Omitted Variables in Fixed Effects Model

[Ivana Ma]

Hello everyone!

I am using panel data and I am trying to run a Fixed Effect model. [Code: xi: xtreg dependent independent1 independent2 dummy population, fe vce(cluster firm_id)]
My sample consists of 1500 firms for 20 years.
My dependent variable and two of my independent variables are all continuous. There is also one dummy variable and a control variable for population.

	Robust
dependent	Coef.	Std. Err.	t P>t	[95% Conf.	Interval]
independent1	.0425467	.0844767	0.50 0.615	-.1233011	.2083945
independent2	.1066156	.0537585	1.98 0.048	.0010749	.2121564
dummy	0	(omitted)
population	0	(omitted)
_cons	.0004769	.0003729	1.28 0.201	-.0002551	.0012089
sigma_u	.0036929
sigma_e	.05388518
rho	.00467477	(fraction	of	variance due	to	u_i)

As you can see, dummy variable as well as population, are omitted....And that's the problem.

The dummy variable is constant for five years (changes happen only once during this period) and the population (I used the logged variable for population) is constant during the decade (due to the fact that I have used the Census data).

What should I do to deal with this problem?

Thank you
Kind regards,
Ivana

[Andrew Musau]

See discussions herein for time invariant variables and fixed effects

http://www.statalist.org/forums/foru...iant-variables

Now, someone may ask you how old you are today to which you may answer 21. However, you may be turning 22 next week but it is still acceptable that you state that you are 21. On the other hand, in most cases It is unreasonable to state that the population of country "X" is that of the last census count 6 years ago (due to births, deaths, and migration over the proceeding years). Annual population estimates are available for almost all countries in the world and therefore you should never regard population as time invariant.

[Ivana Ma]

See discussions herein for time invariant variables and fixed effects

http://www.statalist.org/forums/foru...iant-variables

Now, someone may ask you how old you are today to which you may answer 21. However, you may be turning 22 next week but it is still acceptable that you state that you are 21. On the other hand, in most cases It is unreasonable to state that the population of country "X" is that of the last census count 6 years ago (due to births, deaths, and migration over the proceeding years). Annual population estimates are available for almost all countries in the world and therefore you should never regard population as time invariant.

The link was very helpful!

Thank you very much Mr. Andrew Musau!