Thomas Piketty是谁

万里风中虎

Thomas Piketty是谁 (2014-03-28 09:33:33)转载▼

标签： 杂谈       
纽约客和纽约时报近期都刊登了有关Piketty新书的文章。克鲁格曼在纽约时报的文章甚至认为这本书可能是这个10年中最重要的经济学著作。Piketty这个名字对于中国人来说也许还是相当陌生，但是在法国国内，他可能是法国经济学家中最富盛名的一位。作为巴黎经济学院少数几个中国学生，我感觉有必要介绍一下他。

Piketty是巴黎高师（ENS）学生，不到二十岁拿到高师的数学学位，随后便在Roger Guesnerie（曾经的Econometrics Society主席，现在法兰西学院院士）指导下继续在高师撰写经济学博士论文。博士期间他参加了欧洲数量经济学博士交换项目（EDP），在伦敦政经呆了一段时间（这段经历为后来他参与“创办”巴黎经济学院做了铺垫）。

博士毕业时，22岁的Piketty参加了1993年RES的Tour，随后在MIT拿到教职。两年后，由于“深深眷恋”他的祖国（他对法国文化是个保守的捍卫者，上课常常不自觉从英文变成法文...），放弃MIT教职回到法国国家科学院（CNRS）任职。随后在巴黎社会科学高等研究院（EHESS）和巴黎高师（ENS）任教（EHESS和ENS关系复杂，容后再叙）。

在担任了巴黎高师社会科学系主任两年后，2006年，35岁的Piketty在其博士论文导师Guesnerie以及后来的世行首席经济学家Bourguignon等人的协助下，整合了巴黎高师，巴黎社会科学高等研究院，国立路桥大学（ENPC），法国国家科学院，法国农科院以及后来加入和巴黎一大经济学研究力量，创办了巴黎经济学院（PSE）。

也许和高师人（譬如萨特）的入世精神遗传有关，巴黎经济学院的校训是“经济服务社会”（La science économique au service de la société）。法国很多的经济学家并不以参与社会事务为耻，很多人愿意脱掉白手套，积极参加社会政策的讨论。Piketty参加了07年社会党罗亚尔（现任总统奥朗德的前妻...）的总统竞选阵营，曝光很多。但由于竞选败北，Piketty卸任PSE的主任。

卸任后，他在法国媒体上还是很活跃，政治立场较左，但学术上还是很有成果。他的主要领域就是收入和财富不平等，以及再分配。在PSE他开有两门课，一门是公共财政，一门是不平等经济学（讲义都在他个人主页上）。他工作是通过详实的个税，遗产税数据估算长期各国的收入分配情况。他就是近来流行概念“top 1%收入”的肇始者。

他和Tony Atkinson，Emmanuel Saez等人将早期利用法国数据所做的研究拓展到英国，美国，加拿大，日本等国。后来陆续和其他研究者进一步拓广了研究，其中有关中国的长期收入分配论文是和Nancy Qian一起发表的。在学生Alvaredo的协助下，他还将这些论文研究有关的数据做成了“世界顶端收入数据库”（WTID）。

fcboliver

这个WTID不会成了各国查水表的名单了吧？

晨枫

中国个税、遗产税与西方有本质的不同，他的研究方法如何调整呢？

ccqi

看到了巴黎经济学院，就想到了伦敦政经。这是典型的ssx条件反射？

万里风中虎

晨枫 发表于 2014-4-21 21:16
中国个税、遗产税与西方有本质的不同，他的研究方法如何调整呢？

中国的个税和遗产税不会和西方有本质不同，只是没有财产公开制度，所以个税只针对2500万中产阶级，而遗产税根本不存在。所以，用家庭调查是更真实的方法：

Estimates based on households surveys only. No information available from the income tax. Potential comparisons with the other countries should consider this fact.

Piketty, Thomas and Qian, Nancy (2010). Income Inequality and Progressive Income Taxation in China and India 1986-2015; in Atkinson, A. B. and Piketty, T. (editors) Top Incomes: A Global Perspective, Oxford University Press, chapter 2.

