Rethinking the College Premium in the Age of AI

The college premium is back in the news, with a study showing a drop in the relative wages of recent college graduates. According to the researchers, the change is due to a decrease in the relative demand for college workers from 2022 to 2026, marking the first sustained decline since 1914. They attribute about 30 percent of the drop to the labor market effects of generative AI. 

For over a century, technological advances have benefitted workers with more education and helped sustain our high college wage premium. It’s still too early to tell just how AI will impact the labor market, and there is every reason to believe that education will continue to offer enormous benefits, both in the job market and beyond. But if everyone’s economic well-being is threatened by AI, and the race between education and technology has truly been scrambled, it’s good to revisit higher education’s role in our economy and wage structure. 

We have allowed the college degree to become a stark dividing line in American society, with graduates enjoying far more economic security and mobility, along with greater access to health insurance, paid time off, and retirement benefits. These differences, in turn, have sparked a backlash against higher education and intensified political polarization. We can ensure economic security for workers at all levels of education while still supporting a robust and world-class higher education system. But doing so will require rewiring the relationship between our labor market and higher education system in ways that reduce inequality rather than magnify it.

The College Premium Is a Measure of Wage Inequality 

The “college premium” is a measure of how much more a person with a bachelor’s degree earns than someone with only a high school education. There are various ways to measure it, but each points to a large and enduring earnings gap between workers with and without a college degree, especially since 1980. The San Francisco Fed’s most recent estimate puts the premium around 71 percent. The Bureau of Labor Statistics, using different data, finds annual median earnings for bachelor’s degree holders are about $32,000 higher than those of high school graduates only. The recent study showing a drop in the premium, again using slightly different data and methods, puts the wage advantage at 77 percent in 2026, down from 86 percent in 2022.

Updates to the premium are closely followed by the education policy community because it is considered a critical signal to future students about the value of college. A dip in the premium raises concerns that fewer people may attend college. An increase is counted as evidence that college is still worth the cost. In both cases, a high premium is treated as a positive outcome.  

But is it? 

Another way to understand the college premium is as a measure of inequality. Wage inequality, to be exact, between people with different levels of education. The “premium” refers to the economic benefit that accrues to the individual who completes college compared to those who do not. Measures of wage inequality capture the differences in earnings between the two groups. The premium is the economic reward reaped by the individual graduate; wage inequality is the premium’s impact on society.

The Doom Loop: High Return, High Cost, High Inequality 

Higher education advocates often hold up the wage premium as the reason people should go to college. The Association of Public and Land Grant Universities (APLU), for example, has an entire page highlighting the economic benefits of a bachelor’s degree and the cost of not having one, including the fact that families with a college graduate have a net worth four times that of families without one, and that graduates are less likely to be on Medicaid or receive food stamps. The College Board publishes a report, “Education Pays,” that provides information on the economic returns to bachelor’s degrees, broken out by major, institutional type, and more. There are many others. 

Increasing the share of Americans with a college degree is a worthy goal and should be a focus of public policy. Countries with high levels of educational attainment tend to be wealthier and healthier than those with lower levels, more innovative, more competitive, and more stable. But relying on a high wage premium to sell college has real downsides. For one, framing the benefits of a college degree as a private return-on-investment has provided the intellectual foundation for pushing more of the costs of higher education onto students. Education and skills are treated as a form of “human capital” that generates an individual rate of return over the course of a graduate’s working life, not unlike other investments in land, equipment, or financial assets.

This economic logic has trapped American higher education in a doom loop for the last two decades: The private returns from a college degree are used to justify public disinvestment, forcing institutions to raise tuition and fees to cover gaps in public funding while also making them dependent on a high premium to attract students. The more college is framed as a long-term investment in the future fortunes of individuals, the more sense it makes to replace public subsidies with debt financing models similar to those used in other capital markets. And the more expensive higher education becomes, the more it amplifies existing inequality as the students most likely to reap the college premium are those who can afford to pay the full cost. Students and families struggling to afford college are offered loans to finance their education, but those loans will take a bite out of any premium while leaving the premium’s impact on overall wage inequality untouched. 

Emphasizing the wage premium as a reason to go to college also diverts attention away from how rising income inequality is hurting most Americans, not just non-graduates. Take homeownership rates. In its fact sheet, APLU notes that “In 2000, households headed by individuals aged 25-34 with a bachelor’s degree or higher had a homeownership rate that was just three percentage points higher than for peers whose highest degree was a high school diploma. By 2019, the homeownership gap by education had grown to nearly four times that size.” What the fact sheet doesn’t mention is that actual homeownership rates dropped for both groups, just more severely for those without a bachelor’s degree (from 48 to 44 percent for graduates and from 45 to 32 percent for non-graduates).

