Econometric Modeling for Network Formation and Interactions

Abstract

In this seminar, I will review several of my previous studies on network formation and interactions. In econometric literature, the spatial autoregressive (SAR) model is frequently used as a foundational tool for analyzing the effects of network interactions. However, when network formation is endogenous, SAR models can produce biased estimates. To address this issue, I propose several modeling approaches that integrate the processes of network formation and interaction. These models are estimated using Bayesian MCMC algorithms. Both Monte Carlo simulations and empirical results demonstrate that these approaches effectively mitigate estimation biases arising from network endogeneity. I will also discuss current research challenges and open questions in this area. 

線上視訊請點選連結

附件下載