The Colleges & Majors Delivering Economic Mobility for Low Earners
An analysis of programs at schools serving disadvantaged students with a high return on investment.
Key Points
- This issue brief analyzes the return on investment (ROI) of undergraduate postsecondary education programs by the average socioeconomic status of students attending each institution. ROI compares the main financial benefit of college — the increase in lifetime income attributable to the degree — to the costs, including tuition and foregone earnings.
- Schools serving low-income students tend to offer lower ROI. Nearly half of the programs at institutions in the bottom quartile of student socioeconomic status do not have a financial return.
- Seventeen percent of programs at schools serving mostly low-income students have an ROI exceeding $250,000. Public community colleges and private trade schools offer most of these high-value programs, which are concentrated in health care fields.
- Policymakers should give federally funded institutions incentives to expand programs that serve low-income students well by boosting Pell Grant aid for those pursuing high-ROI credentials.
Introduction
College should, in most cases, provide a positive financial return. While increased earnings potential and employability are not the only goals of higher education, students tend to list them as their top objectives in pursuing education after high school. Financial returns are especially important for students from low-income families, for whom postsecondary education is a key pathway to upward mobility.
But the return on investment (ROI) of college is uneven across different programs and institutions. A student’s choice of major makes a big difference to the return he or she can expect from college. The choice of institution is also important. Schools with higher tuition or lower graduation rates will have lower expected returns, all else being equal.
FREOPP has calculated ROI for 47,000 different undergraduate degrees and certificates receiving federal student aid funding. ROI is defined as the expected increase in lifetime earnings associated with a degree or certificate, minus the costs of obtaining it. FREOPP’s estimates are available at the program level for all sorts of credentials, from Ivy League universities to community colleges and trade schools, and the estimates are adjusted to account for the types of students each school enrolls. This enables an analysis of ROI for schools serving students from many different backgrounds.
A student’s socioeconomic status heavily influences where he or she attends college. According to research by Raj Chetty and colleagues, some elite private schools enroll more students from families in the top one percent of the income distribution than students from the bottom half.
Chetty’s research also shows that schools enrolling students from wealthier families tend to produce higher-earning graduates. Students who attend Ivy League and comparable universities typically end up in the 80th percentile of the earnings distribution as adults, while community college students typically reach only the 60th percentile.
However, one limitation of the Chetty research is its focus on institution-level economic outcomes. The results of FREOPP’s research on college ROI show that the field of study drives most of the variation in the financial returns to higher education. A nursing degree from an open-enrollment community college often produces higher ROI than an art or music degree from a highly selective private university.
These facts are of key importance to policymakers who wish to ensure that postsecondary education provides a reliable path to the middle class for students from all walks of life. This issue brief identifies programs that deliver strong ROI at schools enrolling large numbers of low-income students. By emulating these programs, other schools may be able to replicate their success at providing economic opportunity to prospective students who least have it.
Methods
Identifying high-value programs at schools serving predominantly low-income students requires both a measure of programs’ ROI and an indicator of the socioeconomic status of each school’s student body.
FREOPP has calculated ROI for 30,000 bachelor’s degree programs and 17,000 associate degree and certificate programs. The FREOPP measure of ROI compares the estimated lifetime earnings associated with each program to a counterfactual earnings profile, or what the typical student in that program would have earned had he or she never gone to college.
The counterfactual earnings profile adjusts for the demographic composition of programs. Since women without a college education tend to earn less than men without a college education, programs that enroll mostly women will have lower counterfactual earnings. There is also an across-the-board adjustment for selection bias: since students who choose to attend college tend to have higher socioeconomic status and stronger demonstrated academic ability, one can expect college-goers to have higher earnings potential than high school graduates.
These adjustments attempt to bring FREOPP’s ROI measure in line with the true financial benefit associated with each degree, by comparing students’ earnings with the degree to what those students would have earned in a parallel universe where they never attended college. Of course, there may be unobservable factors affecting earnings potential that I do not account for.
The next challenge is to identify which schools serve predominantly low-income students. The government does not collect high-quality data on this question: income data is only reported for first-time, full-time students at schools with a standard calendar system. These limitations miss millions of part-time and returning students, who are likelier to come from low-income backgrounds.
