Study: Juvenile incarceration yields less schooling, more crime

Teen offenders who serve time finish school less often, become repeat offenders more often.

Press Contact

Abby Abazorius
Phone: 617-253-2709
MIT News Office

Media Resources

1 images for download

Access Media

Media can only be downloaded from the desktop version of this website.

Teenagers who are incarcerated tend to have substantially worse outcomes later in life than those who avoid serving time for similar offenses, according to a distinctive new study co-authored by an MIT scholar. 

“We find that kids who go into juvenile detention are much less likely to graduate from high school and much more likely to end up in prison as adults,” says Joseph Doyle, an economist at the MIT Sloan School of Management and co-author of a new paper detailing the results of the study.

Indeed, the research project, which studied the long-term outcomes of tens of thousands of teenagers in Illinois, shows that, other things being equal, juvenile incarceration lowers high-school graduation rates by 13 percentage points and increases adult incarceration by 23 percentage points.

A key to the study is that it uses the variation in judges’ sentencing tendencies to analyze a large pool of otherwise similar teenagers, thus isolating the effects of the sentences on the kids in question.

“We think this is some of the first real causal evidence on the effects of juvenile detention on kids’ outcomes,” Doyle adds.

Not returning to school

A new article on the study is published in the latest issue of the Quarterly Journal of Economics. The co-authors are Doyle and Anna Aizer, an associate professor of economics and public policy at Brown University.

The study looks at cases involving 35,000 juvenile offenders over a 10-year period in Chicago. The Chapin Hall Center for Children at the University of Chicago created a database on social programs in the state of Illinois that made the research possible.  

The teenagers in question had all committed offenses that provided presiding judges with latitude in determining sentencing levels. The random assignment of judges with different sentencing tendencies to those cases allowed the researchers to conduct a “natural experiment,” as social scientists call it, examining the implications of the varying sentences.

“Some judges are more likely to have children placed in juvenile detention than others, but it’s effectively random which judge you get,” Doyle explains. “Some kids get a judge who will place them in juvenile detention, other ones get a judge who will be less likely to do so, and comparing the outcomes of the kids across the judges, we can actually say what the causal outcome is of placing the kids in juvenile detention.”

Those periods of incarceration appear to make kids much less likely to return to school at all, particularly if they are around age 16 — which appears to be a significant mechanism underlying the later problems of those people.

“The kids who go to juvenile detention are very unlikely to go back to school at all,” Doyle explains. He adds that the later problems people have may also stem from the time spent incarcerated: “Getting to know other kids in trouble may create social networks that might not be desirable. There could be a stigma attached to it, maybe you think you’re particularly problematic, so that becomes a self-fulfilling prophecy.”

“Very disruptive” to be incarcerated

Other public-policy scholars who have read the study say it is a significant contribution to the field.

“I think it’s very credible,” says Janet Currie, a professor of economics and public affairs at Princeton University, who is familiar with the study. “It’s a very disruptive thing for a juvenile to be incarcerated.” And as Currie notes, a combination of policies and social stigma also make it considerably harder for people who have been incarcerated to find steady employment, and gain the increased social stability that might result.

Currie also suggests that future research might shed further light on the subject by looking at programs offering alternatives to juvenile detention — such as psychological counseling or treatment for drug problems — and seeing if those other approaches produce better results.

For their part, Doyle and Aizer readily acknowledge that the study only represents one criminal justice system in action during one time period, and say they would welcome further examination of the issue by other scholars.

“It would be great to see more research like this done in other locations with different criminal justice systems, and see if the results continue to hold,” Doyle says.

Topics: Research, Sloan School of Management, Policy, Government, Social sciences


Well that's good that you have data to prove something that everyone has known about for a long time. Previous studies have shown that kids who commit delinquent acts but never face prosecution do significantly better than those who do the same things and face prosecution. The reality is that the initial decision to prosecute can be damaging in some cases where the child is too young to understand the nature and consequences of his actions. The United States should institute 10 as the minimum age of prosecution and allow children under 14 to demonstrate incapacity based on immaturity, mental illness and/or low IQ. When children are sentenced to youth prison, their cases should be immediately reviewed by a citizen's panel which can reduce the sentence and impose constructive restitution and rehabilitation.

The truth be told, too many judges are focused on their careers and not on the children in front of them, which can lead to many of these problems.

Lock them in schools, not prisons.

This is not true in Missouri, which is a good model for the nation to follow. Best results for the lowest costs...

Google "the Missouri Model".

The results validate what has been assumed and sometimes studied in limited ways for many years. It is an explicit and even implicit assumption of those who work with such adolescents that this would be the case. The problem is that even though we KNOW this change in detention centers/prisons creep slowly, if at all. The interface with possible solutions to their application reflects economics to only some degree. The biggest factor is the mind set/overwork/acceptance of accepting whatever-has-been-will -continue mentality. I was fired from my position as a psychologist at a very large detention center for employing methods which had demonstrated remarkable success. Very very sad. It would be wonderful to do a very well controlled study spread over several years comparing different ideas with the status quo. IT WOULD NOT BE DIFFICULT!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Well, that is probably so, but why did they rather than other go inside? I suspect that it had to do with other issues and past behavior.

I also believe that no matter whether particular kid went in or not, they would have had a worse outcome than others in the same situation.

The study was logically bent for that outcome: blame the system.

Back to the top