Iterative versus Standard Deferred Acceptance: Experimental Evidence

Abstract

We run laboratory experiments where subjects are matched to colleges, and colleges are not strategic agents. We test the Gale-Shapley Deferred Acceptance (DA) mechanism versus the Iterative Deferred Acceptance Mechanism (IDAM), a matching mechanism in which students make applications one at a time. We consider two variations of IDAM: one in which students are only informed of whether they have been tentatively accepted or not (IDAM-NC) and one in which, at each step, they are also informed of the tentative cutoff values for acceptance at each college (IDAM). A significantly higher proportion of stable outcomes is reached both under IDAM and IDAM-NC than under DA. The difference can be explained by a higher proportion of subjects following an equilibrium truthful strategy under iterative mechanisms than the truthful reporting under DA. We associate the benefits of iterative mechanisms relative to DA with the feedback of the outcome of applications provided between steps of the iterative mechanisms. This feedback allows subjects to learn that deviating strategies from truthful do not work as intended.

Keywords: Market Design, Matching, Iterative Mechanisms, College Admissions, Experiments

JEL Classification: C78, C92, D63, D78, D82