Modelling multi-output stochastic frontiers using copulas

The aim is to introduce a new econometric methodology for multi-output production frontiers. In the context of a system of frontier equations, a flexible multivariate distribution for the inefficiency error term is used. This multivariate distribution is constructed through a copula function which allows for separate modelling of the marginal inefficiency distributions and the dependence. Specific attention to the elicitation of a sensible (improper) prior is paid and simple sufficient conditions for posterior propriety are provided. Inference is conducted through a Markov chain Monte Carlo sampler. Bayes factors are used to compare various copula specifications in the empirical context of Dutch dairy farm data, with two outputs.