The Remorse Dilemma: Limitations of Assessing Remorse in Criminal Sentencing

Jason Ruggeberg, 3L, University of Toronto Faculty of Law

This paper explores the delicate balance trial judges must strike when evaluating an accused person’s remorse. Remorse is a key mitigating factor in criminal sentencing which, alongside an offender’s having insight into and taking responsibility for their past conduct, may show that they have “achieved a change in attitude or imposed some self-discipline which significantly reduces the likelihood of further offending.”[1] It therefore warrants a lower sentence in light of the statutory sentencing objectives of specific deterrence, the need to separate offenders from society, rehabilitation, and promoting a sense of responsibility in offenders.[2] Conversely, the failure to express remorse might suggest the need for greater sanctions in some cases, as I will discuss in a forthcoming publication on this blog. 

Despite the fact that remorse impacts the punishment of offenders, it is difficult to define and difficult to apply. Other than by allusions made in section 718(f) of the Criminal Code, which identifies promoting “a sense of responsibility in offenders, and acknowledgment of the harm done to victims or to the community” as an objective of sentencing, remorse is not mentioned in the Code’s sentencing regime.[3] Canadian courts have yet to expansively consider the definition of remorse, so we can look to secondary sources for its interpretation. Susan Bandes, a leading scholar in the study of the role of emotion in law, defines remorse as “recognition of the wrong and the harm it has caused, deep regret for the act, and a desire to atone for the act.”[4]

Canadian courts generally approximate this characterization, treating remorse as distinct from but related to other concepts, such as insight into the offence, acknowledgement of and taking responsibility for wrongdoing, and willingness to make amends.[5] For example, in R v McLellan, the Ontario Court of Appeal stated that “[l]ack of remorse and insight are factors” that the trial judge was entitled to consider, implying they are separate constructs.[6] However, the court went on to say the trial judge did not err in finding the accused lacked remorse because he lacked insight into his conduct and continued to externalize blame, treating insight and accepting responsibility as elements of remorse.[7] Similarly, in R v Magas, the Alberta Court of Appeal treated willingness to make amends as a component of remorse.[8] The court affirmed that the accused’s payment of partial restitution was a mitigating factor, but her failure to make full restitution despite having the means to do so militated against her claim that she was genuinely remorseful.[9] Thus, the courts treat remorse as both constituted in part by, and working synergistically with, related concepts such as insight and acceptance of responsibility.

This paper examines difficulties in evaluating remorse in the courtroom. First, remorse is often assessed using subtle factors like the accused’s courtroom demeanour or paralinguistic cues. Unfortunately, evidence indicates that these factors, like lack of eye contact, are ambiguous.[10] Second, these cues may be perceived differently depending on characteristics of the accused, such as race, culture, socioeconomic status, and mental health.[11]

This analysis is not intended to suggest that remorse should be foregone entirely in criminal sentencing. Rather, it emphasizes that trial judges should be particularly mindful of the complexities of assessing remorse, both when determining whether accused persons are remorseful and when determining the weight that should be attributed to those findings.

A) Cues Used to Assess Remorse Are Ambiguous

The first issue with applying remorse in criminal sentencing is that it is often ambiguous to assess. Only genuine remorse is considered mitigating, which leads judges to scrutinize the sincerity of the accused.[12] In doing so, judges tend to make global assessments of remorse based on demeanour, nonverbal cues, verbalizations, actions outside the courtroom, and corroborating sources.[13] However, the cues used by judges to do so may be unreliable, particularly when applied to different individuals based on relatively brief interactions in the courtroom. This leaves room for personal characteristics of the accused such as culture, race, class, and mental health to influence perceptions of remorse, potentially creating systemic bias in its use.

Interviews with judges indicate that they have­­ difficulty evaluating remorse because demeanour is an unreliable predictor of authenticity.[14] For example, through open-ended interviews with 23 sitting judges in the Connecticut State Superior Court Criminal Docket about their sentencing practices, Zhong found that while some judges were confident in their abilities, others felt that the signs of remorse “can all be faked.”[15] Similarly, Rossmanith found that a common theme across interviews with eighteen judges and magistrates in the Australian criminal justice system was that demeanour was regarded as “an unreliable guide as to whether or not [an offender] was remorseful.”[16] Despite these concerns, judges in both studies still reported incorporating demeanour cues into their assessments of remorse.

