Breaking Barriers to College for High-Achieving, Low-Income Students

The Issue:

College graduates have more career opportunities and earn substantially more over their lifetimes than those who attain only a high school degree. There is even more of an earnings boost from attending a highly selective college. Yet low-income students with strong academic credentials are less likely to attend a highly selective college than students from higher-income homes. This is especially puzzling, given that selective schools tend to offer low-income students more generous financial support and have greater resources that increase the students' chances of completing a degree.

The majority of low-income students with strong academic credentials do not apply to highly selective colleges.

The Facts:

• Students from low-income families are underrepresented in U.S. colleges, with the imbalance being larger at the most selective colleges. Only about 12 percent of college students come from families in the bottom fifth of the income distribution. But, at the most selective schools — those in the ivy league, Stanford, Duke, MIT and the University of Chicago — there are more students from the top 1 percent of the income distribution than the entire bottom half (see here). This gap in college attendance between low- and higher-income students is partly due to differences by income in academic preparation prior to college. But even among well-prepared students, there are substantial gaps in college enrollment and the quality of college attended.
• An important share of students from low-income families attends schools that are less selective than schools they could qualify for. This "undermatching" in effect means that students attend institutions that have average test scores significantly below their own, indicating that they could have attended a more competitive school. In Michigan, for instance, my co-authors and I find large differences in college choices among similarly qualified students from low- and higher-income homes. Among students whose academic achievement makes them likely candidates for the state's flagship university, those from low-income homes lag behind their higher income peers in college attendance. The gaps are wider for more selective schools. For instance, lower-income students are 8 percentage points less likely to attend a highly selective institution, such as the University of Michigan, than higher-income peers (see chart).
• It is not so much that low-income students are rejected by the better schools. Rather, the majority of low-income students with strong academic credentials do not even apply to highly selective colleges. For instance one study found that the majority of low-income students with grade point averages of A- or above and who scored in the 90th percentile of college admission tests did not apply to any selective colleges at all. This mismatch of academic ability and college choice persists even though these students would often qualify for generous financial aid. Low-income students often do not apply to these more selective schools because they are uncertain about whether they are suitable for an elite school; because they overestimate how much college is going to cost them; and because parts of the process — such as filling out financial aid forms — present large procedural barriers. Some of this lack of information stems from the fact that many high-achieving, low-income students come from small school districts where teachers, classmates, and alumni are not familiar with selective colleges (see here). In addition, because of their geographic isolation, colleges’ traditional outreach programs do not reach them. (See also this EconoFact on distance as a barrier to higher education).
• Low-cost low-touch interventions can successfully close the income gap in college selectivity. Providing high-achieving, low-income high school students with better information has been shown to increase their applications to and attendance at more selective colleges. Together with University of Michigan administrators, we designed the low-cost, low-touch HAIL scholarship and outreach program. We identified all rising seniors in Michigan whose grade point average and college test scores met criteria set by admissions officials at the University of Michigan and who were eligible for subsidized meals (which corresponded to a family income below $44,863 for a family of four in 2015). The university sent an informational packet directly to roughly half of the identified high-achieving, low-income rising high school seniors. Information about it was also sent to their parents and principals. Students were told they could succeed at a highly competitive college like the University of Michigan, were encouraged to apply, and promised four years of free tuition and fees if admitted. Sending this HAIL offer to only half of the identified target students allowed comparison between those who received the packet and their peers who did not. Importantly, the program did not involve adding new scholarships: The students who were identified for the study would already have been eligible for at least free tuition and fees in the absence of the intervention.
• The students who received the HAIL offer were substantially more likely to apply to, be admitted to, and attend the University of Michigan. Our study evaluated the effects of the 2015 and 2016 HAIL scholarship offer on application, admission, enrollment, and persistence. The effects of the HAIL scholarship were extremely large and unlikely to have occurred by chance. The likelihood of applying to the University of Michigan more than doubled among the students who received the packet: 67 percent applied, compared to only 26 percent among the students who had similar qualifications but did not receive the HAIL outreach. Overall, the share enrolling at a highly selective program more than doubled, with 28 percent of HAIL recipients enrolling, compared to 13 percent in the control group. The increase in enrollment at the University of Michigan was driven by students who would otherwise not have been in college and those who would have gone to a community college or to a less-selective four-year college.
• The results from HAIL suggest isolation is an important cause of undermatching in Michigan. The effects of the HAIL offer differed substantially across regions of the state and were largest in rural areas. The increase in application and enrollment rates were largest in high schools that previously had no students apply to or enroll at the University of Michigan and where there were the fewest students who qualified for the scholarship. The effects were much larger among students who initially had not considered applying to the University of Michigan. In addition, HAIL narrowed the gender gap in applications to Michigan and widened the female advantage in enrollment. The effect of HAIL increased with time. HAIL scholars persisted in college at substantially higher rates than their peers who were not in the program.