Presentation, Midwest Health Economics Conference, DePaul University, Chicago, Illinois
In this presentation, we discuss our upcoming contribution to the literature measuring and understanding drivers of racial and ethnic sorting (segregation) in the health care setting by constructing dissimilarity indices (DI) for nursing homes. The dissimilarity index, a commonly used measure in residential segregation studies, ranges from 0 to 1 and in our context describes the share of patients in a health care market who would have to switch homes in order for there to be evenness (no segregation) across the providers in that market. Prior work has computed the DI in healthcare settings but provides limited evidence on how the DI varies across geographic areas and/or is based on data more than a decade old. We combine administrative data on all certified nursing homes in the U.S. with health assessments of long-term stay patients residing in those homes for the years 2011 and 2017. Geographically, we define a nursing home market to be a county.