Medicare Hospital Payment Adjustments and Nursing Labor Markets

Numerous studies show that changes in Medicare provider payment methods and levels have important effects on the supply of healthcare, on the coding of patients’ diagnoses, and on physician training and labor supply. In contrast, very few studies examine the effects of Medicare hospital payment on the nurse labor market. We provide new analysis of the link between Medicare hospital payment and the nurse labor market by examining the hospital wage index (HWI) adjustment to Medicare payments, a geographic adjustment intended to compensate hospitals in high-cost labor markets.

We examine two potential consequences of the HWI adjustment. First, as noted by the Institute of Medicine, the process of making the HWI adjustment may unintentionally create “circularity” or endogeneity in the wage index (IOM, 2012). That is, because Medicare calculates the HWI adjustment using wage data reported by hospitals themselves, hospitals in highly-concentrated markets can influence their own wage index and thus drive up their Medicare payments. We examine whether this HWI adjustment process leads to higher nurse wages in markets where hospitals have an increased ability (“opportunity”) and a strong incentive (“motive”) to game the HWI calculation. Second, we examine an adjustment made to the HWI called the Occupational Mix Adjustment (OMA). Concerned that initial HWI calculations rewarded hospitals that hired more expensive higher-skilled labor, CMS implemented the OMA in 2005 to remove skill-mix variation across hospitals from the wage index adjustment. We examine whether employment of higher-skilled nurses decreased after the implementation of the OMA.

We test for both consequences using data from various sources. We construct area-level nurse wage differentials from the 2000 Census and the pooled 2009-2011 annual American Community Surveys, and we measure hospital-level RN employment using the American Hospital Association annual survey data. We use CMS’ hospital impact files to construct measures of hospital market concentration.  To examine the relationship between the HWI adjustment process and nurse wages, we regress nurse wage differentials on hospital market concentration. In this way, our work is similar to prior tests of classic monopsony (e.g., Hirsch and Schumacher, 2005). However, unlike those prior tests, we extend the basic model to allow the effect of market concentration to vary by the relevance of Medicare payments to hospitals’ revenue. Further, we exploit an exogenous change in market concentration arising from the switch from using MSAs to using CBSAs in defining hospital markets. To examine whether the OMA reduced hospitals’ incentive to hire more-skilled labor, we estimate a model of RN employment and test whether hospitals with the ability to influence the OMA and a high share of Medicare days reduced RN employment after the OMA.

We find no support for the circularity critique of the HWI; that is, we find no evidence that existing HWI adjustment methods induce Medicare-intensive hospitals to use their influence on the index to increase wages. However, we do find evidence that hospitals responded to the implementation of the OMA by hiring fewer high-skilled nurses (RNs), which implies that hospitals were gaming Medicare rules prior to 2005.