The New York Times has been printing some of Chetty’s results on longevity for lower and upper-income households, which find that lower-income individuals live longest in expensive coastal places like New York and LA. Paul Krugman’s reaction was, “Public policy Matters!” Steve Sailer’s reaction was that you have to be pretty tough to be poor in New York, so the less tough (and more likely to die) folks leave. My reaction was that a large portion of poor people in New York and LA are immigrants, and immigrants have longer lifespans on average than native-born Americans.
I don’t have Chetty’s longevity data (let-alone his race-and-income adjusted longevity data), but I do have the crude 2013 death rates by county for middle-aged non-Hispanic Whites , used for Case & Deaton’s paper. The overwhelming majority of non-Hispanic white Americans are native-born, so this gives us a clue how native-born death rates (and implicitly, lifespans) vary with immigration. In addition, since these death rates appear to be correlated with support for Trump, and Trump’s campaign has been about “the immigration, stupid” it’s interesting to see how the two factors- immigration and death rates for middle aged non-Hispanic whites- are directly associated or not.
In general, immigrants are in healthier areas of the country:
This is essentially due to immigrants clustering in areas with higher incomes, lower obesity, and lower disability rates (which together explain the large majority of variation in death rates):
Conditional on these factors however, areas with higher percentage foreign born have somewhat higher death-rates for 45-54 year old non-Hispanic whites:
I don’t want to read too much into this (even if this whole blog is about reading too much into things.) I think it does suggest how, from the commanding heights of the economy, the anger over immigration seems misplaced: the areas with more immigrants are doing pretty well for native-born whites, by health as well as income standards. On the other hand, conditioning on health and income factors, areas with more immigrants have somewhat higher death rates (a doubling in the percent foreign born in a county is associated with a 4% increase in non-Hispanic white middle-aged death rates, holding income, obesity rates, and disability rates constant.) This could well be attributable to unobserved differences in cost of living or density (higher percent foreign born counties are also higher population, although higher population itself doesn’t explain differences in death rates once income, obesity, and disability are taken into account.) And it’s perhaps debatable whether obesity and disability are “exogenous to the model”- public policy, the economy, and immigration can all plausibly affect them, through various means, as well.