A comparison of the robust conditional order-m estimation and two stage DEA in measuring healthcare efficiency among California counties

This paper examines cross-county healthcare efficiency rankings using modern non-parametric estimators, while taking into account secondary environmental variables. Results indicate that output-direction efficiency estimates yield counties producing inefficiently for both order-alpha and order-m estimators. After accounting for a variety of secondary environmental variables, unconditional efficiency estimates improve by anywhere between 7.5 and 10-percentage points. Results show that there is little correlation between the highly visible Robert-Woods-Johnson Foundation estimates with those derived here. We also find that counties are more efficient when they possess lower rates of obesity, unemployment, and preventable hospital readmissions. In addition, demographic variables do not play much of a role in explaining cross-county inefficiency. The analysis shows that the two stage DEA is inappropriate and violates several assumptions in comparison to the conditional order-m estimation.