中国地区GDP增长统计的真实性检验

A Regional Level Test on the Authenticity of China's GDP Growth Statistics

本文利用中国28个地区(省、市、自治区)近二十年10个经济基础变量的平行数据,运用主成分分析和固定影响变截距模型的经验分析方法,为中国GDP增长的真实性讨论提供了地区一级数据的检验。我们发现,从整个时期来看,中国地区的GDP统计数据没有出现违背经济规律的统计特征,其中也包括能源增长与经济增长的关系;年度的特殊影响也不太严重,特别是质疑最多的1998年及以后各年并无显著的正影响;地区的特殊影响虽然比较大,但也许能从地区经济结构及变化等方面的差异上得到解释,未必成为GDP数据存在造假的证据。

Using the Panel Data of 10 fundamental economic variables of 28 provinces of China in about 20 years, with the Principle Component Analysis and Fixed-Effect Model, this paper tests the authenticity of China's GDP growth statistics. It reveals that, during all the period, there is no statistical contradiction between the annual change of major economic components and the resulting GDP growth, including energy-GDP relationship. Special effects of regions do exist, but might be explained by the differences within regional economic structures, and its transition and cannot solely be used as evidence for deception. Furthermore,yearly special effects have slightly been found, yet absent in 1998 and later on.