Introduction

People with high-status non-verbal cues like a dominant face, a trim body, and light skin, often are endowed with more privileges, receiving more attention, trust, and deference in social interactions with others (Freeman et al., 2011; Hu et al., 2018; Todorov et al., 2015). Even mere exposure to these cues without face-to-face interactions can influence people’s important decisions about, for example, who to elect as political leaders (Todorov et al., 2015), and who to employ or promote in organizations (Levine & Schweitzer, 2015). Previous research often examines status signals as endogenous physical appearance cues that are difficult to change (e.g., facial appearance, body shape, skin color; Freeman et al., 2011; Hu et al., 2018; Todorov et al., 2015). Less is known about how exogenous status cues (e.g., the clothes one wears, or the car one drives) influence moral judgments, which are important aspects of social norms and social learning (Cialdini et al., 1991; Janoff-Bulman et al., 2009). For example, formal (versus casual) attire is publicly visible, easy to implement, and has some generalizability across cultures (e.g., dress code). It is ubiquitous that people employ formal attire to manipulate interpersonal impressions and strengthen their persuasiveness (Kraus & Keltner, 2009; Kraus & Mendes, 2014; Maran et al., 2021). However, the moral implications of such attire-based impression management tactics are less understood. Therefore, the current research focuses on attire as a cue of social status and examines its intertwined relationship with different types of transgressions in influencing people’s moral judgments.

Under many circumstances, it can be ambiguous whether certain behaviors are right or wrong, and whether they are enacted intentionally or accidentally. Then, people may rely on a target person’s moral claims to infer their subjective moral standards and intentionality (Jordan et al., 2017; Teeny et al., 2023). For example, some morally questionable acts (e.g., un-environmental behaviors) may be viewed as more blameworthy and intentional when the target person also advocates environmental values. Here, we focus on hypocrisy – in particular, a transgression that is accompanied by a contradictory moral claim (e.g., Effron et al., 2018; Laurent & Clark, 2019) – as a unique form of transgression that influences attire-based moral judgments. Below we detail our line of reasoning for two main assumptions: (1) People confer trust and moral credentials to formal (versus casual) attire transgressors with ambiguous intentionality, and (2) the moral favoritism drawn from formal attire is less pronounced when hypocrisy signal intentionality underlying the transgressions.

Status cues and moral favoritism: The case of attire

Many previous studies have shown that information about targets’ status can influence observers’ moral judgments of their transgressions. Some research shows that people judge and punish high- (versus low-) status transgressors more harshly (Fragale et al., 2009; Weiner & Laurent, 2020). Such moral harshness only emerges, however, when the high status is built on dominance (Kakkar et al., 2019) and the transgressions are relatively severe (Karelaia & Keck, 2013). Status can be established through either dominance and coercive gestures, or competence and conferred prestige (Cheng et al., 2013). When high- (versus low-) status holders commit transgressions, people punish dominant ones more harshly, but are willing to give moral credentials to prestigious ones (Kakkar et al., 2019). People consider status holders’ high-severity deviance (e.g., over-reporting travel expenses) as betrayal of expectations and thus punish them more harshly; for low-severity deviance (e.g., being late for meetings), instead, people give them more credentials and judge their misdeeds as more acceptable (Karelaia & Keck, 2013). In the current work, we examine low-severity transgressions (e.g., un-environmental behaviors), for which behavioral intentionality is often ambiguous, and high-status transgressors can receive moral leniency (Karelaia & Keck, 2013; Polman et al., 2013).

Minimal cues can influence people’s status impressions and downstream moral judgments. Previous studies show that formal (versus casual) attire significantly influences competence impressions (Behling & Williams, 2016; Maran et al., 2021), which persist regardless of how long observers are exposed to the targets’ pictures (from 129 millisecond to an unlimited time; Oh et al., 2019). As people give moral credentials to competent status holders and judge their transgressions more leniently (Kakkar et al., 2019), we reason that formal rather than casual attire would induce moral favoritism toward transgressors. Moral favoritism toward formal- (versus casual-) attire targets may not only manifest in people’s judgment of their transgressive behaviors but also their moral characters. Both aspects of moral judgments can have important downstream implications on interpersonal relationships (e.g., third-party observers’ trust of, and interaction with the targets; Abele et al., 2020; Malle et al., 2014; Todorov et al., 2015). Research on impression formation suggests that targets who seem high on status and/or competence are often deemed as high on morality (Bai et al., 2019) and likely to conform to prescriptive moral norms (e.g., being trustworthy; Bai et al., 2019; Fragale et al., 2011; Janoff-Bulman et al., 2009; Stellar & Willer, 2018). We therefore hypothesize a main effect of attire on moral judgments, such that:

H1: People would judge targets in formal (versus casual) attire more favorably following identical transgressions.

Hypocrisy and intentionality

While the above reasoning assumes intentionality to be ambiguous, it stands to reason that attire may influence moral judgments to a lesser extent when social cues suggest that the transgressions might not have been accidental. Such cues may include acts of moral hypocrisy, where the transgressive behaviors are preceded by contradictory moral claims. When targets “say one thing while doing another”, such contradictions often incur perceptions of hypocrisy (Effron et al., 2018; Jordan et al., 2017; Laurent & Clark, 2019). Unless specified otherwise, in the current work, we use the terms of “hypocrisy” or “hypocritical transgressions” to represent transgressions that are accompanied by contradictory moral claims, and the terms of “non-hypocrisy” or “non-hypocritical transgressions” to represent mere transgressions that are not accompanied by contradictory moral claims. We acknowledge that hypocrisy may be related to a broader array of phenomenon beyond the scope of our research. For example, people may be accused of hypocrisy, not only in transgressive contexts but also non-moral contexts (Teeny et al., 2023). Also, it can emerge from not only contradictions between claims and behaviors, but also between earlier and later claims, or earlier and later behaviors (Laurent & Clark, 2019).

We reason that hypocrisy may influence attire-based moral judgments through both perceived severity and intentionality. First, hypocrisy may be seen as more severe transgressions than non-hypocrisy (Dong et al., 2022; Jordan et al., 2017). In addition to committing a transgressive behavior, hypocrisy also violates the social norms and social expectations of being honest and sincere. People may perceive hypocrites as deceptive and manipulative: If their transgressions remain unnoticed, hypocrites can receive undeserved moral credit and be wrongly seen as trustworthy persons given their misleading moral stances (Effron et al., 2018; Jordan et al., 2017). Previous studies established that high-status transgressor received moral leniency for low- but not high-severity transgressions (Karelaia & Keck, 2013; Polman et al., 2013); it is then reasonable to infer that formal (versus casual) attire induces moral favoritism to a lesser extent for hypocrites. We therefore hypothesize an interaction effect of attire and hypocrisy on moral judgments, such that:

H2: Hypocrisy would reduce people’s moral favoritism toward transgressors in formal (versus casual) attire.

