Making Sense of Stigmatized Organizations: Labelling Contests and Power Dynamics in Social Evaluation Processes

How do social audiences negotiate and handle stigmatized organizations? What role do their heterogenous values, norms and power play in this process? Addressing these questions is important from a business ethics perspective to improve our understanding of the ethical standards against which organizations are judged as well as the involved prosecutorial incentives. Moreover, it illuminates ethical concerns about when and how (the exploitation of) power imbalances may induce inequity in the burdens imposed by such social evaluations. We address these questions building on two event-based case studies involving Hells Angels Motorcycle Club Norway, and contribute to organizational stigma theory in three ways. First, social evaluations of a stigmatized organization by multiple audiences are found to interact, collide and combine in a labelling contest. Second, we show that labels employed in this contest are pushed to either negative extremes (‘moral panic’) or positive extremes (‘moral patronage’). Finally, we show when and how power represents a double-edged sword in social evaluation processes, which can be wielded either to the benefit or to the detriment of the actors under evaluation.


Introduction
Social evaluation of organizations-including stigma, legitimacy, reputation, status and celebrity-has long been a core area of organizational research (Pollock et al., 2019). Since such evaluations can have substantial implications for organizations and their members-either beneficial or harmful depending on the type of evaluation (e.g., Pozner, 2008;Shadnam et al., 2020;Van Portfliet, 2020)-their social construction has important ethical dimensions. These relate not only to where the line is drawn between, for instance, acceptable and unacceptable behaviour, but also to who draws that line, based on which ethical principles, and enforced by what sources of power (Greve et al. 2010). From a business ethics perspective, this suggests a need to learn more about (the enactment of) the ethical standards various audiences employ to evaluate and judge others (Shadnam et al., 2020;Thomson, 2018). For instance, analyzing the roles of agency and power in the social evaluation of organizations links to ethical questions about the ends to which this power is deployed, as well as the incentives of social evaluators to engage in a process of identity (de)construction (Ashforth, 2019;Goffman, 1963;Paetzold et al., 2008). Moreover, any organization's activities-especially those leading to controversy-should ideally obtain consent from key stakeholders. Since acquiring such a 'social licence to operate' is "important as a moral ideal", it naturally draws attention to the power, agency and ethical standards of involved 'social licensors' (Demuijnck & Fasterling, 2016, p. 679; see also Mele & Armengou, 2016). These issues lie at the heart of our analysis.
More specifically, we advance our understanding of social evaluation processes by studying how multiple social audiences negotiate and handle stigmatized organizations. Organizational stigma is defined as "a label that evokes a collective stakeholder group-specific perception that an organization possesses a fundamental, deep-seated flaw that deindividuates and discredits the organization" (Devers et al. 1 3 2009, p. 155). It thus constitutes a negative social evaluation with a principally moral basis that can have extremely detrimental implications including social exclusion and marginalization (Helms et al., 2019;Pozner, 2008;Reuber & Fischer, 2010). Extant research on organizational stigma predominantly concentrates on the strategies organizations employ to manage and/or reduce the consequences of their stigma (Helms & Patterson, 2014;Hudson & Okhuysen, 2009;Lashley & Pollock, 2020;Reuber & Morgan-Thomas, 2019;Tracey & Phillips, 2016). While such studies take the existence of stigma as a given, 1 this is theoretically and conceptually constraining since opinions about organizations and their activities are socially constructed and, as such, can be inherently contested (Grandy & Thomson, 2018;Pozner, 2008;Shadnam et al., 2020;Van Portfliet, 2020). Hence, the same organization may be perceived differently by distinct audiences or at distinct points in time (Ashforth, 2019;Helms et al., 2019;Pollock et al., 2019), which requires an evaluation of "when and how multiple sets of actors (…) with potentially inconsistent views get involved and interact with one another" (Shadnam et al., 2020, p. 711).
To the extent that previous work has included external actors, this often remained limited to "a single evaluation by a particular audience" (Pollock et al., 2019, p. 465). Social evaluation processes where multiple audiences work to construct good or bad opinions about particular organizations remain under-explored . As mentioned, our analysis bridges this gap by bringing into focus the power and agency of multiple audiences carrying distinct and often conflicting values and norms. Our central research question asks how their opposing evaluations interact, collide and combine to define what is normal and what is stigmatized (Helms et al., 2019;Hudson & Okhuysen, 2014;Mishina & Devers, 2012;Pollock et al., 2019;Pozner, 2008). We take an inductive approach to this question by analyzing how social audiences outside stigmatized organizations handle and negotiate such organizations.
Building on empirical material deriving from (social) media, websites, official documents as well as multiple rounds of interviews, we conduct two event-based case studies. Both events are centered around the exihibition of photographs from a 5-year project by Norwegian photographer Marcel Leliënhof: "Helvetes Engler [Angels from Hell]: Hells Angels MC Norway". The first event was the April 2013 edition of the Nordic Light International Festival of Photography in Kristiansund (henceforth, Nordic Light), and the second was the May 2014 exhibition "For the Love of Freedom" at the University of Oslo's Museum of Cultural History (henceforth, KHM). The programmes of both events included public interviews and debates with Hells Angels' members to contextualise the exhibited photographs (more details below). This engagement with Hells Angels turned out to be highly controversial and induced fierce societal debates. The involved social audiences varied substantially in their expressed positions as well as their institutional power. These event characteristics provide an ideal setting for our analysis.
Before proceeding, it is important to note that we refrain from normative statements regarding Hells Angels throughout the analysis. This abstention does not imply we advocate a value-free standard of organizational action, nor does it reflect support for the organization at hand. Although potentially uncomfortable, our non-normative position is crucial from an analytical perspective. It allows us to "examine unethical or illegal activities as the sites for contested meanings (…) and open our perspectives to the processes and circumstances that facilitate them" (Hudson & Okhuysen, 2014, p. 245).

