That was exactly my point.

How do you prove that housing supply linked to homelessness is NOT just a correlation?