|
I’m dubious of the papers. The Lancet paper uses data only up to 1st May 2020, so really not much. One uses the Imperial College model, which, after looking at the published source code, I am near certain is gibberish. My biases aside, peer review showed that it is very sensitive to input parameters, which is incidentally what the linked paper says about lockdown effects, so hard to take seriously IMHO. Another says that after subtracting the effect of minor restrictions, the major restrictions had no measurable effect. But surely harsher restrictions are only brought in when rates are going up a lot, so maybe these two effects cancel out? Everyone, except Australia and NZ, avoid harsh lockdowns as much as they can. Meanwhile, in the UK “2nd lockdown”, which was quite light, the overall R coef for Covid dropped below 1, but remained above 1 for the “British variant”, clearly showing that had the lockdown continued as it was, it would not be effective. Also there are just so many conspicuous timings with new cases peaking 1-2 weeks after harsher lockdowns are introduced, I think it’s hard to dismiss out of hand. Yeah, it’s technically possible that the causality isn’t there but I struggle to believe that. |