中国居民“灰色收入”引发争论

中国居民“灰色收入”引发争论
2010-09-01 10:41:28    作者：    来源：《华尔街日报》中文网
If Chinese people don't trust the figures generated by their National Bureau of Statistics, they may have only themselves to blame.
It's become increasingly widely accepted among researchers that better-off Chinese people not only hide their money from the taxman, they also don't honestly answer the survey questions that government agencies use to figure out household incomes. The clear implication is that official income figures are too low.
'Due to people's increasing sense of privacy, there is some omission and understatement in the current income survey, and relatively few members of high-income groups are in the survey sample,' writes one researcher.
This statement comes not from a bomb-throwing critic, but an official within the statistics bureau: Wang Youjuan, who this week published a lengthy essay (in Chinese) on the bureau's website. The essay carries a disclaimer that it represents only the opinions of its author, and not the National Bureau of Statistics.
Wang's essay was a response to a recent, widely circulated paper by an outside scholar, Wang Xiaolu of the National Economic Research Institute in Beijing, entitled 'Gray Income and the Household Income Gap' (in Chinese).
The NERI paper presents the results of an unusual survey that tries to overcome the biases in the official income figures produced by the NBS. 'We contend that the present statistics on residential incomes have major distortions, especially in the part concerning the disposable income of high-income residents in urban areas,' the paper says.
To get around this issue, NERI instead surveyed thousands of people who were personally known to the researchers - and therefore presumably more likely to be honest. That has its own issues: 'Our method is different from that of random sampling, and therefore the data cannot be used directly to calculate the general distribution of urban residential income,' Wang Xiaolu concedes in the paper. (This and other quotes are from an English translation of the NERI paper provided by Credit Suisse.)
But he uses some complicated statistical tricks to nonetheless come up with an income distribution based on patterns of consumption among the people surveyed. The headline-grabbing result: China's average annual per-capita income for 2008 is nearly twice the 15,781 yuan (about $2,300) estimated by the NBS. And this is primarily because of unreported 'gray income' among higher-income groups.
Wang Youjuan of the NBS said that number is clearly too high, and took issue with how NERI constructed its survey, and also with the model used to come up with the income. (Among other things, he thinks that people are likely to lie about their income to acquaintances as well as to strangers.) But he did not reject the approach out of hand, saying that Wang Xiaolu had ideas and techniques that the NBS could learn from. And he emphasized that the NBS is trying to fill in the gaps on its own.
'We are currently organizing a large nationwide sample survey of urban households, and we hope by increasing the sample size and using a simpler questionnaire we can get a higher response rate and more participation by high-income households,' Wang Youjuan writes. The NBS is also working on other surveys that would collect indirect data about nonresponding households and estimate how big the omissions and under-reporting are, he says.
The response of the NBS is notable for its scholarly politeness - the tone is of disagreement among colleagues - and Wang Youjuan closes by thanking Wang Xiaolu for highlighting issues in the income statistics that need to be addressed. The NBS itself produced two different estimates of national household income: Based on the survey, it's a total of 14 trillion yuan ($2.06 trillion), but in the flow of funds accounts, it adds up to 17.9 trillion yuan.
The detailed and technical response is also notable for how it circles around the political implications of Wang Xiaolu's work: that income inequality in China is much worse than it appears to be, and that no one really knows how many wealthy people got their money. Wang Youjuan notes that income tax records, which other countries use to check the results of household surveys, are not extensive enough in China to be used for this purpose.
'Such a concentration of hidden income in high-income groups demonstrates that much of it is not about simple statistical problems in the household survey but potentially income from illegal sources,' Wang Xiaolu writes. 'The facts show that gray income has its origins in the misuse of power and is closely connected with corruption.'
如果中国人不相信由他们的统计局所公布的数字，他们或许只能怪自己。
有这样一种观点正在研究人员中获得越来越多的认可：富裕的中国人不但向税务人员隐瞒了他们的收入，在接受政府部门用来统计居民收入的调查时也说了谎。一个清晰的暗示是，官方的收入数据太低了。
一位研究者写道，由于人们的隐私意识日渐增强，现在的收入调查中会出现一些漏报和少报的情况，而且调查取样中高收入群体的数量也相对较少。
这一言论并非来自投下炸弹的批评人士，而是来自一名统计局官员：王有捐。他上星期在国家统计局网站上发表了一篇长篇文章。文章的免责声明说，它仅代表作者本人观点，不代表统计局观点。
王有捐的文章是对最近一篇广为流传的文章的回应，即中国经济改革研究基金会国民经济研究所学者王小鲁撰写的《灰色收入与居民收入差距》。
王小鲁的文章提供了一个不同于常的调查所产生的结果，该调查试图克服统计局所公布官方收入数据中的偏颇之处。这篇文章说，我们认为居民收入的现有数据存在重大失真，尤其是在城镇高收入居民的可支配收入上。
为了证明这个问题，国民经济研究所课题组对数千人进行了调查，受访者均与调查者相识，因此这大概也让他们更有可能说实话。王小鲁在文章中承认，由于所使用的方法不同于常规的统计抽样方式，因此不能用来直接推断全国总体的城镇居民收入分配状况。
不过他采用一些复杂的统计学方法，却得出了基于受访者消费模式的收入分配状况。这个引人注目的结论是：中国2008年人均年收入几乎是统计局所公布15,781元（约2,300美元）的两倍。而这主要是因为高收入群体的灰色收入没有被包括。
统计局的王有捐说，这一数字有过高之嫌，认为王小鲁课题组使用的调查和估算方法有值得商讨之处。（其中，他认为人们有可能在收入问题上对陌生人撒谎，同样也可能不会对熟人如实相告。）但是他没有对这种方式提出反对，说王小鲁的想法和方式也有可供统计局借鉴之处。他强调，统计局正在促进统计调查方法制度的不断完善。
王有捐写道，我们正在组织开展全国范围的城镇住户大样本调查，希望通过更大的样本量、较简单的问卷、较高的回答率来获得高收入户的参与比例，同时，对无回答住户搜集其住房、社区环境等辅助资料来评估常规调查户样本结构偏差。我们也计划对常规调查户开展一些无记名的邮寄问卷调查来比较其在记账过程中的漏报、少报收入比例情况。
统计局的这一回应因其学者式的彬彬有礼而受到关注，文章的语气就像是同事间在表达不同看法。王有捐在文章最后还感谢王小鲁指出了在收入统计工作中需要关注的问题。在家庭收入的问题上，统计局自身就给出了两种不同的预测：根据调查，全国家庭总收入为14万亿元（约2.06万亿美元），而依据资金账户显示，这一数字为17.9万亿元。
这份细致而专业的回应还因其如何巧妙回避了王小鲁文章所蕴含的政治含义而收到重视：中国的收入不公问题远比它所表现的要严重，没有人真的知道那么多有钱人是如何挣到钱的。王有捐指出，在西方国家多数使用个人收入所得税数据对居民收入调查结果进行评估校正，而我国还缺乏比较完善的个人收入所得税资料，无法使用此方法。
王小鲁写道，高收入阶层出现严重的隐瞒收入现象的原因之一是有大量来源不宜公开的收入，这与腐败和其他非法收入密切相关，也包括各种违规违纪收入。