Recovering Preferences in College Assignment Problems Under Strategic and Constrained Reports
提出一种新方法，在考生面临报告选项数量限制且策略性报告的情况下，恢复其真实偏好，并应用于哥斯达黎加大学数据评估替代录取机制。
ABSTRACT Many countries around the world use centralized admission systems to assign thousands of students to university programs every year. However, the impact of admission policy changes on students' welfare remains largely unknown. To address this, we propose a novel methodology that recovers students' true preferences in a mechanism where students hold private information and face constraints on the number of options they can submit as preferences. This methodology involves two steps: first, the nonparametric recovery of ordinal preferences using a revealed preference approach based on a concatenation algorithm; and, second, the recovery of cardinal preferences using either a Gibbs sampler estimator or a calibration method. Applying our methodology to administrative data from the University of Costa Rica, we evaluate counterfactual policies and suggest two alternate mechanisms based on an ascending auction and competitive equilibrium prices.