Because of these issues, many researchers use the share of students at an institution who receive the federal Pell Grant, a means-tested aid program for low- and middle-income students, to measure the socioeconomic status of a student body. However, the “Pell share” measure has its own problems, chiefly that it simply treats Pell Grant recipience as a binary indicator of student disadvantage. But there is incredible diversity within the Pell Grant recipient population.
Pell share lumps together a student from a poor family earning $25,000 per year — who thus qualifies for the maximum Pell award — with a student from a middle-class family earning $70,000 per year, who qualifies for only a small award. Nearly 40 percent of undergraduates received some Pell aid in 2015–16, so Pell share doesn’t necessarily identify schools serving the most disadvantaged students.
I propose an alternative metric to measure the socioeconomic status of a school’s students: average Pell Grant per full-time-equivalent undergraduate student. Since lower-income students receive larger Pell Grants, a higher average Pell Grant indicates more students from low-income backgrounds. High-income students who receive no Pell aid are included in the average, so schools that enroll mostly rich students will have a low average Pell Grant.
Across nearly 6,000 institutions with available data, the correlation between Pell share and average Pell Grant is 0.85. The two metrics, therefore, measure mostly the same thing, but there are exceptions. At the private nonprofit mega-school Southern New Hampshire University (SNHU), for instance, 47 percent of students receive a Pell Grant, which puts SNHU in the top quartile of institutions by this metric. But SNHU’s average Pell Grant is $1,583, which puts it below the median.
SNHU therefore seems to enroll plenty of middle-income students who qualify for a partial Pell Grant, but fewer students at the low end of the socioeconomic spectrum who receive the maximum grant or close to it. Considering Pell share alone elides this sort of nuance.
Which schools enroll low-income students?
I use average Pell Grant figures to divide institutions into four enrollment-weighted quartiles based on the socioeconomic status of their student body. The average Pell Grant ranges from a minimum of $0 — indicating that no students at this institution receive Pell aid — to a theoretical maximum of $5,943 — the average of the maximum Pell Grant in 2016–17, 2017–18, and 2018–19, the three years from which I collect data — which would indicate that every student at that institution is poor enough to receive the maximum grant.
Institutions in the first quartile have an average Pell Grant ranging between $0 and $1,226. These institutions enroll the highest-income students, on average. Over half of the first-quartile schools are research universities (institutions that award doctorates). The first quartile also includes elite liberal arts institutions such as Oberlin College, where so few students receive Pell aid that the average Pell Grant is below the statutory minimum grant. Around 17 percent of first-quartile schools are public community colleges, especially those in wealthy areas, such as Maryland’s Anne Arundel Community College.
Institutions in the second quartile have an average Pell Grant ranging between $1,226 and $1,704. These schools’ students are not quite as high-income, but still above the average. The second quartile consists mainly of four-year institutions, with more public and fewer private colleges than the first quartile. It includes large public universities such as Oregon State University and the University of California-Santa Cruz. There are also several large, mostly online schools in this group such as Western Governors University and Southern New Hampshire University, which may serve a more nontraditional student population.
Institutions in the third quartile have an average Pell Grant ranging between $1,705 and $2,253, representing the third-richest or second-poorest quartile. Two-fifths of this group are public community colleges — the highest share in any quartile — including many serving underrepresented minorities, such as California’s Compton College. However, 29 percent of this group are research universities, such as the University of Central Florida.
Institutions in the fourth quartile have an average Pell Grant of $2,254 or higher. These schools serve the poorest students on average. The fourth quartile is roughly evenly split between four-year and less-than-four-year schools. Among four-year schools, there are several Historically Black Colleges and Universities such as Grambling State University, as well as research institutions that have made a concerted effort to enroll low-income students like Georgia State University.
This quartile includes the highest share of for-profit colleges — by far — as well as the highest share of trade schools that exclusively grant certificates rather than degrees. Large for-profit institutions such as Grand Canyon University appear in the fourth quartile, along with smaller, specialized for-profits like High Desert Medical College.
Colleges serving lower-income students, as measured by average Pell Grants, tend to offer programs with lower ROI. Fifty-two percent of programs at the richest schools offer a financial return of $250,000 or more, compared to just 17 percent of programs at the poorest schools. In fact, almost half of the programs at schools serving the lowest-income students have negative ROI. This statistic implies that whether a college education will pay off is close to a coin flip for many disadvantaged students.
