Abstract
The purpose of this article is to estimate and compare shifts in (technical) efficiency across OECD countries, caused by the global financial crises and heterogeneity. Technical efficiency of OECD countries is estimated by applying the panel model with arbitrary temporal heterogeneity in time and factor structures (a model with unobservable individual effects) that fits the stochastic frontier analysis. Because of missing values in observations, the bootstrapping-based algorithm allowing for trends in data across observations within a cross-sectional unit is applied for imputations. The parameters are estimated in a semi-parametric way. The proposed estimation derives sufficient results regardless of any assumption on the temporal pattern of country individual effects and contributes to the development of a tool for better understanding of unobserved factors that drive fluctuations in OECD countries.