Monique de Haan and Pieter Gautier and Hessel Oosterbeek and Bas van der Klaauw
On the basis of theory alone, it is hard to advise school districts which school assignment mechanism to adopt. While the Deferred Acceptance (DA) mechanism has some desirable properties (strategy proofness and stability) that the Boston mechanism lacks, Boston may outperform DA in terms of welfare because it allows students to express the intensity of their preferences. We use a unique combination of administrative data and survey data to compare the two mechanisms empirically. We find that DA results in higher average welfare than Boston. This finding contradicts the baseline results from some recent studies that use structural approaches. This contradiction can be attributed to the structural models' assumption that students do not make strategic mistakes. When we impose this assumption, the welfare ranking of the mechanisms reverses.