Models that assume transmission (R_t) falls instantly when a government orders an intervention will likely attribute too much effect to the policy and undercount gradual behaviour change or other drivers. This can mislead policymakers about the marginal value of specific NPIs and the timing needed for lifting measures.
— Modeling assumptions about how quickly behaviour and transmission respond to policy materially change estimated intervention effects and therefore should be central to public debate on lockdown design and exit strategies.
2020.06.08
100% relevant
Flaxman et al. explicitly state their model "assumes that changes in R_t are an immediate response to interventions rather than gradual changes in behaviour," making this assumption traceable to the paper.
← Back to all ideas