Decentralized College Admissions

We study decentralized colleges admissions in the face of uncertain student preferences. Enrollment uncertainty causes colleges to strategically target their admissions to students overlooked by others. Highly ranked students may receive fewer admissions or suffer from a higher chance of coming up empty – “falling through the cracks” – than those ranked below. When students' attributes are multidimensional, colleges avoid head-on competition by placing excessive weights on school-specific measures such as essays. Restricting the number of applications or wait-listing alleviates enrollment uncertainty, but the outcomes are inefficient and unfair. A centralized matching via Gale and Shapley's deferred acceptance algorithm attains efficiency and fairness, but some college may be worse off relative to decentralized matching.

Click here to view paper.