(This is cross-posted from the "Help with Programming" section of this board. In that section, this same question was asked in 2011 but there are no replies until 2017, when I posted the following.)
I wondered if anyone has worked on this. The comparable Stata extension is described at: http://www.stata.com/manuals14/xtxtdpdsys.pdf. It appears in Stata they simply expand the instrument list but the discussion in Baltagi (2013), sec 8.5, does not make this seem like the obvious way to proceed: "...the use of an extended system GMM estimator that uses lagged differences of y(i,t) as instruments for equations in levels, in addition to lagged levels of y(i,t) as instruments for equations in first differences." This also appears to be sec 11.4 in Wooldridge (2010) but he does not link the discussions of estimators to published literature. This seems straightforward, perhaps an extension on UG-423?
The Rats TextbookExample bpanel08-8-1.rpf shows the various estimators on page 177 (Table 1) of Baltagi, 5th ed, but it omits the "System GMM" Blundell-Bond estimator (the example is written for the 4th ed, but the context is clear). Baltagi's 5th ed includes the "System GMM" as the final estimator in the table.
Thanks for any hints.
Richard