Bayesian analysis of Spatial Panel Autoregressive Models with time-varying Endogenous Spatial Weights Matrices and Common Factors

Speaker: Xiaoyi Han
Speaker Intro:

Assistant Professor in WISE

Homepage: hanxiaoyi.weebly.com/

Host:
Description: Abstract: This paper examines the specification and estimation of spatial panel autoregressive (SAR) models with dynamic,time-varying endogenous spatial weights matrices and common factors.  Motivated  by the spillover effects of  state  Medicaid  spending  on  welfare  programs,  we  combine  the  features  of  endogenous time-varying   weights  matrices  and  common  factors  for  the  first  time  in  the  SAR  panel  models.  In this particular  application,  endogeneity of the spatial weights matrices  comes  from the correlation of “economic distance” and the disturbances in the SAR equation.  Common  factors  are introduced to control for common shocks to all states  and  factor  loadings  may  capture  heterogeneity in states’ responses. For the estimation, the Bayesian MCMC method is  developed.  Identification  of factors and factor loadings, and the corresponding model  selection  issues  based upon   the Bayes factor and the deviance information criterion (DIC)  are  also  explored.
Time: 2015-03-18(Wednesday)16:30-18:00
Venue: Room N303 Economic Buildings
Organizer: WISE - SOE

关闭