Cross Subsidization and the Welfare Impact of Specialty Entry

The growth of the specialty hospital industry has been one of the most important changes to the hospital industry in recent years. Specialty hospitals focus on three specific types of care: cardiac surgery, orthopedic surgery and general surgery. Proponents argue that specialty hospitals are focused factories providing higher quality, lower cost care. General hospitals, on the other hand, suggest that these firms enter the most profitable markets and skim the healthiest (lowest cost) patients leaving them without the profit necessary to pay for indigent and uncompensated care. Given the uncertainty on the issue the Federal Government instituted a moratorium on the construction of new specialty hospitals in 2003. The moratorium was subsequently repealed in August of 2006. We add to recent empirical work on the welfare effects of specialty entry by studying the impact of entry on the ability of competing general hospitals to cross-subsidize unprofitable care. The empirical study of the welfare effect of specialty entry to date has indicated a net welfare gain, with market level spending declining and no change in outcomes (Barro et al. 2006). This approach, however, does not account for the potential welfare impact of transferring profits between firms. Standard economic theory treats profits as welfare enhancing regardless of the firm at which they are obtained. This is unlikely to be true in a setting such as health care where diversified, not-for-profit hospitals can use profits from specific services to cross-subsidize less profitable types of care. We empirically investigate the degree to which entry by specialty hospitals led to reductions in the provision of unprofitable care by general hospitals. Estimating the effect of competition on cross-subsidies is not a straightforward exercise because firm entry is endogenously determined. To account for this we take two approaches. Relying on variation in Certificate-of-Need (CON) laws across states, we model the response of hospitals in states with and without entry restrictions to the 2003 moratorium on specialty entry. We use a difference-in-differences approach in which we assume that the moratorium led to a larger relative reduction in specialty hospital entry in markets without CON than in those with CON. Because CON laws are a strong predictor of specialty hospital entry (GAO 2003) and exogenously determined we can directly estimate the effect that specialty competition has on provision of unprofitable services. Taking this approach a step further, we estimate a triple differences model in which hospitals that offer services mostly likely to be affected by specialty competition (i.e. cardiac surgery or catheterization) are compared across states with differing entry restrictions relative to hospitals not offering these services. To exploit the more detailed market level factors that may affect entry we also estimate a market level entry model. Fitting this model for all markets yields a predicted probability of entry that is used to instrument for the endogenously determined actual entry by specialty hospitals