SBEE seminar series: Siqi Pan
03:00 pm -
Costly information acquisition in centralized matching markets
Co-authors: Rustamdjan Hakimov (University of Lausanne & WZB Berlin Social Science Center) and Dorothea Kübler (WZB Berlin Social Science Center & Technical University Berlin)
Every year during school and college admissions, students and their parents devote considerable time and effort to acquiring costly information about their own preferences. In a market where students are ranked by universities based on exam scores, we use a market design approach to explore ways to reduce wasteful information acquisition, that is, to help students avoid acquiring information about their out-of-reach schools or universities. We find that, both theoretically and experimentally, a sequential serial dictatorship mechanism leads to less wasteful information acquisition and higher student welfare than a direct serial dictatorship mechanism.
This is because the sequential mechanism informs students about which universities are willing to admit them, thereby directing their search. Additionally, our experiments show that the sequential mechanism has behavioral advantages because subjects deviate from the optimal search strategy less frequently under the sequential than under the direct mechanism. We also investigate the effects of providing historical cutoff scores under the direct mechanism. We find that the provision of cutoffs can increase student welfare especially when the information costs are high, although the effect is smaller than that of a sequential mechanism.
Siqi Pan is a Lecturer (Assistant Professor) at the Department of Economics at the University of Melbourne. Her research interests include market design, experimental & behavioral economics, game theory, and industrial organization. She holds a PhD from The Ohio State University.
The Social, Behavioral and Experimental Economics seminar series is a joint presentation of the School of Information, the Ross School of Business and the Department of Economics (LSA).
For information on how to watch this lecture and sign up for the SBEE mailing list to receive notice of upcoming events, please visit the SBEE website: https://umbee.github.io/SBEE_Seminars