Theoretical Context
In any social processes where opinions, identities and categories are "dynamically and continually reconstituted", emphasis naturally falls on "activities of people and how these activities contribute to stable categories" (Langley, 2009, p. 415). This article therefore focuses on the social evaluation processes at the heart of Goffman's (1963) canonical conceptualization of stigma. Goffman (1963) argues that stigma arises when stakeholders discredit some social actors, possibly to the benefit of others. The result is a perception of an individual or organization as a deviant, threatening the existing social order (Devers et al., 2009;Hudson, 2008;Mishina & Devers, 2012;Paetzold et al. 2008). Goffman's (1963) types of stigma-i.e. abominations of the body, blemishes of individual character, and tribal stigma-furthermore suggest a strong connection between stigma and ethical standards. For instance, blemishes of individual character-which are central to our analysis-are generally deduced from reprehensible behaviour as judged "relative to the stigmatizers' values, norms and ideologies" (i.e. 'conduct' stigma; Ashforth, 2019, p. 23; see also Pozner, 2008;Thomson et al., 2018). Following this line of argument, the ethical and moral bases of social audiences' evaluations are at the core of our theorizing about organizational stigma. 2 1 Examples include the stigma linked to men's bathhouses (Hudson & Okhuysen, 2009), trade in human cadavers (Anteby, 2010), or cannabis (Lashley & Pollock, 2020).
2 Ethics relates to overriding principles of right and wrong "shared by a group on the basis of mutual and usually reciprocal recognition" (Hazard, 1995, p. 453), while morality concerns "the notions of right and wrong that guide each of us individually and subjectively in our daily lives" (Hazard, 1995, p. 451). Ethics and morality thus are distinct concepts. Yet, they are very closely related since the personal While extant scholarship generally restricts attention to the interaction between the target of stigmatization and one single audience (e.g., Kulik et al., 2008;Roulet, 2015), social actors usually face multiple audiences. 3 This is important since a heterogeneous constellation of audiences may encompass varying opinions about the (in)acceptability of social actors. Different audiences rely on different values and belief systems that carry distinct "decoding capacities" (Goffman, 1963, p. 28). Assessment of social actors thus occurs against differing standards (Helms & Patterson, 2014;Hudson, 2008;Thomson et al., 2018). This can translate into distinctive assessments of social actors' core characteristics and decisions. At the aggregate level, such contradictions trigger "competition between the social audiences and their opposite, mutually exclusive evaluations" (Hudson, 2008, p. 255;Hudson & Okhuysen, 2014).
Previous work has given some attention to the importance of such contestations for the stigma experienced by specific social actors. Yet, the underlying mechanisms, the ensuing ethical implications, as well as how these "varying evaluations should be weighted pose important theoretical and empirical issues" (Pollock et al., 2019, p. 466). Little is known, for instance, about how social audiences use, reproduce, and problematize stigma. Since language is central to social constructions of reality (Gergen, 2010), we argue that it also plays a critical role in the social evaluation(s) of organizations. This follows Shadnam et al.'s (2020, p. 704) notion that audiences deploy "words, metaphors and referential texts to construct an 'account' of [organizational] conduct". More specifically, we take inspiration from the symbolic management literature (Ashforth & Humphrey, 1997;Granqvist et al., 2013) to assess when and how the strategic use of symbols and labels becomes central to managing the impression others have of tainted organizations.
Powerful actors may be expected to take a central role in such social evaluation processes. Ansari et al. (2013) argue that the ideas and positions of powerful actors are more likely to resonate through social networks-a process referred to as catalytic amplification. Demuijnck and Fasterling (2016) and Russell et al. (2016) likewise state that the balance of stakeholder power is a critical determinant of firms' ability to acquire a social licence to operate. Power dynamics and actors' agency have not yet been fully explored in the context of organizational stigma (Hudson & Okhuysen, 2014;Pollock et al., 2019). Nonetheless, successfully managing others' impression of tainted organizations requires the power "to authoritatively represent noncompliance as a more fundamental incongruence with norms" (Shadnam et al., 2020, p. 703).
Clearly, simply stating that 'power matters' leaves open crucial questions about what configuration of actors leads to which outcomes, and why. Building on scholarship of institutional power (Lawrence, 2008;Munir, 2015), we unpack the role of power in social evaluation processes. We thereby pay attention to both formal authority and informal influence (Rhee & Fiss, 2014). 4 In doing so, we can also attend to ethical concerns regarding "the end to which power or the exercise of authority is deployed" (Bayer, 2008, p. 470). Moreover, since power imbalances increase the use of coercive tactics and reduce the willingness to make concessions (Casciaro & Piskorski, 2005;Lawler & Bacharach, 1987), we can address important issues of equity and fairness.

Research Context: The Stigma on Hells Angels Motorcycle Club
Our research does not deal directly with Hells Angels Motorcycle Club Norway (henceforth HAMC). Rather, we study social evaluation processes pertaining this organization. Even so, we must first establish the type and nature of the stigma faced by HAMC (Pollock et al., 2019). Hells Angels Motorcycle Club was founded in 1948 in California. Since then, descriptions in popular culture (Thompson, 1966), media, criminological research (e.g., von Lampe, 2019) as well as police and government reports (e.g., Europol, 2019) generally depict a close-knit community of social and economic outcasts, characterized by lawlessness and extreme hostility to outsiders. The same characteristics are observed in our Norwegian setting (Report to Parliament & no.7., 2010Parliament & no.7., -2011KRIPOS, 2012;Olsen, 2015). HAMC also actively contributes to this negative image by cultivating narratives of an 'outlaw' organization that rejects common social norms and laws (we return to this in more detail below). As a result, HAMC is widely considered as "a dangerous deviant" with deeply engrained objectionable traits-consistent with it being a stigmatized organization (Devers et al., 2009, p. 162;Hudson, 2008). In terms of 4 Informal influence derives from the symbolic and material sources of power embedded in regulations, norms and cognitive scripts. Symbolic sources of power include the meaning and credibility of social actors' identity, whereas material sources of power can be of an organizational or economic nature (Rhee & Fiss, 2014). morality of a member of a given community is derived from the ethical standards of that community (Hazard, 1995). We return to this below.

