1
votes

I'm doing OLS fixed effects regression, and would like to test whether coefficients are the same between the two. One of the regressions has a different dependent variable than the other.

How can I do this?

Specifically, one of my regressions is:

xtreg black MAshock i.year, cluster(fips)

The other regression is:

xtreg white MAshock i.year, cluster(fips)
1
This answer is plastered all over the internet. One place to look: stata.com/statalist/archive/2013-05/msg00119.html, or stata.com/statalist/archive/2006-06/msg00837.html - SOConnell
Even more here, here and here. - Roberto Ferrer

1 Answers

0
votes

you should run one single regression and interact everything with a black dummy variable. the coefficient on that interaction term will test whether the coefficient of interest are the same or not.