Why is there a negative correlation between socioeconomic status and ROI? It is possible that low-income students have lower earnings potential, even with a college degree. While FREOPP’s ROI analysis generally estimates lower counterfactual earnings for schools serving more low-income students, ROI for poorer schools may still be lower even after accounting for this effect. Lower-income students also have lower college completion rates, which reduces the expected return to higher education as there is a greater chance that these students will not realize the earnings benefits that come with a degree.
The good news is that there are many examples of programs at institutions serving low-income students that still deliver high ROI. By focusing on these programs, other institutions can learn what works and revamp their own educational offerings to emphasize programs that deliver a financial return for disadvantaged students. Moreover, policymakers can create incentives for schools to invest in these high-quality programs and deemphasize those that are not serving students well.
How low-income students can reach the middle class
Students can expect to benefit financially from higher education if the ROI of their program is greater than zero. But low-income students are most likely to realize economic mobility from programs that generate a significant boost in lifetime earnings, not those that have a bare minimum financial return. Therefore, the analysis to follow focuses on the subset of programs supplying an expected lifetime ROI of at least $250,000, adjusting for the likelihood of completion (hereafter “high-value” or “high-ROI” programs).
At the richest schools, over 95 percent of high-value programs are bachelor’s degrees, weighting by enrollment. Few schools serving rich students offer subbaccalaureate credentials — associate degrees and certificates — with a lifetime return of $250,000 or more. Bachelor’s degree programs account for more than half the output of high-value credentials at schools in the middle of the socioeconomic distribution as well.
The picture looks different at schools serving low-income students. At the poorest schools, just 30 percent of high-value programs are bachelor’s degrees, and nearly half are undergraduate certificates. Schools serving low-income students are much more likely to move low-income students into the middle class through the skilled trades, while schools serving high-income students rely on the bachelor’s degree to create economic value.
The sectoral composition of schools providing high-ROI credentials also differs by student socioeconomic status. Ninety-seven percent of high-value programs at the richest schools are offered by public or private nonprofit institutions; the vast majority of these are research universities.
Contrast this with schools serving the poorest students. Around half of high-value programs at the poorest schools are offered by public institutions, but for-profit institutions offer another 42 percent. For-profit schools’ role in economic mobility for poor students is almost exclusively confined to less-than-four-year institutions. For-profit schools granting bachelor’s degrees offer comparatively few high-value programs to low-income students.
The reader should bear in mind that schools are classified according to the average socioeconomic status of their student body. Naturally, not everyone attends a school with peers of a similar socioeconomic status. Low-income students may attend schools full of high-income students and vice versa; thus, many low-income students will benefit from the high-ROI programs offered by rich schools. The figures here are meant to show us how schools serving a predominantly low-income population can help their students reach the middle class.
Fields of study that promise economic mobility
FREOPP’s research on ROI emphasizes that field of study is the primary determinant of higher education’s economic value. The field of study explains about half the variation in ROI. The difference in ROI is usually greater for two different majors at the same institution than for the same major at two different institutions. That is, two engineering majors who attended different schools are likelier to have similar outcomes than an engineering major and a sociology major who attended the same school.
For the richest schools, the most popular high-value field of study is a bachelor’s degree in business fields, which accounts for 27 percent of the high-ROI credentials conferred at these schools. Bachelor’s degrees in engineering and nursing also contribute to the body of high-value degrees conferred at wealthy schools.
However, some majors not traditionally seen as lucrative — including communications and the social sciences — also comprise a significant share of the high-value degrees issued at the richest schools. It is possible that access to elite professional networks at schools serving rich students increases ROI in majors that do not typically have a strong financial return. At schools serving lower-income students, these elite professional networks may not be present, or have a smaller effect on students’ career opportunities.
At schools serving students towards the middle of the socioeconomic distribution, bachelor’s degrees in health and nursing make up the largest category of high-ROI credentials. Fields such as business and engineering become relatively less important as school socioeconomic status declines. Bachelor’s degrees in communications and social sciences vanish from the population of high-value credentials entirely.
At the poorest schools, students are most likely to achieve high ROI by pursuing a credential in health care fields. Over half of the high-ROI credentials at schools serving low-income students are in health care. Four-fifths of these are subbaccalaureate credentials, especially associate degrees in registered nursing and undergraduate certificates in licensed practical nursing. Certificate programs in other fields, such as precision metalworking, vehicle maintenance and repair, and ground transportation are important as well.
