This from @BillHanage is the piece I've been waiting to see debunking the apparent UK strategy. I defer to him on the epi analysis (which certainly echoes my own shock at this plan).
I'll add a few thoughts from a policy/strategy/execution perspective.
I'll add a few thoughts from a policy/strategy/execution perspective.
People who spend their lives studying epidemiology understand the uncertainty and limitations of models.
Policymakers desperate for solutions to overwhelming problems sometimes don't. I wrote about this dynamic in US policymaking on Ebola.
cgdev.org/publication/st…
Policymakers desperate for solutions to overwhelming problems sometimes don't. I wrote about this dynamic in US policymaking on Ebola.
cgdev.org/publication/st…
The famous adage is "all models are wrong; but some are useful." Meaning that a model can only be as accurate as the data and assumptions that you build into it. Good models can inform strategy but they're not precisely predictive. If you have blind spots, so will the model.
The UK strategy appear to hinge on reality closely following modeled predictions. But any models for COVID are populated by data and assumptions about a disease we've never seen before and don't understand very well. That creates huge uncertainties and risk of blind spots.
Strategy then hinges on abruptly slamming the brakes once transmission once models say so. Getting the timing right would require an enormous degree of confidence that your epi surveillance provides a near-perfect real-time picture of transmission dynamics - again, dubious.
And for that slam-the-brakes to work, you have to then execute perfectly and rapidly, after having spent weeks telling everyone to go about their lives as normal.
This bakes layers of assumptions and uncertainties into a plan that, to work, requires near-total precision and certainty.
If models are off, or you miss the window for brake-slamming by a week, or the brake-slamming doesn't work as well as predicted - it's a disaster.
If models are off, or you miss the window for brake-slamming by a week, or the brake-slamming doesn't work as well as predicted - it's a disaster.
One of my biggest takeaways from the Ebola models was that they are useful tools that can help with strategic judgment - but they are not a substitute for such judgment. Hope the UK will rethink this.
