Over the past two decades, unethical behavior in and by organizations has received unprecedented scrutiny in the media (Joseph et al., 2009). The proliferation of corporate scandals (e.g., Wells Fargo, Volkswagen, Mylan Epipen), changes in the Federal Sentencing Guidelines for Organizations, and the Sarbanes–Oxley Act have accentuated the need for ethical values and conduct within a business context (McGrath, 2017; Treviño et al., 2008). As a result, research interest in business ethics is also burgeoning (Islam, 2020; Painter et al., 2019). Prior work suggests that judgements related to an individual’s ethicality influence many business-relevant attitudes and outcomes, such as one’s trust in an employee and the organization they represent, as well as decisions related to whether or not to continue a business relationship (e.g., Chen & Mau, 2009). Several studies indicate that, rather than basing such judgements in others’ actions, we often make appearance-based inferences about individuals (e.g., Antonakis & Dalgas, 2009; Oh et al., 2020; Olivola & Todorov, 2010).

Prior appearance-based research revealed that people draw rapid inferences/snap judgements from visually assessing a target person, and that these inferences affect perceptions and choices that are aligned with judgments made in the absence of time constraints (Willis & Todorov, 2006). Although authors suggest that physical appearance can influence impressions on many individual qualities (e.g., personality traits, mood, culture, social status; Knapp et al., 2013; Lower, 2018), most studies on appearance-based inferences have, thus far, focused on a target individual’s competence (e.g., Antonakis & Dalgas, 2009; Ballew & Todorov, 2007; Oh et al., 2020), and few have examined other traits. Considering the growing importance of business ethics in recent years, our research addresses the scarcity of studies on appearance-based ethics-related inferences by examining workplace attire as a main predictor of ethicality perceptions.

Research indicates that clothing is a significant factor in first impressions and initial assessments of an individual (Lennon, 1986; Lower, 2018). Maysonave (2001) argues that in today’s digital world, first and ongoing impressions based on visual aspects are critical. More casual or inappropriate apparel choices on the part of an employee can result in negative evaluations of the employee, as well as their entire corporation, as they can project an image of a lax or unprofessional company. In contrast, professional attire is one of the most important attributes in sha** favorable workplace perceptions (Ruetzler et al., 2012). For instance, Wood and Benitez (2003) suggest that those wearing formal attire at work are ranked higher in credibility, are taken more seriously, and have a higher likelihood of being regarded as “upper management material” by top executives. Moreover, Slepian et al. (2015) and Lukavsky et al. (1995) note that formal clothing engenders respect by signaling ethicality-related characteristics such as professionalism and norm compliance.

Given the long trend towards a more casual workplace (Bhojani, 2019), which may have altered the general public’s attire-based perceptions, we assess whether employees dressed in professional attire are still perceived more favorably. Specifically, we examine whether individuals wearing professional attire (i.e., business formal and business casual attire) are perceived as more ethical than employees dressed in casual attire at work. We further examine whether there are differences in perceptions of ethicality when professionals wear business formal or business casual attire.

Our arguments for these main effects, as well as for attire appropriateness as an explanatory mechanism, and context (i.e., industry) as a boundary condition, are grounded in signaling theory (Connelly et al., 2011; Spence, 1973). Specifically, we argue that the choice and wearing of a certain attire style is a symbol/signal used by the wearer to convey information about themselves to others, such as their values or social identity (Banks et al., 2021; Barney et al., 2020; Rafaeli et al., 1997), which in turn may affect perceptions of ethicality. Signaling theory also suggests that “signalers” are judged by others on how well they conform to social expectations (Banks et al., 2021). In the workplace, these expectations include those related to workplace attire.

Research by Agovino (2019), Karl et al. (2013), and Shinn et al. (2011) indicates that, with some notable exceptions (e.g., the information technology industry), most organizations either explicitly require or expect professional attire at work. As professional attire appears to be the norm or aligned with recent social expectations for the workplace, and given conforming with others’ expectations generally results in positive outcomes for the signaler (Banks et al., 2021; Biddle, 1986), we argue that employees wearing professional attire will be, in general, perceived more favorably and to possess greater ethicality than those who wear casual attire at work. To account for the role of social expectations and conformity/violation of such expectations in sha** perceptions of ethicality, we include “attire appropriateness” (i.e., the perception that one’s attire is appropriate for the situation) as an explanatory mechanism in our model.

In addition, there have been many calls to account for the role of context in sha** organizational perceptions and outcomes (Johns, 2006, 2018). We respond to these calls by examining how a target’s industry affects attire-based perceptions of ethicality. As dress codes/attire-based expectations vary by industry (Agovino, 2019; Shinn et al., 2011), attire-based evaluations for employees working in different industries will likely vary as well. Whereas in banks and professional service firms formal attire is the norm, other industries have casual dress codes (Agovino, 2019). Given what is considered “appropriate” attire for one context may not be so for another, we examine the moderating effects of target industry in the relation between attire style and perceptions of ethicality, with attire appropriateness as a mediator. We thus aim to determine whether we can make general inferences about workplace attire-based social expectations, or whether we need to assess expectations within context and choose workplace attire accordingly, to understand perceptions of ethicality.

To conclude, this paper takes stock of the current state of attire-based perceptions of ethicality and suggests some paths forward for this dynamic topic. From a theoretical standpoint, to the best of our knowledge, our research is the first to examine appearance-based inferences regarding employee ethicality and among the first to examine the role of context in attire-based inferences [see Kwantes et al. (2011) and Howlett et al. (2015) as exceptions]. By examining a potential explanatory mechanism (i.e., attire appropriateness) and a boundary condition (i.e., target industry), our framework is comprehensive enough to allow for an enhanced understanding of the relation between attire style and perceptions of ethicality. From a methodological standpoint, our hypotheses were tested in two different studies, with different adult samples. Subsequently, we conducted three focus groups with employed adult participants to gain additional insight into the underlying mechanisms through which attire style influences ethicality perceptions, and what might be work-relevant outcomes of such perceptions. Given the attention devoted to business ethics in recent years, this study is both timely and practically important.

Theory Background and Hypotheses

Attire-based Inferences of Ethicality

Past research suggests that people often make appearance-based inferences about individuals (e.g., Antonakis & Dalgas, 2009; Oh et al., 2020). For example, Lower (2018) highlights that physical appearance plays an important role in first impression formation, and research by Olivola and Todorov (2010) supports the idea that complex judgments about individuals are made based on appearance in a fraction of a second. What is perhaps more surprising is that these rapid judgements are consistent with those made without any time pressure, and affect important individual, organizational, and country-level outcomes. For instance, appearance-based competence judgments have been found to predict election results in several studies (e.g., Antonakis & Dalgas, 2009; Todorov et al., 2005).

Research on appearance-based inferences reveals that, among other visual cues (e.g., facial traits), attire is a highly influential factor in first impressions and initial judgments (Carr et al., 2009; Howlett et al., 2015; Lennon, 1986; Lower, 2018). In the current paper, we focus on attire, which refers to one’s clothing (Esmail et al., 2020) and excludes the other components of “dress”. Although attire (clothing) and dress are often used interchangeably, the latter is a much broader construct, defined as “an assemblage of modifications of the body and/or supplements to the body” (Roach-Higgins & Eicher, 1992, p 1). Johnson et al. (2014) list as examples of body modifications cosmetic use, piercing, tattooing, cosmetic surgery and even dieting and exercising, and, as examples of body supplements, clothing, accessories, glasses, and hearing aids. According to the Apparel Search Company, an online guide to the Apparel and Textile Industry, attire (clothing) is limited to textiles worn to cover, protect or decorate the body for functional and/or social reasons (Apparel Search Company, 2006). We focus on attire because it covers a larger part of the body than other elements of dress and is thus likely to contribute more to impression formation. Additionally, body modifications and other body supplements have received considerably less attention in the literature, and their ranges are much wider, less clear, and may function in different ways (Johnson et al., 2014).

Studies on the effects of attire on impression formation, social perception and attributions have a long history (Johnson et al., 2014; Lennon & Davis, 1989). Among different attire characteristics (e.g., color, quality of fabric), attire style/mode has emerged as particularly relevant to workplace perceptions and outcomes (Howlett et al., 2015; Karl et al., 2013; Kwantes et al., 2011). Whereas some studies on the effects of workplace attire categorize the latter into formal vs. informal (e.g., Slepian et al., 2015), others recognize that there are three main styles of attire that we encounter in business settings: business formal, business casual, and casual (e.g., Karl et al., 2013; Shinn et al., 2011).