Data Sources
Our analysis relies on three sources of information. First, we collected local and national printed, online and television news items related to both events, as well as specialized reporting on photography, musea and ethics. To obtain further insights into public opinion, we searched for (online) commentary pieces, blog posts, entries on discussion fora, as well as Letters to the Editor. Finally, this data source covers the transcript of a recorded panel debate held during the May 2014 exhibition at KHM (University of Oslo, 2014), as well as the website discourse of Nordic Light and KHM. An overview of our 110 media items is provided in Table A1 in Online Appendix A.
Second, we brought together relevant official documents. This includes reports by the Norwegian police, government white papers, and Parliamentary decisions on organized crime. These document the official position towards HAMC. A non-exhaustive list of these documents is provided in Online Appendix A. Additionally, we collected formal statements by public officials, such as members of the Norwegian police and the Norwegian Association of Municipalities, as well as local and national office-holders.
Third, we conducted multiple rounds of interviews to help unpack the social processes emerging from the discourse of opposing audiences. Our interviews with the photographer and his co-author, the spokesperson of HAMC Norway and the leader of 'Payback' (the interest organization for motorcycle clubs in Norway) covered both events. Regarding Nordic Light, we interviewed the festival's managing and creative directors, the festival board's chairman and one additional board member, the chief of police, the news editor of the local newspaper and the chief county administrator. For KHM, our interviews included the project coordinator and both (co-)curators of the exhibition, as well as the museum's director. In total, we spoke to 16 respondents across several interview rounds.
Our loosely structured face-to-face interviews were conducted between April 2016 and April 2018, and lasted between 30 and 95 min. We began by asking respondents to share their personal experiences and views of the festival/exhibition. During the ensuing conversation, we probed for more detailed information about three main topics: i.e. organizational issues related to the events, perceptions of public reactions, and the roles, views, actions and power of all actors involved (to allow for cross-validation). After initial analysis of the first set of interviews, we returned to the field for additional interviews with (local) politicians, the local police and the events' organizers. The information obtained during our first interviews guided our focus in two ways. Firstly, our initial interviews brought additional actors and stakeholders to our attention. Secondly, several themes reappeared throughout our initial interviews (e.g., the pervasive use of labels, and the apparent importance of power imbalances), which we pursued further in the subsequent data collection efforts.

Data Analysis
We followed an iterative and recursive approach in our analysis (Eisenhardt, 1989). Initially, we coded our media materials and interviews with respect to the key protagonists (individuals and institutions), and developed a timeline of both events. Building on this information structure, we turned to a detailed coding and categorization of each source (i.e. media, interview transcripts, documents, and public debate). We thereby pursued an open coding process focusing on (a) descriptions of HAMC, (b) descriptions of other involved social actors, (c) motivations and arguments provided for these descriptions, and (d) opinions and evaluations of each event and its context. Both authors did this independently.
Archival and interview sources were subsequently reexamined in light of initial findings and, as mentioned, we returned to the field for additional data collection. Moving back and forth between data and developing theoretical insights provided flexibility better to capture the underlying social processes and power dynamics (Eisenhardt, 1989). We also asked an independent researcher to go through all transcribed materials ('peer debriefing'; Corley & Gioia, 2004;Tracey & Phillips, 2016). Subsequent discussions with the peer debriefer about the empirical material and our coding decisions allowed us to further crystallize the conceptual apparatus and the inferences drawn. The resulting data structure is illustrated in Table 1.
Before turning to our findings, two research-ethical issues require discussion. Firstly, since we combined data analysis with additional data collection, there arises a risk of confirmation bias: that is, the risk of seeking out (disregarding) new evidence that supports (invalidates) initial findings. We took a number of steps to avoid such bias. During our interviews, we worked with open-ended questions aimed at extracting respondents' narrative about the events. This benefits truthful and honest answers. Moreover, we kept questions as short and simple as possible since using terminology may involve leading or suggestive wording. This again allows respondents to tell their story with minimal guidance. 5 During data analysis, we independently coded transcripts and documents, and engaged an independent researcher to do the same. This peer debriefer was not informed about our observations nor about the development of our theoretical ideas. This approach forced us to keep an open mind throughout the project.
Secondly, since we study events involving public figures and officeholders, maintaining full anonymity is very difficult. Even so, we only explicitly name the photographer. We refer to all other individuals by their institutional position. When it comes to members of the public, we maintain full anonymity by citing entries on public discussion fora via the article commented upon, and using numbered references for Letters to the Editor. Throughout the analysis, we italicize quotes from our documents and interviews (translated from the original Norwegian).   Figure 1 provides a graphical representation of our key findings, and the remainder of this section discusses each element in turn. First, we describe the actions and decisions of the events' organizers, as well as the (planned) involvement of HAMC. Then, we examine the diverging opinions of social audiences and how these relate to their distinct ethical standards. We thereby show that social audiences express evaluations of social actors through the use of simple and recognizable labels. These labels are pushed to negative or positive extremes as part of a labelling contest. We conceptualize this as, respectively, instances of 'moral panic' (Cohen, 1972) and 'moral patronage'. Finally, we discuss when and how power inequalities between social audiences determine whose assessments gain the upper hand. This observation is important since distinct sources of power may be accessed or mobilized in different contexts.

