Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: alternative specification for Oaxaca decomposition?


From   Maarten buis <maartenbuis@yahoo.co.uk>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: alternative specification for Oaxaca decomposition?
Date   Mon, 1 Jun 2009 08:07:02 +0000 (GMT)

-findit oaxaca- gives you a variety of packages. I think that the 
package you are looking for is described in:
 
Ben Jann (2008) "he Blinder–Oaxaca decomposition for linear regression models" The Stata Journal, 8(4): 453-479.
http://www.stata-journal.com/article.html?article=st0151

Hope this helps,
Maarten

-----------------------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany

http://home.fsw.vu.nl/m.buis/
-----------------------------------------


--- On Mon, 1/6/09, Akresh, Ilana Redstone <redstone@illinois.edu> wrote:

> From: Akresh, Ilana Redstone <redstone@illinois.edu>
> Subject: st: alternative specification for Oaxaca decomposition?
> To: "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
> Date: Monday, 1 June, 2009, 1:06 AM
> Hello,
> 
> I have a paper where I carried out a Oaxaca decomposition
> using the following code:
> 
> reg y x if group == 1
> predict yhat
> sum yhat if group == 2
> scalar a = r(mean)
> 
> sum y if group == 1
> scalar meany1 = r(mean)
> sum y if group == 2
> scalar meany2 = r(mean)
> scalar fractionduetodiscrim = (a - meany2) / (meany1 -
> meany2) scalar list fractionduetodiscrim
> 
> This should then correspond to the equation:
> 
> ybarDIS - ybarADV = [(xbarDIS - xbarADV)BetaDIS] +
> [xbarADV(BetaDIS - BetaADV)]
> 
> Where DIS indicates the disadvantaged group and ADV
> indicates the advantaged group. 
> 
> -----------------------------
> 
> **However, I would like to know how to calculate the
> decomposition as:
> 
> Yadv - Ydadv = <beta>*(Xadv - Xdadv) +
> Xadv(<beta>dadv - <beta>*) + Xdadv
> (<beta>* - <beta>adv) 
> 
> In the case directly above, <beta>* should be a
> vector containing the weighted averages of the 'adv' and
> 'dadv' regression coefficients, with the proportion of 'adv'
> and 'dadv' in the sample as the relevant weights. The first
> component on the right-hand side is the productivity
> advantage of the 'adv' group over the 'dadv' group in the
> absence of discrimination. The second term is the 'adv
> advantage' or the amount by which 'adv' are overcompensated
> relative to a discrimination-free setting, and the last term
> is the 'dadv disadvantage' or the amount by which 'dadv' is
> paid less than the wage that would prevail in the absence of
> discrimination. 
> 
> However, I am totally unclear on how to do this in Stata -
> in other words, what would be the code to get me
> there?  Advice, please?
> 
> 
> *
> *   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/
> 


      

*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index