Ley, EduardoUniversidad Carlos III de Madrid. Departamento de Economía2009-11-272009-11-271992-032340-5031https://hdl.handle.net/10016/5820We study the use of switching regression models to characterize the coefficients in linear production technologies with a finite number of activities. Maximum likelihood-based methods are proposed and different switching specifications are discussed. The viability of these newly proposed technniques is established. The methods developed combine the advantages of the two major approaches to frontier estimation: the functional flexibility of the linear programing-nonparametric and nonstatistical-approach and the statistical nature of the econometric-both parametric and statistical-approach. This combination comes at the expense of some analytical complexity.application/pdfengAtribución-NoComercial-SinDerivadas 3.0 EspañaSwitching regression modelsActivity analysisLinear production modelsSwitching regressions and activity analysisworking paperEconomíaopen access