晨枫

万里风中虎 发表于 2014-4-22 18:18
中国的个税和遗产税不会和西方有本质不同，只是没有财产公开制度，所以个税只针对2500万中产阶级，而遗产 ...

中国个税大体还是“富人税”，普通老百姓不怎么交。所以中国税收中个税不占大头。就凭这一点，就和西方不一样。家庭调查的局限就太大了，在哪个地区，抽查什么样的家庭，这出入就大了。这还不比街头抽查，要随机不容易。

万里风中虎

晨枫 发表于 2014-4-23 09:41
中国个税大体还是“富人税”，普通老百姓不怎么交。所以中国税收中个税不占大头。就凭这一点，就和西方不 ...

个税不是富人税，起征点是3500元，劳务更是扣除800（或20%）以上就要交。有几个人月收入低于这个的？不是普通百姓不交，而是大家都用各种方法在避税或者偷税。

个税的基础是民主制度，政府乐于看到大家偷税，这样就没理由享受民主权利。

而国家就用增值税这个没有立法的税种和企业所得税来替代个人所得税，只要你消费或者有工作，其实都交了税。

现在的情况就是，老百姓承担了流转税税负，却在自以为事地偷逃最重要的个人所得税。政府对得不到收入的个人所得税立法，对收入主体的增值税却不立法，处处把政治权利和税收挂钩。

什么样的百姓就有什么样的政府。

英国人口6000万，工作人口3000万，雇员2500万。交个人所得税的人员2300万。我在英国调查收入就是直接得到税务局600万份纳税单来研究，这些都是公开的。所以，个人所得税普遍征收是现代社会的特征，不是富人税。中国的富人也不交个人所得税，因为可以转成企业利润。自己的企业干嘛要付工资给自己。

中国的现状使得使用税收作为收入调查不可能也不可靠，国家税局和统计局也不会提供这些数据。只有用自己调查的家庭调查数据才是真实可靠的。有社科院和师范大学的CHIPS, 北大的CHARLS,卫生部和北卡的CHNS, 西南财经大学的家庭金融调查，还有人民大学等等的数据库。

你要相信我们这些职业搞调查的人员，这是我们的饭碗。如果我们做的统计的东西都没有考虑这些抽样误差，那我们岂不是和崔永元调查转基因一样会成为笑柄？而且我们的研究结果都是要发表在国际学刊的，如果有低级错误，大家早就笑死了。

如果一个人的结果和其他人不一样（比如说西南财经大学的家庭金融调查，调查就花了1000万人民币），要不就是真有发现像THOMAS PIKETTY一样发表在顶级学刊，要不就是犯了低级错误，要向整个学术界解释就麻烦了，都是功成名就的人，谁会毁了自己一生的学术声誉。

在这些技术问题上，世界上可能没有人比THOMAS PIKETTY和NANCY QIAN做得更好了。

tangotango

晨枫 发表于 2014-4-23 09:41
中国个税大体还是“富人税”，普通老百姓不怎么交。所以中国税收中个税不占大头。就凭这一点，就和西方不 ...

现在的个税起征点3000-4000吧，这怎么算得上富人税？连大学老师给研究生发的津贴（1000左右或者不到）都要交税。只有部分使用现金收入的穷人（不是全部，比如菜场的摊贩肯定也是要交税的）不交。
其实富人反而不用交什么个税，因为可以用公司来避税。车、房都可以挂在公司名下，汽油、家具等开支都可以开发票列入公司成本。所以现在个税主要是中产阶级负担的。

晨枫

万里风中虎 发表于 2014-4-22 20:28
个税不是富人税，起征点是3500元，劳务更是扣除800（或20%）以上就要交。有几个人月收入低于这个的？不是 ...

我不是质疑Thomas Piketty或者Nancy Chang的工作，但是在缺乏全民申报制度的情况下，你如何统计真正的收入？在加拿大的话，自己的公司不发自己工资是可以的，但你要是用公司的资源，比如说车、报销吃饭、补贴住房，这些都是要算作taxable perk。在中国没有这样的算法，所以有中国富人反而不上税的说法。但这样一来，真正的收入怎么计算呢？

谁都愿意相信权威不会犯低级错误。要使数据不受质疑，只要说明一下数据的收集方法就可以了。

liuqing098

万里风中虎 发表于 2014-4-23 10:28
个税不是富人税，起征点是3500元，劳务更是扣除800（或20%）以上就要交。有几个人月收入低于这个的？不是 ...