Prior research reveals that professional attire generally triggers favorable perceptions of the wearer, whether the latter is an employee or a candidate interviewing for a job (Carr et al., 2009; Ruetzler et al., 2012). For instance, Wood and Benitez (2003) suggest that those dressed in formal attire have higher credibility, are taken more seriously and are more likely to be regarded as “upper management material” by top executives. Other authors also argue that formal attire signals positive occupational attributes, such as the following of norms and professionalism in an employee (e.g., Butler & Roesel, 1989; Lukavsky et al., 1995; Slepian et al., 2015). Using this research as our point of departure, in this paper we focus on the effects of one’s workplace attire on others’ perceptions of that person’s ethicality. While prior research has often used the terms ‘ethicality’, ‘ethics’, and ‘morality’ interchangeably (Leban et al., 2020), in this paper we use the term ‘ethicality’ as a comprehensive concept. Specifically, whereas some research conceptualizes ethicality as adherence to a set of community norms and rules (Leban et al., 2020; Moraes et al., 2017), others recognize that it goes beyond norms, rules, laws or societal obligations (e.g., Ambrose et al., 2008; Kohlberg, 1981, 1984; Yukl et al., 2013) and is grounded in personal values and beliefs (e.g., Luedicke et al., 2009). Recent research has revealed specific values associated with the concept of ethicality, such as honesty, integrity, altruism, and concern for sustainability (Yukl et al., 2013). We follow this more comprehensive view of ethicality (e.g., Fichter, 2018; Nygaard et al., 2017; Yukl et al., 2013).

Signaling theory (Connelly et al., 2011; Spence, 1973) provides an explanation for why attire style influences ethicality perceptions. This theory centers around information asymmetry between individuals or larger collectives (e.g., firms) and how “signalers” (i.e., insiders who possess information relevant for making certain decisions) send informational cues (i.e., “signals”) to receivers (i.e., outsiders lacking this information) to reduce information asymmetry and influence desired outcomes. Insiders typically decide to send out positive signals to outsiders and avoid sending negative information that would lead to unfavorable judgements or undesirable decisions (Taj, 2016). Recently, this theory has been applied to examining the effects of leader signaling on followers’ evaluation of the leader’s ethics. Specifically, Banks et al. (2021) propose that ethical leadership behavior comprises signals related to the enactment of prosocial values and expressions of moral emotions targeted at organizational stakeholders such as subordinates/followers. Through these signals, leaders influence followers’ perceptions of leader ethicality. Signals related to the enactment of prosocial values include those that convey “normatively appropriate conduct” on the part of the leader (Banks et al., 2021; Brown et al., 2005), often resulting in more favorable perceptions of one’s ethicality. Banks et al. (2021) argue that the more a leader signals values that are important for a certain social identity, the stronger is the signal and the more such a signal leads to an ethical evaluation of the leader.

Given appearance-based judgments are made when there is little to no other information on a target (Antonakis & Dalgas, 2009; Olivola & Todorov, 2010), information asymmetry is highest in this context. Thus, signaling theory is particularly relevant for studying attire-based perceptions of ethicality. Building on the reasoning of Banks et al. (2021), we argue that wearing attire that conforms to expectations for the workplace in general (i.e., professional attire) signals normatively appropriate conduct, which, in turn, will translate into more favorable ethicality judgements. Whereas ethical leadership literature typically assumes (either implicitly or explicitly) formal leadership (i.e., a supervisor-subordinate relationship), some prior research also emphasizes that leadership can come from anywhere within the organization, and more informal (peer-based) leadership is also important (Shaughnessy et al., 2017; Zhang et al., 2012). As such, rather than assuming formal leadership, our focus is more generalized and involves perceptions of the more neutral category of fellow professional peers.

Indeed, prior literature suggests that perceivers associate several positive attributes related to ethicality with formal attire, possibly due to the halo effect (Latham et al., 1975; Nisbett & Wilson, 1977). Theory on this effect proposes that perceivers form a global evaluation of a target based on a single attribute of the target, which then impacts the perceiver’s evaluations of other unknown target attributes. For instance, Gross and De Dreu (2021) suggest that those perceived as following norms are also perceived as behaving more honestly; Scilhavy and King (2009) argue that individuals scoring high on professionalism have a higher ability to recognize ethical dilemmas, use moral equity judgments, and engage in ethical actions. Other positive occupational attributes associated with formal attire, such as self-discipline, expertise, credibility, trustworthiness, and being responsible (Adomaitis & Johnson, 2005; Kwon & Johnson-Hillery, 1998; Sebastian & Bristow, 2008) have also been positively related with perceptions of different aspects of ethicality (e.g., values, behavior) in prior studies (e.g., Bataineh, 2020; Davis & Rothstein, 2006; Du Toit, 2015). Moreover, research suggests individuals make role and hierarchy attributions based on attire formality, with individuals being dressed in more formal attire being perceived as having higher level positions within organizations, and more authority and responsibility (e.g., Damhorst, 1985). The study of Johnston et al. (2009) also reveals that individuals go as far as inferring/projecting others’ education level based on attire. Specifically, those dressed in more formal attire are perceived to have a higher level of education. Both a higher level of education and being in a position of authority in an organization are indicative of a track record of accomplishment, more exposure to complex situations, and more training in problem-solving in general, and navigating ethical dilemmas in particular (Boni & Lozano, 2007; Deloitte, 2018; Gomez et al., 2020). Research suggests that higher education in general (Boni & Lozano, 2007; Gomez et al., 2020) and business schools in particular (Frémeaux et al., 2018), include various forms of ethical training in their curricula. Reports by Deloitte (2018) also suggest that organizations are increasingly including ethical trainings for those in leadership positions. As such, perceivers may project higher ethicality onto individuals in higher authority positions, as the latter should be better equipped to understand and deal with ethical issues and have higher levels of responsibility. Some prior literature actually reveals an implicit assumption that leaders have high moral ethics (e.g., Luthans & Avolio, 2003).

Additionally, casual attire has been associated with negative attributes, such as perceptions of tardiness, absenteeism, flirtatious behavior (Egodigwe & Alleyne, 2003); perceptions of laxness and decreased productivity (McIntyre, 1998); perceptions of the target as unprofessional, incompetent, underqualified or uncommitted (Maysonave, 2001). Specifically referencing ethicality, Peluchette and Karl (2007) suggest that casual attire may lead to a “casual” work ethic. Taken together, the literature above suggests attire style may trigger a halo effect which extends to perceptions of ethicality, such that employees dressed in formal attire at work are likely perceived as higher in terms of ethicality than those dressed in casual attire.

Taken together, prior studies’ findings on attire formality and occupational attributes (e.g., professionalism, credibility, trustworthiness) and status (i.e., position within an organization) suggest that individuals dressed in more formal attire will likely be perceived as more ethical than those dressed casually. As such, we put forth:

Hypothesis 1

Attire style influences perceptions of ethicality in the workplace, with individuals dressed in more professional attire (business formal) being perceived as more ethical than those dressed in casual attire.

Some prior work within the attire literature suggests that business casual attire is a “happy medium” in the workplace, as it retains some of the advantages of both casual (e.g., comfort, higher morale) and formal attire (e.g., being perceived more favorably on different occupational attributes) (Shinn et al., 2011). Others have argued, however, that the lack of authority the wearer of business casual portrays will still result in some unfavorable judgements and will make it harder for one to progress up the corporate ladder (Black & DiNardo, 1994). As few prior studies have examined differences between business formal and business casual attire and between casual and business casual attire regarding how others perceive those who wear them, and no study has focused on ethicality perceptions, we formally ask:

Research Question 1a

Are individuals dressed in business formal attire perceived as more ethical than those dressed in business casual attire?

Research Question 1b

Are individuals dressed in business casual attire perceived as more ethical than those dressed in casual attire?

The Mediating Role of Attire Appropriateness in the Attire Style—Perceived Ethicality Relation

Normatively appropriate conduct is an implicit explanatory mechanism for how ethical leader behavior affects followers’ perceptions of leader ethicality in Banks et al.’s (2021) review of ethical leadership literature. As mentioned above, the leader signals prosocial values and a social identity by conforming to social expectations in the workplace, and these signals positively affect perceptions of leader ethicality. We argue that wearing “appropriate” attire at work represents normatively appropriate conduct (for leaders, but also peers or individuals interviewing for a job) and will thus result in more favorable ethicality assessments by observers.

