Addressing the challenges of remorse in the criminal justice system

Abstract

Remorse in offenders appears in a number of criminal justice domains. It is a mitigating factor in sentencing, influences parole decision making, may be influential in offender rehabilitation, and may be valued in forensic mental health. However, evaluation of remorse presents challenges in relation to evidence for remorse and expectations about its performance. Nevertheless, remorse is embedded deeply in criminal justice. The consideration of remorse in relation to emotions of shame, guilt and regret may offer an approach for evaluating remorse in sentencing and for addressing remorse in offender rehabilitation. This approach to understanding and working with remorse requires further elaboration, development and testing.

1. Introduction

Remorse is discussed with reasonable frequency in the courts, in other criminal justice settings and in the media, yet remorse in criminal justice has received comparatively little scholarly attention. However, remorse is receiving increased scholarly interest in a range of disciplines (Tudor et al., 2022). In this article, I explore the breadth of criminal justice settings in which remorse is discussed, as well as problems and challenges presented by remorse. I then address reasons that discussion of remorse is unlikely to disappear from criminal justice. Acknowledging that remorse is likely to persist, I examine remorse through the lens of emotion, and I apply this analysis to addressing remorse in the domains of sentencing and offender rehabilitation.

2. What is remorse?

Although there is no clearly agreed definition of remorse, I offer three different attempts at a description of remorse, which includes the Latin root ‘mordere’ (to bite), with the prefix ‘re’ connoting repetition (Gorelick, 1989), and denotes an experience of biting repeatedly on the past (Studzinski, 1989). The first description is from the chapter on remorse in Gaita’s (2004) Good and Evil: An Absolute Conception, in which remorse is a kind of significant moral insight. Gaita described remorse as ‘a pained, bewildered realisation of what it means to wrong someone’ (p. 52). Remorse, according to Gaita, may be an intense, significant and personal experience. Proeve and Tudor (2010) took a second approach, of setting out a working list of cognitive, emotional and behavioural characteristics that a remorseful person might show. According to this list, a remorseful person might: recognise that she has wronged and harmed another person; accept that she was responsible for her action, which was voluntary; experience a sense that life has changed as a consequence of the action; experience internal turmoil; desire to make reparation, for example by expressing remorse, apologising, making restitution and/or behaving differently; desire to be forgiven; show evidence of having acted on the desire to make reparation. In criminal justice settings it is frequently the task of others to discern remorse in a person who has committed a crime. Therefore, both Gaita’s and Proeve and Tudor’s descriptions of remorse present the problem of how to discern the internal experiences of others. In contrast, Weisman (2009) set out the characteristics of a successful performance of remorse in the courtroom. Based on a large number of Canadian cases in which remorse was a concern, Weisman concluded that a person found to be remorseful should show evidence of three characteristics: admitting responsibility for their offences; demonstrating visible suffering in their demeanour, which is empathic and not self-focused suffering; and showing evidence of personal transformation. Although Weisman’s approach may seem to address the problem of judging the internal remorseful experiences of others, it presents other problems, in that the judge has the task of deciding whether suffering displayed by the defendant is empathic or merely self-centred, or whether the defendant has truly admitted responsibility.

There is a degree of overlap in these three approaches to describing remorse, in that remorse seems to have a focus on others, concerns responsibility for one’s actions, and involves personal change. In the next section, I discuss the role played by remorse in various criminal justice settings.

3. Domains of remorse

3.1. Remorse in courts and tribunals

Remorse is widely recognised as a mitigating factor in sentencing, in U.S. federal sentencing guidelines and in numerous decided federal and state cases, in court practice directions and case law in England and Wales, in case law in Canada, in federal and state statutes and case law in Australia, and in New Zealand’s sentencing legislation (Proeve & Tudor, 2010). In some jurisdictions, remorse or its absence may have considerable consequences. For example, remorse is an important influence in American death penalty decisions, where the lack of remorse is cited as a factor in jury decisions to impose the death penalty (Costanzo & Costanzo, 1992; Sundby, 1998).

Remorse has also been cited during proceedings of international tribunals. For example, remorse has featured in case law of the International Criminal Tribunal for the Former Yugoslavia (ICTY; Tieger, 2003). Holá and van Wijk (2022) report that although remorse was not listed in the ICTY statute or rules of procedure or evidence, acknowledgement of responsibility and remorse were factors considered to be relevant by judges in their determinations of punishment. Remorse has also featured in the South African Truth and Reconciliation Commission. Although expressions of remorse were neither expected nor required at the amnesty hearings, the emotional response of perpetrators was an important focus of the proceedings when they were broadcast or reported in the media (Weisman, 2014).

3.2. Remorse in parole hearings

Remorse may play a role in parole decisions in various jurisdictions. According to Bronnimann (2020), remorse informs parole decisions in many states of the USA. With the exception of California and Pennsylvania, remorse is not explicitly a factor to be considered in parole decisions. However, other states’ regulations and statutes contain criteria that can be construed to entail remorse, such as a requirement to pay respect to victims.

In the UK, Tidmarsh (1999) found that remorse did not figure directly in practice directions. However, he noted that, in practice, parole board members were reluctant to release someone who was not remorseful. Remorse was also considered relevant by judges in the International Criminal Tribunal for the Former Yugoslavia when assessing the level of rehabilitation of a prisoner for the purposes of early release decisions (Holá & van Wijk, 2022).

3.3. Remorse in offender rehabilitation

Remorse seemingly has little ‘official’ place in the time between sentencing and applying for parole, when offenders are imprisoned and may participate in offender rehabilitation programmes. The dominant rehabilitation approach in Australia, the UK and North America is the Risk–Need–Responsivity (RNR) model (Andrews et al., 1990; Polaschek, 2012). The RNR model is concerned with: the risk principle, that higher levels of service are reserved for higher risk offenders; the needs principle, that rehabilitation interventions are directed towards criminogenic needs, which are dynamic offender characteristics that are associated with changes in risk of recidivism when changed; and with the responsivity principle, that styles and modes of service should engage offenders to help them learn and change.

Within the model’s principles, remorse could be considered a specific responsivity issue, in that a remorseful offender might potentially be especially motivated to engage in rehabilitation interventions. However, remorse could conceivably be relevant to risk of recidivism, in that lack of remorse is an influential item in determining level of psychopathy, and psychopathy is associated with recidivism and violent recidivism (Cooke & Michie, 1997). However, Hall and Rossmanith (2022) argue that remorse may be subtly present in the expectations and conduct of offender rehabilitation programmes. They point to common concepts that appear during rehabilitation programmes, such as ‘victim blaming’, ‘externalisation’, ‘lack of victim empathy’, and ‘denial and minimisation’. Participants may also compose letters to victims of their crimes, expressing their understanding of the effects of their actions on victims and offering apologies. Considering these prevalent concepts and practices, Hall and Rossmanith suggest that remorse is hiding in plain sight in offender rehabilitation. For example, ‘empathy’ (including ‘victim empathy’) can be understood as appreciating the reality of other people—which is central to feeling remorse. Being aware of the ‘consequences’ of one’s behaviour sounds much like ‘taking responsibility’, another component of experiencing remorse. ‘Denial and minimisation’—denying wrongdoing or minimising one’s own wrongdoing—is akin to not taking responsibility, which may be evidence of remorselessness. Hall and Rossmanith (2022) suggest that, in contrast to sentencing, where remorse might be judged at a single point in time, ongoing remorseful behaviour is evaluated over an extended period, and is shown by participating in official programmes and by good behaviour.

3.4. Remorse in forensic mental health

Remorse may also be present where it is not clearly expected, as O’Donahoo and Simmonds (2022) describe in their description of discussions about remorse in forensic mental health patients. People who have been found not guilty by reason of mental impairment are not expected to assume responsibility for their actions, and responsibility is a key feature of remorse. The majority of participants in this interview study had a diagnosis of schizophrenia or a related psychosis, and most had committed violent offences. Most interviewees did not believe that there would be adverse consequences of not showing remorse. However, most stated that showing remorse was a reasonable expectation of forensic patients. In addition, all but one participant commented on the therapeutic benefits of addressing their remorse, including self-forgiveness, self-acceptance, ‘healing’ and ‘moving on’. Almost universally, participants thought that dealing with remorseful feelings was a necessary forerunner to, but did not guarantee, improved psychological wellbeing and self-forgiveness.

From this consideration of remorse in various criminal justice settings, scholars suggest that remorse may be expected or overtly evaluated in sentencing; it may be overtly or less overtly evaluated in parole proceedings; it may be expected but not overtly discussed in offender rehabilitation; and it may not be expected but nevertheless may be valued in forensic mental health settings. In the next section, I discuss problems presented by the task of discerning remorse.

4. Problems in discerning remorse

At the stage of sentencing, judges (or jurors, in capital cases) may take into account an offender’s behaviour immediately after committing an offence, or his actions prior to trial, such as making restitution or changing behaviour, or behaviour in court, such as the demeanour he presents. Scholars have raised three prominent problems raised in the discernment of remorse: demeanour as evidence of remorse, guilty pleas as evidence of remorse, and the successful performance of remorse.

4.1. Demeanour and remorse

Bandes (2022) argues that demeanour plays an outsized role in capital cases in the USA, in which juries determine whether capital punishment is to be applied. Bandes notes that defendants testify rarely, so that remorse must be discerned by juries from facial and body ‘language’. Judges also seem to rely on demeanour, despite the fact that they have the opportunity to talk to the defendant. Judges take note of behaviours such as crying, eye contact, head posture, fidgeting or disrespectful facial expressions (Bandes, 2022), despite the lack of empirical support for the idea that remorse can be reliably or validly evaluated from demeanour (Bandes, 2016).

4.2. Guilty pleas and remorse

As I noted, remorse is a mitigating factor in sentencing. A guilty plea may afford the defendant a sentencing discount independent of remorse, as it spares the state the cost of a trial and spares victims from a trial. To complicate matters, a guilty plea may be recognised as evidence of remorse (Cameron v The Queen, 2002; R v Thomson, 2000), but a guilty plea may also be seen as motivated by a willingness to facilitate justice (Cameron v The Queen, 2002). Therefore, a guilty plea may or may not indicate remorse, and this begs the question of how to tell the difference between remorseful guilty pleas and those motivated by other concerns. Judges may solve this question by regarding a guilty plea at an early stage of legal proceedings as evidence of remorse (Rossmanith, 2015).

With these concerns regarding guilty pleas as evidence of remorse in mind, there are indications that judgments of remorse may rely unduly on guilty pleas. Wood and MacMartin (2007) found that judges in sexual assault cases viewed a timely guilty plea as the most important mitigating factor, and a not guilty plea as evidence of lack of remorse. In the opinions they studied, judges generally included language affirming that an offender could not be more severely punished for exercising his right to trial, but also language framing a guilty plea as a public confession of wrongdoing and therefore a praiseworthy indicator of mitigation-worthy remorse. Proeve and Tanvir (2022) analysed 262 sentencing transcripts from the higher courts in South Australia over a period of 12 months, in which the judge mentioned remorse or contrition (which was 26% of cases in the higher courts). They recorded the frequency with which the judge noted particular kinds of evidence for remorse. Although judges referred to changing behaviour, seeking help from professionals in relation to offending, and making an apology, the guilty plea was the most frequent meaningful indication of remorse, present in 35% of transcripts. However, guilty pleas were the only evidence of remorse noted in only 6%. There were also cases in which remorse was nevertheless found to be absent, despite a guilty plea. Moreover, although judges in some cases explicitly affirmed the defendant’s right to plead not guilty and to go to trial by jury, judges referred to the lack of a guilty plea in 43% of transcripts where judges found the defendant to be not remorseful, or they precluded remorse from consideration as a mitigating factor. A plea of not guilty was the sole piece of evidence noted in 42% of transcripts in which remorse was found absent or was not considered. Therefore, the guilty plea appeared to be much more influential as an indicator of lack of remorse than as an indicator that remorse was present.

4.3. Performance of remorse

Rossmanith (2015) discussed her findings from interviewing judges about remorse using the lens of performance. Although judges did not refer to witnesses’ evidence as ‘performance’, several invoked theatrical metaphors, and the interviews showed that judges were sometimes strongly moved by their encounters with offenders. Other scholars of remorse (Bandes, 2022; Weisman, 2014) note that judges seem to expect a certain type of performance of remorse, and the components of that performance are often those indicators of remorse (responsibility, suffering, transformation) described by Weisman (2009). Van Oorschot and colleagues (2017) found that remorse in Dutch courts was demonstrated by taking responsibility, making personal change and accepting their burden without complaint. However, they found that performances of remorse that apparently met those criteria could fail to mitigate a sentence, and mitigation might nevertheless occur in the absence of performances of remorse. Remorse has to be performed, and that performance has to be evaluated. However, judges may apply shifting criteria to the evaluation of remorse (Weisman, 2009), and because remorse is performed, expressions of remorse may be discounted as self-serving or coached (Bandes, 2022).

