Flickr user Aapo Haapanen

The Pathways to Desistance longitudinal study followed more than 1,000 juvenile offenders in Phoenix and Philadelphia, tracking them as they left their teen years, as they grew either out of the criminal justice system or became further entangled in it. They were asked at first, as teens between the ages of 14 and 18 in the early 2000s, about their backgrounds and their crimes, their neighborhoods and their expectations. And then they were followed for the next seven years.

The study has produced a raft of research on how teen offenders transition through life and why they might offend again. The latest paper, from Alex Piquero at the University of Texas at Dallas, offers another clue mined from just one question posed in the original three-hour survey of all of these teens.

"How long," they were asked, "do you think you'll live?"

By the end of the study, the teens who expected to die young were much more likely to go on offending at a higher rate and with more serious crimes. Those who could picture themselves living to 70, or 80, or older, had the lowest rate of offending.

Piquero says the study suggests that children who fail to foresee a future around them are less likely to consider the long-term consequences of their immediate behavior. What does it matter, after all, if committing a crime now will prevent you from having a good career at 45 if you don't expect to get there?

The teens in the study who fit this description were also more likely to come from the most distressed neighborhoods, places Piquero described this way: "trash in the neighborhoods, syringes on the street, drugs selling in the neighborhoods." That picture reinforces the idea that we can't think about why people commit crime without considering the context in which they do it.

"Some of these kids only have five options to chose from," Piquero says, contrasting their future paths with other teens whose lives could head in many directions. These are not teens debating between different colleges, or different cities where they might live after college, or different careers once they get there. "These kids who envision very short lives are also having problems in many other life domains, with failed schooling, failed jobs, failed relationship, failed health care.

"These social ills can’t be really divorced from one another."

If a 15-year-old can't see himself as a parent, a mechanic or a teacher at 35, that may well say something about the quality of his school, or the existence of jobs in his neighborhood, or the extent of poverty and crime around him. The policy implication of all of this is that while mentoring might be helpful -- Piquero points to Obama's recent My Brother's Keeper initiative -- so too would solutions that tackle education and jobs in distressed communities. Both strategies could change how disadvantaged children think about the future.

"How do we get kids to see past their constrained environments and social situations," Piquero asks, "to know that there is something that might be better out there?"

In these environments, dying young wasn't an unrealistic expectation for the teens answering the original question. By the time the study was over, 45 of them had died.