Researchers explored the long-term effects of preschool expansion in Japan in the 1960s, revealing significant reductions in risky behaviors amongst teenagers. By analyzing regional differences in the rollout of the program, the study identified links between early childhood education and lower rates of juvenile violent arrests and teenage pregnancy.
The findings suggest that improved noncognitive skills played a key role in mitigating risky behaviors, highlighting the lasting benefits of early-education policies.
Given the complexity of human existence, it can be difficult to accurately define which aspects of youth have certain impacts later on in life. Interventions and experiments to explore theories are nearly impossible due to the long times involved. It is possible to compare different populations, but depending on the matter investigated, there might often be too many confounding aspects to produce reliable data from which to draw conclusions.
So, when exploring a matter like the downstream impact of educational interventions, researchers ideally need multiple nearly identical populations, different only in the presence of the intervention in question.
Building upon previous research which found that day care in Japan improves noncognitive skills of very young children around age 3, my colleagues and I wished to investigate whether these benefits persist over time and if early childhood education can genuinely change behavioral outcomes later in life."
Shintaro Yamaguchi, Professor, Graduate School of Economics, University of Tokyo
Yamaguchi added, "Our latest study examines Japan's 1960s preschool expansion and its long-term effects on adolescent risky behaviors. By leveraging regional variations, we found that increased preschool enrollment significantly reduced violent crime arrests and teenage pregnancies, highlighting the lasting impact of early childhood education, even in a society with low baseline rates of such behaviors."
The team's findings are particularly significant because they come from a universal program that served all children regardless of socioeconomic background. Most previous evidence on the crime-reducing effects of early childhood education comes from small-scale targeted programs for disadvantaged children in the United States.
"A key challenge was ensuring that the effects we observed were genuinely caused by Japan's preschool reform and not by other factors. Since different prefectures expanded preschool at different rates, we needed to rule out alternative explanations for changes in juvenile crime and teenage pregnancy," said Yamaguchi. "To address this, we used a simple but effective approach: We examined adult crime and pregnancy rates before and after the reforms. If preschool had a real impact, the effects should appear only in children who attended, not in older individuals who missed out on the reform."
What's also interesting, and possibly counterintuitive in some ways, is that this preschool rollout program did not increase high school or college enrollment rates. Yamaguchi and colleagues suggest the mechanism behind the reduction in risky behaviors was likely improvement in noncognitive skills rather than additional schooling. This insight is important for understanding how early childhood education produces its long-term benefits.
"Our next step is to further investigate the mechanisms through which early childhood education affects adolescent behaviors. While our current research suggests that noncognitive skills play an important role, we would like to gather more direct evidence about which specific skills are most influenced by preschool education and how they develop over time," said Yamaguchi. "Additionally, we're planning to investigate even longer-term outcomes into adulthood, including effects on health behaviors, family formation and intergenerational outcomes. Understanding these broader and longer-term impacts would provide a more complete picture of the value of investing in early childhood education."
Source:
Journal references:
Ando, M., et al. (2025). Universal early childhood education and adolescent risky behavior. International Journal of Public Economics. doi.org/10.1016/j.jpubeco.2025.105353