Is College Worthwhile?
The culture wars shifted to a new front yesterday following the announcement of President Biden's executive order on student debt relief. Predictably, Fox News called it a vote-buying "handout" that will tax the poor, exacerbate inflation, and incentivize colleges to raise tuition. As Mississippi Governor Tate Reeves tactfully tweeted, "welders, plumbers, laborers, & other Mississippians....will be forced to pay off the debts of Harvard doctorate degree gender studies majors living in California."
On the other end of the political spectrum are expressions of approval (MSNBC), or complaints that even more is needed (The Guardian) in which potential effects on the economy aren't even mentioned. (For reasonably balanced discussions of the new executive order see here or here.)
Meanwhile, millions of high school seniors and their families are entering the late stages of the college application process this month, and some of them are probably still asking: Should we be doing this? In other words, is college a worthwhile investment of time and money?
Americans have been asking some version of this question since the birth of our higher education system in 1636 (i.e., the founding of Harvard). Although people attend college for various reasons, there's some expectation, or at least hope, that the rewards of a college degree will more than offset its costs. Consider this: Roughly 48 million Americans currently carry student loan debt (43 million federal loans; 5 million private). That's nearly one fifth of our adult population. These people didn't take out loans expecting to retire poorer than they would've been with a high school degree alone.
In this newsletter I'll be discussing whether college is a worthwhile investment. There are many ways in which college could be "worthwhile"; my focus is on the prospect of future earnings. Although I'll be dipping into a few weeds, at the end I'll draw some practical conclusions for anyone wondering about the financial payoffs of a college degree.
In several places you'll also see illustrations of an important caveat about statistical data: You have to be careful which statistics you rely on to answer a question. Different stats answer different questions, and sometimes the most straightforward, seemingly appropriate statistics are the ones that lead you astray.
Faltering expectations
In recent years, rising tuition costs, crushing student debt, and an increasingly volatile job market have led more and more Americans to question whether a college degree is a good investment, financially speaking.
It's hard to neatly summarize public opinion – nationally representative surveys differ in question content and format – but here's one detail that stands out: In four prominent surveys conducted over the past 18 months, less than half of respondents chose the most optimistic response to questions about the value of a college degree.
To take just one example, in a June 2022 survey by College Pulse, only 24% of current college students said they were "completely confident" that they would "earn sufficient money to make the cost of college worth it". (Another 24% chose the "very confident" option. 35% were "somewhat confident", and the remaining 18% were either "not too confident" or "not at all confident.") The results were remarkably similar when parents and, separately, high school students responded to the same question.
These data could be read in different ways. For example, College Pulse collapses the responses for "completely", "very", and "somewhat" confident and refers to all of these people as "confident." I think that's a bad move. Because "somewhat confident" is the middle option in a five-point scale, at least some people who chose it may harbor substantial doubt or uncertainty. Still, however you view the data, it's clear that responses are mixed, and most respondents fail to exhibit the highest degree of confidence.
So, what do the data actually show? Is college a good investment in light of future earnings?
Earnings data: First pass
Whether college is a good investment, in any sense, depends on who you are, which college you attend, what you study, etc., but it's worth noting that college graduates tend to earn more than people with high school degrees only. The mean and median annual salaries for college graduates are higher, given the same amount of time since graduation, and estimated lifetime earnings are greater too. For example, median lifetime earnings for men with bachelor's degrees are roughly $900,000 greater than for men with high school degrees only. The difference for women is $630,000 (which confirms the value of a college degree and, at the same time, points to both gender differences in career choices as well as workplace discrimination).
Data from the Bureau of Labor Statistics and other sources add some nuance. The key variable isn't whether or not you have a college degree, but rather how much education you've had, as illustrated by the figure below, generated in 2021 by Anthony Carnevale and colleagues at the Georgetown Center on Education and the Workforce. This figure shows that median lifetime earnings increase with each additional level of education. (My focus will be on the high school student who's pondering whether to work after graduation or pursue an associate's or bachelor's degree.)
Earnings data: A closer look
The stats I just cited are informative but limited.
First, mean and median differences only tell us that college graduates tend to earn more than high school graduates. The distributions overlap, and thus some high school graduates earn more than some college graduates. For example, 16% of high school graduates with no additional education earn more during their lifetimes than half of workers with a bachelor's degree. College does not guarantee a higher paycheck, in other words, even though that's a frequent outcome.
