Earlier first arrest predicts persistent offending

Falk Ö, Wallinius M, Lundström S, Frisell T, Anckarsäter H, Kerekes N (2013). The 1 % of the population accountable for 63 % of all violent crime convictions. Social Psychiatry and Psychiatric Epidemiology.

Longitudinal · N = 2,393,765 · early-onset window within a general population cohort

Among 93,642 individuals with >=1 violent conviction, 1.0% of the population (persistent violent offenders, >=3 violent convictions) accounted for 63.2% of all violent convictions. A first violent-crime conviction before age 19 was an independent predictor of persistence in violence (OR 2.0, 95% CI 1.9–2.0), alongside male sex, personality disorder, and substance use disorder.

Bearing on this claim: Nationwide register cohort: a first violent conviction before age 19 independently predicts persistence in violent crime (OR 2.0).

Cauffman E, Beardslee J, Fine A, Frick PJ, Steinberg L (2024). Trajectories of offending over 9 years after youths' first arrest: What predicts who desists and who continues to offend?. Journal of Research on Adolescence.

Longitudinal · N = 1,216 · adolescents

Latent-class growth analysis identified four trajectories after first arrest: stable low (55%), escalating (23%), short-term recidivist (15%), and persistently high (7%). Younger age at baseline (first arrest) distinguished the escalating/recidivist/persistent groups from the stable-low group, which was the oldest; age at first arrest was among the baseline factors predicting continued offending, alongside peer delinquency, exposure to violence, and callous-unemotional traits.

Bearing on this claim: Prospective first-arrest cohort: younger age at first arrest distinguishes escalating/persistent from stable-low offending trajectories over 9 years.

Hein S, Barbot B, Square A, Chapman J, Foley Geib C, Grigorenko EL (2017). Violent Offending Among Juveniles: A 7-Year Longitudinal Study of Recidivism, Desistance, and Associations With Mental Health. Law and Human Behavior.

Longitudinal · N = 58,678 · adolescents

In the complete population of Connecticut court-referred youth, violent offending peaked at ages 14–15; a high-rate adolescent group (3.7%) accounted for 31.9% of all violent offenses, and 74.2% of that group desisted in adulthood. Earlier age at first offense directly predicted more violent offending and reoffending: each 2.4-month decrease in age at first offense raised the log count of violent offenses by 1.18 and the log odds of violent reoffending by 0.317.

Bearing on this claim: Complete court-referred population: quantifies earlier age at first offense as a direct predictor of violent offending and reoffending; documents that most high-rate offenders desist.

Farrington DP (2019). Childhood risk and protective factors for early desisters, late desisters and life-course persistent offenders. Revista Española de Investigación Criminológica.

Longitudinal · N = 411 · childhood to mid-adulthood (prospective cohort)

In the CSDD (411 males followed age 8–56), childhood individual and school risk factors — notably low popularity and low school attainment — were the most important predictors of whether an offender up to age 20 persisted past that age or desisted; family protective factors (good child-rearing, high parental interest in education) most strongly protected offenders from becoming life-course-persistent. Of offenders still active at ages 40–56, 68.5% had first been convicted between ages 10 and 20, indicating that early-onset careers dominate long-term persistence. Directly rebuts the argument that childhood factors are not predictive of later persistence vs. desistance.

Bearing on this claim: Cambridge Study (prospective cohort to age 56): childhood risk/protective factors predict persistence vs. desistance, and early-onset (age 10–20 first conviction) careers dominate offenders still active at 40–56 — a fourth independent cohort corroborating early onset as a persistence marker and adding the protective-factor dimension.