The Selection Of Chinese Residents' Income Distribution Function And The Estimation Of Income Inequality

Since the reform and opening up, China's economy has shown rapid growth. However, with the rapid economic growth, the income gap between urban and rural areas of the country showing a trend of expansion, polarization has become increasingly serious, causing severe income inequality. Income inequality is a root cause of other inequality, unequal presence also contributed to many social, economic and even political course problem, extremely unfair distribution of income countries often lack cohesion countries, so the question of income distribution is becoming China in the development process problems to be solved.In order to solve these problems efectively, we must first recognize the income inequality.For the study of income inequality, domestic scholars studied more inequality decomposition and calculate point estimation,but for the distribution of income in the form of specific parameters and interval estimation is relatively small to research.This paper based on the Family Nutrition and Health Survey (CHNS) years of micro survey data, studied the dynamic changes of the rural and urban income distributions,and the selection of parameters income distributions,and the interval estimation of our country's Gini coefficient.Firstly, this paper introduced the related properties of Gini coefficient,include the relationship between Gini coefficient and welfare functions,the relationship between Gini coefficient and Lorenz curve,the calculation of Gini coefficient based on continuous or discrete data,and the statistical inference methods of Gini coefficient.The following,this article based on Chinese Nutrition and Family Health Survey (CHNS) data separately for rural and urban survey data were statistically analyzed descriptive.Then this paper used the method of non-parametric kernel density estimation,depicts the dynamic changes from 1989 to 2011 of the Income distribution curve.Then,the article based on microscopic CHNS survey data, discussed the income distribution function suitable for our country.About in the choice of the distribution function,due to the different income data, most also have a distribution function.Therefore, the article was also used to select a single fitting parameter exponential distribution,two parameters lognorm distribution,weibull distribution,gamma distribution, three parameters Burr(SM) distribution,beta? distribution, and four parameters GB2 ditribution..The this paper used maximum likelihood estimation to estimate the parameters,and used a sum of squared error (SSE) as the criteria. The results found that they are no great difference,and further illustrated that using GB2 distribution to measure of our country is appropriate.This paper also given a brief overview about using income distribution function to discuss inequality problem.The sixth chapter of the article discussed the statistical inference and interval estimation of Gini coefficient. This paper use three methods to monte carlo simulation, Simulation results showed that using jackknife to calculate the standard deviation and mean square error is minimum.However, the calculation process is very complicated, it takes a lot of time. Next, the article still based on the CHNS survey data, calculated the standard error of Gini coefficient of rural and urban, As well as the confidence interval of the Gini coefficient 95% confidence.In general,this paper based on the data of China Nutrition and Health Survey of family.Firstly,we analyzed dynamic characteristics of chinese income distribution from 1989 to 2011,then,we use the maximum likelihood estimation to estimate the our country's income indistribution based on individual data,and the last,we calculated the interval estimation of Gini coefficient.