Caution, Caution, Caution – pretty much how you should respond to any news about Coronavirus these days, because the simple fact is that there’s so much we don’t know. And probably the biggest thing we don’t know is this: How many people are actually infected, and don’t know, and have never been tested, because they have no symptoms?

Sweden has been the European outlier in public policy since the outbreak began, with much less stringent restrictions that almost anywhere else. It hasn’t closed schools, for example. And yet, their case numbers don’t seem massively higher, as David Quinn has been pointing out, relentlessly:

But, say the Swedes themselves, that’s not the whole story:

The Swedish Public Health Agency has collaborated on a new study that estimates one-third of all Stockholm residents will have been infected with COVID-19 by May 1. That amounts to approximately 600,000 people in Sweden’s capital city. Sweden’s official figures report only 15,322 positive cases so far in the whole of the country, suggesting significant numbers of “invisible” infections.

The astonishing numbers are in stark contrast to the latest World Health Organization (WHO) estimates that “no more than two to three percent” of the world’s population have been infected.

Sweden’s acting state epidemiologist Anders Wallenstein confirmed the findings during the latest daily press conference about the virus outbreak. It was also revealed that a further 185 coronavirus-related deaths were reported in Sweden in the last 24 hours, bringing the total to 1,765.

This comes a few days after a study from Stanford University suggested that having tested for antibodies in Santa Clara county, California, the number of people who had the disease was something like 50-85 as many people as who had been reported to test positive:

These prevalence estimates represent a range between 48,000 and 81,000 people infected in Santa Clara County by early April, 50-85-fold more than the number of confirmed cases. Conclusions The population prevalence of SARS-CoV-2 antibodies in Santa Clara County implies that the infection is much more widespread than indicated by the number of confirmed cases. Population prevalence estimates can now be used to calibrate epidemic and mortality projections.

Then you have this, from lefty American statistician Nate Silver, most famous, you might remember, for predicting Presidential elections correctly. He’s a serious mind, though, so this is worth considering:

Basically, in English, infections are falling much more steeply in New York City, which has a massive, clustered population, than they are in upstate New York, which is much more rural. If the decline in infections was purely as a result of lockdown, you’d actually expect to see the inverse result, because it’s so much easier to avoid people in rural areas than it is in a metropolis like New York.

And the fall in infection rates in New York could well be explained by a rise in herd immunity.

Herd immunity is a dirty word, of course, because it’s so associated in this part of the world with an initial UK strategy that’s perceived to have failed.

But the basic concept behind it is simple enough: If enough people get the disease and become immune, it becomes much harder for it to spread. The question, from the beginning, has been “yes, but how many deaths does that entail while we build up that immunity?”. The answer to that question has been the great unknown, and the reason why most politicians have been unwilling to risk it.

There are, though, reasons to doubt the Swedes and the Stanford Boffins, as well.

For one thing, Sweden is seventh in the world in deaths per million at the time of writing – doing substantially worse than Ireland is. For another, this is an estimate – it’s not based on any testing that I can find anywhere. And the Stanford study, which did test, has been heavily criticised:

Critics claim the study’s methodology is dangerously flawed and question the political motives of the Stanford-led team. Others have pointed to the study as proof that COVID-19 is merely a partisan-driven flu hoax, as protests broke out this weekend in parts of the country over frustrations with the shutdowns.

In response, on Sunday, the study’s authors said they are planning to soon release a detailed appendix that addresses many of the criticisms and incorporates many of the suggestions into the paper itself.

And there was this, from statistician Alan Cole:

The most interesting figure isn’t Sweden, or California – it’s actually New York, where there has been no widespread antibody testing.

Someone could do the world a huge favour and conduct widespread testing there. Because this is an immensely important argument, and it’s resolution one way or the other will decide an awful lot about what we do next, and what the history books say about what we’ve done so far.