This is from a new CBO report, Changes in the Distribution of Workers’ Hourly Wages Between 1979 and 2009:
Changes in the Distribution of Workers’ Hourly Wages Between 1979 and 2009, CBO Director's Blog: Wages are a key component of the overall economic well-being of individuals and families. Hourly wages and hours worked determine an individual’s earnings, and for most nonelderly adults, earnings constitute the bulk of their family’s income. A CBO study released today, prepared at the request of the chairman and former ranking member of the Senate Finance Committee, documents changes in the amount and distribution of hourly wages received by workers in the United States between 1979 and 2009. It also reviews the leading explanations for changes in the supply of, and demand for, workers with different sets of skills, as well as how labor market institutions affect wages.
The wage rate (the wage per hour of work) received by workers in the middle of the wage distribution (the 50th percentile) increased by about 20 percent over the 1979–2009 period after adjusting for inflation, reaching about $17 per hour in 2009. The dispersion of wages—the gap between wages at the top and bottom of the distribution—also increased over that period, but the pattern of changes at the top and bottom differed. For men and women alike, the gap between the wage rates received by high-wage (90th percentile) and middle-wage workers expanded throughout the 30-year period; the wage rates of high-wage women grew especially rapidly. In contrast, the gap between the wage rates received by low-wage (10th percentile) and middle-wage workers widened for both men and women early in the 1980s but has remained stable for the past 20 years.
Wages are affected by market forces (the level and distribution of skills supplied by workers and employers’ demand for those skills) and institutional factors (such as minimum-wage laws and changes in the share of the workforce represented by unions). Given the complex pattern of changes in the wage distribution between 1979 and 2009, it is not surprising that no single explanation can account for the entire pattern.
In the category of market forces, the growing demand for skilled labor, particularly for highly educated workers, accounts for most of the widening gap during the past 30 years between the wages of college graduates and high school graduates. Although the post–World War II period saw steady growth in demand for college graduates that put upward pressure on their wages, growth in the share of workers with college degrees offset some of that pressure during the early part of that period. Beginning in the early 1980s, however, people entering the workforce did not have significantly more education than those retiring and leaving the workforce. That slowdown in the improvement in educational attainment combined with growing demand for more-educated workers drove the wage premium for college graduates higher—a key reason for the increasing wage dispersion in the top half of the wage distribution
Shifts in international trade might also have contributed to increasing demand for skilled labor, relative to that for less skilled labor, as imports from low-wage countries substituted for some domestic production and employment; however, research on the significance of that effect is inconclusive. In addition, a rising number of foreign-born people in the workforce affected the supply of workers with different amounts of education, but that shift appears to have had only a modest effect on the distribution of wages.
Turning to institutional factors, the federal minimum wage did not keep pace with inflation during the 1980s, and the decreasing real (inflation-adjusted) value of the minimum wage probably increased wage dispersion in the bottom half of the wage distribution during that period. Moreover, a decline in the share of workers who belonged to unions contributed to increasing dispersion in the upper half of the wage distribution for men over that same decade. Neither of those factors is a plausible explanation for the changes in the wage distribution in the 1990s and 2000s, however.
Although this study focuses on hourly wages, changes in the amount and distribution of hourly compensation, which includes both wages and fringe benefits, are also important. Unfortunately, data on hourly compensation are more limited than data on hourly wages. The available data indicate that the dispersion in hourly compensation in the upper half of the distribution was similar to the dispersion of wages, on average, between 1987 and 2007. In the lower half of the distribution, the dispersion of hourly compensation was somewhat greater than that for wages, on average, during the same period. Nevertheless, for both the upper and the lower halves of the distribution of compensation, the changes in dispersion over those two decades were similar to the changes in the dispersion of wages.
There is a lot to say about this, but not much that hasn't been said already, so let me simply point to a previous discussion of this topic: Driving Forces Behind Rising Income Inequality: Tracking the Internet Debate.