FRBSF Economic Letter
1997-25 | September 5, 1997
Rising inequality in wages has been a key feature of the U.S. labor market since the late 1970s. Put simply, rising wage inequality implies that gaps between high-wage and low-wage workers have widened. Perhaps the most striking feature of this trend is its uniformity. Wage gaps between individuals at any fixed points on the wage distribution–for example, those at the high end versus those in the middle, or those in the middle versus those near the bottom–all have widened. Much of this is due to rising returns to measurable skills; for example, the wage gap between high school and college educated workers has increased substantially. However, wage dispersion has increased even within groups of workers with identical measurable skills.
Rising wage inequality is related to a variety of economic and social developments, some positive and some negative. On the positive side, much of the rise in wage inequality in the U.S. has been associated with increases rather than declines in individual and family income. Furthermore, rising wage inequality may be a necessary consequence of a flexible economic system that also produces rapid job creation and low unemployment. However, some observers have expressed concern about potential downsides associated with rising wage inequality in the U.S. These include slippage at the bottom of the income distribution and related social costs such as rising crime and imprisonment, increased homelessness and family instability, deteriorating public health, and erosion of ideals regarding economic freedom and democracy.
In this Economic Letter, I document the extent of rising wage inequality in the U.S. and discuss sources that have been identified and assessed by economists. It is difficult to assign precise contributions to the various sources. However, demand factors related to production technology appear to be more important than factors that determine the available supply of different types of labor. Changing institutional features of the U.S. labor market, such as the minimum wage and the extent of unionism, also may have contributed to rising wage inequality.
Examining the pattern of rising inequality in wages helps to shed light on its causes. A flexible method for measuring the degree to which the dispersion of wages has changed is to compare the distance between individuals at different fixed rankings on the distribution. For example, researchers often examine the wage ratio between individuals at the 90th percentile of the distribution (i.e., 90% of working individuals earn less) and the 10th percentile (10% earn less); we call this the “90-10 ratio.” Between 1979 and 1995, this ratio rose from 3.7 to 4.8 for male workers, an increase of 30% above its base value. The corresponding increase for women was 59% (from 2.6 to 4.1).
A useful comparison of what has happened to the upper and lower portions of the distribution is provided by examining changes in the 90-50 and 50-10 ratios. For men, the percentage increase in the 90-50 ratio between 1979 and 1995 was nearly five times larger than the corresponding change in the 50-10 ratio. Thus, the bulk of rising wage differentials for men occurred among wage earners in the top half of the distribution. For women, this pattern was reversed, with growth in the 50-10 ratio being 1.7 times as large as growth in the 90-50 ratio.
This overall rise in inequality is reflected in rising wage differentials between individuals with differing levels of observable skills. For example, between 1979 and 1995, the average hourly wage gap between high school graduates and college graduates in their 30s approximately doubled. In percentage terms, this gap rose from 26 to 53 percent for men, and from 40 to 78 percent for women.
Overall, these figures reveal substantial growth in wage inequality. Furthermore, inflation-adjusted wages fell for low-wage workers in the U.S. during the 1980s and early 1990s, to the point that low-paid workers in the U.S. have lower living standards than do low-paid workers in many other economically advanced nations (Freeman 1997). These inequality trends are reinforced rather than offset by trends in the receipt of non-wage benefits such as health insurance and pensions: benefit coverage declined most rapidly for the lowest wage groups over the period 1979-1993 (Freeman 1997).
As noted above, much of the rise in wage inequality is associated with a rising return to formal education, which suggests that increased educational investment can offset the factors leading to rising inequality. However, even within skill categories–for example, among male workers who share the same educational attainment and years of full-time work experience–wage gaps have widened. Thus, to explain rising inequality in individual wages, we need to look beyond factors that have raised the wage premiums associated with measurable skills.
Furthermore, although wage inequality has increased in other developed countries, only in the United Kingdom did the increase rival that in the U.S., and low-income workers fared far better in the U.K. than in the U.S. Several developed countries (such as Germany and France) experienced little or no increase in wage inequality during the 1980s (Fortin and Lemieux 1997). Then to some extent, the factors responsible for rising U.S. inequality must be unique to the U.S.
We can group the explanations for rising wage inequality into demand, supply, and institutional factors. Demand factors are those that exogenously determine the quantity of different worker types that employers wish to hire: changes in production techniques that favor workers with higher skill attainment, importation of foreign manufactured goods, and the general decline of manufacturing and rise of service jobs. Supply factors are those that exogenously determine the available quantity of different types of workers: immigration rates, female labor force participation rates, birth cohort size, and factors that change the costs (and therefore attainment) of higher education. Institutional factors include the extent of unionization in the economy and the level of the minimum wage. These two factors are institutional in that they reflect government policy: the minimum wage is set by legislation, and the environment for unionization is determined partly by labor relations laws and their enforcement. Except for the minimum wage, which directly determines the price of low-wage labor, each of these factors indirectly determines wages by affecting the quantities of different types of labor used.
Many economists have focused on demand factors as the key causes of rising inequality, and they argue that the single most important explanatory factor has been changes in production techniques that favor skilled workers (“skill-biased technological change”). Although initial studies estimated the contribution of technological change indirectly, through a process of elimination, recent direct evidence has been consistent with the basic story (Johnson 1997). No consensus exists, however, on whether this effect has arisen due to increased reliance on computer technology in the workplace or due to a more general increase in the complexity of work environments. Furthermore, recent work has questioned the primacy of technological change as an explanatory factor (Freeman 1997). One key piece of contradictory evidence is the absence of rising inequality in some countries–such as Germany–where production technologies are similar to those in the U.S.
