Measuring welfare from ambulatory surgery centers: A spatial analysis of demand for healthcare facilities

The purpose of this study is to estimate structural elements of patients' demand functions for healthcare facilities, particularly hospitals and ambulatory surgery centers (ASCs), towards the goal of answering questions about welfare gains earned from the introduction of ASCs. Spatial variation across patients and facilities, among other variables, is used for identification. Using data that contain the universe of outpatient surgeries in Florida in 2000 and 2001, I estimate by reduced-form quantile regressions of facilities' actual demand curves as functions of travel-cost between patient and facility. I show that there is a strong spatial component to demand. Developing a discrete choice model of demand for healthcare facilities, I estimate structural parameters from patients' demand functions from multinomial (conditional) logit and mixed logit (random coefficient) specifications. Time-to-travel is found to be a significant predictor in patients' choice of healthcare facility. I construct a cross-time substitution matrix to explain how patients substitute between facilities when facilities change their location. Finally, I measure how patient welfare would change if a subset of facilities (ASCs) were removed from patients' choice sets. All of this is done without explicitly including a price variable, but instead using facility fixed effects that absorb price and quality, among other unobserved product characteristics. Welfare loss from the elimination of ASCs is found to be small, about less than five minutes of welfare loss per patient for a given procedure.