Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | john sanders <desmochada@gmail.com> |
To | statalist <statalist@hsphsun2.harvard.edu> |
Subject | st: xtpoisson, fe marginal effects interpretation |
Date | Tue, 3 May 2011 13:52:35 -0400 |
Hi Everyone, I am estimating a count data model with fixed effects, trying to look at how events (i.e. the count variable) across geographic areas (i.e. the level of observation) are related to a continuous variable (the regressor). I first xtset the data: "xtset geography time" I run the following command: "xtpoisson events continuous_variable + controls, fe" which gives me regression estimates. I then use the postestimation command "margins, nu0" which should give the marginal effects under the assumption that the multiplicative fixed effect is equal to 1. This strikes me as a bit strange, as I am used to a linear model where the fixed effect is additive, essentially shifting the intercept, and where we would not think that a fixed effect equal to 0 (or log(1)) makes any a priori sense. The further issue is that I have no idea how to intepret the estimates. Are these counts or rates? Looking at the documentation (as well as Cameron and Trivedi's STATA book) would suggest they are counts (i.e. dy/dx = Beta*E(Y|x,controls,fixed effect)), but I always thought of Poisson estimates as being equivalent to rates. Lastly, since each of these geographic units are of a different population size, I want my coefficient estimates to reflect this. In a linear probability world, I would simply use aweight, but this is not possible using xtpoisson, fe. Any and all help is greatly appreciated. Best, John Sanders * * 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/