We can ask this question about life expectancy first for people in the bottom quarter of the income distribution and then for people in the top quarter of affluence.
According to Stanford economist Raj Chetty’s paper, the poor live longest where there is massive economic inequality, lots and lots of cops, and unaffordable housing: e.g., New York City. Health care access doesn’t much matter to the poor’s life expectancy. Social conservatism and social capital doesn’t matter either. In other words, the poor appear to do best in some kind of plutocratic Giuliani-ville. Ayn Rand would feel vindicated (if she cared about the poor).
(I don’t actually believe this is true in terms of policy advice: I think Chetty’s result may well be an artifact of churn of his populations: healthy young poor immigrants move to ultra-expensive cities like NYC until they are used up, at which point they leave for some place cheaper.)
In contrast, the top quarter of income Americans live longest in economically more equal and socially conservative places, with broad health care access, fewer immigrants, and non-supercharged economies: kind of like Denmark.
From Raj Chetty’s new paper based on your confidential tax returns, The Association Between Income and Life Expectancy in the United States, 2001-2014, here are the correlations between life expectancies for people in the bottom 25% of national income on their 1040s and various characteristics of their “commuting zones” (e.g. extra-large metro areas). The highest life expectancy metro for people in the bottom quarter of income is New York, followed by Santa Barbara, San Jose (Silicon Valley), Miami, Los Angeles, San Diego, and San Francisco. In other words, the poor live longest in super expensive cities with lots of rich people, lots of economic inequality, and lots of cops.
The shortest life expectancy metros for bottom quartile individuals are Gary, Las Vegas, Oklahoma City, Indianapolis, Tulsa, and Detroit.
For example, unsurprisingly, there’s a strong negative correlation between the % of residents who smoke and the life expectancy of lower income residents. (Keep in mind, though, that these are not individual-level correlations. Chetty has individual-level data from 1040s on income and whether or not the individual died in the last year. He doesn’t have data on whether the individual taxpayer smokes, is obese, exercises, or what not. So, he’s correlating individual level data on income and age at death with local averages, such as smoking.)
In summary, local customs regarding health behaviors (smoking, obesity, exercise) are very important for the poor’s life expectancy. Measures of access to health care don’t correlate well with life expectancy.
Income inequality and income segregation by neighborhood are modestly good for the life expectancy of poor people.
Prosperity and population growth aren’t very important.
The unimportance of the % Black Adults figure is an artifact of Chetty presenting to us race/ethnicity adjusted figures. Blacks have shorter life expectancies than, say, Asians, but Chetty has already adjusted life expectancy in each metro area for its racial makeup. (But there are racial interaction effects that he’s missing, which drive some of his outliers.)
Black life expectancy, fortunately, has been improving since NWA broke up, with fewer black on black homicides and fewer deaths from AIDS. Asian life expectancy is expectedly high, while Mexican life expectancy is unexpectedly high. White and American Indian life expectancy has been doing poorly in this century, especially working class and/or Scots-Irish whites.
The Other Factors section at the bottom give away what’s going on in driving local areas’ life expectancies among bottom quartile income individuals. Places with high median home values, which correlate with having lots of college graduates, attract lots of immigrants who come to work hard for the relatively high wages available to people willing to sacrifice living space or short commutes for some period of years. These “sojourners” tend to be healthy and live a long time. Like the man said, if they can make it there they can make it anywhere.
If they can’t make it in an ultra-expensive city, such as because they are in poor health, they tend to leave for some place cheaper. Less expensive cities tend to fill up with people either shed from expensive cities or daunted from even trying.
That’s one reason the Charles Murray / Robert D. Putnam community virtues don’t matter much in this graph: poor people appear to live longer if there is a lot of dynamic churn in the economy, which disrupts community social capital and makes life more economically unequal.
But, we don’t know if that’s a genuine treatment effect or if it’s just an artifact of churn in the population: if somebody moves from Mexico to New York City and sleeps in a bunk bed while hustling at two busboy jobs, they’re probably in vigorous health. But if their health breaks down in NYC and they move to relatively short-lived San Antonio for an easier life, where they die early there, how does Chetty count that?
In contrast, for people in the upper quarter of the national income distribution, the contributory factors for longer life expectancies (other than smoking, obesity, and exercise work) work quite differently.
The longest lived people in the top quarter of income (already adjusted for race) are found in the region around tee-totaling Salt Lake City (no surprise), hardy Portland in Maine, Spokane, Santa Barbara, Denver, Minneapolis, and quite Dutch Grand Rapids: the more conservative part of a Stuff White People Like list of places to live.
For the well-to-do, the worst life expectancy cities are Las Vegas, which you’ll be Leaving a half year earlier than anywhere else, Gary, Honolulu, Brownsville, El Paso, Bakersfield, Miami, Lakeland in FL, and Los Angeles: very Stuff White People Don’t Like.
So, looking at correlations, a culture of non-smoking, non-obese, and exercise helps the affluent as well as the poor.
After that however, things diverge: Health care access measures matter more for affluent than the poor, paradoxically.
The well to do benefit health-wise from Murray-Putnam socially conservative social capital and greater economic equality. The well to do don’t last long in boom towns. Having a lot of immigrants around isn’t good for the upper quarter, but having a lot of college graduates around is good.
In other words, white people tend to do best in the more unfashionable SWPL places like Salt Lake City and the other Portland. Perhaps that shouldn’t be surprising.