Introduction to the Special Issue with some reflections on the role of self-control in Situational Action Theory

In recent years, Per-Olof Wikström’s Situational Action Theory (SAT) (for example, Wikström, 2010; Wikström et al., 2012) has become one of the most-examined crime theories among European criminologists interested in crime causation. Although devised as a tool to explain rule-breaking behaviour in general, most contemporary applications deal with its potential to account for the formation of criminal or delinquent conduct. The theory’s key argument is that the interaction between an individual with a certain crime propensity and a setting with a certain criminogeneity triggers a perception–choice process that immediately governs action. Therewith SAT represents both a broad and a deep theory. ¹ It is broad in the sense that it successfully integrates person-oriented and environment-oriented explanations of criminal conduct. It is deep in the sense that it illuminates the mechanism that brings about criminal behaviour. A distinct feature is the theory’s focus on the interactive interplay of the various factors and processes involved in crime causation. ‘It’s all about interactions’ (Wikström et al., 2012) excellently describes the theory’s central credo. This implies a shift from the study of the unconditional effects of various influencing entities to their conditional effects, which runs through all articles included in this Special Issue of the European Journal of Criminology.

In the first contribution to this issue, Wikström and colleagues provide a thorough introduction to SAT and preferred ways of testing it. Then they employ neural network analyses to examine the interplay of individual propensity and setting exposure in crime causation. Their results corroborate the theory.

The next article gives a comprehensive overview of the cumulative state of research on SAT. In their stocktaking of the first decade of enquiry into SAT, Pauwels and colleagues concentrate on three key propositions: the propensity–exposure interaction, the interplay of crime propensity and deterrence, and the interaction between a person’s morality and his or her ability to exercise self-control. The authors demonstrate that the available findings are more supportive than mixed and sketch out promising avenues for future research.

Several contributions to this Special Issue deal with the role of self-control in SAT. In this theory, self-control is subdivided into a trait (an individual characteristic) and a process (a situational activity) (Wikström and Treiber, 2007). Self-control is both something we have (or have not) and something we do (or do not). The ability to exercise self-control describes an individual’s ability to align one’s behaviour to one’s moral values when faced with external inducements or pressures to choose another course of action. The situational execution of self-control refers to the actual inhibition of a perceived action alternative that collides with one’s personal morals. Consequently, SAT argues that low self-control is more than a vulnerability to the temptations and provocations of the moment, more than the tendency to resist current temptations (situational incentives) and provocations (external pressure) to break social rules, and certainly more than the capacity to anticipate and calculate the long-term consequences of behaviour (Gottfredson and Hirschi, 1990): it is a lack of success at acting in accordance with one’s personal morality when confronted with outer inducements or pressures to breach one’s own moral rules (Wikström and Treiber, 2007). Self-control describes ‘the successful inhibition of perceived action alternatives … that conflict with an individual’s morality’ (Wikström and Treiber, 2007: 244). In crime causation, self-control comes into play only when individuals deliberate, when crime has overcome the moral filter and is perceived as a viable action alternative, and when there is a moral conflict regarding the criminal action alternative. The latter points to incongruent rule-guidance in the sense that personal morality discourages criminal behaviour but the moral context of the setting encourages it.

This conceptualization of self-control has implications that so far have rarely been addressed in the empirical literature.

According to these considerations, self-control becomes relevant when a perceived criminal action alternative contrasts with an individual’s moral beliefs. From this it follows that self-control can be a crime-reducing factor solely when the agent regards crime as morally wrong (that is, has a strong law-relevant morality). However, most empirical studies on the interaction between morality and self-control have found that the crime-dampening effects of the ability to exercise self-control are greater among individuals of lower morality. At first sight, this observation – although interpreted as supportive of the theory by all scholars – seems to be at odds with the dynamics specified in SAT. The researchers conducting these studies usually refer to the workings of the moral filter to bring these results in line with the theory (for example, Hirtenlehner and Kunz, 2016; Svensson et al., 2010; Wikström and Svensson, 2010). They argue that self-control is more influential among individuals characterized by lower morality because these people are more likely to see crime as a real action alternative. This interpretation is grounded in the writings of the founder of the theory: ‘If an individual’s morality does not “permit” him/her to see crime as an alternative, self-control will not enter into the process of choice (and an individual’s ability to exercise self-control will be irrelevant to his/her choice of action)’ (Wikström and Treiber, 2007: 250). Nevertheless, a whiff of ambiguity remains. Resolving this dissonance in future works is highly advisable.

