Measurement of Technical Efficiency in Stochastic Frontier Analysis with Limited and Qualitative Dependent Variables

As vividly demonstrated in Maddala (1983), limited and qualitative dependent variables are often encountered in modern econometric analysis. However, analytical methods for evaluating technical efficiency of stochastic frontier analysis can only be applied to continuous dependent variable. This paper provides closed form formulae for evaluating the technical efficiency of stochastic frontier analysis with limited and qualitative dependent variables. Monte Carlo experiments reveal that the finite sample performances of our formulae are promising