Methods: Department of education and corrections data from a southern state are used in this study. Individuals were linked across these two systems, and those both eligible for dropping out of school, and for contacting the DOC system (N = 223,992) from the period of 1999-2012 were included in this analysis analysis. From this, we tested the direct and indirect effects of predictors on incarceration using a path analysis. Due to the large sample size, we randomly sampled from the eligible dataset (n = 9,753). This mitigates the influence of inflated statistical findings while minimizing sub-sample bias. The primary predictor variables included in the model are truancy, suspension, expulsion, and whether the student failed a grade. The mediating variable examined is whether or not the student dropped out of high school. The dependent variable is whether the individual was incarcerated in the observed time period. The path analysis was conducted using a bootstrap method (1000 draws, 5000 iterations).
Results: Findings indicate that almost all direct and indirect influences were highly statistically significant, as hypothesized. Experiencing a suspension, expulsion, or grade failure led to a direct, significant increase in the incidence of both dropout and incarceration. Furthermore, dropping out was a significant mediator of suspension, expulsion, grade failure, and truancy on incarceration. The exception to expected findings occurs for truancy, which does not yield a significant, direct impact on incarceration. Taken together, the direct and indirect impacts of expulsion on incarceration through school dropout were greater than any other impacts.
Conclusions and Implications: Findings indicate that there is an important causal relationship between school engagement and behavior, school dropout (as a mediating force), and incarceration. The mediated relationship between truancy and incarceration, and the varying degrees of influence by predictor variables in statistical models have implications for programmatic and policy decisions. While all influences are important to consider, the role of school dropout as a mediating force is critical. Interrupting this pathway may be a relatively cost-effective, and feasible means of intervening to prevent the substantial individual, societal, and economic costs associated with adult criminality.