Personally, this is where I hope the costs are going to come down with time. I'm content for this to be something we try, and we can measure its costs and effic

Personally, this is where I hope the costs are going to come down with time. I'm content for this to be something we try, and we can measure its costs and efficacy after some real time and investment. The initial outlay and efforts are likely going to have a lot of bespoke work and trial and error, and lack economy of scale. If after a decade it ends up being very expensive but very effective, or moderately expensive and moderately effective, or expensive and not very effective, we can make better informed decisions -- rather than stick with prison by default, just because it is cheapest for now.