School Choice with Multiple Priorities

TL;DR

This paper explores a school choice model with multiple, potentially conflicting priority orders, introducing a new fairness concept called M-fairness and a mechanism that ensures improved stability and responsiveness.

Contribution

It introduces M-fairness for markets with conflicting priorities and develops a mechanism based on the efficiency adjusted deferred acceptance algorithm that guarantees stronger stability properties.

Findings

The mechanism satisfies responsiveness to improvements.

It achieves improved-group optimally M-stability.

Applicable to markets with conflicting priority orders.

Abstract

This study considers a model where schools may have multiple priority orders on students, which may be inconsistent with each other. For example, in school choice systems, since the sibling priority and the walk zone priority coexist, the priority orders based on them would be conflicting. We introduce a weaker fairness notion called M-fairness to examine such markets. Further, we focus on a more specific situation where all schools have only two priority orders, and for a certain group of students, a priority order of each school is an improvement of the other priority order of the school. An illustrative example is the school choice matching market with a priority-based affirmative action policy. We introduce a mechanism that utilizes the efficiency adjusted deferred acceptance algorithm and show that the mechanism satisfies properties called responsiveness to improvements and improved-group optimally M-stability, which is stronger than student optimally M-stability.