Experience and Advice in a Deferred Acceptance Matching Experiment

Abstract

Recent research suggests that truth-telling in strategy-proof mechanisms may be an artefact of the experimental design, and is vulnerable to misleading advice yet impervious to accurate advice. In this paper, we examine the effect of experience by providing participants with opportunities for practice, in which they receive feedback on the effects of submitting different preference rankings upon their match under the Gale-Shapley deferred acceptance mechanism. We also study the interaction between experience and mixed advice. We find no effect of practice upon the truthfulness of reported preferences, either alone or in combination with advice. However, presenting participants with both misleading and accurate advice significantly reduces truth-telling. The latter effect has previously been documented for the top trading cycles mechanism, but not for deferred acceptance.

Keywords: Market Design, Matching Markets, Economic Experiments

JEL Classification: C91, D47, D90