The Combined Roles of Moral Emotion and Moral Rules in Explaining Acts of Violence Using a Situational Action Theory Perspective

Abstract

The roles of shame and guilt, and their relationships to empathy, have not been modeled adequately as key factors in moral decision-making in the study of violence. The role of moral emotion has been neglected in existing criminological research and this study seeks to develop current explanations of the comprehensive myriad of factors that play a role in moral crime decision-making. This research will test the different roles of empathy, shame, and guilt in violence decision-making using a situational action theory (SAT) perspective. Data taken from the Peterborough Adolescent and Young Adult Development Study (PADS+), a longitudinal study with a large representative sample, provide quantitative questionnaire indices to enable comparison of a persistent and frequent violent offender subsample (N = 48) with the remaining PADS+ study sample (N = 607). A striking majority of violent offenders report that they do not think it is wrong to commit violence, and do not care about it, that is, they lack shame and guilt, and report that violence comes as a morally acceptable and natural action alternative. Furthermore, violent offenders do not register the predicament of their victims; there is a distinct lack of empathy. This article demonstrates a key finding which has rarely been explored to date; regression analyses reveal an interaction effect whereby individuals with weak shame and guilt, combined specifically with weak moral rules, are more likely to commit acts of violence. The study findings provide strong support for the SAT of the role of weak morality in violence decision-making. To reduce the possibility of crime being seen as an action alternative, moral development programs should be developed and administered in childhood.

Keywords

empathy shame guilt moral emotion morality violence decision-making

Introduction

There are few existing criminological theories that directly incorporate the role of moral emotion or morality in the study of crime. Some criminologists, such as Bottoms (2002, p. 24) have argued that “If they are to be true to their calling, all criminologists have to be interested in morality,” but it remains to be explored in depth in criminological research (Tittle, 2007; Wikström, Oberwittler, Treiber, & Hardie, 2012; Wikström & Treiber, 2009a; Wikström, Treiber, & Hardie, 2012). Typically, criminological theories that have focused on individual or person-level factors have ignored the role of morality (including moral emotion);rather, the focus has been on self-control (Gottfredson & Hirschi, 1990), strain (Agnew, 1985), and/or rational choice in weighing costs and benefits (Cohen & Felson, 1979; Cornish & Clarke, 1986). This has led to an underestimation of the role of the more intuitive, unconscious role of moral emotion in criminal decision-making (Treiber, 2011). A purely rational view is unrealistic in real-life terms because emotions undoubtedly play a role in many everyday action tendencies (Blair, 2017; Damasio, 1994), to which criminal behavior is no exception (Clay-Warner, 2014; Dippong & Fitch, 2017). Furthermore, by viewing emotion as a variable that can change depending on the specific moral evaluation in question rather than as a constant, new insight can be gained into the moral decision-making process (Helion & Ochsner, 2016).

Although strain theory (Agnew, 1985, 2001, 2014), reintegrative shaming theory (Braithwaite, 1989, 2000; Braithwaite & Braithwaite, 2001; Sherman, 1993), and E. Anderson (1999, see also C. A. Anderson & Bushman, 2002) do partially acknowledge the role of emotion, often they do not elaborate or provide underdeveloped explanations of the mechanisms that link moral emotion to crime.¹ Of the studies that have examined the relationship between morality and crime outcomes, they have generally used moral values or rules measures and the next useful steps are to measure and to incorporate moral emotions (Wikström & Svensson, 2010). Existing research that will be outlined below often fails to address various issues: first, existing research studies explore one or two emotions in isolation but do not account for the role of the remaining moral emotions. Second, studies do not delve into a detailed discussion or interpretation of their findings; that is, they fail to discuss the proposed relationship between moral emotions and other key variables. For example, shame and guilt are viewed as deterrents in crime prevention (Braithwaite, 1989), rather than as key contributors to the strength of overall morality that inform whether acts of crime are viewed as morally acceptable.

Social learning theory (Bandura, 1977) dictates that young children learn from, mirror, and imitate the values, attitudes, and actions of people that they spend time with and the theory can provide compelling arguments for the importance of the role of moral behavior of significant others in the development of personal morality. Increases in moral emotion capacities are likely to be facilitated by moral displays from significant others in childhood and adolescence, including parenting style, school-based teaching style, and relationships with peers (Svennson, Pauwels, Weerman, & Bruinsma, 2016).

Situational action theory (SAT) supports the notion that morality is a fundamental predictor of crime involvement (Wikström et al., 2012). SAT is the most suitable theoretical model for empirical testing of the role of moral emotion in crime because it fully acknowledges the crucial importance of various moral measures of interest (Pauwels, Svensson & Hirtenlehner, 2018); moral rules, moral emotion, and empathy. A further reason to adopt the SAT framework is that its perception–choice process offers a mechanism by which moral emotion may influence crime decision-making. According to SAT, individual-level factors and environmental-level factors both come together in the perception–choice process that leads to an act of crime (Wikström & Treiber, 2009a). The premise of the perception–choice process is that different people (e.g., with different levels of moral emotion) act in different environments, and their interaction is fundamental to explaining behavior, including crime.² The perception–choice mechanism involves a two-part process; the perception of action options (specifically influenced by moral emotion), followed by the selection (choice) of one of them to act upon (Wikström & Treiber, 2009a).

This research seeks to build upon existing explanations of the roles of shame, guilt, and empathy in offending decisions by examining how they interrelate via moral reasoning. The need for this attempt to disentangle the role that empathy plays in violence has been highlighted (Jolliffe & Murray, 2012). This will be achieved by integrating the various moral measures into SAT’s perception–choice process in an attempt to offer more understanding of the offender decision-making process.

The Roles of Empathy, Shame, and Guilt in Crime

Existing research confirms shame and guilt to be of crucial importance in the explanation of offending (Treiber, 2013), and furthermore, they are found to mediate the effects of other variables (such as parental monitoring) in explaining crime (Svensson, 2004; Svenson, Pauwels, Weerman, & Bruinsma, 2016; Svensson, Weerman, Pauwels, Bruinsma, & Bernasco, 2013). Wikström and Svensson (2008) found that both in an English and Swedish sample weak shame predicted violence, and other research finds that weak shame predicts antisocial behavior (Olthof, 2012). Weak guilt and shame are correlated to aggression (Pornari & Wood, 2010; Roos, Salmivalli, & Hodges, 2011), the intention to engage in a deception (Seiter & Bruschke, 2007), and offending itself (Schalkwijk, Stams, Stegge, Dekker, & Peen, 2014). Shame and guilt have “presumed roles in inhibiting immoral, socially undesirable behaviour and in fostering altruistic, prosocial behaviour” (Tangney, Stuewig, & Hastings, 2011, p. 1). A meta-analysis of 25 studies showed significant negative associations between guilt and delinquency (r = −.278) and between shame and delinquency (r = −.130) regardless of the type of shame and guilt measure or the type of crime (Spruit, Schalkwijk, van Vugt, & Stams, 2016).

However, not all research finds evidence of a link between shame and crime; some contrasting research findings state that shame does not play a role in rule-following behavior because it motivates devaluation, defensiveness, anger, and aggression (Tangney & Fischer, 1995). Similarly, there is work that has linked shame to more, rather than less, aggression and violent behavior Elison, Garofalo, and Velotti (2014).³ Although these findings are in the minority, they oppose the theoretical propositions of the current study and highlight that the relationship between shame and crime may be more complex than the relationship between guilt and crime.

The relationship between guilt and offending behavior has received more research attention than shame. Weak guilt is found to predict violence (Beyers, Loeber, Wikström, & Stouthamer-Loeber, 2001; Krettenauer, Campbell, & Hertz, 2013; Tibbetts, 2003).Weaker guilt is reported by individuals with antisocial personality disorder compared with matched controls (Dinn & Harris, 2000), and individuals with weak guilt are more likely to engage in unethical behavior and delinquency (Cohen, Panter, Turan, Morse, & Kim, 2013; Malti, Gasser, & Buchmann, 2009; Sobhani & Bechara, 2011), in risky and illegal behaviors, have more time spent in prison, more criminal convictions (Tangney et al., 2011), and engage in substance use and criminal behavior (Tangney & Fischer, 1995). Children that commit rule-breaking behavior are found to lack guilt from 3 to 4 years of age (Frick, Ray, Thornton, & Kahn, 2013) and furthermore, feelings of guilt about past violence have been found to prevent subsequent involvement in violence (Bowes & McMurran, 2013). Malti and Krettenauer (2013) carried out a meta-analysis of 42 studies (over 8,000 participants) and found that guilt in particular plays a significant role in the moral behavior (including antisocial behavior) of children and adolescents. The link between guilt and various levels of delinquent behavior has been replicated using various methodologies, including hypothetical scenarios (Krettenauer, Asendorpf, & Nunner-Winkler, 2013) and self-reports and observer reports of guilt (Cohen et al., 2013).

In addition to shame and guilt, empathy is also believed to play a fundamental role in morality (Eisenberg, 2000). The role of empathy in crime has attracted attention in recent years (Posick, Rocque, & Rafter, 2012). In his 2010 American Society of Criminology presidential address, Francis Cullen urged criminologists to “recognize the importance of examining not only negative but positive emotions, such as empathy” (Cullen, 2011, p. 314). Although researchers state that empathy is a core concern in understanding criminality (Hepper, Hart, Meek, Cisek, & Sedikides, 2013), further work can develop understandings of the role of empathy in crime decision-making (Jolliffe & Murray, 2012).

Empathy has been found to be a key factor in distinguishing between prosocial behavior and nonprosocial behavior, including offending (Baron-Cohen, 2011; Bock & Hosser, 2014; Brauer & Tittle, 2017; Frick & Morris, 2004; Hoffman, 2000; Jolliffe & Farrington, 2004; Malti et al., 2009; Marshall, Marshall, Serran, & O’Brien, 2009; Miller & Eisenberg, 1988; Posick et al., 2012; Spinella, 2005). Furthermore, individuals with higher empathy show moderate-strong negative associations with aggressive behavior (Vachon & Lynam, 2016), especially for offenders (Llorca-Mestre, Malonda-Vidal, Samper-Garcia, 2017). Furthermore, those with higher empathy are found to be more successful in reducing aggression and delinquency (Crockett, Clark, Hauser, & Robbins, 2010). Lack of empathy is renowned to be the central characteristic of violent psychopathic behavior (N. Anderson & Kiehl, 2012; Blair, 2017; Blair, Mitchell, & Blair, 2005; Farrington, 2005; Fine & Kennett, 2004). The application of empathy has been found to be particularly impaired in violent situations; nonmoral emotions such as fear and anger can lead to an increase in self-focus and a corresponding decrease in other-focus, that is, in empathy and shame (Barnett & Mann, 2013).

It is important to note that other research finds weak-moderate correlations between empathy and antisocial behavior (Day, Casey, & Gerace, 2010; Jolliffe & Farrington, 2004; Miller & Eisenberg, 1988), including substance abuse and involvement in crime (Jolliffe & Murray, 2012; Jolliffe & Farrington, 2006; Schalkwijk et al., 2014; van Langen, Wissink, van Vugt, Van der Stouwe, & Stams, 2014). It remains to be explored whether this is because empathy has an indirect link to crime via its role in the ability to exercise moral emotion, or whether existing empathy measurement tools are flawed (Vachon, Lynam, & Johnson, 2014).

The Roles of Empathy, Shame, and Guilt in Crime Involvement in a SAT Framework

To date, the role of moral emotion in crime involvement has not been as well developed as other parts of SAT (Wikström, 2005, 2006, 2009, 2010, 2014; Wikström et al., 2012; Wikström & Svensson, 2008; Wikström & Treiber, 2009a; Wikström, Tseloni, & Karlis, 2011). Although the theory prioritizes the importance of clear definitions of concepts and mechanisms, further work is required to build a comprehensive explanation of the specific role of moral emotion in crime.

Moral acts are defined as acts that are governed by what is right or wrong to do in a particular circumstance (Wikström, 2006; Wikström & Treiber, 2009a). Acts of crime are characterized as the breaking of moral rules as defined by law; that is, the law prescribes which acts are legally permissible. This study will focus on violence as an example crime type, but crucially, the role that moral emotion plays in the occurrence of crime is hypothesized to be the same regardless of the crime type (see Wikström & Treiber, 2009a; Bouhana & Wikström, 2008; Wikström et al., 2011).⁴

According to SAT, the moral emotions that are relevant for an explanation of crime decision-making are shame and guilt because these moral emotions are widely recognized to be relevant to moral behavior (Eisenberg, 2000; Tangney & Fischer, 1995; Tangney, Stuewig, & Mashek, 2007). Guilt is defined as a negative feeling, often experienced as a result of an action, which is felt inwardly toward oneself (Wikström et al., 2012). Guilt can be anticipated on contemplation of an action and is measured along a spectrum; strong guilt indicates that one cares very much about the potential breaking of a moral rule, whereas weak guilt indicates that one does not care very much about the potential breaking of a moral rule. Anticipated shame is defined as a negative feeling, not necessarily experienced as a result of an action, which is felt in the presence or consideration of others (Wikström et al., 2012). Anticipated shame can also be measured along a spectrum; strong shame indicates a perception that others care very much about the potential breaking of a moral rule, whereas weak shame indicates a perception that others do not care very much about the potential breaking of a moral rule.⁵

According to SAT, actions are seen as morally acceptable or unacceptable depending on the act-specific moral rule and the act-specific moral emotion (Wikström & Treiber, 2009a). Shame and guilt are hypothesized to contribute to moral rules to form overall individual morality; if one does not feel it is wrong to commit an act and subscribes to that weak moral rule with weak emotional intensity (feels weak shame or guilt regarding the moral act in question) this taken together will play a role in whether the act of offending will take place (see Figure 1). Essentially, moral emotion and moral rules can be viewed as two measures of the overall construct of morality; whether a personal subscribes to a moral rule (moral rules) and the intensity to which they subscribe to the moral rule (moral emotions). Individuals with weak overall morality are more likely to see crime as a morally acceptable action alternative.⁶ Many people simply do not perceive crime to be a potential behavioral action alternative and this is primarily determined by their strong moral emotions and moral rules (Wikström et al., 2012); it follows logic that if crime is not perceived as an action alternative, crime will not occur (Wikström & Treiber, 2009a).⁷

Figure 1.