"个税的基础是民主制度，政府乐于看到大家偷税，这样就没理由享受民主权利。"
这个太深刻了。赞！

万里风中虎

晨枫 发表于 2014-4-23 12:08
我不是质疑Thomas Piketty或者Nancy Chang的工作，但是在缺乏全民申报制度的情况下，你如何统计真正的收 ...

关于福利，包括公司用车房产等是统计的难点，还有农民的耕种收入如何计量等等这都是问题。但是，都有相关的方法。中国的家庭收入调查和你说的方法没什么区别，因为调查的人都是在欧美受教育后。回国设计的这些调查。有的调查直接就是国外的大学教授参与设计的。

每个数据库都有调查方法的说明。比如说用我最熟悉的CHNS (http://www.cpc.unc.edu/projects/china/), 首先要按照人口普查的数据分级：省，市，县，村，家庭，个人。有的调查还要给不同的人给予不同的权重。比如说，你抽样到了比尔盖茨，他的收入就权重很小，因为他这种人太少。数据主要调查个人，家庭和社区三个层面。
The survey covers nine  provinces that vary substantially in geography, economic development, public resources, and health indicators. A multistage, random cluster process was used to draw the samples surveyed in each of the provinces. Counties in the nine provinces were stratified by income (low, middle, and high), and a weighted sampling scheme was used to randomly select four counties in each province. In addition, the provincial capital and a lower income city were selected when feasible. In two provinces, other large cities had to be selected.  Villages and townships within the counties and urban and suburban neighborhoods within the cities were selected randomly. In 1989 to 1993 there were 190 primary sampling units, and a new province and its sampling units were added in 1997.  There are about 4,400 households in the overall survey, covering some 19,000 individuals.  Follow-up levels are high, but families that migrate from one community to a new one are not followed.  Movement within the primary sampling units and some larger urban entities is attempted.

The first round of the CHNS, including household, community, and health/family planning facility data, was collected in 1989. Six additional panels were collected in 1991, 1993, 1997, 2000, 2004, and 2006. Since the 1993 survey, all new households formed from sample households were added. Since 1997, new households in original communities were also added to replace households no longer participating in the study. Also since 1997, new communities in original provinces have been added to replace sites no longer participating. A new province was also added in 1997 when one province was unable to participate.  The dropped province returned to the study in 2000.

The Chinese Ministry of Public Health has committed to continue this project as long as outside support is provided to supplement the extensive support it already provides. The research project is truly a collaborative effort, with funding with funding, expertise, and other skills shared equally by the American and Chinese institutions in significant collaborative effort.

The Household Survey
A complete household roster has been used as a reference for subsequent blocks of questions on time allocation at home (e.g., child care, elderly care, other key home activities) and economic activities. Questions on income and time allocation probe for any possible activity each person might have engaged in during the previous year, both in and out of the formal market. Information on water sources, construction, condition of the home, and ownership of consumer durables is gathered from the respondent. Additional questions probe for all possible items the household might own. Full income from market and nonmarket activities is imputed. The detailed estimation of income that is possible using these data represents a significant advance in the measurement of income in China. Inclusion of nonmonetary government subsidies such as state-subsidized housing is an especially important advance.

Health Services Section
The health services section contains detailed data on insurance coverage, medical providers, and health facilities that the household might use under selected circumstances. Questions about accessibility, time and travel costs, and perceived quality of care are asked. Information on illnesses and on all uses of the health system during the previous month is collected for children under age 7, for adults between age 20 and 45 in 1989, and from all household members in later years. Questions on immunizations, use of preventive health services, and use of family planning services are also asked.

