China's Unreliable Economic Statistics

China's National Bureau of Statistics launched a campaign in late July to boost the patriotic fervor of government statisticians in honor of the 60th anniversary of the founding of Communist China. The campaign, with the lugubrious title "Statistical Feelings: We have walked together—Celebrating the 60th anniversary of the founding of New China," was open to all present and former statistics workers and offered an unspecified "generous cash prize" for entries that were published by the NBS.

Among the patriotic dribble that made the grade were such gems as "I'm just a brick in the statistics building of the People's Republic" and "Love the motherland, love statistics," which included this unintentionally ironic line: "I can rearrange the stars in the sky above because I have statistics."

I'm sure statisticians everywhere need to be cheered up now and then, but one can't help wondering whether this patriotic poet was having a little dig at the state of his profession in modern China.

Over the last six decades, it's fair to say the PRC has had an uneasy relationship with official statistics. China's economic growth figures, in particular, have always been taken with a grain of salt by serious economists. Much of the skepticism of recent years stems from the late 1990s when they were unable to reconcile falling power consumption and widespread state enterprise closures with the rapid growth rates reported by Beijing.

Fresh doubtsOfficial stats have improved over the last decade as the NBS got better at working out what was really going on and began telling the truth, instead of just what it thought the policymakers wanted to hear. But the effects of the global economic crisis and rising political pressure for the numbers to tell a positive story have led to new questions about the accuracy of Chinese statistics.

Quite a few economists just don't believe the rosy picture now being painted by the official numbers.

For one thing, there is a huge gap between the country's gross domestic product as calculated by the NBS and the aggregate national GDP published individually by provincial and municipal governments. This gap exists because the NBS has always discounted data it gets from the provinces on the assumption that local officials are inflating their GDP growth numbers to improve their chances of promotion.

But in the first half of this year the gap was a lot wider than normal. All but seven of China's 31 provinces and municipalities reported GDP higher than the national average of 7.1% in the first half and the total as calculated by each province came in about 10% higher than the NBS's national figure.

In the world's third-largest economy that is a calculation error of around US$250 billion—approximately the entire annual GDP of Thailand.

Widespread skepticismQuestions over funny numbers are even being voiced by state-controlled media, usually the willing accomplices in publicizing questionable figures. Global Times, an affiliate of People's Daily, the Communist Party mouthpiece, reported in July that the public reacted with "banter and sarcasm" to NBS figures showing average urban wages in China rose 13% in the first half of 2009. It quoted an online poll showing 88% of respondents doubted the official data and many thought wages had actually fallen.

An editorial in China Daily, Beijing's English-language mouthpiece, quoted another survey that found 91% of respondents were skeptical of official data, up from 79% in 2007. The same survey found that sex workers ranked alongside farmers, religious workers, soldiers and students as the most trustworthy groups in Chinese society, while government officials ranked far lower and statisticians didn't rate at all.

My hunch is that the numbers are fiddled a bit at the local level and are then adjusted by the NBS to try and capture a more accurate picture of what's going on.

The biggest problem today is not that Beijing is making up numbers in the name of politics, but that statisticians miss huge segments of the puzzle. These include the gray economy and much of the private sector, which don't get counted in any surveys. If you can't see all the stars it makes it harder to rearrange them—even with magical statistics.

Before it's here, it's on the Bloomberg Terminal.