Conceptual model of the relationships between empathy, shame, and guilt and the decision to engage in crime.

This article adds nuance to the existing field of research that finds an association between moral emotion and offending. This article also contributes to the existing SAT framework by introducing the concept of empathy and its relationship to shame and guilt (see Figure 1). The current study defines the ability to exercise empathy as an emotional process and trait (as opposed to a state or emotion).⁸ A strong ability to exercise empathy is hypothesized to play a role in the increased potential to feel shame and guilt during the crime decision-making process. Inversely, if an individual has a weak ability to exercise empathy, this is hypothesized to play a role in the reduced possibility to feel shame and guilt during the crime decision-making process. Without consideration of the consequences of actions for others, or a perception of the judgment that others would make, feelings of shame and guilt may not arise (see Blair, 2017).⁹ Empathy involves a personal evaluation that requires knowledge or perception of how others would view the act in question; it is variable depending on the circumstances.

Empathy plays a different role to shame and guilt in the decision-making process that leads to an act of violence.¹⁰ The current study hypothesizes that a poor general ability to exercise empathy will reduce the possibility to feel shame and guilt.¹¹ If the crime-specific moral rule is weak-moderate, weak shame and guilt, by way of providing a low intensity to which they subscribe to the moral rule, results in weak overall morality and allows crime to be seen as a morally acceptable and viable action alternative. Inversely, it is hypothesized that if one is aware that their subsequent action will cause emotional or physical harm to someone via a strong ability to exercise empathy, this increases the possibility to feel negative feelings of shame and guilt, and overall morality will be strengthened (see Figure 1). In this case, it is hypothesized that violence will not be perceived as an action alternative because it is not seen as a morally acceptable behavior and violence will not occur.

Method

Data were collected with the Peterborough Adolescent and Young Adult Development Study (PADS+); a longitudinal study that follows 700 young people from the age of 12 into adolescence and adulthood and was designed to test the role of morality among other numerous factors in crime involvement. The sample was recruited in 2002 using a random sampling method from a pool of young people listed to begin Year 7 at school (school sampling frame).¹² At age 12 when the study began, there were equal numbers by sex, 83% were White, and family occupational social class was varied (22% lower working class, 39% working class, 34% middle class, 5% upper class). In the current study, seven waves of questionnaire data are used from 2003-2012. The retention rate for the full study sample has remained very high (92% at Wave 7). Peterborough is a medium-sized U.K. city located in Cambridgeshire with a typical crime rate (per 1,000 population) when compared with other similar sized areas in the United Kingdom (Office of National Statistics, 2018). There is variation in area deprivation, that is, it contains areas of high and low disadvantage, and there is a mixed ethnic composition (particularly, people of Pakistani, Indian, Irish, Polish, Italian, and Caribbean descent) within the population. This variation in ethnicity is comparable with the rest of the United Kingdom (10.3% of Peterborough residents responded as non-White compared with 9.1% U.K. average, Office for National Statistics, 2011).

All of the data presented were collected using a structured paper questionnaire which was administered in person at the participants’ schools, homes, or public locations (such as library conference rooms). Questionnaire scales included empathy, shame, guilt, moral rules, violence, and total crime scales (all of which are described in Trivedi-Bateman [2015]). The general empathy scale constitutes 17 items using examples from specific, everyday occurrences (e.g., “Other people’s problems are theirs, not mine”) and was developed specifically for the current study. The scale includes a combination of cognitive empathy items (identifying another person’s viewpoint) as well affective empathy items (feeling emotional congruence with another person’s viewpoint). The scale items refer to a range of small embarrassments to more serious or emotionally painful situations. The response options are “Strongly agree,” “Mostly agree,” and “Mostly disagree,” and “Strongly disagree.” Descriptive statistics indicate that there is variation in the data (normal distribution), that the internal consistency of the scale is reasonably high and that the scale items are closely related with a Cronbach’s alpha of .81.

The shame scale constitutes six items for which participants report how ashamed they would feel (in front of parents, teachers, or best friends) if they were caught shoplifting or breaking into a car. For example, “If you were caught shoplifting and your parents found out about it, would you feel ashamed?” Although the items on the shame scale specifically include shoplifting and car crime as example crime types, the total scale score is assumed to represent “general” shame with regard to committing all acts of crime because they can be grouped together as moral transgressions. The guilt scale constitutes six items for which participants report how guilty they would feel for committing various acts, including moral transgressions not defined by law (legal acts), as well as acts of crime (illegal acts). For example, “Would you feel guilty if you broke into a car and stole something?.” Although the scale items measure anticipated shame and guilt, this is not viewed as problematic for the theoretical aims of the current study; anticipated shame and guilt is found to be a good indicator of experienced shame and guilt in real-life situations, as is evidenced in a comparison of qualitative and quantitative data in Trivedi-Bateman (2015). For the shame and guilt scales, the response options are “No, not at all,” “Yes, a little,” and “Yes, very much.” Both scales were developed in 2004 by Professor Wikström (Director of the PADS+) and administered in Waves 2 to 7.

The population distribution for the shame scale is negatively skewed; indicating that the majority of the participant sample reports strong shame in response to carrying out hypothetical moral transgressions and acts of crime. Data from both scales indicate that there is variation in the data (normal distribution for guilt scale), that the internal consistency of the scale is reasonably high and that the scale items are closely related (see Table 1). The shame and guilt scale alphas (Cronbach’s alpha of .94 for shame scale and .88 for guilt scale) were calculated by summing six total scale item scores for each of the six waves (ages 14-21). A comparison of the shame and guilt scale reveals that the mean guilt score is lower and there are no large differences in the range and standard deviation. A promax factor analysis of all Wave 7 shame and guilt scale items (six shame items and six guilt items) confirms that shame and guilt can be treated as distinct concepts; with “general shame” (four of the shame items loading above .4), and “general guilt” (five of the guilt items loading above .4) emerging as distinct factors.

Table 1.

Comparison of Mean Scores for the Violent Subsample (N = 48) and the Rest of the Sample (N = 607): Empathy, Shame, Guilt, Moral Rules, and Crime.

	Violent Subsample	Rest of the Sample	Difference Between Means	t-Test Significance	η² × 100
General empathy (total score 51)	29.4	33.8	−4.4	.000	3
Shame (total score 12)	8.7	10.8	−2.1	.000	6
Guilt (total score 12)	7.3	9.7	−2.4	.000	8
Moral rules (total score 48)	25.5	30.1	−4.6	.000	3
Assault frequency (unlogged)	50.7	3.8	46.9	.000	18
Total crime frequency	192.3	14.5	177.8	.000	18

Note. –η² × 100 is reported to provide the percentage of variance explained for ease of interpretation.

The moral rules scale constitutes 16 items for which participants report how right or wrong they think certain acts (of differing severity) are in particular circumstances. As with the guilt scale, the item acts include moral transgressions not defined by law (legal acts, for example, “How right or wrong do you think it is to lie, disobey, or talk back to teachers?”) as well as acts of crime (illegal acts, for example, “How right or wrong do you think it is to smash a street light for fun?”). The response options are “Not wrong at all, “A little wrong,” “Wrong,” and “Very wrong.” The moral rules scale is a modified version of the Pittsburgh Youth Study prosocial values scale (Loeber, Farrington, Stouthamer-Loeber, Moffitt, & Caspi, 1998); it was developed in 2001 by Professor Wikström and administered in Waves 1 to 7. Descriptive statistics indicate that there is variation in the data; that the internal consistency of the scale is reasonably high, and the scale items are closely related. Cronbach’s alpha varies from .88 to .90 for Waves 1 to 7 (see Table 1).

Assault frequency (Waves 1-7) and total crime (which covers several crime types including vandalism, arson, and burglary) frequency data have been used in the current study. Moral emotion and moral rules are modeled to predict past committed violence with the understanding that first, moral emotions and moral rules are found to be relatively stable over time (see Wikström et al., 2012) and second, that past violence is an indicator of future violence. The prevalence of assault was 58% (n = 373) in Waves 1 to 7 combined, that is, 58% of participants had self-reported at least one act of assault from age 12 to age 21, ranging from minor to more serious incidents.

Participants were selected for inclusion in the violent subsample if they had reported violence (assault and/or robbery) in the preceding interview wave (Wave 6, 2010, aged 19) and exceeded two or more of the criteria thresholds regarding violence frequency (15+), violence duration (5+ interview waves), and violence age of onset (below age 11). Early age of violence onset (onset before age 11) was included in the sampling criteria because existing PADS+ findings state that early age of onset offenders, compared with late onset offenders, are responsible for the majority of crimes from interview Waves 1 to 5, are more likely to continue offending as they get older, and are more likely to offend with a higher frequency (Wikström et al., 2012). To increase the likelihood that most active violent offenders had been captured in the subsample, participants with medium-high violence (violence frequency of 10+, N = 5) in the immediately preceding waves (Interview Waves 5 and 6) were also added to the sample. Qualitative interviews were conducted specifically to explore the real-life circumstances under which persistent and frequent violent offenders carry out violence. The rationale for the interview subsampling was that to facilitate discussion of recent violence during interview, prolific offenders were required. The majority of the participant sample was male (39 male, nine female). There is no significant association between the ethnicity, χ²(1) = 1.3, p > .05, or family social position (which is measured using parental income, parental education, and parental employment level), χ²(4) = 6.4, p > .05, between the violent subsample (N = 48) and the rest of the PADS+ study sample (N = 607).

Results

For all analyses, the higher the empathy, shame, and guilt score, the stronger the empathy, shame, and guilt reported; and the higher the moral rules score, the more young people think it is wrong to commit moral transgressions and acts of crime. The data analytic plan was as follows: first, investigate whether empathy is required for shame and guilt (ordinary least squares [OLS] regression). Second, explore whether there is a difference in empathy, shame, guilt, and moral rules between the violent subsample and the rest of the sample (estimates of effect size). Third, test whether shame, guilt, and moral rules predict violence involvement in the full sample (OLS regression). Finally, and crucially, explore whether moral emotion, moral rules, and their interaction, predict involvement in violence for the full sample (OLS Regression Models 1 and 2).

Basic control variables are not included in the regression analyses because they are not theorized to be explanatory variables of violence or crime. For example, there is no causal relationship between gender and crime (being male does not dictate that you will commit crime, but rather certain factors that relate to being male can play a role in the development of morality, as well play a role in the choice of settings that males may partake in). In contrast, other variables which are deemed to be theoretically relevant are not included here because they are only specifically important under limited and particular setting circumstances and have been tested extensively elsewhere (see Pauwels et al., 2018 for a review).¹³

Is Empathy Required to Anticipate Shame and Guilt?

Using OLS regression (enter method) for the full sample (N = 655), general empathy significantly predicts shame (R² = .14, β = .376, p < .01) and guilt (R² = .19, β = .440, p < .01). This provides evidence of a basic association that supports the hypotheses of the current study; that general empathy plays a significant role in the possibility to feel shame and guilt. It should be noted that R² values are low; a path model is constructed in Trivedi-Bateman (2015) and reveals the strongest path is from general empathy, mediated by guilt, to violence (as opposed to a direct link to violence). This provides support for the two-part mechanism proposed in the current study; that empathy plays a role in the possibility to feel guilt, and that guilt plays a role in violence.

Is There a Difference in Empathy, Shame, Guilt, and Moral Rules Between the Violent (Prolific Violence) Subsample and the Rest of the Sample (Zero-Moderate Violence)?

The violent subsample participants report significantly lower mean scores for empathy, shame, guilt, and moral rules, than the rest of the sample at Wave 7 (aged 21; see Table 1.).¹⁴ The maximum scale score is reported in brackets next to each data measure. The assault frequency and total crime frequency measures are total additive scores of measures collected at Waves 1 to 7 (ages 13-21). The estimates of effect size or η² (which measure the strength of the difference between the means for the two samples) are higher for shame and guilt than for the empathy measures. This indicates that there is a stronger difference between shame and guilt (compared with empathy) between the violent subsample and the rest of the sample. This provides evidence of a basic association that supports the hypotheses of the current study that empathy plays an indirect role in violence while shame and guilt are more proximal to an explanation of violent behavior.¹⁵ There is a significant negative correlation between moral emotion (shame and guilt combined¹⁶) and assault frequency for the violent subsample (R² = .22, correlation = −.47, p < .001) and the rest of the study sample (R² = .17, correlation = −.41, p < .001). It should be noted that R² values are low; this is because moral rules, as well as settings factors not explored in the current study, are also deemed to be of crucial importance in offending decisions. For the violent subsample in particular, there is a clear negative correlation, that is, individuals with weaker shame and guilt report more self-reported assault. Subsequent analyses will first illustrate the importance of both moral rules and moral emotion in explaining violence, and second, provide evidence for an interaction effect between moral emotion and moral rules in predicting overall crime frequency.¹⁷

Do Shame, Guilt, and Moral Rules Predict Involvement in Violence for the Full Sample?

A regression analysis (OLS) indicates the extent to which shame, guilt, and moral rules together can predict violence involvement for the full PADS+ sample (see Table 2). Model I explains 31% of the variance in assault frequency; guilt is a significant predictor (β = −.507, p < .001), however, shame is not a significant predictor (β = −.062, p = .280).¹⁸ Model II includes moral rules which increase the explained variance to 34%; guilt is a significant predictor (β = −.357, p < .001); however, shame, as with Model I, is not a significant predictor (β = −.026, p = .645), and moral rules is a significant predictor (β = −.238, p < .001).

Table 2.

Regression Predicting Assault Frequency (Logged) by Shame and Guilt (Model I) and Shame, Guilt, and Moral Rules (Model II) (N = 655).

	R ²	B	β	Significance
Model I
	.31
Guilt		−.020	−.507	.000
Shame		−.002	−.062	.280
Model II
	.34
Guilt		−.014	−.357	.000
Shame		−.001	−.026	.645
Moral rules		−.003	−.238	.000

This finding provides support for the hypotheses of the current study which state that the strength of the relationship between shame and violence is moderate compared with the strength of the relationship between guilt and violence. A significant correlation between shame and assault does exist; this indicates that the specific role of shame may not be as straightforward as the role of guilt. It is important to note that although shame is not a significant predictor of violence on the quantitative scales, in contrast, shame is relevant for almost all violent subsample participants in their qualitative recollections of real-life acts of violence (see Trivedi-Bateman, 2015). When moral rules are added to the regression model (Model 2), the explained variance increases slightly; therefore, moral emotion should be explored in association with moral rules (as outlined in the theoretical framework of the current study). This relationship is tested further in the following section by modeling an interaction effect between moral emotion and moral rules.