Individual Survey
Since 2004, all questions that have related to individual activities, lifestyle, health status, marriage and birth history, body shape, and mass media exposure, etc., have been moved to two sets of individual questionnaires: for adults age 18 and older and for children and adolescents under age 18.  Children age 6 and above and all adults are asked to provide their time allocation on household chores and child care, physical activities, and soft drink and sugared fruit drink consumption. Youths age 12 and older and all adults are asked their smoking status; consumption of tea, coffee, and alcohol; diet; and activity level.  Adolescents age 12 and older are asked to provide their conception of their body shape. Both adolescents age 12 and older and women under age 52 married, divorced, remarried, or widowed with children age 6 to 18 in household are asked about mass-media exposure.  Women under age 52 and who are married, divorced, remarried, or widowed are asked to provide their marriage, pregnancy, and birth history and inter-generational linkages to parents and parents-in-law.  Adults age 55 and older are asked to provide their daily living activities and were given a memory test.

Nutrition & Physical Examination
Three days' worth of detailed household food consumption information is collected. In addition, individual dietary intake for three consecutive days is collected for every household member, irrespective of age or relationship to the household head. Adults and children receive detailed physical examinations that included weight, height, arm and head circumference, mid-arm skinfold measurements, and blood pressure (for children age 7 and older and all adults). Limited clinical nutrition and physical functioning data have been collected since 1993. Activities of daily living and related information for older adults and a new set of physical activity and inactivity data have been added for all respondents since 1997.

Community Survey
The community questionnaire (filled out for each of the primary sampling units) collects information from a knowledgeable respondent on community infrastructure (water, transport, electricity, communications, and so on), services (family planning, health facilities, retail outlets), population, prevailing wages, and related variables.

Food Market Survey
In the first two surveys, state and free market data were collected. But by 1997, none of our communities had separate state prices, so only free market stores and large stores have been visited from then onward. In all cases, prices have been collected for a representative basket of commodities.

Health and Family Planning Facility
In 1989, 1991, and 1993, separate visits were made to obtain in-depth data in each community for every identified health service and family planning provider or facility. Information was collected concerning personnel, sources of funds, services available, prices (asked separately for insured and self-pay patients), and distance to the primary sampling units served by the facility. A separate questionnaire was administered to the local family planning official about family planning policies in the community. These were discontinued in 1997, but selected questions have been of these same personnel as part of the community questionnaire since then.

大卫

这个是自由派推的吧，可真能推行麽？

万里风中虎

晨枫 发表于 2014-4-23 12:08
我不是质疑Thomas Piketty或者Nancy Chang的工作，但是在缺乏全民申报制度的情况下，你如何统计真正的收 ...

再回到Picketty和NANCY QIAN用的数据，是Urban Household Income Survey。这个数据库是我一直想要而没有得到的。数据包括1986（1200万观测值）到2003（5000万观测值)的大约3亿五千万的观测值。这是国家统计局的数据，但是一般人看不到，花钱也买不到。

晨枫

万里风中虎 发表于 2014-4-22 23:14
关于福利，包括公司用车房产等是统计的难点，还有农民的耕种收入如何计量等等这都是问题。但是，都有相关 ...

多谢资料。不过这里还是语焉不详啊。比如说，个人收入中，到底是精确到实际收入，还是income bracket？估计是后者，那这bracket有多大？最重要的问题：被抽查者诚实、完整回答的动力是什么？

我们都受到过这样的survey，有些有兴趣填一填，更多的丢到一边不理会了。要仔细分析的话，或许能发现什么样的人更可能回答，什么样的人更可能丢到一边。如何保证这样的bias不成为问题？

很多问题还可能如实回答，比如明年的度假计划、家庭购置大件计划、什么时候想要孩子这些，但每年的收入这样“敏感”的问题，没有切实、具体的动力，很难想象大家会踊跃回答。在这一点上，税务局有独特优势。只要你不犯法，他的数据就是最齐全、最精确的，没有bias。

我是干工程的。老实说，对社科的有些数据来源不大放心。比如那个大跃进中到底死了多少人、现在中国到底有多少基督徒，我对这样的数据是存疑的，因为这都是根据经验模型推断出来的，而经验模型又在很大程度上是根据历史上的经验估计数据而不是真实统计数据估计出来的。理由太简单了：历史上没有这样的真实统计数据。在现实生活中，人们也时常发出对统计数据的疑惑：怎么我和我周围的人都不是这么回事呢？这当然可能是个人的采样空间的偏差，但也可能是统计方法的偏差。事实上，对于中国经济的定量研究经常像医生说红酒对健康有利/不利一样，故事经常要变的，而且每个人的嘴里都不一样。

