2 ms·
You missed the major factor that they ultimately didn't do anything about: frailty. "Because this mortality risk occurred earlier in history for the 1900 birth
by mcguire 13y ago
You missed the major factor that they ultimately didn't do anything about: frailty.
"Because this mortality risk occurred earlier in history for the 1900 birth cohort than for the 1930 birth cohort, the 1900 birth cohort faced higher probabilities of death at the ages between birth and 85. Thus, individuals surviving to age 85 in 1985 may have been more robust than individuals surviving to age 85 in 2015, because it was more difficult to survive to age 85 for the former group."
In actuality, no they didn't try to factor out socioeconomic factors that shouldn't matter (like those you mention) from those that should (like better nutrition). On the other hand, they have a large sample size (I can't find it, but from one aside it looks like around two million records) and they're only comparing two groups: above average and below average. I would imagine boundary cases like your first example would average out.
"One important contribution of this study is to highlight that the segment of the male Social Security–covered worker population experiencing slower mortality improvement is large—that is, the entire bottom half of the population, rather than just a limited group of disadvantaged at the lowest end of the earnings distribution."
And a 4.6 year difference would seem to be pretty significant.
- AnthonyMouse 13y ago> In actuality, no they didn't try to factor out socioeconomic factors that shouldn't matter (like those you mention) from those that should (like better nutrition). You don't think occupation has a significant effect on lifespan? Someone whose career involved breathing coal dust, spraying lead-based paint or pumping leaded gasoline, manufacturing products with asbestos or heavy metals, etc. is significantly more likely to suffer health problems later in life. Even just having to work two jobs and not getting enough sleep will catch up with you over the course of 40 years. > On the other hand, they have a large sample size (I can't find it, but from one aside it looks like around two million records) and they're only comparing two groups: above average and below average. I would imagine boundary cases like your first example would average out. But that's the whole problem. You don't know if they average out or not, and if they don't in the population at large then no sample size is large enough to save you.