Comparing coefficients across regressions

[Tom Ford]

Hello,
I am estimating a Difference in difference model, and I want to examine whether the treatment has significantly different effects on two different outcomes. How can I do this?

In particular I am estimating these two models:

did2s Y1, first_stage(i.unit i.year) second_stage(treatment) treatment(treatment) cluster(unit)    
did2s Y2, first_stage(i.unit i.year) second_stage(treatment) treatment(treatment) cluster(unit)

I would like to test whether the treatment has the same effects across the two dependent variables. How can I go about it?

thank you very much in advance for your help

[Carlo Lazzaro]

Tom:
can -suest- help?

[Tom Ford]

Thank you Carlo for your reply.
Unfortunately the did2s command requires the cluster option, which it is incompatible with suest.


I had a look and tried this on a simpler event study model like so:

    reghdfe Y1 L_ttt_* F_ttt_*, absorb(unit year)
    eststo m1
reghdfe Y2 L_ttt_* F_ttt_*, absorb(unit year)
    eststo m2

But using suest in post estimation I get the following error

suest m1 m2, vce(cluster unit)
unable to generate scores for model m1
suest requires that predict allow the score option

do you know how I could overcome this issue?

thank you very much for your support

[Carlo Lazzaro]

Tom:
quoting from the helpfile of the Stata community-contributed module -did2s- (as you are kindly requested to mention it, for sound reasons that are reported on the FAQ):

cluster(varname) What variable to cluster on (use unit id if you don't want to cluster).

HTH