As a specific example, gaze aversion (whether or not an accused meets the eyes of people speaking with them), is not a reliable determinant of whether remorse is feigned.[17] However, Griffin found that the majority of respondents in a survey of 56 American judges indicated that they frequently used gaze aversion to determine the sincerity of remorse; 73% of respondents stated they used it “more often than not” or even more frequently.[18]

Beyond how (in)accurately specific cues can predict remorse, judges’ interpretations of those cues may also vary. Zhong found that “[m]any behaviours that indicated the presence of remorse to some judges indicated its absence to others.”[19] For example, some judges described silence as an indicator of shyness, fear, poor public speaking skills, or mental illness.[20] Others described it as a sign of remorselessness, disengagement, and distraction.[21] Eye contact, putting down or hanging one’s head, and apology letters were similarly polarizing.[22] These differences likely result to some degree from context; because the judges were interviewed without reference to a particular offender, there was likely greater variance than if each judge was asked to assess the same case. Nonetheless, assuming a relatively random distribution of accused persons from which the judges drew their conclusions, the study still suggests a degree of ambiguity in the indicators of remorse discussed.

Furthermore, although some studies have identified reliable indicators of the presence and sincerity of remorse within a given individual, those indicators may be too subtle to use reliably in the courtroom. For example, Griffin found differences in the verbal content and degree of smiling between videos of feigned and genuine remorse, but the researchers emphasized that the differences were quite small and questioned whether they would be noticeable to an observer.[23] Similarly, although Brinke found that feigned remorse is associated with greater verbal hesitation and differences in emotional expression, these findings were based on frame-by-frame analysis of videos of people telling stories by observers trained to recognize emotional cues.[24]

Moreover, these cues would be less useful in the dynamic conditions of a courtroom than in the research setting. First, the highly formalized environment of the courtroom may affect how accused persons express remorse, and those effects may differ from offender to offender. Second, the relatively brief encounters that a judge has with an accused offer limited opportunity for them to establish a baseline from which to assess these subtle changes. Instead, judges are likely to assess expressions of remorse by comparison to other offenders, which is likely to magnify inaccuracy and bias based on personal characteristics.

B) Characteristics of the Offender Affect How Remorse Is Expressed and Evaluated

The ambiguities involved in assessing remorse are worse for those with protected characteristics. Characteristics such as culture, race, class, mental health, and age may all be linked to whether and how an accused expresses remorse and how judges assess it. Bandes argues that judges may develop expectations that remorse be expressed in a particular way based on their own personal characteristics.[25] Where expressions of genuine remorse do not follow this “script,” judges may fail to recognize them or deem them insincere, meaning it may be more difficult for those from minority communities to benefit from the mitigating circumstance of remorse.

            Cultural differences may affect evaluations of the appropriateness of remorse in a given scenario, as well as its expression. Thus, judges may misperceive valid expressions of remorse if they come from a different background than the accused.[26] For example, in a study on sentence reductions offered to offenders who accept responsibility for their criminal behaviour in the United States, one judge interviewed by Everett and Nienstedt noted “Hispanic males’ difficulty in openly and publicly admitting guilt, ‘to look you in the eye and say they’re sorry.’”[27] Similarly, one of the judges interviewed by Rossmanith noted that indigenous Australians may not make eye contact for cultural reasons.[28] Bennettobserves that such differences in the expression of remorse create a “dual aspect” concern: “[S]ome individuals may be disadvantaged… by cultural norms prohibiting remorse, but also by observers’ failure to ‘read’ displays of remorse correctly.”[29] This concern is particularly salient in a multicultural society like Canada, where sentencing judges may frequently interact with offenders from cultural groups whose expressive norms they are not familiar with.

            Cross-racial differences may cause similar issues. People tend to decode the emotions of those in their own racial group more carefully and may impute implicit stereotypes into their assessments.[30] Although not dealing with judges, Bowers and Steiner found that white jurors were particularly unlikely to perceive black defendants as remorseful compared to black jurors.[31] This type of bias can operate implicitly to resolve ambiguities in perception and evaluation.[32] As such, faced with ambiguous or conflicting cues, judges may implicitly apply negative stereotypes that lead them to conclude an offender is remorseless. Furthermore, the cultural differences discussed above could compound this phenomenon by increasing ambiguity.

