EFFICIENT SEMIPARAMETRIC ESTIMATION IN A STOCHASTIC FRONTIER MODEL

This article considers the semiparametric stochastic frontier model wi th panel data that arises in the problem of measuring technical ineffi ciency in production processes. We assume a parametric form for the fr ontier function, which is linear in production inputs. The density of the individual firm-specific effects is considered to be unknown. We c onstruct an efficient estimator of the slope parameters in the frontie r function. We also give an estimator of the level of the frontier fun ction and investigate its asymptotic properties. Furthermore, we provi de a predictor of the individual effects that can be directly translat ed to firm-specific technical inefficiencies. Finally, we illustrate o ur methods through a real data example.