Feb 04, 2013 07:20 PM

The Real Problem with Those Gini Numbers


The National Bureau of Statistics (NBS) recently released China's Gini coefficient figures, which are a measure of income inequality, for the years 2003 to 2012.

The official reading has been over 0.47 for the last decade. The number peaked at 0.491 in 2008, and dropped to 0.474 in 2012. (The index runs from 0, total equality, to 1, or total inequality.)

Many people have questioned the accuracy of the official figures. How reliable are they? Well, the NBS has China's largest pool of data on urban and rural residents, and I believe its methodology, based on simple random sampling, is fine. The survey result for the low-income bracket is reliable, and different years' data based on the same sampling methodology is comparable.

But the main problem is the NBS' data on urban residents. It is safe to say that the high-income bracket has been hugely underestimated and the sampling does not reflect the reality of current income distribution.

The NBS says the annual disposable income of the richest 10 percent of urban residents was 43,000 yuan in 2008, much lower than my research result of 139,000 yuan. The number in 2011 was 59,000 yuan, which hardly explains what we see today – high housing price in cities, rapidly expanding private car ownership, a total of 35 trillion yuan in private bank deposits, the flow of private funds overseas and the zeal for luxury goods shown by Chinese traveling abroad.

Bureau head Ma Jiantang said that "we feel that our urban Gini coefficient reading based on a survey of urban residents is too low. The main reason is it's hard to access the true figure for the high-income group."

The problem is two-fold. Many high-income residents do not want to take the survey, and the NBS cannot guarantee that their substitutes have the same level of income, a situation that eventually causes part of the high-income group to be omitted. And those high-income survey-takers do not necessarily provide a full picture, especially when they have significant grey income, or income from extralegal sources. Therefore, the result based on reported income is lower than reality.

With that in mind, the next question is, what can statistics officials do and what is beyond their power?

The NBS can improve the quality of its Gini figures by adding high-income samples. One method is to use extra information to adjust the number. For instance, the NBS could collaborate with other government departments that collect information on real estate ownership, car registration and bank deposits to cross check its data on high-income residents. It can also use some macro data to deduce the income information.

Of course, changes to statistical methodology cannot change the fact that the country suffers rampant corruption and some high-income residents have secret sources of income. This is the root cause of unreliable income statistics. The cure for this problem can only be changes in the system that address issues such as unfair distribution of land, financial resources and gains by monopolies, and the embezzlement of public funds.

Statistics officials cannot fix these problems. This can only be achieved by reforms to improve government transparency and efficiency, by putting checks on power, and by exposing systemic malfunctions to rigorous public scrutiny.

The author is a blogger at Caixin and the deputy director and a senior fellow at the National Economic Research Institute of the China Reform Foundation. The article was first published on the Chinese language website of The Wall Street Journal

You've accessed an article available only to subscribers
Share this article
Open WeChat and scan the QR code