re: st: Confusion about fixed effects specification

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

<>
I asked this question once but unfortunately no one replied me back
I have got problem in specification of fixed effects in my
estimations. I am using a three-dimensional panel data (bilateral
country flows, commodity, year) in the gravity model. my data forms
like below:
Malaysia to China-Agricultural,year
.
.
.
Malaysia to China-Manufacturing,year
.
.
.
Malaysia to Germany- Agricultural,year
.
.
.
Malaysia to Germany-Manufacturing,year

as you see in this example I will have 4 cross-sections. but the
problem is that I want my model be fixed just by the country pairs but
not by commodity (e.g. I want to estimate the effects of Malaysia to
China not Malaysia to China-agriculture) how can I estimate to get the
required effects?


If you want the panel variable to be the country-pair,then you must aggregate over commodities. Otherwise you would have repeated time
observations for each value of the panel variable, and Stata won't like that. So just collapse out the commodity aspect of the data. 

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/