Research shows that individuals who begin to be antisocial in childhood and persist into adolescence are less likely than individuals on an adolescent-onset trajectory to desist from such behaviour during emerging adulthood. However, there is also
significant heterogeneity in adult behaviour, even within the high-risk persistent trajectory group. This chapter has discussed theoretical and empirical evidence for the impact of experiences related to drinking and institutional role status during emerging adulthood on such heterogeneity in pathways of aggression from childhood to the early 20s. The opening section described the unique features of the emerging adult life phase and outlined the challenges and opportunities that these features present to youth as they negotiate the post-high school years. Drawing on systems theories, I argued that, like other major normative transitions, the pervasive social and contextual role changes during the transition to emerging adulthood alter the way people fit with and relate to their environments.
Longitudinal studies show that these shifts in person-environment fit can encourage both continuity and discontinuity in trajectories of antisocial behaviour, depending on developmental strengths and weaknesses arising from past levels of aggression, individual characteristics, social functioning and demographic factors. In other words, although different subgroups of emerging adults arrive at the transition differently equipped to negotiate new social and institutional settings in a non- aggressive manner, the evidence suggests that experiences embedded within the new settings, and the ways in which people respond to them, can themselves account for some variation in pathways of aggression from childhood. In general, evidence suggests
that positive life events involving the formation of conventional social bonds, such as marriage and employment, promote desistance, while snares such as substance use inhibit desistance. Theorists argue, however, that the effects of such factors for
disrupting or exacerbating pathways of antisocial behaviour during emerging adulthood differ markedly between members of different antisocial trajectory groups. Moffitt (2002), for instance, speculates that any positive ‘turning point’ effects associated with early adult life events will be most applicable to members of adolescent-onset trajectory groups. This is because the pervasive social and behavioural dysfunction that
characterises the life-course persistent trajectory group will decrease the likelihood that these individuals will encounter such salutary events. The relatively less troubled adolescent-onset young people, on the other hand, will have the capacity to take advantage of turning point opportunities in the early adult years.
In this study I ask (1) whether levels of aggression in early adulthood are associated with different drinking patterns or institutional roles, and (2) whether any such effects vary across bully trajectory groups. For example, is early adult drinking more strongly associated with violence for persistent bullies? Might employment or university study during emerging adulthood disrupt pathways of aggression for some former bullies but not others? To assess this, I test whether institutional role and/or drinking moderate the association between the risk factor (bullying trajectory during school) and the outcome (physical aggression in early adulthood). The following
discussion briefly considers the implications of the evidence reviewed in this chapter for these questions.
The studies reviewed provided ample evidence that drinking and work/study transitions can account for variation in adult aggression over and above the effects of past aggression and other risk factors, such as x and y. As noted in the previous section, positive employment and study experiences were generally associated with less
aggression amongst emerging adults, over and above effects of past antisocial
behaviour. With regard to drinking, Felson and colleagues’ (2008; Felson, Teasdale, & Burchfield, 2008) analyses of North American and Finnish adolescent samples showed that heavy drinking accounted for additional variance in self-reported violence over and above effects of both prior drinking patterns and prior involvement in violent situations. Consistent with these findings, longitudinal studies suggest that drinking can exert short-term contemporaneous effects on aggression during emerging adulthood beyond what would be predicted given child and adolescent histories of aggression (Hussong, et al., 2004; Morizot & Le Blanc, 2007).
Some of the studies showed that drinking and work/study factors functioned to mediate the association between earlier aggression and adult aggression. A consistent finding was that antisocial young people were at greatest risk of making a poor
transition into emerging adulthood. Delinquent adolescents were at greater risk of poor work and study outcomes (for example, fail to complete high school, spend more time unemployed, fail to obtain an advanced educational qualification), and were more likely than non-delinquent adolescents to drink heavily. To the extent that these experiences were associated with concurrent functioning, they served to increase the ‘spread’ between those who were doing poorly prior to the transition and those who were doing well. For example, Roche and colleagues showed that, for males, the relationship between seventh-grade aggression and early adult violence was mediated by advanced education: those who were least aggressive in the seventh grade were most likely to undertake advanced education, and this experience was associated with less emerging adult violence. The implication of these patterns for the present study is that early adult transition experiences might fail to account for ‘turnarounds’ in behaviour for former bullies, instead exacerbating subgroup differences in ongoing trajectories of behaviour.
