Adolescent Crime

Per-Olof H. Wikstrom, David Butterworth, David A. Butterworth

• Publisher: Willan
• Year: 2013
• eBook ISBN: 9781134008261
• Edition: 1
• Topic: Social Sciences
• Subtopic: Criminology
• Index: Social Sciences

This book examines young people's involvement in crime (including crimes of violence, vandalism, shoplifting, burglary and car crime) as both victims and offenders.

Although adolescence is the time when involvement in crime peaks, few previous UK-based studies have attempted to provide a methodical and comprehensive understanding of adolescent offending on a city-wide basis. This book seeks a better understanding of adolescent crime by studying the relationship between individual characteristics (social bonds and morality and self-control) and lifestyles (as defined by delinquent peers, substance use and exposure to risky behaviour settings) and their joint influence on adolescent involvement in crime, against the backdrop of the juveniles' social context - taking into account family, school and neighbourhood influences.

The findings of this study suggest the existence of three main groups of adolescent offenders; propensity induced offenders, life-style dependent offenders and situationally limited offenders, groups of offenders having different causal backgrounds to their crime involvement, and who therefore may warrant different strategies for effective prevention.

Chapter 1

Introduction

There is a lack of current research on young people's offending and its causes in the UK. The main source of information on current juvenile offending comes from two recent national self-report studies (Graham and Bowling 1995; Flood-Page et al. 2000; and see Chapter 4).¹ However, very few academic studies have explored the prevalence and patterns of juvenile offending in UK cities. The only more recent exceptions known to us, of which none have been carried out in England or Wales, are a study of about 1,200 11–15-year-olds from five selected schools in Edinburgh by Anderson et al. (1994), an ongoing (longitudinal) study of 4,300 juveniles, covering most schools in Edinburgh, from which some initial findings are available (e.g. Smith et al. 2001) and some research carried out in the city of Belfast based on a random sample of about 900 14–21-year-olds (McQuoid and Lockhart 1994). These city-based studies, with some exceptions in the Smith et al. study, do not consider explanatory factors in any depth.

In addition to this, there are a few older longitudinal studies (for an overview, see Loeber and Farrington 2001: app. C), of which the most prominent is the so-called Cambridge Study in Delinquent Development, a study of a 1953 male cohort from a London working-class area (see e.g. Farrington 1989, 1992). In this context, also McDonald's (1969) 1964 cross-sectional study of male juvenile delinquency in four different areas of England, and Belson's (1975) 1967–8 cross-sectional study of juvenile theft in London should be mentioned.

Although the British Crime Survey (BCS) has supplied a wealth of information on victimisation, its regular sweeps do not cover young victims (under the age of 16). However, one special study has been made of 12–15 year-olds' victimisation using a special sample from the sample of one of the BCS waves (Maung 1995). Studies of young people's victimisation in particular cities are rare. The already mentioned studies in Edinburgh (Anderson et al. 1994; Smith et al. 2001) and Belfast (McQuoid and Lockhart 1994), in addition to a study of six schools in Middlesborough (Brown 1995), are the prime examples of studies of young people's victimisation conducted in particular cities.

In search of patterns and explanations

The starting point for this research is the sometimes controversial idea that a key objective of social science is to study patterns in social life and to try to offer explanations of these patterns. To do so we need to map out the correlates of social action and to try to understand the causal mechanisms at work. Some of the correlates may be just correlates; others may help to identify what mechanisms cause a particular social action. An important task is therefore to evaluate the correlates (their potential as representing causal mechanisms at work) in relation to theories of what constitutes social action.

The basic position taken here is that social actions, like offending, 1) ultimately are a result of an individual's perception of action alternatives and process of choice, and that 2) a key challenge for social science research is to understand how an individual's wider social situation influences his or her individual characteristics and experiences and exposure to environmental features that, in turn, independently or in interaction, cause action (e.g. acts of crime) through their influence on how an individual perceives ac-tion alternatives and makes choices (Wikström and Sampson 2003; Wikström 2004, 2005, 2006). This is, of course, a monumental task, and no single research project can hope to be able to provide more than a small contribution towards this goal.

