| Spatial Econometrics is an important modern econometrics branch produced many important theoretical and practical results. Spatial econometric theory introduced the pattern of economic phenomena influenced by spatial factors into the model by spatial weights matrix. The paper’s goal is to learn and understand the basic theory of spatial Econometrics and basic model, estimation methods, focusing on spatial data having a nested structure, combined with spatial models and multilevel linear model to analyze the issues of the spatial correlation in different levels.This paper introduces the theory of spatial econometric model, including spatial autoregressive model, spatial error model and spatial Durbin model, spatial weight matrix defining,the theory of Bayesian method to estimate model parameters, and a computer step estimation process. For the problem at different levels of spatial correlation of spatial multilevel data, we defined a plurality of spatial weight matrix to create spatial auto-regression model with plural spatial weights matrix and spatial hierarchical linear models, and analyzed the principles and process of Bayesian method to estimate the parameter of these models, and the estimated amount of properties.This article reveals the differences between spatial multilevel linear model and classical spatial regression model in theory, which reflects the theoretical innovation.The innovation is mainly reflected in: the classical spatial regression model in the analysis of spatial dependence, contains only a level of spatial dependence, this paper based on the detailed interpretation of spatial econometric model, theoretically studied at different levels of spatial dependence problem, reflecting the significance of theoretical innovation spatial hierarchical linear model, in the sense that reveals the different levels influence of spatial correlation of economic phenomena, presents significant applied innovation. |
PDF Full Text Request