The Two Events
At the December 2012 board meeting, Nordic Light International Festival of Photography in Kristiansund finalized its programme for the April 2013 edition. Among the headliners were Marcel Leliënhof's photographs of HAMC. To provide a contextualization for these photographs, the programme featured two posts involving members of HAMC. The first was an onstage interview with the HAMC spokesperson, and the second was a debate between HAMC Norway, the chief of the local police district, and a Norwegian professor specializing in white-collar crime. In January 2013, the outline of Nordic Light's programme by mistake became known to the local newspaper. Local media reports published in subsequent days document a heated debate between the leader of the festival board, intended guests of the planned debate, and the chief of the local police district (Letter to Editor 1; Botten, 2013aBotten, , 2013bJoakimsen, 2013). The local newspaper strongly opposed the invitation of HAMC, and used its position to steer the debate: Tidens Krav believed there could be grounds for raising a debate about whether it is wise to invite HAMC to this town. We took that stand, and had that debate. (Editorial, 1 February 2013) The festival's planned engagement with HAMC took on extra significance as several HAMC members were just then put on trial in one of Norway's biggest ever drug cases. This trial featured prominently in national and local media reports (Nilsen, 2013;Pedersen, 2013;Rise, 2013), as well as in entries on public discussion fora: This trial [in Tromsø] reinforces the strong scepticism and antipathy of 'most people' towards HAMC (the undersigned included). (Hesjedal, 2013, readers' comments) The controversy forced the organizers to make significant changes to the festival's programme within days of its initial publication (Rise, 2013). Marcel Leliënhof's photographs remained on the programme, but the onstage interview and panel debate were cancelled.
Our second event is linked to the official Bicentenary celebrations of the Norwegian Constitution in 2014. On 26 February 2009, the President of the Norwegian Parliament was formally charged with setting up a committee whose remit was to plan these celebrations (Innst. S. nr.162, 2008(Innst. S. nr.162, -2009).
The committee subsequently instructed all state institutions to provide a contribution to the jubilee. KHM's contribution was the exhibition "For the love of freedom", curated by a resident photographer (henceforth Curator 1 KHM) and a resident professor of archeology (Curator 2 KHM). It opened on 16 May 2014-i.e. the eve of Norway's Constitution Day-and one of the exhibition rooms was dedicated to Marcel Leliënhof's HAMC photographs.
As at Nordic Light, the exhibition programme listed additional events involving HAMC members. The highlight was an hour-long public debate on "Freedom and the boundaries of freedom", featuring the rector of the University, a professor in social anthropology, the spokesperson for HAMC Norway, and Marcel Leliënhof with his co-author. The chief of the Oslo Police District had been invited, but declined. HAMC's inclusion in this exhibition was once more highly controversial and fiercely debated. Yet, in stark contrast to the situation at Nordic Light, all scheduled events featuring HAMC members went ahead as planned.

Audience Opinions: Negative and Positive Labels
A central aspect of the social evaluation of stigmatized organizations lies in the opinions of social audiences about observed organizational conduct (see Fig. 1). During both events under analysis, a wide range of social audiencesincluding the events' organizers, police, journalists, academics, politicians and members of the public-contributed to the public debate about HAMC and its participation.  The festival risks being perceived as WANTING these types within their circle, something I definitely believe that 99% of the Norwegian population does NOT. (Hesjedal, 2013, readers' comments, capitals in original) [HAMC] must be met with counter-arguments-not boycott-in various open fora and other democratic arenas." (Båfjord, 2013, readers' comments) As illustrated in Table 2, the discursive evocation of these diverging standards is reflected in contrasting depictions of HAMC. In the light of the analytical distinction between ethics and morality (see note 2), we thereby distinguish between positive and negative evaluations with either an ethical basis (top panel of Table 2) or a moral basis (bottom panel of Table 2). Such a separation makes visible how collective values and norms are articulated and activated by references to the boundaries that safeguard the community and its social order. This constitutes the ethical basis for social evaluations. It also stresses how individual, subjective convictions are used to justify one's statements, actions and assessment of tainted organizations. This constitutes the moral basis for social evaluations. It is clear from Table 2 that statements with ethical as well as moral bases were employed extensively during both events, and we build on both types of accounts in our discussion below. Table 2 first of all indicates that the opponents of HAMC's participation displayed substantial overlap in their expressed opinions across both events. Curator 2 at KHM, for instance, told us that "[HAMC] is undoubtedly an organized criminal organization, both internationally and nationally" (Interview), while a Member of Parliament stated in a national newspaper that "reading the judgments against over 70% of HAMC members [paints] a picture with a lot of crime, violence and threats" (cited in Aftenposten, 28.11.2013). Similar statements in Table 2 highlight the extensive and repeated references to HAMC members as 'criminals' and 'violent' individuals. The organization itself was mostly referred to as a 'criminal organization' and an 'organized criminal network'. By imposing the label 'criminal' onto HAMC and its members, opposing social audiences portrayed the organization as "synonymous with everything that is wrong" (Marcel Leliënh of, Interview).
