Loading...
 
Skip to main content

No Food for Thought

Did the 1918–1920 flu pandemic really reduce life expectancy by a decade?

admin Friday August 8, 2025

Towards the end of Factfulness (published 2018), Hans Rosling gets to "The Five Global Risks We Should Worry About". On page 237 (in chapter The Urgency Instinct), the very first one he mentions is a global pandemic. If he was still alive, he would have been extremely proud to have made that call by now.

But I could not help being skeptical of the introduction, which starts with the following claim:

The Spanish flu that spread across the world in the wake of the First World War killed 50 million people—more people than the war had, although that was partly because the populations were already weakened after four years of war. As a result, global life expectancy fell by ten years, from 33 to 23 […].

Did the Great Influenza epidemic really reduce life expectancy by 10 years⁉️ For sure, 50 million casualties is huge, but how could it reduce life expectancy by 3/10 in a world with a population of 1.8 billion?
The short answer is: No.
The more elaborate answer is that "life expectancy" is based on expectations, so it depends on what your expectations are. The way statisticians compute life expectancy is by projecting from the yearly mortality rate:

The demographic concept of life expectancy measures how long the average person would live if current death rates by age applied throughout their lifespan.

And of course, the high mortality rate from these pandemic years was short-lived. In other words, the way life expectancy is computed makes it misleading during temporary mortality shocks.

Life expectancy has its strengths and is useful when discussing the future, but when looking at the past, it is confusing and not the best metric. A better choice would have been to discuss the average lifespan, which would have avoided this confusion―or should we say exaggeration, Pr. Rosling? 😉 Oddly enough, that example merely reinforces the book, providing a great example to back what he writes a chapter earlier (The Blame Instinct, p. 211):

Our press may be free, and professional, and truth-seeking, but independent is not the same as representative: even if every report is itself completely true, we can still get a misleading picture through the sum of true stories reporters choose to tell. The media is not and cannot be neutral, and we shouldn’t expect it to be.