Research On A Spatially Variable Coefficient Panel Data Model And Its Application In The Analysis Of China's Population Mortality Rat

In recent years,the study of spatial panel data model has received extensive attention.The spatial effect is mainly described in the form of spatial autocorrelation,and the results of spatial heterogeneity are few.On the other hand,geographical weighted regression model has been widely used as a semi-parametric model that can effectively analyze spatial heterogeneity,but it is mainly used in cross-section data,and the analysis results of panel data are few.To solve the above problems,this paper mainly proposes a kind of fixed time effect semi-parameter spatial variable coefficient panel data model,and applies this model to the analysis of provincial mortality.The thesis mainly includes the following two parts of work.Firstly,based on the traditional panel data model of individual and time fixed effects,in order to depict spatial heterogeneity,coefficient terms of explanatory variables and individual effect terms are treated as spatial variable coefficients related to spatial geographic location,while retaining the setting of fixed time effects.For this model,the Profile least squares estimation method is used to estimate the model,and the performance of the estimation method is investigated by numerical simulation.In addition,the problem of hypothesis testing and a kind of generalized model are also discussed.Secondly,the paper studies the temporal and spatial variation of provincial mortality,and uses the proposed model to analyze the influencing factors.Firstly,the Moran Index and Thiel index are used to make an exploratory spatial analysis of the inter-provincial mortality from the characteristics of spatial distribution,temporal evolution and regional differences.Then,the common panel data model and the semi-parametric spatial panel data model proposed in this paper were respectively used to model the mortality and its influencing factors.Based on the modeling results,the influencing factors of mortality were analyzed in depth,and specific suggestions were put forward for different regions.The model proposed in this paper belongs to a class of semi-parametric spatial model,which is the extension of the existing spatial panel data model and the geographical weighted regression model.It broadens the scope of the spatial econometrics model,and has certain theoretical research significance and application value.This paper analyzes the mortality of provincial areas deeply and provides scientific basis for understanding the spatio-temporal changes of regional mortality and formulating relevant policies and measures.