dear kit,
thanks for your comment. i am aware of the risks you are mentioning,
but the point of my model is that the gain in efficiency due to system
estimation should make the difference with respect to
limited-information estimation. in addition, i need to test for cross
equation constraints.
is there a way i can implement system estimation correcting for serial
correlation in stata?
many thanks for your help,
john
On 15/06/07, Kit Baum <[email protected]> wrote:
This is a FAQ. _Why_ do you need to jointly estimate these equations?
Are you imposing cross-equation constraints, or need to test for
same? If not, you can gain consistent point and interval estimates of
each equation by single-equation (limited-information) estimation.
The only thing that systems (full-information) estimation has to
offer is a possible gain in efficiency. On the down side, systems
estimation also implies that any misspecification will pollute the
entire system, not just the equation in which it appears.
Kit Baum, Boston College Economics and DIW Berlin
http://ideas.repec.org/e/pba1.html
An Introduction to Modern Econometrics Using Stata:
http://www.stata-press.com/books/imeus.html
On Jun 15, 2007, at 2:33 AM, John wrote:
> i need to jointly estimate a simultaneous system of equations on panel
> data with a long time series dimension.
> my problem is that i need to correct for error serial correlations and
> as far as i understand this cannot be done under the 3sls "reg3"
> command.
>
> the command "ivreg2" would be useful, but it seems it can only applied
> in a single equation setting.
>
> is there a way to estimate a system of equation correcting for error
> serial correlation?
*
* 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/
*
* 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/
