Intergenerational Risks of Criminal Involvement and Incarceration
- Rucker Johnson, Goldman School of Public Policy, University of California, Berkeley
- Goldman School of Public Policy Working Paper (April 2007)
This paper provides nationally-representative estimates of the cumulative risks of
incarceration and obtaining a criminal record by age 40 for a cohort born between 1951-
1975. I show that men born in the 1960s/early 1970s have significantly greater
cumulative lifetime risks of imprisonment than similarly-aged men born in the 1950s.
This is in part a direct consequence of the transformation of incarceration and sentencing
policy that took off in the 1980s. The racial disparities in lifetime incarceration risks are
alarming. The results highlight that among black low-educated men, one-half either died
or had been incarceration before the age of 40.
Second, this analysis uses an innovative approach to investigate the relative
importance of family background and neighborhood context on deviant behavior over the
life course, including ever being expelled, criminal involvement, ever being incarcerated,
the early formation of risk preferences, and risky health behaviors. Particularly
noteworthy, the analysis of brother and male child neighbor correlations in adult
incarceration history revealed remarkably high correlations of 0.69 and 0.54,
respectively. These results highlight the profound influence that family and/or
neighborhood background has on criminal involvement and risks of imprisonment.
Moreover, the results suggest that neighborhood quality during childhood is a significant
gatekeeper of the intergenerational transmission of deviant behavior and incarceration
risks among males.
Third, this study examines the intergenerational consequences by examining
children of the next generation. I find, using the PSID-CDS data, that the prevalence
rates of parental incarceration at some point during childhood are significantly larger than
point-in-time estimates. I find that 20 percent of black children had a father with an
incarceration history; and among black children with fathers who did not graduate from
high school, an alarming 33 percent of their father’s had an incarceration history.
Fourth, this study is among the first longitudinal child-outcome studies that
examines the role of pre-incarceration risk factors and children’s living arrangements,
parent-child relationships and substitute caregiver-child relationships, to help determine
the impact of parental incarceration on families and children.
I find linkages between exposure to parental incarceration and child behavioral
outcomes. The pattern of results is remarkably similar across all of the empirical
approaches utilized that address omitted variables bias—including hierarchical random
effects models with an unusually extensive set of controls, family fixed effect models,
child fixed effect models, and instrumental variables estimates. This study bears
evidence on the extent to which parental incarceration has exacerbated racial disparities
in childhood and in early adulthood.
Download a PDF (449KB)