Some of the groups that the Dale & Krueger study reviewed included a large sample of students from a 1989 cohort of multiple races, and just the Black + Hispanic students from that same cohort. It found that selective of college as measured by average SAT score of college attended or Barron’s index had negligible impact on 2007 earnings among the full sample, after the controls which included both student characteristics and selectivity of colleges to which the students applied. However, they found selectivity of school attended was far more positively correlated with income among the Black + Hispanic subgroup. This effect was particularly noteworthy among URMs who did not attend HBCUs. The difference was even more stark for college tuition. Higher tuition was associated with much higher income for URMs after controls (as high as more selective college), but was associated with lower income for ORMs after controls.
For the ORM students, income was higher among students attending highly selective colleges without controls, but with the self revelation controls (controls for selectivity of schools to which the student applies), that increased income dropped to ~0. However, for URM students, there was essentially no difference between basic controls and self revelation controls. In short, for ORMs increased income was associated with applying to selective colleges, rather than attending, which might signal certain personality or family background characteristics. And for URMs, the income benefit seemed to requiring attending selective colleges. Applying to selective colleges without attending. had little impact for URMs.
The authors also mentioned that a statistically significant positive income effect occurred for the subgroup of students whose parents did not attend college. This income benefit only occurred for the 1989 cohort – not the 1976 cohort. The 1989 cohort income benefit for first gens appeared to be much smaller than the income benefit for URMs. The author’s speculate that the URMs and less educated family subgroups may benefit from the network connections. This also fits with URM income seeming to be as correlated with high tuition as selectivity. There may be benefits from networking with wealthy high SES kids and/or being exposed to the way they look for jobs with internships and such, even if not at the most selective college to which the student is accepted.
"One possible explanation for this pattern is that while most students who apply to selective colleges may be able to rely on their families and friends to provide job-networking opportunities, networking opportunities that become available from attending a selective college may be particularly valuable for black and Hispanic students, and for students from less educated families. "