�������

In the ultimate measure of health — how long you live — it makes a big difference where you live. Especially if you are poor.

Men in the bottom five percent of the income distribution who live in New York, N.Y. can expect to live five years longer than men with comparable incomes in Gary, Ind., according to a new study led by Stanford economist Raj Chetty.

The , published in the Journal of the American Medical Association, is the first to examine how the link between income and life expectancy varies from one area to another within the United States.

"These are really enormous differences across places," said Chetty, a senior fellow at the Stanford Institute for Economic Policy Research.

To place a five-year life expectancy disparity in perspective, the Centers for Disease Control estimates that life expectancy in the United States would increase by just over three years if cancer was eliminated as a cause of death.

"You can think of the fact that there is a five-year gap between Gary and New York as if low-income people in New York don't get cancer at all while people in Gary do," Chetty said.

The researchers used anonymous data on deaths from the Social Security Administration and more than 1.4 billion individual tax records from 1999 to 2014 to obtain more precise estimates than any previous research on the differences in life expectancies among income groups in the U.S.

Being richer was associated with living longer at every level of the income distribution. And the gap between the richest 1 percent and the bottom 1 percent in the nation was vast. At 40, the richest men could expect to live to 87 while the bottom 1 percent had a life expectancy of just above 72 — equal to the average in a developing country like Sudan.

Women had longer life expectancies than men, but the gap between genders narrowed substantially at higher income levels. Women at the top of the income distribution could expect to live close to 89. The life expectancy of women at the bottom was 79 years — a 10-year gap equivalent to the drop in longevity associated with a lifetime of smoking.

Moreover, inequality in life expectancy between the rich and poor in the U.S. widened during the 2000s. High-income people gained about three years of life expectancy while the poorest saw little to no improvement from 2001 to 2014. But this statistic masked broad differences at the local level. In places like Birmingham, Ala., the poor were gaining just as much in life expectancy as the rich overall (three years over the period). But in other cities such as Tampa, Fla., the life expectancy of the poor was actually decreasing over the 2000s.

"We find very large differences across areas for the poor but very small differences across areas for the rich. Where you live matters much more if you are poor than if you are rich," Chetty said.

After taking into account differences in life expectancy that stem from ethnic and racial composition, the researchers found that some of the lowest levels of life expectancy for the poor were in the industrial Midwest.

Eight of the 10 states with the lowest levels of life expectancy for the poor formed a geographic belt from Michigan to Kansas, encompassing Ohio, Indiana, Kentucky, Tennessee, Arkansas and Oklahoma. In contrast, the poor had significantly longer life expectancies in California, New York and Vermont.

Differences in health behaviors seem to play a more important role in Chetty’s findings than measures of health care coverage and access to medical care.

"The places with shorter life expectancy tend to be places with higher rates of smoking, higher rates of obesity, and lower rates of exercise," Chetty said.

Local levels of income inequality and residential segregation by income did not seem to be associated with differences in health among the poor. Instead, measures of affluence, larger fractions of college graduates, higher share of immigrants, and greater government expenditures were strongly linked with longer life expectancy for the poor. The researchers found that poor people tend to live particularly long in cities like San Francisco and New York, which have high costs of living and highly educated populations.

"We don't know exactly why that is, but what I find striking is that the results go against the view that the poor do better in cities that are more affordable with less inequality," Chetty said.

It may be that these cities are often the first to enact public health policies such as smoking bans, restrictions on trans fats, or imposing taxes on sugary drinks – moves that affect the health of the poor as well as the rich.

It is also possible that people are influenced by others’ good habits. If there are more people around you who are exercising and eating well, you may spend less time sitting on the couch eating junk food.

The results of the study imply that the relationship between income and health is not set in stone and that there is significant room to improve the health of the poor by focusing on the issue at the local level. Health behaviors and why they vary from one area to another seem to be a particularly promising area for future research.

"There is a lot discussion about inequality and health in the U.S. as a whole but I think that conversation should be occurring at the local level," Chetty said.

The study's findings also have implications for national policies like Social Security and Medicare. The fact that the rich have longer lifespans than the poor means that low-income people are paying into the system for a long time but don't get to enjoy the benefits as long. This is particularly important to discussions of raising the retirement age.

"If we think about a policy like indexing the retirement age to life expectancy, we need to think hard about which life expectancy we are talking about. If we just use average life expectancy in the U.S. we are going to essentially start hurting the poor especially in certain areas — like Detroit — relative to the rich," Chetty said.

The study was co-authored with Michael Stepner (MIT), Sarah Abraham (MIT), Shelby Lin (McKinsey and Co.), Benjamin Scuderi (Harvard), Nicholas Turner (Office of Tax Analysis, U.S. Treasury), Augustin Bergeron (Harvard) and David Cutler (Harvard).

The study was funded in part by grants from the U.S. Social Security Administration, the National Institutes of Health, the Social Sciences and Humanities Research Council of Canada, the Smith Richardson Foundation, and Laura and John Arnold Foundation.

More information and data are available at 

Miriam Wasserman is a freelance writer.

Find more coverage of this research in: