基于CD生产函数的中国省域GDP数据质量评估

Accuracy Evaluation of Provincial GDP Based on Chinese Provincial Cobb-Douglass Production Function

2000年以来有关中国政府统计数据质量的质疑此消彼长,新常态经济背景下GDP增长率更是人们关注的焦点。目前,中国已正式采纳IMF的SDDS。中国政府统计工作需要进一步研究适用于中国的数据质量评估方法。然而,有关中国省域GDP数据质量评估的研究相对较少,其评估方法与中国GDP数据质量评估方法也不尽相同。本文从省域差异、经济理论和动态变化的视角,以CD生产函数为基础,考虑各要素投入和科技进步与GDP之间的关系以及省域间差异的影响,建立适用于省域GDP数据质量评估的统计模型。综合要素投入产出弹性系数和异常值动态识别的评估结果表明,中国省域GDP数据质量在总体上较好,少数省域年度GDP被高估的可能性大。

The problem about governmental statistical data quality of China in recent 20 years is a permanent fact. Growth rate of GDP is also the focus in the context of new normal of economic. Currently, SDDS of IMF has been accepted in China, it is critical for Chinese official statistics to further study the method of data quality. The research about data quality evaluation of Chinese provincial GDP is little, and the evaluation method is differ from the whole nation GDP. This paper improves the statistical model form of CD production function,based on the perspective of provincial, economic theory and development, and make it applied to the data quality evaluation of provincial GDP. Based on factor input-output elasticity coefficients and dynamically identifies of outlier, the results show that the quality of Chinese provincial GDP data is generally good, and there is the possibility to overestimate GDP of a few provinces in a certain year.