Second, people often perceive intentional wrongdoings as worse than unintentional ones, even when they cause identical objective harm (Ames & Fiske, 2013; Monroe & Malle, 2017). Increased perception of intentionality may also attenuate favorable moral judgments based on attire. When people evaluate low-severity transgressions and receive little information about intentionality, they are likely to believe high- (versus low-) status transgressors are more trustworthy persons and less intentional in their wrongdoings (as reasoned in H1). However, in the context of hypocrisy, a target person’s moral claims can signal their subjective awareness of right and wrong (Jordan et al., 2017; Laurent & Clark, 2019), and leave less wiggle room for observers to interpret their transgressions as unintentional. As compared to non-moral claims, moral claims often convey stronger attitude strength: When a target person makes moral (versus non-moral) claims, people form expectations that their attitudes should be hard to change and their behaviors should be consistent with such claims (Brannon et al., 2017; Dong et al., 2021; Teeny et al., 2023). Hypocritical transgressions can then seem more intentional and induce harsher moral judgments especially when being interpreted as selfish (Malle et al., 2014; Monroe & Malle, 2017). If the inconsistency between words and deeds are not directly questioned, hypocrites have a chance of gaining undeserved trust by preaching moral values (Jordan et al., 2017). They can then exploit such trust to serve self-interest or deter punishment for transgressions (Dong et al., 2022; Jordan et al., 2017; Lönnqvist et al., 2015). The negative evaluations of hypocrisy can be especially true for high- (versus low-) status targets, who have more self-interest at stake and may use hypocrisy as a strategy to maintain their status and reputation (Dong et al., 2021, 2022). We therefore hypothesize a mediating role of intentionality in the effect of attire and hypocrisy on moral judgments, such that:

H3: Intentionality would mediate the effect of hypocrisy on attire-based moral favoritism.

The present research

Four studies examined how formal versus casual attire influences moral judgments for hypocritical versus non-hypocritical transgressions. We propose an interaction effect between attire and hypocrisy, such that people would evaluate targets in formal than casual attire more favorably (H1), while such favoritism would be diminished when the targets transgress hypocritically (versus non-hypocritically; H2). Furthermore, perceived intentionality would partially account for the effect of hypocrisy on attire-based moral favoritism (H3).

Across all studies, we consistently adopted the attire stimuli from Oh et al. (2019) for three main reasons. First, their study provided validated competence ratings of targets in both formal and casual attire, enabling us to select the targets who were both representative of different ethnic groups and revealed a large competence difference in their formal- versus casual- attire ratings. Second, the pictures depicted real persons who do not only have attire but also hair and facial features, which may help reduce demand characteristics of stimuli featuring only attire difference (Maran et al., 2021). Third, formal versus casual attire was manipulated in a broad sense (e.g., wearing a suit versus a jacket, or a tie versus not), which can apply across various occupations and reduce the impact of other confounding factors (e.g., occupation-based attire manipulation; Ratcliff et al., 2011).

Study 1 first established people’s moral favoritism toward transgressors in formal (versus casual) attire (H1), while Studies 2 to 4 further examined the diminished moral favoritism following hypocrisy (versus non-hypocrisy; H2). Moreover, to investigate the presumed mechanism of intentionality (H3), we measured perceived intentionality in Study 2 and manipulated hypocritical transgressions as intentional versus unintentional in Study 4.

As indicators of moral judgments, we examined both moral blame of transgressive behaviors (Studies 1, 2 and 4) and moral impression of the transgressors (Studies 2, 3 and 4). Whereas earlier research examined moral impression more broadly as good or bad intent (also termed as “warmth” or “communion”; for a review, see Abele et al., 2020), we follow more recent recommendations to differentiate morality from sociability (as sub-dimensions of warmth impression; Goodwin et al., 2014; Landy et al., 2016), and to further divide morality into what people ought to be (i.e., trustworthiness) and what people ideally can be (i.e., compassion; Cornwell & Higgins, 2015; Landy & Uhlmann, 2018). Put differently, we measured three impression dimensions of trustworthiness, compassion, and sociability, as a function of attire manipulation. Furthermore, the attire and hypocrisy manipulations may have a stronger influence on trustworthiness than other two impression dimensions. Status and competence perceptions are more related to the “self-control” component of morality (Fragale et al., 2011; Stellar & Willer, 2018), and hypocrisy is essentially seen as deceptive and manipulative (Effron et al., 2018; Jordan et al., 2017). Both concepts are closely related to perceptions of trustworthiness.

We conducted a-priori power analyses with G*Power (Version 3.1; Faul et al., 2009) in Studies 1, 3, and 4, and relied on a methodological paper (Pan et al., 2018) to determine our sample size in Study 2. Across the studies, the main statistical analyses and visualization were performed in R, with R packages “tidyverse” (Wickham et al., 2019) “rstatix” (Version 0.6.0; Kassambara, 2020) “psych” (Version 2.0.9; Revelle, 2020) “ggstatsplot” (Patil, 2018) and “effectsize” (Ben-Shachar et al., 2020). All measures and manipulations were disclosed in the respective studies. We did not exclude participants from further analysis if they completed the whole study. We pre-registered the study design, planned sample size, inclusion/exclusion criteria, and planned analyses of Studies 1 (at https://tinyurl.com/4vkzs55e) and 4 (at https://tinyurl.com/2p2hb2kx) on the Open Science Framework. All pre-registered analyses were reported with no deviations. The data, analyses, codebook, and experimental materials have been uploaded on the Open Science Framework and can be accessed at https://tinyurl.com/5ytcrmu3.

Study 1: Attire, non-hypocrisy, and perceived intentionality

Study 1 focused on non-hypocritical transgressions. To test H1, namely, the moral favoritism effect based on formal attire, participants read transgressive descriptions accompanied with transgressor pictures, either in formal or casual attire, and then answered questions about moral blame and intentionality. We predicted that people would blame formal- (versus casual-) attire transgressors less and also perceive them as less intentional.

Method

Participants

We recruited 120 American participants (53 males; Mage = 33.2 years, SD = 11.7) from Prolific. An a-priori power analysis suggested a minimum sample of N = 102 to detect a medium-size attire effect on moral judgments (t-test; d = 0.5; α = 0.05; with 80% power). We estimated an effect size d = 0.5 based on a comprehensive review of different social psychology research topics. The review on topics related to “attitudes” suggested an average r = .27 (i.e., an equivalent d = 0.56) based on a summary of 2,476 studies (Richard et al., 2003).

Design and procedure

Study 1 selected six target persons from Oh et al. (Oh et al., 2019; see Supplementary Materials for specific stimuli). Participants were randomly assigned to either a formal-attire or a casual-attire condition (each n = 60). Both conditions presented the six targets, each matched with a unique transgression. The transgressions were selected from the work of Monroe and Malle (2017). Accordingly, the six selected transgressions had moderate ratings of intentionality and provided ambiguity for subjective interpretations.