This strategic use of labels with strong negative connotations contrasts sharply with the labels brought forward by, for instance, several supporting respondents from academia and the arts. Their accounts offered a neutral categorization of HAMC members (though not necessarily of HAMC itself). Curator 1 at KHM told us that s/he wanted to "relate to them as human beings, not as potential criminals" (Interview). A similar sentiment was brought forward by the spokesperson of Nordic Light who argued that "we are entirely impartial; we only relate to the photographs" (cited in Tidens Krav, 22.01.2013). In keeping with the perceived importance of maintaining impartiality, Table 2 indicates that significant stress was placed on the idea that HAMC members should be treated as individual human beings rather than a homogenous group of criminals:   Nordic Light "These are in their essence not evil people." (Interview, Creative director Nordic Light) "In fact, not all HAMC members are criminals" (Båfjord, 2013, readers' comments) "These guys are not exactly mom's best kids." (Creative director Nordic Light, Interview) "I think it is documented beyond any doubt that [HAMC] is full of heavy criminals. I think there are people in HAMC who are capable of doing anything-and who have done it." (Board member Nordic Light, Letter to Editor 1) "Crime by MCs is on the rise. HAMC is not just a club for motorcycle enthusiasts." (Member of Parliament, cited in Botten, 2013b)" "It is entirely unacceptable to me to be part of something that can shine a positive light on a criminal organization." (Expert in white-collar crime, cited in Joakimsen, 2013) "It is reprehensible that Nordic Light invites a criminal organization." (Expert in white-collar crime, cited in Lillegård, 2013) "The essence and core of HAMC is amphetamins and money laundering." (Båfjord, 2013, readers' comments) "Organized crime, murder, extortion, threats, drug trafficking are not even close to my concept of culture." (Båfjord, 2013, readers' comments) "Here the festival risks being perceived as WANTING these types within their circle." (Hesjedal, 2013, readers' comments) "[HAMC] is one of three well-established criminal networks characterised by violence, murder, drugs, prostitution, and so on." (Hesjedal, 2013, readers' comments) "When HAMC, Al Qaida, neo-Nazis, closed religious communities engage in marketing, it will become very pathetic and dull." (Hesjedal, 2013, readers' comments) "That this Leliënhof has managed to get a book published with mediocre photographs and act as a useful idiot for a criminal organization, we can't change that." (Tolpinrud, 2013a(Tolpinrud, , 2013b, readers' comments) "Well known that HAMC reflects a heavily criminal environment." (Volunteer at Nordic Light) KHM "I want to relate to them as human beings, not as potential criminals" (Interview, Curator 1 KHM) "The stereotypical Easy Rider figure and the idea of a free life outside the established society is for many a strong symbol of freedom." (Website KHM "I would call them modern Vikings. They construct their own way of life." (Curator 1 KHM, cited in Uniforum 16.5.2014) "The Dalai Lama and HAMC are in some sense two sides of the same coin, because they stand up to power without being afraid to say anything else than what kind-hearted citizens do." (professor of social anthropology, public debate at KHM) "Bikers become a very important corrective, and a reminder that there are other ways to live life." (professor of social anthropology, public debate at KHM) "Easy Rider stereotype." (Lecturer, cited in Sandsmark, 2014) "This is undoubtedly an organized criminal organization, both internationally and nationally." (Curator 2 KHM, Interview) "I would think that there have never been so many people with criminal backgrounds inside a Norwegian museum" (professor of social anthropology, public debate at KHM) "By reading the judgments against over 70% of the members, we get a picture with a lot of crime, violence and threats." (Member of Parliament, cited in Aftenposten, 28.11.2013) "[HAMC maintains] a cynical and anti-social view of society, where fear is used to maintain internal justice in the clubs, violence is used against competing clubs, and crime is used to gain status." (Member of Parliament, Letter to Editor) The notion of 'freedom' also triggered references to Easy Rider and Viking stereotypes as labels with positive connotations to describe HAMC members (though, again, not the organization). KHM's website, for instance, stated that the "stereotypical Easy Rider figure (…) is for many a strong symbol of freedom", while Curator 1 at KHM argued that "I would call [HAMC members] modern Vikings. They construct their own way of life" (cited in Uniforum, 16.5.2014). Overall, our findings indicate that audiences use simple and recognizable labels to support their claims about the (in)acceptability of social actors. This reflects the fact that labels are a powerful means to signify membership within a particular category (Ashforth & Humphrey, 1997;Granqvist et al., 2013). They are not only characterized by their explicit meaning, but often also carry an implicit meaning or 'connotation' linked to their ethical or moral basis (Becker, 1963). Hence, they represent an opportunity to convey both explicit and implicit understandings, which claims-makers can-and do-exploit to express their opinions about social actors. Yet, importantly, the mere use of labels can have significant ethical implications. Warren and Laufer (2009), for instance, illustrate how the practice of labeling countries as 'corrupt' has self-fulfilling effects by reducing others' willingness to invest there (thus helping to maintain the corrupt status quo).
Psychological research likewise suggests that social interactions are affected negatively when a stigmatizing label is attributed to interaction partners (Sibicky & Dovidio, 1986). More recently, Cho (2015) highlights that the reverse may also happen, since positive 'sustainability' labeling causes an increase in consumer's product evaluations. Labels and labelling thus are not innocuous from a business ethics perspective, and can have a major impact on the subsequent actions and statements of social actors (see also below).
We should note at this point that HAMC's self-presentation as an 'outlaw' organization (see above) appears to have influenced how it was perceived, evaluated and described by other audiences. Marcel Leliënhof, for instance, stated that: "HAMC does not want to be perceived as a sewing club, and many of the members have lived hard lives" (cited in Kjentfolk, 21.5.2014), while Curator 2 at KHM stressed that: "they have a different view of law and order in society than most people" (Interview). The same happened during the public debate at KHM: "[HAMC] has chosen to stand on the sidelines of the system" (professor of social anthropology, public debate at KHM). Unfortunately, we are unable to engage in a more in-depth exploration of when, how and why an organization's self-presentation affects the opinions of other actors. One way to approach this would be by adding at least one case where the involved organization does not embrace transgressive behaviour and celebrate its own stigma. While this was not available to us, such a comparative approach constitutes an important extension to our work in future research.

Labelling Contest: Moral Panic and Moral Patronage
Although differences of opinion are a predictable and uncontroversial observation, our findings show that these opposing evaluations led to a process of contestation between diverse sets of claims-makers (Fig. 1). This ensuing labelling contest was characterized by the calculated and deliberate use of forceful symbols and labels. Cohen's (1972) concept of moral panic is of central relevance here. Cohen (1972, p. 9) maintains that moral panic arises when an "episode, condition, person or group of persons (…) [is] defined as a threat to societal values and interests". This is particularly likely when a person or group is perceived as posing a menace to society in the eyes, and according to the understandings, of influential audiences (such as social and political elites or the media; Cohen, 2002, p. xxvii). Under such conditions, Cohen (cited in Clegg, 2009, p. 319) argues that stakeholders "create stylized and stereotypical representations, raise moral fears and pronounce judgment". Hence, any episode of moral panic entails a social actor being deprived of individuality by stereotyping, categorization and caricature. Targeted actors are labelled based on negatively evaluated features (e.g., 'criminal') that reveal a conflict with the standards and norms of the evaluating audience (Devers et al., 2009;Mishina & Devers, 2012). A moral panic thereby directly exploits the notion that fear-including the fear of negative labelling-triggers "conformity and a cognitive constriction", which leads to the "reproduction of traditional practices" (Gill & Burrow, 2018, p. 445).
Our analysis indicates that HAMC's proposed participation in both events induced such moral panic. It triggered claims among media, politicians and police labelling HAMC as a dangerous deviant and a direct threat to public order.
[HAMC maintains] a cynical and anti-social view of society, where fear is used to maintain internal justice in the clubs, violence is used against competing clubs, and crime is used to gain status. (Member of Parliament, Letter to Editor) In the last two years alone, about 20 members of HAMC have been convicted. The convictions include aggravated violence, rape, serious drug crimes and threats. (Tidens Krav, 24.1.2013) The essence and core of HAMC is amphetamins and money laundering. (Båfjord, 2013, readers' comments) This highlights that the symbols and labels employed during moral panic are pushed to extremes (Ashforth, 2019;Cohen, 2002). In our setting, these included references to "ISIS terrorists" (news editor of the local newspaper, Interview), the Balkan mafia (Letter to Editor 9), or the perpetrator of the terrorist attack in Norway on 22 July 2011 (Letter to Editor 5; Båfjord, 2013, readers' comments). Such extreme labels are intended to draw strong boundaries and justify drastic action (e.g., revoking financial support), leading to a metaphorical demonization of the social actor (Pontikes et al., 2010).
Nonetheless, this moral panic was countered by a discourse of freedom and sovereignty. A professor of social anthropology, for instance, argued during the public debate at KHM that "the Dalai Lama and HAMC are in some sense two sides of the same coin, because they stand up to power without being afraid". As mentioned previously, Curator 1 at KHM similarly maintained that HAMC members could be viewed as "modern Vikings" since they "construct their own way of life" (cited in Uniforum, 16.5.2014). With symbols and labels thus pushed towards extreme positive depictions, this can be seen as a form of 'moral patronage'-a mirrorimage of moral panic. By linking HAMC to ideal representations of freedom (not just freedom of expression, but freedom in general), moral patronage in our setting appears to elevate HAMC and its members symbolically to the state of near-normality.
People are people. When you get to know people, whether they are a politician, an actor or an HAMC member, most of them are quite ordinary. (Marcel Leliënhof, TV2 God Morgen Norge, 3.12.2012). They live ordinary A4 lives like the rest of us. (Marcel Leliënhof, cited in Ekeland, 2013) The idea of a free life outside the established society.

(Website KHM)
Although HAMC is still presented as marginalized, and its members as societal outsiders, this is given a positive connotation by linking it to individuality and autonomy. The organization becomes a symbol of freedom rather than a threat to social order. 6 Such claims reflect the ethical position that deviance must be tolerated in a modern, rational and democratic society. That is, abstract ideals of freedom and justice must include 'deviants', and social order can only be preserved through such tolerance. Clearly, moral patronage in this sense still conveys the moral superiority of the bestowing audiences, and thereby shows distinct traces of paternalism.
Both moral panic and moral patronage reflect the "initiative taken by someone within the community or acting upon [its ethical standards]" (Hazard, 1995, p. 457). In other words, they express how social audiences react to specific situations as a function of these audiences' ethical standards. This is consistent with evidence in the business ethics literature that social actors "primarily view ethical scenarios through their understanding of the provisions of their own standards or codes of ethics" (Claypool et al. 1990, p. 704). We thereby observe that moral panic is triggered when a legalistic ethics dominates, whereas moral patronage arises when a democratic ethics takes centre stage. Indeed, as illustrated in Table A2 in the Online Appendix and elaborated upon in the next section, social audiences with a legalistic ethics were the leading force at Nordic Light, whereas audiences with a democratic ethics commanded proceedings at KHM.
Furthermore, our findings in this section confirm Ashforth's (2019, pp. 23-24) theoretical proposition that "a sense of moral superiority can fuel (…) over-the-top behavior", which may "veer to the extremes: good or bad". While using symbols and labels, by construction, seeks to influence the societal perception of targeted organizations (see above), veering towards extremes will incidentally, but unavoidably, promote a destructive social climate. Our analysis thereby links to a large literature on the relation between moral absolutism-defined as the "tendency to engage in rigid, 'black-and-white' moral thinking in terms of others' behavior" (Moss & O'Connor, 2020, p. 7)-and support for extreme interventions such as violence or castigating opponents as morally decrepit (e.g., Giner-Sorolla et al., 2011;Moss & O'Connor, 2020). From a business ethics perspective, this connection raises important questions on whether-and, if so, when-it can be acceptable to take on extreme inclusionary/exclusionary stances (Bayer, 2008;Pozner, 2008;Warren, 2007). Furthermore, such questions arguably take on increased relevance since going to extremes may have wide-ranging implications beyond the originally targeted organization(s). The rapidly developing literature on stigma-by-association in organizational settings indeed highlights that negative labels have contagious effects (Kulik et al., 2008;Kvåle & Murdoch, 2021), which would encourage the social consequences of extreme stances to spread far and wide.

