There's something about Atlanta, Chicago and Detroit that traps children in poverty. Poor kids who grow up there are less likely to reach the middle class as adults than poor kids who grow up in, say, Salt Lake City, according to a now-famous study released three years ago that found wide geographic variation in economic mobility across the country.
Precisely what that something is, though, the authors couldn't say.
Patrick Sharkey, a sociologist at New York University, suspected part of the answer: These places have pockets of deep violence and high crime rates, and the fraying neighborhoods that come with them. In Chicago, Sharkey had found that black children performed worse on vocabulary and reading assessments when a homicide occurred in their neighborhood in the previous week. If violence impairs how children learn and navigate life, he figured, it could be a key to economic mobility. "All of the evidence generated to this point," Sharkey says, "suggests that it should be."
In new research with Ph.D. student Gerard Torrats-Espinosa, Sharkey is now making the bold argument that higher levels of violent crime lower children's prospects for upward mobility, explaining a significant part of the geographic variation in that landmark 2013 study. Their work is among the first to claim what's causing differences across the country in mobility, filling in some of the mystery created by the original research.
Violence itself has harmful effects. But Sharkey and Torrats-Espinosa are saying that violent places do, too, and that the environments created by recurring homicides and assaults — whether or not children are the direct victims of it — hinder their lives in ways that become apparent in their economic standing as adults.
Using data shared from the original 2013 study, led by economists Raj Chetty and Nathaniel Hendren, Sharkey and Torrats-Espinosa found that poor children across the country have worse outcomes in more violent places, like Chicago and Detroit. They have higher economic mobility in safer places like Fairfax County, Va.:
But they also discovered, within individual counties, that children who reached their crucial teen years at more violent times fared worse than children growing up in the very same places during periods of greater peace.
Their research exploits the fact that the children in the study came of age at the beginning of a sweeping national decline in crime. And they further test their theory by tracking changes in economic mobility as crime spiked during local crack epidemics, and as it waned following federal grants for more local policing. Poor children whose teenage years coincided with the peak of the crack epidemic in Chicago, for instance, lived through an intensely violent period in the city's history. These children achieved less mobility later in life than children who grew up in Chicago just a few years later.
In all the ways Sharkey and Torrats-Espinosa sliced their data, the analyses pointed to the same pattern: More violence meant poor kids were less likely to advance up the economic ladder, measured by their own incomes at age 26. It's hard to imagine that something else explains the consistent link they observed. That's why they argue that changes in violence aren't just associated with differences in mobility rather, they cause these differences in mobility.
Hendren, one of the authors of the original economic mobility study, says he was impressed by the findings. The most convincing piece, he says, is the discovery that adding more police — through federal grants given out by the Community Oriented Policing office created in 1994 — resulted in greater upward mobility. Of course, all policing strategies aren't equally beneficial (and some may be harmful), but this point begins to hint at how we could wield specific policies to improve odds for poor kids.
All of these findings notably apply to poor kids but not wealthy ones. Changes in violent crime don't appear to affect the economic outcomes of kids who start life closer to the top. This makes sense given previous research by Sharkey and others showing that the biggest drops in crime since the 1990s have occurred in the poorest, most segregated neighborhoods.
This economic mobility data is divided by county, not by neighborhood. But there's reason to believe that the kids living in the poorest neighborhoods within these counties were affected by these changes in violence the most. The original data, based on tax records, also isn't broken down by race. But other research also suggests that changes in violence disproportionately affect black children.
"Everything that I’ve done, everything that I’ve seen leads me to conclude that the strongest benefits — even though we can’t see it in this dataset — have occurred for African-American kids," Sharkey says.
But why exactly would less violence mean more upward mobility for them? Sharkey and Torrats-Espinosa offer two arguments. The first is that exposure to violence has direct effects on children: It heightens their stress levels, distracts them in class, disrupts their learning and impairs their cognitive development.
The children in this data who were exposed to less crime weren't more likely to attend college. But they were less likely to drop out of high school, suggesting in some ways they may have done better in school as a result (the income gains from having a high-school diploma as opposed to none at all could also explain improvements in economic mobility for kids who started off near the bottom).
The second explanation is that violence breaks down entire communities, stripping them of the things that poor kids need to get ahead, like good schools and job networks.
"Violence undermines community life as a whole," Sharkey says. Violence makes it less likely that business will open, providing jobs; that good teachers will choose these schools; that public spaces will be safe; that families will invest in the neighborhoods around them.
The children in the earliest cohort in this study, born in 1980, experienced American cities at such a low point.
"Urban communities around the country fell apart during this period," Sharkey says. "In the late '80s, when we’re looking at these kids growing up, kids were living in war zones. They were living in places that looked like they were bombed out, where people had retreated from the streets, where businesses wouldn't dream of investing and opening up a store, where institutions were dilapidated."
Children born even a few years later reaped the gains of the decline in crime. This suggests that disparities in economic mobility would be even more stark today if Chicago, Detroit and Atlanta remained as violent as they once were. Although Atlanta is still one of the most violent places in the country, the homicide rate there has fallen by more than 50 percent from its peak in the early 1990s. In Chicago and Detroit, homicides have dropped more than 25 percent.
As Sharkey and Torrats-Espinosa resolve some of the mystery around why differences in economic mobility appear, their work will inevitably point back to another riddle: There is no definitive explanation for why violent crime has declined so dramatically across the country since the 1990s. The best answer is a set of theories, and many of them are contested. But Sharkey argues that we can still pursue individual ideas that we know work — like summer jobs programs or Obama's favorite Becoming a Man intervention — without holding the answer to that larger question.