Doctoral Dissertation Research in Economics: Evolutionary Preference Revelation Dynamics under School Choice Mechanisms

在美国，孩子们传统上是根据他们居住的地方分配给公立学校的。然而，越来越多的公立学区现在允许家长指明他们希望自己的孩子上哪所学校。由于每所学校只能支持有限数量的学生，通常不可能让每个学生都首选学校。为了解决这种短缺，政策制定者采用了学生分配机制，根据报告的学校偏好和每所学校合法确定的学生优先事项，将每个学生分配到一所学校。在这项研究中，我们进行了实验室实验，以考察三种广泛使用的择校机制在连续时间变量下的动态行为。在这些连续时间机制中，参与者异步调整他们的偏好报告，并接收关于他们分配的学校的即时反馈。通过提供相当大的学习和调整机会，这些连续时间机制可以促进理性行为，实现更高的配置效率。因此，这项研究可以帮助政策制定者为他们所在的地区选择可能的最佳择校机制，并改善就读公立学校的儿童的教育结果。在可操作的机制下，如广泛使用的波士顿机制，参与者经常有动机策略性地谎报他们的偏好。为了鼓励真实的偏好报告，机制设计者通常采用防策略机制，在这种机制下，参与者永远不会有动机错误报告偏好，因此真实的偏好报告是一种弱主导策略。与这些理论预测相反，之前的实验研究发现，即使是策略防御机制也经常无法诱导真实的偏好揭示。与以前的研究不同，我们考虑的是分配机制，在这种机制下，参与者异步调整他们的偏好报告，并接收关于他们分配的学校的即时反馈。特别是，我们考虑了三种广泛使用的择校机制的连续时间变化：延期接受机制、最高交易周期机制和波士顿机制。由于这些持续时间机制提供了相当大的学习和调整机会，它们可能会促进理性偏好揭示行为，从而在策略验证机制下导致更稳定和更有效的结果。

Children in the United States are traditionally assigned to public schools based on where they live. However, a growing number of public school districts now allow parents to indicate which schools they want their children to attend. Since each school can support only a limited number of students, it is often impossible to give every student her top choice of schools. To resolve such shortages, policy makers employ student assignment mechanisms that assign each student to a school based on the reported preferences over schools and the legally determined student priorities at each school. In this study, we conduct laboratory experiments to investigate dynamic behavior under continuous-time variants of three widely employed school choice mechanisms. In these continuous-time mechanisms, participants adjust their preference reports asynchronously and receive instantaneous feedback regarding their assigned school. By providing considerable opportunity for learning and adjustment, these continuous-time mechanisms can promote rational behavior and achieve higher allocative efficiency. Accordingly, this study can help policy makers to select the best possible school choice mechanism for their district and improve educational outcomes for children who attend public schools.Under manipulable mechanisms, such as the widely employed Boston mechanism, participants frequently have an incentive to strategically misreport their preferences. To encourage truthful preference reports, mechanism designers often employ strategy-proof mechanisms under which participants never have an incentive to misreport preferences, so truthful preference reporting is a weakly dominant strategy. In contrast to these theoretical predictions, previous experimental studies find that even strategy-proof mechanisms frequently fail to induce truthful preference revelation. Unlike previous studies, we consider assignment mechanisms under which participants adjust their preference reports asynchronously and receive instantaneous feedback regarding their assigned school. In particular, we consider continuous-time variations of three widely employed school choice mechanisms: the deferred acceptance mechanism, the top trading cycles mechanism, and the Boston mechanism. Since these continuous-time mechanisms provide considerable opportunity for learning and adjustment, they may promote rational preference revelation behavior, leading to more stable and more efficient outcomes under strategy proof mechanisms.