Strategy-proof improvements upon deferred acceptance: A maximal domain for possibility

Onur Kesten, Morimitsu Kurino

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)

Abstract

In one-sided matching/assignment problems, an important debate centers around whether it is possible to improve upon the Gale-Shapley student-proposing deferred acceptance mechanism (DA) via alternative strategy-proof mechanisms. In unrestricted preference domains, no strategy-proof mechanism Pareto improves upon DA whether students have outside options or not. When standard exogenous outside options, e.g., private school, do not necessarily exist, we show that endogenous outside options, i.e., problem-specific schools that could always accept a deviating student, arise and the impossibility of obtaining a strategy-proof improvements over DA prevails. It is, however, possible to construct natural subdomains allowing for positive results, where some students' preferences are in part induced by an exogenous hierarchy of quality tiers. We then identify maximal domains on which it is possible to improve upon DA without sacrificing strategy-proofness. This result may help better assess the underpinnings of the three-way tension among efficiency, individual rationality/stability, and strategy-proofness in matching problems.

Original languageEnglish
Pages (from-to)120-143
Number of pages24
JournalGames and Economic Behavior
Volume117
DOIs
Publication statusPublished - 2019 Sep

Keywords

  • Outside options
  • Pareto dominance
  • Strategy-proofness
  • Student-proposing deferred acceptance mechanism

ASJC Scopus subject areas

  • Finance
  • Economics and Econometrics

Fingerprint Dive into the research topics of 'Strategy-proof improvements upon deferred acceptance: A maximal domain for possibility'. Together they form a unique fingerprint.

Cite this