Objective: Our study examines changes in physicians' work hours in response to a coverage expansion. Methods: We use as a natural experiment the Children's Health Insurance Program (CHIP), which was established in 1997 and significantly expanded children's eligibility for public health insurance coverage. The magnitude of the CHIP expansion varied across states and over time, allowing its effects to be identified using a state-year fixed effects model. We focus on pediatricians, and we measure their selfreported work hours using multiple waves (pre-and post-CHIP) of the physician survey component of the Community Tracking Study. To address endogeneity concerns, we instrument for CHIP enrollment using key program features (income eligibility cutoffs and waiting times).
Results:We find a large negative relationship between the magnitude of a state's CHIP expansion and trends in pediatricians' work hours. This relationship could be due to key supply-side features of CHIP, including relatively low provider reimbursements and heavy use of managed care tools. He, F., White, C.
Keywords:
E2
IntroductionThe Congressional Budget Office (CBO) has estimated that providing insurance coverage for all of those who are currently uninsured would increase the demand for health care services by 2 to 5 percent (Congressional Budget Office, 2008). But, for the quantity of services to actually increase, providers have to supply the additional care, and it is not clear whether, or how, they will do so. Based on a purely demand-driven model of physician labor supply, the CBO and other researchers have hypothesized that a large coverage expansion could lead, in the short run, to physicians working longer hours to accommodate such an influx of patient demand. On the other hand, a supply-driven model hypothesizes that physicians could choose to work fewer hours with a large coverage expansion, due to constraints on the supply side. To test between the two hypotheses, we analyze the impact of the Children's Health Insurance Program (CHIP), a public health insurance program for children in low-and middle-income families.Established in 1997, CHIP has several features that make it well suited to our purposes. First, the CHIP expansion was large, resulting in almost one in five children becoming newly eligible for public coverage-that expansion in eligibility is roughly similar in magnitude to the expansion for adults that will occur under the Affordable Care Act (ACA). Second, states play a central role in designing and financing their CHIP programs, and have some latitude to determine eligibility criteria. As a result, eligibility criteria and other key program features vary widely from state to state, which we exploit in our analysis. Third, CHIP was established relatively recently. This allows access to fairly rich and recent data sources, and also makes it more plausible that any findings would be relevant to upcoming reforms.We use a state-year level panel data analysis, with observations spanning the period before and after the establi...