Earlier this week, New York State conducted a study in an attempt to determine how much of its population has been infected with COVID-19. Around 3000 people from 19 counties across the state were tested to see if they had developed antibodies designed to attack the virus, which the body will usually develop once it has been infected.
The study estimates that in New York City, the epicenter of the virus in the US, around 1 in 5 people had antibodies in their systems, with this estimate being nearly 1 in 7 across the state. This would imply around 2.6 million people have been infected across New York State, which is around ten times the number of cases officially reported there.
Before I go much further, I should point out that the number of infections, and the number of deaths, are both going to be understated to some degree. The number of infections will be understated as people who are self-isolating with virus symptoms would not have been able to take part in the study, and the number of deaths will be understated as it only includes hospital deaths and not deaths elsewhere. It’s hard to know which of the two will be a bigger understatement.
The study suggests that around 1 in 140 people infected with the virus will die. Given the reported figures from countries that are very good at testing (notably Australia, where their reported death rate was around 1 in 100 as of Tuesday), this figure is probably not too far off the mark.
If around 1 in 140 people infected die from the virus, we can quickly estimate how many people have been infected with the virus across Europe based on their respective death tolls.
There have been nearly 116,000 deaths across Europe, which on our estimate would suggest that over 16 million people across the continent have been infected with the virus. While this may sound like a lot, it only represents around 3% of the population. Of course, this varies from country to country, with the countries facing the earlier outbreaks having the highest infection rates.
In any event, even if we assume that presence of the antibodies implies immunity from the virus and from the ability to spread it (which currently is unclear), and even when we allow for a degree of under-reporting of deaths and cases, there is no prospect of herd immunity developing anywhere in Europe.
There is a wide margin of error for this study, and a number of countries (including the UK) will be attempting similar studies in an attempt to get a better estimate of infection rates. And while this knowledge is highly unlikely to bring an end to the lockdown, it could help to inform what a lockdown-lite might look like.