Good day,
I'm a newbies to STATA and really need some help in my
paper..
With the typical "Gravity Model", where real trade
volume is a function of domestic income, foreign
income, and distance between 2 countries. X=f(GDPd,
GDPf, D)
For simplicity and purpose of illustration, let
X=f(GDPd, GDPf), therefore,
lnX_ijt=a+b(GDP_it)+c(GDP_jt)+u_ijt
In my paper, I further divide the Export, GDPd and
GDPf into Agri, Manufacturing and Service sectors, so
to analyze the effect of the growth of various sectors
on trade. Therefore the equation becomes..
|lnX(A)_ijt| |A_i/A_i-1 A_i/M_i A_i/S_i |
|lnX(M)_ijt|=a+B'|M_i/A_i M_i/M_i-1 M_i/S_i |
|lnX(S)_ijt| |S_i/A_i S_i/M_i S_i/S_i-1|
|A_i/A_j A_i/M_j A_i/S_j|
+ C'|M_i/A_j M_i/M_j M_i/S_j|+other var+u_ijt
|S_i/A_j S_i/M_j S_i/S_j|
where A_i(j) = growth rate of Agri in country i(j)
A_i/A_i-1 = ratio of growth rate of Agri
A_i/M_i = ratio of the growth rate of Agri/Manu
in country i
*ratio of growth rates was used as IV to avoid
the correlation between the variables and the error
terms
I hope my model make sense, if it does, here's my
problem.. how to regress it in STATA?
1. do I input all the data (lnX of A, M and S)
together and regress it? if so, how can i separate the
effects of each one?
2. Or should i regress each (A, M and S) equations
separately? when do i need to add dummy variables? I
think there's no need of dummy var when using fe or re
right?
3. Can anyone suggest me some good readings on panel
dynamic estimations? specially on how to
operationalize/step wise procedures on those
estimation methods. Any suggestions and comments on
the model is highly appreciated!
Thank you very much!
Wilson
__________________________________
Do you Yahoo!?
New and Improved Yahoo! Mail - 100MB free storage!
http://promotions.yahoo.com/new_mail
*
* 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/
