A0864
Title: Spatial panel data models with time-varying network structures and multi-dimensional fixed effects
Authors: Xiaoyu Meng - Nankai University (China)
Zhenlin Yang - Singapore Management University (Singapore) [presenting]
Abstract: Specification and estimation of spatial panel data models with time-varying network structures and multi-dimensional fixed effects (FEs) are considered. The former capture the (time-varying) endogenous, contextual and correlated social interaction effects, and the latter control the unobserved group, individual, and time heterogeneities. A general strategy is proposed for the identification of common parameters as well as group, individual and time FEs. A simple and general method is introduced for model estimation and inference, where the concentrated quasi-scores (with group, individual and time effects being concentrated out) are adjusted to account for the effect of estimating the incidental FEs. Consistency and asymptotic normality of the proposed estimators are established. Monte Carlo results show excellent finite sample performance of the proposed methods.
