There’s some data on this. The late Angus Maddison devoted his life to assembling estimates of demographic changes and economic growth. The OECD published a roundup near the turn of the millennium.
Life Expectancy at Birth, Group A Countries: (Japan, Europe and Europe’s Offshoots)
1000…24
1820…36
1900…46
1999…78
So between 1000 and 99 we gained 54 years. 59% of that happened in the last 100 years – so it seems that antibiotics and the advance of medicine is a big part of human progress in life expectancy.
19% occurred between 1820 and 1900 – together that implies that 78% of the progress has occurred since 1820. That’s the bulk of the story, though of course the remaining 22% (1000-1820) would be interesting to learn about as well.
In 1976 the US published the Bicentennial edition of Historical Statistics of the United States: it’s available online at the Census website. Table B 126-135 shows life expectancy profiles for Massachusetts – we can sample 1850 and 1900-02. How does life expectancy at birth compare with life expectancy at age 20?
Years of remaining life at various ages: Massachusetts
Age 1850 1900-1902
Male Female Male Female
0 38.3 40.5 46.1 49.4
20 40.1 40.2 41.8 43.7
40 27.9 29.8 27.2 28.8
Sheesh, this is tough dataset to interpret. Females in Massachusetts live longer than males as early as 1850. From what I understand, females were increasingly valued in New England during post-Colonial times as literacy rates increased, because the mother traditionally was responsible for the kids’ education.
Anyway, the data seems consistent with the infant mortality reduction story. Total gain in life expectancy at birth was 7.8 years for males and 8.9 years for females. After age 20 males gained only 1.7 years while females gained 3.5 years. But after age 40 there’s a reversal! Life expectancy at age 40 drops after 1880 for both genders. Is this an artifact of the data? Perhaps the effect was this driven by immigration, migration or shifting urbanization rates. Maybe we could google for another dataset.
For the moment, we might note that 60% of the progress in life expectancy has occurred in the past 100 years or so, and that all ages of benefited. But before 1900, it might have mostly been an infant mortality story. Data! We need more data!