In trying to explain attire appropriateness, Griffiths (2008) noted, in a study of orchestra players, that observers displayed a shared understanding of the “code” of dress for a concert situation, likely emerging from shared cultural labels attached to attire style and prior experiences of the socio-cultural practices of concert situations. Because black concert dresses are traditionally associated with female orchestra players, observers appraised the concert dress as the most appropriate attire from the three shown to them. The author also argued that participants may have exhibited a preference for prototypes—what is most frequently seen in a western classical performance situation. Other authors also tried to describe how appropriateness evaluations are formed; Johnson and Roach-Higgins (1987) referred to prior experiences related to similar circumstances, behavior sequences of events that help us classify appearance into appropriate or inappropriate for certain interactional situations, based on predetermined, identifiable categories. Drawing on inference theory, Shao et al. (2004) discussed how existing knowledge structures and judgement heuristics expedite the defining of expectations about objects and events and their labeling and categorizing, and how individuals make inferences based on knowledge and judgement of what is considered appropriate workplace attire.

Despite employing “appropriateness” as a variable in their models and attempting to explain how it functions, previous studies did not define it. To address this limitation, based on the explanations above, we offer the following conceptualization for appropriateness of workplace attire: fit/compliance with shared social expectations of workplace attire, formed based on prior experiences in a given situation and what is most frequently seen in that particular situation.

Research has found support for the significant effects of attire appropriateness on important workplace outcomes (Shao et al., 2004). For instance, Bardack and McAndrew (1985) showed that appropriateness of clothing affects the decision about whether a person should be offered a job. The findings of Ray (1986) support this idea; the author showed that a job applicant rated as “low” on appropriateness of attire received recommendations to hire that were significantly different from those rated “moderate” and “high” for appropriateness. In a study examining the effects of attire style on the musical evaluation of female classical soloists, Griffiths (2008) also found a strong concept of what constitutes appropriate dress for a female recitalist on the part of observers, with the target dressed in concert attire being rated significantly more positively than those dressed in nightclubbing attire or jeans. These findings suggest that individuals wearing “appropriate” attire are rated more favorably on a variety of criteria (e.g., hireability, musical performance). Thus, a halo effect is likely to appear here as well, with those wearing more “appropriate” attire being perceived as more ethical.

The experimental study of Gross and De Dreu (2021) suggests that conforming to rules and other social norms provides a basis/is a pre-requisite for honest, ethical behavior. Whereas our focus here is not actual ethical behavior, but perceptions of ethicality, we argue that the two outcomes are assumed to be linked. For instance, Falconi (1996) states “If you look sharp, you are more likely to act sharp” (p 13), Kaplan-Leiserson (2000), claims that “The way you look directly affects the way you think, feel, and act… When you dress down, you sit down […] Manners break down, you begin to feel down, and you’re not as effective” (p 39) and Lee (2005) concurs: “When we’re dressed in appropriate business attire, we tend to act in appropriate business fashion. When we’re not, we don’t” (p 36). In their review of workplace attire literature, Shinn et al. (2011) suggest that advocates of formal attire codes believe that employees wearing casual attire at work not only leave their professionalism open to question, but can also jeopardize the reputation of their employer. They provide Target Corporation as an example of a company that recognized this threat to its reputation and revised its dress code from business casual to jacket and tie in 2004. Although no prior research has examined if violations of workplace attire expectations affect perceptions of ethicality, following the reasoning above, we argue that individuals wearing “appropriate” attire will be assumed to act ethically and thus will be perceived as more ethical. On the contrary, those wearing inappropriate attire, as “rule-violators”, will be punished by being rated lower on ethicality.

In terms of what attire style may trigger higher attire appropriateness perceptions in the workplace (in general), research has documented significant shifts in what has been viewed as appropriate over the past two decades. Although, traditionally, professional attire was the norm, in the 1990s the boom of hi-tech and dotcom companies has led to more casual workplace attire (Karl et al., 2013; Shinn et al., 2011). These companies have encouraged this change based on the idea that more casual attire increases employee morale and productivity (Hunsberger, 2005 in Karl et al., 2013). Both Karl et al. (2013) and Shinn et al. (2011) indicate, however, that while some organizations presently accept casual attire throughout the week or have “dress casual” days, many organizations have shifted back to requiring professional attire at work, in light of this attire style being increasingly related to employee productivity and professionalism, which, in turn, relate to workplace outcomes tied to profitability. Since professional attire appears to be aligned with more recent social expectations for the workplace, assuming that individuals wearing more professional/formal attire at work will be rated higher on attire appropriateness than those wearing casual attire seems warranted.

Taken together, our arguments regarding the relations between attire style and attire appropriateness and attire appropriateness and perceptions of ethicality, respectively, suggest a mediating effect of attire appropriateness in the attire style—perceptions of ethicality relation. Thus, we formally forward:

Hypothesis 2

Perceptions of workplace attire appropriateness mediate the relation between attire style and perceptions of ethicality.

Context and Attire-based Inferences of Ethicality

Studies emphasizing the role of context in organizational behavior have a long history (Johns, 2018). Mowday and Sutton (1993) define context as “stimuli and phenomena that surround and thus exist in the environment external to the individual, most often at a different level of analysis” (p 198). Thus, context can take the form of many different situational characteristics (e.g., leadership/supervision, group dynamics, organizational culture). Some scholars have argued that situational characteristics are even stronger than individual attributes in sha** workplace perceptions and outcomes (Davis-Blake & Pfeffer, 1989). In the attire literature, Damhorst (1985) revealed, for instance, that the context within which attire is perceived has a notable impact on a target’s evaluation. In their dyadic study, attire was assessed in social interactions and context was operationalized as the attire style of another individual the target person was compared with. Their findings revealed, for instance, that male and female targets wearing business formal attire were, in general, described as higher in rank when their adjacent companion wore casual attire and males in business formal attire were described as giving more directive, rewarding, or punishing communications to females wearing casual attire. Prior leadership research also suggests that perceptions of different phenomena are not universally held, but are shaped by context (e.g., Tskhay et al., 2014, 2017). For instance, social expectations of a leader/leadership prototypes are influenced by national culture (Globe Project, 2020). Thus, it is possible that the norms or cultures of an organization or industry might also affect perceptions of what “looks like” an ethical person. That is, perceptions of ethicality may be different across different industries, where norms are different. Signaling theory suggests that observers may use signals when comparing individuals to prototypes/their cognitive templates and making judgements about a target. Such signals may be nonverbal, such as the attire individuals wear (Barney et al., 2020).

Industry type is a particularly relevant contextual factor for our study’s focal phenomenon, as prior research in the attire literature (e.g., Agovino, 2019) emphasizes that dress codes and habits vary greatly by industry, and that what is expected or considered “appropriate” in one setting is not the same in another. Some authors even offer attire manuals that provide norms by industry to help individuals dress “appropriately” in the workplace (see Rothman, 2016 for an example). In professional service firms and the floor of the U.S. Congress, for instance, employees are expected to wear formal attire (Agovino, 2019; Shinn et al., 2011). High-tech firms, however, do not have such strict norms and rather encourage a relaxed, casual workplace attire, even among individuals occupying high positions in the company’s hierarchy (Agovino, 2019; Karl et al., 2013; Schmearer, 2019). For instance, Mark Zuckerberg, Facebook’s founder and chairman, is notorious for wearing a T-shirt and hoodie at work. Because of their divergent attire norms and also because, in presenting a history of workplace attire styles, Shinn et al. (2011) suggest that, over time, dress code trends have primarily been influenced by the technology sector and the professional services (which include accounting, finance and banking), we focus on these two industries to capture contextual effects potentially affecting our focal relation.

Prior literature reveals several reasons why technology/IT companies have different attire norms than companies pertaining to the professional services. Some authors point to identity differences between the two industries, suggesting that what employees wear is reflective of the culture they work and live in. The professional services industry uses the “power suit” (i.e., formal business attire) as a symbol of the power and prestige it views as core identity values (Hollander, 1994), whereas the IT industry uses casual attire to symbolize its investment in a product, idea, and mission, rather than image and “the trap**s of wealth” (Schulte, 2021, p 552). Casual attire is also used in IT to symbolize values such as non-conformity, creativity and entrepreneurship, and a focus on productivity (Ashton, 2019; Schulte, 2021). Greenfield (2012) notes that casual attire (the hoodie) has become central to tech employees’ identity and aims to project the image of a different, more modern businessperson. Schulte (2021) supports this idea and suggests that this type of attire symbolizes other markers of power and prestige (rather than wealth), such as academic credentials and a whiz kid reputation.

Another reason for the different attire norms between the IT and professional services industries involves the types of interactions employees traditionally engaged in, specifically the direct interactions with clients. As Ashton (2019) points out, IT employees were, traditionally, members of staff who did not venture into the public domain or meet with clients. As such, there was an unwritten rule that they could wear casual attire at work. The author notes that “techies” have kept true to that principle and that even those who now directly interact with customers (while performing on-site services or attending pre-sales meetings) and interview for a job will rarely dress in formal attire.

