You really can’t trust China’s official GDP statistics. That is the gist of an explosive new research paper by economists at the Chinese University of Hong Kong and the University of Chicago.
The economists estimate that from 2008 to 2016, official GDP statistics were inflated by an average of 1.7 percentage points per year. If their findings are accurate, that would mean that China’s economy is more than 10% smaller than currently reported by the government.
China’s national GDP statistics are primarily based on data collected by local statistical agencies. As the economists point out, the officials at these agencies are under pressure to make make their province’s numbers look good. That is because they are evaluated by local government officials whose promotions are partly based on how well provinces perform on these metrics.
What statistics on the Chinese economy can you trust?
The government has admitted that these incentives sometimes lead to overblown local GDP numbers (paywall). The national agency that produces China’s overall GDP tries to account for this, but according to these economists, it stopped doing a good job around 2008. Perhaps not coincidentally, this was just when the economy started slowing down from its boom in the early and mid-2000s.
The economists rely on publicly available numbers they believe are more difficult to fudge. Most importantly, they examine the growth in revenue from taxes. They argue that tax fraud and evasion are difficult in China, and since that difficulty has not changed much over time, tax revenues for a given sector should grow similarly to that sector’s GDP growth. Up until 2008, that was true. It hasn’t been since then.
Adjusting GDP to be more in line with this tax growth is a big component in how the economists come to make their revised national and provincial estimates. Inner Mongolia and Liaoning are among the provinces they find most overestimate their local GDP, inflated by 20% and 17% respectively.
Bringing real numbers to light
The research is part of larger academic literature showing GDP numbers are often unreliable, particularly in autocratic countries.
Now, researchers in the new field of “forensic economics” are finding better ways to analyze economic prosperity. Tax revenue, satellite data on the brightness of a country’s nighttime lights, and electricity consumption have all been used to assess the accuracy of official growth figures.
This unbiased data offer us a more realistic understanding the global economy.