Grant: Travel Awards - Institute Community Support

Since its introduction in 1978, quantile regression has been widely applied by researchers to examine patterns observed at the extremes an outcome variable. Efficiency measurement is an area where quantile regression could provide valuable insights into the attributes of top-performing care providers, and factors that enable them to be more efficient. This paper presents an application of panel quantile estimation - specifically, one that takes a correlated random effects approach - to assess the efficiency of nursing homes in Ontario, Canada. Data used for this study was taken from Statistics Canada's Residential Care Facilities Survey. Specifically, our analysis was conducted on a longitudinal dataset consisting of 615 nursing homes, observed over 15 fiscal years (1996-2011). Results from the regression models found chain affiliation, urban location, and operation size to be consistent and positive predictors of efficiency. Contrary to what was found in prior research studies, for-profit ownership did not have a significant effect on the efficiency of top-performing nursing homes in Ontario. The analysis presented in this paper not only provides a novel approach to efficiency measurement, but also demonstrates a non-uniform effect of technical inefficiency across the quantiles. Specifically, high- and low-performing nursing homes uses different mixes of nurses and care aides. The application of panel quantile regression in efficiency analysis overcomes many shortcomings of other existing methods. This study demonstrates that correlated random effects quantile regression produced reliable estimates of efficiency.