Power Context
Despite the general similarity in depictions of HAMC at Nordic Light and KHM, the outcomes were very different in terms of HAMC's participation (see above). What can explain this difference? What determines the successful development of, or resistance to, instances of moral panic/ patronage? Our respondents suggest a prominent role for the power constellation between opposing claims-makers. Marcel Leliënhof, for instance, told us that "the Museum of Cultural History is too powerful an institution to be pressured" (Interview), while a Member of the Nordic Light board argued that boundaries: are obviously drawn by the police, in alliance with the local newspaper, sponsors and squeamish politicians on several levels (…). There is a lot of power in that square. (Letter to Editor 1) Hence, the balance of power between social audiences was perceived to be crucial for their ability to take a stand, defend their position, and make it count. Table 3 provides an overview of the symbolic, material and normative sources of power available to representatives of art, state and academia across our two cases. In this section, we discuss how (lack of) access to these sources of power determines the ability of a given audience to authoritatively speak for-or against-a tainted organization ( Fig. 1; Shadnam et al. 2020). Table 3 first of all indicates a broadly-based agreement that Nordic Light and KHM as a matter of principle "must have artistic freedom" (Editorial, Fotografi 12.4.2013). There was formal as well as legal support for the fact that choosing the "content for such an exhibition [at KHM] (…) falls clearly under the university's authority to assess" (Secretary of State in the Department of Education, cited in Mjaaland & Helsingeng, 2013). This illustrates the strong adherence to the principle of institutional autonomy of arts and academia in Norway, which lends these actors important symbolic and normative sources of power (Table 3). Even so, it does not provide equal power to representatives of the cultural sector throughout Norway. The managing director of Nordic Light, for instance, stated that "in Oslo they would not even bat an eyelid". The implication is that the power derived from the value of artistic expression is stronger in Oslo (as capital city) compared to small towns such as Kristiansund (confirmed by the creative director of Nordic Light and the chairman of the festival's board).
Although power deriving from the principle of institutional autonomy is important, our findings show that it is insufficient on its own. Material resources matter too. While KHM has financial security within the university as "a separate faculty and thus its own kingdom" (Director KHM, Interview), Nordic Light's limited financial and organizational resources left it vulnerable to financial threats from opposing local politicians and sponsors: As an owner who provides support and grants, you can have expectations. As an owner, the chances of pursuing objectives for the community are stronger. (Chief county administrator, Interview) Afterwards, it became more difficult to get sponsorship funds for the festival. (Board member Nordic Light, Interview) In the context of the Constitution Jubilee, normative sources of power embedded in the Constitution further fortified the position of KHM. The Minister of Education and Research, for instance, stated unequivocally that the Constitutional right to freedom of expression precludes interference with KHM's plans: It is critical that both politicians and citizens participate in the debate on freedom of expression, and I think that the best starting point for such debates are free institutions and free citizens (Document no., 1515(Document no., :93, 2013(Document no., -2014. Power thus emerges from our analysis as a double-edged sword. In contested situations characterized by diverse sets of claims-makers, it can have both positive and negative implications for the social evaluation of tainted organizations. This observed importance of power inequalities and access to (multiple) sources of power relates to Nietzsche's view of power as "a capacity to define reality" (Haugaard & Clegg, 2009, p. 2). Nietzsche maintains that a claims-maker with the ability to "define the real and the moral" holds the keys to influence societal perceptions of (un)acceptable organizational practices (Haugaard & Clegg, 2009, p. 2). While the previous section thus highlighted which ethical standards gained the upper hand during each event, power  "The celebration of the Constitution is a celebration of freedom of expression, and I think it is very positive that universities-including their musea-and university colleges participate in this celebration." (Document no., 1515(Document no., :93, 2013(Document no., -2014 "Universities or university colleges may not be instructed regarding the academic content of their teaching and the content of research or artistic or academic development work." (Section 1-5, University and university college Act) "And then they drew the freedom-of-expression card, and there is almost nothing that can top that at a university." (Curator 2 KHM, Interview) inequalities help explain why particular ethical standards came to dominate the labelling contest in a given context. Nonetheless, since power is often misused and can facilitate bias by impairing individuals' cognition and judgments, the presence and importance of power inequalities for social evaluations has an important additional ethical dimension. As discussed in Albrecht et al. (2015, p. 807), power can indeed lead to "flawed assessments of others' interests" or the use of "stereotypes in forming opinions of others" (see also Keltner & Robinson, 1997;Goodwin et al., 1998). Such (mis)uses of power naturally raise critical ethical concerns about the normatively appropriate apportionment of the consequences of (negative) social evaluations (Moberg, 1994;Warren, 2007), as well as the need-and possibilities-to empower weaker stakeholders (Civera et al., 2019;Hess, 2007). We return to this in the next section. Clearly, there is some risk of circularity when inferring social audiences' power from their observed influence over social evaluation processes, particularly when power configurations in society depend on the broader social context. In our setting, however, the relative power of the various claims-makers was in large part defined ex ante. The Norwegian Constitution and University of Oslo are much more powerful than a photography festival and artists, with politicians, the police and media located in-between (Table 3).

Ethical Implications
Our analysis highlights a key role for contestation and power dynamics in the social evaluation of stigmatized organizations. In this section, we discuss four main ethical aspects of these findings.
First, our analysis focuses on the opinions expressed by social audiences about a tainted actor. However, any process of contestation between multiple claims-makers entails the risk that opposing sides of the debate negatively evaluate each other. A key ethical concern thus relates to the negative labelling of opposing audiences (alongside any labelling of the tainted actor). Recent work by Thomson et al. (2018), for instance, suggests that community polarization about the presence of polluting industries leads to labels being attached to opposing social audiences. That is, "some supporters devalued the stance of the 'nonsupporters' position or labelled them as 'liars', 'greenies' or having a 'rent-acrowd' attitude" (Thomson et al., 2018, p. 223). Our data likewise indicate such labelling spillovers onto opposing social audiences. For instance, opponents of HAMC's participation labelled the photographer and events' organizers as "head-shakingly naïve" (Werp, 2014) and "useful idiots" (e.g., Curator 1 KHM, Interview; Løberg, 2013, readers' comments;Tolpinrud, 2013aTolpinrud, , 2013b. While such effects are reminiscent of stigma-by-association (Kulik et al., 2008;Kvåle & Murdoch, 2021), they are conceptually distinct in two ways. That is, the content of the labels attached to opposing social audiences differs from that of the tainted organization, and this content varies depending on the target (as well as targeting) audience. From a business ethics perspective, this observation calls for a better understanding of the 'prosecutorial incentives' leading to such behaviour (Warren, 2007). From the perspective of organization stigma theory, it brings to light a distinct type of spillover effect from organizational stigma, and highlights a need to distinguish between multiple sets of targets in future research (i.e. tainted actors and opposing audiences).
