Re: st: mfx- marginal effects for log linear model

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

--- On Wed, 7/7/10, Nyasha Tirivayi wrote:
> I would like to determine the marginal effects at means for
> the following model
> 
> lnY= a + D + ln X
> 
> where D is a dummy. How do I determine the marginal effects
> at means for the log transformed variables and the dummy in
> Stata? Which mfx commands in stata 10 do I use?

Computing marginal effects in such a model is surprisingly 
difficult. Notice that you model the average ln(y) rather 
than the average of y, and since the logarithm is a 
non-linear transformation, that results in all kinds of 
complications. For that reason I would avoid marginal 
effects for this model.

Instead, I would directly interpret the coefficients: the
coefficient of ln(x) is an elasticity, the expected 
percentage change in y for a percent change in x, and 
100*the coefficient of D is the percentage change in y
for a unit change in D.

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/