Actually, the longitudinal studies that back that up are available. Colorado has done detailed work. I think they use a minimum of five different studies, some from their own state and some from the national level, in which they are looking at use. They, I would say, have a fairly sophisticated method of testing and double-testing, using sources from different places. I would think we could probably learn something from that.
On September 11th, 2017. See this statement in context.