Association for Public Policy Analysis & Management

This paper considers the extent to which education affects crimimnal participation of the next generation and argues that previous literature has underappreciated the extent to which investments in human capital play an enduring role in reducing crime. In particular, we study intergenerational effect of parents’ education on crimes committed by the next generation. The key to disentangling the causal effect of parental education on their children’s propensity to commit crime is to identify an exogenous source in variation that affects parental education without directly affecting their abilities or directly affecting potential determinants of their children’s outcomes. Exploiting changes in compulsory schooling laws that occurred between 1914 and 1974 in the United States, which affected, in particular, the lower tail of the socioeconomic distribution, this research evaluates whether children of parentsaffected by the policy are less likely to engage in delinquent behavior than children of parents not affected by compulsory schooling laws.

This study contributes to the literature of the intergenerational effects of education in four ways. First, to the best of our knowledge, this is the first research that studies the intergenerational effects of education on crime in the United States. Second, our outcome of interest is self-reported delinquency, which includes delinquent acts that are more prevalent among adolescents than crimes observed from administrative data on arrests which represent only a fraction of crimes committed. Third, this study explores whether parental education affect potential behavioral and non-cognitive outcomes among the children of parents who were affected by more strict compulsory schooling laws, as potential mechanisms.

The empirical results suggest that compulsory schooling laws decreased the propensity to damage property among their children by 1.8 percentage points (relative to the average share of respondents who report having damaged property of 18%), decreased the propensity to assault by 1.3 percentage points (relative to the average share of respondents who report having committed assault of 27%) and the probability to shoplift by 1.3 percentage points (relative to the average share of respondents who report having shoplifted of 17%). These effects are important in comparison to standard policy mechanisms used to control crime – crime elasticities are larger, for example, than elasticities of crime with respect to police or the prison population.

We address several potential mechanisms through which compulsory schooling laws may affect criminal behavior of their children such as higher human capital transmission, lower fertility which results in more resources per child, or through parental attitudes towards crime. We find that increasing parental education results in a cohort that has fewer siblings, one that expects to ultimately obtain higher levels of education, one that watches less television and one that feels more in control over their lives. On the other hand, we do not detect differences in parental investments in formal training as proxied by summer educational activities. Interestingly we also do not find that children of more educated parents get more sleep, a mechanism which has been linked to a range of negative outcomes among poor youth in the psychology literature.