This is the process that Schulenberg, Maggs and O’Malley refer to as increased heterogeneity (2003).
On the other hand, some of the studies suggested that consequences of
experiences during the transition to adulthood—be they positive or negative—did not affect the antisocial behaviour of the entire sample of emerging adults in the same way. The clearest evidence for subgroup differences was found in studies relating to work and study transitions. Although not all the studies reviewed specifically examined moderating effects, or within-subgroup variation in offending as it related to work/study variables, two general patterns were apparent. First, there was no indication that that less optimal work/study outcomes (for example, periods of unemployment) were associated with sudden late onset of difficulties in youth with no history of antisocial behaviour. Rather, such ‘off-diagonal’ beneficial and detrimental effects were most salient for young people with histories of delinquency. For some former delinquents, work and study experiences contributed to ‘turnarounds’ in behaviour. Some of the persistent serious adolescents in the Stouthamer-Loeber et al. (2004) study, for example, were employed and/or studying at ages 20–25, and these young people were more likely to have desisted than those delinquents who were not employed or studying.
Second, Roisman and colleagues’ analyses showed that even these off-diagonal effects did not apply to all former delinquents. In contrast to Moffitt’s proposition that life-course persistent delinquents will not benefit from early adult turning points, the authors found that positive work and study experiences during emerging adulthood were associated with desistance only for young people who had been persistently aggressive since childhood. The authors argued that, while the most aggressive
individuals were least likely to experience positive occupational outcomes, variation in antisocial behaviour within this group was largely attributable to those who for some reason did manage to successfully engage in post-high school realms of work and study.
This finding is more broadly consistent with some theorists’ arguments that promoting factors show a stronger effect under high-risk conditions than low-risk conditions. That is, while the most aggressive young people may be less likely to encounter turning points, those who are lucky enough to do so have the most to gain from the experience (Rutter, 1987, 1994). These findings imply that engagement in different post-high school institutional roles might provide persistent bullies with opportunities to break away from prior difficulties and establish more positive behavioural patterns.
There was also some evidence that effects of snares like drinking can vary depending on developmental history of aggression. Felson and colleagues’ analyses (2008) showed that the exacerbating effect of drinking for late adolescent violence was
greatest for those individuals who had been most violent in the past. The authors argue that these findings are consistent with theories suggestingthat alcohol increases the risk of aggression primarily for the violence-prone. In this case, drinking might be more strongly associated with aggression for persistent rather than adolescent-onset bullies. However, as discussed earlier, the proximal effects of features of the social
environments in which drinking takes place during emerging adulthood might precipitate violence even for young people with fewer aggressive tendencies. Moffitt (1993) argues that substance use will act as a snare to inhibit desistance even for the less-aggressive adolescent-onset delinquents. Although differential relationships between substance use and adult desistance within the Dunedin study trajectory groups has not been examined, Hussong and colleagues’ analyses demonstrated a time-varying increase in antisocial behaviour associated with drinking that did not appear to vary based on earlier antisocial behaviour.
To summarise, the evidence reviewed in this chapter confirms that emerging adult transition experiences can affect aggressive pathways, and suggests a number of different possibilities about the ways in which drinking and institutional role might
affect continuity and discontinuity from school bullying to adult aggression in the present study. However, as outlined in the research model at the end of Chapter 1, proximal effects of early adult transition experiences are only one link in the chain of events from childhood bullying to adult aggression. Regardless of whether effects of emerging adult drinking and/or work and study experiences are additive, accounting for variance in adult aggression for the whole sample, or interactive, accounting for
variance in adult aggression to different degrees across subgroups of emerging adults, the developmental theory and evidence strongly suggests that the amount young people drink and their different work and study situations after high school will be outcomes of cumulated earlier strengths and weaknesses.To what extent might positive or negative effects of the transition experiences reflect earlier differences in pathways of
functioning during school? More broadly, what other school-years experiences might contribute to adult aggression?
