Early childhood parenting and adolescent bullying behavior: Evidence from a randomized intervention at ten-year follow-up

Abstract

Rationale: Bullying in education is a significant public health problem with long-run life cycle consequences, but efforts to reduce bullying were met with mixed results. We ask whether improving early childhood parenting can reduce subsequent bullying behavior by the children. Method: We analyze ten-year follow-up of a randomized intervention where the parents of preschool children received a training aimed at improving parenting techniques, such as disciplinary strategies. 280 parents in 17 preschools were cluster-randomized at the preschool level (intervention group, cluster = 11, n = 186 vs. control group, cluster = 6, n = 94). Their children were followed ten years later, at ages between 12 and 16, with overall attrition rate of 12.5%. Ordinary least squares is used to estimate the intent-to-treat effects, controlling for the baseline characteristics of the child and the mother. Multiple imputation and inverse probability weighting are used to account for attrition and wild-cluster bootstrap test is used for statistical inference with few clusters. We estimate treatment-on-the-treated effects using two-stage least squares to account for noncompliance in the intervention group. Finally, we examine the associations between bullying and externalizing behavior. Results: Adolescent children are significantly less likely to bully their peers if their parents received the parenting training when their children were in preschools. The effects are stronger for boys and for aggressive forms of bullying such as 'beating' and 'threatening'. Treatment-on-the-treated effects are greater in magnitude and more significant than intent-to-treat effects, suggesting that the effects are driven by program participants. No effect is found for cyberbullying and for victimization. The results were robust to using multiple imputation and inverse probability weighting. We also find suggestive evidence that the effects on bullying are not explained by changes in externalizing behavior. Conclusion: We show that improving early childhood parenting can have important public health benefits through reduction in adolescent bullying behavior.