People use life expectancy statistics completely incorrectly

You often hear things like "In the past, the average life expectancy was only 30 years".

And then people act like in that era, people had kids at 10, were old at 30, and living past 40 was really rare.

This is absurd and wrong

If that were the case, how come some Roman emperors lived to 70? They'd be living more than double the life expectancy - the equivalent of living 2*77 years today.

What's really happening is that life expectancy statistics are dominated by infant mortality. In ancient times, people who lived past childhood would regularly live into their 50s or 60s - but half of everyone died in childhood, which brought down the average.

In modern times, a small increase in the percentage of the population not having access to health care can cause a huge drag on life expectancy. i.e. having infant mortality rates go up by 10% among 10% of your population will introduce a lot of zeros to the number, and drop it. But when people see the drop from 77 to 76, they assume that hospital treatment has gotten worse, or that we are falling being, or that people are living less healthy lives.

It's arbitrary

Imagine Christians forced the US to change our definition of when life begins - now, life (& statistics keeping) would begin at conception. Infant mortality would shoot up drastically, and life expectancy would drop to 30.

Would our quality of life be lower? No, it'd be exactly the same. ( Also, everyone's age would be 9 months longer, and instead of birthdays we'd use conception days. )

Imagine if Catholics forced us to start counting all the virtual children they believe are killed by contraception. Life expectancy would drop to something like 0.001 years per "life" - but nothing would change about day to day life.

On the other end of it, imagine if we only counted children once they'd passed a "coming of age" ritual at age 13. In that world, life expectancy would be 80+ in the US, and even in the ancient world it would have been 60+. In fact, that is closer to what ancient times were like - people didn't pay much attention to young kids, because they had a lower chance to survive. They'd even delay naming them until they had a higher chance of making it.

Sea Turtle life expectancy

Out of 100 eggs that sea turtles lay, only a few baby turtles will make it into the sea - and most of those will be eaten before they make it to a year old. Say one in a hundred hatched turtles lives to one year, and all of those make it to 100 years of age. That makes the average life expectancy of a sea turtle be one year.

Does that mean they live their entire life cycle in a year? No! A one-year-old sea turtle is just a baby.

So it's foolish to use life expectancy at birth as a way to understand a life cycle. It has nothing to do with it. (The normal "life expectancy" of grasshoppers is like 1 day, since 99.9% of them don't make it to adulthood)

What we should talk about

People use life expectancy at birth to try to understand how societies work. This is wrong. They try to say that 1/2 life expectancy is middle aged, near the end is "old", etc. But for sea turtles, 10 years may be "adult" and old may be 50 years old - but we'd never get that from the knowledge that their life expectancy at birth is just a year.

Almost every sea turtle you have ever seen has wildly exceeded the average life expectancy. (But the average sea turtle's existence does last for exactly their average life expectancy; we simply don't see most of the ones with short expectancies)

This is true of humans, too - almost everyone you see will live way longer than the human life expectancy at birth.

What we should actually think about is life expectancy at a certain age. Once turtles make it through the first dangerous year, their life expectancy probably shoots way up to 60 years or so - and this number actually means something.

If you are already 20 years old, life expectancy at birth means nothing to you. You should only care about life expectancy at 20. A healthy Roman 20 year old probably had another 40 years of expectancy - even though the life expectancy of his society was only 30.


Say you live in a town that has a life expectancy of 80.

But then your township is merged with another nearby town. This other town has many people who don't have health care, so they have higher infant mortality. So when they reculaculate life expectancies for this new town, it drops to only 60. Has you life gotten any worse?

They also revise the historical life expectancies - so during your childhood, your town had a life expectancy of 60 (one of the worst in the country), down from 80.

Should you revise your opinion of what life was like at that time? You are now considered to have been born in the town with one of the lowest life expectancies in the country. Was it really terrible? You remember it was nice...

This is just a wrong way to think. In general, the weight you put on statistics should be proportional to how much they matter to you - you should care about life expectancy at your own age, not at birth, and stats about people near you, not ones who are arbitrarily included within your group.