Abstract
It is shown that the bootstrap method should be applied with care for the Data Envelopment Analysis (DEA) estimator of average technical efficiency if the production frontier is stochastic. A stochastic production frontier leads to an inconsistency of the DEA estimator, which in turn leads to inconsistent and potentially highly misleading bootstrap confidence intervals. A Monte Carlo simulation study reveals that the empirical coverage accuracy of the bootstrap confidence intervals approaches zero as the sample size increases, even for small contributions of frontier variance to total frontier and efficiency variance.