Re: st: coldiag2 to check for collinearity.

[Maarten buis <maartenbuis@yahoo.co.uk>]

--- On Mon, 5/7/10, natasha agarwal wrote:
> I issue the following commands
> 
> xtreg lnrval lnk lnw lnvfdi y14-y18, fe vce(robust)
> 
> Table 1. Intra-and/or-Inter-Industry  FDI-induced
> spillovers within a Province
> Dependent Variable: Log of Value Added of Domestic Firms
>              
>         (1)             (2)
<snip>
> Log of FDI     
>    0.294***           0.022
>     (0.013)          (0.020)
<snip>    [model (1) without time dummies; model (2) is
with time dumies, MB]     
> Joint Significance of Time Dummies
> (F-test)    113.82***    
> 
> 
> I understand that the highly significant coefficient on the
> time dummies in the regression is because they are highly
> collinear with lnvfdi as the results show above.

The time dummies do what you wanted them to do: control for
idiosyncratic changes over time (what economists call the 
business cycle). If you don't include the time dummies then
most of those changes over time will be assigned to FDI.
You wanted to control for that, and it turned out that there
isn't any effect left. That is your result, whether you like
it or not. By calling this collinearity, you suggest that 
somehow that effect still exist, but was hidden by some
technical problem with a fancy name. This is not the case,
your result is that FDI has very little (or no) effect.

Hope this helps,
Maarten

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany

http://www.maartenbuis.nl
--------------------------


      

*
*   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/