IV Estimation of Heterogeneous Spatial Dynamic Panel Models with Interactive Effects

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Jia, Cui, Guowei, Sarafidis, Vasilis, Yamagata, Takashi
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910805972221952
author Chen, Jia
Cui, Guowei
Sarafidis, Vasilis
Yamagata, Takashi
author_facet Chen, Jia
Cui, Guowei
Sarafidis, Vasilis
Yamagata, Takashi
contents This paper develops a Mean Group Instrumental Variables (MGIV) estimator for spatial dynamic panel data models with interactive effects, under large N and T asymptotics. Unlike existing approaches that typically impose slope-parameter homogeneity, MGIV accommodates cross-sectional heterogeneity in slope coefficients. The proposed estimator is linear, making it computationally efficient and robust. Furthermore, it avoids the incidental parameters problem, enabling asymptotically valid inferences without requiring bias correction. The Monte Carlo experiments indicate strong finite-sample performance of the MGIV estimator across various sample sizes and parameter configurations. The practical utility of the estimator is illustrated through an application to regional economic growth in Europe. By explicitly incorporating heterogeneity, our approach provides fresh insights into the determinants of regional growth, underscoring the critical roles of spatial and temporal dependencies.
format Preprint
id arxiv_https___arxiv_org_abs_2501_18467
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle IV Estimation of Heterogeneous Spatial Dynamic Panel Models with Interactive Effects
Chen, Jia
Cui, Guowei
Sarafidis, Vasilis
Yamagata, Takashi
Econometrics
This paper develops a Mean Group Instrumental Variables (MGIV) estimator for spatial dynamic panel data models with interactive effects, under large N and T asymptotics. Unlike existing approaches that typically impose slope-parameter homogeneity, MGIV accommodates cross-sectional heterogeneity in slope coefficients. The proposed estimator is linear, making it computationally efficient and robust. Furthermore, it avoids the incidental parameters problem, enabling asymptotically valid inferences without requiring bias correction. The Monte Carlo experiments indicate strong finite-sample performance of the MGIV estimator across various sample sizes and parameter configurations. The practical utility of the estimator is illustrated through an application to regional economic growth in Europe. By explicitly incorporating heterogeneity, our approach provides fresh insights into the determinants of regional growth, underscoring the critical roles of spatial and temporal dependencies.
title IV Estimation of Heterogeneous Spatial Dynamic Panel Models with Interactive Effects
topic Econometrics
url https://arxiv.org/abs/2501.18467