We randomized the sequence of the six targets in each attire condition. Participants first saw a target’s picture and evaluated their competence on three items (“competent/capable/efficient”; α = 0.95 for the 18 items across the six targets; rated on a 9-point scale ranging from 0 = Not at all to 8 = Extremely; selected from previous work, e.g., Goodwin et al., 2014; Landy et al., 2016). They then read about the target’s transgressive behavior (e.g., “gave a customer incorrect change”), and indicated moral blame of the behavior on six items (“condemnable/immoral/unethical/honorable/acceptable/tolerable”; on a 9-point scale from 0 = Not at all to 8 = Extremely; adapted from Effron & Monin, 2010). After reversely coding the last three items, the moral blame measures yielded an α = 0.95 for the 36 items across the six targets. As an exploratory measure, participants also rated the intentionality of the transgression on one item (α = 0.56 across the six targets; on a 9-point scale from − 4 = Definitely unintentional to 4 = Definitely intentional). Participants’ responses regarding the six targets and specific items were averaged.

Results

Manipulation check

An independent sample t-test showed that targets in formal attire (M = 4.90, SD = 1.09) were seen as more competent than those in casual attire (M = 4.43, SD = 1.15), t(118) = 2.30, p = .023, d = 0.42, 95% CI [0.06, 0.79]. Thus, the manipulation of attire successfully varied competence perception, as intended.

Moral blame and perceived intentionality

As pre-registered and shown in Fig. 1, an independent sample t-test showed that formal-attire transgressors were blamed less than their casual-attire counterparts, t(113) = 2.04, p = .043, d = 0.38, 95% CI [0.01, 0.76]. For perceived intentionality, we also performed an independent sample t-test, and found that people considered formal-attire transgressors as more unintentional than casual-attire counterparts (see also Fig. 1), t(113) = 2.63, p = .009, d = 0.49, 95% CI [0.12, 0.84], even though perceived intentionality was low in both conditions.

Fig. 1
figure 1

Moral blame of transgressions (left) and intentionality perception (right) of formal-attire versus casual-attire transgressors in Study 1. When both targets committed identical transgressions with ambiguous intentionality, people blamed formal- (versus casual-) attire targets less (left), and perceived them as more unintentional (right).

Full size image

Discussion

Study 1 supported our first hypothesis (H1) and showed that people judged transgressors in formal (versus casual) attire more leniently, and also attributed less intentionality to them. These findings are in line with previous research, suggesting that people give moral credentials to high- (versus low-) status targets (Kakkar et al., 2019; Karelaia & Keck, 2013; Polman et al., 2013) for their low-severity transgressions. In our study, such credentials can be gained through simply replacing casual with formal attire, which further reduced blame for transgressions and attributed intentionality when intentionality information was missing. The moral favoritism reflected on intentionality attribution also suggested that status cues may not only influence judgments of moral severity but also underlying intention reasoning about transgressions (see also Dong et al., 2021; Fragale et al., 2009).

Study 2: Attire, hypocrisy, and the mediation of intentionality

Study 1 demonstrated people’s moral favoritism toward transgressors in formal (versus casual) attire (H1); in Studies 2 to 4, we further examined whether hypocrisy reduce this attire effect on moral favoritism (H2) and the potential mechanism of intentionality (H3).

In Study 2, we first aimed to replicate Study 1 findings. But different from Study 1 where people made independent judgments about formal- (versus casual-) attire transgressors, Study 2 asked people to evaluate formal- (versus casual-) attire transgressors in a comparative setting. This was intended to simulate situations where multiple targets commit identical transgressions and may share responsibilities (see also Gray & Wegner, 2009). Moreover, in addition to the targets’ transgressive behaviors, we presented their contradictory claims and measured how moral judgments change “before” versus “after” the presence of hypocrisy. We presume that people would initially favor the transgressor in formal (versus casual) attire (H1), but this favoritism would be diminished when both targets behave hypocritically (H2).

Study 2 examined moral favoritism regarding both moral blame of transgressions (as in Study 1) and moral impression of transgressors. We mainly focused on moral impressions as trustworthiness but kept measuring compassion and sociability for comparison. Keeping other impression dimensions may also help explain whether attire-based moral favoritism should be attributed to an overall positive impression (e.g., Bocian et al., 2018). If so, formal (versus casual) attire should induce comparable moral favoritism on the three dimensions.

Moreover, Study 2 examined our presumed mechanism of intentionality more formally with a sufficient sample size for mediation analyses. Based on our line of reasoning, people would judge casual- (versus formal-) attire transgressor as more intentional and self-interested, but less so when both targets enact hypocrisy. Inferences about intentionality would account for the changes in people’s moral judgments.

Method

Participants

We intended for a sample size that was necessary to detect the predicted within-participants mediation effects of intentionality. We referred to a published simulation work (Pan et al., 2018), which allowed us to determine our sample size based on key input parameters of (1) X→M effect size, (2) M→Y effect size, and (3) the correlation coefficients of the repeated measures for sample size determination. We anticipated a small effect of competence on the mediators (i.e., X→M; β = 0.14; as in Fragale et al., 2009), a medium-size effect of the mediators on moral judgments (i.e., M→Y; β = 0.29; Guglielmo & Malle, 2010), and a moderately strong correlation between the repeated measures (ICC = 0.5). Accordingly, the simulation yielded a required N = 398 with 80% power at an alpha level of 0.05, to examine a within-participants mediation effect with the bootstrapping method. We therefore targeted N = 400 from Prolific and included all 403 American participants (194 males; Mage = 33.6 years, SD = 11.3) in further analyses.

Design and procedure

We employed a within-participants design, measuring the interested constructs twice (except for competence at the beginning) — “before” and “after” the presence of hypocritical claims. Each time, participants first indicated their perceived intentionality and self-interest of a transgressive behavior, and then answered questions about moral blame and impressions of the two transgressors regarding the three traits (i.e., trustworthiness, compassion, and sociability). All the measures were administered on a 9-point comparative scale, where one casual-attire target (Definitely Person A = 0) and one formal-attire target (Definitely Person B = 8) represented the two ends of the scale, with the midpoint 4.0 representing a neutral judgment (i.e., Neither Person A nor Person B). Targets’ faces were counterbalanced, such that the target in formal (casual) attire in Condition 1 (n = 202) appeared in casual (formal) attire in Condition 2 (n = 201). We aggregated data from these two Conditions and averaged items within one trait in the main analyses.

