When Americans think about government policies that promote social equality, they tend to think about policy that targets the poorest. That’s how we get Head Start for disadvantaged kids (and not universal preschool), Pell Grants for college students from lower-income families, and other means-tested benefits. Then there are other government programs that are available to all, like Social Security and Medicare, but that are structured to redistribute income in favor of lower-wage workers at older ages.
Widely publicized work published at the end of April in the Journal of the American Medical Association by Stanford economist Raj Chetty and his colleagues challenges the way we think about social welfare policies in two distinct ways. First, they suggest that local public policies that apply to everyone in the area explain more about mortality differences between income groups than access to health care among the disadvantaged does. Second, they show that universal programs supporting the elderly provide more benefits to the rich than to the poor.
How does their work suggest conclusions that many might find counter-intuitive? Very carefully. For starters, they linked deidentified tax records to Social Security Administration death records: 1.4 billion records from 15 years. The income levels of those who didn’t file tax returns were determined from W-2s and unemployment benefits. In other words, the researchers had a lot of hard data about income and survival in the United States. Next, they showed that the life expectancy gap between the richest 25% of the U.S. population and the poorest 25% was more than 10 years from 2001 to 2014.
Let that sink in for a moment: Not only is income a matter of life and death in the United States, it makes a difference of more than a decade (closer to 15 years for men). There’s a lot people can do in a decade. Alice Park, writing for Time magazine, put it this way: “Money may not buy you happiness (or love) but it might just buy you more time to find them.”
Perhaps even more sobering than this disparity in life expectancy is the fact that the advantage of the rich has grown over time: Life expectancy improved about two and half years more in the top quarter of the income distribution than it did in the bottom quarter from 2001 to 2014. Again, don’t let the statistic go by too quickly: It isn’t just that income inequality has grown in our country; Chetty and his colleagues show that the consequences of income inequality in terms of life and death have also intensified.
Which, of course, raises the question of what we can do about it. This is where the study holds a few surprises. The data suggest that local policies (e.g., city-wide bans on smoking in public) may explain more about mortality differences between income groups in a given area than seemingly more likely candidates, such as the proportion of residents without health insurance or the local degree of income inequality.
Local policies may explain more about mortality differences between income groups than seemingly more likely factors.
While their data do not prove the efficacy of particular policies, they nonetheless build a compelling case when they show that income matters less for life and death in places that actively promote healthier lifestyles, but that geographic differences in the longevity of the poor were not significantly correlated with access to medical care, physical environmental factors, income inequality, or labor market conditions. In other words, the mortality gap between the rich and the poor varies a lot from one area to another, but that isn’t due to things like medical care, pollution, and unemployment rates. I, for one, would have thought that there would be bigger gaps where more of the poor weren’t getting routine medical care and where more of the poor were unemployed. Instead, gaps were larger where there were fewer immigrants, fewer college-educated people, and lower government expenditures.
I believe these three factors are closely linked. Much of the press coverage of this study has focused on the last two, speculating about how high proportions of college-educated people and relatively large government expenditures translate into more equal survival chances. Some believe college-educated people model healthy behaviors and support legislation restricting public smoking and taxing sugary drinks. Other local policies that are not directly aimed at promoting health like universal preschool and paid sick leave may also reduce health disparities by easing the stress associated with low-income work.
It seems to me that areas with higher proportions of immigrants may do well for similar reasons. One of the leading explanations for why immigrants are healthier than the native-born despite their lower socioeconomic status is that they have low rates of smoking. They may also eat better than those who have been exposed to the American diet for longer. Even though I readily admit that immigrants’ health advantage has other sources, it seems clear that role-modeling of healthy behaviors doesn’t have to come from the college-educated.
The study suggests that areas where healthy lifestyles are directly and indirectly promoted have smaller gaps in years lived between the rich and the poor, and I think that conclusion is supported by all three of the local-level factors the authors identify. That means promoting good health for everyone can help the poor—perhaps even more than targeting health policy toward the poor alone. More generally, when Chetty and his co-authors compiled life expectancy disparities for 714 different areas in the United States, they showed that inequality had been reduced substantially in some of them over time. Therefore, it is not immutable.
Finally, if we step back from the local findings and reconsider the authors’ most basic result—that richer people outlive poorer people by over a decade—it becomes easy to see why universal programs for the elderly provide more benefits to the rich than to the poor: The rich draw benefits for a longer period of time. This does not mean that Social Security and Medicare are more important to them; these programs may in fact serve as a social safety net supporting the health and the longevity of the poor. But it does mean that even though the formula that determines monthly Social Security benefits favors the poor, total benefits are tipped back in favor of those who live longer.
Plus, it seems quite reasonable to assume that an important reason why lower-wage workers are more likely to start drawing benefits at age 62 rather than full retirement age is that they are in poorer health. Monthly benefits are substantially smaller when they start earlier, so health disparities associated with income again make the system less progressive.
This work highlights the fact that the rich live long lives everywhere in the United States, but that there is substantial variation in the lot of the poor. It challenges us to think in new ways about what kinds of policies hold the most promise for closing the gap.