[medvezhenok]

I'm not saying whether or not we should do it. I'm saying that the estimates given (8B) don't take into account behavioral changes based on the program. If we expand the scope (as you said), we're looking at ~80B instead of 8B.

But that's just one example - there are many other possible behavioral changes that can happen that will alter the cost computed up front even more (i.e. we're not even considering fraud with people applying for free housing and how much $ it will take to catch the fraud and deal with it).

I think we should advocate for solutions, as long as we keep in mind the true potential costs.