Substitution effects in college admissions

I show how local supply changes create ripple effects in a national educational market. Admitting an applicant to a program will free up a slot to be filled at her next-best alternative. To investigate such substitution effects I re-engineer the centralized admission system of the Danish tertiary education sector and simulate equilibria under counterfactual supply. I estimate potential earnings with a regression discontinuity design and quantify market clearings in terms of earnings. On average, a change of 10 slots leads to 15 applicants moving and substitution effects explain 40 percent of the variation in earnings. Substitution externalities are generally positive but vary in sign and magnitude. I document a trade-off between earnings and inequality.