Discrimination in Sentencing: Showing Remorse and the Intersection of Race and Gender

The Role of Remorse in Jury Decisions

Legal scholars have long debated whether jurors should consider the presence or absence of offender’s remorse when making sentencing decisions. Proponents of a remorse principle emphasize the moral aspect of remorse (Tudor 2008). In this line of reasoning, remorse not only merits acknowledgment within the legal system but also holds significant potential for fostering reconciliation between offenders and victims (Bibas and Bierschbach 2004). Conversely, those who deem remorse irrelevant justify their objections by highlighting the challenges of discerning genuine remorse, especially when individuals are motivated to engage in surface-level displays of emotion (Duncan 2002). Opponents also cite the lack of empirical evidence supporting a correlation between remorse and recidivism against the remorse principle (Bagaric and Amarasekara 2001).

Despite these unresolved disputes, empirical work consistently shows that rulings are shaped by judges’ perceptions of remorse (Bandes 2016; Weisman 2014). Affect control theory, a sociological account of emotions, proposes that we rely on others’ emotional expressions as cues to assess an individual’s character or identity (Heise 2007; MacKinnon and Robinson 2014; Robinson and Smith-Lovin 1999). Thus, when signs of remorse are absent, jurors are more inclined to view the defendant as criminal. These criminality assessments, in turn, lead to harsher punishment (Robinson et al. 1994; Tsoudis and Smith-Lovin 1998, 2001; Zhao and Rogalin 2017). Given the influence of emotions on character (identity) assessments, that we cannot know what kind of person an offender really is unless we know how they feel after wrongdoings (“people should not feel good for being bad”), these findings offer empirical insights into the decision-making process of jurors. Conspicuously absent from this literature is a discussion of how race and gender may complicate the signals of showing and perceiving remorse. Because each demographic characteristic carries its own set of preconceived assumptions about expected emotional behaviors, jurors may interpret—and reward—Black and White women and men’s displays of remorse differently.

In the next two sections, we briefly review literature on emotional norms. Although the larger strand of the literature focuses on gendered emotion norms, we also look into a smaller yet emerging strand that discusses emotions related to race. Following the lead of Bonilla-Silva (2019), we conceptualize remorse as a gendered and racialized emotion, laying the theoretical foundation for our predictions about the unequal payoffs of showing remorse. ¹

Gendered Emotion Norms

Cultural beliefs about individuals’ emotional behaviors are largely gendered. With the exception of anger and pride, people believe women experience and express more emotions than men (Shields 2002). Gendered emotional stereotypes are also consistent with the notions of conventional masculinity and femininity. Whereas women are expected to display positive and communal emotions, men are prescribed to display powerful emotions (Hutson-Comeaux and Kelly 2002). When people violate these gendered norms, such as women professionals expressing anger and exhibiting domineering behaviors at work, they often incur social sanctions and backlash (Brescoll and Uhlmann 2008).

When it comes to remorse, however, there are exceptions to these gendered emotional norms with important implications for inequality. Specifically, although men are less often permitted to show weak and powerless emotions (e.g., fear and deference), due to cultural changes in ideal manhood (Bridges and Pascoe 2014), men’s expressions of feelings do not always contradict perceptions of competence (Warner and Shields 2007). In fact, not only are people unbiased against men who display controlled expression of deep feelings (Shields 2002; Zawadzki, Warner, and Shields 2013), but respondents also evaluate men who had moist eyes in the death of an intimate partner more positively than similarly sorrowful women (Fischer et al. 2004). In the context of criminal justice, research further demonstrates displaying remorse reduces participants’ negative evaluations of the offenders, but it does more so for men than for women (Zhao and Rogalin 2017). Thus, violating gendered emotion norms can yield benefits instead of sanctions, but largely for privileged groups (i.e., men).

Emotion Norms Related to Race

Compared to well-documented gender stereotypes about emotion, scholarship on emotions and race is scarce. ² We take, however, as our point of departure a growing literature that emphasizes the ways in which Black people put up an emotional shield to appear emotionally stoic and in control (Collins 2004; Jackson 2018; Majors and Billson 1992). Although not based in the study of emotions, early work on criminalization documents the display of “toughness” by working-class Black men in urban cities (Anderson 1999; Ferguson 2001). For example, when discussing how violence emerges in inner-city contexts, Anderson (1999) argues that Black men with lower socioeconomic status adopt a “cool pose” to convey a sense of control (Majors and Billson 1992). By acting emotionally aloof, working-class Black men pursue a form of masculinity that is unattainable for them through conventional approaches, such as employment and wealth. Ironically, realizing the constraints of these controlling images, Black people strategically defy racialized emotional stereotypes by displaying emotional sensitivity and vulnerability to communicate their moral standing to criminal justice gatekeepers, such as police officers (Stuart and Benezra 2018).

