The number of people who have died from covid-19 is expected to be around 17 million
To see all of our estimates of the true death toll from the pandemic, check out our tracker.
WCOVID-19 POOL first hit northern Italy in February 2020, The Economist noticed that something strange was happening. There were 1,140 deaths attributed to the virus in parts of the province of Bergamo in March, for example, but the overall increase in the number of deaths compared to the same period the previous year was 2,420. Similar phenomena were quickly noticed in America, Great Britain, Spain and other countries.
The official death toll from Covid was – and continues to be – inaccurate. They typically rely on tests for the virus, which can be spotty, and use definitions that exclude the collateral consequences of the pandemic, such as additional deaths in overcrowded hospitals. Fortunately, there is a better way to calculate the death toll. To measure those from events such as wars, researchers don’t check how many dead have bullets in them, but on something much more reliable: how many more dead in these than we know? would wait there in normal years. This is called “excessive deaths”.
The gap between official and reliable counts of covid-related deaths means that while the global toll officially passed 5 million on November 1, the actual number of people lost to the virus is, The Economist‘s count, closer to 17m.
Undercoverages are much higher in the poorest countries. In Bulgaria, deaths from the pandemic are probably almost twice as high as the official tally. In Russia, official statistics seem to underestimate the actual number of deaths from covid by a factor of 3.5; in India, the government figure underestimates the death toll by a factor of ten. In addition to leading to a misunderstanding of the true spread of the virus, this phenomenon means that the impact of the pandemic is most underestimated in countries with the least resources to fight it (see map).
There are still a few reasons to use official death tolls. They can show trends over time within countries, for example. But all too often, they are used to compare rich and poor countries or to estimate the number of deaths in the world, measures for which they are unnecessary. They are however still in use because they are more up to date than OCD and unlike them they are available for all countries.
The Economist has found a way to remedy this situation and is the only source of daily updated estimates of excessive deaths worldwide. To find the true toll of the pandemic, The Economist first collected mortality data for all countries where it was available (with the help of researchers and volunteers), then built a machine learning model to predict it where it is unknown. (We have made these estimates, along with our data, code, and models, available for free, and have presented our work to the World Health Organization, among others.) Our estimates, at the time of writing, find the global death toll most likely to be 16.8m, with a confidence interval between 10.3m and 19.5m. The most likely date for which deaths from the pandemic topped 5 million fell in the first week of December 2020.
The estimates of our models are subject to uncertainties, sometimes enormous, especially in the places where the data are available. But the uncertainty around a correct number is both justified and appropriate when the data is imperfect. It’s also better than the alternative: complete trust in one number is sure to be wrong. ■