Panel data regression with one time-invariant variable

[Jasper de Boer]

Hi all,

Currently I am using a panel data set and I am trying to find the effects of a time-invariant variable (measured at one point in time and assumed to be stable over time) on several time specific variables. So, my dataset exists of several observations that all have values for:

id characteristic_x income_2010 income_2011 income_2012 income_2013

Now I would like to find the effects that characteristic_x (the time-invariant variable) has on the change in income over time. However, just running a panel data regression does not work as Stata omits the time-invariant regressors automatically.

Does anyone know how to take care of this problem?

Thanks in advance

[Carlo Lazzaro]

Jasper:
welcome to this forum.
You do not describe what you typed an what Stata gave you back (as you should in order to increase your chances of getting helpful replies. See the FAQ).
That said, my guess is that you have applied a fixed effect estimator which, as expected, gets rid of time-invariant predictors. The trivial fix is to switch to random effect estimator (see -hausman-).