.png){kind=icon}
.png){kind=icon}
Return
Panel models with interactive effects
.png){kind=icon}
PRE
AI
DOI：10.1016/j.jeconom.2018.06.017.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The multiplication of individual specific effects, (lambda) under tilde (i), and time-specific effects, f(similar to t), (lambda) under tilde (i)f(similar to t), provides a more general formulation than the traditionally used additive form to capture the unobserved heterogeneity in panel data modeling. It is also a useful approach for dimension reduction for modeling cross-section dependence. However, (lambda) under tilde (i) and f(similar to t) are unobservable. We explore the implications for econometric modeling under various formulations of the interactive effects models and suggest a quasi-likelihood approach as a common framework to study issues of estimation and statistical inference when regressors are either strictly exogenous or predetermined and under different combinations of the data size of cross-sectional dimension, N, and time series dimensions, T. We also suggest some computationally simpler estimation methods in light of the quasi-likelihood approach. Monte Carlo studies are conducted to highlight the issues involved. (C) 2018 Elsevier B.V. All rights reserved.
Keywords:
Interactive effects
Static and dynamic models
Initial observations
Asymptotic bias
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
4
Papers:
5.2K
Citations:
3.0W