Strong differences in Individual vs population impacts
Surprisingly, life expectancy is closely monitored, but life-span variation isn't
Double burden of inequality
Dying earlier + greater variation in time of death, whereby the latter is mostly dying much earlier than at life expectancy of your cohort, which is more pronounced in low socioeconomic groups.
First row: Life-expectany increases, better educated/richer/non-manual workers life longer
Second row: life-span variation increases as well but more so for manual workers, in the lower income & education percentiles
[all for Finns]
Policy implication #1: invest in health & social policies to reduce mortality in lower socioeconomic groups
Policy implication #2: invest to drive down (premature) midlife-death will reduce life-span variation
Nice summary, let the average (life span) be experienced by many.
JvN directs the readers' attention to the danger if a mathematical discipline moves farther and farther away from its empirical source or any empirical content at all.
He qualifies this if that researchers with an "exceptionally well-developed taste" could make this a success.
If the distancing from empirics happens, he predicts a fraying of the discipline "into a multitude of insignificant branches" and it "becomes a disorganized mass of details and complexities." Ultimately the discipline faces danger of degeneration due to "abstract inbreeding".