The idea that poor education is related to crime is pretty intuitive: good teachers are less likely to want to serve in high-crime areas, poor education makes finding non-criminal employment more difficult, crime and poor education both cluster around low-income, high-poverty neighborhoods, etc. But does tackling one problem help address the other as well?

A  new study by economists Costas Meghir of Yale, Mårten Palme of the University of Stockholm, and Marieke Schnabel of University College London suggests that yes, you can kill two birds with one education reform stone. They looked at an education reform introduced in Sweden in 1950. Prior to the reform, sixth graders were sorted based on grades into a track that went on to middle and high school, preparing them for college afterward, and a track that received only one or two years more education. The reform replaced this with a system where all students went to middle and high school to prepare them for college. The Swedish government tried the new policy in a sample of school districts, selected to be representative of the country as a whole. Comparing the outcomes of students in this sample and students outside it, then, allowed researchers to isolate the effect the reform had on students' life outcomes.

A previous study by Meghir and Palme found that students whose fathers had been sorted into the non-college track earned more and went further in school under the new system, while students whose fathers were in the college track made slightly less, as they had to compete with students who a generation previous wouldn't have made it past middle school. Overall, though, students made more money, and education outcomes were considerably more equal. The new study tries to determine whether these improvements translated into lower crime.

The answer: It sure did, and the results lasted a long time. The authors find that students in the reform group were 5 percent less likely to be convicted of a crime than students in the old system, and the effect was stronger if a student's parent was on the non-college track. But, more startlingly, the children of students affected by the reform were less likely than the children of students not in the reform group to get convicted of a crime. The effect is small -- the children were 0.6 percent less likely to be convicted of a crime, and 1.02 percent less likely if their grandfathers were on the non-college track -- but it translates into a 2.5 percent reduction in overall crime. This is especially surprising given that these children don't earn more or get more education than students whose parents weren't affected.

So what's going on? The researchers investigate a few possible explanations. Perhaps being more skilled and having more money allowed students to move out of high-crime neighborhoods where their children would have been more susceptible to criminal elements. But the researchers find that moving from low-income to high-income areas didn't affect the likelihood of a reform beneficiary's child engaging in crime.  What the economists did find evidence for is the theory that the reform made students more capable parents and role models, resulting in their children avoiding crime. The reform students' children are avoiding crime, the researchers conclude, because of "increased parental resources and improved paternal role models."

It's important not to extrapolate too wildly from studies like this. It may be that these crime-related gains are unique to the kind of reform introduced in 1950s Sweden, which, given that the United States makes kids stay in school until they're 16 today, is not exactly implementable here. It could also be that these gains are due to the structure of Swedish society, which, especially in 1950, was much more ethnically and culturally homogenous than the United States. But if nothing else it's a good reminder that education reform can have effects far beyond just educational outcomes.