing with international labor standards. The ILO stresses non-enforceable compliance with its Conventions, and works closely with individual member countries to arrive at mutually satisfactory outcomes. Given the diversity of domestic circumstances, especially the level of economic development, which affect the appropriate level of labor standards, this flexibility is essential.

The ILO’s supervisory capacities are limited to cases where a country has ratified the Convention in question, except for violations of freedom of association. To the extent that countries with the worst abuses fail to ratify Conventions, the ILO cannot address the most important cases. The ILO has considered proposals to extend the special procedure on freedom of association to other core labor rights such as prohibition of forced labor and non-discrimination, but this has encountered opposition and little progress has been made so far, according to OECD (1996). The ILO is also considering asking member countries to submit regular reports explaining the obstacles to ratification of Conventions in cases where they have failed to do so. This appears to be a promising approach to increasing the number of ratifications, which is a priority for the ILO. Further consideration and debate on these matters is scheduled for the Governing Body meeting in March 1997 and the International Labor Conference in June 1997.

The ILO’s role could also be enhanced through the use of its standards in new private sector initiatives, such as product labeling and codes of conduct for multinationals operating in developing countries.

Product Labeling. Freeman (1994) argues that labor standards in developing countries are largely private goods that consumers in the developed countries can obtain in the marketplace if they are willing to pay for them, provided they have the requisite information. Freeman (p. 81) points out that some consumers will willingly pay more for goods produced in conformity with basic labor standards. Through organized boycotts or individual purchasing decisions, consumers can set a market price for labor standards. The main market failure is consumers’ lack of information about workplace conditions. Freeman notes that public or private organizations could help fill this gap. Indeed, human rights organizations already do this.

A nice illustration of this nascent consumer movement is provided by the recent controversy in the United States over clothes made in Honduras.⁴¹ The National Labor Committee (NLC), a human rights organization, attacked the host of a popular television program, for clothes bearing her name produced by Hondurans working for “slave wages.” Following much press attention, the TV host claimed ignorance of these practices and publicly endorsed efforts to improve working conditions where her clothes are produced. This story shows that human-rights activists and the media may sometimes be very successful in drawing attention to alleged abuses.⁴² Rodrik (1996) cites other recent examples of companies which have certified that their products are not produced by child labor or under other objectionable conditions, e.g., Levi Strauss, Walmart, lkea.

While recognizing that product labeling may be useful in some cases, Rodrik (1996) and OECD (1996) view it as at best a partial solution. Rodrik raises the following objections to labeling.

1. Monitoring by consumers is likely to be difficult,

2. Compliance is limited since individual consumers have an incentive to buy non-certified cheaper goods (the free rider problem),

3. Labor standards are too broad to be captured by simple labels,

4. At the domestic level, regulations go well beyond labeling, presumably because labels are viewed as insufficiently effective and stringent.

Rodrik’s scepticism about the effectiveness of labeling, however, seems to underestimate the power of the consumer movement, illustrated by the above example. Freeman (1994) points out that producers may be very responsive to such concerns as even moderate losses of market share could strongly affect profitability. Moreover, firms show through their philanthropic contributions and some of their advertising that public image is important to them. They therefore have an incentive to monitor themselves, thus obviating some of Rodrik’s concerns. Also, the zeal of human rights activists may offset the difficulties posed by the free rider problem, according to Bhagwati and Srinivasan (1996). In fact, a danger is that consumer boycotts could be based on unfounded or exaggerated claims about labor practices. As Hansenne (1996) points out, the ILO can play an important role by disseminating accurate information. Finally, the consensus for more binding regulations that may prevail at the domestic level, does not exist at the international level. Under these circumstances, the flexible and voluntary nature of labeling is a huge advantage relative to government regulation.

Corporate Codes of Conduct. A complementary proposal, forcefully made by Jagdish Bhagwati, is for multinational companies to adhere to the same labor standards abroad that they follow at home, either unilaterally or through an OECD code.⁴³ An advantage of such a code is that the most potent objections to North-South trade occur when “plants are closed by one’s own multinationals and shifted abroad” (Bhagwati and Srinivasan, p. 178). Bhagwati (1995) draws a parallel with the Sullivan Principles, which set guidelines for U.S. firms operating in South Africa under apartheid.

While advocates of labor standards argue that the Bhagwati proposal is too weak, Fields (1995, p. 15) objects that the Bhagwati proposal may be too stringent. As stressed earlier, in some cases it is completely impractical and counterproductive to mandate that U.S. companies follow American labor standards in developing countries; e.g., U. S. minimum wages obviously should not be applicable to Mexico. Fields asks why it should not be up to Mexico to set such standards.

Certainly, some developed-country standards setting minimum wages and working conditions are inappropriate for developing countries. But in other cases, such as child labor laws and worker safety, the case for the Bhagwati proposal is stronger. While developed country standards may impose costs from a strictly economic point of view, instituting codes of conduct regarding basic labor rights provides offsetting benefits by responding to the moral and political problems posed by certain practices.

Codes of conduct regarding basic labor rights, and some particularly controversial workplace practices, appear to be one of the least intrusive and most promising approaches towards addressing the calls for international harmonization. Compliance should be voluntary, but the importance of the moral pressures and product image for companies should not be underestimated. Bhagwati (1995) argues that the ILO rather than the WTO is best suited for this task, as the ILO has experience and expertise in this area, and a role for the WTO would open the door for trade sanctions.