Good day everyone.

I am running an experimental analysis (difference-in-differences) and I have a little but a big problem. The interaction term (treatment year and treatment group in this case ty##tg) is omitted because of collinearity. I have tried every possible means to address the problem but to no avail (note: there are enough observations for the interaction term and even after implementing the filters (ifs)).

Please below is the command line the note from stata.

Code:
reg arbink ty##tg man age i.sni2. i.FA. i.year. i.educ if inlist(qun, 1,5) & bet2050==1, vce (cluster PID)


note: ob.ty#1.tg identifies no observations in the sample
note:1.ty#tg omitted because of collinearity

Please, any help as to how to deal with this?

Thank you in advance.

Evans.