This paper considers the semiparametric stochastic frontier model with panel data which arises in the problem of measuring technical inefficiency in production processes. We assume a parametric form for the frontier function, which is linear in production inputs. The density of the individual firm-specific effects is considered unknown. We construct an effient estimator of the slope parameters in the frontier function. We also give an estimator of the level of the frontier function and its asymptotic properties are investigated. Furthermore, we provide a predictor of the individual effects which can be directly translated to firm-specific technical inefficiencies. Finally, we illustrate our methods through a real data example.
Park, B. u., & Simar, L. (1992). Efficient semiparametric estimation in stochastic frontier model (STAT Discussion Papers 9201). https://hdl.handle.net/2078.5/32579