We investigate the change in the effect of market structure on planned hospital quality for three high-volume treatments, using a quasi difference in differences approach based on the relaxation of patient constraints on hospital choice in England. We employ control functions to allow for time-varying endogeneity from unobserved patient characteristics. We find that the choice reforms reduced quality for hip and knee replacement but not for coronary bypass. This is likely due to hospitals making a larger loss on hip and knee replacements, since robustness checks rule out changes in length of stay, new competitors’ entry and hospital-level mortality as possible confounders.
|CHE Research paper
|Centre for Health Economics, University of York
- difference in difference
- control function