st: Combined marginal effects for two parts model

[Husaina Banu Kenayathulla <husaina13@yahoo.com>]

Hi, 
  
I am wondering whether someone can help me if stata code for calculating combined marginal effects for two-parts model. 
  
For the first part of the hurdle model, I run the probit model (whether households incur educational expenditures or not on all the households with school-age children). On the second part of the hurdle model, I run the OLS (dependent variable-log of educational share on those households with school age children that incurred positive educational expenditures only). 
  
This is the command that I used: 
  
global xlist  lnpcaphhexp lntotmem  M0TO4 M5TO9 M10TO14 M15TO19 M20TO24 M25TO60 M60MORE  F0TO4 F5TO9 F10TO14 F15TO19 F20TO24 F25TO60  hhmale Chinese Indian Indigenous Christian Hinduism Buddhism otherel  urban Central  South East EM
 
probit educ_prob $xlist
mfx
regress lned5to19share $xlist if educ_prob==1
mfx 
  
But, I am interested to calculate the combined marginal effects for both part of the model(for instance, for variable M5to9; M5to9 is the proportion of male children age 5 to 9 in the household).  
  
Any help or advice would be appreciated, 
  
Thanks 
Husaina   
 

*
*   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/