LIML in the static linear panel data model

We consider the static linear panel data model with a single regressor. For this model, we derive the LIML estimator. We study the asymptotic behavior of this estimator under many-instruments asymptotics, by showing its consistency, deriving its asymptotic variance, and by presenting an estimator of the asymptotic variance that is consistent under many-instruments asymptotics. We briefly indicate the extension to the static panel data model with multiple regressors.