Sorting into payment schemes and medical treatment: A laboratory experiment.

In this study, we introduce the opportunity for physicians to sort into capitation or fee-for-service payment. Using a controlled medically framed laboratory experiment with a sequential within-subject design allows isolating sorting from incentive effects. We observe a strong preference for fee-for-service payment, which does not depend on subjects' prior experience with one of the two payment schemes. Further, we identify a significant sorting effect. Subjects choosing capitation deviate ex ante less from patient-optimal medical treatment than subjects who sort into fee-for-service payment. Particularly the latter become even less patient-oriented after introducing the choice option. Consequently, the opportunity to choose between fee-for-service and capitation payment worsens patient treatment, if at all. Our results hold for medical and for nonmedical students.