The relationship between childhood physical abuse and juvenile delinquency

Abstract

Intergenerational research in criminology often focuses on the cycle of violence. Despite the increased attention to intergenerational transmission processes, there is often disagreement about the measurements used and the analytical methods applied. Using data from the German longitudinal prospective panel study Crime in the Modern City (CRIMOC), the current study sought (1) to make a statement concerning the extent of the participants’ self-reported experiences of physical violence in childhood as committed by their own parents, (2) to examine the connection between the participants’ experience of violence and subsequent delinquency in adolescence, and (3) to determine how the definition or operationalization of physical abuse has influenced results regarding the reported connection with delinquency. The findings illustrate the suspected connection between physical abuse and delinquency or violent delinquency, especially for those who have experienced severe violent treatment by their parents. These results remain stable even with different definitions and operationalizations of physical abuse.

Keywords:

• Cycle of violence
• juvenile delinquency
• life-course
• physical abuse
• violence

Disclosure statement

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Notes

1 Wetzels (1997) used the following 10 items: (1) attacked with a thrown object, (2) attacked vigorously or pushed, (3) knocked down, (4) beaten with an object, (5) beaten or beaten up, (6) beaten with a fist or kicked, (7) choked, (8) injured with an object/weapon, (9) burned/scalded, and (10) threatened with a weapon, knife, or firearm. Possible answers included five grades of frequency: never, rarely, sometimes, frequently, and very frequently. Pfeiffer et al. (1999) used Wetzels’ items 1 to 8 but a frequency scale ranging from 1 (never) to 4 (frequently). Baier et al. (2009) used Wetzels’ items 1 to 6 but a 6-point scale for frequency (1 = never, 2 = once or twice, 3 = 3 to 12 times, 4 = several times a month, 5 = once a week, and 6 = several times a week). Hellmann (2014) used Wetzels’ items 1 to 4 but added a fifth item ("bitten") and used a frequency scale from 1 (never) to 5 (very frequent). In the Lower Saxony surveys, the first 6 items of Wetzels (1997) were used; as for frequency of abuse, the possible answers ranged from 1 (never) to 5 (very often) in 2013 and 2015 (Bergmann et al., 2017) and from 1 to 6 in 2017 (1 = never, 2 = once or twice, 3 = 3 to 12 times, 4 = several times a month, 5 = once a week, and 6 = several times a week) (Bergmann et al., 2019).

2 In this context, severe corporal punishment was defined (strictly speaking) as "more frequent than sometimes," thus equating the severity of maltreatment with the frequency of its occurrence.

3 In Germany, five different types of schools follow elementary school: Hauptschule, offering a lower level of education and ending after grade 9; Realschule, a medium-level school that ends after grade 10; Gymnasium, offering the highest educational level and ending (for our cohort) after grade 13; Gesamtschule (a combination of Realschule and Gymnasium), which enables more pupils to achieve a higher educational level; and Förderschule, where pupils with learning disabilities receive special support. Two thirds of all the 7th graders in Duisburg were included in the study; the other third could not be reached because some schools refused to participate.

4 Data collection in the year 2006 proved to be the most challenging. Because respondents in the lower educational level schools were on school leave and school attendance was compulsory for all adolescents up to age 18, we attempted to retrieve these school-leavers in selected classes at vocational schools. As a consequence, the cross-sectional data include additional cases of individuals who attended these classes but who had not participated before. Because the latter could not be matched to previous cases, they have no impact on the panel dataset.

5 Smith and Thornberry (1995) also distinguished among serious, moderate, and minor forms of delinquency. CRIMOC does not include the offenses categorized in their study as “serious”; furthermore, their definitions of serious, moderate, and minor delinquency differ from those in German criminal law. Therefore, our study concentrates on the measures of “overall delinquency” and “violent delinquency.”

6 A later change in the German school system meant that the Gymnasium no longer reached the 13th grade but only grade 12.

7 These results are comparable to the trajectories reported by Reinecke (2014). The slight deviations are due to the different composition of the dataset. Moreover, the accepted models have been compared to a seven-class trajectory model derived from the CRIMOC data, developed by Seddig and Reinecke (2017) and Boers et al. (2014) and based on incidence rates. They identified the following seven classes: non-offenders (43%), early desistance offenders (9%), late-onset offenders (6%), adolescent intensive offenders (15%), persistent offenders (8%), low-level offenders (14%), and early intensive/desistance offenders (6%). Additional cross-tabulation of our five-class solution (reported here) with this seven-class solution illustrates considerable overlap, as follows: our high-rates class consisted of 61.5% persistent offenders, 1.9% early intensive/desistance offenders, and 36.5% of adolescent intensive offenders; our non-offender class consisted mainly of 85.1% non-offenders, 9.2% early desistance offenders, and 3.7% low-level offenders; our low-decreasers class consisted mainly of 43.7% low-level offenders and 33.7% early desistance offenders; our high-decreasers class included 47.8% adolescent intensive offenders, 20.8% early intensive/desistance offenders, 16.4% low-level offenders, and 12.0% persistent offenders; and finally, our increasers class consisted of 33.5% adolescent intensive offenders, 31.7% late-onset offenders, and 20.3% low-level offenders.

8 Cross-tabulation of both classifications (not displayed here) resulted in a high association between the two analyses. For instance, 72.1% of the violent non-offender class belonged to the general delinquency non-offender class. Two thirds of the violent decreasers class belonged to the high-decreasers class and 16.7% belonged to the low-decreasers class in the general delinquency LCGA. Also, 64.8% of the violent high-rate offenders belonged to the comparable class in the general delinquency model.

Additional information

Funding

This study was supported by the German Research Foundation (DFG).