Journal Issue

Wages, Inequality, and International Integration

International Monetary Fund. External Relations Dept.
Published Date:
January 1995
  • ShareShare
Show Summary Details

DOES the growing integration of the world economy offer an opportunity or pose a threat to the world’s workers? The outcome depends on whether governments pursue market-based policies that take advantage of international opportunities.

Earnings differentials across countries are huge. Adjusted for differences in their currencies’ purchasing power, the wages of engineers in Frankfurt, Germany, are 56 times those of unskilled female textile workers in Nairobi, Kenya. The most prosperous group of workers in the world, the skilled workers of the industrial countries, now earn some 60 times more than the poorest group, the farmers of sub-Saharan Africa. These differences reflect, in part, the occupational pay structure within each domestic economy—skilled workers earn several times the wage of unskilled workers in almost every country. But they can be traced primarily to huge international differences in returns to work (see chart).

Recently, there has been rapid movement toward market-based development strategies and growing international integration of markets. In the 1970s, about two thirds of the world’s labor force lived in countries that were largely insulated from international markets by prohibitive trade barriers and capital controls, or by planned trade. Today, three giant population blocs with nearly half the world’s labor force—China; the Baltic countries, Russia, and the other countries of the former Soviet Union; and India—are entering the global market, and many other countries, from Mexico to Turkey, have already established deep linkages. By the year 2000, less than 10 percent of workers may be living in countries that are disconnected from world markets.

Earnings differ tremendously

Source: Union Bank of Switzerland, 1994.

1Thousand dollars per year, converted at purchasing power parity exchange rates.

What does a more market-driven and economically integrated world mean for workers? Does growing integration offer an opportunity or pose a threat, especially to those in the world’s poorest regions? Will inequality between rich and poor countries continue to grow, or will we witness a worldwide trend toward convergence?

Two views on convergence

The prospect of China and other labor-rich countries entering the global market strikes fear in the hearts of unskilled workers around the world, from Indonesia to the United States, who see China’s huge, low-wage workforce as a threat to their incomes. This widespread perception, loosely based on economic theory, has fueled protectionist sentiment in both rich and poor countries. But the predictions of the theory are not so clear cut. There are two views of wage convergence which typically influence the debate over the effects of international integration.

The first view, based on the well-known Heckscher-Ohlin model of trade, predicts that increased trade between industrial and developing countries will narrow the gap between wages in the North and South. The logic behind this is that trade affects the relative rewards of factors of production by changing the relative price of goods. Opening up to trade increases the price of labor-intensive goods in poor, labor-rich countries, which, as a consequence, shift their resources to the production of labor-intensive goods. This, in turn, raises demand for labor in poor countries, and hence raises relative wages. As relative prices of goods converge in rich and poor countries, so do wages. Whether wages converge upward (as poor countries’ wages and living standards move toward those in rich countries) or downward (the reverse), depends very much on specifics—for example, on whether countries are building their endowment of skills and hence shifting their comparative advantage toward higher wage products. But, broadly, this model predicts that convergence will involve both a degree of poor-country wages being pulled up and rich-country wages being pulled down, at least for unskilled workers.

A different view of convergence suggests that there are advantages to backwardness that arise because poor countries can take advantage of technical innovations developed in rich countries. This view of convergence pictures poor countries as “catching up” to rich ones, with wages of poor workers in the developing countries being pulled up to industrial country levels (see box).

What is the evidence?

These two views on convergence have different implications for what happens to the wages of rich-country workers, but both point toward an equalization of incomes between North and South. Is this borne out by the evidence?

There are three empirical trends that are relevant to this question: increasing inequality within the industrial countries, reflected either in widening wage inequality or in rising unskilled unemployment; a disparity of experiences among the developing countries, with inequality falling sharply in some and increasing in others; and divergence in incomes among the richest and the poorest countries.

Since the 1970s, the industrial countries have struggled with two worrisome labor market trends: rising wage inequality in Australia, Canada, and the United States, and growing unemployment throughout Europe. In the United States, the jump in inequality has been remarkably sharp: average wages of young men with only a secondary education fell by 20 percent between 1979 and 1987, even as those of young male college graduates increased by 11 percent in real terms. In Europe, wage-setting mechanisms meant to reduce inequality have contributed instead to high unemployment: there are now 20 million unemployed workers in the European Union, about 12 percent of the labor force—most of them unskilled workers.

There is no doubt that part of these developments is related to increased competition from developing countries’ imports—the difficult question is how much. Most analyses conclude that trade with developing countries can explain only 10 to 30 percent of the industrial countries’ labor market difficulties, but some studies come up with more extreme results—on both sides of the argument. Total imports from developing countries were only about 3.5 percent of GDP in industrial countries in 1992; imports of manufactures were only about 2 percent of GDP. Even allowing for the unusually labor-intensive nature of the goods concerned, the direct effect of this trade on industrial countries’ workers must have been limited. Factor content calculations made by Adrian Wood (North-South Trade, Employment and Inequality: Changing Fortunes in a Skill-Driven World, Oxford, England, Clarendon Press, 1994) suggest that trade with developing countries during the past two decades reduced the demand for unskilled workers by between 3 million and 9 million, or 1 percent and 3 percent of total employment (2 percent to 5 percent of the unskilled labor force), depending on whether the factor proportions used in the computations are those of industrial or developing countries.