Do Moral Emotions, Moral Rules, and Their Interaction Predict Involvement in Violence for the Full Sample?

The current study hypothesizes that moral emotions are of crucial importance, but only when combined with other relevant factors (Lapsley & Carlo, 2014), such as moral rules (and the setting, which is explored elsewhere; Wikström et al., 2012). An interaction term offers an exploration of whether the effect of an explanatory variable (moral emotion) on the outcome variable of violence is contingent upon another explanatory variable (moral rules). This method allows testing of whether moral emotion is only relevant to violence when moral rules are specifically weak. If this is the case, it raises clear lines of enquiry and focus for how and where to target prevention. Very little is known about whether the criminogenic effect of moral emotion is dependent on moral rules; Do strong moral rules protect from the criminogenic effect of weak moral emotion? Are young people with weak moral rules more susceptible to the criminogenic effect of moral emotion?

OLS regression method for testing the interaction is deemed to be the most suitable and appropriate for data of this kind (Svensson & Oberwittler, 2010). However, use of interaction terms does not come without its problems (Hirtenlehner, Pauwels, & Mesko, 2015; Svensson & Oberwittler, 2010). First, the skewed distribution of the outcome variable of self-reported violence violates key assumptions (multivariate normal distribution and homoscedasticity) of linear regression analyses. It has been proposed that spurious interaction effects may arise due to the distribution of self-reported violence (Osgood, Finken, & McMorris, 2002). However, OLS regression is particularly robust to this (Hirtenlehner et al., 2015) and due to the non-normal distribution of self-reported violence, robust standard errors have been calculated (Hannon & Knapp, 2003). When standard errors are reestimated to be heteroscedasticity-consistent using the robust technique, both moral emotion and moral rules have a significance value below .05. In addition, quadratic terms (that account for the nonlinear relationship of moral emotion and moral rules to self-reported violence) are calculated to enable estimation of an interaction effect that is not obscured. The dependent variable is also not log transformed to prevent further difficulties in identifying an interaction effect (McClelland & Judd, 1993). Finally, the predictor variables have been mean-centered to allow for more meaningful interpretation of the regression analyses.

For the full PADS+ sample, Regression Model I analyses moral emotion (β = −.114, p = .042) and moral rules (β = −.230, p < .001) as predictors of assault frequency (unlogged) and explains 11% of the variance (see Table 3).¹⁹ Model II gives a higher explained variance of 19% and incorporates the interaction term (β = .313, p < .001), as well as moral emotion (β = −.043, p = .426) and moral rules (β = −.165, p < .001). This increase in explained variance can be attributed to the interaction of moral rules with moral emotion. Of key relevance, when the interaction term is added, moral emotion is no longer a significant predictor of assault; this indicates that it is the interaction between moral emotion and moral rules that is the fundamental predictor of violence.

Table 3.

Regression Predicting Assault Frequency (Unlogged) by Moral Emotion and Moral Rules (Model I) and Moral Emotion, Moral Rules, and Their Interaction (Model II) (N = 655).

	R ²	B	β	Significance
Model I
	.11
Moral emotion		−0.131	−.114	.042
Moral rules		−0.183	−.230	.000
Model II
	.19
Moral emotion		−0.049	−.043	.426
Moral rules		−0.131	−.165	.002
Moral emotion × Moral rules (interaction)		1.19	.313	.000

Conclusion and Discussion

The current study makes three overarching conclusions; violent offenders report a weak ability to exercise empathy and a weak ability to anticipate shame and guilt, weak empathy is related to a reduced ability to experience shame and guilt, and weak moral emotion makes a significant contribution to predicting violence, specifically when moral rules are also weak (this is a key finding which has not been explored to date).

This research develops explanations that have previously received little attention in relation to moral emotion. It develops a detailed explanation of the role of moral emotion in crime involvement by first exploring the role of ability to anticipate empathy in the possibility to feel shame and guilt, and second, by exploring the contribution of shame and guilt to the strength of overall individual morality, which ultimately plays a role in whether crime is perceived for action.

While many theories acknowledge the role of emotions and moral reasoning, SAT serves to be the most useful in aiding understanding of the implications of the interrelationships between empathy, moral emotion, moral reasoning, and its impact upon offending decisions. In summary, the primary theoretical proposition of the current study is that weak moral emotion demonstrates a low intensity to which individuals subscribe to a moral rule (to lead to weak overall morality), and weak morality primarily determines whether crime is perceived as a viable and morally acceptable action alternative; this proposition is supported and evidenced using empirical data. A comparison of the violent offender subsample and the rest of the sample reveal that violent offenders report weaker general empathy, shame, guilt, and moral rules. Regression analyses provide evidence that empathy can predict shame and guilt and guilt and moral rules can predict violence.

The interaction effect between moral rules and moral emotion is one of the key findings of the current study and has rarely been analyzed or presented in other existing criminological research. The interaction finding sheds light on the nuanced processes at play during crime decision-making and emphasizes the importance of the role that moral emotion plays in weakening individual morality and its association to crime involvement. Crucially, it emphasizes this to be specifically important for those with weak moral rules. In conclusion, moral emotion has a stronger effect on violence for young people with weak corresponding weak moral rules; moral emotion alone does not predict crime but rather, it makes an essential contribution to overall morality, which is a key predictor of crime frequency. The results support the key hypotheses of the study and provide a firm context for the importance of further and future work to explore the role of moral emotion in violence.

There are numerous strengths to the current study. SAT, combined with robust empirical data achieves an explanatory line of enquiry (and the identification of the mechanisms at play), as opposed to a less fruitful enquiry of basic associations. Use of self-reported data enables and captures a more valid picture of real-life crime in comparison with incomplete officially recorded police data due to the well-documented existence of a hidden or dark figure in police-recorded statistics (see Wikström, 1985). However, it should be noted that self-report data may not be entirely bullet-proof; given the constructs of interest in the current study, allowance should be made for the possibility that young people who report low levels of guilt and shame may be more willing to self-report crime.²⁰ A further strength to the current study is that PADS+ data are highly reliable (Wikström et al., 2012) and this is particularly important for generalisability of these study findings to the population. Although the questionnaire scales constitute items which refer to hypothetical situations, and therefore measure a general ability to exercise empathy, and a general ability to anticipate shame and guilt; further research indicates that in-depth interview data, which provides more detail and more specificity, confirm the importance of the situational application of empathy, shame, and guilt in specific real-life violent events (see Trivedi-Bateman, 2015).

There are numerous limitations to the current study. First, the study does not take into account the role of wider societal and cultural factors such as cultural practices and norms and religiosity (all of which are likely to play a role in the development of morality and the settings that people partake in). Second, as there are no control variables used in the analyses in the current study, the generalisability of the findings may be limited. Third, since the violent subsample includes high-frequency crime offenders, the specific subsample comparison analyses may be more applicable to serious and persistent offenders as opposed to individuals that commit minor acts of violence.

