Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Maarten buis <maartenbuis@yahoo.co.uk> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: st: RE: RE: estimation with a time trend. |
Date | Mon, 5 Jul 2010 16:39:30 +0000 (GMT) |
--- On Mon, 5/7/10, natasha agarwal wrote: > The time dummies are highly collinear with one of my > variables in the model. However, exclusion of the same > would lead to an error as we wanted to control for > business cyclical effects. For the very reason, we > decided to introduce a rough measure of time dummies > that would be a trend. The point of a cyclus is that it implies things going up and than going down again and than going up again and etc. etc. A trend implies that something either goes up or goes down. So it is odd to capture something like a cyclus with a trend. If your time frame is short enough compared with the period of the cycle (how long it takes to for the cycle to finish an entire first up and than down and than up again cyclus), you can get away with that, but that is something that you will need explicitly explain in your paper. I would look for official data on the timing of the business cycle and see if my timeframe fits entirely within a single such period and present this as additional evidence that my time trend captures the business cycle. -- 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/