Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


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

st: Is demeaning variables that cover different time periods, but are used in the same regression, a problem?


From   Bertel Teilfeldt Hansen <geilfeldt@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   st: Is demeaning variables that cover different time periods, but are used in the same regression, a problem?
Date   Fri, 14 Jan 2011 16:22:39 +0100

Dear statalisters!

I am currently writing a thesis about countries' level of democracy
and their economic growth, where democracy is operationalized as a
stock-variable (something that accumulates over time). To eliminate
time-invariant and unit-specific errors I wish to employ country-fixed
effects (time-demean the variables in my regression).

Here's the problem, though: For theoretical reasons, I would like to
demean each country's stock of democracy from 1800 till 2009, whereas
I - because of scarcity of data - can only demean other variables from
1960 till 2009. This means that I want to include a variable that has
been demeaned with regards to a long period of time, and other
variables that have been demeaned with regards to a shorter period of
time in the same regression.

Now, this is no problem to do technically, but what I'm unsure of is
how big of a statistical problem it is. I have a hunch that it is
somewhat of a problem, but how grave it is, I really don't know.

Kind regards

- Bertel Teilfeldt Hansen, University of Copenhagen
*
*   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 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index