Maximum Likelihood Estimator for Spatial Stochastic Frontier Models

This research is devoted to analysis of efficiency estimation in presence of spatial relationships and spatial heterogeneity in data. We presented a general specification of the spatial stochastic frontier model, which includes spatial lags, spatial autoregressive disturbances and spatial autoregressive inefficiencies. Maximum likelihood estimators are derived for two special cases of the spatial stochastic frontier. Small-sample properties of these estimators and comparison with a standard non-spatial estimator were implemented using a set of Monte Carlo experiments. Finally, we tested our estimators on a real-world data set of European airports and discovered significant spatial components in data.