摘要
本文应用蒙特卡罗模拟方法,在定义单次模拟程序时,假设数据产生机制是一个超越对数随机前沿生产函数的IO模型,由此创造出模拟样本,并用一个超越对数的OO模型(scaling-property模型)计算出有关参数、特别是非效率项的估计值。又进一步判定了所得到的估计值和原来IO模型中的"真实"非效率项的一致性。研究发现,真实非效率项与从scaling-property模型中计算出来的非效率估计值之间的各种相关系数均为负值。因此,效率秩估计值和"真实"效率秩是不一致的。
Based on the Monte Carlo simulation, the data generation process is assumed as a translog production stochastic frontier model with 10 technical inefficiency in one simulation program. The simulation sample is created, and the scaling-property model is employed to fit it. So we can obtain the estimates of the parameters in the model and relative technical inefficiency. After that, we test the consistency between the estimates of inefficiency and the true inefficiency terms. What we find is that the correlation coefficients are negative, and the estimate of efficiency ranking is inconsistent with the true efficiency ranking.
出处
《统计研究》
CSSCI
北大核心
2007年第11期80-84,共5页
Statistical Research
基金
西安交通大学"985工程"二期(项目编号:07200701)资助。
关键词
随机前沿生产函数
一致性
蒙特卡罗模拟
Stochastic frontier production model
Consistency
Monte Carlo simulation
