Volume Responses to Hospital Revenue Targets: Evidence from Maryland's Global Budget Revenue

This paper looks at Maryland's 2014 implementation of an annual total hospital revenue cap--a payment change that builds on four decades of all-payer rate regulation. Under the Global Budget Revenue (GBR) system, the rate-regulating authority prospectively determines each hospital's budget based on baseline utilization to slow total hospital spending growth in the state. Prior to the payment system change, hospitals, limited by per admission revenue caps, expanded the volume of admissions. Maryland hopes that hospitals will meet their budgets by improving population health and investing in care coordination with community providers to reduce hospital volume. While this strategy may be feasible in the long-term, it is unlikely to be the mechanism by which hospitals meet their targeted revenues in the first few years of the payment system change. This paper will examine the immediate inpatient volume response overall, by service-line, and by patient composition.

I leverage variation in exposure to the policy by hospital in a quasi-experimental design. First, Maryland hospitals with baseline revenues deviating below historic mean revenues will feel the payment change more acutely than hospitals with deviations above their historic means. Second, due to technology and drug advances, demand for certain service lines is growing faster than demand for service lines experiencing less innovation; therefore, hospitals with a higher share of fast-growing demand (measured by year over year revenue growth) will also feel the target more sharply.

In a second identification strategy, I leverage the discontinuity of the GBR system at the D.C. - Maryland border. Maryland residents in close proximity to D.C. may more easily access inpatient care outside of the revenue constrained Maryland hospitals. Although the state of Maryland is responsible for constraining spending growth for the entire Maryland Medicare population, hospitals themselves have little incentive to prevent (and may in some cases encourage) Maryland residents to seek care in neighboring states. For these residents, the impact of the payment change may not be as strong.

This paper will add to the existing literature by examining implications outside of the Medicare population, focusing on changes to the denominator of inpatient admissions, and examining heterogeneity in treatment effects. No study to date has looked at heterogeneity in treatment effects by hospital-physician integration under the GBR system, but several studies suggest that aligning financial incentives for physicians with hospital incentives will be important for the success of the model.