School to jail transition: early warnings from primary school — by János Köllő and István Boza 

The answer is never simple. Family background, local opportunities, peer groups, personal choices, and chance all play a role. In our study, we focus on one part of this larger story: what can be learned from primary schools. We ask

To answer this question, we use linked Hungarian administrative data. We follow a 50 percent random sample of eighth-grade students who took part in the 2008 National Assessment of Basic Competences, and we observe whether they were incarcerated at least once by their early twenties. Imprisonment is rare: only slightly more than one percent of the students in the sample were incarcerated during the observation period.

This makes the differences we find all the more important. They do not point to schools as the sole or direct cause of later criminal behavior, but they do show that some school environments are much more strongly associated with later risks than others. The first main result is that disadvantage matters most when it is concentrated. Students from schools with both weak test results and a high share of low-status families were more likely to be imprisoned later than students from more favorable school environments. School performance was especially important among low-status schools: in these settings, better test results were associated with lower later incarceration risks. At the same time, family background mattered most in schools with weaker educational performance. In other words, social disadvantage and weak school performance appear to reinforce each other.

Several early warning signs are visible already at the student level. Students who had repeated a grade, had low mathematics grades, were classified as multiply disadvantaged, or were absent from the national competence test were much more likely to be incarcerated later. Not completing the family background questionnaire was also associated with higher risk. We interpret these indicators not as isolated causes, but as signs of a broader weakening of the relationship between the student, the family, and the school. When students are already hard to reach at age 14 or 15, later difficulties may be more likely.

Our study also highlights the role of school practices. Higher grade repetition rates, separate classes for low-achieving students, the absence of tutoring outside regular lessons, and the failure to test special education needs students for pedagogical purposes were all associated with higher later incarceration risks. These practices are not harmful by definition. Repeating a grade or grouping students by ability may be less problematic if students receive strong support. In the Hungarian setting we study, however, such practices often appear together with social segregation and limited educational resources. In this context, they may mark a school environment that pushes vulnerable students further to the margins instead of helping them catch up.

The later life paths of future prisoners also underline how early educational disadvantage can accumulate. Many left education early and most did not obtain a secondary-level qualification. Their labor market careers were often fragmented: they spent less time in regular employment, were more likely to appear in public works or casual work, and tended to hold short, low-paid jobs. This does not mean that poor schooling mechanically leads to prison. But it does suggest a chain of disadvantage: weak school attachment, early exit from education, unstable employment, and limited contact with more secure social and labor market environments.

The policy message is therefore not that schools should be treated as extensions of the criminal justice system. Quite the opposite. Our findings suggest that prevention can begin much earlier, through educational support. The data already collected by education authorities can help identify schools where risks are concentrated: schools with many disadvantaged students, weak test results, high absence rates, frequent grade repetition, limited tutoring, and signs of exclusion. These schools could be priority targets for pedagogical support, mentoring, remedial programs, and wider social assistance.

The broader lesson is that crime prevention is not only a matter for police, courts, or prisons. It is also about whether vulnerable students remain connected to school, whether they receive help before they fall behind irreversibly, and whether schools can offer credible routes toward further education and stable work. Our study shows that the warning signs are often visible years before the first prison spell. The challenge is to notice them early enough, and to respond with support rather than exclusion.