Leeat Yariv, Marina Aganov, and Ahrash Dianat, California Institute of Technology; Larry Samuelson, Yale.
Exchange in many markets involves two-sided matching processes: hospitals/interns, workers/employers, organ donors/patients, students/schools, Uber drivers/passengers. Applications use algorithms for efficient matching, based on each side’s rank order preference for the other. Theory assumes each has complete information on others’ preferences. But recent theory enables the prediction of final outcomes in match/rematch processes that are expected to emerge as preferences are revealed by decisions in repeat interactions. How would you test that prediction? In the field, we don’t know individual preferences for predicting final allocations. But in the lab, one can use payoffs to induce preferences, and vary payoff structures and initial allocations. In the baseline experiments, subjects know all others’ value types. In the incomplete information setting, all know firms’ value types, but workers’ value types are strictly private. With target prediction well defined, information and the sophistication requirement of firms’ decisions about workers are varied across treatments. By inaugurating the study of matching markets under incomplete information, this research holds promise for improving and expanding the many applications.