These questions pertain to longitudinal mediated pathways from school bullying to adult aggression. Consistent with the overall developmental perspective, evidence suggests that the effects of distal factors for adult aggression may unfold over time in complex ways (Dodge & Pettit, 2003). Some school-years factors may exert direct effects upon future aggression, while others may operate indirectly, relating to events occurring later in the chains of events that eventuate in violent behaviours. In the next chapter I provide a theoretical background to such possible pathways in this study. This involves taking a step back to the school years to flesh out the earlier parts of Figure 1.1. I discuss measures of socio-emotional functioning at Times 1 and 2 (see Figure 1.1), and review findings from earlier waves of the Life at School study showing how these factors contribute to continuity in bullying over the primary school to high school transition. Might these processes also carry through to affect the likelihood of
present study are theoretically related to increased risk or drinking and different post- secondary work and study roles? Chapter 3 addresses these issues.
CHAPTER 3:
CHILDHOOD AND ADOLESCENT MEDIATORS IN THE PATHWAY FROM SCHOOL BULLYING TO ADULT AGGRESSION: SOCIO-EMOTIONAL FUNCTIONING DURING
PRIMARY SCHOOL AND HIGH SCHOOL
So far in this thesis I have mostly focused on bullying as a distal predictor of adult aggression, and considered ways in which proximal influences associated with institutional role and drinking in the transition to emerging adulthood might disrupt this relationship. Assuming that bullying is likely to be related to aggression, a further issue that arises is that of documenting the intervening processes involved.
Important as early aggression is for predicting later aggression, it is unlikely that bullying will be the only distal variable in the pathway leading to adult aggression. Developmental studies instead implicate multiple distal risk factors and varying paths to early adult aggressive behaviour. These risk factors reside both in the individual and within his or her interpersonal, social, and institutional contexts. Certain distal risk factors (e.g. impulsivity) are sometimes shown to lead directly to aggressive behaviour (Nagin & Tremblay, 1999). However, these same factors are also linked to other diverse adult outcomes, such as delinquency, substance use, educational failure, poor mental health, and unemployment, which may themselves be the proximal conduits for violence. Moreover, risk factors within the individual and features of his or her social contexts may be correlated with each other and mediate each other to lead to aggressive outcomes (Broidy, et al., 2003; Tremblay, 2000). Investigating the ways in which early bullying and other risk factors mediate each other to adolescent experience, reverberate in early adulthood, and subsequently lead to aggression is an important task. This is because understanding when and why and for whom discontinuities in aggression occur at various points along the pathway from childhood to emerging adulthood, is critical to informing intervention efforts to effectively re-route these pathways.
Mediating factors of interest in this study relate to aspects of socio-emotional functioning within the school context during primary school and high school.
Specifically, I focus on school adjustment, (including academic functioning and school liking/connectedness), impulsivity and shame management. These constructs reflect healthy socio-emotional functioning in childhood and adolescence, and also predict a range of adult outcomes, including levels of antisocial behaviour, social functioning, mental health, and occupational attainment. The first construct relates to children’s self- regulatory abilities, particularly with regard to regulation of behaviour and emotion, and is represented by shame management (regulation of emotion) and impulsivity
(regulation of behaviour). The second construct relates to connectedness and bonding within the institutional context of the school, and is represented by measures of school adjustment.
