Criminal exposure in vulnerable areas

• Andresen, Anette Marie & Skardhamar, Torbjørn (2026). Rethinking Criminal Onset: Age at First Offence, Persistence, and Sex Differences. Journal of Developmental and Life-Course Criminology. ISSN 2199-4641. 12(1). doi: 10.1007/s40865-026-00315-x. Full text in Research Archive Show summary

  How frequent is criminal onset at each age, and to what extent does onset age predict continued offending across the life course for men and women? Criminological theory and empirical research put emphasis on young offenders, as early onset is a clear risk-factor for persistence, and often assume later onset implies lower risk of continued offending, yet the prevalence and persistence of adult onset remain disputed. It is also contentious whether there is any meaningful onset after reaching adulthood. While it is clear that the risk of onset declines with age, the dominant empirical approach in the literature is to identify the most common trajectory groups, using methods that are less likely to identify segments of unusual offending patterns, such as onset after normative ages. Using nationwide administrative records of solved offences for the resident population of Norway, we provide descriptive accounts of first registered offences in 2010 and subsequent continuation across all ages (N = 22 610; observation window 1982–2020). We find registered onset across all ages for both men and women, and while the risk of continuation declines with age, a non-ignorable number of offenders continue after onset even in older ages among both sexes (21% among men and 23% among women with onset at age 40 or older).

• Friestad, Christine & Skardhamar, Torbjørn (2025). Criminal Careers in the Making? Characteristics and Trajectories of Juveniles Charged With a Sexual Offense. Sexual Abuse. ISSN 1079-0632. doi: 10.1177/10790632251393988. Full text in Research Archive Show summary

  The present study investigates the onset, persistence and diversity in sexual offending by exploring Norwegian registry data on the social and criminal trajectories of juveniles charged with a sexual crime before the age of criminal responsibility, compared to juveniles charged with other types of offenses. The sample consisted of all persons aged 15 during the period 1997–2005 (N = 388,814). Of those, 19,559 juveniles (5%) were charged with a crime, mostly (83%) of a non-violent nature. For the remaining 16.9% (n = 3,309), the charge involved violence, either non-sexual (n = 2,991, 15.3% of those charged) or sexual (n = 318, 1.6% of those charged). Proportional hazard models were used to investigate background characteristics and risk of recidivism. Early onset of offending was related to a more disadvantaged background. Juveniles charged with violent and sexual offenses were generally younger and had more charges against them than juveniles charged with any other crimes. Early sexual crime carried higher hazard rate (HR) of future sexual crime (HR = 3.6) than early violent (HR = 1.9) and other crimes (HR = 1.4). The risk of later violent crime was similar among those with early onset violent (HR = 2.8) and sexual (HR = 2.5) crime, and higher than for early onset general crime (HR = 1.7). Since violent crime has a higher base rate than sexual crime, violence is more dominant in the criminal careers of people with early onset sexual offending. However, the majority of adult sexual criminal charges are raised against persons without a juvenile criminal record.

View all works in NVA

• Skardhamar, Torbjørn (2025). Utsatte områder i Norge? Full text in Research Archive
• Andresen, Anette Marie & Skardhamar, Torbjørn (2025). Crime throughout the life course: onset and continuation in offending. Full text in Research Archive
• Skardhamar, Torbjørn (2025). Machine learning and risk prediction of violence - theoretical considerations. Full text in Research Archive
• Skardhamar, Torbjørn (2025). Kriminologisk livsløpsperspektiv. Full text in Research Archive
• Skardhamar, Torbjørn (2025). Svenske tilstander_. Full text in Research Archive
• Andresen, Anette Marie & Skardhamar, Torbjørn (2025). Age of onset and continued offending. Full text in Research Archive Show summary

  This study aims to provide new evidence of age of onset and the amount of continued offending after onset at different ages. Early onset has been identified as a clear risk-factor for continued offending. Individuals with a later onset is associated with lesser risk for continued offending, but to what degree is uncertain. This uncertainty may be the result of a long standing discussion of whether late or adult onset exists at all. Defining late or adult onset has been proven difficult, where one reason may be the variety data used in the efforts to investigate the phenomenon. This study utilize total population data from Norway to investigate the existence of onset across all ages. We do so in an attempt to overcome some of the challenges in previous studies. The data include official offence data from 1992 to 2020. In this study we explore criminal onset among all full time Norwegian residents above the age of 10 with their first registered charge in 2010. The dataset also allows for descriptive information on continued offending after onset at different ages. The result confirm criminal onset is most common during late teens/early adolescence. Criminal onset, however, occurs in all ages, although less common with the increase in age. Similarly for continued offending, continued offending is most common among earlier onset, but there are instances of quite serious careers at later age of onset.

• Skardhamar, Torbjørn (2025). Life course trajectories vs vriminal career trajectories. Full text in Research Archive Show summary

  Individuals’ offending over time is often described using trajectory modeling, with offending considered as a function of age. This approach follows naturally from a developmental and life course perspective. Alternatively, offending trajectories can be described as a function of time since the onset of offending, thereby avoiding the modeling of criminal careers before any observed offending behavior has begun. In this paper, I demonstrate how these different approaches can lead to diverging results, and argue that chronological age is not always the most relevant time scale. This empirical analysis utilizes total population registry data from Norway.

• Frith, Michael James (2025). Age Matters: Using Age Windows to Measure Offending · CrimRxiv. doi: 10.21428/cb6ab371.d38781a0. Full text in Research Archive Show summary

  While long-term population register data in the Nordic countries offer unique opportunities for criminological research, systematically recorded and linkable crime data are often limited to more recent and shorter time periods. As such, criminological research, particularly intergenerational crime research, often rely on observing offending within relatively short age windows, for example, up to age 30, as a substitute or proxy for measuring longer-term offending. This study uses Norwegian population register and Police charges data to assess how offending measured at specific ages and within three-, five- and ten-year age windows corresponds to broader involvement in offending (between ages 15-42, 20-47, and 25-52). While offending at individual ages shows relatively weak correspondence to longer-term involvement in offending, Pearson correlations show that offending during the late twenties and early thirties, particularly the 10-year window covering ages 20–29, are strongly linearly related and can explain over 80% of variation in long-term offending totals. In contrast, the rank-order and prevalence of offending correspond less strongly but relate most closely at earlier ages (e.g., 15–24), however, these measures depend much more heavily on longer observation windows. The findings highlight both the utility and limitations of using age-limited observation windows in crime research.

View all works in NVA