Establishing causation: elite college and earnings

Question

Graduates of elite universities earn more than others, but you cannot randomly admit students. You want the causal effect of attending an elite school on later earnings.

Describe how you would estimate it from observational data. Name at least two identification strategies and the key assumption each relies on.

Accepted Answer

The obstacle is confounding by ability and background: elite schools admit already-talented, well-connected students who would earn more anyway, so a raw graduate comparison overstates the school's effect. Several designs address it. 1. Regression discontinuity at the admission cutoff. Where admission hinges on a score threshold, compare applicants just above the cutoff (admitted) with those just below (rejected). Near the line, admission is essentially random. Key assumption: applicants cannot precisely control their score around the cutoff, and nothing else about them changes discontinuousl…

Establishing causation: elite college and earnings

Your answer

Discussion

Learn the concepts

Related questions