Dear all,
I am trying to estimate the following model:
y_it=a*y_it-1+b*x_it-1+c*z_it-1+error
I have two questions that I would appreciate if you could help me...
1) How can I check if the variable z_t-1 is an endonegenous variable?
I followed the following procedure: first I run xtabond2 with the
option iv(z); second, I run xtabond2 with the option gmm(z,lag(. .));
at last I performed the Sargan test for the additional moment
conditions(Sargan2-Sargan1). As the p-value is about 80% I conclude
that I should include the additional moments and hence z is not
exogenous. Is this procedure correct? If not, how can I check for
endogeneity of z?
2) The variable x is highly persistent in time (autocorrelation:0.98).
Ss past values of x are not good instruments for differenced equatons,
does the following formalization deals with this characteristic of the
variable x?
xtabond2 l.y l.x l.z, gmm(y z, lag(2 4) eq(diff)) gmm(x, lag(2 2)
eq(diff)) gmm(x, lag(2 4) eq(level)) twostep
I've only used x as an instument for level equations . Is that correct?
Thanks in advance for your attention,
Ana
*
* 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/
