Estimation of a Panel Stochastic Frontier Model with Unobserved Common Shocks

This paper proposes a panel stochastic frontier model with unobserved common shocks to control cross-sectional dependence among individual firms. The novel feature is that we separate technical inefficiency (decision-dependent heterogeneity) from the effects induced by individual heterogeneity (decision-independent) caused by unobserved common shocks. We propose a feasible maximum likelihood method that does not require estimating the effects of unobserved common shocks and discuss its asymptotic properties. Monte Carlo simulations show that the proposed method has satisfactory finite sample properties when cross-sectional dependence exists. Application is illustrated by comparison of the efficiency of savings and commercial banking industries in the US.