st: PSMATCH2 Mahalanobian matching

[Aakanksha Pande <apande@fas.harvard.edu>]

Hi,
I am using PSMATCH2 to perform Mahalanobian matching with calipers defined
by .25* std dev of propensity score and would appreciate clarification on
the following three questions, for which I am having a hard time finding
answers to.
My code is:

set seed 1234
generate x= uniform()
sort x
label variable x "Random variable"

bs, reps(1000): psmatch2 treatment,  ///
out(y) ///
 mahalanobis(x1 x2 x3 x4) ///
 caliper (.11116)

My unanswered questions are on:
1.Does this code do Mahalanobian matching with replacement? If so, does
psmatch2 adjust for the number of times the control observation is used as
a match for different treatments? If not, how does one do so.
2. Bootstrap: Can I use these bootstrap standard errors even though they
are not based on a kernel estimator as suggested by Abadie and Imbens?
3. Is the matching based on Mahanalobian distance a one to one or many to
one match? If many to one, does psmatch2 appropriately weight for number of
matches?

Thank you for your help.

best
Aaka
*
*   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/