Second, a college graduate's earnings are influenced by variables such as student demographics, institution attended, and choice of major. Disaggregating by these variables, as I'll show near the end, reveals substantial differences.
Practically speaking, prospective applicants and their families aren't considering every institution of higher education as both desirable and attainable. Rather, they're wondering whether a particular college or set of colleges would be worth the investment. This brings to me a report by Michael Itzkowitz, a senior fellow at the center-left think tank Third Way, published in 2019 and referenced in a NY Times article earlier this week.
Itzkowitz points out that when comparing colleges in terms of how much their students eventually earn, mean and median values can be misleading. Most people are familiar with these statistics, and they somehow feel like the right ones to use. However, means can be skewed by a small number of extremely high earners, while both means and medians might, for example, underestimate the value of institutions that serve more students in lower-paying fields (e.g., education). Thus, Itzkowitz notes that a more accurate metric is the percentage of students from each institution who earn more than the typical high school graduate who hasn't attended college.
Itzkowitz analyzed the percentages of former college students earning more than the average high school graduate. Comparsions were made six, eight, and 10 years after enrollment in college. (Students were included in the college group whether or not they actually graduated.) Analyses focused on students who received federal grants or loans to pursue associate's or bachelor's degrees. Students in grad school during the six to 10 year period were excluded. Over 3,800 institutions were represented.
Among the many findings, what struck me most are those from the 10 year mark (the greatest amount of time, in this dataset, for college students to have found their career groove). For instance, more than half of students from 28.9% of colleges were earning less than the average high school graduate. That's startling. Concretely speaking, there are 1,104 institutions of higher ed in the U.S. where more than half of the students who enroll end up earning less than the average high school graduate 10 years later. (Institutional-level data are available here.)
One message from stats like this for students, families, and high school counselors is that from the perspective of future earnings, choice of college matters. Although students may enter college with the expectation that a degree will promote financial security, Itzkowitz concludes that "many will leave with employment opportunities that fall short of that initial expectation, as thousands of schools are providing minimal economic benefit to those who enroll."
Earnings data: Some qualifications
Itzkowitz's conclusion is bracing and strongly supported by the evidence. At the same time, two qualifications can be added.
First, from the six year to the 10 year mark, there's a decline in the percentages of former college students who earn less than the average high school graduate. This decline is seen for public, private, and for-profit colleges. In other words, some college attendees do catch up, and other studies show this trend continuing after the 10 year mark. (I don't mean to sound too cheerful though. For students who graduate from college in four years and immediately enter the work force, the 10 year mark is their 6th year of work. That feels like a long time to wait before earning more than you would've with your high school diploma – especially since, at this point, you haven't yet recouped the costs of attending college.)
Second, not everyone ranks earnings at the top of their list for why they purse higher education. Edward Wingenbach, president of Hampshire College, was interviewed by the NY Times in connection with the fact that only 46% of Hampshire students earn more than the average high school student six years after enrolling. Dr. Wingenbach pointed out that in surveys of incoming Hampshire students, "being well off financially" only ranked seventh among their future goals. He also noted that a majority of these students come to Hampshire because they wish to be activists, artists, educators, or entrepreneurs, and these career paths don't tend to have early career payoffs. (See Appendix below for additional details.)
Earnings data: Further qualifications
So far I've focused on differences between colleges in their impact on future earnings, but other variables are critical too. Here I'll describe three.
First, among those who enter the workforce with a bachelor's degree, choice of major strongly impacts lifetime earnings. In the figure below, the median value for architecture and engineering majors ($3.8 million) is nearly twice as high as for education majors ($2 million). In fact, among those with only a high school degree (not depicted in this figure), workers in computer and mathematical fields have higher median lifetime earnings ($2.6 million) than folks with bachelor's degrees in education. (If you're one of my former education students from SMU, I just want to say: This is so wrong. You deserve more money!)
Second, at all levels of education, gender disparities are found in lifetime earnings, as illustrated in the figure below. In a word, women earn less than men. Roughly speaking, there are two reasons for this: (a) women are more likely to pursue occupations in lower-paying fields such as counseling and education, and (b) within the same field, women tend to earn less than men, even when controlling for differences in variables such as experience.
Third, comparable differences are seen in racial/ethnic group comparisons, with white workers showing higher lifetime earnings at all educational levels (except among master's degree holders, where Asian workers earn the most.) As with gender, racial/ethnic group differences can be attributed to a combination of occupational choice and workplace discrimination.
