Patricia,
I wonder if matching is appropriate in your case. Why don't you
consider a duration-model approach, and more precisely the
timing-of-events method from Abbring and van den Berg?
Here are two interesting articles on that topic:
Abbring J. and G. van den Berg, 2004, "Analyzing the effect of
dynamically assigned treatments using duration models, binary
treatment models, and panel data models". Empirical Economics, 29(1).
A preprint version is available from Jaap Abbrings homepage:
http://www.econ.vu.nl/medewerkers/jabbring/papers/empec_final.pdf
Fredriksson, P. and P. Johansson, 2004, "Dynamic Treatment Assignment
-- The Consequences for Evaluations using Observational Data".
Manuscript, Department of Economics, Uppsala University.
http://www.nek.uu.se/StaffPages/Publ/P371.pdf
Best wishes,
jean
*
* 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/
