The pitfalls of higher selectivity: Evidence on university degree completion

We study the effect of selectivity on university degree completion. Using data from the centralized university admission system in Chile, which includes students’ ranked options,we use a regression discontinuity design with students at the margin of admission between relatively more and less selective university programs. We find that students marginally admitted to a more selective program are 2.3 percentage points less likely to graduate from their initial program, compared to marginally rejected students who were instead admitted to the relatively less selective option. We also find a 1.9 percentage point drop in the likelihood of obtaining any university degree. For students who graduate, we find a one-month increase in time to graduation. This decline in overall graduation rates is driven by lower-income students, who face greater challenges in transferring and completing a degree elsewhere. We find suggestive evidence that these results are driven by higher academic rigor, rather than by ordinal rank.