In a recent speech, President Bush acknowledged a problem that economists and policy makers have long been concerned about: the growth of economic inequality. The president went on to offer this explanation for the growth in unequal economic outcomes:
“The reason is clear: We have an economy that increasingly rewards education and skills because of that education.”1
In fact, there is no evidence that the recent growth in income inequality is due to increased rewards to education. To the contrary, for the first few years of this business cycle, the college premium-the wage advantage to college educated workers-actually fell slightly, implying other factors are responsible for inequality’s recent resurgence.
First, let us be very clear: the earnings advantage to persons with higher levels of education is one of the most persistent findings in economics. And for many years, that advantage was growing: not only did college educated workers earn more than persons with only a high school degree, but the difference grew quickly, particularly over the 1980s.
Second, there are clear and obvious advantages-economic and social-to achieving higher levels of education, both for individuals and for society.
That said, policy makers too often, like the president, reduce the growth of inequality to this sole factor, even when the evidence points elsewhere. Here, we collect a number of those facts.
The chart shows the trend in the college wage premium just noted, adjusted for the set of relevant factors that labor economists typically control for in this type of analysis, including age, race, gender, region where workers live, and marital status. The college premium increased from 25% in the late 1970s, to 45% in the late 1990s and has since flattened, even trailing off a bit in the early 2000s, before ticking up slightly in 2005. Thus, the argument that employers’ skill demands were bidding up the wage premium to those with more education was arguably true (other factors were involved here too, noted below) over the 1980s and 1990s, but not in the 2000s.
Next, though the inequality debate references relative wages-the earnings of one education group compared to another-note that the actual inflation adjusted earnings of college graduates rose very little over this business cycle. Between 2000 and 2006, hourly wages of workers with a college degree (B.A. only) grew by only 1.4%. This wage estimate omits fringe benefits, and as these costs rose over the period, a full compensation measure would show more growth, but Gould (2006) shows that the share of college-educated workers with employer-provided health care fell by about 3.6 percentage points, 2000-05. (Gould finds larger coverage losses among less-educated workers.)
In a similar vein, Allegretto (2007) shows that the employment rates of young college graduates (the share of young college graduates that are employed) are still recovering from the cyclical decline in their employment rates that occurred between 2000 and 2003. Bureau of Labor Statistics data for employment rates of persons 25 and up show that since the peak of the last business cycle (March 2001), the only group whose employment rates are up is high school dropouts. Those of college graduates are down 1.2 percentage points.
Together, these data show that the wages of college graduates have improved only slightly since 2000 and have not improved more than high school educated workers. Also, college graduates are now less likely to be employed than in 2000.
None of this implies that college graduates are worse off than other groups-in some cases, as in employer-provided health coverage, they have lost less ground than those with less education. But neither are they insulated from negative trends that have beset the workforce at large, a key point to which we return below. Equally important, there does not seem to be an insatiable demand by employers for college graduates that is leading them to improve the wages and benefits of such workers.
We noted above that while increased skill demands were certainly one force in play behind the trend in the chart, other factors were also driving up the wage gap. For example, less educated workers are more likely to earn the minimum wage, and its real value fell steeply over the 1980s. Persistent, growing trade deficits and globalization also dealt a blow to the U.S. manufacturing sector over these years, leading to the loss of high-quality jobs for non-college-educated workers and putting downward pressure on wages among similarly skilled workers. Similarly, union representation has been declining, another important reason for greater inequality. Finally, unemployment was high, on average, over the years when the premium grew most quickly. When unemployment fell sharply in the latter 1990s, the growth of the wage gap slowed, suggesting that the absence of full employment, and the diminished bargaining power associated with slack labor markets, is another factor that must not be overlooked.
In other words, wage inequality is driven by a slew of factors, of which differences in education is but one. More recently, in the 2000s, the evidence shows no evidence of increasing skill demands, or at least no evidence that these demands are not being met by enough skilled workers. Instead, in recent years, it appears the inequality has largely been driven by increased concentration of income and wealth at the very top of the scale. The current inequality is much more than a simple skills story-persons like CEOs and holders of large capital assets hold a privileged position that has enabled them to steer the bulk of growth their way.
Sure, many of these persons have college degrees, but as our and others’ data show, a college degree, even an advanced degree, does not guarantee entry in this rarified club (see, for example, our analysis of recent inequality trends). In fact, our own research shows that half of the growth in wage inequality over the 1980s, and most of the growth in the 2000s, occurred within education groups, meaning that inequality’s growth is currently being driven by the gains of some college graduates relative to others with the same education credentials.2
Policy makers and analysts must avoid reducing the inequality debate to a sole explanation regarding education. For any individual making a decision about their future, and for any society contemplating useful investments, education is an obvious and important area to consider. But it is not solely responsible for the growth of inequality, not over the longer term, and especially not in recent years. Thus, other policies-minimum wages, a level playing field for union organizing, health care and pension provision, work supports for low-income workers, full employment, and responding to the downsides of globalization-also need to be pursued.
 See Mishel, Bernstein, Allegretto (2006), State of Working America, Table 3.23.