Please use this identifier to cite or link to this item:
Full metadata record
DC FieldValueLanguage
dc.contributor.authorPesaran, Mohammad Hashemen_US
dc.description.abstractThis paper presents a new approach to estimation and inference in panel data models with amultifactor error structure where the unobserved common factors are (possibly) correlatedwith exogenously given individual-specific regressors, and the factor loadings differ over thecross section units. The basic idea behind the proposed estimation procedure is to filter theindividual-specific regressors by means of (weighted) cross-section aggregates such thatasymptotically as the cross-section dimension (N) tends to infinity the differential effects ofunobserved common factors are eliminated. The estimation procedure has the advantage thatit can be computed by OLS applied to an auxiliary regression where the observed regressorsare augmented by (weighted) cross sectional averages of the dependent variable and theindividual specific regressors. Two different but related problems are addressed: one thatconcerns the coefficients of the individual-specific regressors, and the other that focusses onthe mean of the individual coefficients assumed random. In both cases appropriate estimators,referred to as common correlated effects (CCE) estimators, are proposed and their asymptoticdistribution as N tends to infinity, with T (the time-series dimension) fixed or as N and T tends to infinity (jointly)are derived under different regularity conditions. One important feature of the proposed CCEmean group (CCEMG) estimator is its invariance to the (unknown but fixed) number ofunobserved common factors as N and T tends to infinity (jointly). The small sample properties of thevarious pooled estimators are investigated by Monte Carlo experiments that confirm thetheoretical derivations and show that the pooled estimators have generally satisfactory smallsample properties even for relatively small values of N and T.en_US
dc.publisher|aCenter for Economic Studies and Ifo Institute (CESifo) |cMunich-
dc.relation.ispartofseries|aCESifo Working Paper |x1331en_US
dc.subject.keywordcross section dependenceen_US
dc.subject.keywordlarge panelsen_US
dc.subject.keywordcommon correlated effectsen_US
dc.subject.keywordestimation and inferenceen_US
dc.titleEstimation and inference in large heterogeneous panels with a multifactor error structureen_US
dc.typeWorking Paperen_US

Files in This Item:

Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.