The Returns to Elite University Education: a Quasi-Experimental Analysis

Abstract

I take advantage of a discontinuity in the probability of admission to a highly selective private university to estimate causal returns to investing in elite university education. I use a newly assembled data set that combines individual administrative records about high school attendance, university admission, university attendance, and tax returns. I find a discontinuity in income of 38 log points at the admission cutoff. The fuzzy regression discontinuity estimate for the elite enrollment effect is 58 log points. This should be interpreted as the average treatment effect for students applying to the elite university who are close to the cutoff and chose to enroll. When I take into account the evidence that students enrolling in the elite university tend to make different field choices, the net institutional enrollment premium is 41 log points. Cumulated over 15 years, the net-of-tuition elite premium is €246,991. I explore potential channels explaining the sizeable enrollment effects and I find that students just above the admission cutoff are 15 percentage points more likely to complete a university degree, they are 26 percentage points more likely to graduate on time and attend university with substantially higher quality peers.

Recruiting and supporting low-income, high-achieving students at flagship universities

“…There are some comparison schools in these cities as well but they tend to be located outside the urban centers and reflect the fact that these programs faced budget constraints that allowed them only to treat a subset of qualifying schools. Figures 1 and 2 suggest that there is plausibly-exogenous variation in treatment status based on geography that allows us to identify β 1 in equations (1) and (2). While in our main estimates we do not explicitly take advantage of this, we provide specification checks that restrict our comparison schools so that they are not in districts or counties with treated schools and find similar results.…”

“…Columns (2) and (3) of Table 5 provide estimates of the impact of attending an LOS or CS high school on enrollment at a flagship. We find an increase in attendance of 2.2 percentage points in UT-Austin due to LOS exposure.…”

“…For example, in the first panel column (1) shows the effect of the LOS program on the likelihood students begin at UT-Austin and graduate from a flagship. The estimates in column (2) in the third panel show the effect of LOS on the likelihood a student first enrolled in an ERU and then graduated from a non-flagship four-year school. All graduation outcomes are within six years.…”

“…In the NLSY97, only 2% of low-income students attended a flagship public school, while among the highest-income students 16% did. 2 The likelihood of attending a private school also increases with income, and the proportion of students enrolling in a two-year school declines with income. There is substantial evidence of large impacts of college quality on completion .…”

The Returns to College(s): Estimating Value-Added and Match Effects in Higher Education

“…Many of the colleges in our data and across the U.S. practice "holistic admissions" with no sharp quantitative admissions cutoffs. This hampers a regression discontinuity research design, which has been increasingly employed to study admission impacts in non-U.S. settings (e.g.Saavedra, 2009; Hastings et al, 2014; Kirkeboen et al, 2016; Canaan and Mouganie, 2018;Anelli, 2019; Jia and Li, 2019), as well as at specific cutoff-using U.S. institutions in binary treatment frameworks that pool various mixtures of other 4-year institutions, 2-year colleges, and non-attendance into a composite counterfactual, rather than isolating particular institutional comparisons (e.g.Hoekstra, 2009; Zimmerman, 2014; Goodman et al, 2017).6 Subsequent variants of the matched applicant approach include Arcidiacono (2005), Long (2008), Fryer and Greenstone (2010), Cunha and Miller (2014), Arcidiacono et al (2016), and Ge et al (2018).…”