Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: RE: Year dummies in panel regs (Econometrics question)


From   "Swanquist, Quinn Thomas" <qswanqui@utk.edu>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   st: RE: Year dummies in panel regs (Econometrics question)
Date   Sun, 22 Jul 2012 18:53:35 +0000

Year FE actually capture both observed and unobserved year fixed effects. The latter is the one we worry about. For example if you think that yearly inflation and investor sentiment in a given year (assuming these impact all observations within a year equally) is related to the effect of X on Y, then you need year effects in the model.

When making the decision to include these in your model or not, you need to evaluate these year indicators as you would any other control variable. That is, is there a an unobserved year fixed effect that could explain the relation between X and Y? Also, you need to ensure that there is sufficient within year variation in your variable of interest, otherwise the year fixed effect will capture most of the effect (multicollinearity). In general, including year fixed effects won't hurt your consistency but might reduce efficiency. Not including year fixed effects when they are needed creates biased estimators.

Quinn Swanquist
qswanqui@utk.edu



-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Narahari H.S.
Sent: Sunday, July 22, 2012 12:46 AM
To: statalist
Subject: st: Year dummies in panel regs (Econometrics question)

Hello all,

This is not a Stata specific question but a general econometrics question.

I am curious to know why year dummies are included in panel regressions. I have asked a few professors and googled a bit but the only answer I get is "to capture unobserved year effects". Is that all there is to it or is there any other reason why they are included? Also, what is the mistake in not including them in the regression model?

Thanks in advance
Hari

*
*   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/




*
*   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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index