Empirical Research Based On The Bootstrap Method For Panel Data Of The Total Retail Sales Of Social Consumer Goods In Guangdong

Total retail sales of social consumer goods is an important indicator to reflect the degreeof consumption and socio-economic climate, there are certain reference value on theeconomic development of Guangdong to study the indicator. The panel data model canincrease the capacity of the samples, will help to properly analyze the relationship betweeneconomic variables. View of the above advantages, we expand the bootstrap method used inGranger causality test to the Panel data fixed effects model，and study on total retail sales ofsocial consumer goods of Guangdong. Built in the original bootstrap used in Grangercausality test, we construct a panel data test statistic, and discuss the method to determine thepanel data model.we use unit root tests to determine the data stationarity,judge panel data model by thediscussed method, determine statistic for the hypothesis test’s feasibility by chi-square test,and then use the value of LR sequences to build the confidence interval to do Grangercausality test. Get results: First, Guangdong’s total retail sales and gross domestic product andpromote each other. Second, secondary industry and tertiary industry of Guangdongdevelopment Branch to promote the growth of total retail sales of social consumer goods;Third, Guangdong urbanization is conducive to the growth of total retail sales of socialconsumer goods; Fourth, rural residents and urban residents in low per capita income is amajor factor restricting the growth of domestic demand, the proportion of income is asignificant factor of the total retail sales of social consumer goods; Fifth, the total fixed assetinvestment can promote the growth of total retail sales of social consumer goods. And view ofthe advantages of panel data model with the bootstrap, the conclusions obtained using thismethod is more reliable and more reasonable than the time series model.