Culture, race, and class may also affect other cues of remorse such as verbalizations and corroboration. As Bandesexplains: “When judges expect particular verbal formulations, they may be unintentionally penalizing defendants who are inarticulate, uneducated, or simply unversed in certain styles of speaking and writing that resonate with upper class judges.”[33] Thus, judges may be less moved by the expressions of remorse of accused who are from a different class or culture, potentially viewing them as less convincing. Furthermore, accused with less means or who are from communities with less access to rehabilitative programming may be unable to “walk the walk” before sentencing to corroborate their remorse. This could compound a judge’s existing concerns about sincerity stemming from ambiguous cues or cultural differences, leading them to erroneously conclude that an accused lacked remorse.

Psychiatric conditions can have effects on perceptions of remorse similar to culture, race, and class, but may be less predictable. Like culture, psychiatric conditions can alter how people express themselves. For example, mood disorders may cause both “over the top” emotionality and a flattened affect, whereas autistic people may be more “matter of fact” in their descriptions of offences.[34] Other conditions, like schizophrenia, may also impair the accused’s ability to articulate their thoughts in a manner that neurotypical judges can easily understand.[35] This again raises the issue of offenders not matching the expected “script” when expressing genuine remorse. Medication may further complicate the situation, as psychotropic drugs can blunt an accused’s emotions and render their facial expression an unreliable indicator of their state of mind.[36] Given the complexities in how psychiatric conditions manifest, it would be unrealistic to expect judges – even if reliably informed regarding an accused’s psychiatric conditions – to accurately assess how those diagnoses influence their demeanour.

Furthermore, like race and culture, psychiatric disorders are associated with stereotypes that may implicitly inform judgments. People suffering from mental illness may be viewed as more unpredictable, less capable of change, and less capable of remorse.[37] Thus, notwithstanding judges’ ability to recognize how psychiatric conditions may affect expressions of remorse in the first place, their implicit beliefs could still lead them to devalue genuine expressions of remorse.

Lastly, age differences may also modulate how specific cues of remorse are expressed. Bandes notes that youth are often unwilling to show vulnerability in public and may be particularly intimidated by the artificial and formal nature of the courtroom setting.[38] Youth may also take longer to comprehend the gravity of their crime than adults, despite their cases often proceeding faster.[39]  Thus, youth may not be at a stage of reflection that allows them to express remorse during sentencing but come to feel remorseful once they have fully processed the event post-trial.

Pre-sentence reports may compound bias based on the characteristics discussed above. Courts may order a probation officer to prepare a pre-sentence report on the offender where the sentencing judge would benefit from additional information about an offender’s “background and… prospects for the future.”[40] Unless otherwise specified by the court, pre-sentence reports must include information about the offender’s character, behaviour, attitude, and willingness to make amends.[41] Probation officers working with offenders with psychiatric diagnoses noted the same issues reported above: offenders’ difficulty in expressing remorse in conventional ways (and officers’ difficulties in understanding their expressions) as well as stereotypes of unpredictability, inability to change, and inability to feel remorse.[42] Stereotypical views about offenders can also form based on characteristics such as cultural background and personal history and then manifest in the narratives conveyed by pre-sentence reports.[43] Narratives may be particularly negative where probation officers are unable to imagine the offender’s inner world, connect emotionally, or read their expression of taking responsibility, which may itself be the product of differences in culture, class, and mental health.[44]This means that pre-sentence reports may fail to find remorse even when it is present due to the same biases impacting a judge’s assessment of remorse. These reports may therefore reinforce potentially biased interpretations of ambiguous signs that judges already form. This could be particularly problematic for offenders who plead guilty, because judges would have less opportunity to create an independent opinion of the offender before reading the report. In some cases, their first impressions of the offender during the sentencing hearing could be coloured by the probation officer’s interpretations, potentially leading to confirmation bias. Thus, uncertainty surrounding the assessment of remorse may be compounded when pre-sentence reports are available and the offender pleaded guilty.

