Title: A vine copula approach to estimation of a production system with technical and allocative inefficiency
Authors: Jian Zhai - University of Sydney (Australia) [presenting]
Artem Prokhorov - University of Sydney (Australia)
Abstract: Models are considered where an equation for the stochastic frontier production function is complemented by the first order conditions for cost minimization. Such models allow for both technical and allocative inefficiencies. A recent paper proposed a new family of copulas that is appropriate for modelling dependence between that two types of inefficiency. The copula is called APS-d, where d is the number of inefficiency terms. We consider a vine copula construction that makes use of multiple APS-2 copulas and a Gaussian copula. We argue that this construction is natural for this type of productivity models and achieves a more comprehensive coverage of dependence than APS-d, where d can be substantially greater than 2. We also study the extra precision in the estimation of technical inefficiency scores, permitted by the use of dependence information among the error terms in such models of production.