海角大神

When you want to know what is going on in the Chinese economy, where do you look?

Paul Cavey studies excavator sales.

What might seem an arcane detail is actually 鈥渁 real indicator鈥� of economic trends, says Mr. Cavey, a China analyst at Macquarie Bank. Digger sales point to likely construction, which is a big driver of Chinese economic growth, he notes. 鈥淵ou can鈥檛 draw macro conclusions from those figures,鈥� he acknowledges, 鈥渂ut they serve as a sense check of official data.鈥�

That鈥檚 important when the official data paint a conveniently rosy picture, such as Tuesday鈥檚 announcement that GDP growth in the last quarter of 2011 was 8.9 percent, marginally more than expected, which suggests China will have a soft landing as its economy cools.

That news was good, and it mattered: It lifted stock markets across Asia. But was it true?

Not for Derek Scissors, a researcher at the Heritage Foundation in Washington who dismissed the figure as invented 鈥渘onsense 鈥� just a PR exercise and a piece of theater.鈥� Mr. Scissors pointed to discrepancies between reportedly solid GDP growth last year and sharp slowdowns in the growth of auto sales, ship orders, and oil imports.

Other experts who also earn their living by poring over Chinese government figures are less dismissive. 鈥淓conomic statistics are more of an art than a science, but Chinese statistics are reliable enough for practical purposes,鈥� says Arthur Kroeber, head of the Dragonomics economic consultancy in Beijing.

Can you trust the data?

鈥淭here鈥檚 huge uncertainty over what the data means and whether it can be trusted,鈥� Tom Orlik told reporters recently, introducing his book 鈥淯nderstanding China鈥檚 Economic Indicators.鈥�

聽鈥淲hen people think of the production of economic data in China,鈥� he added, 鈥渢hey think about a couple of drunken Communist party officials 鈥� picking the number that best fits their own political needs.鈥�

Once, he recalled, that was not far off the truth. The manipulation of economic data for political ends has a particularly tragic history in China: 30 million people died of famine at the end of the 1950s when rural officials fabricated grain production figures in order to pretend to meet the government鈥檚 absurdly ambitious targets, providing a wholly illusory impression of the nation鈥檚 grain stocks.

Forty years later, during the 1998 Asian financial crisis, provincial officials were still massaging the output figures by which their superiors judged them, said Mr. Orlik. Since then, however, 鈥渨e have seen significant moves to improve China鈥檚 data collection system鈥� by centralizing it, he added.

鈥淐hina鈥檚 economic statistics are actually very impressive,鈥� he says, 鈥渨ith relatively timely, accurate, and comprehensive data published on a range of key indicators.鈥�

Not that this is true across the board. China鈥檚 National Bureau of Statistics makes no serious effort to measure the politically sensitive unemployment level, for example: It has officially stood at around 4 percent for a decade, which nobody believes.

Nor does anyone looking for an apartment believe the official accounts of only moderate increases in housing prices 鈥� another politically sensitive question 鈥搘hich can only have been based on heavily skewed housing samples.

Scissors sees such figures as evidence that 鈥渢he statistics bureau is political. It鈥檚 not all lies all the time, but the data is so unreliable that I wonder why we are analyzing it so carefully,鈥� he says.

Even some top officials apparently harbor doubts. Li Keqiang, expected to be China鈥檚 next premier, told the American ambassador in 2007 that the official GDP figure was 鈥渕anmade鈥� and 鈥渇or reference only,鈥� according to a US embassy cable released by Wikileaks.

But a number of foreign analysts point out that the Chinese government has managed its economy well over the past three decades, that it must have used accurate economic data to do so, and that economic policy generally reflects the published figures, which suggests the government is not using a secret second set of accounts.

Rather, they suggest, officials simply cannot always get a firm grip on reality.

Back to reality

Data on fixed asset investment, for example, give a very different picture of housing starts than reports of the area under construction. That, says Cavey, is because developers are required to build on land they buy within two years. 鈥淪o they tell the government they are building on all their land, so as to avoid having it confiscated, when in fact they are not investing very much at all.

鈥淭he standards for reporting are not as high as they are elsewhere in the world,鈥� he adds. Rich people lie about their income to hide bribes, small businesses such as shops and restaurants under-report their activity, and 鈥渋t is very difficult to have a system that effectively tracks an economy whose structure is changing all the time.鈥�

鈥淭he worst fears of political manipulation of the data are misplaced, but the numbers still need to be taken with a grain of salt,鈥� Orlik believes. In the end, he says, reality will prevail over any official attempts to misrepresent it.

鈥淧eople have a lived experience of what鈥檚 going on in the economy,鈥� he points out. 鈥淭hey either have a job or they don鈥檛 鈥� they are either seeing wages rise faster than inflation or lower than inflation. That will be much more important for them than whatever the Chinese government says.鈥�

And as more and more people share their experiences and feelings on the Internet, he predicts, 鈥渢his will mean there is new pressure on officials to produce statistics which are accurate.鈥澛