More recent scholarship directs its attention to the emotion of anger, which holds particular significance for middle-class Black Americans (Dow 2016; Durr and Wingfield 2011; Wilkins 2012; Wingfield 2013, 2021). Because our culture paints Black people as vicious and out of control, Black professionals feel compelled to conceal feelings of frustration and irritation, engaging in emotion management to avoid racial stigma (Evans and Moore 2015; Jackson 2018). This kind of emotion management, however, is not expected of their White counterparts because racial power dynamics within organizations legitimize expressions of anger and agency from White men (Pierce 1999; Wingfield 2010, 2021). Thus, similar to professional women who face a double-bind, where their expressiveness and emotional control are both negatively evaluated (“too emotional” vs. “too cold”; Kelly and Hutson-Comeaux 2000), Black professionals also find themselves caught in a dilemma: Their emotional outbursts in response to racism are often interpreted as having an “attitude,” whereas appearing pleasant and collegial can reinforce racial hierarchies (Ray 2019; Wingfield and Alston 2014).

The extant literature so far presents a sketch of emotionality with a narrow range of appropriate emotional expressions for Black men and women. It has yet, however, to explicitly examine the emotion of remorse. As a negatively valenced emotion, remorse is often perceived as a proxy for an individual’s moral character (Tangney, Stuewig, and Mashek 2007). Given the disproportionate effect of mass incarceration on Black people, prejudice against Black people could potentially influence the assessment of remorse in racist ways. As such, we should consider remorse as a racialized emotion: Whereas a Black defendant’s expression of remorse is deemed dubious and likely to be discounted, emotional pleas from a White defendant are more likely to be interpreted as signs of rehabilitation (Hanan 2018).

An Intersectionality Perspective

To predict sentencing outcomes for offenders from various racial and gender groups, we consider possible intersectional processes that may create characteristic cultural binds and freedoms. Following Ridgeway and Kricheli-Katz (2013), we use the term cultural binds to refer to challenges stemming from the overlapping stereotypes related to gender, race, and emotion that place group members at a disadvantaged position. Similarly, we use the term cultural freedoms to refer to positive outcomes resulting from nonprototypical and disadvantaging status (i.e., Black, woman).

We start with White women and men. Broadly speaking, emotional norms tend to perceive women’s expression of sadness as part of their emotional disposition, whereas they often view men’s tears as more sincere and positive (Zawadzki et al. 2013). Consequently, respondents discount emotional distress exhibited by women while seeing similarly remorseful men as less criminal, thus bestowing on them an “emotional display premium”—higher returns for expressing emotions (Zhao and Rogalin 2017). We expect gender disparity in remorse displays to mostly apply to White women and men, though. White people represent the dominant racial group by which others are contrasted (cf. Fiske et al. 2002). This suggests that White women and men are judged and treated in more stereotypical terms. Given men’s emotional display premium and the prototypicality of White men in the categories of race and gender (Purdie-Vaughns and Eibach 2008), we expect that White men displaying remorse will receive more favorable evaluations. These evaluations, in turn, are likely to result in lighter sentences.

We propose competing hypotheses for Black men. On one hand, displays of remorse go against stereotypical expectations associated with Black men. A growing body of empirical research identifies strong effects of counter-stereotypes in reducing prejudicial attitudes (Blair 2002; Livingston and Pearce 2009; Pedulla 2014). For example, Black men tend to be more readily accepted as leaders when they exhibit traits such as having a “baby face” (Livingston and Pearce 2009) or identifying as gay (Pedulla 2014)—attributes that often carry negative consequence for White men in the labor market. Perhaps by displaying distress, Black men activate a “disarming mechanism”— namely, “physical, psychological, or behavioral traits that attenuate perceptions of threat by the dominant group” (Livingston and Pearce 2009:1229). If this holds true, Black men will be equally, if not more, rewarded in sentence reduction compared to groups who face fewer negative stereotypes. That is,

