st: re: 3sls and fixed effects

[Christopher Baum <kit.baum@bc.edu>]

<>
Biljana said

The size is following:

Number of unique values of gvkey_n is  615
Number of records is  6492

I have totally 6492 firm-year observations.

The time period is 30 years (fyears)

That gives me matrix of:

2*((G-1) + (T-1))

2*(614+29)=1286

Is that too large?



I believe it is. You must be running Stata 10.x, as you can't use factor variables. In Stata 11, -help limits- does not say anything about the number of variables in an estimation command. I could swear that there was such a limit in Stata 10 and previous-- I think it was 800. It is not in the hardcopy manual for Stata 10 (nor 11), and I do not have access to Stata 10 at present. See what -help limits- on your version 10 system says.

But in any case if you have 615 firm dummies, I suggest you follow my prior advice and center all of the variables you want to include in the analysis by firm. Then you don't have to include i.gvkey_n. You can still include the time dummies, as there are only 30 of them. Centering both y and the other X variables manually will have the same effect as including a full set of firm dummies, and should get around the problem you're encountering.

Kit

Kit Baum   |   Boston College Economics & DIW Berlin   |   http://ideas.repec.org/e/pba1.html
                              An Introduction to Stata Programming  |   http://www.stata-press.com/books/isp.html
   An Introduction to Modern Econometrics Using Stata  |   http://www.stata-press.com/books/imeus.html


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/