Labor studies.

• Author: Freeman, Richard B.
• Position: Research on wage inequality - Program Report

Two developments have made the period since I last summarized the work of the NBER's labor program particularly exciting: first, several topics of NBER research have attracted great public attention. Rising wage inequality, the subject of many Bureau Working Papers between 1991 and 1994, is now recognized as one of the nation's major economic problems. The effect of labor market institutions and government policies--ranging from minimum wages to training--on employment and wages also has been the subject of considerable public discourse. Further, internationalization of the U.S. economy, NAFTA, and development of the Common Market all have increased the interest of business and the public in how labor markets overseas work.

Second, newly available microdatasets for firms and individuals in many countries, and continued improvements in computer technology have opened new vistas for NBER researchers. In the 1980s, most researchers exploited U.S. microdatasets, which provided information on individuals; in the 1990s, many economists have analyzed datasets from other countries, and cross-sectional and longitudinal data for firms. Now researchers are beginning to examine the "intersection" of these two forms of data: that is, longitudinal data that track the economic performance of both individual workers and the establishments in which they work.

In the early 1990s, the NBER's labor studies program focused in particular on the following areas: earnings inequality; the behavior of labor demand; the effect of government programs; training; international comparisons; and social or family effects on individuals.

Earnings Inequality(1)

Bureau studies of earnings inequality, and of the fall in the real earnings of less-skilled workers, have documented a period of rising inequality, both among and within observable skill groups, and in hours worked. What emerges in these studies is a multicausal explanation of the rise in earnings inequality: in the 1980s, and to some extent earlier, the forces of supply and demand and institutional changes all operated in the same direction to raise inequality.

On the supply side, the key factors were: 1) a reduced rate of growth in the supply of college graduates relative to less-educated workers, caused in part by young Americans' decisions about investing in education; and 2) the influx of less-educated immigrants. On the demand side, shifts toward more-educated workers within industries, shown by Alan B. Krueger to be related in part to technological changes associated with computers, and by myself, George J. Borjas, and Lawrence F. Katz to be related to the growth of the trade deficit, are widely cited factors. In terms of institutional changes, David Card, Thomas Lemieux, and I have shown that the decline in unionization and the reduction in the real value of the minimum wage contributed to the increase in inequality. Katz and Kevin M. Murphy conclude that these factors help to account for the rise in earnings differentials among workers who differ in measured skills. In addition, measured shifts in demand and supply have been shown to affect the earnings of blacks and women by John Bound and Harry J. Holzer, and by Francine D. Blau and Lawrence M. Kahn, respectively. Steven G. Allen and several others also have examined the interindustry wage structure, which is an important component of wage differences. Still, these analyses have been unable to pin down the factors underlying the rise in inequality among workers with the same measured attributes.

Labor Demand Behavior

The surge in interest in the demand for labor,(2) focused on everything from the use of outside contractors to the dynamics of adjustments in firm size and employment growth, perhaps spurred by changes in the minimum wage, has yielded some results that have challenged conventional wisdom. Daniel S. Hamermesh has developed models that explore the dynamics of demand, among other things. Steven J. Davis (and others) have shown that small employers do not create more net jobs than larger firms in the manufacturing sector, because the job destruction rate as well as the job creation rate is greater in small firms than in large firms. Eli Berman, Bound, and Zvi Griliches report that labor demand within industries has shifted toward more skilled workers, which helps to explain the rise of inequality. Bound and Holzer also have examined the effect of industrial shifts on the employment of black workers relative to white workers. And, Douglas L. Kruse has developed evidence that profit-sharing arrangements raise firm productivity.(3)

Several researchers, in particular Card, Krueger, and Katz, have examined the demand-side effects of recent increases in the minimum wage.(4) Their results are striking: they find that the recent increases have had no discernible adverse effect on employment, which raises questions about the appropriate model for analyzing modest changes in wages in the labor market for low-skilled workers.

Training and Human Capital Formation

Continuing the long NBER tradition, stretching back to the pioneering work of Gary S. Becker and Jacob A. Mincer, researchers have explored the effects of various government programs and policies on investment in schooling, earnings profiles, and human capital formation in general.(5) James J. Heckman and Steven J. Cameron have examined the effects of the widely used General Equivalency Degree, and found that it is far from the equivalent of a high school education in terms of earning power. Thomas J. Kane and Cecilia Rouse have found that two years of junior or community college earns roughly the same return as two years in a four-year college, though. Ronald G. Ehrenberg and his coauthors have studied the effects of government programs on the decision to attend college.

Finally, Card and Krueger, and Michael A. Boozer, Krueger, and Shari Wolkon have examined issues relating to school quality. Their results suggest that differences in school quality play a greater role in earnings than had been found in past studies of educational production functions.

In addition, various NBER studies used different instrumental variables to determine the effect of schooling, and examined randomized experiments to assess training programs. These papers are summarized in the methodology section of this Program Report.

Other Government Programs

In addition to studying minimum wages and training or education programs, Bureau researchers have analyzed the labor market effects of a diverse set of governmental programs.(6) Janet Currie has examined the effects of several programs, including Head Start, on the...