Testing Social Support Theory: A Multilevel Analysis of Recidivism

PurposeTests of social support theory have relied on aggregate crime rates as the outcome of interest, but such a focus ignores the potentially important macro-level processes and effects on individual-level behavior We thus perform the first multi-level investigation of social support theory.MethodsMultilevel modeling is used to explore whether the two varieties of county-level social support – the presence of charitable organizations and AFDC expenditures – are associated with recidivism in a sample of Florida prison releasees.ResultsResults show that while social support explains little variation in individual-level recidivism, a combination of private and public social support may reduce the likelihood of reconviction for drug offenses.ConclusionFindings provide mixed evidence for the prospect that social support—whether governmental or nongovernmental—is associated with recidivism among recently-released inmates.

► A multi-level test of social support theory focusing on individual-level recidivism.
► We test both public and private sources of social support.
► Social support explains little variation in individual-level recidivism.
► Interaction of support types reduces the likelihood of recidivism for drug offenses.