Do Elite Universities Practise Meritocratic Admissions? Evidence from Cambridge

The merit-vs-diversity balance in university-admissions remains a controversial issue. Statistical analysis of these problems is jeopardized by applicant characteristics observed by admission-officers but unobserved by researchers. Using administrative microdata from the two-stage Cambridge admission-process, we compare post-entry exam-scores of directly admitted h-type students with g-types entering via the "pool" — a second-round clearing-mechanism. Better performance by the latter implies higher admission-standards for g-types, irrespective of the unobservability problem. We find strong evidence of higher admission-standards for males in STEM/Economics, and a weak one for private-school applicants. The gender-gap weakens over time for a cohort, and is non-evident in Law/Medicine