QMLE for Unbalanced Dynamic Network Panel Data
Published:Dec 31, 2025 09:47
•1 min read
•ArXiv
Analysis
This paper addresses the challenge of estimating dynamic network panel data models when the panel is unbalanced (i.e., not all units are observed for the same time periods). This is a common issue in real-world datasets. The paper proposes a quasi-maximum likelihood estimator (QMLE) and a bias-corrected version to address this, providing theoretical guarantees (consistency, asymptotic distribution) and demonstrating its performance through simulations and an empirical application to Airbnb listings. The focus on unbalanced data and the bias correction are significant contributions.
Key Takeaways
- •Addresses the problem of unbalanced panel data in dynamic network models.
- •Proposes a QMLE and a bias-corrected estimator.
- •Provides theoretical guarantees (consistency, asymptotic distribution).
- •Demonstrates performance through simulations and an empirical application to Airbnb data.
Reference
“The paper establishes the consistency of the QMLE and derives its asymptotic distribution, and proposes a bias-corrected estimator.”