Lockdown Effectiveness: Much More Than You Wanted To Know
Read the original on Astral Codex Ten →
Follow-up reading
↳ Things I Learned Writing The Lockdown Post — author's postscript on writing it · Jul 2021
Summary
Comprehensive MMTYWTK on whether lockdowns worked. Carefully decomposes 'lockdown' into specific NPIs vs. the Oxford stringency index, separates mandate effects from voluntary behaviour change (Goolsbee-Syverson: only 7 of 60 points of traffic drop was legal restriction), treats the pandemic as an imperfect control system, works through Flaxman (and its debunking), Brauner, Bjornskov, Lemoine, and US-state stringency-vs-death data, then runs his own Guesstimate QALY cost-benefit (median ~$98k/QALY; the emotional-cost framing much harsher — dozens of lockdown-months per healthy-life-month saved) and lands on well-hedged, Pareto-frontier conclusions.
Why this score
Quality 82 · Excellent. High-Excellent / strong-classic band — exemplary comprehensive, numerate, honest-about-uncertainty deep dive on a maximally contested topic, with real original analysis (his own US-state and cost-benefit models). A notch below the Ivermectin MMTYWTK (88) in definitiveness but clearly above Aducanumab (78). 82.
Claude’s paradigm shift 52 · Moderate. Notable — the control-system + Pareto-frontier framing is a fresh contribution, but the piece mostly synthesises others' analyses in its own time.
Real-world impact 3 · Moderate. Influential within the lockdown discourse; a retrospective analysis rather than policy-driving. 3.