Aside from the different attire style norms, as noted above, prior literature also suggests a difference in the strength of attire norms between the two industries. Because, in general, professional services firms have more clear and strict norms for formal attire, whereas tech firms offer more flexibility/a wider range of casual attire outfits (Agovino, 2019; Schmearer, 2019), we argue that judgements of attire appropriateness will be more extreme in the professional services industry, as norms are clearer and stricter and rule/norm violations are more obvious in this context. In other words, the relation between attire style and attire appropriateness should be stronger for professional services firms rather than high-tech firms.

Considering the linkage between attire appropriateness and perceptions of ethicality proposed earlier, as well as our arguments above for how industry as context may impact the relation between attire style and attire appropriateness, we put forward:

Hypothesis 3

Target industry moderates the mediated relation between attire style and perceptions of ethicality, with attire appropriateness as mediator, such that the relation between attire style and attire appropriateness will be stronger for individuals working in industries with formal attire expectations (e.g., professional service organizations) and weaker in industries with casual attire expectations (e.g., IT).

Overview of Studies

Our studies examined the effects of attire style, attire appropriateness, and context (industry) on perceptions of ethicality. In Study 1, we used written vignettes to describe working professionals wearing three different attire styles (i.e., business formal, business casual, and casual) when working in two different industries (i.e., professional services and IT). The vignettes allowed us to assess perceptions of appropriateness and ethicality. Study 2 replicated Study 1, except that, instead of using written vignettes, we used photographs of working professionals to depict different attire styles, and we included a written description of where they worked. We supplemented these studies with three focus groups (referred to as Study 3 hereafter), to better understand our findings from studies 1 and 2, and to gain an insight into workplace-relevant outcomes of ethicality. All studies were approved by the Institutional Review Board at the first author’s university and subjects’ consent was received before participation.

Study 1

Procedure

Written vignettes (see “Appendix A”) were used to describe working professionals in one of three attire styles (business formal, business casual, casual) and in one of two industry conditions (professional services and information technology [IT], respectively), resulting in a 3 × 2 experimental design. We included vignettes for both males and females. Industry type was manipulated in the first part of the vignette, and attire style in the last sentence, where a target employee was described in one of the three attire styles. Attire was described according to categories and descriptions used in Karl et al. (2013). The vignettes for the different conditions were written such that they were the same length and similar in description. Participants were randomly assigned to one of the conditions in an online survey. The survey instructions read, “In the next section you will read a short scenario where you observe another person while working. This person is also an employee, and they will be dressed in business formal, casual or business casual attire.” Descriptions for these attire styles were then provided (see “Appendix A”), followed by the condition vignette and survey questions. An open-ended question was included in the survey asking participants to explain their responses. Lastly, demographic information was collected. To increase data quality, we screened for careless responding by randomly including a careless response item in the survey (Meade & Craig, 2012). The careless response item was: “Please select Moderately Disagree for this question.” Out of 365 participants, 42 (11.5%) failed this response (i.e., they responded other than “Moderately Disagree”) and were therefore excluded from our sample.

Participants

We collected data from LinkedIn contacts and the alumni of a university in the Northeast United States. The latter approach allowed us to reach a wide range of working professionals. The Office of Development and Alumni Engagement at the first author’s university sent out an email to alumni, inviting them to volunteer to participate in our study.

From our original sample (n = 380), we excluded possible students (n = 15), to solely examine the perceptions of those with working experience. From the remaining sample of 365 participants, we also removed participants who failed the careless response item (n = 42). The final sample therefore consisted of 323 participants, of which 153 were female, 166 male and 4 preferred not to answer. Our sample’s mean age was 49.33 years, with a range of 21–90 years and a standard deviation of 16.01 years. 296 participants reported being Caucasian, 2 Black or African American, 3 Asian, 10 belonging to an “Other” category, and 12 preferred not to answer. Participants came from a number of industries (82 education, 30 banking/finance, 27 government, 16 nonprofit, 15 health care, 7 manufacturing, and others, such as communications, information technology, transportation, social work, construction, engineering, human resources, law enforcement, retail, and sales). Company sizes, in the form of employee numbers, ranged widely: 1–100 employees (79 participants), 101–200 employees (24 participants), 201–300 employees (11 participants), 301–500 employees (26 participants), 500 + employees (126 participants). 12 participants reported an “other” company size and 45 provided an “unknown size” or did not respond.

Measures

Ethicality

To assess perceptions of ethicality, we adapted the 15-item Ethical Leadership Questionnaire (Yukl et al., 2013; Cronbach’s alpha = 0.98) by changing the point of reference from a leader to a person in the workplace. Specifically, participants were told to rely on their “gut feeling” when indicating how well each statement described the person in the photograph on a Likert scale (1 = Strongly disagree to 6 = Strongly agree). A sample statement reads, “The person you observed shows a strong concern for ethical and moral values.”

Attire Appropriateness

To measure attire appropriateness, participants were asked to respond on a Likert scale (1 = Very inappropriate to 5 = Very appropriate), “Based on the short description you just read, to what extent is the person you observed dressed appropriately for this setting?” Participants were randomly assigned to view either the appropriateness or ethicality question first.

Analyses and Results

To ensure target gender did not affect our results, we tested for the effects of target gender on attire appropriateness and ethicality. Given we did not find such effects, we proceeded with collapsing the data. All analyses were performed in R. Table 1 reports means, standard deviations, and correlation coefficients for our study’s variables.

Table 1 Means, standard deviations, and intercorrelations for Study 1 and Study 2
Full size table

To test Hypothesis 1 and explore Research Question 1, which focused on the relations between attire style and perceptions of ethicality, we ran a one-way ANOVA. The Shapiro–Wilk normality test was violated (p < 0.05), which indicates the normality assumption was violated. We therefore used a nonparametric ANOVA test, the Kruskal–Wallis rank sum test, which was significant (H(2) = 15.50, p < 0.001). This suggests there are significant differences among at least some of the conditions being compared. To determine where the differences between conditions occurred, post hoc tests were conducted using the Dunn test in the FSA package in R (Ogle et al., 2021). Results showed that casual attire was perceived as less ethical than business formal attire (p < 0.05), which supports Hypothesis 1. Regarding Research Question 1, casual attire was perceived as less ethical than business casual attire (p < 0.001; Research Question 1b); however, there was no difference in perceptions of ethicality between business casual and business formal attires (p = 0.13; Research Question 1a).

Hypothesis 2 stated that the effect of attire style on perceptions of ethicality is mediated by attire appropriateness. We tested this hypothesis using the mediation package in R (Tingley et al., 2014). The bootstrap** method (1000 bootstrapped samples) was used to adjust the standard error estimates in three mediation models, where we contrasted two conditions at a time: (1) casual versus business casual attire, (2) casual versus business formal attire, and (3) business casual versus business formal attire (see Fig. 1).

Fig. 1
figure 1

Results from Mediation analyses for three comparisons: Casual versus Business Casual; Casual versus Business Formal; Business Casual versus Business Formal

Full size image

Results revealed that, compared to casual attire, business casual attire was positively related to perceptions of attire appropriateness, which was positively related to perceptions of ethicality (ACME = 0.51, 95% CI 0.32 to 0.75, p < 0.001; ADE = 0.01, 95% CI − 0.37 to 0.38, p = 0.996; Total Effect = 0.52, 95% CI 0.21 to 0.84, p < 0.01). Moreover, compared to casual attire, business formal attire was positively related to perceptions of attire appropriateness, which was positively related to perceptions of ethicality (ACME = 0.55, 95% CI 0.35 to 0.78, p < 0.001; ADE = − 0.21, 95% CI − 0.59 to 0.18, p = 0.27; Total Effect = 0.34, 95% CI 0.04 to 0.65, p < 0.05). However, compared to business casual attire, business formal attire was not perceived as more or less appropriate (ACME = 0.04, 95% CI − 0.07 to 0.15, p = 0.47; ADE = − 0.22, 95% CI − 0.52 to 0.09, p = 0.16; Total Effect = − 0.18, 95% CI − 0.49 to 0.13, p = 0.25). For the first two mediation models, we found support for full mediation. Therefore, Hypothesis 2 was supported.

Lastly, we tested Hypothesis 3 (i.e., the moderated-mediation model) with the user-defined function PROCESS in R. Similar to the mediation analyses, we contrasted two conditions at a time: (1) casual versus business casual attire, (2) casual versus business formal attire, and (3) business casual versus business formal attire (see Table 2 and Fig. 2).

