The school choice problem (SCP) looks at assignment mechanisms matching students
in a public school district to seats in district schools. The Gale–Shapley deferred
acceptance mechanism applied to the SCP, known as the student optimal stable
matching (SOSM), is the most efficient among stable mechanisms yielding a solution
to the SCP. A more recent mechanism, the efficiency adjusted deferred acceptance
mechanism (EADAM), aims to address the well-documented tension between
efficiency and stability illustrated by SOSM. We introduce two alternative efficiency
adjustments to SOSM, both of which necessarily sacrifice stability. Our discussion
focuses on the mathematical novelty of new efficiency modifications rather
than any practical superiority of implementation or outcome. That is, our
contribution lies in process rather than outcome. Yet we argue that the
demonstration of multiple processes yielding common outcomes is, in itself, a
measure of the quality of that outcome. More specifically the consistency
of outcome from different processes strengthens the argument that Pareto
dominations of SOSM can be supported as “fair” despite the resulting priority
violations.
Keywords
mechanism design, assignment, matching, school choice