Estimating China’s Omicron Death Toll

Triangulating multiple data sources points to a number around 6 million

View from China with an Austrian School of Economics Perspective

Just when we thought that Covid-19 was passing into the mists of history, the NY Times editors apparently decided to remind everyone that there are still a few mysteries in need of clarification.

One of these is China’s overall death toll due to the Omicron tsunami which swept through China in December 2022.

In its article, the NY Times uses public reports on the deaths within the ranks of China’s Academy of Engineering and its Academy of Sciences, many of whom are elderly. These show an increase of over 300% in death rates. While such data cannot be assumed to be representative of the society as a whole, it does provide a data point. It also strongly hints that substantial excess deaths took place.

So how many really died due to Covid-19?

In our previous article on the lessons to be learned from China’s massive Covid control experiment we mentioned data from a Beijing-based government institute with approximately 1,000 retirees over 65. As of early January they had recorded 21 deaths. Now that number is up to 35. The normal annual death toll is about 20. If we take these numbers as a proxy and assume a somewhat elevated monthly death toll during the winter months, then we get excess deaths of approximately 3.1% due to the Omicron tidal wave. If for the sake of argument, we assume this number to be representative, then we should be able to calculate out total excess deaths by multiplying this number (3.1%) with China’s total estimated population aged over 65.

If we put these numbers together, let’s see what we get.

According to figures released by China’s National Bureau of Statistics in May 2021, 13.5% of mainland China’s total population or 190.6 million people were 65 or older. This data can be found both on Wikipedia as well as on the Statistics Bureau website; the source data is here. 3.1% of this number would represent 5.9 million excess deaths, or around 0.42% of the population. This of course disregards deaths among those under 65; it also disregards Covid-related deaths which took place prior to December 2022. However, we can guess that these represent at most a very minor portion of the total.

On average around 867,000 people die each month in China. If we factor in a winter-related increase of 12% (based on US numbers), we get an expected value of 1,942,000 deaths over the course of December and January. An additional 5.9 million additional deaths would amount to an increase of 304%, or, to put it another way, a quadrupling of the normal number of deaths.

Is this number plausible?

Given that we don’t have any comprehensive data, we can only take the data points we do have and compare them.

We have the above data from Beijing which point to an increase of approximately 300%. We have the data compiled by the NY Times estimating an increase in deaths among the elderly amounting to over 300%. We have significant anecdotal accounts of the death of grandparents in our circles of colleagues and friends. And as noted in that previous article, we have information from the crematoria in Shanghai. These are so backlogged that as a rule, multiple corpses are now being cremated together, thus commingling the ashes. It is hard to imagine such drastic measures being taken if deaths were only, say, double the normal volume.

How would this 0.42% number compare to statistics on Covid-related deaths from other countries?

If we look at, say, the United States, total cumulative deaths recorded as being Covid-related currently amount to approximately 0.33% of the total population of 336m.

If we look at the European Union, the numbers are similar: 1.21 million out of a population of 447 million, or 0.27%.

While both are somewhat lower than the 0.42% estimate for China, they are at least in the same range. Of course unlike in China, these deaths were spread out over the course of over two years, leading to a societal impact which was likely far less palpable.  

On the other hand, if we look at Japan, we can calculate a total Covid-related mortality rate of only 0.056% (69,770 recorded deaths out of a total population of 125,472,175). This is a tiny fraction of the percentage recorded for the United States and the EU. Was Japan simply recording its data differently, for example by excluding deaths associated with co-morbidities? Multiple explanations are imaginable, but at a minimum this data would seem to be an outlier.

Our best judgment: The estimate of 0.42% is at least plausible.

Follow us on Twitter @AustrianChina.