similar interaction effect first-differenced OLS/probit diff-in-diff

[Anna C]

hi,

I get the exact same interaction coefficient when running a first-differenced OLS model, compared to running a standard difference-in-difference approach using a probit model, recalculating the marginal interaction effect using inteff:

reg dh_60p_new dtreatment, vce(cluster clusterid)
probit h_60p time treatment time_treat, vce(cluster clusterid)
inteff h_60p time treatment time_treat

Would anyone know why this is the case?

Many thanks,
Anna

[Jorge Eduardo Perez Perez]

All the independent variables are discrete, so the marginal effect of time_treat in the probit model can not vary across its distribution.