Abstract

Why do people report their preferences non-truthfully under strategy-proof mechanisms? Realizing that truth-telling is a dominant strategy requires an ability to think contingently. A recent literature finds that people often suffer from a failure in contingent thinking (FCT) in simple choice tasks (Niederle and Vespa, 2023). We examine whether FCT contributes to the non-truth-telling in a student-school strategy-proof matching mechanism. We obviate the need for subjects to think contingently by asking them to choose one school in each possible contingency. Compared to the usual (contingent-thinking-requiring) method of submitting one ranking for all contingencies, we find no treatment effect on truth-telling. Only when we remove information on applicant priorities in various schools and replace it with probabilities of the various contingencies, which is similar to how contingencies are presented in the FCT literature, do we find a positive treatment effect on truth-telling. We conclude that while FCT is present in subjects, it is not a cause of non-truth-telling in the canonical student-school matching context.

Keywords:

strategy-proofness, contingent thinking, school choice, market design

JEL Classification:

C72, C91, D47, D91