Given its controversial implications for trade policy, the contribution of international trade to rising inequality has been hotly debated. The basic argument is that factors such as declining transportation costs and ongoing technology transfers have reduced U.S. import prices for goods whose production requires abundant unskilled labor; this in turn has led to downward pressure on the employment and wages of low-skilled workers. Recent research mostly suggests that the effect of trade on overall inequality has been small. A key piece of evidence is that increasing relative demand for skilled workers occurred in all industries, not just those subject to import competition.
A related concern is the potential erosion of the middle-class job base due to ongoing job shifts from manufacturing to services. However, recent evidence suggests that this shift has had only a small effect on wage inequality (Valletta 1997). Furthermore, the shift from manufacturing to services reflects more fundamental demand factors, such as underlying changes in production techniques and productivity.
Supply side factors appear to be less important than demand factors in explaining rising U.S. wage inequality. Most frequently mentioned is immigration. Available evidence suggests that its overall impact on inequality is small (Topel 1997). However, like trade, immigration mainly affects the wages of those at the low end of the distribution, whose relative earnings have declined. This consideration is likely to keep trade and immigration in the policy spotlight.
Increased female labor force participation is another oft-cited source of growing inequality in male wages. However, available evidence (Topel 1997) suggests that its contribution has been small: for example, the surge in female labor supply occurred in the 1970s, before the noticeable acceleration in male wage inequality that occurred in the 1980s. A similar argument applies to cohort size effects on rising wage inequality: the large “baby boom” cohort entered the labor market in the 1970s, and any resulting increase in wage dispersion would have occurred before the 1980s acceleration in inequality.
Recent analysis suggests that institutional changes may have contributed to growing wage inequality in the U.S. The key factors are the declining real minimum wage and declining union membership. During the 1980s, the inflation-adjusted minimum wage declined by more than 30%, and union membership among men fell nearly one-third, from 31% to 21% (Fortin and Lemieux 1997). Fortin and Lemieux (1997) cite evidence showing that these declines explain large shares of rising inequality for both men and women in the U.S. during the 1980s.
One caveat to Fortin and Lemieux’s analysis, however, is that it ignores any positive employment effects that might be associated with reduced unionism and minimum wages. Employment growth was relatively rapid in the U.S. during the 1980s, as was the rise in inequality. This contrasts with countries such as Germany and France, where government policies treated unions and minimum wage workers more favorably than in the U.S. Inequality grew relatively slowly in these countries, but employment growth and unemployment performance was weaker than in the U.S. Direct evidence on a tradeoff between employment growth and inequality is not available, however, nor is evidence on the effect of less easily measured institutional factors, such as wage-setting norms within and across firms.
Is the trend toward increasing U.S. wage inequality likely to continue? A mitigating factor arises from the tendency for the supply of workers to respond to wage differentials across skill levels. For example, the large wage premium earned by college-educated workers in the 1980s partially reflects a slowdown in college attendance in response to a much smaller college premium during the 1970s. College attendance rates, however, have not responded rapidly to the rise in the college wage premium in the 1980s and 1990s. This may be explained by steadily rising costs of a college education, which appear unlikely to diminish soon. As a result, the supply of college graduates probably will not outpace rising demand for their skills in coming years (Johnson 1997). In addition, a substantial portion of rising inequality is not associated with observable skills, and the form of investment that would help to offset this portion is unclear.
More generally, the demand and supply factors identified above do not appear to be abating. Furthermore, recent welfare reform initiatives are likely to increase the supply of low-skilled job-seekers. This will place added downward pressure on the wages of those at the bottom of the distribution, at least in the short run. If the erosion of wages for low-income individuals continues, rising inequality is likely to remain an important policy issue in the U.S.
Fortin, Nicole M., and Thomas Lemieux. 1997. “Institutional Changes and Rising Wage Inequality: Is There a Linkage?” Journal of Economic Perspectives 11(2), Spring, pp. 75-96.
Freeman, Richard B. 1997. When Earnings Diverge: Causes, Consequences, and Cures for the New Inequality in the U.S. Washington, DC: National Policy Association.
Johnson, George E. 1997. “Changes in Earnings Inequality: The Role of Demand Shifts.” Journal of Economic Perspectives 11(2), Spring, pp. 41-54.
Topel, Robert H. 1997. “Factor Proportions and Relative Wages: The Supply-Side Determinants of Wage Inequality.” Journal of Economic Perspectives 11(2), Spring, pp. 55-74.
Valletta, Robert G. 1997. “The Effects of Industry Employment Shifts on the Wage Structure, 1979-1995.” Federal Reserve Bank of San Francisco Economic Review, 1, pp. 16-32.
Opinions expressed in FRBSF Economic Letter do not necessarily reflect the views of the management of the Federal Reserve Bank of San Francisco or of the Board of Governors of the Federal Reserve System. This publication is edited by Sam Zuckerman and Anita Todd. Permission to reprint must be obtained in writing.
Please send editorial comments and requests for reprint permission to
Attn: Research publications, MS 1140
Federal Reserve Bank of San Francisco
P.O. Box 7702
San Francisco, CA 94120