Another primary limitation involves the issue of establishing causation between moral emotion and violence. Although causation, that is, the role that a change in moral emotion can play in a change in violent behavior, can only be tested using real-life experiments (involving manipulating people’s emotions and environments), this is unethical, impractical, and likely unfeasible. However, the current study offers a suitable alternative, which involves the development of testable implications and assumptions and testing of them, and therefore enables reasonably firm conclusions to be made regarding the processes and mechanisms by which violence occurs.

Future research can accommodate plausible mechanisms that explain a link between morality and crime by incorporating the role of the setting, the importance of which is evidenced in numerous studies that have tested the SAT framework (Pauwels et al., 2018). Although the current study focuses its research efforts on exploration of the explanations and related mechanisms of why crime (and specifically violence) occur, the ultimate hope for crime research is that it will be channeled toward and translated to prevention efforts. This is imperative to investigate the processes by which persistent violent offenders continue to commit a large volume of violent crime, which constitutes a large proportion of total crime. To reduce the possibility of crime being seen as an action alternative, moral emotion development in childhood and adolescence is instrumental.

Footnotes

Declaration of Conflicting Interests

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

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the Economic and Social Research Council (ESRC), with additional funding from Darwin College and the Cambridge Institute of Criminology. I thank my colleagues from The Peterborough Adolescent and Young Adult Development Study (PADS+) who provided insight and expertise that greatly assisted the research, particularly Professor Per-Olof Wikström, Dr. Kyle Treiber, and Dr. Beth Hardie.

Fully funded by the Economic and Social Research Council (ESRC).

ORCID iD

Neema Trivedi-Bateman

Notes

Author Biography

Neema Trivedi-Bateman is lecturer in Criminology at Anglia Ruskin University, Cambridge. Her research focuses on the role of moral emotions in young people’s violence. She is interested in adolescent psychological development and in identifying interventions, particuarly for disadvantaged young people.

References

Agnew

(1985). A revised strain theory of delinquency. Social Forces, 64, 151-167.

Agnew

(2001). Building on the foundation of General Strain Theory: Specifying the types of strain most likely to lead to crime and delinquency. Journal of Research in Crime and Delinquency, 38, 319-361.

Agnew

(2014). Social concern and crime: Moving beyond the assumption of simple self-interest. Criminology, 52, 1-32. doi:10.1111/1745-9125.12031

Anderson

C. A.

Bushman

B. J.

(2002). Human aggression. Annual Review of Psychology, 53, 27-51.

Anderson

(1999). The code of the streets. Atlantic Monthly, 273(5), 81-94.

Anderson

Kiehl

(2012). The psychopath magnetized: Insights from brain imaging. Trends in Cognitive Sciences, 16, 52-60. doi:10.1016/j.tics.2011.11.008

Bandura

(1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191-215.

Barnett

Mann

(2013). Empathy deficits and sexual offending: A model of obstacles to empathy. Aggression and Violent Behavior, 18, 228-239. doi:10.1016/j.avb.2012.11.010

Baron-Cohen

(2011). Zero degrees of empathy: A new theory of human cruelty. London, England: Penguin Books.

10.

Beyers

Loeber

Wikström

P.-O. H.

Stouthamer-Loeber

(2001). What predicts adolescent violence in better-off neighborhoods? Journal of Abnormal Child Psychology, 29, 369-381.

11.

Blair

(2017). Emotion-based learning systems and the development of morality. Cognition, 167, 38-45.

12.

Blair

Mitchell

Blair

(2005). The psychopath: Emotion and the brain. Hoboken, NJ: Blackwell Publishing.

13.

Bock

Hosser

(2014). Empathy as a predictor of recidivism among young adult offenders. Psychology, Crime & Law, 20, 101-115.

14.

Bottoms

(2002). Ideology, crime and criminal justice. London, England: Willan Publishing.

15.

Bouhana

Wikström

P.-O. H.

(2008). Theorizing terrorism: Terrorism as moral action. London, England: UCL Jill Dando Institute of Security and Crime Science.

16.

Bowes

McMurran

(2013). Cognitions supportive of violence and violent behavior. Aggression and Violent Behavior, 18, 660-665. doi:10.1016/j.avb.2013.07.015

17.

Braithwaite

(1989). Crime, shame, and reintegration. Cambridge, UK: Cambridge University Press.

18.

Braithwaite

(2000). Shame and criminal justice. Canadian Journal of Criminology, 42, 281-298.

19.

Braithwaite

(2001). Shame, shame management and regulation. In Ahmed

Harris

Braithwaite

(Eds.), Shame management through reintegration (pp. 3-58). Cambridge, UK: Cambridge University Press.

20.

Brauer

J. R.

Tittle

C. R.

(2017). When crime is not an option: Inspecting the moral filtering of criminal action alternatives. Justice Quarterly, 34, 818-846.

21.

Clay-Warner

(2014). Crime and emotions. In Stets

J. E.

Turner

J. H.

(Eds.), Handbook of the sociology of emotions: Volume II (pp. 473-493). Dordrecht, The Netherlands: Springer.

22.

Cohen

Panter

Turan

Morse

Kim

(2013). Agreement and similarity in self-other perceptions of moral character. Journal of Research in Personality, 47, 816-830. doi:10.1016/j.jrp.2013.08.009

23.

Cohen

L. E.

Felson

(1979). Social change and crime rate trends: A routine activity approach. American Sociological Review, 44, 588-608.

24.

Cornish

Clarke

(1986). The reasoning criminal: Rational choice perspectives on offending. New York, NY: Springer.

25.

Crockett

Clark

Hauser

Robbins

(2010). Serotonin selectively influences moral judgment and behavior through effects on harm aversion. Proceedings of the National Academy of Sciences of the United States of America, 107), 17433-17438. doi:10.1073/pnas.1009396107

26.

Cullen

(2011). Beyond adolescence-limited criminology: Choosing our future-the American society of criminology 2010 Sutherland address. Criminology, 49, 287-330. doi:10.1111/j.1745-9125.2011.00224.x

27.

Damasio

(1994). Descartes error: Emotion, reason, and the human brain. New York, NY: Quill.

28.

Day

Casey

Gerace

(2010). Interventions to improve empathy awareness in sexual and violent offenders: Conceptual, empirical, and clinical issues. Aggression and Violent Behavior, 15, 201-208. doi:10.1016/j.avb.2009.12.003

29.

Dinn

Harris

(2000). Neurocognitive function in antisocial personality disorder. Psychiatry Research, 97, 173-190. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11166089

30.

Dippong

Fitch

(2017). Emotions in criminological theory: Insights from social psychology. Sociology Compass, 11(4), e12473.

31.

Eisenberg

(2000). Emotion, regulation, and moral development. Annual Review of Psychology, 51, 665-697.

32.

Elison

(2005). Shame and guilt: A hundred years of apples and oranges. New Ideas in Psychology, 23, 5-32. doi:10.1016/j.newideapsych.2005.07.001

33.

Elison

Garofalo

Velotti

(2014). Shame and agression: Theoretical considerations. Aggression and Violent Behavior, 19, 447-453.

