Predictors of Increased Privitization and Competition in Medicaid

The Affordable Care Act (ACA) proposes expansions to Medicaid eligibility and raises Medicaid fee-for-service reimbursement schedules. Expanding eligibility, while raising the cost of traditional fee-for-service Medicaid, is likely to increase the use of privatized managed care. Between 1993 and 2013, the share of Medicaid enrollees in managed care increased from 14% to over 60% in 2012. The welfare implications of this trend towards privatized managed care are not well understood in the literature. The first challenge to studying these effects is the absence of claims data for enrollees under privatized managed care. The second challenge is the endogeneity between observed outcomes and the decision to transition enrollees from fee-for-service to managed care. This paper uses state-level fiscal shocks as an instrument for changes to states’ demand for managed care. I contribute a theoretical model of the state's decision to constrain utilization at the intensive and extensive margins in Medicaid programs. I create a new dataset from several sources. National market composition and enrollment data at the county (2003-2009) and state level (1993-2009) are provided by CMS. In addition, the project leverages detailed fiscal data from the National Association of State Budget Officer’s Reports, which detail both realized and anticipated outlays, as well as data from the Internal Revenue Service on tax revenue. Data on mandates related to Medicaid managed care were systematically collected from state Medicaid offices. I present a series of descriptives that characterize each state's evolving use of managed care in Medicaid over time in the U.S.. Consistent with previous work, I find substantial variation across states and time in Medicaid managed care. I posit that this variation is the result of negative fiscal shocks and enrollee composition. To test this hypothesis, I exploit variation in several measures of fiscal planning: (1) outlays of state general funds, (2) changes in the balances of state emergency funds, and (3) tax revenue for identification. Using detailed county and state level data, I present evidence suggesting that low-cost Medicaid enrollees (i.e., non-disabled adults and children) are shifted to managed care following a negative budget shock. In counties where managed care enrollment for these groups is high prior to the shock, my results suggest a higher likelihood of managed care enrollment among high-cost populations, such as the disabled and aged. These trends are particularly prominent in states with a conservative majority in the legislature at the time of the fiscal shock.