State-Government partisanship and Health Care Financing

Background: The policy preferences of state governors and legislators have the potential to impact health care financing. This occurs directly through their decisions, votes, and vetoes related to Medicaid generosity, state-employee health insurance, indigent programs, and other state-level programs. Also, it occurs indirectly, through the reaction of health care users, providers, and payers (e.g., out-of-pocket spending levels and cost-shifting) to the changes implemented by policy makers. This paper studies the impact of policy preferences on health care financing by estimating the effect of partisan control of the state government on key health care financing measures.

Methods: We estimate the effect of partisan control on health care financing measures using regressions with state and year fixed effects as well as time-varying covariates (income per capita, state-level gdp, demographics) on a panel of 49 states in the period 1999-2009. We measure partisan control using indicators for independent governors, democratic governors with democratic-controlled legislatures, democratic-governors with republican-controlled legislatures, republican governors with republican-controlled legislatures, and republican governors with democratic-controlled legislatures. We use a broad set of healthcare financing outcomes including state-level all-payer expenditure (aggregate and per capita), Medicare beneficiaries and expenditure (aggregate and per beneficiary), Medicaid beneficiaries and expenditure (aggregate and per beneficiary), and private expenditure (aggregate and per capita). We also do analyses at the component-payer level, where the components include durable medical products, home health, hospital care, and physician care services, among others. State Congress partisan characteristics come from US Census Bureau’s Composition of State Legislatures by Political Party Affiliation. Governor partisanship identifiers come from Klarner politics, Indiana State University. State level all-payer health care spending per capita comes from the Centers for Medicare & Medicaid Services (CMS), Health Expenditures by State of Residence database and includes all categories except dental services. We check the sensitivity of the results to different timeline assumptions.

Findings: The two independent governors in the period produced various and complex effects.  State-years in which one party controls the executive and the other controls the legislature show no statistically significant effect relative to the reference (no party controls both legislative chambers). State-years in which the democrats control both branches of government (i.e., the state executive and both legislative chambers) show an increase in Medicaid enrollment, no effect on Medicaid expenditure, and a consequent decrease in Medicaid expenditure per capita. State-years in which the republicans control both branches of government show an increase in Medicare expenditure, coming from Part D (drug coverage) and two other components. This is surprising given that Medicare is a federal program. However, documents show that, in an attempt to balance budgets, some state governments allowed citizens that were receiving state funds for medication to receive drugs, instead, under Medicare Part D. Finally, in general, the component-payer level results point to the existence of a complex policy-making process that involves increases and decreases in different cost components.

Conclusion: The partisanship of the members of the state executive and legislative branches has a significant and robust effect on key health care financing metrics.