Explanatory Power of Different Data Envelopment Analysis Models for Determining Airports' Cost-Efficiency

Abstract

For analyzing the cost-efficiency of airports and the determinants of that efficiency, several data envelopment analysis (DEA) methods are available. The variety across the available approaches is a result primarily of different assumptions in the model specification. Although it is well established that the outcomes of these models vary depending on what the analyst assumes with respect to input and output orientation and economies of scale, there is still a lively debate on the impact of bootstrapping and the choice of input variables on the explanatory power of such models. This paper analyzes the cost-efficiency of Italian and Norwegian airports over time. A series of two-stage DEA approaches is applied, with truncated regression models in the second stage to evaluate the explanatory value of the chosen methods and also to establish robust results on the impact factors of airports' cost-efficiency. With regard to the latter it is found that the impact of airport size is limited to technical efficiency and that the catchment area is the most significant impact factor on cost-efficiency.