迭代递延接受机制

The iterative deferred acceptance mechanism

Games and Economic Behavior · 2022
被引 33 · 同刊同年前 2%
人大 AABS 3

中文导读

针对德国、巴西、中国等国大学招生中使用的序贯机制,指出其在激励和结果上的缺陷,并提出一种新的迭代递延接受机制,参与者每步选择最偏好学校的简单策略构成事后均衡,产生学生最优稳定匹配。

Abstract

Lately, there has been an increase in the use of sequential mechanisms, instead of the traditional direct counterparts, in college admissions in many countries, including Germany, Brazil, and China. We describe these mechanisms and identify their shortcomings in terms of incentives and outcome properties. We introduce a new mechanism, which improves upon these shortcomings. Unlike direct mechanisms, which ask students for a full preference ranking over colleges, our mechanism asks students to sequentially make choices or submit partial rankings from sets of colleges. These are used to produce a tentative allocation at each step. If at some point it is determined that a student can no longer be accepted into previous choice, then she is asked to make another choice among colleges that would tentatively accept her. Participants following the simple strategy of choosing the most-preferred college in each step is an ex-post equilibrium that yields the Student-Optimal Stable Matching.

迭代延迟接受机制序贯机制学生最优稳定匹配大学录取