Early, high‑visibility epidemic models that pool data across jurisdictions can act as accelerants for large‑scale interventions by producing timely, dramatic counterfactual claims (e.g., 'lockdowns were necessary and sufficient'). Those models produce powerful policy effects but also compress complex behavioural change into intervention dates and rely on fixed epidemiological parameters.
— If models routinely become decision engines in crises, we need governance rules for model provenance, sensitivity disclosure, and institutional checks to avoid lock‑in on fragile assumptions.
2020.06.08
100% relevant
Flaxman et al. (Nature, June 2020) used partial pooling across 11 European countries, back‑calculated infections from death data, and concluded lockdowns were the dominant driver of R_t reductions—an archetypal example of a model that accelerated adoption of national lockdowns.
← Back to All Ideas