Scott Merryman (generously) replied:
> I believe the problem is not that you have unbalanced panels, but you have
> missing observations for your time variable.
[...]
> If there is data where there are missing year values then the solution
> would
> be to replace the missing year values; if there is no data then simply
> drop
> the missing observations.
>
> If you take a look at lines 73 through 76 of the -xtabond2- code you will
> see:
>
> count if `t' >= .
> if r(N) {
> di as err "Missing values in time variable (`t')."
> exit 459
>
> ...this is where you are getting tripped up.
>
> Again, it is the missing values on time variable that is causing the
> problems, not that the data is an incomplete or unbalanced panel.
Thanks for this, Scott: I'm indebted to you. As my last post indicated,
I'd worked out that the missing values in my time variable was the
problem. My solution was indeed to drop these observations and save it as
a new dataset and I'm happy with that as a solution.
The next step is to be satisfied with the validity of the models I'm
getting -xtabond2- to pump out. For instance, I've just run a model using
six GMM instruments, none of which appear as regressors. Is this OK? None
of them are correlated with the error term but _are_ correlated with the
LDV (so are valid as instruments). I didn't include the regressors as
instruments, as all were correlated with the error term. Hansen's J-test
was significant: I haven't a clue whether this is good or bad news, even
after reading the help file (my fault; _not_ David Roodman's)! I think
I've run this model correctly as the coefficients look sensible, but I'm
entering new territory here.
Anyone, of course, is invited to reply.
CLIVE NICHOLAS |t: 0(044)191 222 5969
Politics |e: [email protected]
Newcastle University |http://www.ncl.ac.uk/geps
*
* 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/