34.

Farrington

(2005). The importance of child and adolescent psychopathy. Journal of Abnormal Child Psychology, 33, 489-497. doi:10.1007/s10802-005-5729-8

35.

Fine

Kennett

(2004). Mental impairment, moral understanding and criminal responsibility: Psychopathy and the purposes of punishment. International Journal of Law and Psychiatry, 27, 425-443. doi:10.1016/j.ijlp.2004.06.005

36.

Fraser

(1996). Guilt and shame in children: Relationships with empathy and aggression. Halifax, Nova Scotia, Canada: Dalhousie University.

37.

Frick

Morris

(2004). Temperament and developmental pathways to conduct problems. Journal of Clinical Child & Adolescent Psychology: American Psychological Association, Division, 53, 3354-3368. doi:10.1207/S15374424JCCP3301_6

38.

Frick

Ray

Thornton

Kahn

(2013). Annual research review: A developmental psychopathology approach to understanding callous-unemotional traits in children and adolescents with serious conduct problems. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 55, 532-548. doi:10.1111/jcpp.12152

39.

Gottfredson

Hirschi

(1990). A general theory of crime. Palo Alto, CA: Stanford University Press.

40.

Hannon

Knapp

(2003). Reassessing nonlinearity in the urban disadvantage/violent crime relationship: An example of methodological bias from log transformation. Criminology, 41, 1427-1448.

41.

Helion

Ochsner

(2016). The role of emotion regulation in moral judgment. Neuroethics, 11, 297-308. doi:10.1007/s12152-016-9261-z

42.

Hepper

Hart

Meek

Cisek

Sedikides

(2013). Narcissism and empathy in young offenders and non-offenders. European Journal of Personality, 28, 201-210. doi:10.1002/per.1939

43.

Hirtenlehner

Pauwels

Mesko

(2015). Is the criminogenic effect of exposure to peer delinquency dependent on the ability to exercise self-control? Results from three countries. Journal of Criminal Justice, 43, 532-543.

44.

Hoffman

(1982). Development of prosocial motivation: Empathy and guilt. In Eisenberg

(Ed.), The development of prosocial behaviour. New York, NY: Academic Press.

45.

Hoffman

(2000). Empathy and moral development: Implications for caring and justice. New York, NY: Cambridge University Press. doi:10.1017/CBO9780511805851

46.

Jolliffe

Farrington

(2004). Empathy and offending: A systematic review and meta-analysis. Aggression and Violent Behavior, 9, 441-476. doi:10.1016/j.avb.2003.03.001

47.

Jolliffe

Farrington

(2006). Examining the relationship between low empathy and bullying. Aggressive Behavior, 32, 540-550.

48.

Jolliffe

Murray

(2012). Low empathy and offending. In Loeber

Welsh

B. C.

(Eds.), The future of criminology. Oxford, UK: Oxford University Press.

49.

Krettenauer

Asendorpf

Nunner-Winkler

(2013). Moral emotion attributions and personality traits as long-term predictors of antisocial conduct in early adulthood: Findings from a 20-year longitudinal study. International Journal of Behavioral Development, 37, 192-201. doi:10.1177/0165025412472409

50.

Krettenauer

Campbell

Hertz

(2013). Moral emotions and the development of the moral self in childhood. European Journal of Developmental Psychology, 10, 159-173. doi:10.1080/17405629.2012.762750

51.

Lapsley

Carlo

(2014). Moral development at the crossroads: New trends and possible futures. Developmental Psychology, 50, 1-7. doi:10.1037/a0035225

52.

Lewis

(1971). Shame and guilt in neurosis. Madison, CT: International University Press.

53.

Llorca-Mestre

Malonda-Vidal

Samper-Garcia

(2017). Prosocial reasoning and emotions in young offenders and non-offenders. The European Journal of Psychology Applied to Legal Context, 9, 65-73.

54.

Loeber

Farrington

Stouthamer-Loeber

Moffitt

Caspi

(1998). The development of male offending: Key findings from the first decade of the Pittsburgh Youth Study. Studies on Crime and Crime Prevention, 7, 141-171.

55.

Malti

Gasser

Buchmann

(2009). Aggressive and prosocial children’s emotion attributions and moral reasoning. Aggressive Behavior, 35, 90-102. doi:10.1002/ab.20289

56.

Malti

Krettenauer

(2013). The relation of moral emotion attributions to prosocial and antisocial behavior: A meta-analysis. Child Development, 84, 397-412. doi:10.1111/j.1467-8624.2012.01851.x

57.

Marshall

Serran

O’Brien

(2009). Self-esteem, shame, cognitive distortions and empathy in sexual offenders: Their integration and treatment implications. Psychology, Crime & Law, 15, 217-234. doi:10.1080/10683160802190947

58.

McClelland

G. H.

Judd

C. M.

(1993). Statistical difficulties of detecting interactions and moderator effects. Psychological Bulletin, 114, 376-390.

59.

Miller

Eisenberg

(1988). The relation of empathy to aggressive and externalizing/antisocial behavior. Psychological Bulletin, 103, 324-344. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/3289071

60.

Office for National Statistics. (2011). 2001 Census aggregate data (Edition: May 2011). Colchester, UK: UK Data Service. doi:10.5257/census/aggregate-2001-2

61.

Office of National Statistics. (2018). Crime in Peterborough compared with crime in other similar areas. https://www.police.uk/cambridgeshire/Peterborough/performance/compare-your-area/

62.

Olthof

(2012). Anticipated feelings of guilt and shame as predictors of early adolescents’ antisocial and prosocial interpersonal behaviour. European Journal of Developmental Psychology, 9, 371-388.

63.

Osgood

Finken

McMorris

(2002). Analyzing multiple-item measures of crime and deviance II: Tobit regression analysis of transformed scores. Journal of Quantitative Criminology, 18, 319-347.

64.

Pauwels

Svensson

Hirtenlenher

(2018). Testing situational action theory: A narrative review of studies published between 2006 and 2015. European Journal of Criminology, 15, 32-55.

65.

Pornari

C. D.

Wood

(2010). Peer and cyber aggression in secondary school students: The role of moral disengagement, hostile attribution bias, and outcome expectancies. Aggressive Behavior, 36, 81-94. doi:10.1002/ab.20336

66.

Posick

Rocque

Rafter

(2012). More than a feeling: Integrating empathy into the study of lawmaking, lawbreaking, and reactions to lawbreaking. International Journal of Offender Therapy and Comparative Criminology, 58, 5-26. doi:10.1177/0306624X12465411

67.

Roos

Salmivalli

Hodges

(2011). Person × context effects on anticipated moral emotions following aggression. Social Development, 20, 685-702. doi:10.1111/j.1467-9507.2011.00603.x

68.

Schalkwijk

Stams

G. J.

Stegge

Dekker

Peen

(2014). The conscience as a regulatory function: Empathy, shame, pride, guilt, and moral orientation in delinquent adolescents. International Journal of Offender Therapy and Comparative Criminology,60, 675-693. doi:10.1177/0306624X14561830

69.

Seiter

Bruschke

(2007). Deception and emotion: The effects of motivation, relationship type, and sex on expected feelings of guilt and shame following acts of deception in United States and Chinese samples. Communication Studies, 58, 1-16. doi:10.1080/10510970601168624

70.

