Marco Shiva <[email protected]> wrote that
> I would like to ask if xtabond allow us to declare a predetermined
> variable without asking stata to estimate its contemporaneous
> coefficient (just in case some endogenity problem exist or remains).
-xtabond- forces the lags of the predetermined variables to start with
current observation and work backward to the requested number of lags.
However, if you only want to drop the contemporaneous variable due to
endogeneity, then you could specify the endogenous option within the pre()
option. This option specifies that the ``predetermined" variables should be
treated as endogenous variables. For instance, instead of
. xtabond y l(1/2).x pre(w,lag(2,1)) lags(2)....
specify
. xtabond y l(1/2).x pre(w,lag(2,1) endogenous) lags(2)....
The contemporaneous variable is still included in the model, but the
set of instruments is designed to account for the endogeneity of the
variable.
--David
[email protected]
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
