Maximum likelihood-based estimation of dynamic panel data models

مرکز و کتابخانه مطالعات اسلامی به زبان های اروپایی

منو

" Maximum likelihood-based estimation of dynamic panel data models "
Gareth Thomas M. Ahn

Document Type	:	Latin Dissertation
Language of Document	:	English
Record Number	:	55156
Doc. No	:	TL25110
Call number	:	‭3178276‬
Main Entry	:	Gareth Thomas
Title & Author	:	Maximum likelihood-based estimation of dynamic panel data models\ Gareth Thomas
College	:	Arizona State University
Date	:	2005
Degree	:	Ph.D.
student score	:	2005
Page No	:	134-134 p.
Abstract	:	This work studies various maximum likelihood (ML) estimators of dynamic panel data models with a large number of cross-sectional units and a small number of repeated time-series observations for each cross-sectional unit. Investigation of several different existing ML estimators for data with independent and identically distributed error terms is undertaken and their properties examined. It is found that when data follow unit-root processes, the ML estimators have singular information matrices. This is not a non-identification problem because the ML estimators are still consistent. Nonetheless, the estimators have nonstandard asymptotic distributions and their convergence rates are lower than root N. For this reason, the sizes of the Wald unit-root tests are severely distorted even asymptotically, and they reject the unit-root hypothesis too often. However, likelihood ratio (LR) tests for unit root follow mixtures of chi-square distributions. Monte Carlo experiments show that the LR tests are much better sized than the Wald tests, although they tend to slightly over-reject the unit root hypothesis in small samples. It is also shown that the LR tests for unit roots have good finite-sample power properties. A ML estimator of a panel data model with either heteroskedastic error terms or moving average of order one (MA(1)) is also examined. Under these error term structures the information matrix of the ML estimator is in general no longer singular when the data follow unit root processes. Asymptotically Wald unit-root tests are nondistorted and likelihood ratio tests for unit root follow standard distributions. The ML estimator for heteroskedastic error terms or MA(1) error terms is applied to the Solow growth model. Monte Carlo simulations comparing this estimator to the Least Squares Dummy Variable and Minimum Distance estimators provided by Nazrul Islam (1995), the Arellano-Bond Generalised Method of Moments (GMM) estimator used by Caselli, Esquivel and Lefort (1996) and the Blundell-Bond (1998) GMM estimator show the ML estimator consistently outperforms the existing estimators under a wide variety of parameter values. The ML estimator is applied to a data set of 97 different countries and obtain estimates of the parameters in the Solow Growth Model.
Subject	:	Social sciences; Maximum likelihood-based estimation; Dynamic panel data; Unit-root processes; Solow growth model; Wald tests; Economic theory; Simulation; Economic growth; Studies; 0511:Economic theory
Added Entry	:	M. Ahn
Added Entry	:	Arizona State University