[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: RE: Decomposition of effects: logit/probit models
> -----Original Message-----
> From: email@example.com
> [mailto:firstname.lastname@example.org] On Behalf Of
> Sent: 26 November 2003 09:24
> To: email@example.com
> Subject: st: Decomposition of effects: logit/probit models
> The commands -decomp- and -decompose- perform the the
> Oaxaca/Blinder decompositions regarding especially gender or
> race differences in earnings or wages.
> I would like to perform an analysis applying the
> Oaxaca/Blinder decomposition-logic in the case where the
> dependent variable is binary, say, work fulltime (1) or not (0).
> A paper by Picot and Myles* describes generically how to
> perform such an analysis. My question is: Has anyone written
> a do-file (or an ado-file) to carry out such a task?
> Best regards,
> Christer Thrane
> Lillehammer University College
> *Picot, G., & Myles, J. (1996) Social Transfers, Changing
> Family Structure and Low Income among Children. Canadian
> Public Policy, 22, 244-267.
I don't have do files, but it should be relatively easy to program (the
greater complications are providing standard errors for the components).
By the way, a better reference for the method is
Gomulka J and Stern N (1990) 'The employment of married women in the
United Kingdom 1970-83",
Economica 57, 171-199.
They were the orginators of the decomposition, and the Appendix to their
paper sketches out the econometric theory justifying it (and gives
Professor Stephen P. Jenkins <firstname.lastname@example.org>
Institute for Social and Economic Research
University of Essex, Colchester CO4 3SQ, U.K.
Tel: +44 1206 873374. Fax: +44 1206 873151.
* For searches and help try: