The relation between self-conscious emotions and delinquency: A meta-analysis

Highlights

• Feelings of guilt and shame were associated with less delinquency.

• Stronger associations were found for guilt than for shame.

• Guilt was stronger associated with self-reported than official delinquency.

• Self-conscious emotions are as important as empathy and moral judgment.

• The relation between shame and delinquency may depend on culture.

Abstract

Self-conscious emotions are expected to be related to delinquency, as they guide moral decision making. In the current study, two separate multilevel meta-analyses were performed to examine the overall relation between guilt, shame and delinquency. In addition, possible moderating factors were examined. In total, 25 studies with 24 independent samples, reporting on 75 effect sizes, were included. The results showed significant negative associations between guilt and delinquency (r = −.278), and between shame and delinquency (r = −.130), indicating that higher levels of guilt and shame were related to less delinquency. Implications for theory and practice concerning the role of self-conscious emotions in delinquency and offender treatment are discussed.

Introduction

When it comes to the understanding of delinquent behaviors, cognitive elements of moral development, such as moral judgment, have been studied extensively (Stams et al., 2006). On the last few decades, much attention has been given to the role of emotions in immoral behavior. A range of moral emotions are considered relevant in this regard, with empathy, guilt, and shame among the most studied. Although all distinctive in nature, they are highly associated (Hoffman, 1998, Trivedi-Bateman, 2015, Tangney, 1991). More precisely, Tangney (1991) assumed that true interpersonal guilt relies both on the ability to identify, share and respond to someone's distress, i.e., being empathic (Cohen & Strayer, 1996), and on the understanding of one's own role that has led to the distress. This self-reflective and evaluative state in combination with an understanding of moral rules and standards distinguishes guilt and shame from empathy. Therefore, guilt and shame are generally referred to as self-conscious emotions (Lewis, 2000).

Self-conscious emotions are expected to be related to a range of moral behaviors, including delinquency. There is general agreement that self-conscious emotions guide moral decision making, and therefore influence moral behavior (Eisenberg, 2000, Pizarro, 2000). People continuously evaluate their thoughts and actions from their personal moral reference of values and standards (Lewis, 1991, Schalkwijk, 2015). Negative self-conscious emotions, such as guilt and shame, are evoked when the evaluations of actions or thoughts are in conflict with the person's moral values and standards. Since negative self-evaluations are hurtful, people avoid behaviors that evoke negative self-conscious emotions (Schalkwijk, 2015, Tangney and Dearing, 2002a, Tangney et al., 2007). On the other hand, behaviors can be immediately reinforced if followed by positive self-conscious emotions, such as pride (Tracy and Robins, 2007, Eisenberg, 2000). As a result of this regulating effect on moral behavior, self-conscious emotions often affect antisocial behaviors, including delinquency (Eisenberg, 2000, ⁎Gold et al., 2011, Jackson, 2009, ⁎Murphy and Harris, 2007, Ribeiro da Silva et al., 2015, Schalkwijk, 2015, Stuewig and Tangney, 2007, Tangney et al., 2011, ⁎Tibbetts, 2003).

To date, no systematic reviews or meta-analyses describing the relation between self-conscious emotions and delinquency are available. This lack is remarkable as more and more research is focused on the identification of criminogenic risk factors for delinquency and recidivism, to provide leads for treatment and offender therapy (Andrews & Bonta, 2010). Self-conscious emotions are of particular interest in this regard, as they are assumed to be critical in the rehabilitation of offenders (Tangney et al., 2011) by helping the offender to take responsibility for his acts and to repair the harm done to victims or society (Braithwaite, 1989). Therefore, the aim of the current meta-analysis was to systematically review the literature, assess the strength of the relation between self-conscious emotions and delinquency, and to examine factors that could moderate this relation. Self-conscious emotions include various emotions, such as shame, guilt, remorse, regret, pride, embarrassment and humiliation (Tracy, Robins, & Tangney, 2007). Since research into the relation between self-conscious emotions other than guilt and shame is sparse, only guilt and shame will be addressed in the present study. Notably, remorse and regret should be considered as central to guilt, because the experience of guilt is intrinsically connected to the wish to have behaved differently (Tangney et al., 2011), whereas embarrassment should be considered a distinct emotion if compared to shame and guilt (Keltner & Buswell, 1996).

Although guilt and shame are both negative self-conscious emotions, they are not equally linked to delinquency throughout literature (Eisenberg, 2000, Stuewig and Tangney, 2007, Tangney et al., 2011). The key difference according to Lewis (1971) is that guilt relates to the evaluation of behavior, whereas shame relates to the evaluation of one's whole being. Guilt is an emotion that mostly emerges within an interpersonal context, caused by an action that inflicts harm, loss, distress or pain on (significant) others (Baumeister, Stillwell, & Heatherton, 1994). In these situations there is often a possibility to repair the damage by helping the other person or expressing feelings of guilt and remorse. Social bonds between the offender and the victim can become stronger after these reparative actions and help prevent a negative self-evaluation (Baumeister et al., 1994, Stuewig and Tangney, 2007). Experiencing feelings of guilt encourages people to internalize the blame of the inflicted harm on others and to take responsibility of their actions, resulting in restorative behaviors. Since guilt-proneness goes together with more internalized blaming and higher levels of empathic functioning (Mandel and Dhami, 2005, Stuewig et al., 2010), it is expected that higher levels of guilt feelings are associated with less delinquency (Parrott and Strongman, 1984, Van Langen et al., 2014).

The relation between shame and delinquency is equivocal (Stuewig & Tangney, 2007). Since the feeling of shame is a negative self-evaluation of one's whole being, shame can be an extremely painful emotion (Elison et al., 2014, Tangney et al., 2011). Therefore, the anticipation of shame-feelings has a strong inhibiting effect on delinquent behaviors; predicting that a certain behavior will lead to shame feelings should cause people to refrain from that behavior (Schalkwijk, 2015, Tangney and Dearing, 2002a, Tangney et al., 2007). On the contrary, Lewis (1971) proposed that shame may lead to such strong feelings of worthlessness and powerlessness that the self-concept needs to be protected from those negative self-evaluations. The defense mechanism resulting from this need is to externalize the blame of the actions and behaviors (Schalkwijk, 2015, Stuewig et al., 2010, Tracy and Robins, 2003). The anger that comes with the externalizing blame has often been related to aggressive behaviors (Harper et al., 2005, Tangney et al., 2011). The pathway from shame to externalizing the blame, anger, and antisocial behavior has mainly been described for aggression and not for delinquency. However, it can be argued that since an external locus of control is associated with delinquency (Parrott & Strongman, 1984), this path may also hold for delinquent behaviors. Altogether, it is expected that the relation between delinquency and guilt or shame differs. A protective role of guilt regarding delinquency is hypothesized. For shame, the relation with delinquency is less clear. Therefore, two meta-analyses will be conducted, assessing the relation between guilt and delinquency, and shame and delinquency.

The strength of the relation between self-conscious emotions and delinquency may be influenced by other factors, such as characteristics related to self-conscious emotions, delinquency, studies, and samples. Considering factors related to self-conscious emotions, the measurement of guilt and shame may possibly moderate the relation with delinquency (Else-Quest et al., 2012, Kim et al., 2011). Some instruments measure self-conscious emotions in specific contexts (i.e., contextual guilt and shame). For example, the Test of Self-Conscious Affect (TOSCA-3; Tangney, Dearing, Wagner, & Gramzow, 2000) presents specific scenario's in which a person has committed an immoral act, upon which the respondent indicates the likelihood of reactions that represent the experience of guilt and shame. Other instruments (for example, the Personal Feelings Questionnaire-2; Harder & Zalma, 1990) measure self-conscious emotions independent of context (i.e., generalized self-conscious emotions), for example, by asking respondents about the frequency of guilt and shame experiences. Further, measures of guilt and shame are often correlated (⁎Stuewig et al., 2015, Tangney, 1996), making it difficult to assess the unique influence of guilt and shame on delinquency. However, as we expect that guilt is stronger associated with delinquency than shame, “shame-free” guilt may show stronger associations with delinquency than measures of guilt that include shame. Thus, whether a study controls for the covariance between guilt and shame could moderate the relation between self-conscious emotions and delinquency.

Considering factors related to delinquency, one of the possible moderators is the type of delinquency (Stuewig and Tangney, 2007, Stuewig et al., 2010, Tangney et al., 2011). We would expect a moderating effect of general versus violent delinquency, because there are some differences between delinquency and aggression in the mechanisms underlying the relation with self-conscious emotions, especially for shame (Stuewig and Tangney, 2007, Tangney et al., 2011). The associations between experiencing shame, anger, and aggressive behavior (Stuewig et al., 2010, Tangney et al., 1996) could implicate that shame is stronger related to violent delinquency than to general delinquency. Considering sample characteristics, gender may be an important moderator too. In adolescence and adulthood, small gender differences have been found in guilt and shame (Else-Quest et al., 2012). Up until puberty, research has found little differences between the extent to which boys and girls experience self-conscious emotions. In puberty, however, differences become more visible. Young people of both sexes become somewhat less prone to experience self-conscious emotions, with males becoming less so than females (Bybee, 1998, Else-Quest et al., 2012). Also, Tibbetts (2003) and Schalkwijk, Stams, Stegge, Dekker, and Peen (2014) found different relations between self-conscious emotions and delinquency for males and females. Age is an important sample characteristic, because Shulman, Cauffman, Piquero, and Fagan (2011) showed that moral disengagement in convicted juveniles tends to decline over time, and with it, offending.

The main question that will be addressed in the current meta-analysis is how strongly guilt and shame are related to delinquency. Further, the possible moderating effects of characteristics of self-conscious emotions, delinquency, studies, and samples will be examined.