Stability and Deferred Acceptance: Strategic Behavior in Two-Sided Matching

When matching mechanisms yield assignments that are not stable, offers unravel, and, eventually, participants abandon the match. This logic is thought to provide a reasonable explanation of why, empirically, the stable Deferred Acceptance (DA) mechanism persists where non-stable alternatives, such as Priority mechanisms, are abandoned. Theory, however, tells us that both Deferred Acceptance (DA) and Priority mechanisms can yield unstable matches in incomplete information equilibrium. If, however, under DA, match participants deviate from equilibrium by truthfully revealing, then the empirical difference between DA and Priority can be explained. In an experiment, we find evidence that, while participants are able to learn to deviate from truth-telling under a Priority mechanism, they fail to do so under DA. This implies that out-ofequilibrium truth-telling might help to explain why DA persists in spite of being unstable in equilibrium. JEL: C78, C92, D02