Bo Cutter <[email protected]> asked:
> I am using the xtgee command to get robust standard error estimates for a
> random-effects panel data model. (i.e. xtgee..., family(gauss) link(ident)
> robust). p.446 of manual S-Z (last sentence, first paragaph) has a sentence
> that implies these robust standard errors are correct even if the
> panel-specific effect is correlated with the regressors. Can anyone tell me
> if I am interpreting this statement correctly ? And do you know what the
> reference behind this statement is ?
And Bobby <[email protected]>, correctly noted that :
> Yes, this statement is true. In general, the robust standard errors
> are a true indicator of the sample-to-sample variability of your
> parameter estimates even in mis-specified models.
> As for a reference, try
> Domowitz & White (1982). Misspecified models with dependent
> observations. Journal of Econometrics, 20, 35-58.
But it should be clarified that in this case -xtgee , robust- is providing
consistent estimates of the standard errors of estimates of population
parameters that are probably NOT what Bo wants from his model.
Let me explain via an example. Consider
y_it = B x_it + u_i + e_it
where u_i is i.i.d over the panels and e_it is i.i.d over all the
ovbservations. If x_it is correlated with the unobserved u_i then two
things are true:
i) xtgee will not provide consistent estimates of the parameter B for
the model that assumes x_it is uncorrelated with u_i.
In particular, if x_it is correlated with u_i, the estimate that
-xtgee- produces for B cannot be interpreted as the marginal effect
of x_it on y_it.
ii) But there exists a true value of B in the mis-specified
model in which x_it is correlated with u_i. The robust option
provides consistent estimates of the standard
errors of the estimate B in the mis-specified model.
--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/
