收入的尺度分布一直是认识经济规律的有效渠道。目前，国内外的学者在收入分布的实证研究和理论研究方面都做了很多工作。收入分布实证研究的突出特点是尽管研究对象不同，但收入分布常常表现出相似的分布特征。即收入的高端服从幂律分布，中低端服从对数正态分布或Gamma分布。收入分布的理论研究表明人类活动的某些共性是相同分布形式的主因，掌握了这种规律就能预测经济的发展，进而控制发展中的不利因素。但相比较国外丰富的研究成果，我国的收入分布工作由于数据缺乏等原因，相对研究较少。本工作中我们利用中国家户收入调查（China Household Income Projects，CHIPS）得到的1998年至2002年中国个人收入数据分析了中国收入分布。实证工作表明，中国的收入分布也呈现出高端幂律分布、中低端对数正态分布的分布特征。该分布形式已经与成熟市场经济国家的收入分布形式非常相似。并且，将该收入分布与前期工作结果相比较，我们发现改革开放后，中国的收入分布经历一个从一点式分布（决定平均）到对数正态分布，再到中低端对数正态分布高端幂律分布的演化过程。中国的收入分布形式可以作为我国市场经济深化程度的一个客观反映。在得到中国收入分布形式之后，我们进一步探讨该分布形式的形成机制。通常，对数正态分布的产生机制是一种收入的单期增长率服从独立同分布的累积随机过程（Mulitiplier Stochastic Process，MSP），即Gibrat模型。但实证结果中的幂律尾的出现说明了Gibrat模型的假设存在问题。因此，为了检验Gibrat假设在实际过程中是否成立，也为了得到收入增长过程的真实特征，我们利用同样的数据分析了中国收入增长率分布的分布形式。实证工作表明，收入增长率不是独立同分布的，而是随着收入初始水平的不同而变化的。收入增长率的方差随着收入水平的增长先减小，再保持较小值不变，最后在收入达到某个界限后迅速增大。通过与金融学中的资本资产定价模型（Capital Asset Pricing Model，CAPM）对比，我们解释了该收入增长率分布特征的经济意义。高端收入的增长率方差很大，表明该部分群体主要依靠资本投资获得收入，中低收入群体的收入增长率方差小，表明该部分群体主要依靠劳动获得收入。我国收入不平等程度逐渐恶化的原因，是资本投资的回报率过大，而劳动收入的回报率过低。因此，调整资本和劳动的收入报酬比率是解决我国收入不平等问题的关键。进而，我们在累积随机过程模型的框架下设计了符合收入增长规律的收入模型。仿真结果印证了实证分布形式。对于研究收入分配问题的基础——收入分布，其研究初衷是收入的不平等程度。对于尺度分布成因研究的新方向——增长率分布，它的本质是流动性。所以应该将流动性和收入不平等程度这两个最基本的经济研究对象直接建立其联系，进行研究。对流动性的讨论不但可以直接揭示收入不平等现状的成因，还可以解决连接在两个问题中的收入分布成因、增长率分布特性等一系列经济问题，具有巨大的理论研究意义。因此，在本文的最后一部分我们详细了收入不平等和收入流动性的定义、研究历史及相关的实证工作，并讨论了收入分布和收入流动性这两个指标的关系。我们从绝对收入流动性矩阵和收入不平等程度测度Gini系数的定义出发，推导了Gini系数在绝对收入流动性矩阵中对应的表达式，分析了收入流动性矩阵对不平等程度的作用途径，进而构造了对应于收入不平等程度变化的新的收入流动性测度。同时，利用1998-2002年的收入面板数据验证了该理论分析。实证分析结果表明，我国收入不平等程度增大的同时代表机会平等的收入流动性也在持续增长。由于向收入高低两端的流动增长较快，并在势头上超过了向中等收入水平的流动，整体收入不平等程度恶化了。该现象表明在我国的税收和福利等政策对于收入不平等的二次调节虽然发挥了一定的作用，但其力度不够。为了停止或逆转收入不平等程度继续升高的趋势，我们需要加大这些政策的调节力度。

Income distribution is an effective way to get to know the economics laws. Up to now, there is a lot of empirical works and theoretical works have been done by researchers. The conspicuous character of empirical works is that income distributions of different countries or regions have a uniform distribution, i.e. Log-normal distribution in the low-middle part and Power law distribution in the higher part. While, the theoretical works indicate that some common feature of human race is the main reason of uniform income distribution. If we could understand the reason, we can have a well prediction of the future and avoid the disadvantage factor in our developing process.But compared with luxuriant works of income distributions in other countries, China’s income distribution has been rarely studied due to the lack of data. In this paper with the income data from China Household Income Projects (CHIPS) during the years 1998-2002, we examine the functional form of Chinese income distribution. The fitting results show that Chinese income distribution is a log-normal function with a remarkable power-law tail. This functional form indicates that income distribution of China has changed a lot from its appearance in the early stage of China’s reform and becomes very close to that of developed countries. This fact demonstrates the profile of income distribution can be seems as an index of the degree of China’s marketizing process. In the next step, we try to analyze the generation mechanism of income distribution. In general, the Log-normal distribution are generated by a multiplier stochastic process of independent isolated distributed income growth rate, as described by Gibrat model. But the emergence of Power law tail in Lognormal distribution indicated the Gibrat model failed to explain the real situation. In order to test the hypothesis of Gibrat’s model, we examine the features of income growth rate distribution. With income data from China Household Income Projects (CHIPS) in 1998-2002, we study the features of income growth rate distribution. The fitting results suggest that the assumption of Gibrat’s model that growth rate is independent of initial incomes only holds in medium income groups. The standard deviation of income growth rate decrease with income in lower income groups and increase with income in higher income groups. Especially these higher income groups cluster exactly in the region of power law tail of income distribution. Considering standard deviation is related to people’s risk aversion in the story of capital asset pricing model (CAPM) in Finance, diverse risk aversions are introduced to explain the differentiation of growth rate’s standard deviation. People get the income from asset investment usually has a great standard deviation of income growth rate. While people get the income from working has a small and stable income growth rate. It further give the reason of income inequality, which is the input-output ratio of asset investment is unreasonable higher than labor payment. Furthermore, a multi-industries model is proposed to explain the current functional form of Chinese income distribution. The simulation results has reproduced the empirical income distribution, which demonstrates our model can successfully explain why there are power-law distribution and log-normal distribution coexisted in income distribution.The aim of income distribution research is to explain the reason of income inequality, and the essence of income growth distribution is income mobility. Therefore, we should put the two indices together, and find the relationship between them. In the last part of this paper, we studied the effects of income mobility on income inequality. In the first step, an expression of Gini index are derived from the absolute income mobility matrix based on an analysis of their definitions, by which we get the idea what kind of income dynamics can improve the income inequality. And as a result, a new income mobility index with respect to income inequality is built by us. All these theoretical analysis are tested with the personal income data in the period of 1998-2002. From the empirical results, it is very gratifying to find there is a consecutive growing trend of income mobility behind the large income inequality, which indicates a promotion of equal opportunity. But since the flows to the higher and lower income levels have been over that to the middle income level, income inequality grows up. Thus it is important to make an adjustment on tax and welfare policies which have the strong effects on increasing trend of income flow to extremes, accompany with the equal opportunity and free competition promoting in the process of our economic reform.