Table 2 Results for Moderated-Mediation Analyses for Study 1, by Condition
Full size table
Fig. 2
figure 2

Results from Moderation-Mediation analyses for the Casual versus Business Formal Condition, Study 1

Full size image

First, when comparing casual and business casual attire conditions, there was no support for moderated-mediation, as the 95% bootstrap confidence interval for the index of moderated-mediation (index = 0.12) included zero (− 0.11, 0.37). However, when comparing the casual and business formal attire conditions, there was support for moderation-mediation (index = 0.47), as the confidence interval did not include zero (0.23, 0.76). The results suggest that, compared to business formal attire, casual attire is perceived as less appropriate (a1 = − 2.43, 95% CI − 2.79 to − 2.02). Moreover, lower levels of perceptions of appropriateness are associated with lower levels of perceived ethicality (b1 = 0.32, 95% CI 0.18 to 0.46). Finally, the indirect effect of attire style on perceptions of ethicality through perceived appropriateness holds for both levels of the moderator (i.e., target industry), but the relationship is stronger for professional services (industry coefficient (standard error) and confidence intervals: professional services = − 0.77 (0.18), 95% CI − 1.14 to − 0.44; IT = − 0.30 (0.11), 95% CI − 0.55 to − 0.12). Lastly, when comparing business casual and business formal conditions, there again was support for moderation-mediation (index = 0.31), as the confidence interval did not include zero (0.11, 0.57). The conditional effects for the different levels of the moderator showed that (1) compared to business formal, business casual is less appropriate in the professional services industry (coefficient = − 0.68 (0.23), 95% CI − 1.13 to − 0.22), and (2) compared to business formal, business casual is more appropriate in the IT industry (coefficient = 0.45 (0.23), 95% CI 0.001 to 0.91). Overall, this study’s findings suggest that industry type influences the relation between attire style and perceptions of ethicality, with perceptions of appropriateness as a mediator.

Study 2

Procedure

Study 2 differed from Study 1 in two ways. First, we used photographs (see “Appendix B”) instead of written vignettes to better control for and depict attire style conditions. We purchased professional photographs of models from Shutterstock. Second, instead of using only one person in a photograph to represent a condition, we used four models for each condition so that participants would be less biased by characteristics of a single model. In each condition, the models (two males and two females of average height and weight) were standing in various poses so that body language would not influence perceptions (Pease & Pease, 2006). They were dressed in neutral colors (e.g., blue and black) rather than bold colors (e.g., red). The photographs were cropped to exclude faces and avoid potential confounds, such as attractiveness (Klein & Shtudiner, 2020). Included with each photo was a short description of the industry for context (professional services or IT). This description was the same written vignette used in Study 1. Therefore, we again used a 3 (attire style: casual, business casual, business formal) × 2 (industry type: professional services/IT) design, where participants were randomly assigned to one of the six conditions. The survey instructions read, “Based on the short description you just read and the photo you saw, please indicate how well each of the following statements describes the person you observed by selecting one of the following response choices. Rely on your “gut feeling” when responding.” Participants then answered survey questions about perceptions of appropriateness and ethicality. Lastly, demographic information was collected. Similar to Study 1, we included a careless response item in the survey (Meade & Craig, 2012) that read: “For this question, please select Disagree.” Out of 434 participants, 10 (2.3%) failed this response (i.e., they responded other than “Disagree”) and were therefore excluded from our sample. The final sample size was 424 participants.

Participants

To test our hypotheses, we used a sample of working professionals recruited through Amazon’s Mechanical Turk (MTurk). To ensure high-quality data, our participants were required to meet the following conditions: (1) they had to be located in the U.S., (2) they had to be full-time working employees, and (3) they had to have HIT approval rates (proportion of completed tasks) of at least 95%. The final sample consisted of 424 participants (175 female, 249 male; Mage = 36.47 years, SDage = 11.06 years, range = 21–72 years; 372 Caucasian, 25 Black or African American, 16 Asian, 4 American Indian or Alaska Native, 7 “Other”). Participants came from several industries (118 manufacturing, 81 banking/finance, 52 health care, 51 education, 18 government, 12 nonprofit, and others, such as telecommunications, information technology, transportation, construction, hospitality, logistics, marketing, software development, retail, and sales). Company sizes ranged from 1 to 100 employees (90 participants), 101–200 employees (79 participants), 201–300 employees (83 participants), 301–500 employees (63 participants), 500 + employees (106 participants), and “unknown size” (3 participants).

Measures

Ethicality

We used the same, adapted Ethical Leadership Questionnaire (Yukl et al., 2013) as Study 1 (Cronbach’s alpha = 0.93).

Attire Appropriateness

Similar to Study 1, participants were asked, “Based on the short description you just read and the photo you saw, to what extent is the person you observed dressed appropriately for this setting?” Participants were randomly assigned to view either the attire appropriateness or ethicality question first. Participants were asked to explain their response through an open-ended question, and demographic data was collected at the end of the survey.

Manipulation Check

To make sure the photographs captured the three attire styles well, we first performed a manipulation check. Specifically, we used a sample of 53 students (Mage = 22.72 years; SDage = 2.42 years; 26 females; 20 Caucasian; 14 Black or African American; 8 Asian or Asian Indian; 5 Hispanic, Latino, or Spanish Native; 6 Other) from two universities in the Northeast United States to classify photos into the three styles. Participants were asked to select the attire style category in which the photograph representing the models best fit, based on the attire style descriptions (see “Appendix A”). Data were analyzed using a goodness of fit Chi-square analysis. Because there were three levels (of attire styles), we compared observed frequencies to expected frequencies, where the expected frequencies for each level were equal to 53/3 = 17.67. Given multiple comparisons, we used a Bonferroni correction so the new p value for the results had to be less than 0.017 to be significant. Significant results demonstrated that expected and observed frequencies were different, meaning that the pictures had been properly classified. For example, for the female casual condition, 51 participants rated the picture as casual, 2 participants rated the picture as business casual, and 0 participants rated the picture as business formal. We concluded the picture was properly classified as casual (\({\chi }^{2}\left(2\right)=94.45, p< 0.001\)). All results were significant, indicating participants accurately identified attire styles.

Analyses and Results

Table 1 reports means, standard deviations, and correlation coefficients for Study 2 variables. The same statistical analyses as in Study 1 were performed in Study 2.

First, to test Hypothesis 1 and explore Research Question 1, which focused on the relations between attire style and perceptions of ethicality, we ran a one-way ANOVA. As in Study 1, the Shapiro–Wilk normality test was violated (p < 0.05). Moreover, the Levene test for homogeneity of variance was also violated (p < 0.01). We therefore again used a nonparametric ANOVA test (i.e., the Kruskal–Wallis rank sum test), which was significant (H(2) = 6.80, p < 0.05). The post hoc Dunn test showed that casual attire was not perceived differently from business formal attire (p = 0.25). Therefore, Hypothesis 1 was not supported. Regarding Research Question 1, the data suggest business casual is perceived more favorably compared to casual attire (p < 0.05; Research Question 1b), but not necessarily more favorably than business formal attire (p = 0.28; Research Question 1a), which aligns with the results from Study 1.

Mediation analyses were again used to test Hypothesis 2 (i.e., the effect of attire style on perceptions of ethicality is mediated by attire appropriateness). Results (see Fig. 1) revealed that, compared to casual attire, business casual attire was positively related to perceptions of attire appropriateness, which was positively related to perceptions of ethicality (ACME = 0.35, 95% CI 0.23 to 0.49, p < 0.001; ADE = − 0.04, 95% CI -0.22 to 0.15, p = 0.720; Total Effect = 0.31, 95% CI 0.13 to 0.50, p < 0.01). Moreover, compared to casual attire, business formal attire was positively related to perceptions of attire appropriateness, which was positively related to perceptions of ethicality (ACME = 0.35, 95% CI 0.23 to 0.47, p < 0.001; ADE = − 0.19, 95% CI − 0.39 to 0.01, p = 0.06; Total Effect = 0.16, 95% CI 0.04 to 0.36, p = 0.09). However, compared to business casual attire, business formal was not perceived as more or less appropriate (ACME = 0.00, 95% CI − 0.06 to 0.06, p = 0.96; ADE = − 0.15, 95% CI − 0.33 to 0.02, p = 0.08; Total Effect = − 0.15, 95% CI − 0.33 to 0.03, p = 0.09). For the first mediation model, we found support for full mediation, and for the second mediation model, we found support for partial mediation. Therefore, Hypothesis 2 was supported.

