Does the Neighborhood Context Shape Violent Victimization Independent of Delinquency? A Mediation Analysis of Victimization among German Urban Adolescents

ABSTRACT

Research findings suggest that there is a strong relationship between offending and being a victim of crime, whereby victimization is often a consequence of delinquent behavior. Therefore, impacts of explanatory variables on victimization might only occur as a noncausal by-product induced by their influences on delinquency. This study investigates the influences of neighborhood characteristics on violent victimization while also considering the role of delinquency as a potential mediator. Using a sample of 3,065 15-year-old students from a long-term study on crime conducted in Germany, structural equation modeling is applied to examine the influence of neighborhood characteristics on offending and victimization. Additionally, influences on victimization are decomposed into direct and indirect effects to examine mediation via delinquency. The results show that – when analyzing offending and victimization simultaneously – influences on victimization are actually mediated through delinquency. Neighborhood context (disorder, leisure opportunities, perceived safety) mainly shapes offending. This investigation cannot provide evidence for direct influences of neighborhood characteristics on victimization that are independent of a person’s delinquent behavior. Effects on violent victimization emerge from mediation via delinquency, particularly violent offending. These findings help to further disentangle impacts on victimization risk and point out implications for crime and victimization prevention within neighborhoods and cities.

KEYWORDS:

• Social disorganization
• neighborhood disorder
• victimization
• routine activity theory
• structural equation modeling

Acknowledgments

This research article uses data from the project “Crime in the Modern City” funded by the German Research Foundation. The views expressed herein are those of the author and do not necessarily reflect the policies of the funding organization. The author would like to thank the anonymous reviewers for their helpful and constructive comments on a previous version of this article.

Disclosure statement

The author declares no potential conflicts of interest with respect to the research, authorship, and/or the publication of this article.

Notes

1. Thus, a mathematically positive association is assumed.

2. Some cases from the original data set were excluded for this analysis because the respondents either reported living outside the city or provided no information about their district, yet, this information is required for the following analyses. Thus, the data set used in this paper contains 3,065 cases. The multivariate models are based on slightly fewer cases (n=3,033) due to missing values on independent variables.

3. The BBSR categorizes German cities as big cities (>100,000 inhabitants), medium cities (20,000-100,000 inh.), and small cities/group of communities (5,000-20,000 inh.) whereby these are again divided by size (smaller and larger big/medium/small cities).

4. Usually, also socioeconomic status (SES) or educational status belong to the standard array of control variables. The analysis was also controlled for SES and school type, yet, no significant associations with neither offending or victimization could be observed. Thus, for the sake of parsimony, these controls were left out since they do not contribute to the estimated models.

5. However, this association must be regarded in a nuanced way. While officially registered crime statistics consistently show elevated levels of delinquency for migrant youths, analyses of self-reported delinquency present this association in a different light. In reference to the present data about adolescents in Duisburg, Walburg (2014) argues that crime differences between native and non-native juveniles diminish when taking into account, for instance, educational participation or social marginalization. Moreover, migration status can even serve as a protective factor (due to, for example, religious norms, fewer risky activities, low alcohol consumption). The present analysis is in line with this result: The current study finds lower offending and victimization rates for Turkish youths when taking the residential area into account (see Table 2).

6. All specific indirect effects can be derived from the direct effects by multiplying the path coefficients of all connecting variables. Therefore, regarding factor 1, the indirect effects are [0.127×0.292=0.037], [–0.197×0.265×0.292=–0.015], and [–0.197×–0.018=0.004], which add up to the total indirect effect.

7. The variables are equally generated as described before: The frequencies of the according items were summed and then recoded on a scale of 0 to 12+.