An experiment on referrals in health care

How referral fees affect primary care physicians’ behavior is not well understood. We conduct a behavioral experiment on referral fees in which subjects take on the role of physicians. An illustrative theoretical model guides our experimental design and the formulation of behavioral hypotheses for heterogeneously altruistic physician-types. We exogenously vary the level of referral fees paid by specialists to primary care physicians and study the impact of these fees on primary care physicians’ diagnostic effort and referrals. In a separate experimental task, we evaluate primary care physicians’ altruism. In line with our model, introducing medium-sized referral fees significantly increases referrals made by barely altruistic primary care physicians of patients who need specialist treatment, when compared to the baseline without referral fees. High referral fees significantly increase referrals irrespective of physicians’ level of altruism. Surprisingly, diagnostic effort is not significantly affected by referral fees. Implications for economic efficiency are discussed.