For linear panel data models with fixed effects, cluster-robust covariance estimation does not use variability over time. The extant heteroskedasticity-robust methods available under strict exogeneity do not generalize to dynamic models. We propose novel robust covariance estimators under a strong version of serial uncorrelatedness, where serial uncorrelatedness is required to identify dynamic panel models. Asymptotics are established, and simulations verify theoretical findings. The estimator can apply to the popular dynamic IV-GMM estimators and be a sharper alternative for cluster-robust covariance estimators in panel data models with limited cross-sectional information.
Bibliographical noteFunding Information:
The authors thank the Editor and two anonymous referees for helpful comments. This work was supported by the Institute of Information & Communications Technology Planning & Evaluation (IITP) Grant funded by the Korean Government (MSIT) (No. 2022‐0‐00302, Development of global demand forecasting and analysis/prediction system of market/industry trends).
© 2023 Oxford University and John Wiley & Sons Ltd.
ASJC Scopus subject areas
- Statistics and Probability
- Social Sciences (miscellaneous)
- Economics and Econometrics
- Statistics, Probability and Uncertainty