We selected two targets from Study 1 who had the highest competence difference when they were in formal versus casual attire. Participants first saw the two targets — one in formal and the other in casual attire — and were asked to evaluate their relative competence on five items (“competent/capable/efficient/intelligent/skillful; α = .90) as manipulation check. Then, in the “before” condition, we introduced that both Person A and Person B bought a Range Rover vehicle for daily use, which is well known for high CO2 emission. We then measured participants’ perceived intentionality on three items (“intentional/deliberate/unintended” with the last item reversely coded; α = .76), self-interest on three items (“self-serving/egoistic/self-centered”; α = .89; both adapted from previous work, e.g., Fragale et al., 2009), and behavioral blame on six items (as in Study 1; α = .74) regarding the un-environmental purchase, and their impressions of the targets. The impression measures comprised three traits, trustworthiness (“trustworthy/principled/honest/loyal”, α = .82), compassion (“compassionate/caring/helpful/empathic”, α = .91), and sociability (“sociable/extroverted/playful/friendly”, α = .87), each with four items. In the “after” condition, we then added information about the targets’ hypocrisy by showing that both targets work for an environmental organization, which “convinces people to reduce their CO2 emission by taking public transport to work”. We checked our manipulation of hypocrisy (“How inconsistent or consistent is their behavior versus objective at work?” on a 9-point scale ranging from − 4 = Extremely inconsistent to 4 = Extremely consistent), and again asked participants to compare the targets on intentionality (α = .75), self-interest (α = .87), behavioral blame (α = .68), and moral/warmth impressions (trustworthiness, α = .85; compassion, α = .90; sociability, α = .86) in the listed order and based on identical items in the “before” condition.

Results

Manipulation check

In a one-sample t-test, participants deemed the formal-attire targets as more competent than the casual-attire ones (M = 5.02, SD = 1.43), t(402) = 14.29, p < .001, d = 0.71, 95% CI [0.60, 0.82], as compared with the scale midpoint of 4.0. In another one-sample t-test, participants perceived the targets’ behavior as strongly inconsistent with their goal at work (M = -2.88, SD = 2.00), t(402) = -28.92, p < .001, d = 1.44, 95% CI [1.30, 1.58], as compared to the midpoint of 0.

Moral blame

Both compared with the scale midpoint of 4.0 in one-sample t-tests, people blamed the casual- (versus formal-) attire target more, only “before”, t(402) = -3.10, p = .002, d = 0.15, 95% CI [0.06, 0.25], but not “after”, knowing the two targets’ hypocrisy, t(402) = 1.63, p = .104, d = 0.08, 95% CI [-0.02, 0.18]. Through another paired-sample t-test, we found that moral blame in the “before” and “after” conditions also differed significantly, r = .42, t(402) = 4.47, p < .001, d = 0.22, 95% CI [0.12, 0.32], meaning an increased blame of formal- (relative to casual-) attire targets’ un-environmental behavior (see Fig. 2).

Fig. 2
figure 2

Comparative behavior blame (upper left), trustworthiness impression (upper middle), compassion impression (upper right), and perceptions of intentionality (lower left) and self-interest (lower right), as a function of hypocrisy in Study 2. After (versus before) seeing signs of hypocrisy, people blamed formal- (relative to casual-) attire transgressors more, and perceived them as less trustworthy and compassionate. As potential mechanisms, signs of hypocrisy increased perceived intentionality and self-interest. Except for moral blame after hypocrisy (upper left), other scores significantly differed from the neutral judgment (ps < 0.002), meaning that despite the decrease, moral favoritism toward formal- (relative to casual-) attire transgressors persisted after hypocrisy.

Full size image

Moral impression

We first employed repeated-measures ANOVA to examine how hypocrisy influenced impressions on different dimensions. We fitted hypocrisy and traits (i.e., trustworthiness, compassion, and sociability) as independent variables and impression scores as the dependent variable, and found an interaction effect emerged between hypocrisy and traits, F(2, 401) = 14.12, p < .001, \({\eta p}^{2}\)= 0.07, 95% CI [0.03, 0.12]. As shown in Fig. 2, on both trustworthiness (r = .79), F(1, 402) = 28.73, p < .001, \({\eta p}^{2}\)= 0.07, 95% CI [0.03, 0.12], and compassion (r = .64), F(1, 402) = 32.99, p < .001, \({\eta p}^{2}\)= 0.08, 95% CI [0.03, 0.13], but not on sociability (r = .49), F(1, 402) = 0.08, p = .775, \({\eta p}^{2}\) < 0.001, people significantly decreased their favorable impressions of the formal- (versus casual-) attire transgressor “after” (versus “before”) seeing both targets’ signs of hypocrisy. We then conducted multiple one-sample t-tests, to compare impression scores (of different traits, both before and after hypocrisy) with the scale midpoint 4.0. We found that regardless of hypocrisy, people generally favored the formal- (relative to casual-) attire target on impressions of trustworthiness and compassion, t(402) > 6.62, ps < 0.001, d > 0.33, but not sociability, t(402) < 1.16, ps > 0.250, d < 0.06. These findings suggested that moral hypocrisy attenuated but did not fully eliminate the attire effect on moral favoritism.

Mediation of intentionality and self-interest

As in behavior blame and moral impressions, paired-sample t-tests showed that people perceived the transgression as more intentional, t(402) = 5.48, p < .001, d = 0.27, 95% CI [0.17, 0.37], and self-interested, t(402) = 5.78, p < .001, d = 0.29, 95% CI [0.19, 0.39], when the formal- (relative to casual-) attire target transgressed hypocritically versus not (see also Fig. 2). Despite so, one-sample t-tests showed that people generally perceived the formal- (relative to casual-) attire target as less intentional and self-interested, all as compared to the scale midpoint 4.0, t(402) > 9.22, ps < 0.001, d < 0.46.

We then investigated perceived intentionality and self-interest simultaneously as mediators in moral judgments “before” versus “after” hypocrisy. We adopted the MEMORE SPSS macro, which allowed us to fit models with “multiple mediators operating in parallel and serially” and “discuss the comparison of indirect effects in these more complex models” (Montoya & Hayes, 2017). This way, in the effect of hypocrisy on moral judgments, our model allowed multiple mediations, with intentionality and self-interest as independent parallel mediators and the serial mediation of intentionality → self-interest. With 5,000 bootstrapping resampling (see the specific indirect effects in Table 1), we found a significant mediation effect of intentionality for trustworthiness impression. Moreover, across (1) behavioral blame, (2) impression on trustworthiness, (3) impression on compassion, and (4) the aggregated overall appraisal, our statistical tests showed significant serial indirect paths of hypocrisy → intentionality → self-interest → moral judgments.

Table 1 Mediation analyses in Study 2.
Full size table

Discussion

As in Study 1 and supporting our first hypothesis (H1), people favored transgressors in formal (versus casual) attire. Study 2 also supported our second hypotheses (H2), showing people’s reduced favoritism (regarding blame, trustworthiness, and compassion) toward formal- (versus casual-) attire transgressors after both targets enact hypocrisy. These effects were not only manifested on moral blame of transgressions but also moral impression of transgressors (Landy & Uhlmann, 2018). Consistent with previous studies on the multi-dimensional nature of moral impressions (Goodwin et al., 2014; Landy et al., 2016; Stellar & Willer, 2018), we found differential attire effects on different impression dimensions, which could not be attributed to an overall positive/negative impression and was mainly reflected on trustworthiness and compassion but not on sociability.

The attire-based moral favoritism before showing hypocrisy and reduced moral favoritism after showing hypocrisy also manifested in attributions of intentionality and self-interest (H3), such that people perceived formal- (versus casual-) attire transgressors as more intentional and self-interested after (versus before) hypocrisy. These findings were consistent with our reasoning that hypocrisy increases perceived severity and intentionality, and thus diminishes attire-based moral favoritism. However, in this study, hypocrisy did not fully eliminate people’s moral favoritism toward formal- (versus casual-) attire transgressors. That is, people still perceived formal- (versus casual-) attire hypocrites as more trustworthy and compassionate, and less intentional and self-interested.

Further mediation analyses suggested that intentional and self-interested attributions accounted for a significant portion of variance in moral favoritism judgments. In contrasts between formal- (versus casual-) transgressors, the reduced moral favoritism toward formal-attire hypocrites was related to increased perception of intentional pursuits of self-interests. However, the mediation analyses were correlational and conditional on our model assumptions, which could not preclude other possible models. For example, stronger perceptions of self-interest or moral wrongness may have amplified intentional attributions (Fragale et al., 2009; Knobe, 2003).

Study 3: Attire and other forms of hypocrisy and non-hypocrisy

Study 2 replicated the attire effect on moral favoritism judgments (H1) and further showed its diminished effect after (versus before) people were informed about the transgressors’ contradictory moral claims (H2). However, it can be argued that people do not consider contradictory claims after transgressions as signs of hypocrisy but genuine attitudinal change (Barden et al., 2005). Study 3 therefore examined hypocrisy as a preceding moral claim followed by a behavior violating the claim. As in Study 2, we presume that people would evaluate formal- (versus casual-) attire targets more favorably when both make moral claims (H1), and the moral favoritism would diminish when both targets enact hypocrisy (i.e., say one thing and then do another; H2). Study 3 further explored the attire effect on moral favoritism when targets enact moral integrity (i.e., say one thing and then do as they say). Since behavioral blame did not fit with moral integrity, we only examined moral impressions in Study 3.

Method

Participants

An a-priori power analysis yielded N = 198 to achieve 80% power for a small competence by behavior interaction effect (ANOVA; \({\eta p}^{2}\)= 0.01, α = 0.05) as a within-participants factor. Hence, two hundred and one American participants (110 males; Mage = 33.6 years, SD = 10.3) were recruited on Prolific and were all included in further analyses.

Design and procedure

We employed a 2 (attire: formal versus casual) by 2 (time: “before” versus “after” the presentation of a subsequent behavior following a moral claim) by 2 (subsequent behavior: hypocrisy versus integrity) mixed design, with only subsequent behavior as a between-participants factor. In each condition, participants read about four target persons in a random sequence and evaluated them on a scale from 0 = Not at all to 8 = Extremely.

Participants were first presented with the moral claim and were told that the four persons often “SHARE articles supporting #Feminism on LinkedIn”. In this “before” condition, participants first saw the four targets — two in formal and two in casual attire — and evaluated their competence as manipulation check (α = 0.92 for 12 items, that is, 3 items for each target), moral impressions of trustworthiness (α = 0.95 for 16 items) and compassion (α = 0.96 for 16 items), and non-moral impression of sociability (α = 0.91 for 16 items) on identical items as in Study 2.

In the “after” condition, then, participants were randomly assigned to one of the two subsequent behavior conditions. In the integrity condition (n = 101), participants read that the targets also “DONATE to #Feminism causes”. In the hypocrisy condition (n = 100), the targets “LIKE articles supporting #AntiFeminism”. One manipulation check question was inserted (“How inconsistent or consistent are the causes that the users SHARE versus DONATE to/LIKE?”; on a 9-point scale from − 4 = Extremely inconsistent to 4 = Extremely consistent). We again presented the targets’ pictures and asked participants to evaluate them on the same items as in the “before” condition, in terms of trustworthiness (α = 0.99 for 16 items), compassion (α = 0.99 for 16 items), and sociability (α = 0.97 for 16 items) after knowing the targets’ subsequent deeds of hypocrisy or integrity.

Results

Manipulation checks

As expected, a paired-sample t-test showed that targets in formal attire (M = 5.02, SD = 1.28) were considered as more competent than targets in casual attire (M = 4.06, SD = 1.46), t(200) = 9.93, p < .001, d = 0.70, 95% CI [0.55, 0.86]. Also, two one-sample t-tests respectively showed that as compared to the midpoint 0, participants perceived the targets’ behavior (in)consistency in the hypocrisy (M = -1.62, SD = 2.70; t(99) = -5.99, p < .001, d = 0.60, 95% CI [0.39, 0.82]) and integrity (M = 2.72, SD = 1.45; t(100) = 18.87, p < .001, d = 1.89, 95% CI [1.56, 2.21]) conditions as intended.

Moral impression

We conducted repeated-measures ANOVA with subsequent behavior as between-participants factor while attire, time, and traits as within-participants factors. In addition to a strong main effect of subsequent behavior (integrity condition: M = 4.73, SD = 1.57; versus hypocrisy condition: M = 3.14, SD = 1.57), F(1, 199) = 102.92, p < .001, \({\eta p}^{2}\) = 0.34, 95% CI [0.24, 0.43], a main effect of attire showed that people perceived targets in formal attire more positively (M = 3.99, SD = 1.13; versus casual attire, M = 3.88, SD = 1.21), F(1, 199) = 5.12, p = .025, \({\eta p}^{2}\) = 0.03, 95% CI [< 0.001, 0.08]. A significant two-way interaction between attire and traits, F(2, 320) = 9.19, p < .001, \({\eta p}^{2}\) = 0.05, 95% CI [0.01, 0.11], suggested that attire mainly influenced impressions of trustworthiness (p < .001, \({\eta p}^{2}\) = 0.09), but not compassion (p = .269, \({\eta p}^{2}\) = 0.01) or sociability (p = .692, \({\eta p}^{2}\) = 0.001).