Sherman

(1993). Defiance, deterrence, and irrelevance: A theory of the criminal sanction. Journal of Research in Crime and Delinquency, 30, 445-473. doi:10.1177/0022427893030004006

71.

Sobhani

Bechara

(2011). A somatic marker perspective of immoral and corrupt behavior. Social Neuroscience, 6, 640-652.

72.

Spinella

(2005). Prefrontal substrates of empathy: Psychometric evidence in a community sample. Biological Psychology, 70, 175-181. doi:10.1016/j.biopsycho.2004.01.005

73.

Spruit

Schalkwijk

van Vugt

Stams

G.-J.

(2016). The relation between self-conscious emotions and delinquency : A meta-analysis. Aggression and Violent Behavior,28, 12-20. doi:10.1016/j.avb.2016.03.009

74.

Svensson

(2004). Social control and socialisation: The role of morality as a social mechanism in adolescent deviant behaviour. Stockholm, Sweden: Stockholm University.

75.

Svensson

Oberwittler

(2010). It’s not the time they spend, it’s what they do: The interaction between delinquent friends and unstructured routine activity on delinquency: Findings from two countries. Journal of Criminal Justice, 38, 1006-1014.

76.

Svenson

Pauwels

Weerman

Bruinsma

(2016). Explaining individual changes in moral values and moral emotions among adolescent boys and girls: A fixed-effects analysis. European Journal of Criminology, 14, 290-308.

77.

Svensson

Weerman

F. M.

Pauwels

L. J.

Bruinsma

G. J.

Bernasco

(2013). Moral emotions and offending: Do feelings of anticipated shame and guilt mediate the effect of socialization on offending? European Journal of Criminology, 10, 22-39. doi:10.1177/147737081245439

78.

Tangney

Fischer

(1995). Self-conscious emotions: The psychology of shame, guilt, embarrassment, and pride. New York, NY: Guilford Press.

79.

Tangney

Stuewig

Hastings

(2011). Assessing jail inmates’ proneness to shame and guilt: Feeling bad about the behavior or the self? Criminal Justice Behavior, 38, 710-734. doi:10.1177/0093854811405762

80.

Tangney

Stuewig

Mashek

(2007). Moral emotions and moral behavior. Annual Review of Psychology, 58, 345-372. doi:10.1146/annurev.psych.56.091103.070145

81.

Tibbetts

(2003). Self-conscious emotions and criminal offending. Psychological Reports, 93, 101-126.

82.

Tittle

(2007). The explanation of crime: Context, mechanisms and development. Contemporary Sociology: A Journal of Reviews, 36, 484-485.

83.

Treiber

(2011). The neuroscientific basis of situational action theory. In Walsh

Beaver

(Eds.), The Ashgate research companion to biosocial theories of crime (pp. 213-248). Farnham, UK: Ashgate Publishing.

84.

Treiber

(2013). A neuropsychological test of criminal decision making : Regional prefrontal influences in a dual process model. In van Gelder

Elfers

Nagin

Reynald

(Eds.), Affect and cognition in criminal decision making (pp. 193-220). Abingdon, UK: Routledge.

85.

Trivedi-Bateman

(2015). The roles of empathy, shame, and guilt in violence decision-making (Doctoral dissertation). University of Cambridge. doi:10.17863/CAM.16553

86.

Vachon

Lynam

(2016). Fixing the problem with empathy: Development and validation of the affective and cognitive measure of empathy. Assessment, 23, 135-149.

87.

Vachon

Lynam

Johnson

(2014). The (non)relation between empathy and aggression: Surprising results from a meta-analysis. Psychological Bulletin, 140, 751-773. doi:10.1037/a0035236

88.

van Langen

Wissink

van Vugt

Van der Stouwe

Stams

. (2014). The relation between empathy and offending: A meta-analysis. Aggression and Violent Behavior, 19, 179-189. doi:10.1016/j.avb.2014.02.003

89.

Wikström

P.-O. H.

(2005). The social origins of pathways in crime: Towards a developmental ecological action theory of crime involvement and its changes. In Farrington

(Ed.), Integrated developmental and lifecourse theories of offending (pp. 211-245). New Brunswick, NJ: Transaction.

90.

Wikström

P.-O. H.

(2006). Individuals, settings, and acts of crime: Situational mechanisms and the explanation of crime. In Wikström

P.-O. H.

Sampson

(Eds.), The explanation of crime: Context, mechanisms, and development (pp. 61-107). Cambridge, UK: Cambridge University Press.

91.

Wikström

P.-O. H.

(2009). Crime propensity, criminogenic exposure and crime involvement in early to mid adolescence. Monatsschrift Fur Krimonologie Und Strafrechtsreform, 92, 253-266.

92.

Wikström

P.-O. H.

(2010). Explaining crime as moral actions. In Hitlin

Vaysey

(Eds.), Handbook of the sociology of morality (pp. 211-240). New York, NY: Springer.

93.

Wikström

P.-O. H.

(2014). Why crime happens: A situational action theory. In Manzo

(Ed.), Analytical sociology: Actions and networks (pp. 74-94). Hoboken, NJ: John Wiley.

94.

Wikström

P.-O. H.

Oberwittler

Treiber

Hardie

(2012). Breaking rules: The social and situational dynamics of young people’s urban crime. Oxford, UK: Oxford University Press.

95.

Wikström

P.-O. H.

Suède . Brottsförebyggande rådet. (1985). Everyday violence in contemporary Sweden: Situational and ecological aspects. Stockholm, Sweden: National Council for Crime Prevention, Sweden, Research Division.

96.

Wikström

P.-O. H.

Svensson

(2008). Why are English youths more violent than swedish youths? A comparative study of the role of crime propensity, lifestyles and their interactions in two cities. European Journal of Criminology, 5, 309-330. doi:10.1177/1477370808090835

97.

Wikström

P.-O. H.

Svensson

(2010). When does self-control matter? The interaction between morality and self-control in crime causation. European Journal of Criminology, 7, 395-410. doi:10.1177/1477370810372132

98.

Wikström

P.-O. H.

Treiber

(2007). The role of self-control in crime causation: Beyond gottfredson and hirschi’s general theory of crime. European Journal of Criminology, 4, 237-264. doi:10.1177/1477370807074858

99.

Wikström

P.-O. H.

Treiber

(2009a). Violence as situational action. International Journal of Conflict and Violence, 3, 75-96.

100.

Wikström

P.-O. H.

Treiber

(2009b). What drives persistent offending? The neglected and unexplored role of the social environment. In Savage

(Ed.), The development of persistent criminality. Oxford, UK: Oxford University Press.

101.

Wikström

P.-O. H.

Treiber

Hardie

(2012). Examining the role of the environment in crime causation: Small area community surveys and space-time budgets. In Gadd

Karstedt

Messner

S. F.

(Eds.), The SAGE handbook of criminological research methods (pp. 1-42). London, UK: Sage.

102.

Wikström

P.-O. H.

Tseloni

Karlis

(2011). Do people comply with the law because they fear getting caught? European Journal of Criminology, 8, 401-420. doi:10.1177/1477370811416415