Background and research questions
Situational Action Theory (SAT) is a recently developed general action theory of crime that integrates and synthesizes existing individual and ecological explanations (Wikstrom 2006, 2009, 2010a, 2014a, 2014b; Wikstrom, Oberwittler, Treiber, and Hardie 2012). In explaining individual differences in committing or not committing criminal acts, SAT postulates that high self-control and high morality are key characteristics that prevent individuals from committing such acts. Contrary to other theories of crime, SAT explicitly postulates conditional effects of self-control on offending: according to the theory, only individuals that consider crime as an alternative are guided by their level of self-control in their actions. Whether or not one sees crime as a behavioural alternative is determined by one's moral evaluation of action alternatives. Individuals that have strong moral standards do not see crime as an action alternative and do not, therefore, offend, regardless of their level of self-control.
Like some other theories of crime causation, SAT includes propositions on the role of peers. According to SAT, peers affect individual behaviour when they are part of the micro-situational behavioural settings that individuals are exposed to. SAT postulates that adolescents are more likely to offend when (1) they spend more unsupervised time with peers, and (2) they have more rule-breaking peers.
In the present article, we aim to empirically test these core assumptions of SAT by estimating cross-sectional models with cross-lagged effects on data from two-wave panel data collected with specifically designed measurement instruments. We first test the association between offending, on the one hand, and morality, self-control, and their statistical interaction, on the other. This is the first key statement of SAT, referred to as the principle of the conditional effects of self-control. Second, we investigate the role of the two peer influences mentioned above, testing simultaneously their impact on the variation in frequency of offending. Finally, we explore the interaction of peer influences with both elements of propensity to offend (morality and self-control). This interaction is in line with the assumption of SAT that there is an interaction between exposure to criminogenic settings and individual propensity to crime, a second key statement of SAT.
Theory and hypotheses
The fundamental arguments of SAT are (1) that acts of crime are a special case of moral actions (acts of moral rule breaking stated in law) and therefore need to be explained as such; and (2) that people engage in acts of crime because they (a) come to see such acts as viable action alternatives, and (b) choose (habitually or deliberately) to carry them out (Wikstrom 2006, 2009, 2010b, 2014a, 2014b; Wikstrom and Butterworth 2006; Wikstrom et al. 2012). SAT proposes that the convergence (in time and space) between a person's propensity (individual tendency to commit offences) and exposure to particular settings initiates a perception choice process, the outcome of which is an action (or inaction) that can be an offence (or not). The theory explicitly states that the individual's propensity and exposure to criminogenic settings interact to determine whether an act of crime is committed. The two causally relevant elements of propensity are morality (individual moral rules) and self-control.
According to SAT, two types of characteristics define the criminogenic potential of a setting: characteristics that define the moral context of the environment and characteristics that indicate exposure to a seducing physical environment (settings characterized by temptation and provocation). Causally relevant characteristics of a criminogenic setting are the amount of time spent with friends unsupervised by adults in settings with poor collective efficacy, and exposure to peer delinquency (Wikstrom 2009; Wikstrom, Ceccato, Hardie, and Treiber 2010; Wikstrom et al. 2012). A core statement of SAT is that, when individuals have a low level of morality and self-control, and when they spend more time in criminogenic settings, the likelihood that they will offend increases.
Again, Wikstrom has emphasized in many publications (Wikstrom 2009, 2010a, 2010b, 2014a, 2014b; Wikstrom and Butterworth 2006; Wikstrom and Svensson 2010; Wikstrom and Treiber 2007) that the key elements of a propensity to crime are an individual's level of morality and self-control and not her or his motivation to offend.
Morality can be described as the moral rules of an individual that guide the individual's action alternatives. Moral rules guide the process of seeing crime as an action alternative in the perception-choice process. If moral standards are poor, a person may see crime as a viable alternative.
Self-control is defined as "the successful inhibition of a perceived action alternative or the interruption of a course of action that conflicts with an individual's morality" (Wikstrom and Svensson 2010: 397). A person whose self-control is low will offend more quickly in the face of temptations and provocations. In line with these SAT statements, we formulate the following hypothesis 1: Adolescents with low morality and low self-control will be more involved in offending than adolescents with high self-control and high morality.