In the Peterborough Youth Study we attempt to move knowledge a little further forward by studying the relationship between individual characteristics (social bonds and morality and self-control) and lifestyles (as defined by delinquent peers, substance use and exposure to risky behaviour settings) and their joint influence on juvenile involvement in crime, against the backdrop of the juveniles' social context (their family's social position and, to some extent, their wider social context as represented by their neighbourhood and school contexts). We assume that individuals' social lives will differ and interact with their individual characteristics and that this will have some bearing on their involvement in crime as offenders or victims. However, it is out of the scope of this research empirically to address the role of individual perception of action alternatives and the processes of choice (although we will theoretically discuss this as the main mechanism linking individuals and behavioural contexts to their actions).

The problem of correlation and causation

Criminological research has demonstrated hundreds, if not thousands, of stable correlates to adolescent crime involvement. These correlates are commonly referred to as risk (or protective) factors, and sometimes treated as established causes rather than mere correlates for which causation has to be established. This problem has increasingly been recognised as one that has to be dealt with in order to advance our knowledge about the causes and explanation of crime involvement.

Farrington (2002a: 664) has defined ‘risk factors’ as ‘prior factors that increase the risk of occurrence of the onset, frequency, persistence, or duration of offending’. There appears to be an increasing consensus in discussions of risk (and protective) factors, with regard to their associations with offending, that there is a need for a framework by which one can distinguish risk factors that are 1) causally related to offending from those that are 2) symptomatic of offending and, finally, those that may be 3) both symptoms as well as causes, such as excessive alcohol use (Farrington 2002a). Farrington (2002a) reviewed the criminological literature on temperament and personality factors (e.g. impulsivity; see also Tremblay and LeMarquand 2001), family factors (e.g. parental supervision; see also Wasserman and Seracini 2001) and school factors (see also Herrenkohl et al. 2001), which have been found to be risk factors in the development of offending, and noted:

In explaining the development of offending, a major problem is that most risk factors tend to coincide and tend to be interrelated. For example, adolescents living in physically deteriorated and socially disorganized neighbourhoods disproportionately tend also to come from families with poor parental supervision and erratic parental discipline, and tend also to have high impulsivity and low intelligence. The concentration and co-occurrence of these kind of adversities make it difficult to establish their independent, interactive, and sequential influences on offending and antisocial behaviour (2002a: 680).

Therefore, one point which seems to be widely accepted by those working within this criminological tradition is that ‘a major problem of the risk factor prevention paradigm is to determine those risk factors that are causes from those that are merely markers or correlated with causes’ (Farrington 2000: 7, emphasis added).

Conversely to risk factors, ‘protective factors’ are characteristics that are seen in some way as operating to reduce the likelihood of offending. Farrington (1998) has suggested that there are three distinct definitions of what constitutes a ‘protective’ factor. The first of these is that a protective factor is just the opposite of a risk factor. Hence if low self-control were a risk factor for offending, then high self-control could be a protective factor. Secondly, protective factors may simply ‘stand-alone’ — that is, they do not have a linear relationship with offending. Thirdly, protective factors may operate in interaction to mitigate the effects of other risk factors (see also Lösel and Bender 2003).

It seems entirely plausible to suggest that a fruitful way to sort true risk and protective factors (i.e. causes) from correlates is through greater theoretical attention being paid to what actually puts the ‘risk’ into risk factor, or what it is about protective factors that actually provides the ‘protection’. For example, Sampson and Laub (1993) suggest that marriage, per se, does not act as a ‘protective’ factor but, rather, it is the strength of the attachment to one's partner that accompanies marriage which acts to lessen the likelihood of offending. As Rutter (1987: 329, emphasis in original) concludes in his discussion of protective factors, ‘The focus of attention should be on the protective processes or mechanisms, rather than on variables’ (see also Hedström and Swedberg 1998; Wikström 2004).