Lastly, we tested Hypothesis 3 (i.e., the moderated-mediation model) using the same approach used in Study 1 (see Table 3). For all three condition comparisons, there was no support for moderated-mediation, as the 95% bootstrap confidence intervals for the index of moderated-mediation (standard error) included zero: (1) casual versus business casual = 0.00 (0.10), 95% CI − 0.19 to 0.20; (2) casual versus business formal = 0 0.11(0.10), 95% CI − 0.07 to 0.31; and (3) business casual versus business formal = 0.11(0.06), 95% CI 0.00 to 0.22. Therefore, results from Study 2 suggest that industry type does not moderate the indirect effect of attire style on perceptions of ethicality through the mediating variable, attire appropriateness.

Table 3 Results for Moderated-Mediation Analyses for Study 2, by Condition
Full size table

Study 3

Studies 1 and 2 demonstrated that individuals’ ethicality perceptions differ by attire style. In both studies, business casual was perceived as more ethical compared to casual attire; however, business formal was perceived as more ethical compared to casual attire only in Study 1. Across both studies, there was no difference between business casual and business formal for perceptions of ethicality. Moreover, we found support for mediation in both studies, where the relation between attire style and ethicality perceptions was mediated by attire perceived appropriateness. Lastly, we found inconsistent results for the industry moderation effect.

We conducted three focus groups to further explore these inconsistent results, so that we can (1) better understand the differences in perceptions between casual, business casual, and business formal attire styles, (2) better understand the effects of industry type on perceptions, and (3) explore workplace outcomes associated with ethicality perceptions, to highlight the importance of such perceptions and pave the way for future research in this underdeveloped area.

Procedure

Although there are inconsistent recommendations on the number of focus groups that should be used in a study (Nicholson et al., 2020), research suggests that three to four groups are adequate for exploring various phenomena (Krueger, 1994; Nyamathi & Shuler, 1990). As such, we employed three focus groups to address issues related to possible inconsistencies across groups with respect to group dynamics, and ensure that different voices are heard, but without information overload.

Based on best practices in the literature (Broom, 2005), we pretested a script (see “Appendix C”) using a pilot focus group to ascertain if the questions were understood, engaged participants, and generated useful information. Minor modifications to the script were made and, using well-established guidelines (Nicholson et al., 2020), we conducted the focus groups. Each had 3–5 participants, was conducted online synchronously via Zoom Video Conferencing, lasted 45–60 min, and was recorded with participants’ consent. One author facilitated the focus groups, and another was an observer. One week before the focus groups, participants were provided a short description of focus group guidelines and attire style definitions and were instructed to reflect on their work experiences and how employee attire influenced their perceptions of employees. All focus groups started with ice-breaking conversations, to set an open and safe climate, and a review of the ground rules (see “Appendix C”).

Participants

Participants were recruited from two MBA courses, where they were offered extra credit in exchange for their participation. We sampled from this population because these participants had working experience across a variety of industries, and they represented the general working population that we were interested in studying. In total, there were 12 participants in three focus groups. The average age was 32.78 years (SDage = 13.33 years), and there were seven females and five males. Four participants identified as White, four identified as Black or African American, two identified as Asian, and two choose not to report their ethnicity. Participants came from several industries, including accounting, banking and finance, carpentry, education, government, health care organizations, and non-profit organizations. Company sizes ranged from 1 to 100 employees (1 participant), 201–300 employees (1 participant), 301–500 employees (1 participant), and 500 + employees (6 participants). One participant was unsure of their company size and two did not report their company size. Lastly, the average number of years spent in their industry was 7.10 (SD = 9.12 years).

Analyses and Results

Focus group videos were transcribed and shared among the first three authors, who independently identified categories of responses for each focus group question and then reached consensus regarding themes for each question. Next, the three authors independently viewed the transcripts and, for each participant, classified comments into themes. An 80% initial agreement rate (range: 70–95%) was achieved across the eight focus group questions, indicating a high level of interrater reliability. The authors then met, discussed differences in coding themes, and reached 100% consensus. Although we asked several questions, many of the responses overlapped. We therefore focus only on the most relevant questions, comments, and discussions for our focus group objectives stated previously (See Table 4 for a summary of results).

Table 4 Focus Group Key Findings and Sample Participant Quotes
Full size table

First, the question What are your thoughts on each attire style? revealed that casual attire was perceived more negatively than positively (5 participants reported negative perceptions and 2 reported positive perceptions). Participants across focus groups mentioned that an advantage of casual attire is that it makes people seem more down-to-earth and approachable. However, casual attire was more often perceived negatively, and was associated with perceptions such as being unprepared or lacking respect for the job. On the other hand, business casual was the only attire style that received only positive remarks. In fact, business casual was described as “a nice happy medium” by one participant.

For business formal attire, participants were spilt as to whether it elicited positive or negative impressions (6 positive perceptions and 5 negative perceptions). For instance, one participant reported positive perceptions of trust: “I think people perceive you by how you dress, and a lot of people trust you if you dress formally.” However, others had negative perceptions of business formal attire, and even specifically associated suits and ties with scandals, crime, and unethical behavior. One participant noted, “He has a suit and tie on, he’s probably unethical as well.” Another participant referenced white collar crimes explicitly and added, “People can commit crimes in a suit and tie.”

These differences in perceived ethicality that different attires generate may help explain the inconsistent results across Studies 1 and 2. Business formal attire is sometimes viewed positively and sometimes viewed negatively. Perhaps the positive signals that business formal attire used to send have changed for some people due to scandals and the unethical behavior of those who often wear formal attire. On the other hand, if there is a halo effect, the all-positive attributes associated with business casual would explain why business casual was perceived as more ethical compared to casual attire, which was mostly perceived negatively.

The second inconsistency we wanted to explore was the moderating effect of industry type (Does attire appropriateness depend on the situation?). Through the focus groups, we learned there are other situational variables that may be influencing perceptions of appropriateness and therefore ethicality, such as whether someone is being interviewed or is an existing employee, the level of management of an employee, and the type of client an employee is working with. In Studies 1 and 2, these last two points were not explicitly addressed in the descriptions of the employee being observed, which may have left them open to interpretation by participants. We elaborate on these additional findings below.

The value of contextualizing other variables that may include appropriateness attributions was highlighted by several participants. For example, some focus group participants argued there are differences in expectations for those interviewing for a job versus current employees. “That’s what I’ve always been taught [to dress up for an interview]. But once you have a position, it’s different. Unless you are giving a presentation, once you are there, you must know what you are doing … [so you don’t have to dress up].” Another participant shared a story about her teacher who was always dressed casually and who had a lot of tattoos. She asked how he ever got a job, to which he replied that he never looked and dressed like that for his interview, but instead wore a suit for his interview and waited until he got the job to start dressing differently and getting tattoos.

Additionally, other participants referenced the effects of roles and management level in influencing workplace attire expectations. One participant pointed out that, in an engineering company, floor employees will be dressed more casually, whereas high-level management employees are dressed more formally. Thus, even within the same industry, there are different expectations and perceptions of what is appropriate to wear. Lastly, some participants noted how they would change their attire contingent upon the client with whom they were interacting. Thus, the moderating effect of industry type may be more intricate than we expected.

Finally, we were interested in the outcomes of ethicality. For the last question What are likely outcomes of ethicality or the lack of it?, participants reported a wide range of negative workplace outcomes that could result from being perceived as unethical, including (1) disciplinary action/termination, (2) unwillingness to work with an individual, (3) legal issues, (4) negative co-worker relations, (5) loss in reputation/trust/credibility, and (6) unlikely hiring or promotion. For example, participants reported that unethical behavior could result in legal troubles related to fraud, getting fired from a job, being asked to resign, or progressive disciplinary action, where unethical behavior starts with a verbal or written warning and progresses towards termination. Some participants noted it can affect interpersonal working relationships, such that employees may feel uncomfortable and may not want to work with an unethical co-worker.

In qualitatively reviewing the data, we noticed an additional theme that added to our understanding of the inconsistent results: conversations about the effects of Covid-19 over time. For example, one participant reported a change in norms towards laxer, less business professional attire: “I think that people are becoming more practical, especially after Covid and realizing that […] there is a workplace benefit to […] not be choked by a tie.” Another participant noticed a similar trend, “…we interviewed five people and none of them wore suits on the virtual Zoom call. Instead, they all looked very professional, but they did not wear a suit, so I thought that was an interesting trend versus what I saw last year.” These comments and conversations suggest another reason for the discrepancies in our data. Study 1 was conducted at the beginning of Covid-19, whereas Study 2 was conducted nearly two years into the pandemic. If expectations and norms evolved to be more accepting of more casual attire, so too might people’s attire-based perceptions of appropriateness and ethicality.

