Combining Outcome-Based and Preference-Based Matching: A Constrained Priority Mechanism

Abstract We introduce a constrained priority mechanism that combines outcome-based matching from machine learning with preference-based allocation schemes common in market design. Using real-world data, we illustrate how our could be applied to the assignment of refugee families host country locations, and kindergarteners schools. Our allows planner first specify threshold $\bar g$ for minimum acceptable average outcome score should achieved by assignment. In context, this corresponds probability employment, whereas student it standardized test scores. The is considers both outcomes preferences assigning agents (refugee students) based on their preferences, but subject meeting planner’s specified threshold. strategy-proof efficient always generates not Pareto dominated any other respects