Re: st: Year Fixed Effect Interpretation

[Austin Nichols <austinnichols@gmail.com>]

Vinod <Vinod.Venkiteshwaran@tamucc.edu> :
Simply add year as a regressor to measure the average trend; the year
dummies then measure deviations from trend, and some number of them
will be dropped, depending on what is need to avoid perfect
collinearity.  If you have another regressor that exhibits a perfectly
linear trend (e.g. because it is constructed as a linear interpolation
across two points) then the regression is not estimable, in theory or
practice.

On Wed, Sep 7, 2011 at 4:18 PM, Venkiteshwaran, Vinod
<Vinod.Venkiteshwaran@tamucc.edu> wrote:
> Hello
>
> I am running a pooled regression with dummies(t-1 dummies) as controls for year effects. I have chosen to drop the first year so that I can interpret the coefficients on the remaining year dummies in terms of the first year. The estimation results shows that the coefficients on the dummies display a strong downward trend. At this stage I have to include an additional independent variable that is collinear with these year dummies so Stata automatically drops another dummy , for the last year, before running the regression. My focus is on whether or not the coefficients on the dummy variables systematically increase or decrease over time. However, since the dropped dummies are at the beginning and end of the sample the trend in the dummy coefficient is no longer present. I have tried to drop two year dummies at the beginning of the sample but still the time trend seems to have vanished!
>
> Any suggestions on how the coefficients of year dummies are interpreted when the two dropped years are at either end of the sample, together at the beginning or the end, or somewhere in the middle?
> Thank you
> Vinod
>

*
*   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/