The key assumption in a Differences in Differences design is that both the control and the treatment group have parallel tendencies, that is, they'd follow similar paths over time if there was no treatment. I can see how it is important and unparallel tendencies surely make LATE estimation biased, but is there any pratical example of this problem? Any papers on it or that suffer from it?


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.