The Road to Desistance: The Relationship between Formal Institutions of Social Control, Informal Social Bonds, and Intermittency in Offending

ABSTRACT

The current study examines the relationship between formal social control, informal social bonds, and intermittency in offending. Using data from a Canadian population who have been formally flagged as representing a disproportionate risk to public safety because of their frequent and serious offending over the life course, (N = 376), we use multiple-failure survival analysis to examine whether formal social control (probation supervision) and informal social bonds (to employment, family, and other prosocial community supports) are related to intermittency in offending. We also investigate whether social bonds moderate the relationship between probation and intermittency. Findings indicate that probation, employment and romantic relationships reduce the hazard of reoffending. Longer periods of probation were related to a lower hazard of reoffending for those who have been employed for longer than six months. While probation length has an effect on intermittency, the mere presence of social bonds reduces the hazard of reconviction. Given the importance of prosocial relationships outside of formal systems of control, we argue for policies that are non-punitive, particularly since lengthy periods of probation can also extend system involvement, which can stall the desistance process.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ Given the imperfect nature of these data, if there was incomplete information about any given ACE, the person with the missing information was coded as not having been exposed; this is consistent with previous studies that have used the ACE score (see Felitti et al. 1998; Chapman et al. 2004; Dube et al. 2001).

² The log-likelihood test was not significant in any of the five models (Model 1: LR test =2.21,${\chi }^2\left(\mathit{df}=1\right)$=2.21, p= 0.137; Model 2: LR test =2.06, ${\chi }^2\left(\mathit{df}=1\right)$=2.06, p= 0.151; Model 3: LR test ; ${\chi }^2\left(\mathit{df}=1\right)$=2.24, p=0.135, Model 4: LR test ${\chi }^2\left(\mathit{df}=1\right)$=2.19, p=0.139; Model 5 LR test ${\chi }^2\left(\mathit{df}=1\right)$ =2.01, p=0.157) indicating that the addition of age-squared did not improve model fit; as such the models we present here do not include age-squared.

Additional information

Notes on contributors

Tamara Humphrey

Tamara Humphrey is an Assistant professor in the Department of Sociology at the University of Victoria. Her interests focus on how formative life events and structural disadvantage impact trajectories of offending and desistance from crime.

Grace Li

Grace Li is a PhD candidate in Sociology and Statistics at the University of Victoria, Canada. Her research interests include demography and quantitative methods. She is particularly interested in developing and evaluating methods for modelling life transitions and trajectories.

Erin Gibbs Van Brunschot

Erin Gibbs Van Brunschot is Professor of Sociology and Vice-Dean of the Faculty of Arts at the University of Calgary. Her primary research interests are in the realms of risk, criminal behaviour and social control, with specific interests in how individual, organization/agency and state orientations to crime and security diverge and converge. Her recent work has included a focus on offender management, technology and policing, and disinformation regarding security issues.