How Was Medical Care Utilization Affected by the Medicaid Primary Care Fee Bump?

Historically, Medicaid has offered lower reimbursement rates to providers for the same services relative to other payers. This disparity in reimbursement discourages providers from participating in Medicaid, and many primary care physicians either do not accept Medicaid or are not currently accepting new Medicaid patients. The result is limited access to physician services by Medicaid recipients, which adversely affects participants' health and results in higher rates of ER use, thereby increasing Medicaid costs.

We investigate the effect of the largest ever increase in Medicaid primary care reimbursement rates on the use of medical services. In particular, the Medicaid fee bump, a provision of the Affordable Care Act, mandated that states raise Medicaid payments to match Medicare rates for primary care services in 2013 and 2014. The fee bump was not re-authorized by the federal government and ended on December 31st, 2014. Thus, the fee bump policy resulted in substantial variation in Medicaid reimbursement rates; a large increase in 2013 followed by a steep decline in 2015.

We employ two complementary identification strategies to address the potential endogeneity of states' provider payment policies. Our first analysis includes a regression discontinuity design that uses the Medicaid fee bump to identify the break in the utilization trend of medical services. Our second strategy is a fixed-effects model that exploits within-state variation in Medicaid payments to primary care physicians over the 2008-2015 time period for identification. In the latter, we incorporate hand collected state-level data on fee schedules for five primary care services in order to trace out variation in state-level Medicaid payment policies before and during the Medicaid fee bump.

Our results suggest that more generous Medicaid payments to primary care providers increase the number of office-based primary care visits. This increase is much larger for mid-level providers than physicians, indicating that the use of nurse practitioners and physician assistants is important to how practice groups respond to changes in payment levels. We also find that higher Medicaid fees are associated with improvements in access to timely care and an increase in the utilization of prescription drugs, suggesting that at least some of the additional primary care services by Medicaid enrollees were necessary. The improvement in access to care is more pronounced in counties that are underserved by primary care physicians.

We investigate whether the ACA Medicaid expansion affected our results by estimating our models using the time period before 2014, and find similar results. In addition, estimated effects in the sample of those who were eligible for Medicaid both before and after the ACA Medicaid expansion (proxied by low-income children and adults with children) are broadly comparable to our estimates from the full Medicaid sample. Thus, we are confident that our results are driven by the primary care rate increase, and not from the ACA Medicaid expansions.