Koop, GarySteel, Mark F.J.Osiewalski, JacekUniversidad Carlos III de Madrid. Departamento de Estadística2009-02-172009-02-171992-12https://hdl.handle.net/10016/3677In this paper we describe the use of Gibbs sampling methods for making posterior inferences in stochastic frontier models with composed error. We show how the Gibbs sampler can greatly reduce the computational difficulties involved in analyzing such models. Our fidings are illustrated in an empirical example.application/pdfengAtribución-NoComercial-SinDerivadas 3.0 EspañaComposed error modelsBayesian econometricsGibbs samplingRejection samplingworking paperEstadísticaopen access