By Christopher Monckton of Brenchley

Some good news: more than a million people have now recovered from the Chinese virus.

And some more good news: lockdowns are being unwound by little and little. Even HM Government, which has moved with all the vim, dash and rapidity of a glacier flowing uphill over a vat of superglue, is talking of setting out an unlocking plan sometime next week. Maybe. Once it has had a nice cup of tea. Here are the dates on which various territories locked down, and the dates on which some began to unlock:

Georgia, one of the last states to go into a strictish lockdown, is among the first to unlock. The Governor, Brian Kemp (Republican) has issued a down-to-earth, practical, quite detailed and very clearly-explained unlock strategy. Here are a couple of slides encapsulating some aspects of that strategy. More at his website:

It is not only the “Democrats” who are beside themselves with fury, on the ground that there may be a second peak if the state is unlocked. Health professionals are also muttering into their beards. But the Governor is banking on people following the rules he has set out, and using their common sense. Georgia, then, will join Sweden as one of the places to watch.

Some more good news (h/t Mosher, who has kindly been supplying first-class information on the pandemic). Research by the London School of Hygiene and Tropical Medicine shows that lockdowns have discernibly halted the infection’s exponential spread in some countries, though not in all. Take the United States:

Mr Trump declared a state of emergency on March 13. Four key states – New York, California, Illinois and New Jersey – locked down between March 19 and March 22. About a week later, a peak in new infections (which the School estimates occur a couple of week before the cases were reported) was reached in the U.S.A.

Looking at six populous states, the lockdown had no apparent effect in California, Illinois or Massachusetts, and the peak in Pennsylvania was ten days after the lockdown, but in densely-populated New York and New Jersey the peak was reached within a week of the lockdown.

In the UK the lockdown came into full effect on March 24 and the peak in new infections was on April 4, 11 days later. However, the half-dozen most-affected regions all showed near-immediate peaks following the UK-wide lockdown:

One notable feature of all these curves of daily cases is that, though the approach to the peak is steep the decline from it is slower. The reason is that lockdowns delay the acquisition of “herd immunity” and, therefore, the symmetrical shape of the curve either side of the peak that, as my good friend Willis Eschenbach has rightly pointed out, is characteristic of a pandemic following the logistic curve does not arise.

For contrast, here is Sweden, which has not locked down at all. The School thinks a peak has been reached nonetheless:

One more piece of good news: our daily graphs show that in the United States estimated active cases (on the cautious, weekly-averaged basis) are at last declining.

Fig. 1. Mean compound daily growth rates in estimated active cases of COVID-19 for the world excluding China (red) and for several individual nations averaged over the successive seven-day periods ending on all dates from April 1 to April 30, 2020.

Fig. 2. Mean compound daily growth rates in cumulative COVID-19 deaths for the world excluding China (red) and for several individual nations averaged over the successive seven-day periods ending on all dates from April 8 to April 30, 2020.

Ø High-definition Figures 1 and 2 are here.

Sweden now has the highest growth-rate in estimated weekly-averaged active cases among all the countries we are following, and the third highest death-rate. Its Public Health Authority is no longer holding daily press conferences, and the director of the Authority is currently preparing a report on why so many people have died in care homes (a problem that has afflicted Britain and many other European countries, with the notable exception of Germany).

Finally, I apologize for having mangled yesterday’s equation (1). I explained that once the deaths are falling by one-nth per day, on the assumption that deaths will continue to decline at that rate, one can estimate the total deaths T from any day d simply as the product of n and that day’s deaths m. The corrected equation for the sum of the relevant infinite series is:

Ø A growing number of commenters here are providing valuable information about best practice in public policy and in approaches to treatment of the virus. Keep this information coming and I shall feature the best information here from time to time, as I have today.