In a new NBER working paper released this month, economists from the University of Texas and Texas A&M scrutinize some of the factors motivating racial and ethnic differences in college application choices, using data from the entire population of high school graduates in Texas over the past two years. In disaggregating rather than lumping together minorities (as some other studies do), the authors find that Hispanic students are in particular less likely to apply to any college, even after controlling for college readiness and high school quality.

Three of the factors that seem to be most important when it comes to predicting where a student decides to apply are:

  • Distance between home and college
  • Presence of same-race/ethnicity students on college campus
  • Successful degree completion of same-race/ethnicity students from high school

The latter two are especially strong predictors for Hispanic students with low incomes. They are also factors about which students are likely to have highly uncertain information (which previous studies have cited as a basis for low-income, high-achieving students “undermatching” to colleges). To tease out informational asymmetries between students, the authors compare students automatically accepted to an in-state institution under the Texas Ten Percent plan (who face no information constraints about their admission to the state’s top schools) and those who barely miss the academic qualifications (and thus face the typical uncertainty of the college admissions process).

While perhaps not surprising, this paper’s finding – that certain types of traditionally underserved students prioritize racial and ethnic “familiarity” in their college application decisions – is an important addition to the recent literature examining the factors behind the student-college matching process. In a 2013 paper, Caroline Hoxby and Christopher Avery analyzed data on every student who took the SAT or ACT in 2008 and showed that low-income, high-achieving students tend to “undermatch” (i.e., not apply to colleges they are academically qualified for) as a result of being isolated from other high-achieving students. In another paper, Amanda Pallais found through a quasi-experiment that low-income students do apply to more selective universities when exposed to decreases in application costs.

So when it comes to optimizing the academic match between students and colleges, previous studies suggest that colleges and universities would be well advised to broaden their recruiting efforts (to reach students in more isolated schools) and decrease or even waive the application and testing costs for underserved students. But the Texas study also indicates that the social match for students may be even more important.

Both colleges and the high schools have a role to play in enhancing social match. Colleges such as the University of Texas at Austin have designed various student success programs targeted towards increasing the comfort level – and ultimately the academic success – of lower-income and minority students. And, as the Texas study shows, minority students who are still in high school and deciding on where to apply are often heavily influenced by where their older same-race/ethnicity classmates have enrolled – and succeeded. In short, efforts to strengthen the student-college match process, which tend to focus on the logistical barriers to access, can’t ignore the social barriers.