More importantly, we found a significant three-way interaction of attire by time by traits, F(2, 354) = 6.36, p = .002, \({\eta p}^{2}\) = 0.04, 95% CI [0.01, 0.08], but not a four-way interaction of attire by time by traits by subsequent behavior, F(2, 354) = 1.01, p = .358, \({\eta p}^{2}\) = 0.01, 95% CI [< 0.001, 0.03]. Further simple effects analysis showed that the effects of attire were attenuated both when subsequent acts of hypocrisy and integrity followed initial moral claims (i.e., in the “after” versus “before” condition). However, the above attire by time interaction effect only manifested on trustworthiness, F(1, 199) = 10.03, p = .002,\({\eta p}^{2}\)= 0.05, 95% CI [0.01, 0.12], and not on compassion, F(1, 199) = 0.33, p = .569, \({\eta p}^{2}\) = 0.002, 95% CI [< 0.001, 0.03], or sociability, F(1, 199) = 0.003, p = .957, \({\eta p}^{2}\) < 0.001, 95% CI [< 0.001, 0.01]. Specifically, as shown in Fig. 3, people evaluated targets in formal (versus casual) attire as more trustworthy, while only “before” but not “after” knowing their hypocritical transgression (“before”: F(1, 99) = 8.39, p = .005, \({\eta p}^{2}\) = 0.08, 95% CI [0.01, 0.20]; “after”: F(1, 99) = 1.65, p = .202, \({\eta p}^{2}\)= 0.02, 95% CI [< 0.001, 0.10]) or integrity behavior (“before”: F(1, 100) = 16.37, p < .001, \({\eta p}^{2}\) = 0.14, 95% CI [0.04, 0.27]; “after”: F(1, 100) = 4.26, p = .042, \({\eta p}^{2}\) = 0.04, 95% CI [< 0.001, 0.14]).

Fig. 3
figure 3

Trustworthiness impression as a function of attire and sign of integrity (upper) or hypocrisy (lower) in Study 3. After (versus before) following a preceding moral claim, people’s trustworthiness impression on formal- (versus casual-) attire actors of integrity increased to a lesser extent (upper). After (versus before) violating a preceding moral claim, people’s trustworthiness impression on formal- (versus casual-) attire transgressors decreased to a greater extent (lower).

Full size image

Discussion

As in Study 2, people evaluated targets in formal attire as more trustworthy than those in casual attire (H1), and the attire effect decreased with the presence of signs of hypocrisy (H2). In relation to Study 2, we may conclude that hypocrisy attenuates attire-based moral favoritism, regardless of whether a moral claim preceded (Study 3) or followed (Study 2) a contradictory transgression. Different from Study 2, however, these effects mainly manifested in moral impressions of trustworthiness but not compassion. This is conceptually consistent with some previous work, suggesting trustworthiness (not compassion or sociability) as the core component of moral impression (Landy & Uhlmann, 2018), and the trustworthiness implications of status cues (Fragale et al., 2011; Fragale et al., 2011; Stellar & Willer, 2018) and hypocritical behavior (Effron et al., 2018; Jordan et al., 2017).

Interestingly, formal- (versus casual-) attire targets lost their moral favoritism when their good deeds were preceded by moral claims. Put differently, whereas both formal- (versus casual-) attire targets were praised for acts of moral integrity, casual- (versus formal-) attire targets received more credits for managing to do as they say. People may also expect better word-deed consistency given the implied high competence of formal attire (Brannon et al., 2017; Dong et al., 2021). Whereas moral claims are often used as a behavioral strategy to signal virtues and impress others (Jordan et al., 2017), formal- (versus casual-) attire targets can lose their unique advantage when their good deeds are preceded by such “cheap talk” (Farrell & Rabin, 1996). Cheap talk of high- (versus low-) status targets may incur more blame for their bad deeds and less praise for their good deeds. Together, moral privilege of formal (versus causal) attire targets can be modulated by subsequent behavior and be compromised both when their good and bad deeds are preceded by moral claims. Even though we did not directly measure intentionality in Study 3, these findings are generally in line with our reasoning that moral claims signal subjective awareness of right and wrong (Jordan et al., 2017; Laurent & Clark, 2019), and therefore leave less wiggle room for attire-based favorable judgments.

Study 4: Attire, hypocrisy, and manipulated (un)intentionality

Correlational analyses in Study 2 suggested a role of perceived intentionality in the observed effects. To establish a causal effect of intentionality, Study 4 manipulated (un)intentionality underlying transgressions with contradictory moral claims. In this design, people may consider transgressions as hypocritical only when the hypocritical transgressions were intentional rather than unintentional. Based on our line of reasoning, we posit an interaction effect between attire and intentionality information, such that moral favoritism toward formal- (versus casual-) attire targets should be reduced when their seemingly hypocritical transgressions were intentional (versus unintentional).

Study 4 further explored observers’ actual perceptions of hypocrisy in different conditions. None of our previous studies provided solid evidence on that people perceived our manipulated hypocrisy as indeed hypocritical. We therefore presume that people would perceive formal- (versus casual-) attire hypocrites as more hypocritical when their transgressions are considered intentional (versus unintentional).

Method

Participants

As pre-registered, we intended 280 American participants from Prolific. The sample size was determined by an a-priori power analysis, suggesting a sample of N = 256 to detect the attire by intentionality interaction effect (ANOVA; \({\eta p}^{2}\)= 0.03 from Study 3) with 80% power at an alpha level of 0.05. We had 283 online participants (118 males; Mage = 34.6 years, SD = 11.3), who were all included in further analyses.

Design and procedure

We employed a 2 (attire: formal versus casual) by 2 (intentionality: intentional versus unintentional) between-participants design. Participants were randomly assigned to one of the four conditions and evaluated three targets in a randomized order. The three targets were again selected from Study 1 who had high competence ratings and represented different ethnic groups.

First, participants were assigned to either a formal-attire (n = 143) or a casual-attire (n = 140) condition. After knowing each target’s picture and environmental stance (e.g., “Riley often tweets pro-environmental activities, using hashtags like #Environmentalist #ClimateChange #Nature #Sustainability”), participants were asked to evaluate the target’s competence as manipulation check (as in Studies 1 and 3; α = 0.94 across 9 items, that is, 3 items for each target). Participants then read about the targets’ un-environmental behaviors (e.g., “Riley littered his snot-wiping paper”). In the intentional condition (n = 142), the un-environmental behaviors were depicted as intentional (e.g., “He did so intentionally because he found no trashcan nearby”). In the unintentional condition (n = 141), the un-environmental behaviors were described as accidental (e.g., “He did so accidentally because the paper fell out of his pocket when he took out his cellphone”).

After reading the above information, participants indicated their perceived intentionality as manipulation check (e.g., “How unintentional or intentional do you think Riley’s behavior of littering his snot-wiping paper is?” on a 9-point scale ranging from − 4 = Definitely unintentional to 4 = Definitely intentional; α = 0.88 across 3 items), moral blame of the transgression (“immoral/unethical/blameworthy”; α = 0.91 across 9 items, that is, 3 items for each target), and impressions of the transgressor as in Studies 2 and 3 (α = 0.94 for trustworthiness; α = 0.93 for compassion; α = 0.88 for sociability). We also explored how “hypocritical” participants perceived the transgressor to be in different conditions (on a 9-point scale ranging from − 4 = Not at all to 4 = Extremely; α = 0.79 for 3 items). As in Studies 1 to 3, we averaged responses regarding the three targets and items within constructs.

