Marijuana: An Update

Update to Scott's 2013 marijuana cost-benefit analysis. New post-legalization data (Highway Loss Data Institute, ~5-6% rise in auto-insurance claims; a fatality study's non-significant +2.7%) contradicts his earlier prediction that marijuana would substitute for alcohol and REDUCE crashes. He concedes he was wrong, recomputes (~1,800 extra US deaths/yr if nationally legalized), shifting the net effect of legalization from slightly positive to slightly negative (especially vs. decriminalization), then lists nine intuition-guiding considerations (still less bad than alcohol; bystander vs self harm; the cancer wildcard; taxes won't be spent well) and declines to land on a firm position.

Quality 64 · Strong. Strong (64). Lifted above a routine short update by genuine intellectual honesty -- a public 'I was wrong' on a consequential empirical prediction, cleanly recomputed ('score one for obvious things over clever substitution effects') -- plus the thoughtful nine-point list. Capped in Strong by its brevity and dependence on the larger parent MMTYWTK.

Claude’s paradigm shift 42 · Moderate. Moderate (42). An evidence update rather than a new idea; the small generalizable lesson (obvious effects beat clever substitution stories) gives modest novelty.

Real-world impact 2 · Minor. 2 -- minor/within-blog. Contributes to the marijuana-legalization cost-benefit discourse but with a within-blog footprint; the heavier real-world weight sits with the parent analysis.