统计中估计和模型方法我也知道一点，同时也知道一点统计是有很多假设的，比如正态分布和定常过程，问题在于，有人的地方就有江湖，有江湖的地方就有互动，首先这定常过程就是一个很可疑的假设，否则把过去几个月的华尔街数据回归一下，预测明天应该买卖什么股票，人人都发大财了。

还是要多谢老虎的回答。再一次说明：我不是要挑战经济学的定量方法，只是想对于数据的“代表性”有更多认识。

Dracula

晨枫 发表于 2014-4-23 21:31
多谢资料。不过这里还是语焉不详啊。比如说，个人收入中，到底是精确到实际收入，还是income bracket？估 ...

经济学经验研究中使用的数据要比自然科学质量低不少。关于收入分配的研究也不例外。Piketty也很清楚这一点，在他的论文和新出的这本书里反复提到。不过，在他做政策建议，主张提高对富人的税收，增加再分配的时候，似乎就忘了他所依赖的数据不是那么可靠。

你如果对他的研究和收集数据的方法感兴趣的话，可以读一下他的论文。

Income Inequality and Progressive Income Taxation in China and India, 1986-2015

http://piketty.pse.ens.fr/fichie ... yQian2009_AEJPP.pdf

这是其中关于中国数据来源的部分。

The Chinese data used in this paper comes from the urban household income surveys collected by China's National Statistical Bureau (NSB). These surveys are designed so as to representative of urban China. Between 13,000 and 17,000 households are surveyed each year until 2002, where the sample rose to 45,000-50,000 in 2002 and 2003. The micro-files for these surveys are unfortunately not available for all years, and we asked the NSB to provide us with annual, 1986-2003 tabulations based on the micro-files. We asked for two series of tabulations: household tabulations and individual tabulations. Household tabulations report the number of households whose total household income falls into that bracket, their average total income and household size, as well as their average income broken down by income sources (wage income, business income, capital income and transfer income) for a large number of income brackets (and in particular a large number of top income brackets). Individual tabulations report the number of individuals whose individual income falls into that bracket, their average age, years of education, income and household size, as well as their average income broken down by income sources for a large number of income brackets. In practice, some forms of income cannot be properly attributed to a specific individual within the household (this is particularly true for transfer income and capital income). Hence, the total income aggregates reported in household tabulations are larger than in individual tabulations, and various adjustments are necessary when one uses the latter. However, the  important advantage of individual tabulations is that China's income tax applies to individual income (rather than household income).

We used standard Pareto interpolation techniques to approximate the form of the Chinese household and individual distribution of income, and we then used these structural parameters to compute top fractiles’ incomes and to make income tax simulations. The Chinese data appears to be very well approximated by a Pareto distribution (for any given year, Pareto coefficients are extremely stable within the top decile), although there is some presumption that top incomes are underestimated in the survey data. For each year of the 1986-2003 period, we computed income thresholds and average incomes for fractiles P0-90, P90-95, P95-99, P99-99.5, P99.5-99.9 and P99.9-100. Projections for the 2004-
2015 period were made by assuming nominal income trends by fractile similar to those observed during the 1996-2003 period.

We did not attempt to use similar tabulations from rural household surveys. According to the 2000 China Population Census, over 97 percent of households in rural areas are agricultural households, and are hence exempt from income tax. Average rural income was in 2001 more than 3 times smaller than average urban income. So given that our focus is on top incomes and progressive income taxation, the exclusion of rural households should not be too problematic. In fact, our simulated income tax revenues (based solely on urban household surveys) appear to be reasonably close to actual income tax revenues, both in levels and in trends.

