The College Wage Premium
Does higher education have a "value" problem?
The recent backlash against a college education has at its core the notion that college is no longer “worth it”. The argument is that a combination of rising tuition prices, low completion rates, lack of preparation for the work world, and demand for occupations that do not require a college degree has sharply lowered the returns to attending college, resulting in an explosion of student debt and loss of confidence in the value of a college degree. This critique is wrong in (almost) every particular, as we will lay out in coming posts. This post will focus more narrowly on the question of the financial returns to a college education, that is the earnings a college grad can expect in contrast to the earnings a student could expect if they stopped education after high school or perhaps pursued vocational/technical degrees.
The New York Fed maintains an excellent interactive website “The Labor Market for Recent College Graduates” where you can explore median wages for recent college graduates in contrast to high school diploma holders, compare the range of wages paid, outcomes by major, and unemployment and underemployment.
In 1990, the median wage for recent college grads was 38% more than the median wage for a high school diploma holder. That premium had risen to 67% by 2023, which reflected both a 10% rise in real (inflation-adjusted) wages for college grads and almost a 10% decline in real wages for high school grads.
Wages vary (a lot) among college grads
Of course, not all college graduates earn the median wage, some earn much more and others much less: using the New York Fed data for 2020, a 75th percentile college degree holder was earning 233% more than the high school median, while a 25th percentile college degree holder had only a 17% premium over high school median. In addition, the highest earning high school grads will earn more than low earning college grads. A study published by the Center for Education and Work at Georgetown University found that about 16% of high school grads and 28% of Associate Degree holders earn more than about half of college grads.
But one should not think that these higher paying jobs are easily available to anyone finishing high school. Some of the high earning high school grads are employed in skilled trades, and these skills can take as long as a college degree to obtain. For example, becoming a journeyman electrician or a skilled plumber requires four to five years of schooling and apprenticeships! Beyond these occupations that take several years to master, some of the high earning high school grads are receiving what an economist would call a “compensating differential”, that is, higher pay for dangerous or difficult work. With some months of training a high school grad could become a crude oil hauler and make more than $100,000 a year… sitting atop thousands of gallons of volatile fuel in heavy traffic every day.
Much of the variation in earnings for college grads is driven by choice of majors (not choice of universities). Looking at the “outcomes by major” tab from the NY Fed data tool: early career engineers are the 75th percentile types, earning twice the salary of 25th percentile types in education, liberal arts, hospitality, or theology degrees. Some of the variation in earnings is driven by where graduates choose to work. The returns to college degrees in cities are much higher than they are in rural areas, where fewer firms employing college graduates choose to locate.
And while this is very hard to measure carefully, some of the earnings variation is no doubt due to variation in the cognitive and non-cognitive abilities of graduates as well as differences in their personal traits and underlying preferences/values. Some graduates are smarter than others. Some graduates work harder than others. Some graduates have better networking/relationship-building skills or have stronger pre-existing social capital (i.e. family-based connections that open doors to internships and first job offers) than others. Some graduates focus on earnings, others on raising a family, others on pursuits that are meaningful to them but not especially remunerative. We should not think college a failure because it inspired some graduates to serve as pastors or create art or raise children or provide care for the aged.
Grad school and on-the-job experience
The comparisons above have all been for recent college grads 22-27 years old. Mid-career the salary gaps between college grads and other workers explode, for two reasons. First, an undergraduate college degree opens the door to graduate degrees, which can generate earnings that are, on average, three times those of a high school degree holder and much higher than those for certain professional degrees in health, law, and business.
Source: Jonathan James, “The College Wage Premium” https://www.clevelandfed.org/publications/economic-commentary/2012/ec-201210-the-college-wage-premium
Second, while all workers earn higher wages as they gain experience, recent evidence from David Deming demonstrates that college degree holders enjoy greater wage growth from experience than do high school degree holders, even accounting for the ability of the graduate in question. He attributes that fact to the notion that over the course of a career, the superior learning ability of college graduates enables them to master jobs of rising complexity and earnings power. (In the figure below, the solid line represents how wages grow with experience within the same job; the dashed lines represent “between-jobs” wage growth that occurs over time when job switching.)
Source, David Deming, “Why do wages grow faster for educated workers?” 2023. https://www.nber.org/papers/w31373
Where do skeptics have a point?
All this evidence points strongly to the idea that college education is imparting a significant benefit to its graduates, something that is valued in the form of higher wages in the market place. Still, if you were inclined to be skeptical of the value of attending college, what data might you point to?
First, after decades of rising rapidly, the college wage premium calculated over all workers (not just recent grads) has flattened out over the past ten years. The decline after 2020 seems to be primarily driven by a post-pandemic rise in wages for high school educated workers rather than a fall in wages for college educated workers. And the overall wage premium remains very significant: somewhere between 75 and 80% over this entire period. But there is no mistaking the flattening of the premium. We’ll tackle this question, and how it relates to the expansion of college-going and increases in retention and completion rates, in an upcoming post.
Source: Bengali, Sander, Valleta, and Zhao, “Falling College Wage Premiums by Race and Ethnicity”, 2023
Second, as we’ve noted, wages at the lower end of the earnings distribution for college grads are below those of many high school grads. This is likely linked to the issue of “underemployment”, that is, college grads doing jobs that they could have done had they simply finished high school. The New York Fed website tracks these data and reports that a third of all college grads and over 40 percent of recent college grads fit this category: working in occupations where more than half the workers report a college degree is not needed. That seems like a problem! We’ll tackle this issue in a follow-up post.
Third, we cannot say with certainty from the above evidence that going to college caused wages to rise. It might be that there is some unobserved characteristic (say, cognitive or non-cognitive ability) that both enables students to succeed in college and leads to higher wages. True ability is notoriously hard to measure and so this is a hard criticism to refute. (We can say that many of the highest earning professions are not available to high school grads, because lack of a college degree forecloses the option of attending graduate school.) We’ll come back to this question of the causal returns to college in later posts, as well as the idea that students don’t actually learn anything in college but that college represents only a massive experiment in signaling true ability to employers.
Fourth, we haven’t said anything yet about the rising cost of college and whether the return on that investment is still positive. This is tricky because the eye-popping sticker prices you may have seen often have little to do with students’ actual cost of attendance. We will dig into this in some detail as well.
What have we learned? And coming attractions…
Taking all this together, it seems clear that the typical college grad earns much more than the typical high school grad, and these differences get larger as one acquires more education and gains more experience in the workforce. Nevertheless, there are significant numbers of college grads who, for a variety of reasons, work in the jobs that also employ many high school grads, and there are also significant numbers of college grads who earn little more than high school grads.
We think the most important questions about the value of a college education and whether universities could do much better by their students should focus on this sizeable but atypical group of underemployed graduates. We’ll tackle this issue in the coming weeks.
But before that, we’ll examine three questions closely related to the college wage premium.
How do college grads perform on measures other than salary and wages?
If the earnings power of a college degree is so great, why has student loan debt grown so rapidly and perilously?
If employers are willing to pay college graduates a sizeable wage premium, why do employers also complain so much about students’ lack of preparation?
“Finding Equilibrium” is coauthored by Jay Akridge, Professor of Agricultural Economics, Trustee Chair in Teaching and Learning, and Provost Emeritus at Purdue University and David Hummels, Distinguished Professor of Economics and Dean Emeritus at the Daniels School of Business at Purdue.