Average Wages and Inequality
Marvin Kosters

[page-number of print ed.: 32 = blank page]
[page-number of print ed.: 33]

Average Wages and Inequality
Marvin Kosters

I want to use this opportunity to make two simple points. The first is that the average worker in the US economy has experienced significant gains in real wages over the past couple of decades. The second is that although wage inequality has increased, the wider wage differentials that emerged reflect mainly the increasing value that employers place on skills and the increase in inequality has not resulted in significantly deeper cleavages between workers with different schooling levels. These two points can be only briefly stated and illustrated here. They are developed and documented in detail in my recent monograph, Wage Levels and Inequality: Measuring and Interpreting the Trends, (The AEI Press, 1998).

Wages of the average worker are often described as stagnant or declining after adjustment for inflation. Data on wages and compensation of several types are available, with data from different sources having strengths and weaknesses. The data that are usually used to support the view that the average worker is falling behind, however, have deficiencies that are particularly serious for measuring trends in the level of real wages. For this purpose, data on average compensation per hour and wages and salaries per hour are more satisfactory in terms of the components of pay they include and the comprehensiveness of their coverage of workers. These latter data show significant improvement in workers real pay levels over the years.

Data that show the average level of total real compensation of workers, and of the wage and salary component of their pay, are charted in Figure 1. In contrast to the view that average real wages have fallen, these data show an increase in total real average hourly compensation of about 30 percent since 1973 and an increase in the wage and salaries component of more than 20 percent. Inflation has been taken into account in these measures by adjustments that

[page-number of print ed.: 34]

Figure 1:
Adjusted Measures of Real Wages and Compensation
Comparisons from 1973

[page-number of print ed.: 35]

apply procedures for measuring consumer prices that are consistent over time. [For a detailed discussion see Marvin H. Kosters, Wage Levels and Inequality: Measuring and Interpreting the Trends (Washington, DC: The AEI Press, 1998), pp. 24-27 and references noted there.]
The validity of these measures is supported by analogous data on productivity growth and by other independent measures of employment costs. Although these measures of workers' real pay increased less rapidly than during the 1950s and 1960s, significant gains in average pay were nevertheless realized. Moreover, although these measures of average real pay show a cyclical pattern, there has been no apparent tendency for acceleration or deceleration of the rate at which gradual increases in pay have accumulated during the past 25 years.

Wage inequality has increased since the mid-1970s. One measure of wage differentials that has been closely related to measures of overall wage inequality is the ratio of the wages of college graduates to the wages of workers with only high school credentials. As shown in Figure 2, the premium that employers have been willing to pay for the higher skills of workers with college credentials increased for about 20 years into the early 1990s. A major portion of that rise in the college/high school wage premium, of course, should be regarded as a recovery from a level that was temporarily depressed during the 1970s by the extraordinarily large number of college graduates in the baby boom generation who entered the work force at that time.

The increase in the economic reward to education since the mid-1970s created a strong incentive for young peole to invest in more schooling. A growing proportion of high school graduates who attend some post-secondary schooling represents one important way that youth have adjusted their behavior to take advantage of the higher payoff for schooling. The strong and persistent uptrend in enrollment beyond the high school level shown in Figure 2 is one dimension of a more general rise in workers willingness to invest in themselves to improve their skills and their economic circumstances. The remarkable similarity in these patterns of

[page-number of print ed.: 36]

Figure 2

College Enrollment and College Wage Premium

[page-number of print ed.: 37]

increased payoff to schooling and higher school enrollment illustrates the importance of self-correcting behavioral responses to changes in economic rewards.

In addition to the economic incentives produced by wider wage differences between workers with different skill levels, and their consequences, it is also important to consider the social and political implications of wage inequality. In particular, increased wage inequality could be associated with a growing gap between incomes of workers with different educational qualifications. For example, workers with high school credentials could have so much more difficulty earning a middle-class income that they would have little in common with college graduates. To examine this issue, we can look at what has happened to the degree of overlap in wages for different schooling categories.

One measure of the extent of the education-income gap is the percentage of high school graduates who earn less than the median wage for college graduates. This measure, charted in Figure 3, shows that there was more overlap between wages of workers in these two schooling categories when the average wage gap was smaller. Even in the early 1990s, however, about 20 percent of high school graduates earned more than the median wage of college graduates, a proportion that was about the same as prevailed during the late 1960s. Despite considerable widening of the difference in average wages between workers with high school and college credentials, because wages within each schooling category are quite widely dispersed, their wages have evidently not become much more sharply divided.

Two views about developments in the labor market are quite common. The first is that average pay has stagnated or declined. The second is that growing wage inequality has increased the economic and social gap between workers with different education. The idea

[page-number of print ed.: 38]

Figure 3:

Percentile Location in High School Wage Distribution of the Median College Wage

[page-number of print ed.: 39]

that either of these developments would be more tolerable if the other had not occurred links these two views. Based on the data that I have analyzed, I argue that the average worker's payhas not stagnated or declined, and the adverse social consequences of larger skill premiums have often been exaggerated.