It Seems Not Relevant To The Study Of Spatial Econometric Models

In recent years, spatio-temporal data analysis and modeling have been attracted more and more attention of statisticians and econometrics, many types models have been proposed. At present, the spatial panel data model is the most commonly used to fit spatio-temporal data, but time heterogeneity has not been fully considered. To describe the temporal heterogeneity and spatial effect simultaneously, some spatial seemingly unrelated regression models have been proposed, and spatial effect was modeled by spatial lag or spatial error autoregression. For cros-sectional data, spatial error component was an alternative way to describe spatial effect and has its advantage. There is no research on combining seemingly unrelated regression model and spatial error component method. To solve this problem, two new types spatial econometric models were proposed and their statistical inference were investigated.Based on the spatial error component method and seemingly unrelated model, spatial error component seemingly unrelated regression model was proposed, A estimation method is introduced to estimate the unknown parameters of the model. Some simulations are conducted to examine the performance of the proposed procedures and the results show that the proposed methods have good finite sample properties.To capture the underlying relationships between response variables and their associated covariates, partially linear seemingly unrelated regressions with spatial error component was proposed, and the profile least square estimation was applied to estimate the unknown parameters and nonparametric components, the results of some simulations show that the proposed methods have good finite sample properties.To explore spatio-temporal data, two new spatial econometric models were proposed and studied, The results will extend the field of spatial econometrics, and provides statistical methods for real data analysis.