In offender rehabilitation, the performance requirement of remorse also presents difficulties (Hall & Rossmanith, 2022). Remorse might be demonstrated by making personal change, such as attendance at a drug or anger management programme. However, access to therapeutic programmes may be limited according to the prison or the classification of the prisoner. Performing remorse consistently in prison settings is difficult across various settings of prison life. For example, prisoners spend time in the prison yard, where extreme enactments of masculinity may be required, but may, in the very different context of the psychologist’s office, be expected to demonstrate sensitivity and self-insight.

5. Why does remorse persist in criminal justice?

As I have discussed, remorse appears in a number of places in criminal justice, but the task of discerning remorse presents significant problems. Remorse persists in criminal justice nevertheless, and in this section, I propose three reasons for the persistence of remorse.

5.1. Remorse and moral communities

Weisman (2014) states that moral communities exert pressure on their members to show remorse, and to not show remorse. In short, being remorseful when we transgress the rules of our moral communities is a way in which we demonstrate membership of a moral community. Those who are remorseful can be part of our moral community, whereas those who do not may be rejected from the moral community.

On the other hand, in some circumstances remorse might go against membership of a moral community. For example, Holá and van Wijk (2022) discussed the celebrated case of the former president of Republika Srpska, the only prominent Bosnian Serb official to plead guilty at the International Criminal Tribunal for the Former Yugoslavia. This individual revealed in interviews after her release from prison that she admitted guilt for instrumental reasons. After her return home, she retracted her most significant admissions, and reiterated a nationalist worldview that showed little, if any, political change. Holá and van Wijk suggest that, in a deeply divided post-conflict setting, it may be beneficial not to show remorse.

5.2. Remorse is difficult to banish from the criminal justice system

A second reason for the persistence of remorse is that it is practically difficult to banish from criminal justice, because remorse is connected to other valued qualities such as empathy, conscience and atonement, and it is a sign of the possibility of change (Bandes, 2022). Dillon (2022) suggests that remorse is woven into any discussion of sentencing and the purposes and effectiveness of sentencing. Because we live in a culture that hopes for change in offenders, we will inevitably search for signs of change. The question, then, as Bandes (2022) suggests, is how to regulate the evaluation of remorse.

5.3. Remorse is part of human relationships

Judeo-Christian morality, as a foundation for Western legal systems, is an important cultural source of the enduring presence of remorse. However, another source may be the encounters with one another’s human reality that can occur in criminal justice. Gaita (2004) contends that remorse is a recognition of the reality of others whom we have wronged. We might say that, during sentencing, an offender judged to be remorseful apparently shows that recognition of the reality of others affected by their actions, so that judges can speak of remorse that ‘“cries out” . . . and that is felt by the judge across the room’ (Rossmanith, 2015, p. 182). This recognition of others’ reality may occur at other times, during parole hearings, meetings with probation officers and even during the entire sentence, served in incarceration or in the community.

To the extent that criminal justice is concerned with human relationships with one another, and not only with the control of behaviour, criminal justice systems will be concerned with remorse (Rossmanith et al., 2022). Because remorse is intertwined with individual relationships, with broader relationships in moral communities and with values that are prized in those particular moral communities of criminal justice, remorse is likely to persist in criminal justice. If remorse will persist in criminal justice, as Bandes (2022) suggests, the question is how to regulate the evaluation of remorse.

In the following sections, I explore an approach to evaluating remorse in relation to two aspects of criminal justice—sentencing and offender rehabilitation. Two aspects of Weisman’s (2009) successful remorse performance in relation to sentencing are empathic suffering and personal transformation. Secondly, offender rehabilitation self-evidently requires personal change. Emotions are patterns of perception, experience, physiology, action and communication, which hold people together in social groups, are involved in decision making and behavioural regulation, and are the driving motivational forces of life (Niedenthal & Ric, 2017, pp. 2–3). Therefore, emotion would seem to be the kind of human experience that promotes empathic suffering and motivates personal change. I explore the notion of remorse as an emotion in the next section.

6. Remorse as an emotion

Of the notions of remorse I discussed previously, Gaita’s (2004) description of remorse as a recognition of the reality of wronged others is seemingly more moral encounter than emotion. Weisman’s (2009) description of a remorseful performance includes observable emotional expression, outward manifestations that an offender assumes responsibility, and observable behavioural change, but is not a description of emotional experience. Finally, the list of remorseful characteristics (Proeve & Tudor, 2010) includes cognitive, behavioural, feeling and motivational elements. According to this list, remorse would seem to be emotional. However, limited evidence suggests that remorse may not be a distinct emotion, but it is closely related to three emotions.

Shaver et al. (1987) asked participants to rate a large number of emotion names on the extent to which they are prototypical emotions and, using statistical clustering, determined the similarity of different emotions to each other. One small cluster contained four emotions only: guilt, shame, regret and remorse. By asking participants to rate emotion experiences on a variety of items, Davitz (1969) developed phenomenological descriptions of various emotions, including remorse. There was substantial overlap in features of remorse with shame and guilt. For example, remorse, shame and guilt shared a sense of regret. Overall, around half of the features of remorse overlapped with guilt or with shame. So, remorse seemed to be related to shame, guilt and regret. A related question is whether remorse is distinct from these three emotions. Proeve (2001) compared self-reported content and emotion characteristic ratings of remorse, regret, shame and guilt experiences. All four emotions were characterised by ‘regret’ thinking, wishing that things were otherwise and that one could turn back time. Overall, shame experiences had the most unique features, so there were more similarities between remorse and guilt or regret. However, Bhushan et al. (2020) developed scenarios to distinguish shame, guilt and remorse. They found that shame and remorse overlap and are distinct from guilt. From these studies, remorse appears to be related to, but not clearly distinct from, shame, guilt and regret. Therefore, it may be more useful to understand remorse experiences in terms of the emotions of shame, guilt and regret. For all three emotions, there is evidence that they are distinguishable emotions. I describe shame, guilt and regret in turn.

There is a relatively large literature concerning shame, which suggests that it includes these features: judgment of the self as inferior, incompetent or bad in some way (e.g. H. B. Lewis, 1971; Nathanson, 1992; Tangney, 1999; Wurmser, 1981); thinking of oneself as judged by others (e.g. Barrett, 1995); averting one’s gaze from others (e.g. Izard, 1977); hiding oneself from others (e.g. Barrett, 1995; Deigh, 1992; Nathanson, 1992; Tangney, 1995); and externalising strategies to control shame, such as denying a negative action, or blaming others (e.g. H. B. Lewis, 1971; Nathanson, 1992).

Shame can be distinguished from guilt, which has the following characteristics: an appraisal that one has violated a personal standard (e.g. Barrett, 1995; H. B. Lewis, 1971); appraisal that he or she is responsible for the violation of personal standards (e.g. Baumeister et al., 1994; H. B. Lewis, 1971; Tangney, 1991); and atonement or reparation, such as apologising, or undoing damage as a result of one’s actions (e.g. Barrett, 1995; Baumeister et al., 1994; H. B. Lewis, 1971). Although people experiencing guilt may focus on aspects of the self that led to a failure of action (M. Lewis, 1995), they do not make global negative evaluations of the self (Teroni & Deonna, 2008).

Regret may apply to a wide range of objects, in that one can regret one’s own behaviour, shortcomings or mistakes, the behaviour of others, one’s circumstances, and whether one is responsible, or not, for the situation (Landman, 1993). Phenomenological characteristics of regret include thinking of a mistake made, feeling one should have known better, feeling like correcting one’s mistake, and wanting a second chance (Zeelenberg et al., 1998). Regret and guilt can be distinguished, in that guilt is predominantly felt in situations of interpersonal harm, whereas regret is felt in both situations of interpersonal harm and harm to oneself (Zeelenberg & Breugelmans, 2008). In relation to criminal offending, we might focus on a particular form of regret, in which a person regrets a past action simply because of the bad consequences for oneself that it has produced. This is a self-oriented type of regret, which has been referred to as prudential regret (Williams, 1981). It is useful in criminal justice settings to distinguish prudential regret, as judges tend to distinguish distress for personal consequences from distress related to criminal acts (Duncan, 2002; Rossmanith, 2015).

Because remorse as an emotion is related to, but not clearly distinct from, shame, guilt and regret, it may be useful to evaluate remorse in criminal justice in terms of shame, guilt and prudential regret, all of which have greater evidence of being distinguishable emotions than remorse. In the following sections, I apply this approach of evaluating remorse through the emotion categories of shame, guilt or regret to evaluating remorse in sentencing and in offender rehabilitation.

