WHO’s Dubious Model that Claims the Real Pandemic Death Toll is 15 Million – and 5 Million of Them Are in India

First we had the Economist claiming to be able to work out how many had really died in the pandemic, then the Lancet joined in. Now it’s the turn of the World Health Organisation. While the Economist and Lancet claim the true toll is around 18 million (though find a very different distribution across countries), the WHO goes for 15 million. Once more we find that the (massive) gaps in reported data are filled in with modelling: “The methods rely on a statistical model derived using information from countries with adequate data; the model is used to generate estimates for countries with little or no data available.”

The estimates for India are particularly inflated and have drawn sharp criticism from the Indian Government. The WHO claims that India experienced 3,400 deaths per million over the two years (note the figures quoted in most reports as WHO estimates for 2020-21 are an average of the two years), which amounts to 4.69 million total deaths – almost a third of the global total. That’s nearly 10 times more than India’s official Covid death toll.

India’s official Covid death toll in 2020 is 148,994. The Government said this week that its official estimate of additional deaths in 2020 compared to 2019 is 474,806, which is 3.2 times higher than the official Covid toll. It hasn’t yet provided its estimate for additional deaths in 2021, but we know that the official Covid death toll for 2021 is 332,492. If we assume the same degree of undercounting then the number of additional deaths in 2021 would be 1.06 million. (Note that India has around 10 million deaths each year, so this represents about a 10% excess mortality in 2021.) Adding the two together gives 1.54 million additional deaths for 2020 and 2021. The WHO’s estimate of 4.69 million is three times higher than this. No wonder the Indian Government is disputing the findings.

Looking just at 2020, the WHO estimates 600 excess deaths per million, whereas the Indian Government reports 474,806, which is 344 per million – almost half the number.

For 2021, India’s official Covid death toll is 2.2 times higher than in 2020 (332,492 vs 148,994). However, the WHO’s estimate of excess deaths is 4.7 times higher for 2021 than 2020 (2,800 vs 600 per million). This means the WHO thinks India’s undercounting became more than twice as bad in 2021 compared to 2020. These are not small differences and it’s hard to see them as plausible.

Another curiosity is Germany. According to Our World in Data, Germany had excess deaths of 1,080 per million in 2020-21. Yet according to the WHO, the figure is 2,330 per million. Why is the WHO saying Germany’s excess deaths are well over double the OWID estimate? It means Germany moves from having half the excess deaths of the U.K. to being worse. Germany has very complete records of deaths, so there is no excuse for this kind of discrepancy.

Once again, these estimated, modelled figures are looking very dubious.

Like the Lancet study, the WHO’s estimates make restriction-lite Sweden look good and lockdown heavy countries like Peru look bad, so for sceptics it is tempting to see them as a good thing. And of course, it’s nice suddenly to be better than Germany.

However, we shouldn’t pick our data based on what best supports our argument but on what is most plausible and likely to be true. And as the Indian Government has pointed out, that is not this dubious set of modelled estimates from the WHO.