Timing of Preference Submissions under the Boston mechanism [SSRN]
Designing Heaven’s Will: Assigning Civil Servants by Drawing Lots in Imperial China, with Inácio Bó
This paper presents the evolution of a centralized matching mechanism that was in place for the assignment of civil servants from the late 16th century to early 20th century in China. Based on original documents and historical studies, we provide the first formal description of this procedure: candidates were assigned to jobs through a sequential lottery-based procedure, and they could not be assigned to their home regions, known as the rule of avoidance. By constructing a general model of sequential random matching, we first show that the procedure was not efficient, defined as maximality of the matching. We then show prioritizing “hard to match” candidates and jobs improves efficiency. This provides theoretical foundation for a policy change in 1824. We further provide conditions under which the sequential lottery-based procedure is efficient.
Time-constrained School Choice, with Juan Pereyra
We study a time-constrained sequential mechanism where students are asked to submit one choice and they can revise their choices in the subsequent rounds. A critical distinction between this mechanism and the standard school choice mechanism is that it provides students with information about the match along the matching process and allows them to make necessary changes. We show this information however can create double-edged effects when students only have a fixed number of times to revise. On the one hand, information helps students with lower priorities to be matched to lower ranked choices. On the other hand, these students with lower priorities are less likely to obtain their more preferred choices, because information also helps students with higher priorities to revise more effectively.
A classic trade-off that school districts face when deciding which algorithm to use to match students to schools is that it is not possible to always respect both priorities and preferences. The student-proposing Deferred Acceptance algorithm respect priorities (fairness) but can lead to inefficient allocations. The Top Trading Cycle algorithm respects preferences (efficiency) but may violate priorities. We identify a new condition on school choice markets under which both algorithms yield the same allocation. Our condition differs from earlier conditions in the literature in that it places restrictions on how preferences and priorities relate to one another, rather than placing restrictions on either side separately. Such condition is natural as priorities are often the reflection of what school districts view as legitimate preferences.
University Admission Practices – UK [link], MiP Country Profile 7, 2012
University Admission Practices – Ireland [link], MiP Country Profile 8, 2012
Matching Practices for Elementary Schools – Ireland, [link], MiP Country Profile 10, 2012
Matching Practices for Secondary Schools – Ireland, [link], MiP Country Profile 11, 2012