Emily Badger is a former staff writer at CityLab. Her work has previously appeared in Pacific Standard, GOOD, The Christian Science Monitor, and The New York Times. She lives in the Washington, D.C. area.
Yes, homicide-related death rates are higher in urban areas, but you're twice as likely to die in a car crash outside a city.
In one popular explanation for the mass exodus from urban America over the last several decades, people left the city because the city wasn't safe. In suburban and rural America, by contrast, the cars drive slower down cul-de-sacs, random crime is less common, and gunfire is scarce. You've probably heard this before.
Here, however, is the data: Yes, homicide-related death rates are significantly higher in urban parts of the country. But that risk is far outweighed by the fact that you're about twice as likely to die in a car crash in rural America than you are in the most urban counties. Nationwide, the rate of "unintentional-injury death" – car crashes, drownings, falls, machinery accidents and the like – is about 15 times the rate of homicide death. Add together all the ways in which you might die prematurely by intentional or unintentional injury (as opposed to illness), and your risk of death is actually about 22 percent higher in the most rural counties in America than in the most urban ones.
All together, your risk of injury death actually goes up the more rural the community where you live.
This finding comes from a new study by researchers from the University of Pennsylvania and the Children's Hospital of Philadelphia, published in the Annals of Emergency Medicine. The study looked at every injury death in America between 1999 and 2006 (excluding death by terrorism – the researchers considered Sept. 11 too much an outlier to contribute to our understanding of these public-health patterns). That number totaled 1,295,919 deaths. Each was tagged to the county where the injury took place, with counties classified on a 10-step continuum from urban to rural.
The main finding inverts many of our assumptions about danger and place: "When considering all mechanisms of injury death as an overall metric of safety," the authors write, "large cities appear to be the safest counties in the United States, significantly safer than their rural counterparts."
On the below two illustrations from the paper, the map of population density by county (top) directly contrasts with the map of death rates (bottom):
Obviously, this is a limited way of defining safety. Plenty of things that don't kill us impact our perception of where it's "safe" to live. And this study looks only at injury death, not injuries (including attempted homicides or suicides) that simply land a person in the hospital. But given the depth of all this data, drawn from the National Center for Health Statistics, this comparison of risk is just about the most comprehensive, quantitative retort that exists to the common assumption that cities are not safe.
Across the whole population, the top three causes of death were motor vehicle crashes, firearms and poisoning. But start to break these numbers down by region, age group and even race, and the picture gets more interesting. Motor vehicle crashes, for example, lead to 27.61 deaths per 100,000 people in the most rural counties. But that number is just 10.58 deaths per 100,000 people in the most urban counties.
Other risks you might expect are more common in rural areas, like injury from machines and environmental events like flooding, animal attacks or exposure to the cold. As for guns, the risk of firearm-related death is actually pretty consistent across the country, population-wide. But firearm deaths are significantly higher in rural areas for children and people over age 45. In the city, they're much higher for people aged 20 to 44.
Race also played a curious factor. Rural counties with large black populations had lower risk of injury death than rural counties with fewer blacks. For Latinos, the pattern was the opposite.
In all of these analyses, the researchers also tried accounting for the presence of trauma centers, to determine if rural risks were higher because the injuries were worse (or more common), or because the health care was harder to find. The results changed little when they did that. But the authors argue that these findings should cause us to rethink where trauma care resources are located. Currently, injury risk is the highest in the places least likely to be served by emergency doctors and specialists.
More fine-grained research could measure the actual distance to trauma centers. The researchers also propose studying rates of alcohol and substance abuse and driving patterns to better get at why these disparities exist between rural and urban communities. In the meantime, though, this first pass at the data may change how we think about whether it's "safe" to live in the city (or it may change the reasons people use to leave it).