SUBSAMPLING APPROACH FOR STATISTICAL INFERENCE WITHIN STOCHASTIC DEA MODELS

In the current literature, many different stochastic extensions of the DEA framework can be found. Such generalized approaches enable one to model the uncertainty inherent to the form of the production possibility set and the value of the technical efficiency measure at any given point of the former. In the paper we provide a detailed discussion of some statistical model and the subsampling algorithm which are used in statistical inference. The methodology is then illustrated with an empirical example using the real world data from the Polish energy sector.