Do Nonprofit Hospitals Provide Less Charity Care When Faced with Financial Losses?

This paper estimates the effect of the sharp reduction in nonprofit hospital incomes following the 2008 financial crisis on the supply of charity care. Nonprofit hospitals have a substantial share of the health care market in the United States, accounting for almost 70% of all inpatient cases in acute care hospitals (Frank and Salkever 1994). With tax subsidy policies favorable to these non-for-profit organizations, they play an important role to supply public services (often in charity medical care) under contract to government. Though existing literature has identified financial income as an important contributing factor that could lead to the substantial differences in the provision of such public goods by nonprofit hospitals, the evidence of income effect in regards to whether hospitals reduce their supply of charity care in response to the loss of income, however, is mixed, with some studies pointing in the opposite direction. In this study, we construct a new data set consisting of audited financial accounting statements of all hospitals in the United States in 2007 and 2008 and use the recent financial crisis as natural experiment to examine whether and to what extent an exogenous shortfall in hospital income affects the supply of charity care. The great recession of 2008 not only put a major dent in hospital income but also led to a slowdown in growth of funding to improve health in many other countries, which was not anticipated by these health services providers. Perhaps the most striking result from this research is a non-result: the lack of a causal link between hospital financial condition and the supply of the mission good. Using a quasi-experimental design with both matched and unmatched control groups, we find that the decline in a hospital’s financial income did not cause it to reduce its spending on uncompensated care. In fact, we show that the income effect on the supply of charity care are indeed small in both difference-in-differences analysis and cross-sectional regressions after controlling for hospital size, type, payer mix, asset liquidity, financial risk, operating efficiency, and labor costs. Given the fact that the market structure of the health services industry did not change, at least over the period from 2007 to 2008, this evidence suggests that holding constant the crowding-out effect there is no income effect on the supply of charity care. Additionally, from the results of instrumental variable (IV) regressions, hospitals with more Medicaid reimbursement, less Medicare revenue, higher financial risk, lower operating efficiency, higher labor costs, and being located in rural areas are associated with more uncompensated care expenses. Finally, we note that the evidence of the relationship between charity care and hospital size and asset liquidity is mixed.