Risk selection and cost shifting in a prospective physician payment system: Evidence from Ontario

We study the risk-selection and cost-shifting behavior of physicians in a unique capitation payment model in Ontario, using the incentive to enroll and care for complex and vulnerable patients as a case study. This incentive, which is incremental to the regular capitation payment, ceases after the first year of patient enrollment and may therefore impact on the physician's decision to continue to enroll the patient. Furthermore, because the enrolled patients in Ontario can seek care from any provider, the enrolling physician may shift some treatment costs to other providers. Using longitudinal administrative data and a control group of physicians in the fee-for-service model who were eligible for the same incentive, we find no evidence of either patient 'dumping' or cost shifting. These results highlight the need to re-examine the conventional wisdom about risk selection for physician payment models that significantly deviate from the stylized capitation model.