Simulation And Application Of Threshold Point Inspection Method For Spatial Panel Data

The development of China’s regional economy shows obvious characteristics of convergence,that is,the characteristics of "one glory and one glory,one loss and one loss" usually appear among the economies in the region.Based on such an economic fact,when we analyze the economic data of multiple regions,if we only consider the influencing factors of the region itself,the analysis results will often be biased.The newly emerging spatial econometric theory in recent years is based on the consideration of the above-mentioned situation and is based on the econometric theory.At the same time,in the process of empirical economics,we often encounter economic variables with structural change points,such as GDP.The change of these variables is often not purely linear,but there is a structural change point.When the variable is greater than a certain value,its growth rate will change significantly.Threshold regression model is one of the most widely used inspection research methods for structural changes in practice.This article attempts to add spatial influence factors to the panel threshold model,establish an exploratory panel threshold model,and use this model to analyze realistic economic problems.The article is divided into six parts: The first chapter is the introduction.This chapter reviews the current development of spatial econometrics and traditional threshold models,raises research questions,and analyzes the significance of the research.Then according to the research questions,determine the goals and content of the research.Finally,the research methods and innovations of this article are clarified,and the technical route and structure of this article are summarized.The second chapter is a literature review.It gives a detailed overview of the theory and development of spatial econometric models.It mainly includes the theoretical basis of spatial econometrics,the types of spatial panel static data models,and the estimation methods of various models.Finally,the construction ideas and estimation methods of traditional threshold models are introduced.The third chapter is model construction.In this chapter,based on the theoretical foundation described in Chapter 2,a detailed analysis of the threshold model measurement theory of spatial panel data is made,and two models of single threshold and double threshold are constructed respectively.Finally,the feasibility of the theory is tested using simulated data.The fourth chapter is the empirical part.Combined with the model constructed in Chapter 3 to test the actual data,explore the threshold effect of China’s R & D investment on economic development and explain the model results.The fifth chapter is the last part of this article,which summarizes the full text and points out the conclusions and deficiencies.From the perspective of the model,the spatial panel threshold regression model constructed in this paper has achieved satisfactory results in both simulation data and empirical data analysis.From an empirical point of view,the following conclusions are reached through quantitative and qualitative analysis:(1)The impact of R & D personnel on economic development is very small,and R & D personnel input has little effect on per capita GDP before the threshold is reached;when the threshold is reached After that,the negative impact of R & D personnel investment on per capita GDP will increase to a certain significance,but the impact will still be small.(2)R & D capital investment always has a positive impact on the growth of per capita GDP,but the impact will be reduced to a certain extent after reaching the threshold.(3)From the perspective of spatial effects,the spatial spillover or lag effects of all models are significant,so we need to consider multiple impact paths when considering the relationship between R & D input and economic growth.