Panel data is often grouped together. Not controlling for potential dependency within the group often leads to biased standard errors. Thus, it would be very useful if the plm package made it possible to group standard errors in case dynamic panel is estimated by the widely used GMM. This is especially referring to the Arellano, Bond (1991) and Blundell, Bond (1998) approaches.
Panel data is often grouped together. Not controlling for potential dependency within the group often leads to biased standard errors. Thus, it would be very useful if the plm package made it possible to group standard errors in case dynamic panel is estimated by the widely used GMM. This is especially referring to the Arellano, Bond (1991) and Blundell, Bond (1998) approaches.