Estimating the Revenue Efficiency of Public Service Providers in the Presence of Demand Constraints

Evaluating the performance of public service providers is often complicated by the fact that they must choose input levels before demands for their services are known. We consider an even more complicated situation in which service providers have no opportunity to directly influence demands. This means that their predetermined inputs may be more than what is required to meet realised demands. In such cases, conventional measures of revenue efficiency used in the operational research literature will generally mis-classify rational and efficient managers as inefficient. We develop a more appropriate measure of revenue efficiency that accounts for exogenously-determined demands. We explain how data envelopment analysis (DEA) methods can be used to estimate our measure, and also how they can be used to assess the consequences (if any) of providers having to choose input levels before demands are known. The methodology is applied to hospital and health service providers in Queensland (Australia). We obtain estimates of revenue efficiency that are quite different from estimates obtained using a conventional approach. Our results also indicate that hospital and health service providers were not disadvantaged by having to choose input levels before demands were known