Efficient Semiparametric Estimation in a Stochastic Frontier Model
针对面板数据的随机前沿模型，假设前沿函数参数形式已知但企业个体效应密度未知，构造了前沿函数斜率参数的有效估计量，并给出个体效应的预测方法，用于衡量生产技术无效率。
This article considers the semiparametric stochastic frontier model with panel data that 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 to be unknown. We construct an efficient estimator of the slope parameters in the frontier function. We also give an estimator of the level of the frontier function and investigate its asymptotic properties. Furthermore, we provide a predictor of the individual effects that can be directly translated to firm-specific technical inefficiencies. Finally, we illustrate our methods through a real data example.