7. Remorse as an emotion in sentencing

Because remorse is a recognised mitigating factor in sentencing, so that there are incentives to offer a convincing performance of remorse, evaluation of genuine remorse is a concern in the courtroom. In this section, I consider the implications of understanding remorse in terms of shame, guilt and regret for types of evidence that judges may examine in coming to their findings about remorse. On the way to considering evidence of remorse, I suggest that it is helpful to ask the question ‘What does remorse do in sentencing?’ or ‘What is the work of remorse in sentencing?’. The work of remorse in sentencing can be deduced by examining legislation, case law and the comments of judges regarding remorse. I offer some brief examples of the work of remorse from these various sources.

In some Australian legislation, remorse or contrition is a mitigating factor as long as it includes acceptance of responsibility and acknowledgement of injury or damage (Crimes (Sentencing Procedure) Act, 1999; Sentencing Act, 2017), and may involve making reparation (Crimes Act, 1914; Crimes (Sentencing Procedure) Act, 1999; Sentencing Act, 2017). Case law in Australia and other common law jurisdictions states that remorse encompasses an appreciation of one’s impact on victims, a desire to repair damage and clear one’s conscience (R v Phillips, 2012), and realisation of the gravity of one’s actions, changed attitude and some imposed self-discipline (R v Anderson, 1992). Third, judges interviewed by U.S. researchers stated that remorse involves a distressing recognition of wrongfully harming another, acceptance of personal responsibility, regret, appreciation of impact on victims and desire to make reparation. For at least some judges, remorse is also believed to be predictive of greater rehabilitation and lower recidivism (Zhong et al., 2014). Finally, Weisman’s (2009) analysis of Canadian cases highlighted themes of acceptance of responsibility, empathic suffering and personal transformation. There are common themes in these various sources, that remorse should: demonstrate that a defendant takes responsibility for actions; involve acknowledgement of injury and impact to victims; involve suffering by the offender, which is empathic suffering; and involve the desire to make reparation. In addition, remorse should predict greater potential for rehabilitation and lower recidivism. If these themes represent the work of remorse in sentencing, but remorse is not a distinct emotion, then it may be useful to examine the extent to which the emotions of guilt, shame and regret might perform the work of remorse in sentencing.

Studies of the phenomenology of guilt, reported above, suggest that appraisal that one is responsible for the violation of personal standards is a core feature of guilt. Furthermore, atonement or reparation has been noted commonly as a feature of guilt. Other studies show that the tendency to experience guilt is associated with increased empathy, with increased ability to take the perspective of others and with self-reported moral behaviour (Leith & Baumeister, 1998; Tangney, 1991). The finding that a tendency to experience guilt was associated with increased empathy for others, and the ability to take their perspective, has been replicated in an offender population (Bumby, 2000). With regard to recidivism, Tangney et al. (2014) found that a tendency to experience guilt was associated with lower recidivism, and young offenders with a greater tendency to experience guilt, or who experience guilt more frequently, show lower recidivism (Hosser et al., 2008; Spice, 2013; Spice et al., 2015).

Studies of shame phenomenology show that shame involves judgment of the self, aversion of gaze or withdrawal from others, and effort to control shame by denying actions or blaming others. The tendency to experience shame is associated with lower empathy and greater personal distress, but is unrelated to taking the perspective of others (Leith & Baumeister, 1998; Tangney, 1991). Shame-related withdrawal is also associated with lower empathy for others, as well as greater anger, hostility, physical aggression and unethical decisions (Cohen et al., 2011). In an offender population, Bumby (2000) found that shame-proneness was associated with increased personal distress and externalising of responsibility. With regard to recidivism, Tangney and colleagues found that the tendency to experience shame had both adaptive and maladaptive relationships with future offending. Shame was associated with greater blaming of others or circumstances, and this externalisation predicted greater recidivism. However, shame also motivates social withdrawal, which predicted lower recidivism (Tangney et al., 2014). However, frequent feelings of shame in young offenders predicted higher recidivism (Hosser et al., 2008).

Regret involves thinking of a mistake made, feeling like correcting one’s mistake and wanting to get a second chance. Prudential regret, that type of regret for past actions because of the consequences for oneself, is evidently not particularly related to assuming responsibility for one’s actions or appreciating harm to others. However, there are interesting findings regarding this kind of regret and recidivism. A measure of prudential regret (although referred to as remorse) predicted lower likelihood of chronic, serious offending in young offenders (Corrado & Peters, 2013). In addition, prudential regret in adult property offenders was associated with lower odds of reconviction after 10 years (LeBel et al., 2008).

