Robust non-parametric estimation of cost efficiency with an application to banking industry

The paper modifies the methodology of Simar and Wilson 2007 [J Econometrics 136] and 1998 [Manage Sci 44] to propose a new algorithm for robust estimation of cost efficiency in data envelopment analysis in terms of bias correction and estimating returns to scale. Simulation analyses with multi-input multi-output Cobb-Douglas production function with correlated outputs, and correlated technical and cost efficiency demonstrate consistency of the new algorithm both in absence and presence of environmental variables. Finally, we offer real data estimates for Japanese banking industry. An R package `rDea', developed for computations, is available from GitHub and CRAN repositary.