C) Conclusion

                  Though remorse is fundamental to criminal sentencing, this paper has explored how its application is fraught with ambiguity and the potential for unfair application. Judges may overemphasize ineffective means of assessing remorse to the detriment of offenders from different cultures and mental health profiles. These issues leave the assessment of remorse on unsure footing. On the one hand, remorse is a powerful indicator that an accused person has reflected on their wrongdoing and wishes to change. Proportionality therefore dictates that a remorseful offender should receive a lesser sentence. On the other hand, if remorse cannot be assessed reliably, giving it too much weight may contribute to disproportionate sentences and exacerbate existing disparities in the sentencing of minority offenders. The analysis in this paper does not suggest remorse should be removed from criminal sentencing. Instead, it indicates that judges should understand the drawbacks of assessing remorse based on subtle cues, especially with respect to disadvantaged offenders.

[1] R v Anderson, 2991 CanLII 6002 (BCCA) at para 45, cited in R v Friesen, 2020 SCC 9 at para 165 [Friesen] [emphasis removed]; see also R v Reeve, 2020 ONCA 381 at para 11 [Reeve].

[2] R v McGee, 2020 BCCA 362 at para 23; Reeve, supra note 1 at paras 11-14.

[3] Criminal Code, RSC 1985, c C-46, s 718(f) [Criminal Code]; R v Al-Saedi, 2017 ONCJ 204 at para 20.

[4] Susan A Bandes, “Remorse and Judging” in Steven Tudor et al, eds, Remorse and Criminal Justice: Multi-Disciplinary Perspectives (London: Routledge, 2021) 18, online:  <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3535062> [Bandes, “Remorse and Judging”].

[5] See e.g. Friesen, supra note 1 at para 165; Reeve, supra note 1 at para 11.

[6] 2012 ONCA 717 at para 32 [emphasis added].

[7] Ibid at para 36. See also R v Al-Isawi, 2017 BCCA 163 at paras 83-84, holding the trial judge did not err in giving minimal weight to remorse as a mitigating factor (in part) because the accused continued to blame corruption amongst law enforcement officials for his conviction.

[8] 2012 ABCA 61.

[9] Ibid at paras 14-20.

[10] See e.g. Rocksheng Zhong et al, “So You’re Sorry? The Role of Remorse in Criminal Law” (2014) 42:1 J Am Academy Psychiatry & L 39, online: < https://jaapl.org/content/jaapl/42/1/39.full.pdf>.

[11] For an overview, see Susan A Bandes, “Remorse and Criminal Justice” (2015) 8:1 Emotion Rev 14 at 16–17, online: <https://journals.sagepub.com/doi/pdf/10.1177/1754073915601222?casa_token=_CfGMK06KXoAAAAA:IavRFhWHJy2lZ_Wkv0z9k-c> [Bandes, “Criminal Justice”].

[12] Reeve, supra note 1 at paras 10–11.

[13] Desiree Adams Griffin, Exploring remorse behaviors: Verbal and nonverbal indicators of authentic, exaggerated, and feigned remorse (PhD Dissertation, University of Alabama, 2011), online: <https://www.proquest.com/docview/904545137/A5B17587589C4A84PQ/1?accountid=14771&sourcetype=Dissertations%20&%20Theses>; Kate Rossmanith, “Affect and the Judicial Assessment of Offenders: Feeling and Judging Remorse” (2015) 21:2 Body & Society 167, online: <https://librarysearch.library.utoronto.ca/permalink/01UTORONTO_INST/fedca1/cdi_crossref_primary_10_1177_1357034X14558073>; Zhong, supra note 10.

[14] See e.g. Rossmanith, supra note 13; Zhong, supra note 10.

[15] Zhong, supra note 10 at 43.

[16] Rossmanith, supra note 13.

[17] Griffin, supra note 13 at 59–68, 115–16.

[18] Ibid at 84, 92, 116.

[19] Zhong, supra note 10.

[20] Ibid at 44.

[21] Ibid.

[22] Ibid at 43–44.

[23] Griffin, supra note 13 at 26–44, 56–57, 111–12.