Alternatively, expressing remorse may not necessarily yield any benefit for Black men. Despite recent shifts in the performances and understanding of masculinity (Bridges and Pascoe 2014), the cultural transformations toward more “sensitive” styles of masculinity are primarily available to young, White, heterosexual men (Messner 1993). Due to their subordinate race status, Black men run a greater risk for violating proscriptive gender norms, and they may consequently be excluded from any premium associated with the display of remorse. It is also important to note that expressions of remorse allow individuals to lay claims to a moral judgment of their character, but they too acknowledge “personal culpability” (Tangney et al. 2007). Hence, it stands to reason that:

The story is somewhat more complex for Black women. Black women do not fit neatly into the prototypical categories of “woman” or “Black” (Purdie-Vaughns and Eibach 2008; Sesko and Biernat 2010). ³ Being in this “off-diagonal” social position affords certain freedoms, such as a reduced susceptibility to the application of race and gender stereotypes to Black women (Ridgeway and Kricheli-Katz 2013). For example, research finds that compared to White women, women of color face less backlash when they deviate from stereotypic gender expectations (Livingston, Rosette, and Washington 2012; Ong 2005, Tinkler et al. 2019; Wingfield 2010). Note, however, that the Black women in these studies are often described as highly competent. Proving high ability seems to be a key factor for professional Black women to leverage their authority and agency. Indeed, Black women face harsher judgment when they make mistakes (Rosette and Livingston 2012) and are overcriminalized for minor offenses (Potter 2015). They also experience unique biases when assuming roles and positions that expect a more traditionally feminine demeanor (Anderson et al. 2018; Kilbourne, England, and Beron 1994). Whereas White women’s tears may invoke empathy, tears from women of color are often seen as signs of vulnerability (Srivastava 2006). Given the legal context of our study and the inherently feminine nature of remorse expression, we hypothesize that:

In summary, the literature on gendered and racialized emotional norms leads us to expect that White men will receive the largest sentence reduction when displaying remorse, followed by White women and Black women. We predict two contrasting outcomes for Black men: Remorseful Black men will either receive equal or the most rewards (i.e., counter-stereotypical benefit) or receive the fewest rewards (i.e., cultural binds).

As a first step in understanding how remorse works, in this study, we explore two potential mechanisms that could help explain any observed disparities in sentencing reductions. First, participants may encounter difficulties in accurately detecting expressions of remorse in others (Duncan 2002), or they may carry systematic biases that affect their interpretations of a defendant’s emotional displays which, in turn, may translate into unequal sentencing. Alternatively, even in situations where a fair and accurate perception of remorse is achieved across different groups, unequal returns for displaying remorse may still arise if these emotions have varying impacts on how defendants are evaluated in terms of criminality. These evaluations can subsequently wield a significant influence over participants’ sentencing decisions. In other words, affect-based identity assessments (Heise 1979, 2007) may emerge as a mechanism by which identical displays of remorse result in sentencing disparities.

Participants

Participants were recruited via Amazon’s Mechanical Turk (MTurk) in the fall of 2018. To be eligible for this study, participants needed to be residents of the United States. Compared to traditional laboratory samples comprised mainly of undergraduate students, the MTurk data pool is substantially more diverse (see Weinberg, Freese, and McElhattan 2014). Additionally, MTurk workers invest considerably more effort and attention when participating in survey experiments (Anson 2018). ⁴ Furthermore, research consistently demonstrates that MTurk samples yield reliable and high-quality participant pools for studies in social psychology (Manago, Mize, and Doan 2021; Paolacci and Chandler 2014). A total of 1,155 respondents completed the study in exchange for a $1.50 flat fee, with each condition comprising 103 to 125 participants. We excluded responses from 37 participants who failed manipulation checks regarding offenders’ race and gender (3.2 percent) and three participants who had missing values on their demographic information. ⁵

Table 1 reports respondent characteristics for our sample. Of the 1,115 participants included in our final analyses, 42 percent identified as women, and 75 percent identified as White. Although conducting separate analyses by respondent race/ethnicity is critical, the number of Black respondents in our data set is too small to estimate effects (N = 95). Compared with the U.S. population, our sample is somewhat more educated, more liberal, and less religious. ⁶ Approximately 55 percent were college educated, 53 percent identify as liberal, and 47 percent report having no religious beliefs. ⁷ We had slightly fewer women in the no-remorse-displayed conditions than what we would expect due to chance. ⁸ But all other demographic covariates appear well balanced across groups, suggesting the randomization process was successful.

OPEN IN VIEWER

Table 1.

