st: xtabond2

[Steven Archambault <archstevej@gmail.com>]

Dear Statalist,

I have been learning how to use the xtabond and xtabond2 commands. I
am very thankful for to helpful articles on the topic.

http://ideas.repec.org/p/cgd/wpaper/103.html

http://www.fordham.edu/economics/mcleod/Elitz-UsingArellano–BondGMMEstimators.pdf

In the second paper the example includes exogenous instruments
[iv(fin_integr trans_index flows_eeca l.growth uncert tot dev_m2)]
which are applied to all of the gmm variables and their lags in the
model. The entire line of code is as follows.

xtabond2 inv l.inv fdi loans portfolio l.growth uncert tot dev_m2, gmm
(inv fdi loans portfolio, eq(diff) lag (3 3)) iv(fin_integr
trans_index flows_eeca l.growth uncert tot dev_m2) small noconst

My question is whether or not it would make a difference to apply the
exogenous instruments to each of the gmm endogenous variables (inv fdi
loans portfoli) if in fact some instruments only apply to one or two
of the gmm variables. For example, maybe fdi has different indep
instruments than loans. Can we specify which instrument goes to which
variable? And if not, does it matter?

Thanks for your help here!

Regards,

Steve

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/