While bachelor’s degrees outside of health care are rarer among the body of high-ROI credentials at schools serving low-income students, there are a handful of bachelor’s degrees in fields such as accounting, finance, and mechanical engineering that contribute to economic mobility at these schools.
Most valuable programs at schools serving low-income students
In a previous issue brief on the ROI of state university systems, FREOPP introduced the “Most Valuable Program” (MVP) ranking. The MVP list aims to measure not only the economic value of a degree or certificate program, but the number of students that program serves. Programs that generate high returns but lack the ability to scale are unlikely to be engines of economic mobility.
To construct the MVP list, I multiply the ROI of each program by the number of graduates between 2015 and 2017. The resulting figure is a measure of the total economic value that program creates. It indicates which programs are strong at both inclusive enrollment and financial returns.
The table below lists the top 25 MVP programs from schools in the fourth quartile of student socioeconomic status. It indicates which programs create the most economic value for students from disadvantaged backgrounds. For instance, Tulsa Welding School’s undergraduate certificate in precision metal working offers ROI of $213,310 and graduated 5,944 students over the study period. Multiplying these figures together yields a total value created of nearly $1.3 billion.
Undergraduate certificate programs dominate the MVP list. Alongside Tulsa Welding School’s precision metalworking certificate, the campuses of the New England Tractor Trailer Training School, offering certificates in commercial truck driving, appear several times. The short duration of these certificate programs — often less than a year — and their lower costs relative to a four-year degree may make them a more attractive option for students from families with limited resources or doubts about their ability to finish college.
Seven bachelor’s degree programs appear on the MVP list. Five of these are programs in registered nursing, many at campuses of the California State University (CSU) system, known for its high admissions rates and generous need-based grant aid. Another MVP program is a financial management degree at CUNY Baruch College. The school’s proximity to New York City’s financial sector, along with its senior-year internship program, may assist students in landing well-paying jobs in their field after graduation.
Improving ROI for low-income students
The gap in ROI between higher education programs serving low-income and high-income students is cause for concern. While schools serving high-income students produce a positive financial return more than 80 percent of the time, nearly half of the programs at schools serving low-income students are negative-ROI. Students at these schools deserve better options.
Fortunately, there are promising models to improve outcomes. Many trade schools — both public and private — offer vocation-oriented instruction with high returns. Bachelor’s degree programs in fields that directly prepare students for a job, especially in nursing and finance, also appear to serve low-income students well.
FREOPP’s ROI calculations rely on federal data, so they do not account for postsecondary pathways that operate outside of the federal student aid system. These include promising models such as apprenticeships and short-term skills-training programs. Congress should explore avenues to ensure that alternative postsecondary models operate on a level playing field with traditional higher education when it comes to federal funding. For instance, a bill introduced by Representative Elise Stefanik (R-NY) would expand Pell Grant eligibility to short-term programs that deliver a significant increase in earnings for graduates, potentially enabling more low-income students to access these options.
Policymakers should also create incentives for schools to offer high-ROI programs. Federal student aid funding is largely not contingent on student outcomes, meaning schools get paid regardless of how well they serve students. There is little incentive for schools to invest in high-quality programs, which may be more expensive to operate. For instance, American nursing schools turn away 80,000 qualified applicants every year due to lack of capacity, even though nursing is a reliably high-ROI field.
FREOPP’s white paper on higher education reform proposes requiring schools to compensate the government for unpaid student loans, which would deter schools from offering low-value programs that leave graduates unable to repay their debts. The proposal reinvests the savings from this policy into additional Pell Grant funding for students who pursue programs that deliver high earnings for moderate prices. This policy will enable more low-income students to afford high-quality programs.
High-quality schools will also benefit: if more financial aid is available, more students will enroll in high-ROI programs. This will yield more revenue for schools that offer these programs and enroll Pell Grant students. According to FREOPP’s calculations, community colleges and trade schools — which disproportionately enroll low-income students — will benefit most, since their tuition is lower, but only if they also offer programs that deliver strong outcomes.
The American higher education system promises economic mobility but often fails the most vulnerable students. Fortunately, there are plenty of success stories that illuminate a path forward, so long as schools are given the incentive to build upon what works.
