我国居民收入分布的问题研究

发布时间：2017-12-30 21:13

本文关键词：我国居民收入分布的问题研究　出处：《安徽大学》2016年硕士论文　论文类型：学位论文

【摘要】：本文利用中国健康和营养调查数据库(CHNS)1988—2010年的调查数据,首先对我国城乡居民收入的分布特征、不同分位点处收入的变化以及收入不平等问题做简单的统计描述,统计结果表明我国居民收入分布呈右偏态,不同分位点处收入的差异明显。在以往收入分布的研究中,学者一般预先设定收入的分布形式,然后采用参数估计的方法估计未知参数,而这种设定有很多局限性,研究发现,无需预先设定具体分布形式的非参数核密度估计方法对收入分布的研究更为适合。实证分析显示,1988-2010年间,不同分位点处城乡居民收入的变化趋势不同,分位点越高,收入的增长速度越快,这说明我国居民收入差距在不断扩大,收入分配领域出现“穷者愈穷,富者愈富”的“马太效应”。然后,本文将分位数回归方法应用到我国居民收入分布的研究中,以2010年的收入数据为例,构建线性分位数回归模型,然后讨论性别、受教育程度、城乡户籍、年龄、地区因素对收入的影响,重点讨论了同一影响因素在不同分位点处对居民个人收入影响的差异性及其出现的原因。较为全面的论证性别、年龄、受教育程度、城乡户籍和地区等因素对收入的影响机制及其变化规律。本文的研究对我国现阶段的收入分布格局以及收入差距的研究有一定的借鉴意义。本文主要分三个部分来进行论述。第一部分,先对现阶段我国居民的收入分配现状进行概括,以期对近年来居民的收入分布状况有宏观印象。第二部分,将参数估计方法与非参数核密度估计方法进行比较,找出更适合研究居民收入形态的方法,并利用非参数核密度估计方法对我国城乡居民收入大致的分布形状、不同分位点处收入的变化进行研究,得出居民收入分布右偏以及“两头大,中间小”的收入极化趋势。第三部分,对比线性均值模型,本文论证分位数回归模型对收入分布研究的适用性,对可能影响居民收入的年龄、受教育程度、性别、地区和城乡因素进行描述,然后利用这些变量建立分位数回归模型,论证影响因素的显著性和在不同收入分位点处影响的差异性。最后,针对各个影响因素在不同分位点处对收入的影响程度不同,有针对性的提出一些增加居民收入的建议。
[Abstract]:Based on the survey data of China Health and Nutrition Survey (CHNSN) from 1988 to 2010, the distribution characteristics of urban and rural residents' income in China were first analyzed in this paper. The change of income at different loci and income inequality are simply described. The statistical results show that the income distribution of Chinese residents is right skewed. In the previous research of income distribution, scholars usually set the distribution form of income in advance, and then use the method of parameter estimation to estimate the unknown parameters. However, this setting has many limitations. It is found that the non-parametric kernel density estimation method is more suitable for the study of income distribution without pre-setting the specific distribution form. Empirical analysis shows that this method is more suitable for the study of income distribution. From 1988 to 2010, the change trend of income of urban and rural residents at different loci is different. The higher the locus is, the faster the income growth rate is, which indicates that the income gap of residents in our country is constantly expanding. In the field of income distribution, "the poorer the poor, the richer the rich" "Matthew effect". Then, the quantile regression method is applied to the study of income distribution in China, taking the income data of 2010 as an example. A linear quantile regression model was constructed, and then the effects of gender, education level, urban and rural household registration, age, and regional factors on income were discussed. This paper mainly discusses the difference of the influence of the same influencing factors on the residents' personal income at different loci and the reasons for their emergence. The author fully demonstrates the gender, age, and education level. The influence mechanism of urban and rural household registration and region on income and its changing law. The research in this paper has certain reference significance for the study of income distribution pattern and income gap in our country at present. This paper is mainly divided into three parts. The first part. Firstly, the present situation of income distribution of residents in China is summarized in order to have a macroscopic impression on the income distribution of residents in recent years. The second part, the parameter estimation method and the non-parametric kernel density estimation method are compared. Find out a more suitable way to study the income patterns of residents, and use the non-parametric kernel density estimation method to study the distribution shape of income of urban and rural residents in China, the change of income at different loci. In the third part, compared with the linear mean model, this paper demonstrates the applicability of the quantile regression model to the study of income distribution. The factors of age, education level, sex, region and urban and rural areas that may affect the income of residents were described, and then the quantile regression model was established by using these variables. Demonstrate the significance of influencing factors and the difference in the impact of different income loci. Finally, the impact of different factors on income at different loci is different. A number of targeted suggestions to increase residents' income.
【学位授予单位】：安徽大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：F224;F126.2

【参考文献】