Association for Public Policy Analysis & Management

We seek to estimate the fraction of the racial achievement gap that can be explained by differences in incarceration patterns of adults by race. Adults may be parents or role models that have a positive influence on children during their development. The thought experiment is as follows: suppose incarceration policy disproportionately removes black and Hispanic fathers or role models from the lives of their children (relative to fathers or role models of white children). What impact does this have on the relative academic achievement of their children as they proceed through school?

To answer this question we will build a panel of counties/jurisdictions where state/federal prisoners were sentenced (i.e., where they lived before incarceration) and relate variation in the prison population to variation in academic achievement years later. We estimate counts of sentenced prisoners by geography and race from the National Corrections Reporting Program and the Federal Justice Statistics Program. To measure academic achievement we use the Stanford Education Data Archive (SEDA), which includes detailed race and achievement data by school district across the United States from 2009 to 2015. In addition to estimating descriptive relationships, we also use variation in sentencing policy across geographies and time to identify the causal impact of policy changes.

This work makes important contributions to two literatures. First, it adds evidence to the growing literature on the causes and correlates of the racial achievement gap in education. If parental/adult incarceration is related to the racial achievement gap, adjustments to criminal justice policy could have implications beyond the offender. For instance, the removal of mandatory minimum sentences for crimes that disproportionately impact black or Hispanic offenders will allow parents and role models to remain in the household and their children to thrive in school.

The results of this study also speak to an additional consequence of incarceration of adults. If incarceration of parents or role models negatively impacts the educational attainment of their children, this could imply future social costs if their children later participate in taxpayer-funded social welfare programs or become criminals themselves (though estimating those costs is beyond the scope of this work). If it exists, such a cost should be included in the calculus of broad policies regarding incarceration for the marginal offender.