Respondent Characteristics (N = 1,118)

	Proportion or Mean (SD)
Women (reference = man)	.42
Race
White	.75
Black	.08
Age	36 (10.83)
Education: % with BA	.55
Marital status
Married	.40
Never married	.42
Living with a partner	.11
Divorced/separated	.07
Employment status
Working full-time	.63
Working part-time	.10
Self-employed	.16
Full-time student	.02
Political ideology
Conservative	.30
Moderate	.16
Liberal	.53
Religion
Catholic	.23
Protestant	.21
Jewish	.02
None	.47
Income (< $50,000)	.56
Region
Northeast	.22
Midwest	.18
South	.39
West	.22

Procedure

Participants were told that researchers are interested in understanding the decision-making process in a legal case. Upon consent, each respondent read a version of the vignette: a transcript from a (fictional) videotaped confession of a college student convicted of DUI after causing injury to another driver while driving drunk. Participants were then presented with a series of questions gauging their impressions of the offender. They were next asked to provide sentencing recommendations. Finally, participants completed their demographic information and responded to two open-ended questions about their decision-making processes.

Measures

Manipulation Check

Table 2 provides descriptive statistics for all variables across the study conditions. We began our analysis by assessing the effectiveness of our emotion manipulation. In line with our qualitative findings, t tests comparing the levels of perceived emotional distress show that participants in the remorse conditions rated the offenders as significantly sorrier, sadder, more concerned, and more remorseful compared to participants in the nonremorseful conditions (t scores range from –24.60 to –34.09, all ps < .001).

OPEN IN VIEWER

Table 2.

Descriptive Statistics for Study Variables by Study Conditions

	Experimental groups								Information-vague groups
	No remorse				Remorse				No remorse	Remorse
	Black		White		Black		White
	Men	Women	Men	Women	Men	Women	Men	Women
Emotional distress	5.33 (2.37)	4.49 (2.49)	4.81 (2.35)	4.50 (2.29)	8.92 (1.91)	9.02 (1.86)	9.00 (1.86)	8.63 (2.00)	4.91 (2.12)	8.83 (1.66)
Criminality evaluation	5.98 (2.12)	6.23 (2.14)	6.26 (1.89)	6.04 (1.97)	4.98 (1.86)	4.97 (1.89)	4.89 (1.75)	5.20 (2.07)	6.48 (2.05)	5.49 (1.88)
Jail (1 = yes, 0 = no)	.57 (.50)	.60 (.49)	.55 (.50)	.65 (.49)	.52 (.50)	.45 (.50)	.45 (.50)	.47 (.50)	.59 (.49)	.60 (.49)
Jail length	4.21 (2.34)	4.66 (2.42)	4.66 (2.38	4.76 (2.44)	4.44 (2.13)	4.04 (2.35)	4.00 (2.26)	3.77 (2.27)	4.51 (2.36)	4.34 (2.39)
N	111	105	125	107	111	111	108	118	103	118

Note: the values in the table represent means and standard deviations.

Effects of Showing Remorse on Sentencing Length

We ran a series of regression models predicting the interactive effects of offenders’ race, gender, and displays of remorse on sentencing outcomes while controlling for participants’ demographic characteristics. We employed logistic regression for the binary outcome (1 = jail time) and OLS regression for the continuous outcome (i.e., mandatory jail length in days). The decision for incarceration effectively captures the rate at which offenders received jail sentences instead of noncustodial punishments (jail rate), whereas the mandatory jail length measures the severity of incarceration (jail length). Because interactions between experimental factors can be challenging to interpret, we followed Mize (2019) and plotted the marginal effect (ME) of showing remorse across five groups in Figure 1. ¹²

OPEN IN VIEWER

Figure 1.

Marginal Effect of Showing Remorse on Sentencing Outcomes by Study Groups: (a) Sentencing Outcome: 1 = Jail Time, 0 = No Jail Time. (b) Sentencing Outcome: Mandatory Jail Length (in Days)

Figure 1 presents the estimated marginal effect of showing remorse, indicating the change in the outcome when an offender goes from being perceived as relaxed to being perceived as remorseful while holding all other variables constant. Figure 1a demonstrates the effects of displaying remorse on the jail rate, and Figure 1b shows the impact of remorse on jail length.

