China's Baffling GDP Figures

Why are provincial and national economic data so different?


Kim Kyung-Hoon/Reuters

In the April, China's central government and provincial authorities released data on the country's economic performance for the first quarter of 2013. On April 15, the National Bureau of Statistics announced that the country's year-on-year real GDP growth rate had been 7.7 percent.


But something baffling arises when one compares the number published by the central government with those published by provincial authorities. By April 29, all of the 31 provincial governments had released economic statistics. Among them, in all defiance of the law of averages, no single province claimed a real GDP growth rate lower than the nationwide number. What gives?

Although apparently surprising, the contradiction between the national and provincial data is nothing more than a continuation of a long-time trend. In 2012, the real GDP growth rates released by all provincial authorities except Beijing and Shanghai exceeded 7.8 percent, the nationwide figure.

In 1985, the national statistics bureau and provincial statistic agencies were separated into two non-interfering systems. Since then, economic data from each province have been gathered and calculated by each provincial agency independently. The national bureau has been charged only with accounting for national economic data, without relying on data briefed by each province. Since 1985, a discrepancy between national and provincial data has been continuously widening. In 2009, 2010 and 2011, the sum of GDP figures released by all provincial agencies outnumbered the figure from the national bureau by 2.68 trillion RMB (US$430 billion), 3.5 trillion RMB and 4.6 trillion RMB, respectively. In 2012, the gap grew to 5.73 trillion RMB, approximately 11 percent of China's national GDP.

This discrepancy is partially explained by technicalities. When collecting primary data, the central bureau and provincial agencies sometimes use different basic "reference point" data  and statistical calibrations. Second, certain inter-provincial economic activities are claimed by all the provinces involved, leading to double counting.

Yet it is hardly possible that technical errors alone can account for such a consistent, substantial, and ever-growing discrepancy. Chinese media and citizens have inevitably speculated that economic data have been artificially modified.

Part of this skepticism traces to the incentive system for Chinese officials. Economic performance is still used as a major criterion for the evaluation and promotion of provincial bureaucrats. This leaves economic statistics particularly subject to political manipulation to serve officials' personal interest. As @xuehanbo observes on Sina Weibo, China's Twitter, "The reason why the data published by provincial agencies are falsely high is actually simple: the local authorities are still crazy for GDP growth, which is the key standard used to appraise officials' performance." During a meeting in April 2012, Ma Jiantang, chief of the National Statistics Bureau, explicitly admitted that data fabrication has been the major source of statistical distortion.

In order to solve the problem, the National Statistics Bureau has proposed several initiatives to reform the bifurcated statistical reporting system. In 2004, the Bureau proposed to monitor all economic data reported by provincial agencies. In 2006 and again in 2011, the Bureau advocated for the merger of the separate systems and offered to take direct charge of provincial economic statistics reporting.

Yet none of these proposed initiatives has been turned into sustained practice. For one, the root cause of statistical corruption would not have been eliminated with reforms. Since the current system of evaluation and promotion of officials provides the primary incentive for statistical manipulation, it is in fact this system that needs to be changed. An recent editorial by the Beijing Youth Daily (@ 北京青年报) suggests as much:

There is only one way to end the contradiction between provincial and national GDP statistics: to reform the official evaluation system and increase citizens' say in the process so that local authorities are not responsible for GDP growth alone, but also the environment, high-quality public services, and sustainable development. Only then will there be hope of freeing ourselves from the awkward GDP fights between localities and the central government.

This post also appears at Tea Leaf Nation, an Atlantic partner site.