Bias in Conditional and Unconditional Fixed Effects Logit Estimation

Bibliographic Data

ID6332005
AuthorsEthan B Katz (0009-0002-0526-8375, Harvard University, corresponding author), Ethan Katz
Year2001
Volume9
Issue4
Pages379-384
Publication date2001-01-01
Peer ReviewedYes
Open AccessYes
TypeARTICLE
VenuePolitical Analysis (JOURNAL)
Journal identifiersISSN: 1047-1987 • E-ISSN: 1476-4989
PublisherCambridge University Press (CUP) (PUBLISHER)
DOI10.1093/oxfordjournals.pan.a004876
OpenAlexW2077379689
LanguageEN
Citations received43
References cited10

Fixed-effects logit models can be useful in panel data analysis, whenNunits have been observed forTtime periods. There are two main estimators for such models: unconditional maximum likelihood and conditional maximum likelihood. Judged on asymptotic properties, the conditional estimator is superior. However, the unconditional estimator holds several practical advantages, and therefore I sought to determine whether its use could be justified on the basis of finite-sample properties. In a series of Monte Carlo experiments forT< 20, I found a negligible amount of bias in both estimators whenT≥ 16, suggesting that a researcher can safely use either estimator under such conditions. WhenT< 16, the conditional estimator continued to have a very small amount of bias, but the unconditional estimator developed more bias asTdecreased

Econometrics · Estimator · Fixed effects model · Logistic regression · Logit · Maximum likelihood · Panel data · Statistics · Economic and Environmental Valuation · Housing Market and Economics · Mathematics · Spatial and Panel Data Analysis

Tools

Open DOISci-Hub