Consideration of the three emotions of guilt, shame and prudential regret in relation to the work of remorse suggests that, overall, guilt would seem most closely to do the work of remorse. Prudential regret does not include empathy for victims or necessarily involve taking responsibility for one’s actions, and therefore would not seem to do the work of remorse. Finally, shame would seem to be an unreliable emotion for doing the work of remorse, as shame and its associated reactions might result in lower empathy, or in blaming others.

If a judge were most concerned with remorse-related emotions as predictors of lower recidivism, then the studies described above suggest that prudential regret, like guilt, might indicate lower risk of recidivism, while shame may not. However, conclusions about remorse-related emotions and recidivism should be tempered by the knowledge that available studies are few and limited in various ways. Most studies are based on juvenile offenders (Corrado & Peters, 2013; Hosser et al., 2008; Spice, 2013; Spice et al., 2015) and may not apply to adult offenders, while the study of shame and recidivism in adults (Tangney et al., 2014) shows a complex relationship, in that withdrawal in response to shame was associated with lower recidivism, while the shame-related reaction of externalisation was associated with increased recidivism. In broader context, a review of predictors of sexual recidivism (Hanson & Morton-Bourgon, 2005) shows that variables that would seem to be associated with lack of remorse (lack of victim empathy, denial, low motivation for treatment) showed little or no relationship with recidivism. Therefore, the question of remorse and related emotions and their implications for recidivism should be considered with caution. Evaluating remorse in terms of shame, guilt and regret has implications for evidence of remorse in the courtroom. If guilt is most suited to doing the work of remorse in the courtroom, and shame is less suitable, then this has implications for demeanour evidence of remorse. Psychological research has shown no clear display of guilt. However, shame has a clear non-verbal display, of hunched posture and downcast gaze (Martens et al., 2012). Therefore, if a defendant shows non-verbal behaviour of this nature, it might well reflect shame rather than guilt. In contrast, other behavioural signs of ‘remorse’ are consistent with guilt—for example, a thoughtful apology made to a victim of crime which seems to grasp the reality of their experience. Additionally, clear indications of reparation, such as changing one’s behaviour in relation to drug use, or seemingly genuine engagement with professional help, are also consistent with guilt.

I conclude this section with two general observations regarding presentation of remorse in the courtroom. First, reliance on observable behaviours such as apology or making restitution as indications of remorse may be preferable, for judges and for mental health practitioners, to the difficult and perhaps impossible task of discerning a defendant’s genuine emotional state, under conditions in which convincing but insincere performance of remorse may considerably benefit the defendant. In addition, apology and restitution, regardless of motivation, may have the effect of validating the experiences of those who are harmed by the defendant’s actions. Second, some defendants with conditions such as intellectual disability, autism or schizophrenia may find it difficult to provide socially acceptable performances of remorse. In such cases, mental health practitioners may assist courts by explaining the connections between these conditions and their deficient performance of remorse.

7. Shame, guilt and regret in offender rehabilitation

In this section I want to consider the evaluation of emotional reactions to offending in offender rehabilitation through the framework of shame, guilt and prudential regret. Earlier, I discussed the idea that the expectation of remorse is present, if hidden in plain sight, in offender rehabilitation (Hall & Rossmanith, 2022) and the finding that inmates of forensic mental health facilities value the discussion of remorse (O’Donahoo & Simmonds, 2022). Other interview studies show that offenders discuss regret, shame and guilt in relation to their offences or in relation to desisting from crime (Elisha et al., 2012; Farrall & Calverley, 2006). In my own experience, offenders’ reactions to their own offending emerge, with little or no prompting, in individual or group discussions with mental health professionals. I suggest there are additional reasons that emotional reactions to offending should be explored during offender rehabilitation. First, as offending behaviour is the context for many encounters between professionals and offenders, the moral and emotional reactions of offenders must surely be noted and influence the engagement and relationship of professionals with their offender clients. Second, to engage with offenders in reflection about their offences is, hopefully, to engage with them as moral beings rather and not solely as offenders with collections of risk factors and criminological needs, whatever the value of attention to risk and criminological needs for offender rehabilitation. Third, I suggest that shame, guilt and prudential regret connect with broad notions of offender rehabilitation, and that each emotion may have different implications for interventions with offenders.

