Panel Data Fixed Effect - Country Code omitted because of collinearity

[Anisul Islam]

In a regional level study, my panel ID is institution code and panel time variable is year.

I have used a separate independent dummy variable for the individual country of the region. I used i.countrycode after regular independent and control variables.

In fixed effect panel estimate, the stata output is showing 'Country Code omitted because of collinearity', however, the remaining portion of the output is okay.

What should I understand from this? Is there anything wrong?

[Carlo Lazzaro]

Anisul:
welcome to this forum.
Nothing weird with your output: -fe- machinery wiped out time-invariant predictors (as countries are, because they do not change within the same panel as time goes by).
For the future, please follow and act on FAQ recommendations: post what you tyoed and what Stata gave you back. Thanks.

[David Benson]

It's because when you run xtreg, fe i(countrycode) it is creating the country dummy variables for you, so you've essentially done it twice.

In other words:

reg   dep_var control_vars i.countrycode

xtreg dep_var control_vars, fe i(countrycode)

Both specifications are identical, although the first one will show you the coefficients for the various countries.

[Anisul Islam]

I typed following:
xtreg OUT DTA SZE RSK LIR i.CCODE, fe cformat(%9.3f) pformat(%5.2f) sformat(%8.2f)

and output is in image-