Using Market Design to Improve College Transfers of Student-Athletes

Présentation de Robert G. Hammond - Department of Economics, Finance, and Legal Studies, University of Alabama.
Présentation de Robert G. Hammond

Gregory Dowd - Department of Economics, North Carolina State University
Umut Dury - Department of Economics, North Carolina State University.
Robert G. Hammond - Department of Economics, Finance, and Legal Studies, University of Alabama.
Utku Unverx - Department of Economics, Boston College.

Abstract: Several areas of economics have become the source of collaboration between policymakers and economists. Of these, matching markets are one of the most productive areas of research. We study the transfer market for student-athletes administered by the National Collegiate Athletic Association (NCAA) in the US. Colleges annually recruit high school athletes for their athletic teams, but some years after enrolling, there are student-athletes who would like to transfer to another college for various reasons. As of 2019, around 30% of the student-athletes in NCAA Division I basketball have transferred. After transferring, a player usually needs to sit out one year before playing for their new college. To gain immediate eligibility, a player must submit a waiver request. But waiver requests are administered in an opaque way that lacks consistency. In this paper, we propose a new design to the NCAA transfer market. Within a matching with contracts framework, the theoretical model considers a two-sided market composed of colleges and students wishing to transfer. We construct a new mechanism under which immediate eligibility of a student is conditioned on the set of incoming students to his home college. Our proposed solution improves in a meaningful way upon current practice of the NCAA. The empirical analysis includes data from the initial recruitment period and from the transfer period. The initial recruitment data allow us to estimate students' and colleges' preferences. The transfer data allow us to measure what improvements are possible with our approach. We coffirm that our proposed solution leads to large effciency increases.