Results

Manipulation checks

As intended, independent sample t-tests showed that targets in formal attire (M = 5.22, SD = 1.33) were seen as more competent than those in casual attire (M = 4.88, SD = 1.31), t(282) = 2.15, p = .033, d = 0.26, 95% CI [0.02, 0.49]. Likewise, targets in the intentional condition (M = 3.19, SD = 0.87) were perceived as more intentional than those in the unintentional condition (M = -2.91, SD = 1.08), t(282) = 52.39, p < .001, d = 6.24, 95% CI [5.67, 6.80].

Moral blame

As predicted, an ANOVA (with attire and intentionality as predictors and moral blame as the dependent variable) revealed a significant interaction between attire and intentionality, F(1, 279) = 27.42, p < .001, \({\eta p}^{2}\) = 0.09, 95% CI [0.04, 0.16]. As shown in Fig. 4, further simple effects tests showed that targets in formal (versus casual) attire received more blame for intentional hypocritical transgressions, F(1, 140) = 30.80, p < .001, \({\eta p}^{2}\) = 0.18, 95% CI [0.08, 0.29], but not for their unintentional ones, F(1, 139) = 2.64, p = .107, \({\eta p}^{2}\) = 0.02, 95% CI [< 0.001, 0.08]. When transgressions with ambiguous intentionality (in Studies 1 to 3) became clearly unintentional (in Study 4), neither the formal-attire nor the casual-attire transgressors were blamed.

Fig. 4
figure 4

Moral blame (upper left), trustworthiness (upper right) and compassion impressions (lower left), as well as hypocrisy perception (lower right), as a function of attire and (un)intentionality manipulation in Study 4. For intentional (versus unintentional) hypocritical transgressions, people blamed formal- (versus casual-) attire targets more (upper left), and saw them as less trustworthy (upper right), less compassionate (lower left) but more hypocritical (lower right) persons.

Full size image

Moral impression

Another ANOVA with attire and intentionality as between-participants factors and traits as within-participants factor revealed a three-way interaction effect, F(1, 839) = 4.11, p = .043, \({\eta p}^{2}\) = 0.01, 95% CI [< 0.001, 0.02], such that the attire by interaction effect manifested in impressions on trustworthiness, F(1, 279) = 10.68, p = .001, \({\eta p}^{2}\) = 0.04, 95% CI [0.01, 0.09], and compassion, F(1, 279) = 5.37, p = .021, \({\eta p}^{2}\) = 0.02, 95% CI [< 0.001, 0.06], but not sociability, F(1, 279) = 1.29, p = .257, \({\eta p}^{2}\) = 0.01, 95% CI [< 0.001, 0.03]. As also shown in Fig. 4, people perceived formal-attire hypocritical transgressors as less trustworthy and compassionate than their casual-attire counterparts, but only when they intentionally transgressed (trustworthiness: F(1, 140) = 13.70, p < .001, \({\eta p}^{2}\) = 0.09, 95% CI [0.02, 0.19]; compassion: F(1, 140) = 7.35, p = .008, \({\eta p}^{2}\) = 0.05, 95% CI [0.004, 0.14]), and not when they unintentionally transgressed (trustworthiness: F(1, 139) = 1.06, p = .306, \({\eta p}^{2}\) = 0.01, 95% CI [< 0.001, 0.06]; compassion: F(1, 139) = 0.64, p = .425, \({\eta p}^{2}\) = 0.01, 95% CI [< 0.001, 0.05]).

Perceived hypocrisy

On perceived hypocrisy, an ANOVA with attire and intentionality as predictors yielded a significant interaction effect, F(1, 279) = 12.36, p < .001, \({\eta p}^{2}\) = 0.04, 95% CI [0.01, 0.10]. Specifically, formal- (versus casual-) attire hypocritical transgressors were seen as more hypocritical only when they transgressed intentionally, F(1, 140) = 12.36, p < .001, \({\eta p}^{2}\) = 0.08, 95% CI [0.02, 0.18], but not when they transgressed unintentionally, F(1, 139) = 2.89, p = .091, \({\eta p}^{2}\) = 0.02, 95% CI [< 0.001, 0.09]. Other ancillary results can be found in the Supplementary Materials. Across behavior blame, person impressions of trustworthiness and compassion, and perceived hypocrisy, people perceived targets in formal versus casual attire differently depending on their (un)intentionality underlying the seemingly hypocritical transgressions.

Discussion

In Study 4, we manipulated (un)intentionality information underlying hypocritical transgressions and found the expected interaction effect between attire and intentionality on moral judgments. Moreover, perceived hypocrisy increased particularly for formal- (versus casual-) attire intentional (versus unintentional) transgressors. Different from previous studies, we also found that (1) people did not favor targets in formal (versus casual) attire in the unintentional transgression condition, and (2) even perceived targets in formal (versus casual) attire more negatively in the intentional transgression condition. This latter finding suggested that hypocritical transgressions with clear information about intentionality did not only diminish (Studies 2 and 3) but even reversed (Study 4) the moral favoritism toward formal- (versus casual-) attire targets. The idea that people judged high- (vs. low) status transgressors more harshly also emerged in some previous studies, especially when people deemed the transgressions as relatively severe (Dong et al., 2022; Karelaia & Keck, 2013).

However, the results in Study 4 may not be directly comparable to those in Studies 1 to 3. These previous studies had a controlled non-hypocrisy condition without moral claims, while all the transgressions in Study 4 featured hypocrisy and moral claims. We will systematically discuss the implications of these seemingly divergent findings later in the General Discussion. Together with Studies 1 to 3, the findings consistently suggest that the attire effect on moral favoritism emerge for transgressions without contradictory claims (Studies 1 to 3) but not for transgressions with contradictory claims (Studies 1 to 4), regardless of whether the word-deed contradictions are introduced as intentional or unintentional (Study 4).

General discussion

In contemporary societies, people often wear a tie or a suit to strategically signal their high status and make a competent impression (Maran et al., 2021; Oh et al., 2019). These exogenous cues can have profound influences on observers’ decisions, both when they consume information from traditional news outlets and social media coverage (Levine & Schweitzer, 2015; Maran et al., 2021; Todorov et al., 2015). However, relatively less is known about how exogenous status cues such as different attire would sway observers’ judgments when they are exposed to transgressive contents. In the current research, four studies (with Studies 1 and 4 being pre-registered) examined when and how formal versus casual attire influences observers’ moral judgments in transgressive contexts. We specifically focus on hypocritical transgressions that feature contradictory moral claims. We propose an attire effect on moral favoritism depending on hypocrisy and intentionality information, such that people favor transgressors in formal (versus casual) attire (H1), except for when they transgress hypocritically (H2). And the moral favoritism toward attire and reduced moral favoritism toward hypocrites are related to people’s perceived intentionality underlying the transgressions (H3).

