Dzemski, AndreasOkui, Ryo2020-08-172020-08-172020-081403-2473 (print)1403-2465 (online)http://hdl.handle.net/2077/66141JEL code: C23, C33 C38We study kmeans clustering estimation of panel data models with a latent group structure and N units and T time periods under long panel asymptotics. We show that the group-speci c coe cients can be estimated at the parametric root NT rate even if error variances diverge as T ! 1 and some units are asymptotically misclassi ed. This limit case approximates empirically relevant settings and is not covered by existing asymptotic results.15engPanel datalatent grouped structureclusteringkmeansconvergence ratemissclassificationText