Geographic Market Definition: The Case of Medicare-Reimbursed Skilled Nursing Facility Care
Bowblis, John R., North, Phillip, Inquiry
Correct geographic market definition is important to study the impact of competition. In the nursing home industry, most studies use geopolitical boundaries to define markets. This paper uses the Minimum Data Set to generate an alternative market definition based on patient flows for Medicare skilled nursing facilities. These distances are regressed against a range of nursing home and area characteristics to determine what influences market size. We compared Herfindahl-Hirschman Indices based on county and resident-flow measures of geographic market definition. Evidence from this comparison suggests that using the county for the market definition is not appropriate across all states.
Researchers and policymakers are interested in the role that market structure plays in the health care market because more competitive markets could result in better access to care at higher quality and lower costs. Market structure is important to antitrust authorities when they need to estimate the competitive effects of a potential merger; while in comparison, local health agencies may need to study any access and cost issues that might arise if a facility enters or exits the market. From the perspective of researchers, to fully understand the impact of changes in government policy, knowledge of how firms respond to competition is needed. In all of these cases, the concern is how the number of firms and the distribution of those firms' market share impact an outcome. (1) To measure market structure, it is pivotal to have an accurate definition of the relevant geographic market.
In principle, there is a general consensus on how to define geographic market, but there is wide variety in how to implement market definition empirically (Baker 2007). Multiple papers have addressed geographic market definition in general, but most literature that focuses on market definition in the health care sector is limited to the hospital industry (Lindrooth 2008). Many of these techniques require the use of patient-level data, something that was not often available to nursing home researchers until recently. Furthermore, the study of market definition in the nursing home industry is underdeveloped, with the majority of past research using geopolitical boundaries to define the relevant market.
The objective of this paper is to provide a comprehensive evaluation of the different empirical techniques to define relevant geographic markets for nursing home care, and to inform researchers and policymakers which method would be more appropriate given their research objectives and the constraints of their data. The particular focus is on nursing home residents who are reimbursed by Medicare for skilled nursing facility (SNF) care. Although nursing homes provide both long-term and skilled nursing care, we focus on skilled nursing for multiple reasons. First, the medical situation needed for a patient's admission and the processes used to admit an individual are different for skilled nursing compared to long-term care. Second, the reimbursement mechanisms for both types of care are different. Medicare and private health insurance are the primary payers of skilled nursing, whereas Medicaid and private means are the primary payers for long-term care. Finally, Medicare SNF care prices are set by the federal government and the resident does not have any out-of-pocket costs for the first 20 days. This reduces the influence of price on the geographic market.
We provide a general overview of the empirical techniques used to define the relevant geographic market in the health care sector, present a conceptual model for determining the size of the market, and empirically test the conceptual model. Using merged data from the Minimum Data Set (MDS), the Online Survey Certification and Reporting (OSCAR) system, and the Area Resource File for the states of California, Florida, Georgia, Ohio, Illinois, New York, New Jersey, and Texas, we empirically compare the use of geopolitical boundaries to actual patient flows. …