Psm

[Serena Gallo]

Hi everyone,
I want to apply propensity score matching and then implement the regression with the matched sample. But I have some concerns.
I want to estimate the effect of governance and financial characteristics on firm innovation. Firm innovation is a treatment variable that takes 1 if a firm performs innovation and 0 otherwise.
The sample between treated and untreated is strongly unbalanced and by applying t-test results, firms have different governance and financial characteristics.
Is correct the following command?

psmatch2 treatment  x1 x2, outcome(treatment ) neighbor(1) caliper(0.15)

I have concerns because my dependent variable is treatment. Can I include treatment as an outcome variable?

[David Radwin]

The FAQ asks posters to state the authors and source of user-written programs. psmatch2 is by Leuven and Sianesi and is available from SSC.

To answer your question, no, you cannot include the treatment as an outcome variable. You may benefit from reading a few articles and/or books on matching theory and practice. For suggestions, see
https://www.statalist.org/forums/for...99#post1244099 .

[li qizheng]

I have the similar question, and wondering the latest application in staggered did. Thank You!