基于模型比较视角的我国区域技术效率再估计

Re-Estimation of Chinese Regional Technical Efficiency Based on the Perspective of Model Comparison

基于模型视角,对CSSW、CSSG、BC及KSS四种经典的随机前沿测算模型进行了比较研究,并对1997—2010年间我国区域技术效率进行了再估计。研究发现:(1)蒙特卡罗仿真表明,当模型不存在内生性时,并不存在哪一种模型更为可靠;当模型存在内生性时,鉴于在技术非效率项和内生性处理的优越性,KSS模型是目前相对较好的测算模型。(2)格兰杰因果检验、Hausman-Wu检验及随机误差项正态分布检验共同表明,在利用对数型柯布—道格拉斯生产函数进行估计时,模型本身存在内生性,前人利用BC模型得到的结论可信度较低。(3)我国整体及区域技术效率水平仍然较低,并呈现涨跌互动的波动趋势,而非单调递增或递减;2008年之前东部地区技术效率水平最高,随后被中部地区赶超,并呈现差异扩大化趋势。

There are two completely contrary conclusions, i.e. increasing or decreasing monotonously on the region- al technical efficiency in the views of scholars. It is necessary to explore the reasons and re-estimate Chinese re- gional technical efficiency. From the aspects of models, this paper compares the four classical stochastic frontier models named CSSW, CSSG, BC and KSS, and estimates regional technical efficiency from 1997 to 2010 again. The major results are shown as follows. （1） The results of Monte-Carlo simulation indicate that, when does not exist endogeneity, there is not a model more reliable. In contrary, considering the superiorities in the technical inef- ficiency term and endogeneity processing, the KSS estimator is the relatively best technical efficiency model. （2） Granger causality test, Hausman-Wu test and normality test of random error term make it clear that there is the endogeneity problem when using the Logarithmic Cobb-Douglas production function. In other words, the previous results based on BC are unbelievable. （3） During the years of 1997-2010, the overall and regional technical effi- ciencies present the fluctuant character and differ from the previous conclusions significantly. Prior to 2008, the technical efficiency of eastern region is the highest, followed by central and western regions, while the central re- gion catch up with the eastern and the gap is going to widen after 2008