Second, the contested nature of social evaluations provides actors with an incentive to be fast and direct when advancing their evaluation of tainted actors. More assertive positions offer greater benefits in terms of gaining the upper hand in contests. These incentives invoke ethical questions about (commonly used) strategies that intentionally exploit feelings of shame, guilt or other "primitive and destructive emotions" (Burris, 2008, p. 475;Kvåle & Murdoch, 2021). Such questions are particularly important, because social evaluation of stigmatized organizations implies a process of identity (de)construction by others (Ashforth, 2019;Goffman, 1963;Paetzold et al., 2008), which can have farreaching consequences. Answering these ethical questions requires further research into the factors and perceptions that influence how social audiences process discreditable actions.
Third, our analysis indicates that the power constellation between claims-makers determines whose claims gain the upper hand. This not only raises important questions regarding the nature and meaning of the existing social order. The possibility of strategic manipulation by the strongest partybased on a desire to strengthen one's own legitimacy and/or moral superiority (Shadnam et al., 2020), or to promote specific societal interests at the expense of others-can also provide a form of social control that (re)produces inequality and injustice (Tyler, 2020;Tyler & Slater, 2018). Since social evaluations are never impartial and will always be perceived as arbitrary or burdensome by at least some audiences, this may seem inevitable. Nonetheless, with the burden of stigma effectively imposed by the most powerful audiences, concerns regarding the (in)equity of these burdens should be carefully considered (Moberg, 1994;Warren, 2007). Moreover, assessing the possibilities for empowering weaker stakeholders is critical to shed light on the mechanisms behind the arbitrariness and inequality in social evaluations (Bondy & Charles, 2020;Demuijnck & Fasterling, 2016), as well as the ethical consequences of it.
Finally, our findings provide new insights for managers aiming to understand when and how interactions between stakeholders can exacerbate positive and negative evaluations of their organization, when stakeholders will cut ties, and how they may act after public endorsement has been withdrawn (Norheim-Hansen & Meschi, 2020). By increasing our understanding of how individuals and organizations "think and act when faced with ethical situations" (O'Fallon & Butterfield, 2005, pp. 375-376), our work links to the literature on ethical decision-making. This literature thus far predominantly studies moral judgments-i.e. "deciding which course of action is morally right" (O'Fallon & Butterfield, 2005, p. 376)-in terms of social actors' own behaviour. Yet, our findings stress the relevance of extending this field of enquiry to include also moral judgments about others' behaviour.

Conclusion
This article aimed to answer two main questions. The first question asked how interactions between social audiences shape the societal perception of tainted organizations. Such analysis of the nature and characteristics of interactions between social actors-while taking the stigmatized organization itself out of the equation-has not previously been conducted. Our findings show that social audiences act as claims-makers and use simple and recognizable labels to evaluate the (in)acceptability of social actors. This results in a labelling contest where symbols and labels are pushed to extremes (Ashforth, 2019)-reflective of either moral panic (when negative) or moral patronage (when positive). These findings contribute to organizational stigma theory in two ways.
First, labelling theory has been central to scholarship on organizational stigma at least since the path-breaking work by Devers et al. (2009). Yet, stigmatizing labels are generally portrayed as uncontested 'identity markers' (Elsbach, 2004) that are applied as "an act of social control" (Thomson, 2018, p. 195). We push this idea further by allowing for multiple, opposing labels that reflect audiences' diverging perspectives. A key insight from this broader approach is that a struggle for dominance can ensue whereby audiences' labels are pushed to extremes-positive or negative. Second, scholarship on organizational stigma predominantly studies the interaction between a stigma 'target' and its evaluating 'audience' (Hudson & Okhuysen, 2009;Paetzold et al., 2008;Reuber & Morgan-Thomas, 2019). Our analysis shifts focus to the interaction between multiple and heterogenous audiences (Shadnam et al., 2020), which helps to reveal how the notion of stigmatized organizations is negotiated and handled in social practice.
Our second question dealt with the role of audiences' power in social evaluation processes (Lawrence, 2008;Munir, 2015). Previous studies highlight the role of power (im)balance between the tainted organization and its stigmatizing audience(s). Hudson and Okhuysen (2009, p. 150), for instance, argue that bathhouses face varying levels of stigmatization because the "power of the audiences that stigmatize bathhouses varies as well". In similar vein, Demuijnck and Fasterling (2016, p. 680) maintain that "a 'social licensor' must ideally have some power to grant and reject the Social Licence to Operate". 7 Our findings, unsurprisingly, confirm that claims-makers with greater symbolic, material or normative sources of power are more influential in the social evaluation of an organization. More interestingly, our analysis emphasizes how the broader societal context determines the power constellation between claimsmakers. This elicits the insight that, depending on who constitutes the most powerful audience in a given context, power can have both dark and bright sides in social evaluation processes. In our Norwegian setting, it thereby appears that a context characterized by principles of democratic ethics awards relatively more power to academic institutions (as at KHM), whereas the reverse holds when the context favours a legalistic ethics focused on law and order (as at Nordic Light).
Finally, our study analyzed events involving a (selfdeclared) marginal organization that is extreme in its level of stigma. While this generates a strong setting to investigate the relationships of interest, it naturally raises questions regarding the transferability of our findings. Nonetheless, we believe that our inferences about how instances of moral panic or moral patronage develop depending on contextspecific power constellations are transferable to others settings, social audiences and areas of society. One such setting would be the corporate world, where debates about the ethical implications of corporate actions and social responsibility become increasingly important (Smith & Rönnegard, 2016). The same is true within businesses, where power and ethics often circumscribe the relationship between management and employees (e.g., literature on ethical leadership) as well as between (groups of) employees (Reiley & Jacobs, 2016).
Funding Open access funding provided by University of Bergen (incl Haukeland University Hospital).
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.