Abstract
Chile is experiencing a period of rapid aging, which increases the need of long-term care services in the country. Nursing homes have been the traditional alternative to deal with the increase of elderly population in the country, with services supplied by a mix of for-profit and nonprofit providers. Additionally, population exhibits a high degree of geographical concentration. The study aims to identify the determinants of the geographical location of nursing homes in Chile at municipality level. The analysis takes into account the different location criteria for different types of nursing homes as well as potential spatial effects. The paper uses spatial analysis tools to identify clusters of nursing homes and population characteristics and to estimate the determinants of nursing homes availability and coverage in the country. The analysis-based on spatial global and local tests, and spatial autoregressive models- show the existence of clusters of nursing homes as well as clusters of municipalities according to elderly population, income, poverty, population density, and public health insurance coverage. Residuals from ordinary least squares regressions were spatially autocorrelated, showing the need of using spatial models. Estimations show that availability and coverage of nursing homes are positively related with municipality income, and that for-profit and nonprofit facilities respond differently to different factors. A negative coefficient was found for poverty in nonprofit nursing homes, raising doubts about the effectiveness of giving public subsidies to incentive the installation of facilities in areas with high needs and low income.