The same goes for earnings. While the college premium nearly doubled between 1980 and 2000, a significant share of the increase, especially among men, was due to the decline of earnings among non-graduates. Real earnings for men with a high school education fell by about 15 percent from 1979-1995 and are still below their 1979 level today.  In an economy where the majority of people are falling behind previous generations, college graduates are falling less fast. No wonder everyone is so angry. 

A High College Premium Does Not Increase College Completion

While a high premium may convince some to go to college who might not have otherwise, there is little evidence that it is necessary to boost a country’s educational attainment level or that a low premium suppresses skills formation. According to the OECD’s annual “Education at a Glance” series, there is no correlation between the size of the wage premium and a country’s “tertiary” graduation rates (any postsecondary degree, including bachelor’s, master’s, and doctorates). Using the OECD’s measure, the U.S. premium is 71 percent, well above the OECD average of 54 percent. The U.S. is also squarely in the middle of the pack in terms of attainment. Among Scandinavian countries, tertiary graduates earn around 22 percent more than their high school-only counterparts, but their attainment rates are comparable to, or higher than, those of the U.S. In countries with economic policies more similar to the U.S., such as Canada, the U.K., and Australia, the tertiary wage premium still does not exceed 40 percent, while the share of the population graduating from college is higher than here. These countries show that, with the right mix of policies, the wage premium does not have to be high for people to decide to attend college. 

But Americans tend to treat wage and income inequality, including the college premium, as beyond the control of policymakers. Markets determine wages, and inequality is the consequence of technological advances that increase the demand for more-educated workers, driving up the college wage premium; this, in turn, increases the supply of college graduates, or so the thinking goes.

These things are true—but only up to a point. As noted, other advanced economies facing the same technological and competitive pressures as the U.S. increased their college attainment rates, while maintaining comparatively low college premiums. The supply and demand for skills are certainly important determinants of wages, but they don’t have to be the only ones. Because labor markets are not perfect, wage levels are as much an expression of the relative power of employers and workers as they are of any market-clearing equilibrium at work, especially at lower wage levels. That is why most wealthy countries rely on a mix of market and non-market wage-setting strategies to ensure inequality does not get too high. Minimum wage laws, for example, establish a floor for all workers, regardless of education level. In the U.S., construction workers benefit from prevailing wage policies that ensure good wages for publicly funded projects. In California, an industry standards board established a $20.00 an hour wage floor for fast food workers, and in Minnesota, a similar effort has increased wages for nursing home workers. In some countries, workers and business representatives negotiate wage floors across all industries. 

In other words, there are many ways to influence wages in an economy. The high level of wage inequality in the United States is a policy choice, not an inevitable outcome of advancing technology or global competition, nor an indication of an efficient labor market correctly pricing skills. 

Reimagining Higher Education as an Antidote to Inequality 

Reversing the higher education doom loop will require a set of mutually reinforcing policy reforms that reduce inequality and raise wages for workers making the least—those in the denominator of the college premium. This can be accomplished through increases in federal, state, and local minimum wages and through the wider use of sectoral wage boards for workers in low-wage industries such as in retail, health care, early education, and others. In addition to wages, policies that increase economic security but can be hard for lower-wage workers to access, such as health insurance, paid leave, and secure retirement, need to be expanded. Finally, we need to reinvest in public higher education and make it affordable for everyone. The first two actions will shrink the college premium, not by reducing the earnings of college graduates, but by pulling up earnings for everyone else. The third will break the loop at the point where the high cost of college drives the need for a high premium, broadening access to higher education and the many benefits, both individual and societal, that come with it. 

That new study referenced at the beginning of this essay that found a sharp decline in the college premium attributed about thirty percent of the drop to AI. Some portion of the remaining 70 percent is likely the result of a sharp increase in wages for non-college workers that began in 2020 and continued through at least 2022, as well as increases to the minimum wage in a number of states. Is that bad news for college graduates? It reduced the college premium, to be sure, but by raising the floor for all workers. 

One of the ironies of the last two decades of higher education policymaking is how often it is grounded in the goal of economic mobility—not in the sense of lifting all boats but in allowing individuals to move out of bad economic conditions that will remain the same for those they leave behind. As AI upends our labor market, the answer is not to double down on college as the gatekeeper to economic well-being and essential benefits, but to create more economic security for everyone. In fact, if other countries are any guide, reducing inequality may well be the key to increasing college completion.