General Discussion

Given the pervasive impact of unethical behavior and the well-publicized corporate scandals of the past, this research aimed to ascertain the influence of attire style on perceptions of ethicality in the workplace. Studies suggest individuals often act based on what they believe to be true, regardless of facts (Ambrose & Schminke, 1999), so it is important to understand what factors influence individuals’ beliefs regarding other’s ethicality. Overall, the results across all studies tell an interesting story. First, business casual may be the happy medium. It was the only attire style associated with positive attributes exclusively in the focus groups, and it was perceived as more ethical compared to casual attire in the experimental studies. Albeit, we do leave the caveat that there were no differences in perceptions of ethicality between business casual and business formal attires, so putting business casual on a pedestal may be premature. Second, we found support for mediation—the relation between attire style and perceptions of ethicality can be explained by attire appropriateness.

However, two results from studies 1 and 2 were inconsistent: business formal was perceived as more ethical than casual attire only in Study 1 and the moderating effect of industry did not hold across studies. The focus group input suggests that Covid-19 challenges, poor economic conditions, and time of data collection could account for the discrepancies we found across the studies, particularly for the first inconsistent finding. Study 1 data were collected in the first year of the pandemic (December 2020), whereas Study 2 data were collected nearly two years into the pandemic (November/December 2021). Given the radical changes that took place in the workplace environment (i.e., remote work), workplace attire expectations also changed. For example, popular press articles reference how clothing retailers need to adapt to “new demands of the hybrid work week” (Maheshwari, 2022). Since the start of the pandemic, the line between where you work and live has thinned, and this trend appears to also apply to attire and where you live. Retailers and social media have coined the terms “power leisure” and “work leisure” in reference to these new attire styles (Maheshwari, 2022). Moreover, unemployment levels have also changed. What employees wear at work may be a reflection of the labor market, such that people wear more professional attire “as a form of risk aversion” during times of high unemployment, but dress more casually when labor shortages exist, as it was when Study 2 data were collected. It is therefore possible that a reason for the inconsistent results we found in our studies was related to these changes in macro factors (e.g., labor market), norms, and expectations over time. That is, whereas business formal was perceived more ethically compared to casual attire in Study 1, this effect faded away in Study 2, possibly due to the changing perceptions throughout Covid-19 and the changing economy.

Our focus group discussions surrounding Covid-19 support this idea. For instance, in our results section we mention a quote from one participant who noted that, after Covid-19, employees are becoming more practical and “there is a workplace benefit to … not be[ing] choked by a tie.” Another participant also noted that none of the candidates they interviewed for a job during the pandemic wore suits on the virtual Zoom call, although they did come across as professional.

A further breakdown of the focus group quantitative data revealed some possible changes in ethicality perceptions over time. Whereas in both Study 1 and Study 2 business casual was perceived most favorably and casual least favorably, from Study 1 to Study 2, all measures of ethicality increased (both studies used 1–6 Likert scale), and casual attire made the largest increase in perceptions of ethicality compared to business casual and business formal. In fact, in Study 2, we found that casual attire was perceived as less ethical compared to business casual only (there was no difference between casual and business formal attires). Perhaps, due to Covid-19 and the changing labor markets, casual became more acceptable, and was therefore not perceived differently compared to business formal in regard to ethicality. Given our limited data in this respect (in terms of both samples and times of the studies), we offer this conjecture with caution.

The second inconsistency surrounded the moderating effect of industry type on the relation between attire style and perceptions of ethicality, with attire appropriateness as mediator. Specifically, in Study 1 we found that the indirect effect of attire style (casual versus business formal) on perceptions of ethicality through perceived attire appropriateness was significant for both types of industry, with the relation being stronger for professional services compared to IT, as Hypothesis 3 proposed. Likewise, we found that, compared to business formal, business casual was considered less appropriate in the professional services industry than in the IT industry. In Study 2, however, we found no support for moderated-mediation. Indeed, results from the subsequent focus groups suggest further research is needed to understand how the context and industry type influence perceptions of ethicality. Specifically, inconsistent results may be due to the industry types we chose to study. For instance, we realized, through the focus groups, that IT and professional services industries may be more similar than we originally thought in terms of attire expectations. Focus group participants often referenced other industry types that are more casual and less professional, such as construction. Moreover, we learned from the focus group that other variables are influencing perceptions, such as level of management, the type of client an employee is working with, and whether someone is being interviewed or is currently employed. As previously mentioned, in Studies 1 and 2, these points were not explicitly addressed in the descriptions of the employee being observed, which may have left them open to interpretation by participants. Therefore, we conclude that, even though the results of Study 1 suggest an industry effect, whether this effect is present, as well as the relation between attire style and perceptions of ethicality may depend on other contextual variables, such as type of organization (startup vs. established) and the target’s position within the organization (executive vs., for instance, manual worker) that we did not explicitly control for.

Theoretical Contributions and Implications

Some of our studies’ findings bring support to prior attire literature findings. For instance, previous research has found negative perceptual outcomes associated with casual attire, such as unprofessionalism, incompetence, and poor work ethic (Maysonave, 2001; Peluchette & Karl, 2007). On the other hand, casual attire has been associated with friendliness (Shinn et al., 2011). Participants from the focus group in the current study also mentioned the approachability aspect of casual attire. Thus, casual is this “double-edged sword” in that it is associated with approachability, but also with perceptions of incompetence and laziness (Shinn et al., 2011; Wood & Benitez, 2003). Moreover, business formal has been associated with credibility and higher levels of management (Wood & Benitez, 2003), and we found consistent observations from the focus group. However, in the middle is the “happy medium,” business casual. Our findings are also consistent with this, as business casual was perceived as the most ethical in Studies 1 and 2. Furthermore, in the focus group, one participant explicitly referred to business casual as the “happy medium.”

However, what is unique to our research is that we explicitly study attire-based perceptions of ethicality, which have not been previously investigated. Consistent with signaling theory (Connelly et al., 2011; Spence, 1973), we found that attire style signals to observers (participants) information that they then use to make judgements when all else is uncertain. Moreover, consistent with the halo effect (Latham et al., 1975), it seems as though positive attributes associated with attire style, such as the “happy medium” for business casual, carry over and have similar effects on perceptions of ethicality. That is, just as business casual is the happy medium between casual and business formal for many other perceptual outcomes, it is also the happy medium when it comes to perceptions of ethicality.

Consistent with prior literature, our studies suggest that trends and changes in workplace attire expectations change over time (Shinn et al., 2011). Whereas formal attire was the norm pre-1990s, the expansion of the internet and the recession in the early 1990s resulted in a shift towards more casual attire, followed by another shift back towards more formal attire. During this time, business casual attire became popular and a “happy medium.” Our studies support the trends and shifts that change over time due to changes in the market, economy, and societal norms, as suggested by previous research (Maheshwari, 2022; Shinn et al., 2011). However, what is unique to our study is the explicit description of the effects of the pandemic on these changes. Not only did we find changes across our empirical studies and therefore the pandemic, but participants from the focus group explicitly mentioned observations of changes in norms and attire style throughout the course of the pandemic. Specifically, whereas business casual was always the happy medium, perceptions of business formal and casual changed. Business formal became less popular, and casual attire became more acceptable to the point where there were no longer differences in perceptions of ethicality between business formal and casual as we moved deeper into the pandemic.

Our findings have direct and relevant impact on all forms of research concerning ethicality at work and in business. For example, with its emphasis on signaling theory, our work informs research concerning ethical leadership (e.g., Banks et al, 2021). Similarly, it informs business ethics research focused on diversity, equity, inclusion, and belonging (DEIB), particularly among groups for whom conceptions of appropriate attire may differ (e.g., Fujimoto et al., 2022). One especially intriguing area of inquiry is the hiring process, especially as the use of artificial intelligence -which ostensibly should lessen or eliminate bias related to perceptions based on attire—is on the rise (e.g., Figueroa-Armijos et al., 2022).

Our findings also provide evidence that may be useful to adjacent streams of research. Many of the participants explicitly or implicitly stated that attire acts as a signal, supporting our argument for signaling theory. One participant shared that dressing appropriately signals being a “rule follower.” If someone is a rule follower, we might associate this with being ethical. In a similar vein, our findings may also lend support to role theory (Biddle, 1979; Katz & Kahn, 1966). According to this theory, individuals within a social system (i.e., an organization) fulfill roles that align with others’ expectations. When role players conform to expectations, they better integrate into the social system and are typically viewed more positively by others. Our results support these tenets by showing that individuals wearing what was perceived as “appropriate” attire for their role (i.e., conforming to norms and expectations) were rated as more ethical.