In their contribution to this special issue, Kroneberg and Schulz take up this subject. They submit that SAT’s conceptualization of self-control implies that high self-control actually increases the likelihood of offending among individuals of a very low, not to say antisocial, morality. For individuals holding strong deviant moral beliefs (for example, adolescents who view violence as the only right way to respond to insults), a well-developed ability to adhere to one’s personal morals may result in higher levels of criminal behaviour. In Kroneberg and Schulz’s words, ‘self-control is a double-edged sword with respect to crime and delinquency’. It is irrelevant for individuals with high morality, reduces crime for individuals with medium morality, and fosters crime for individuals with really low morality. This view is certainly contrary to the common understanding of the operation of self-control in our discipline, namely that high self-control makes crime less likely (Pratt and Cullen, 2000). But, although their own findings do not support the assumption of a criminogenic self-control effect among respondents characterized by the lowest levels of morality (in fact self-control unfolds its greatest crime-reducing impact in this subgroup), this line of thinking has great potential to fertilize future inquiries. It may well be that these and other authors observe larger crime-inhibiting effects of the ability to exercise self-control among respondents of lower morality because the sampling frames underlying their studies do not include individuals holding strong antisocial moral beliefs. As outlined by Pauwels and colleagues in their review of previous research on SAT, hitherto most tests have relied on samples from Europe, where the proportion of individuals with explicitly deviant moral values may be infinitesimally small. In any event, a promising avenue for future research dedicated to the interplay of morality and self-control will be to move away from the testing of linear interaction effects and switch to the examination of nonlinear moderation relationships.

The principle of moral correspondence posits that controls are irrelevant when there is congruent rule-guidance, that is, when both personal morality and the moral context concordantly encourage or discourage criminal behaviour. It appears very plausible that self-control does not affect the course of action when both personal morals and the moral context exclude engaging in criminal behaviour, because in this case crime does not enter the set of perceived action alternatives. If crime is not considered as an action alternative, controls are not involved in the process of choice.

However, things become more complicated when both personal morals and the moral context support criminal behaviour. In this case, the reason for the insignificance of controls is not totally clear. Is it because when personal morality and moral context concordantly demand criminal behaviour, law-abidance is not a perceived course of action, as Kroneberg and Schulz point out? This interpretation is in line with the described workings of the moral filter. When individuals consider no other action alternatives than criminal ones, controls may lack any relevance. Or is it because of the inherent qualities of criminal behaviour that are outlined by Gottfredson and Hirschi (1990) and pictured by Katz (1988)? Does the capacity of criminal behaviour to bring quick and simple pleasures without much effort trump the influence of controls and render crime irresistible? Or does self-control remain out of play because in this constellation rule-guided creatures tend to act habitually?

On the other hand, the empirical findings presented by Schepers and Reinecke in this Special Issue demonstrate that a strong ability to exercise self-control also reduces offending under these conditions. Based on student samples from Germany, the authors show that self-control has its largest crime-inhibiting effect among respondents characterized by both low personal morality and a weak moral context. This observation challenges the moral correspondence hypothesis, but may again be due to the failure to include individuals with genuinely antisocial morals and predominant exposure to settings where crime is normative in their study. Finally, it may also be argued that, when both a person’s moral rules and the moral norms of the setting encourage breaking a rule of conduct, self-control as the capability to adhere to one’s own morals may actually foster breaking this rule. In this context, self-control may increase the likelihood of criminal behaviour because it helps to bring behaviour into line with a deviant morality.

SAT clearly maintains that without moral conflict or incongruent rule-guidance there is nothing to control, which squares with its definition of self-control, but does this really leave us with the full picture of the crime-reducing potential of an individual’s self-regulatory forces? Are there other higher-order standards (aside from a person’s moral rules) that result in crime-dampening self-control effects, as outlined by Hay and Meldrum (2016)? ² Could SAT’s understanding of self-control be too narrow? ³ We cannot claim to solve these problems here, but recommend that future research in the framework of SAT contributes to clarifying these issues.

