A simultaneous system of dynamic spatial stochastic frontier models with dependent error components and inefficiency determinants

In this paper, we develop a system of simultaneous stochastic frontier models with inefficiency determinants, spatio-temporal effects and correlated inefficiency as well as correlated random errors among frontiers. The dependence among the errors of the different equations can stem from either shocks external to the system, interrelated inefficiency mechanisms, or a combination of both. Estimation is performed using a copula-based quasi-maximum likelihood approach. Simulation results confirm the good finite sample properties of the proposed estimator. To demonstrate the effectiveness of the proposed model and estimation technique in empirical settings, we analyze the key role of some sustainability-related factors in determining the efficiency level of Italian cultural and creative sectors.