Computation And Empirical Research On Precautionary Savings

For a long time, China’s economic development shows a high savings rate and low consumption rate in parallel, and the growth rate of economic and income level is lower than savings. Economics scholars generally believe that the reason for this phenomenon is that people have too strong precautionary savings motive. According to the precautionary savings theory, it is due to the uncertainty of income and consumption in the future. Although the uncertainty in the theory has been extending to the terms of the length of life, medical expenses, etc., but in the final analysis, the presence of the only factor of the precautionary savings is uncertainty in the theory. Therefore, based on the uncertainty, the theoretical analysis and the empirical research of the precautionary savings is particularly important.In this paper, we make the empirical research according to ask questions, analyze and solve problems, and follow the guideline of closely integrated of the qualitative analysis and the quantitative analysis. The core of the paper is to conduct empirical research of Buffer-Stock Model by using kernel density estimation method and application the software MATLAB2010b to estimated the uncertainty of various expenses. To seek influencing factors of the precautionary savings, we regard the per capita amount of household savings as the explained variable, and regard the per capita amount of payment on education, foodstuff, clothing, housing, household equipment and supplies, traffic and the communication, entertainment, medical treatment and other as the explaining variable. According to the data of the 31 provinces, municipalities and autonomous except Hong Kong, Macao and Taiwan, We conducted an empirical study by establishing the panel data model and taking advantage of Buffer-Stock Model.The main innovation and found in this paper:when studying the precautionary savings, precautionary buffer savings model is superior to multiple linear regression model; The description of uncertainty by Kernel density estimation method is more accurate; housing expenses, medical expenses and education spending remains one of the important factors affecting our residents precautionary savings; The impact of food expenditure and clothing expenditure is relatively small on the residents of precautionary savings, and other expenses are not significant.