Physician Sensitivity to Short-lived Incentives: Evidence from a Medicare Fee Bump

A long literature has diagrammed providers’ varied responses to movements in public payer reimbursements. However, existing studies primarily concern a fee change that is either permanent or absent a specified expiration date. A notable feature of several contemporary payment reforms is a well-defined policy timeline. Within such a context, both the arrival of the reimbursement change and its sunset are public knowledge – and presumably taken into account by relevant providers. Given that many physician services are not one-off transactions but instead a series of patient-provider interactions, a temporary incentive presents an interesting circumstance – one with which we have very little empirical evidence to draw upon. An example of a transient payment initiative has in fact already come and gone. For a period of just two years (2013 – 2014), the Affordable Care Act raised Medicaid fees for primary care providers to be on par with Medicare rates, with the hope of bolstering the supply of Medicaid services. Notwithstanding pervasive doubt among experts and advocacy groups, the policy’s lone empirical study actually shows significant behavior change. Within this paper, we leverage a previously unstudied change to Medicare’s reimbursements for physician services embedded within the Medicare Modernization Act. The positive fee change was both targeted and temporary, with the policy’s start and end dates legislated simultaneously – giving providers complete information when calculating their optimal response. We then assess if the provider behavior captured in the recent Medicaid study is evident in a setting occurring a decade earlier. Using difference-in-differences, we document increased physician willingness to supply more services to the Medicare market following the new incentive. The strongest policy effect is within the domain of new patient visits, with nearly double the proportion going to Medicare beneficiaries after the fee bump was implemented. However, the observed responses are heterogeneous and largely concentrated among primary care physicians. The data hint at several provider strategies that may be occurring in combination to accommodate more Medicare business – including reducing pediatric services. At the same time, we fail to detect clear welfare improvements for Medicare enrollees, and our results suggest a negative income effect on physician labor supply attributable to more generous fees. For these reasons, our work benefits from a wider gaze of potential policy implications, and our findings imply an income transfer from taxpayers to providers with minimal social benefit.