First, when mock jurors were not exposed to the offender’s demographic information (i.e., “generic” offenders), displaying remorse had no effect on jurors’ sentence rulings (ME_{jail rate} = .02, SE = .06; ME_{jail length} = –4.03, SE = 9.23; ps > .10). It appears that when race and gender are not primed, jurors make similar sentencing decisions regardless of the emotions displayed. Although not hypothesized, these null results are intriguing given the well-established empirical pattern that remorse plays a significant role in sentence rulings (Bandes 2016; Weisman 2014). We revisit this point in the discussion section.

Moving to our experimental groups, once information about the offenders’ race and gender became explicit and salient, perceived emotional displays substantially altered sentencing outcomes. Notably, when remorse was displayed, the incarceration rate (Figure 1a) significantly decreased among White men (ME = –.11, SE = .06), White women (ME = –.21, SE = .06), and Black women (ME = –.16, SE = .06; all ps < .01). Showing remorse also significantly reduces suggested jail time (Figure 1b) for White men (ME = –20, SE = 8.92) and White women (ME = –31, SE = 9.61; all ps < .05). For Black women, displaying emotional distress shortened mandatory prison time, although the effect is only marginally significant (ME = –16, SE = 9.41; p = .08). Importantly, in neither outcome measure did Black men receive sentencing leniency for being remorseful (ME_{jail rate} = –.05, ME_{jail length} = 6; ps > .10). Not only is the marginal effect of remorse significantly smaller for Black men, but it is also statistically insignificant.

Figure 2 further illustrates our results by showing the predicted values of sentence outcome for each group. Consistent with our earlier findings, the sentencing gaps between the remorse and no-remorse conditions are smallest for Black men and largest for White women. Upon closer examination, it becomes evident that White women’s advantages in receiving leniency are primarily driven by the harsher sentences they face when displaying no remorse. Although participants sentenced remorseful White men and women at a similar rate, the penalty for violating emotional norms is much greater for White women. Of consequence, though, is a greater sentence reduction when White women conform to the norms of displaying remorse. In the case of Black women, the outcomes are contingent: They tend to face less backlash when breaching emotional norms (Figure 2a) while also receiving fewer benefits when displaying emotions deemed appropriate (Figure 2b).

OPEN IN VIEWER

Figure 2.

Predicted Jail Rate and Length by Study Groups: (a) Sentencing Outcome: 1 = Jail Time, 0 = No Jail Time. (b) Sentencing Outcome: Mandatory Jail Length (in Days)

To summarize, when offenders’ race and gender are ambiguous and implicit, respondents are able to keep their racial and gender stereotypes at bay, punishing offenders for their crime rather than for the emotions they displayed. The display of remorse, however, substantially altered sentencing outcomes when participants are primed with offenders’ demographic characteristics. Everything else being equal, participants reward remorseful White women and men and to a lesser extent, Black women by prescribing punishment that would keep them out of prison and suggesting shorter periods behind bars. In support of Hypothesis 2b (cultural binds) and in rejection of Hypothesis 2a (counter-stereotype benefits), we find Black men received the least payoffs for complying with emotional norms. White women received the largest sentence reduction for showing remorse, largely due to the higher social sanctions imposed on nonremorseful White women (Hypothesis 1). Regarding Black women, they received slightly less sentence reduction compared to White women, although the difference is small (Hypothesis 3).

Why is it that showing remorse affects only some offenders but not others? To further advance our inquiry into the “how” questions, we now turn to two explanations offered in the literature, bias in emotion readings and bias in identity assessment, that may contribute to disparities in sentencing.

Bias in Reading Emotions

First, given the difficulties of discerning remorse (Duncan 2002), unequal rewards for expressing remorse can be driven by participants’ inability to detect emotions in others. Specifically, if people fail to recognize remorse in Black men, they may make sentencing decisions as if the offenders are not sorry. To address this concern, we performed analysis of variance tests with perceived levels of distress as the dependent variable and the race and gender of the offenders as the independent variables. ¹³ In the no-remorse condition, we find no racial differences in the perceived levels of distress (F _race = 1.29; F _{gender × race} = 1.36; all ps > .05), although respondents viewed men as slightly more distressed than women (F _gender = 6.41; p < .05). In the remorseful conditions, neither gender nor race mattered. When participants read confessions embedded with remorse cues, they rated the same levels of emotional distress irrespective of the offenders’ race and gender (F _gender = .57; F _race = .70; F _{race × gender} = 1.69; all ps > .05).

