Application Of Quantile Regression Method In Residents’ Income

The income of residents is an important indicator to measure the living standards of residents.It is generally calculated on a household basis.It is not only a basic indicator for measuring the living standards of residents in different countries and regions,but also in macroeconomic theory and empirical research.It has a very important position and plays an important role in the formulation of national macroeconomic policies.Based on the survey data in the CHNS database,this paper firstly makes a simple statistical description on the per capita income of households in China from 1988 to 2014,including the distribution pattern of residents’ income,the change of each quantile and the description of income disparity.But we find that the traditional method of normal distribution assumption is not in conformity with the actual situation.Ordinary least-squares regression to investigate the condition of mean reversion also is not the result of what we want.Naturally quantile regression is considered to be an appropriate research method.Then,based on the data of personal income in 2014,we summarize the nonparametric and semiparametric theory and quantile regression analysis technology.Firstly,the mean linear regression model was established as a reference,and then the quantile regression model,non-parametric quantile regression model and semi-parametric quantile regression model were constructed successively.Finally,we obtain the most suitable regression model for the income distribution of residents in China.