The nature of the empirical findings obtained so far may reflect the common practice of measuring self-control in studies on SAT. All previous analyses of conditional self-control effects suffer from a deficient measurement of self-control. There clearly is a discrepancy between the conceptualization and operationalization of self-control in research related to SAT. Hitherto, all tests have relied on attitudinal measures of the capability to exercise self-control, which are strongly inspired by the Grasmick et al. (1993) scale, notwithstanding the fact that this scale was developed to capture self-control in the sense of Gottfredson and Hirschi (1990) and at best depicts the ability to resist current temptations and provocations. So in fact these studies tap into an individual’s ‘tendency to avoid acts whose long-term costs exceed their momentary advantages’ (Gottfredson and Hirschi, 1994: 3). This may explain why crime-reducing self-control effects have been found for individuals with the lowest levels of morality as well as for individuals subject to low personal morality and a weak moral context. When persons perceive crime as a viable action alternative, their tendency to take the long-term consequences of behaviour into consideration may indeed decrease their level of crime involvement. However, since the corresponding instruments provide only a very inaccurate and insufficient operationalization of the ability to withstand external pressures to act against one’s own personal morals (Wikström and Treiber, 2016), the findings gained with it may not tell the full story of the dynamics underlying the workings of self-control as they are specified in SAT. Therefore we strongly advise the next generation of studies to reconsider the operationalization of self-control and to draw on measures that better match the conceptualization of self-control in SAT.

Apart from the criminogeneity of the setting a person takes part in, an individual’s crime propensity (formed by his or her personal morality and his or her ability to exercise self-control) is one of the key inputs to the perception–choice process. To determine an individual’s level of crime propensity, many studies operating in a SAT framework rely on an additive proxy measure in which personal morality and self-control ability (in fact Z-scores of two scales) are summed up (for example, Hirtenlehner and Treiber, 2017; Svensson and Pauwels, 2010; Wikström et al., 2012). However, given SAT’s definition of the ability to exercise self-control, such a procedure is not unproblematic. It appears very convincing that individuals characterized by both a high law-relevant morality and well-developed self-control ability rank on the crime-averse end of the propensity continuum. But are people with a low or deviant morality and a low capacity for self-control really the most crime-prone ones? Can’t it be that individuals with a low or deviant morality and a strong ability to implement their morals into behaviour are more inclined to perceive and choose crime as an action alternative? Again, in light of the currently employed instruments to operationalize the ability to exercise self-control, it does not come as a surprise that the additive proxy measure of crime propensity works well and produces the expected results. But, when making use of more refined measures that really tap into the capability to act in accordance with one’s own moral rules, a different procedure of combining morality and self-control ability may be necessary.

Serrano-Maíllo’s Latin American investigation addresses the interplay of an individual’s tendency to see crime as action alternative and his or her ability to exercise self-control. Embedded in a moral filter framework, this study determines the frequency of perceiving crime as an alternative by the level of temptation to engage in criminal behaviour. The results suggest that self-control decreases offending particularly among individuals who are more likely to perceive crime as an action alternative. This accords well with the reasoning that moral rules represent the first line of defence against acts of crime and controls are only the second line of defence (Brauer and Tittle, 2017).

Gerstner and Oberwittler shift the focus on the role of peers in SAT. Drawing on both peer-reported and perceptual measures of peer delinquency, they investigate the interplay of crime propensity, unstructured socializing and association with crime-prone peers. Employing the direct measurement of peer delinquency, they find significant evidence for a three-way interaction according to which unstructured socializing with same-aged friends exerts its greatest influence on offending when the number of delinquent peers is large and the level of crime propensity is high.

In the last contribution to this Special Issue, Pauwels replicates a randomized scenario study aimed at analysing SAT’s perception–choice process that was originally conducted by Wikström and colleagues (2012). The results show that the impact of a scenario’s criminogeneity (determined by the levels of provocation and monitoring) on the likelihood of choosing a violent response depends on an individual’s propensity for crime.

In sum, the works assembled here illustrate that SAT has succeeded in bringing scholars together under a unifying framework that is highly promising. Given the results obtained so far, there is no reason for a paradigm shift. However, the necessity of some refinements and clarifications as well as methodological improvements cannot be denied. We hope to stimulate a fruitful further development of both the theory and related research with this special issue of the European Journal of Criminology.