The aim of the current chapter is to review evidence suggesting how interrelationships among these socio-emotional factors influence the unfolding of bullying and aggression across childhood, adolescence and adulthood. The bulk of the chapter focuses on describing how shame management, school adjustment and
impulsivity are related to each other and to bullying over the transition from primary school to high school. This focus is warranted because although theoretically all of the aforementioned socio-emotional variables could be equally important in the prediction of adult aggression, previous research with these data carried out by Life at School researchers shows that shame management is the main predictor of bullying during primary school and high school. Moreover, school adjustment and impulsivity have been shown to relate to both bullying and shame management during the school years (Ahmed, 2001).
In the first section of the chapter I provide a brief overview of empirical
more specific discussion of shame. The second section reviews the emotion of shame and its role in interpersonal functioning and aggressive behaviour. Much of the research in this area is found in the clinical literature, linking shame to a range of internalising and externalising disorders. Research on emotion regulation, on the other hand, views both the experience and expression of emotion as sets of ongoing, mutually influential person-environment transactions. From this perspective, the consequences of any emotion for individual wellbeing are dependent upon the way that emotion is regulated.
Ahmed’s (2001) shame management theory draws on these concepts to
explicitly describe socially adaptive and maladaptive ways of regulating shame. Ahmed developed and tested the theory in the first wave of the Life at School study, when the young people in the current sample were in primary school (Ahmed, 1999, 2001; Ahmed & Braithwaite, 2004a, 2004b). I introduce this theory in the third section and review empirical findings from previous Life at School research in the fourth section. Shame management theory purports to explain how the way in which children manage feelings of shame relates to their bullying behaviour. Moreover, it is designed to draw together many of the disparate social and individual correlates of bullying (e.g. parenting styles, self-esteem, empathy, family conflict, school experiences) into a cohesive framework that explains how these factors lead to bullying by the way in which they help or hinder socially adaptive shame management.
Ahmed showed that children’s shame management skills mediated the
relationships between impulsivity, school adjustment, and bullying. That is, impulsivity and school adjustment influenced the degree to which children were able to regulate feelings of shame in a socially functional manner, with shame management
subsequently predicting bullying. Longitudinal analyses using data from Times 1 and 2 (Ahmed, 2006; Braithwaite, 2006; Braithwaite, et al., 2003) showed that continuity in bullying between primary school and high school was mediated by continuity in shame
management. Moreover, findings support the possibility that continuity in adaptive shame management is itself mediated by continuity in positive school adjustment.
In the fourth section I consider the significance of these interrelationships among socio-emotional variables and bullying during the school years for continuity of aggression into adulthood. Longitudinal studies show that, although many of the factors that predict aggression at earlier ages are also associated with violence at later ages, the effects of these risk factors are unlikely to operate independently of either earlier aggression or later proximal experiences. On the basis of this literature, I argue that the socio-emotional factors that maintain bullying during school will have consequences for adult aggression, but pathways are likely to be complex. I outline evidence supporting several possible mechanisms. For instance, it is possible that socio-emotional variables could make direct contributions to the prediction of adult aggression, beyond the risk conveyed by bullying. Alternatively, it is also possible that their effects could be indirect, mediated by (a) bullying, and/or (b) emerging adult drinking and institutional role.
In the final section of the chapter I return to the research model presented at the end of Chapter 1 (Figure 1.1), drawing together the argument presented across these opening chapters to outline the specific research questions that will be addressed by the data analyses that follow.
Impulsivity, school adjustment and bullying
Impulsivity
Developmental researchers have repeatedly found that aspects of behavioural undercontrol, including hyperactivity, impulsivity, risk-taking and inattention, predict aggressive and violent behaviour in both childhood and adolescence (Caspi, Henry, McGee, Moffitt, & Silva, 1995; Farrington, 2005a). More specifically, numerous
studies in several disciplines document both the concurrent and longitudinal predictive relationship between impulsivity and a range of problem behaviours in children and adolescents (for example, Luengo, Carrillo-de-la-Peña, Otero, & Romero, 1994; Olson, Schilling, & Bates, 1999; White, et al., 1994). Lösel and colleagues (2007), for instance,