When things do not change: non-intact families and adolescents'risks of substance use across 30 European countries and two decades

Substantial evidence indicates that children in non-intact families experience higher risks of substance use compared to those living with both parents. The "institutionalization hypothesis" suggests that this penalty for children living in non-intact families should weaken-or even vanish-as new family behaviors become more prevalent and socially accepted. Our study tests the institutionalization hypothesis by examining the relationship between family arrangements and adolescents' susceptibility to using cannabis, alcohol, and tobacco, utilizing a unique dataset that spans 21 years across 30 European countries. We measure the diffusion of new family behaviors with a comprehensive country-year index encompassing the rise in divorces and extra-marital births, and the decline in marriages, distinguishing its between-country and within-country components. Our findings indicate that adolescents from non-intact families, either in single-parent families, stepfamilies, or no-parent families, are more likely to engage in the use of cannabis, alcohol, and tobacco. Importantly, their higher risks of substance use appear to be strikingly persistent regardless of the between-country and within-country diffusion of new family behaviors. Thus, the paper provides robust evidence against the institutionalization hypothesis.