I am working with a panel data with two periods (dummy variable time equals 1 for post-intervention data) and two groups (dummy variable intervention equals 1 for the group which received the treatment). As such, I ran the following on STATA:

xi: ivregress 2sls outcomevar intervention time (instrument=did)

where did=intervention*time.

It works perfectly well. However, I was wondering how I could assess the effect of female vs male respondents for example. Should I use a triple-difference to do so? Or run the regression seperately for female and male respondents?

Thank you

  • $\begingroup$ Cross-posted on Statalist. $\endgroup$ – Dimitriy V. Masterov Dec 8 '17 at 2:33
  • $\begingroup$ Hi Dimitriy, it's a different question - on statalist, I am asking about DID heterogeneous effects, whereas here, I am using about IV regression $\endgroup$ – Nick Dec 8 '17 at 5:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.