总结一下，他用的数据来自国家统计局1986-2003的问卷调查。调查的问题是收入的bracket。然后假设数据遵循Pareto分布，用bracket的数据来估计这个分布。这种数据的问题自然多多，Piketty对此很清楚。在下一篇我给出链接的文章里，他还讨论过为什么他认为税务局提供的数据优于问卷调查，感兴趣的话，你可以读一下。但是研究中国收入分配的变化毕竟是个非常有意思，非常重要的问题，在没有更好的数据之前，他的研究还是会给我们提供不少有用的信息，另外如果导致Survey数据误差的因素，比如non response，measurement error等等如果不随时间变化的话，那么尽管数据的level不那么可靠，数据的trend 还是比较可靠的。当然这是一个很大的假设，社会经济的变化可能导致误差发生变化，下面的那篇文章里还提到survey的方法变化同样可能导致数据发生大的变化，尽管收入分配变化不大。因此，我觉得在这些数据为基础上的政策建议要谨慎一些。

还有一篇文章，可以说是他在收入分配经验研究工作的综述。

Top Incomes in the Long Run of History

http://piketty.pse.ens.fr/fichie ... ez2011_20110913.pdf

里面还有不少篇幅讨论使用税务局数据估计收入分配的局限性，这些不是我的研究领域，但是读来觉得很有意思。

晨枫

Dracula 发表于 2014-4-23 10:49
经济学经验研究中使用的数据要比自然科学质量低不少。关于收入分配的研究也不例外。Piketty也很清楚这一 ...

"在没有更好的数据之前，他的研究还是会给我们提供不少有用的信息"

完全同意。这样的研究肯定是必要的和有价值的，但是对其结论也是要小心对待的。不知道他的分析有没有工程中常有的灵敏度分析，也就是说，如果原始数据有x%的误差，对定性结论有什么影响？

中国有钱人很多对自己的财产和来源很忌讳，还是好奇：这样的自愿性、合作性的survey到底有多准确？可巧研究的重点正是在于高收入人群。

Dracula

晨枫 发表于 2014-4-24 02:36
"在没有更好的数据之前，他的研究还是会给我们提供不少有用的信息"

完全同意。这样的研究肯定是必要的和 ...

他的研究中计算极富人群的收入，不是直接从数据里来的。而是用bracket的数据先估计一个Pareto分布，然后用这个分布计算出来的。Pareto分布是Pareto100多年前发现这个概率分布描述收入分布非常好，据Piketty说，后来的大量研究都证实了这一点。

Survey里，富人肯定会under-report，就是税务局的数据，也肯定有很多合法非法的偷税漏税，（他还提到税法本身随时间变化也是个问题），survey还有一个问题就是non-response，如果non-response rate富人穷人是一样的话，也不太要紧，如果不一样的话，就会带来很大误差。要纵向比较的话，如果这些导致误差的因素不随时间变化，到也关系不大。不过我对这个假设不是多么有信心。横向各个国家比较的话，就是使用税务局的数据我觉得也不是特别靠谱。象欧洲国家，北欧国家偷税漏税的比例很低，而南欧国家意大利，希腊等各个阶层都对偷税漏税习以为常。Piketty的看法是，偷税漏税的问题很重要，但是按他的估计其magnitude不足以改变其主要结论。

thacaos

Dracula 发表于 2014-4-24 03:36
他的研究中计算极富人群的收入，不是直接从数据里来的。而是用bracket的数据先估计一个Pareto分布，然后 ...

这个Pareto分布不知道在中国有没有验证过？从个人经验看，中国与西方国家对露富的态度应该有很大差异。

晨枫

Dracula 发表于 2014-4-23 13:36
他的研究中计算极富人群的收入，不是直接从数据里来的。而是用bracket的数据先估计一个Pareto分布，然后 ...

如果富人underreporting导致他的结论要对富人征税更多，这个误差倒是也不大重要。非法漏税本来就是一个没法统计的事情，要不然也不会一直非法下去，而是早就判刑、罚款了。还是那句老话，我对他的研究和结论不质疑，只是想更多了解他的数据。

他的结论是对富人猛征税可以提高国家的稅入和社会公平？他的富人是什么标准？

小木

方法我倒是不怀疑。帕托累分布大概也是对收入最好用的分布了。
09年的文章，使用的86-03年的数据，这个数据滞后性有点厉害。尤其在中国，03年还是上一波繁荣的前夜。这个研究提供的信息拿来套现在的情况，是要非常小心的。