Moving from one US metro area to another can increase a 65-year-old’s life expectancy by 1.1 years, according to an innovative study coauthored by MIT economist Amy Finkelstein.

Researchers have long observed significant regional variation in life expectancy in the US, often attributing it to “health capital”—different tendencies toward obesity, smoking, and related behavioral factors. But by analyzing health records of Medicare recipients who did and did not move from one region to another, Finkelstein and her colleagues were able to separate the role of health capital from the role of environment itself.

“The idea is to take two elderly people from a given origin—say, Boston,” Finkelstein says. “One moves to low-­mortality Minneapolis; one moves to high-mortality Houston. We then compare how long each lives after they move.”

All told, the study found that urban areas including New York, San Francisco, Miami, and Chicago have positive effects on longevity while a large swath of the South and Southwest have negative effects. These results deliver important new information about large-scale drivers of US health outcomes—and raise the question of what it is about different places that affects older people’s life expectancy. One clear possibility is the nature of available medical care. Others include climate, pollution, crime, traffic safety, and more.

The scholars estimate that health capital accounts for about 70% of differences in longevity while location effects account for about 15%. “Yes, health capital is important, but yes, place effects also matter,” Finkelstein says.

She and colleagues are now working on a pair of new studies about health care to see what impact place-based differences may have; one study focuses on doctors, and the other looks at the prescription opioid epidemic.