Title: Parametric stochastic frontier models and probability statements with spatial errors
Authors: William Horrace - Syracuse University (United States)
Christopher Parmeter - University of Miami (United States)
Ian Wright - University of Miami (United States) [presenting]
Abstract: The presence of spatial correlation in the error terms for stochastic frontier models yields an intractable likelihood where the number of integrals grows with the sample size. Sequential conditioning is used in order to factor the joint distribution which produces a likelihood that has only a single integral irrespective of the sample size. This leads to a likelihood function which is numerically more feasible to solve. Additionally, certain probability statements are generalized to account for spatial dependence. Maximum likelihood estimates are discussed.