Dear Statalisters,
I need some advice on using mfx to compute elasticity in a tobit model.
Assuming that I specify the following model:
y = a + b ln(x) + c (ln(x))^2 + D Z
where x is income, Z is a vector of control variables in a linear form
(non-ln) and D is its corresponding parameters to be estimated.
I wish to calculate income elasticity and test whether the income
elasticity in this tobit model is significantly greater than unity. By
using mfx with dyex option, the command computes elasticity without
recognizing the square term of the x variable. My questions are
(i) is there a tip or trick to make mfx recognize the square term in
calculating elasticity and test it?,
(ii) whether the resulting variance-covariance matrix is correct because
the vector Z is not in a natural log form,
(iii) if mfx is not appropriate to calculate and test the elasticity, is
there any other command that will do the job.
Thank you very much in advance for your kind assistance.
/Thamana
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/