McNeill (2012) discussed four forms of offender rehabilitation: psychological, concerning individual-level change; legal, concerning requalification of offenders as citizens; moral, concerning satisfaction of moral demands; and social, concerning acceptance and social recognition of the offender. Given the interesting, but not well-established, findings regarding shame, guilt and regret with recidivism, which are described in the previous section, and evidence of limited effects on recidivism of rehabilitation programmes that address related concepts such as victim empathy (Beaudry et al., 2021), it may be that these emotions are less relevant to individual-level psychological rehabilitation. However, shame, guilt and regret, as different moral responses to wrongdoing, are very relevant to moral rehabilitation. In addition, these emotions and their associated speech and behaviours are likely to be judged differently by other people, and therefore affect social recognition and acceptance of offenders, so that they are relevant to social rehabilitation.

McNeill (2009) discussed the distinction between bonding and bridging social capital. Bonding social capital refers to close ties with family and friends, while bridging social capital refers to ties with a wider network of acquaintances and colleagues. Bonding social capital is associated with reduced recidivism, an association that may operate indirectly through employment, because social support from family members motivates offenders towards more conventional lifestyles (Berg & Huebner, 2011). McNeill (2009) suggested that it is important to help repair the bonding social capital of family ties, as well as assist development of bridging social capital with employers, non-government organisations or faith communities.

Guilt, shame and prudential regret may each be associated in different ways with social capital. Guilt is tied to specific behaviour for which one is responsible, is linked to failures that flout norms, and is commonly concerned with private norms, without global negative evaluation of self (Teroni & Deonna, 2008). Guilt is also associated with approach actions of repair and reparation, and because of the orientation towards reparation, the offender is likely to be motivated to repair bonding social capital and to engage with bridging social capital. With these considerations in mind, helpful interventions for a person experiencing predominant guilt might include exploring an offence pathway, to address the question ‘How did I come to do that crime?’ as well as reflecting on and returning to living according to one’s values. In addition, as noted in self-forgiveness literature (Cornish & Wade, 2015), the guilty person may want to explore ways of reparation, including direct reparation to those harmed, or indirect reparation through actions such as voluntary work. Engagement with social capital may be facilitated by understanding one’s offending narrative and then exploring ways of explaining that narrative to others. This might include answering difficult questions from family members or friends who seek to reconcile their understanding of the person they knew with their knowledge of that person’s offending.

Shame is related to the entire self, which is judged as bad or incompetent. Shame is commonly associated with loss of reputation and perceived judgment by others, and it leads to action tendencies such as avoidance of others (Teroni & Deonna, 2008). Because there may be withdrawal from others, there might be little basis for building or rebuilding social capital of either kind. Highly ashamed offenders may benefit from psychotherapeutic intervention oriented to moving towards self-compassion or self-acceptance, and from shame to guilt (Fisher & Exline, 2010). To move from shame to guilt, it is important to learn to define oneself separately from one’s offences, while also taking responsibility for past actions and their effects. Additionally, highly ashamed offenders may have an enhanced sense of scrutiny and condemnation by others. In such cases, behavioural experiments from the cognitive–behavioural tradition may be helpful in order to challenge such perceptions and beliefs.

Prudential regret may be associated with the narrative of a person who made mistakes, but is not necessarily a bad person. For an offender who experiences prudential regret, social capital may be not particularly affected if one’s support network has not been directly affected by the offending. However, if people in the support network were directly affected, or they have a strong moral position regarding that offending, then social capital may be damaged. With regard to interventions, the prudentially regretful offender is likely to be open to problem-solving approaches, seeking work and pursuing other aspects of social reintegration. The Good Lives Model (Purvis et al., 2011) may provide a suitable approach in this situation, as it is designed to allow offenders to explore ways of achieving personally meaningful outcomes in ways that do not involve offending.

9. Conclusion

Remorse is present in a number of domains of criminal justice, including sentencing, parole and offender rehabilitation. Despite problems presented by the task of evaluating remorse, including evidence for remorse and the apparent necessity to perform remorse, it is likely to persist in criminal justice, given its relationship to fundamental values in criminal justice and its role in relationships and moral communities. It may be helpful to address the evaluation of remorse using a framework of three reactions to offending: guilt, shame and prudential regret. This framework has implications for evidence of remorse in sentencing and for intervention approaches in offender rehabilitation. The framework requires further elaboration, theoretical development, and testing of its relevance to sentencing, rehabilitation and additional domains of criminal justice.

Ethical standards

Declaration of conflicts of interest

Michael Proeve has declared no conflicts of interest

Ethical approval

This article does not contain any studies with human participants or animals performed by the author.