We are not "living longer"

[Igel]

48. Except that the perception may be skewed

but it wasn't wrong.

When you look at Egyptian burial grounds, for instance you find a hell of a lot of bodies from those who didn't make it though their 30s and 40s. If they weren't dropping dead, why were they buried? That helped make the average life expectancy in ancient Egypt what it was.

Infant mortality was high. Dispose of that, life expectancy increases.

But malnutrition, disease, accidents, war and other kinds of violence led to early deaths for a lot of people.

One part of the flawed perception is that "average" means "everybody." If the average age of death is 40, nobody should live to 41. That's foolishness, but not for the reasons in the OP. Mostly if you managed to get to 50 you would live to 70. For workers, the poor, they tended to drop like flies before that. When you get to 50, you have adult kids to help out and don't have to fend for young'ns, and soon after that you no longer have to fend for your parents (most likely), and things get a bit easier. Fertility drops of and so maternal mortality rates drop. You're no longer of military age. You've seen most diseases pass through at least once.

Even for the period that the OP says most increases in life expectancy were from childhood mortality decreases there was a good increase in expetation of life. That's increased since then. Fewer wars, less disease, etc.

Expectation of life =/= life expectancy at birth. I can look up my expectation of life now. Let's say I'm 55.

http://www.cdc.gov/nchs/data/nvsr/nvsr64/nvsr64_11.pdf says for me it's 25.5. That's average. 50-50 chance I'll live longer. Above median family income, above median education says "better than that." Sedentary lifestyle says "below median." Father died at 85, not from natural causes; mother, at 88, is dying from Alzheimer's. No way to know how it'll average out for me.

My life expectancy at birth was far lower.

For many of the subgroups with lower expectations of life--African-Americans, for instance--some of that could be rectified through lower crime-related fatalities or better compliance with health-care directives. (It's not nearly so much access as it is just doing things that in general promote healthier outcomes or just following physician's instructions. This is one place "more education = better health, on average" comes into play in a big way.) Latinos have better longevity (and lower infant mortality and lower maternal mortality) for reasons that nobody's sorted out. Yet. They pretty much trash the "low income = early death" scenario that works for other groups, though, and forms the basis for a lot of political rhetoric.)