Essays on the measurement and causes of inefficiency in the public sector with application to education

This study consists of five essays examining the measurement and causes of inefficiency in the public sector. Public sector service provision differs from private production theory due to the impact that fixed environmental factors have on production. These essays extend the nonparametric programming technique for frontier estimation to the public sector and provide analyses of the provision of educational services in New York State school districts. The first essay analyzes the existing nonparametric model that purportedly controls for exogenous inputs; it is shown that this model is flawed, leading to biased efficiency estimates. A new model is developed to properly control for environmental factors. The results of a simulation analysis provide evidence that the new model does a better job controlling for exogenous factors. The second essay focuses on the conceptual notion of inefficiency; it is shown that the existing radial index of efficiency does not really measure inefficiency since additional slack may exist in some, but not all inputs, after radial efficiency is achieved. A two-stage procedure for efficiency evaluation is suggested as a solution to the problem. The third essay extends the newly developed model to allow measurement of cost inefficiency, which results if a local government is technically and/or allocatively inefficient. In addition to the cost efficiency index, a new measure is developed that indexes environmental harshness. Essay four extends my efficiency model to allow measurement of returns to scale in the provision of public services. Due to the presence of exogenous factors of production, returns can be conceptualized not only along but between production frontiers. This essay provides the means to measure scale economies in the public sector. Application to New York State reveals large scale economies; this suggests that school districts which operate along an increasing returns to scale portion of the production frontier would be able to realize larger increases in outcomes for a given amount of state aid ceteris paribus. Finally, the last essay empirically tests bureaucratic models of supply by linking existing theoretical models with the measurement of cost efficiency. This essay provides useful extensions to both the measurement and theory literature by analyzing the causes of inefficiency