中文摘要
本文分析了台湾用于学生与高中匹配的机制的福利性质。我在一个理论模型下研究台湾机制,该模型中学生具有相同的排序,且由单次抽签决定学校的偏好。在这一设定下,对于进行策略操纵的学生而言,所有台湾机制在事前都弱帕累托优于延迟接受机制。此外,在我的模型中,总存在一种台湾机制,为不进行策略操纵的学生提供高于延迟接受机制的福利。利用剑桥公立学校数据,我表明在一系列扣除规则下,台湾机制的表现优于延迟接受机制。
Abstract
This paper analyzes the welfare properties of the mechanism Taiwan uses to match students to high schools. I study the Taiwan mechanism under a theoretical model where students share the same ordinal rankings, and a single lottery determines schools’ preferences. In this setting, all Taiwan mechanisms ex ante weakly Pareto dominate deferred acceptance for strategizing students. Additionally, in my model, there always exists a Taiwan mechanism that provides greater welfare than deferred acceptance to students who do not strategize. Using Cambridge Public School data, I demonstrate that the Taiwan mechanism outperforms deferred acceptance for a range of deduction rules.