Wang borrowed methods used by sociologists, letting surveyors ask their close relatives and friends. They didn't ask what income was legal and what wasn't-- they simply asked what the yearly family income was and what they consumed, and then according to various income levels, calculated their consumption levels.

One of the main consumption indicators is the Engel's Coefficient, which measures the proportion of income spent on food. "I only had just over two thousand samples, so I clearly couldn't extrapolate the data to the whole nation. But as long as these figures are relatively reliable, it doesn't matter that there are only two thousand. They can still be tested against the available data."

The results of the investigation were fascinating. The National Bureau of Statistic's and Wang's data were identical for the bottom ten percent, but as one climbs the income ladder, the discrepancies grew. At the top, they were massive.

In the report, Wang writes, "for the top ten percent of urban families-- 19 million, with 50 million total family members-- disposable income is 97,000 yuan, three times higher than the under 29,000 yuan previously thought. Figures for upper-middle income families are also higher than the public figures. Rural residents make a gross income of 13.5 trillion yuan, and not 8.3 trillion. Unaccounted urban income for the whole nation could be as much as 4.8 trillion yuan, accounting for 26 percent of GDP.

But Are these figures reliable?

Wang Xiaolu says, "I've used various methods to verify my data, for example, we would ask how many people have cars, how many live in commercial housing... we knew what income levels can support these kinds of consumption habits, and compared that to how their bank loans were distributed. From these, we calculated their income levels ourselves. The end result was that the data was all reconcilable. Of course, the next step in verification should be taken, and future research may discover mine to have ended with slightly higher or lower figures."

When asked if both were possible, Wang says yes. The 4.8trillion yuan that he believes to be unreported income remains to be confirmed, but with a larger data pool, it can be. But his figure likely to be more conservative than anything else.

Wang Xiaolu believes that there are five main sources of grey income: government financial organs, the universal nature of financial corruption, administration of licenses and permits, the loss of land value, and monopolized industry income. He adds, "the reason for the wide income distribution gap does not lie in the freeing of the market. It is a systemic problem. This leads to corruption and grey income."