Life expectancy in the 1800s not as bad as reported

I can't count how many times I've seen people write about how life expectancy has gone up dramatically because of modern [medicine|sanitation|technology|whatever]. Often they report (as this site does) (edit: here's another) that the average life expectancy has increased to near-80 from something under 50 (43 in the first example, 62 in the 1930s in the second). For some time I've wondered whether this was life expectancy at birth (and therefore affected by infant mortality), which would be quite different than everyone keeling over before their grandchildren are out of diapers. This site gives the answer:

Life Expectancy by Age, 1850–2004 — Infoplease.com.

...and it's clear that infant mortality was the issue in the 1800s, not general early death. Looking at white males (the first table) the life expectancy at birth in 1850 was indeed a dismal 38.3 years. But that's incredibly misleading.

At age 10 the life expectancy had increased to 58.0 years.

The final number goes up for each subsequent column, but the big jump in life expectancy is from 0 to 10:
  • 0 to 10 -- 19.7 years
  • 10 to 20 -- 2.1 years
  • 20 to 30 -- 3.9 years
  • 30 to 40 -- 3.9 years
  • 40 to 50 -- 3.7 years
  • 50 to 60 -- 4.0 years
  • 60 to 70 -- 4.6 years
  • 70 to 80 -- 5.7 years.
The difference in life expectancy for a newborn and a ten year-old doesn't mean that a newborn could actually expect to live an average of 38.3 years. It means that a newborn faced two distinct possible futures: he had somewhere between a 60 and 66% chance of living 58 years; and he had somewhere between a 34 and 40% chance of not living to see double-digits. The variance is due to the unknown average age of death for under-10s. If they all died as infants, then roughly 34% died that way. If they all died as 9 year-olds (far less likely it would seem) then 40% of them died that way. In any case, it's fair to say that the majority of the reported prolonging of life is due to reducing childhood mortality.

Comparing just life expectancy for 20 year-olds, in 1850 a young man could expect to live to 60.1. In 2004, that same man could expect to live to 76.7. That's a significant improvement, but considering that in 1850 the germ theory of disease was just being formalized, it seems a little less impressive.

Posted