Perhaps therefore the fundamental issue that arises as a result of the rather atheoretical nature of the risk and protective factor paradigm is the question of their underlying cause or causes. As Farrington (1998: 262) notes: ‘What is needed is a coordinated program of research to determine how many key theoretical constructs underlie all these variables, what they are, how they are causally related, and how they should best be operationally defined and measured.’ Furthermore, while person-centred risk factors have been the focus of extensive investigation, Farrington (2003: 227) highlighted that much less is known about peer, school and neighbourhood risk factors.

Although the consistency of between-individual differences is frequently accepted (for example, the chronic/non-chronic offender dichotomy), the possible reasons for such continue to appear to be hypothesised as being the result of stable individual traits. This individual-oriented view has diverted attention from the social contexts within which individuals actually live their lives. As Magnusson (1988: 23) has commented, there is a ‘dynamic, continuous, and reciprocal process of interaction between the individual and the environment’. While self-selection is no doubt likely to play a role in determining the social contexts in which individuals choose to engage as they age, there can be little doubt that such ability to choose those contexts is severely limited during the early years of individual development.

Hence the risk and protective factor approach has not only tended to downplay the consideration of the environment in producing contexts of action which lead to within-individual differences in the likelihood of criminal propensity becoming substantiated as (criminal) acts at different times (Farrington 2000; Wikström 2004), it has also largely ignored the influence of environmental factors as a context of development in the acquisition of characteristics (e.g. see Bronfenbrenner 1979; Martens 1993; Wikström 2005) that give rise to between-individual differences in offending.

In this study we will demonstrate associations (correlations) between variables and show how they independently, jointly and through interaction can predict outcomes (e.g. crime involvement). It is important to stress once again, as discussed above, that correlation (and prediction) do not necessarily imply causation and that findings of statistical studies, like the present one, cannot by themselves establish causation and offer explanation. One of the authors has elsewhere argued that causation can only be established through experimentation, and that explanation can only be offered by analytic work aimed at specifying plausible mechanisms (processes) that link the putative cause to the effect (Wikström 2006). In the concluding chapter we will discuss the empirical findings of this study in relation to a theoretical framework (the Situational Action Theory of Crime Causation) that will allow us to interpret the findings in terms of how they may contribute to the explanation of adolescent offending.

Aims of the research: key questions and constructs

The overall aim of this research is to contribute to a better understanding of adolescent involvement in crime (primarily as offenders but, to some degree, also as victims). We aim to achieve this by studying the relationships between family social position (parents' occupational social class, family structure and family ethnicity), the adolescents' social situation (family and school bonds), their individual dispositions (morality and ability to exercise self-control) and lifestyles (as implicated by their peers' delinquency, their own activities and alcohol and drug use) and how these factors relate to their involvement in crime as offenders (and partly as victims). We are particularly interested in testing the idea that the adolescents' individual characteristics (their social situation and dispositions) interact with their lifestyle in producing crime involvement. Specifcally, we hypothesise that the impact of adolescent lifestyles on their crime involvement is dependent on their individual characteristics. We are also interested in exploring the influence by the adolescents' family social position on their individual characteristics (morality and self-control) and lifestyle. We are postulating, and aim to test whether the data are consistent with the assumption, that any impact of the adolescents' family social position on their crime involvement is mediated through an impact on key individual characteristics and lifestyles. Finally, we are also interested in exploring any influences of the wider social context (as defined by neigh...

Table of contents

• Cover
• Half Title
• Full Title
• Copyright
• Contents
• List of figures and tables
• Acknowledgements
• 1 Introduction
• 2 The Peterborough Youth Study
• 3 The City of Peterborough and its neighbourhoods
• 4 Involvement in crime and substance use
• 5 Family social position
• 6 Individual characteristics: social situation and dispositions
• 7 The community context: neighbourhoods and schools
• 8 Lifestyles
• 9 Youth routines and involvement in crime: some preliminary findings from the space-time budget study
• 10 Key findings and their explanations
• Appendices
• References
• Index