POT KETTLE POT KETTLE POT

China’s well-hidden, ill-gotten gains

Aug 13th 2010, 11:55 by S.C. | HONG KONG

WOULD you lie about how much you earn? Would you lie about how much you spend? Would you, ultimately, fib about how much you spend on abalone and shark’s fin? If you are a well-connected member of China’s new rich the answers are yes, probably not, and no, according to professor Wang Xialou of the National Economic Research Institute at the China Reform Foundation. He is making waves with new estimates of China’s “shadow income”: the vast sums that richer households keep off the books and under the table. A piece in Caixin yesterday drew on his own study released by Credit Suisse on August 6th (summarised and posted in full on Sinocism). Mr Wang calculates that the disposable income of China’s households is 9.3 trillion yuan ($1.4 trillion) higher than official estimates suggest. That’s equivalent to 30% of China’s 2008 GDP.

Mr Wang asked his team of researchers to interview people they knew (over 4,000 families) in the hope that familiarity would breed honesty. The families were asked about their income and spending patterns. Their answers were truthful, Mr Wang assumes, but they may not have been representative of China as a whole. 

He therefore had to match his sample of families with the much bigger sample in the official household survey carried out by the National Bureau of Statistics. Food provided the link. As people get richer, they spend a smaller share of their dosh on nosh, according to Engel’s law, an economic regularity named after Ernst Engel. In Mr Wang’s survey, for example, the poorest people devoted 48% of their outlays to food, whereas the richest spent only 29% (an Engel’s coefficient of 0.29).

Mr Wang matched families from the two surveys according to their Engel’s coefficients. He would, for example, take the 565 families in his survey who spent 29% of their outlays on food, and line them up with families in the official survey with the same coefficient of 0.29. In theory, their incomes should match. But in practice, it was quite different. In the official survey, these people say their disposable income is 43,614 yuan. But in Mr Wang’s survey, such families admit their income is really 139,000 yuan.

The fibs get bigger as families get richer. The richest 10% of urban families underreport their income by about 69%, he finds. They are about 26 times better off than the bottom tenth, not merely nine times, as the official figures suggest. Indeed, China appears less egalitarian on this measure than America. The top 10% of Chinese households nationwide take home 51.9% of the income, whereas their counterparts in America hog only 47.2%, according to calculations by Credit Suisse, which sponsored the study.

If China is earning 9.3 trillion yuan of "grey income", does that mean the country’s economy is 30% bigger than we thought? Not quite. China’s GDP figures do not rely on the NBS household survey, but on “flow-of-funds” data collected from enterprises. And some of the shadow income pocketed by Chinese families may already show up elsewhere in the national accounts, as corporate income or government income—it may be misreported, not unreported. The upshot is that China’s economy was about 10% bigger in 2008, according to Credit Suisse.

These shocking results rely on a number of assumptions. Engel’s law, for example, is not absolute. As Mr Wang admits, lots of other factors can interfere with it. So it’s possible that two families with the same Engel’s coefficient really do have different incomes, because of where they live, their family circumstances or their gastronomic passions. Mr Wang’s results suggest, for example, that the people of Beijing, Shandong, Hubei, Guangdong, Chongqing and Henan are China’s foodies, spending more on dining than the national average. Mr Wang does his best to control for such things. But there’s still plenty of play in Engel’s law that he cannot explain or control for.

Mr Wang also assumes that people lie about how much they earn, but not about how much they spend or eat. Or if a household does underreport its spending, he assumes that it also downplays its food expenses proportionately, so that their Engel’s coefficient is unaffected. 

If people underreported their overall spending, but told the truth about their food spending, their Engel’s coefficients will be artificially high. Mr Wang would therefore have paired them with the poorer households in his survey. In those circumstances, their underreported income would go undetected. If that’s the case, then 9.3 trillion may be an underestimate! China’s shadows may be even deeper than they seem.