An empirical evaluation of the physician offsetting response

Using data on office visits for Medicare-treating physicians, this study examines the effects of several policy changes that reduced real reimbursement for some, but not all physicians, from 2012 to 2017. Using genetic propensity score weighted Difference-in-Differences (DID) models that include individual physician and county fixed-effects, this study finds strong evidence that those who experienced a reimbursement reduction increased their service provision, indicative of a supplier-induced demand response. Our estimates imply that 27–33% of a reimbursement reduction will be offset through changes in provider behavior.