The mediating effect we found may also bring support/be relevant for other organizational science theories. For example, Implicit Personality Theory (IPT) (McCrae et al., 2019) maintains that individuals exacerbate the strength of relations among traits of individuals they observe, hold biases regarding the perceived or assumed covariation among traits and characteristics, and develop impressions based on insufficient data about strangers (Riva et al., 2019; Schneider, 1973). In our studies, subjects were asked to judge the ethicality of individuals based only on either photographs or written vignettes that depicted attire. As attire appropriateness was found to mediate the attire style—perceived ethicality relation, it may be that attire appropriateness is part of a larger IPT held by respondents.

Practical Implications

Our research suggests that people use signals, such as attire style, to form impressions. This happens consciously and unconsciously. In fact, one focus group participant noted his “natural bias towards formal attire.” However, the problem is that most of these impressions occur unconsciously, which may be the case with the current study. Though, in the focus groups, no one explicitly associated ethicality with certain attire styles, studies 1 and 2 show these associations are, in fact, made. From a practitioner standpoint, this is extremely concerning. Ethicality assessments based on attire may be short-sighted and biased, resulting in suboptimal decision-making and poor judgment. In fact, such biases may result in poor employment decision-making (e.g., staffing, promotions, and compensation) and may even adversely impact protected classes, leading to damage to employees’ lives, low retention, low morale, legal liability, and decreased organizational reputation. Stereotypes have negative organizational outcomes including poor selection decisions (Rice et al., 2016; Riva et al., 2019), age discrimination (Ilișanu & Andrei, 2018), negative attitudes towards individuals with disabilities (Nelissen et al., 2016), and lower work-life conflict resolution (Li et al., 2017).

In a very practical and poignant example that exemplifies the role of attire in creating perceptions of ethicality, one participant shared a story of when she was in a training session. At the training, everyone was shown a picture of two people, where one was dressed nicely in a suit and the other one was not, and they were asked who they would trust more. The person in the suit was viewed as more trustworthy; however, the catch was that he was Ted Bundy, and the other person was an Ivy League school graduate and lieutenant governor of Pennsylvania. Examples such as this demonstrate that an attire-driven halo effect indeed plays an important role in creating perceptions of ethicality.

Indeed, participants from the focus groups believed we should not make judgements based on attire. One participant noted that what you wear may not be a signal of your competence, desire for a job, or the extent to which you care about something, but instead may be the result of inexperience and simply not being taught what to wear. Moreover, formal attire is often more expensive than casual or business casual attires. Those of lower social economic status may not be able to afford a “dress to impress” attire style and may subsequently be excluded from job opportunities, given their financial situation rather than their actual competence and ethical mindset. As integrity is among the most important traits when hiring new employees across all levels within an organization (Career Advisory Board, 2016), our research is timely and relevant, as it reveals that a job candidate may be discriminated against based on attire style. Thus, it is important for organizations to address these biases, by, for example, increasing awareness on them, training those involved in recruitment on overcoming them, using objective and validated assessments such as job-related work sample tests and structured interviews, and emphasizing the need for more data-driven employee assessments.

Limitations

As with any other study, our research has certain limitations worth noting. First, MTurk respondents self-selected into the sample pool, and alumni respondents were a select sample (they attended a single university), which limits the generalizability of our findings. Perhaps results would have been different if employees working in Silicon Valley companies were surveyed. Moreover, MTurk professionals received incentives to participate, and university alumni did not. This represents a potential confounding variable across studies that serves to reduce the generalizability of our findings to other populations and industries.

Another limitation relates to differences in methodological approaches (i.e., research design) between studies. Results differed across the written vignette and photograph samples, suggesting that the research prompt might have affected the results. It is entirely possible that the photographs offer more cues with respect to the variables studied (e.g., color of attire, age of the target), whereas written vignettes allow participants to “fill in the blanks” regarding missing cues. Moreover, for our criterion, we used a forced Likert scale, where participants had to lean towards ethical or unethical for their responses. Some participants commented that they felt their option did not accurately represent their true feelings, which may have affected results. Thus, it is possible that survey methodology affected results. However, in pilot studies using a 1–5 Likert scale, where participants could report neutral feelings, we found similar results.

Another limitation to our study is that we did not include a manipulation check to measure and identify the extent to which the attire styles were actually perceived as the norm for the different industries. We based our manipulations on prior research, which suggested what attire styles were the norms for given industries. However, it is possible that participants do not have these perceptions. If this is the case, this may be another reason for the inconsistent results we found regarding the moderating effect of industry type. Regarding the manipulation check for Study 2, we would like to note that we used students to verify the photos. It is possible that generational differences influence perceptions, so working professionals would have been a better population to conduct the manipulation test.

Lastly, to uncover potential reasons behind the inconsistent results found in studies 1 and 2, we used focus groups after conducting experiments. In retrospect, it is entirely possible that conducting focus groups before designing the experimental studies would have been useful in identifying and better understanding the mechanisms through which attire style influences perceptions of ethicality. As such, we recommend that future research in this area considers conducting focus groups before survey data collection.

Future Research Directions

Given our inconsistent findings regarding the moderating role of industry type, future research could explore the factors that may be causing these inconsistencies. Researchers could, for example, test the effects of or control for the level of management and employee role (e.g., type of client an employee is working with, whether the employee is being interviewed or is currently employed). Additionally, to assess the generalizability of our findings, other industries could be compared (e.g., law, construction). To better understand the effects of the labor market, perceptions of ethicality based on attire style could be studied under different conditions (e.g., low and high unemployment levels). Lastly, future researchers could specify and test the directions of the relations. For example, is there a positive relation between attire style and appropriateness for professional service organizations and a negative relations for IT/tech organizations? In other words, would formal business attire be perceived as appropriate in a professional services organization, and inappropriate in an IT/tech company?

Furthermore, future research could explore additional mediating factors that help explain how individuals form impressions of others with respect to ethicality. The mediating effect of perceived attire appropriateness may be related to IPT research. To ascertain trait clusters individuals believe are covaried, IPT could be tested by asking subjects to assess the personality, intelligence, or cognitive complexity and ethicality of strangers with different workplace attires. This research could ascertain if individuals with more complex personalities, intelligence, or cognitive complexity are less dependent on stereotypes or implicit co-variations among variables (e.g., individuals dressed in a way that violates norms and expectations aren’t necessarily less ethical than individuals that dress congruent with expectations).

Future research could also test moderators at different levels of analysis. At the individual level, for instance, rater individual differences (e.g., openness to experience, conscientiousness) may affect perceptions. It is entirely possible that those who are more open-minded (i.e., liberalism facet) are more open to all attire styles and rate them as “appropriate”. This could result in a lower (and perhaps insignificant) effect on perceptions of ethicality. At the organizational level, studies suggest that culture type, such as mercenary (i.e., focus on profit) or ethical (i.e., focus on ethical standards, ethical behavior, and managers acting as role models) affects a variety of workplace perceptions (Goffee & Jones, 1996; Ruiz-Palomino et al., 2013). For our focal phenomenon, organizational culture could either strengthen or weaken the individual level relation between attire style and perceptions of ethicality.

Additionally, future research could examine differences between explicit and implicit perceptions and biases (see Rubinstein et al., 2018). Our studies revealed a bias in that casual attire was perceived as the least ethical. However, it is not clear if this bias is explicit or implicit (i.e., whether it is related to something that people consciously believe or not). Although our focus group participants believed it was wrong to judge people based on what they are wearing, specific examples revealed that attire-based judgements were still made, and that casual attire was perceived as the least ethical. As such, we recommend that future research clarifies what type of bias this is and how to best address it/what may buffer against it.

Lastly, our research examined the impact of attire style on others’ perceptions of an individual’s ethicality. However, attire style may have effects on the target itself, too (see Karl et al., 2013). For example, more formal attire may make individuals feel more ethical, which may then affect their behavior. In other words, attire style may have a self-fulfilling prophecy effect, such that wearing certain clothes may increase ethical behavior in an organization, an important consideration that warrants future investigation.

Conclusion

People do not respond to reality; they respond to their perceptions of reality, formed through the signals they receive. Our results support this idea and signaling theory, as they show that attire style acts as a signal that, directly and indirectly, impacts perceptions of ethicality. Specifically, we found that participants perceived business casual as a “happy medium,” as they rated this attire style as more ethical compared to casual attire, and they mentioned only positive attributes related to this style. However, positive and negative attributes surrounded both casual and business formal attire styles. Furthermore, these perceptions were mediated by attire appropriateness. Although we did not find support for the moderating effect of industry type, focus group conversations suggest that other factors, such as Covid-19, changes in the labor market, and other factors (e.g., level of management), may influence attire-based perceptions of ethicality. Overall, our research supports some prior attire literature findings, introduces unique findings, and emphasizes important practical implications related to attire style’s impact on perceptions of ethicality and important related workplace outcomes.