In sum, earnings tend to be lower for college graduates who enter low-paying fields, and/or for women, and/or for members of racial/ethnic minority groups. However, overall, earnings are still higher among these groups than among people who do not attend college.
Conclusion: Is college worthwhile?
This newsletter looked at future earnings, which is only one of the many potential benefits of a college education. Other benefits include more diverse career options, greater job stability and satisfaction, access to health care, and networking opportunities. With respect to earnings, here's what I would tell students, families, and high school counselors:
1. College graduates tend to earn more than those with a high school degree alone.
2. The more education, the better. For example, people with bachelor's degrees tend to earn more than people with associate's degrees, while people with master's degrees tend to earn more than both.
3. College degrees don't automatically translate into greater earnings. The first two points are generalizations that gloss over a number of key variables. As noted below, these variables include college attended, major, and student demographics.
4. Colleges vary widely in the future earnings of those who attend. A substantial number of students from some colleges earn less than the average high school graduate. Over time, this gap diminishes. (Data for individual colleges can be found here.)
5. Future earnings are impacted by choice of major, and by demographic variables such as gender and race/ethnicity, although "the more education the better" maxim typically applies in all cases. Choosing certain majors and/or being a woman and/or being a racial/ethnic minority diminishes but does not completely negate the typical earnings advantage of a college education. (Data for majors and demographic variables can be found here.)
Postscript: When is college not worthwhile?
Overall, the message from this newsletter is that college tends to be a good investment, at least from the perspective of future earnings. But that "tends to" is a deliberate hedge. None of the data I've discussed allow for absolute statements that certain people, under certain conditions, will necessarily earn more if they attend college.
We might ask then whether there are people for whom college would never be a worthwhile investment. I think the answer is pretty much the same – absolute statements aren't warranted by the data. I would simply recommend extreme caution in certain cases. For example, if a majority of students from some college aren't earning as much as the typical high school graduate six to 10 years after enrolling, I wouldn't apply to that college unless I could figure out what's wrong (e.g., low graduation rates) and then ask myself how likely I am to be among the fortunate few.
Thanks for reading!
Appendix: Economic mobility
I'm including some information here for those interested in other ways to evaluate links between college and future earnings. This is in an appendix because, as you'll see, my discussion is more relevant to evaluating institutional performance than to an applicant's choice of which institution to attend.
In the NY Times interview mentioned above, Hampshire College president Edward Wingenbach noted that although not many former Hampshire students earn more than the typical high school graduate six years after enrolling, Hampshire's performance in economic mobility is above average. The measure of economic mobility he was referring to, also developed by Michael Itzkowitz, takes into account how many low- and moderate-income students a college serves, as well the time it takes former students to recoup their college-related expenses.
Itzkowitz's economic mobility index (EMI) diverges in startling and important ways from popular approaches to determining the financial "value" of attending a specific institution. For instance, U.S. News ranks Harvard #3 in its "Best Value Schools" category for 2022. In the EMI, also for 2022, Harvard ranks #847.
Why such a huge difference? In the U.S. News system, the academic quality of the school (as determined by....you guessed it...U.S. News rankings) contributes to 50% of a school's Best Value ranking, while the other 50% derives from three financial aid variables related to scholarships and grants. So, to simplify a bit, schools like Yale, MIT, and Harvard top the U.S. News list of Best Values because the prominence of each school compares favorably to the extent of students' tuition responsibility.
At the same time, schools like these perform poorly on the EMI because they enroll so few low-income students. Those they do enroll usually recoup any education-related costs they incur (Harvard ranks fourth in the nation in how quickly they do so), and their future earnings tend to be strong. There just aren't very many low-income students at these schools, hence their EMI ranking is low.
Now you can see why the EMI is suitable for evaluating the financial impact of a college but not necessarily relevant to the individual applicant. It's good to see from EMI data that HBCUs and Hispanic-Serving Institutions, for example, tend to have strong rankings. This tells us something about institutional performance. However, from an applicant's perspective, impact on future earnings is probably more relevant than EMI ranking. For instance, a college with a strong EMI ranking (e.g., Hampshire College) might be serving a lot of low-income students without having much impact (at least in the short term) on their future earnings relative to a high school degree alone. On the other hand, a college with a low EMI ranking (e.g., Harvard) might be serving relatively few low-income students, but providing those students with strong financial support en route to a lucrative career.