The net effects of morality and self-control on offending have been tested extensively in the literature predominantly from the English-speaking world (Goode 2008; Pratt and Cullen 2000). However, the role of self-control in SAT is other than the role in the general theory of crime (Gottfredson and Hirschi 1990), in which low self-control is the primary direct cause of all offending. In SAT, an individual's capacity to exert self-control is causally relevant only if his or her moral rules are in conflict with the temptations and opportunities to commit offences in a certain setting (Wikstrom and Svensson 2010; Wikstrom and Treiber 2007). The existence of this interaction has recently been supported by empirical studies (Svensson, Pauwels, and Weerman 2010; Wikstrom et al. 2012; Wikstrom and Svensson 2010). Thus, hypothesis 2 is formulated as follows: For adolescents with low morality, the effect of low self-control on offending will be stronger than for adolescents with high morality (in other words, as morality gets weaker, the effect of low self-control on offending increases). This proposition is vital for and exclusively an element of SAT. For that reason, we will pay special attention to it by examining the existence of this interaction effect in more detail.
In SAT, peers are seen, on the one hand, as part of the (moral) setting that may affect a person's propensity to offend and, on the other hand, as part of the moral context in the interactive model that explains the emergence of acts of crime. Wikstrom (2014a, 2014b) assumes that spending more time with peers in public spaces without the presence of adults increases the probability of committing acts of crime. This notion is in line with previous work by other scholars who stressed the importance of time spent with peers in shaping the behavioural setting of adolescents (Agnew and Peterson 1989; Bemasco, Bruinsma, Pauwels, and Weerman 2013; Bemasco, Ruiter, Bruinsma, Pauwels, and Weerman 2013; Osgood and Anderson 2004; Osgood, Wilson, O'Malley, Bachman, and Johnston 1996; Svensson and Oberwittler 2010; Weerman, Bernasco, Bruinsma, and Pauwels 2013). From this perspective, it is the exposure to settings that are low in social control and/or high in offering opportunities to offend that increases someone's level of offending. Spending time with peers in an unstructured environment and without supervision of adults may seduce adolescents to take opportunities to offend for symbolic rewards like status and reputation. This is supported by previous research (Haynie and Osgood 2005; Osgood et al. 1996) that revealed a strong correlation between various unstructured activities (e.g., driving around with friends, spending time in the city centre) and offending.
Further, Wikstrom argues that the rule breaking behaviour of peers is one element of the criminogenic moral setting. The positive relationship between someone's own delinquency and the delinquency of peers is a classic finding in criminology (Haynie and Osgood 2005; Matsueda and Anderson 1998; Warr and Stafford 1991; Weerman 2004, 2011; Weerman and Smeenk 2005): high levels of one's own offending are correlated with high levels of peer offending. In SAT, this relationship is important because it concerns the interaction between an adolescent's low morality and a criminogenic moral setting. More generally, when adolescents interact with peers who have law-violating moral rules, the chance that they will commit crimes increases. Thus, hypothesis 3 is as follows: Adolescents who have more rule-breaking peers and spend more unsupervised time with peers will be more involved in offending, independent of their propensity to crime.
Finally, Wikstrom assumes, in SAT, that individuals with a certain propensity (low morality and low self-control) are more vulnerable to the influence of a criminogenic setting. His arguments imply an interaction between the individual's propensity and his social environment. There exists an extensive literature on the interaction between self-control and peer influences, but not all peer influences have been studied simultaneously (Chappie 2005; Gardner, Dishion, and Connell 2008; Hay and Forrest 2008; LaGrange and Silverman 1999; Longshore 1998; McGloin and Shermer 2009; Meldrum and Hay 2012; Ousey and Wilcox 2007; Thomas and McGloin 2013). Few studies have investigated the interaction between morality of the adolescent and peer influences (Hannon, DeFronzo, and Prochnow 2001; Mears, Ploeger, and Warr 1998). In line with SAT and based on the available literature, we argue that adolescents...