[kdtop3]

I find the graph that others link to [1] to be confusing at best. It seems to be tabulating *absolute harm* over all users. So if a substance has many users, it has more total harm. But then the substances are ranked in a way that seems to imply the intrinsic danger of it all by itself. Thus alcohol may cause more damage on a national level than crack cocaine. But tracking the lives of 1 user of alcohol vs 1 user of crack cocaine would show, IMHO, the cocaine use to lead to more problems.

A similar analogy would be to compare the relative danger of automobiles to atomic bombs.

[1] https://www.economist.com/graphic-detail/2019/06/25/what-is-...