Our findings first substantiated people’s moral favoritism toward transgressors in formal (versus casual) attire (H1). When behavioral intentionality is ambiguous, people evaluated formal- (versus casual-) attire transgressors as less intentional and self-interested, and judged them more leniently in terms of moral blame and moral impression (Studies 1 to 3). Moreover, hypocrisy attenuated moral favoritism based on formal attire (Studies 2 to 4; H2). Moral claims increased perceived intentionality for transgressions and reduced the situational ambiguity that allowed for unintentional attributions and favorable judgments (Studies 2 and 3). Moral favoritism toward formal- (versus casual-) attire transgressors was even reversed when hypocritical transgressions were unequivocally intentional to pursue self-interest (i.e., suggesting high levels of hypocrisy; Study 4). Altogether, these findings support the argument that signs of hypocrisy attenuate, or even reverse, attire-based moral favoritism.

Theoretical and practical implications

Wearing formal (versus casual) attire makes people act like high-status people, being more assertive and dominant, and thinking more abstractly and globally (Kraus & Mendes, 2014; Slepian et al., 2015). People with high social status (e.g., organizational leaders and politicians) can also actively choose different attire, to draw interpersonal attention or public appeal (Maran et al., 2021; Ratcliff et al., 2011). However, less is known about whether formal (versus casual) attire gains interpersonal privileges following transgressions. Filling this gap, we indeed found privileged moral judgments of formal- (versus casual-) attire transgressors. People formed competent impressions from formal (versus casual) attire, which prompted moral (especially trustworthy) impressions. The attire effect on morality was independent from its effect on sociability (Studies 2 to 4) and may not be simply attributed to an overall positive impression (Bocian et al., 2018). The attire effect on moral favoritism was also consistent with previous studies showing that people judge esteemed transgressors more leniently (Kakkar et al., 2019; Polman et al., 2013).

Our findings can also add insights into the social signaling literature. Both status cues and verbal communications can be used to attract cooperators and gain social esteem (Kraus & Mendes, 2014; Maran et al., 2021; Shank et al., 2018). Observers may license transgressors, either when they possess high-status characteristics (Kakkar et al., 2019; Polman et al., 2013), or when they successfully disguise their misdeeds under a veil of moral or prosocial words (Lönnqvist et al., 2015). However, employing both strategies may backfire and even induce more negative moral reactions than transgressing blatantly.

Moreover, the intertwined relationship between attire and hypocrisy provides a unique case of impression formation and updating. People can form initial impressions of others simply based on appearance cues and then update such impressions given additional descriptions of behaviors. Importantly, information across different channels may not only influence the magnitude of moral appraisals independently but also the moral implications of each other. In the case of moral judgments, our work revealed the significant role of visual cues (e.g., attire) and their differential effects contingent on hypocritical communications. People even judged formal- (versus casual-) attire targets more harshly when their hypocritical transgressions were depicted as clearly intentional (versus unintentional, Study 4; see more detailed discussion in the next section). As such, depending on textual information about behaviors, people can weigh visual inputs to different extents and even interpret them differently.

Our findings can have practical implications, for both senders and receivers of status signals. Organizational and political leaders often atone for their misdeeds while adjusting their tones, gestures, and clothing. Aiming to mitigate public blame and restore trust, it can be important to discern whether the misdeeds are preceded by contradictory moral stances, and how intentional the misdeeds are perceived. Furthermore, the attire effect on moral judgments is important to consider when justice outcomes are at stake. For example, legal decisions should be made in relatively standardized attire conditions wherever applicable. Though the influence of attire on moral judgments may be small, it can be difficult to eliminate, especially when transgressors are evaluated in a comparative setting with shared responsibilities (our Study 2), and even when people are incentivized to ignore status cues (Oh et al., 2019).

Limitations and future directions

Some limitations of the current research should be noted, however. First, our examined transgressions were mostly low on severity (e.g., un-environmental behaviors), which allowed subjective interpretations of intentionality. In this context, attire-based moral favoritism is consistent with previous studies showing moral leniency toward high- rather than low-status transgressors (Kakkar et al., 2019; Karelaia & Keck, 2013; Polman et al., 2013). Other studies show that people judge high- (versus low-) status or competent transgressors more harshly, however (Gray & Wegner, 2009; Guglielmo & Malle, 2010; Malle et al., 2014; Monroe & Malle, 2017; Weiner & Laurent, 2020). This latter line of research often focuses on high-severity transgressions that are deemed as wrong and intentional (e.g., killing, cheating, or stealing). We speculate that the moral and intentionality ambiguities of transgressive behaviors play a role in this discrepancy (Karelaia & Keck, 2013; Monroe & Malle, 2017). The mechanisms for judging transgressions with ambiguous versus absolute intentionality can be different (Malle et al., 2014; Monroe & Malle, 2017), such that status cues may prompt moral leniency for ambiguous transgressions (Studies 1 to 3) but moral harshness for clearly intentional transgressions (see also our Study 4 findings). When transgressions become clearly intentional, hypocritical, and severe, people no longer give formal- (versus casual-) attire targets moral credentials but make harsh judgments due to their betrayal of high expectations (Dong et al., 2021; Karelaia & Keck, 2013). However, it is important to note that our results do not provide direct evidence for the above line of reasoning, since we only manipulated (un)intentionality but not the ambiguity of intentionality.

Second, the current research examined formal versus casual attire as a manifestation of high versus low status. Complementing previous research using descriptive status information (Dong et al., 2021; Kakkar et al., 2019; Polman et al., 2013), it sheds light on the visual and subtle influence of status cues. However, the current attire stimuli also induced other confounding factors. It is not clear (1) whether the attire effect extends to female targets, whose competent attire prompts more complex social appraisals (Howlett et al., 2015; Oh et al., 2019), or (2) how attire as status cues influences moral judgments through other impression dimensions like wealth, dominance, and attractiveness. Also, we conceptualized formal attire as a prototype of competence- and prestige-based status in moral judgments; however, little is known about whether the current effects generalize to other status cues (e.g., endogenous ones like competent- versus incompetent-looking faces).

Conclusion

Formal (versus casual) attire prompts favorable moral evaluations for transgressors and their misdeeds, except when they are hypocritical and make contradictory moral claims. The current research demonstrates that moral underpinnings of formal versus casual attire are contingent on perceived intentionality of concurrent behavior. Following subtle manipulation of attire, people can use double moral standards and attribute different intentions and motives. Broadly, our findings can have implications on various situations where people make, or formally need to make, moral judgments. From everyday social encounters where people decide how to get along, to online social media where people decide what to comment, and to organizations and the court where people decide who to discipline and punish, the mere presence of subtle status cues can influence whether people trust norm violators and make favorable justice decisions.