You seem to be basing all your arguments on national averages. That data shows none of the variances in life expectancy depending on race, class, region, and income.

Other measures of the social ills correlated with being very poor? In cities, if you are also black or hispanic , a high chance of death by violence. A relatedly high rate of incarceration, often for nonviolent crimes. High rates of children in the foster system.

Just because the averages nationwide are slowly ticking up over time, does not mean that there are not significant populations slipping backwards.