Alternatively, jurors may hold a cynical view when it comes to Black male offenders, suspecting that they are feigning remorse with the intention of securing lighter sentences (i.e., “fake it to get out”). This skepticism could, in turn, reduce Black men’s likelihood of receiving sentence leniency. To investigate this possibility, we compared the perceived levels of emotional sincerity across experimental groups. Our findings revealed no discernible difference in perceived emotional authenticity: Respondents viewed emotions displayed by all offenders as equally sincere (F _gender = 1.90; F _race = 3.70; F _{race × gender} = 1.86; all ps > .05).

Finally, emotion scholars suggest the existence of an ingroup advantage in emotion recognition (Elfenbein and Ambady 2002). To check for potential ingroup bias, we examine whether respondents rate co-race or co-gender offenders as more remorseful. ¹⁴ Again, we found no evidence to suggest a systematic ingroup favoritism. A White respondent looking at a White defendant read their expression similarly to when paired with a Black defendant. The same goes for Black and women participants, where no ingroup bias was detected.

As an added precaution, we performed parallel analyses on all study variables to assess the degree of consistency in ratings within and across demographic groups. ¹⁵ Our results demonstrate striking consistency in participants’ ratings. The high level of agreement increases our confidence in conducting experimental studies with a diverse MTurk participant pool. More importantly, it suggests that the phenomena being investigated are likely to be cultural rather than political or attidudinal.

To recap, respondents in our sample were accurate and consistent in their readings of remorse. They also viewed remorseful expressions from Black men as genuine as those from other groups of offenders. By ruling out biases in perceptions of remorse, we can now turn our attention to an identity-based mechanism, where emotions and decision-making are connected through character assessment.

Bias in Identity Assessment Process

According to affect control theory (Heise 1979, 2007), emotions serve as signals to others about an individual’s identity in a given situation. When individuals engage in negative behaviors, their identities may be tarnished, but they can restore their identities by expressing appropriate emotions that align with observers’ expectations. To investigate this process, we perform a two-step regression analysis where we first examine the effect of displaying remorse on the evaluation of criminality; we then explore the interactive effects of criminality and the offenders’ race and gender on sentencing outcomes. ¹⁶ We present our findings from the first step in Figure 3. ¹⁷ For the second step, only race emerged as a significant factor interacting with criminality, Thus, we plotted (a) the estimated jail probability and (b) the suggested mandatory jail length by the race of the offenders in Figure 4, with 95 percent confidence intervals over a range of criminality.

OPEN IN VIEWER

Figure 3.

Marginal Effect of Showing Remorse on Criminality Evaluation by Study Groups

OPEN IN VIEWER

Figure 4.

Predicted Sentencing Outcomes over a Range of Criminality Evaluation by Offenders’ Race: (a) Sentence Outcome: 1 = Jail Time, 0 = No Jail Time. (b) Sentence Outcome: Mandatory Jail Length

Figure 3 plots the marginal effects of displaying remorse on perceptions of criminality. As shown in Figure 3, displaying remorse significantly lowered the evaluation of criminality for all offenders (i.e., all marginal effects are negative). Consistent with prior studies (Robinson et al. 1994; Zhao and Rogalin 2017), participants evaluated remorseful offenders more favorably than nonremorseful offenders, expecting them to have no prior criminal history, a reduced likelihood of reoffending, and to be a lawful citizen. Notably, we find no significant group differences in the magnitude of the effect of remorse on criminality evaluation (χ² range = .00–1.60; ps > .10). In the eyes of observers, all offenders seemed to restore their tarnished identity by displaying remorse.

And yet, not everyone benefits equally from a positive evaluation of their identities. The racial disparity is evident in Figure 4, particularly among offenders with low (perceived) criminal propensity (score 3 and below). Respondents assigned significantly harsher sentences to a low crime-prone Black offender compared to a White offender with an identical criminality evaluation. The curvilinear shape suggests that when jurors adjusted their criminal assessment from high (9) to low (2), a White offender’s sentence decreases substantially, from 130 days to 50 days. The same adjustment, however, has minimal impact for Black offenders, with sentences changing from 110 days to 90 days. In other words, a Black person viewed as neither dangerous nor criminal received a punishment on par with highly threatening individuals. Although our data preclude us making definitive conclusions about the factors influencing participants’ sentencing decisions regarding remorseful (and therefore nonthreatening) Black men, the stark disparity highlights a painful truth in our findings: Emotional cues from Black men and women are often discounted, and leniency is predominantly extended to remorseful White men and women.