But these estimates do not account for either defensive labor-saving innovation by firms that respond to the threat of cheap imports by finding new production methods that use less unskilled labor, or the displacement of unskilled labor in services and non-traded sectors that supply intermediate inputs to manufacturing. There is no precise way of quantifying these effects. But for trade with developing countries to account for all of the labor market developments in the industrial countries, the estimated upper range of the direct impact of trade would have to be quadrupled.

Rising wage inequality within skill groups, and increases in the ratio of skilled to unskilled employment in all sectors also suggest that some other force is at play. In particular, technological change seems to be increasingly labor-saving—perhaps partly because of increased international competition.

The Heckscher-Ohlin trade theory suggests that the impact of trade on the wage distribution in developing countries should be the mirror image of what we observe in the industrial countries. While the relation between increased trade and falling poverty is well established, the impact of openness on wage disparities in developing countries has not been homogeneous. In East Asia, and in countries such as Morocco, increased trade has coincided with increased equality. But in some Latin American countries, such as Argentina, Chile, and Mexico, a more recent wave of trade liberalization has coincided with increased wage and income inequalities. The word “coincided” needs to be emphasized in both cases since analysts are still disentangling the various effects at play. In Mexico’s maquiladora (assembly) enterprises, the ratio of nonproduction to production wages rose from 2 to 2.5 between 1985 and 1988. And, in Chile, the wages of university graduates rose by 56 percent relative to those of high school graduates between 1980 and 1990.

Can one conclude that trade increases wage inequality in industrial and developing countries alike? Most likely no. Rising trends in inequality may be related to other factors, such as the introduction of new labor-saving technologies, particularly the computer revolution, or the rise of the service sector. For developing countries, there may be additional explanations: shortages of well-trained, skilled workers immediately following liberalization; complementarities between inflows of new capital and skills; and, in some middle-income countries, competitive pressures from lower-wage unskilled workers in poorer countries.

Are poorer countries catching up with richer ones?

Are there advantages to backwardness? Or are rich countries getting richer while the poor get poorer? And, what is the role of international integration in allowing the poorer countries to catch up? The debate about convergence is based on a search for such historical regularities. Careful empirical work suggests that absolute divergence in output per person is a dominant feature of the world economic scene, but that conditional convergence forces are also at work.

Divergence in per capita income is the dominant feature of modern economic history. According to estimates made by Lant Pritchett (“Divergence, Big Time,” unpublished, World Bank, 1995), the ratio of per capita income in the richest versus the poorest countries has increased from 11 in 1870, to 38 in 1960, and to 52 in 1985, and overall international inequality rose slightly between 1960 and 1986. This divergent relation between growth performance and the initial level of per capita income is empirically valid on average over a sample of 117 countries. The first column in the table reports the result of relating per capita income growth to the initial level of per capita income for the sample of countries. On average, the empirical results suggest that countries that started richer grew faster.

However, cross-country econometric work also indicates that “conditional convergence” is occurring. When per capita income growth is related not only to the initial level of income but also to the main determinants of growth—investment rates and school enrollment rates—then a lower per capita GDP gives rise to a faster growth rate (see table, column 2). This means that if all countries had similar rates of accumulation of human and physical capital, poor countries would grow (slightly) faster than rich ones, and, therefore, differences in income per capita would be reduced over time. This weaker type of convergence is generally attributed to the advantage conferred by backwardness, because the low income level of poor countries does not allow them to invest as much in human and physical capital as the richer countries.

How should divergence and conditional convergence be reconciled? Countries that are initially poor tend to invest less, and to have lower educational attainments (see table, columns 3 and 4). This closes the circle. Poor countries tend to grow more slowly than rich ones in spite of the (small) advantages conferred by backwardness.

Explaining growth of per capita GDP, 1960–88


The richer

accumulate faster
Effect ofAverage growth of

GDP per capita
Average growth of

GDP per capita

Primary school

Initial level of GDP per
capita relative to leader0.40-0.324.4314.57
Average level of investment-0.07--
Average enrollment in
primary school-0.03--
Source Lant Pritchett. “Divergence, Big time” unpublished. World Bank. 1995
Source Lant Pritchett. “Divergence, Big time” unpublished. World Bank. 1995

Where do we stand?

Divergence in incomes per capita, not convergence, has been the dominant feature of modern economic history. But it need not remain so. Recent evidence on trends in inequality in both the industrial and the developing economies offers both hopes and fears for the future. Good government policy, in the domestic and international realms, has the potential to reverse the long-run trend of widening international inequality—a reason for hope. But failure to implement the right policies could make things worse. This means governments must:

• pursue market-based growth paths that generate rapid demand for labor, expansion in the skills of the work force, and rising productivity;

•take advantage of new opportunities at the international level, by reacting to new market opportunities and attracting capital—and managing the dislocations that changing trade patterns sometimes bring;

• construct a framework for labor policy that complements informal and rural labor markets, supports collective bargaining in the formal sector, provides safeguards for the vulnerable, and avoids biases that favor relatively well-off workers; and

• in those countries struggling with the transition to more market-based and internationally integrated patterns of development, try to ensure that the transition happens as fast as possible without large or permanent costs for labor.

This article draws on work by the authors for the World Development Report 1995, Workers in an Integrating World, Oxford University Press, New York, June 1995.

Other Resources Citing This Publication