国家与省级汇总GDP数据的差异问题研究

On Deviation of National GDP Statistics from Total Provincial GDP Statistics

我国各省GDP的汇总数据十年来持续大于国家数据,两者的误差的绝对值和相对值都在增大。各省汇总生产法GDP的正偏差大部分来源于第二产业,从支出构成看,资本形成总额各省汇总值远高于国家值。通过使用Benford法则和历次经济普查数据对全国各省GDP数据进行检验,并没有发现各省GDP数据显著性的作假、篡改或粉饰的证据。总体来说,第三产业增加值国家统计数据被低估了,第二产业增加值省级汇总被高估,数据修正后,GDP国家值与省级汇总值偏差被缩小。国家与各省汇总GDP产生的偏差主要是由省级GDP核算中的重复计算、统计资料缺失、数据来源不同、统计口径不一致、地方行政干预及统计协调欠缺等引起的。国家数据与省级数据的差异的缩小要从统计制度设计和方法技术改进两个维度进行。

Over the past decade, the total provincial GDP statistics were always larger than national statistics, and the absolute valve and relative value of the error had been increasing in China. From the perspective of production, the positive deviation of total provincial GDP statistics mostly lied in the secondary industry. From the perspective of expenditure, the total provincial value of capital was significantly larger than national value of capital. Based on the Benford law and the statistics of general survey of economy, the paper examined the quality of provincial GDP statistics. The results showed that no significant fraud was found in the provincial GDP statistics. In general, the national statistics of added value of tertiary industry were underestimated while the total provincial statistics of added value of secondary industry were overestimated, and with the correction of statistics, the deviation of national statistics from total provincial statistics was lessened. The deviation of national GDP statistics from total provincial GDP statistics mainly resulted from repeated calculation, lack of statistical data, different sources of data, inconsistent accounting method, local intervention and lack of coordination mechanism. It＇ s necessary to design statistical system and improve calculating method to lessen the deviation of national statistics from provincial statistics.