[24] Leanne ten Brinke et al, “Crocodile Tears: Facial, Verbal and Body Language Behaviours Associated With Genuine and Fabricated Remorse” (2012) 36:1 L & Human Behaviour 51 at 54–58, online: <https://www.proquest.com/docview/914688382?accountid=14771&parentSessionId=jzNU1k6RjOODlIx9JARUS%2BkSMojqPf7%2F%2Beu30dtDl%2Fc%3D&pq-origsite=primo&sourcetype=Scholarly%20Journals#s20> [Brinke].

[25] Such as class or education, see e.g. Bandes, “Criminal Justice”, supra note 11.

[26] Christopher Bennett, “The Role of Remorse in Criminal Justice” in Michael Tonry, ed, Oxford Handbook Online in Criminology and Criminal Justice (New York: Oxford University Press, 2016) at 16, online: <https://eprints.whiterose.ac.uk/102428/3/The%20Role%20of%20Remorse%20in%20Criminal%20Justice%20final.pdf>.

[27] Ronald S Everett & Barbara C Nienstedt, “Race, Remorse and Sentence Reduction: Is Saying Sorry Enough?” (1999) 16:1 Justice Q 99 at 117–18, online: <https://journals-scholarsportal-info.myaccess.library.utoronto.ca/pdf/07418825/v16i0001/99_rrasrisyse.xml_en>.   

[28] Rossmanith, supra note 13.

[29] Bennett, supra note 26 at 16.

[30] Bandes, “Criminal Justice”, supra note 11.

[31] William J Bowers, Benjamin D Steiner & Maria Sandys, “Death Sentencing in Black and White: An Empirical Analysis of the Role of Jurors’ Race and Jury Racial Composition” (2001) 3:1 U Pa J Const L 171, online: <https://scholarship.law.upenn.edu/jcl/vol3/iss1/3/>.

[32] M Eve Hanan, “Remorse Bias” (2018) 83:2 Mo L Rev 302 at 334–35, online: <https://scholarship.law.missouri.edu/mlr/vol83/iss2/6>.

[33] Bandes, “Remorse and Judging”, supra note 4 at 30.

[34] Colleen M Berryessa, “Exploring the Impact of Remorse on Recommendations for Sentencing Diversion for Defendants with Psychiatric Diagnoses” (2023) 39:4 J Contemporary Crim Justice 491, online: <https://journals-sagepub-com.myaccess.library.utoronto.ca/doi/full/10.1177/10439862231189416>.

[35] Ibid.

[36] Georgina Stobbs & Mark R Kebbell, “Jurors’ Perception of Witnesses with Intellectual Disabilities and the Influence of Expert Evidence” (2003) 16:2 J Applied Research in Intellectual Disabilities 107, online: <https://psycnet.apa.org/record/2004-11101-002>.

[37] Ibid.

[38] Bandes, “Criminal Justice”, supra note 11.

[39] Ibid; see Youth Criminal Justice Act, SC 2002, c 1 [YCJA] (“the criminal justice system for young persons must be separate from that of adults, must be based on the principle of diminished moral blameworthiness or culpability and must emphasize the following…(iv) timely intervention that reinforces the link between the offending behaviour and its consequences” s 3(1)(b)).

[40] Criminal Code, supra note 3, ss 721(1), 721(3); R v Junkert, 2010 ONCA 549 at para 590; YCJA, supra note 39, s 40. Pre-sentence reports may be particularly appropriate for youth (see e.g. YCJA, supra note 39, s 39(6), presumptively requiring the court to consider a pre-sentence report before imposing a custodial sentence on a young offender) and first-time offenders charged with serious matters (see R v Bates, 1977 CanLII 2054 (ONCA)).

[41] Criminal Code, supra note 3, ss 721(3); YCJA, supra note 39, s 40(2).

[42] Berryessa, supra note 34.

[43] Colleen M Berryessa, “Modeling ‘Remorse Bias’ in Probation Narratives: Examining Social Cognition and Judgments of Implicit Violence During Sentencing” (2002) 78:2 J Soc Issues 452, online: <https://spssi.onlinelibrary.wiley.com/doi/pdf/10.1111/josi.12508>.

[44] Ibid.