OTT0010.1177/2631787720975192Organization TheoryRoulet and Pichler research-article9751922020 View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Apollo

Theory Article

Organization Theory Volume 1: 1–30 Blame Game Theory: © The Author(s) 2020 Article reuse guidelines: sagepub.com/journals-permissions Scapegoating, Whistleblowing DOI:https://doi.org/10.1177/2631787720975192 10.1177/2631787720975192 and Discursive Struggles journals.sagepub.com/home/ott following Accusations of Organizational Misconduct

Thomas J. Roulet and Rasmus Pichler

Abstract Research on organizational misconduct has examined how audiences generate discourses to make sense of behaviour that may transgress the line between right and wrong. However, when organizations are accused of misconduct, the resulting ambiguity also opens opportunities for organizations and their members to generate discourses aimed at deflecting blame. Little is known about how actors who are at risk of being held responsible actively respond to misconduct accusations by engaging in discursive strategies. To address this question, we build on crisis communication and discourse theory to integrate processes of scapegoating and whistleblowing into a holistic model. We develop a blame game theory – conceptualizing the sequence of discursive strategies employed by an organization and its members to strategically shift blame by attributing responsibility to others or denying misconduct. Our model identifies four blame game pathways as a function of two types of ambiguity: moral ambiguity and attributional ambiguity. We highlight accusations of misconduct as pivotal triggering events in the social construction of misconduct. By conceptualizing the discursive dynamics of strategic reactions to accusations of misconduct, our blame game theory contributes to the literature on organizational misconduct and has implications for research on social evaluations.

Keywords blame game, crisis, discourse theory, misconduct, scapegoating, whistleblowing

University of Cambridge, Cambridge, UK

Corresponding author: Thomas J. Roulet, University of Cambridge, Trumpington St, Cambridge, CB2 1TN, UK. Email: [email protected]

Creative Commons Non Commercial CC BY-NC: This article is distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 License (https://creativecommons.org/licenses/by-nc/4.0/) which permits non-commercial use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 Organization Theory

A man may fail many times but he isn’t a ways, and this ambiguity makes it difficult for failure until he begins to blame external audiences to evaluate the accusation somebody else. (Faulkner, 2011). The accused organization therefore has both the incentive and opportunity Jean-Paul Getty (1892–1976), American petrol to try to influence audiences’ evaluation of the tycoon and industrialist situation. In other words, the ambiguity of the situation creates the potential for a strategic An extensive body of research has looked at response by the accused actor. It is important to organizational misconduct – ‘a behavior in or take this potential into account to advance our by an organization [that is judged] to transgress understanding of how the social construction of a line separating right from wrong’ (Greve, misconduct unfolds following an accusation. Palmer, & Pozner, 2010, p. 56). As the line To account for the potential of strategic between right and wrong is blurry and can be responses by actors accused of misconduct, we manipulated, misconduct can be considered to suggest adopting a discursive perspective of mis- be socially constructed (Palmer, 2012). A criti- conduct. Because ‘morality originates from and cal point in this social construction is when an is situated in every day discourse’ (Shadnam & organization is accused of misconduct. When Lawrence, 2011, p. 384), we see discourse as such an accusation is made, it is still to be deter- central to the process of the social construction mined whether misconduct really took place of misconduct. Audiences collectively construct and, if so, who was responsible (Faulkner, misconduct through discourses (Clemente & 2011). The organization’s behaviour is subse- Gabbioneta, 2017). Misconduct is constructed quently evaluated by external audiences – when audiences generate discourses that desig- groups of actors who actively try to make sense nate a behaviour as crossing the line between of a situation (Radoynovska & King, 2019; right and wrong, and consequently as condemn- Roulet, 2020) – and judgements are made about able (Roulet, 2020). After an accusation of its wrongfulness (Mohliver, 2019; Palmer, organizational misconduct has been made – for 2013). This audience evaluation is therefore example, through the media pointing out a trans- often associated with negative outcomes such gression (Palmer, Greenwood, & Smith-Crowe, as a reputational penalty for the accused organi- 2016; Roulet & Clemente, 2018) – there is the zation (Coombs & Holladay, 2006). opportunity for such discourses to emerge (Adut, To date, however, the literature on misconduct 2005). In addition to evaluating whether the has neglected the role of accusations in the social behaviour crossed the line between right and construction of misconduct, even though accusa- wrong – thereby constructing its moral status – tions are pivotal triggering events that expose this these discourses also determine who should be social construction (see Palmer, 2014). We there- held accountable for the misbehaviour and thus fore know little about how the dynamics that attribute responsibility (Allport, 1979). The iden- unfold following an accusation of misconduct tification of responsible parties relies on dis- contribute to the social construction of miscon- courses because of ambiguity in the attributional duct and, in particular, about how the accused process (Jacquart & Antonakis, 2015): external organization itself responds to allegations of mis- and internal audiences can both only make more conduct and influences its social construction. At or less plausible inferences about who is respon- the same time, the nature of the situation follow- sible for misconduct (Johansen, Aggerholm, & ing an accusation of misconduct makes it highly Frandsen, 2012). Therefore, both the moral eval- likely that the accused organization will respond uation of the scrutinized behaviour and the attri- strategically and influence the social construction bution of responsibility for that behaviour will of misconduct (Butler, Serra, & Spagnolo, 2020). originate from discourses that emerge following Accusations of misconduct are highly ambig- the accusation. This characteristic in turn enables uous in that they can be interpreted in multiple the accused actors to produce their own Roulet and Pichler 3 discourses with the aim of influencing the social Longueval, 2020) and whistleblowing (Kenny, construction of misconduct by audiences. 2019). We integrate both into the notion of a We therefore turn our focus to the discursive blame game, seeing them as two sides of the reactions of the accused organization and its same coin, as essentially blame-shifting discur- members following an accusation of misconduct. sive strategies through which organizations and As the ambiguity inherent in such accusations their members react to accusations of miscon- opens space for discursive reactions, organiza- duct. Coombs (2007a) identifies scapegoating as tions at risk of being held responsible commonly one of the primary organizational reactions: that generate discourses that attribute responsibility to is, blaming actors who are not necessarily others in order to avoid reputational penalties responsible for a negative outcome but whom the from being caught misbehaving (Coombs, organization can condemn to deflect blame away 2007a). Unfolding crises can also trigger blaming from itself (Boeker, 1992; Grint, 2010). For processes between organizational members when example, organizations can denounce so-called organizations are held responsible for misconduct ‘rogue’ employees. One example of scapegoat- (Gabriel, 2012; Gephart, 1993). This is because it ing is the case of rogue trader Jerome Kerviel, is not just organizations that suffer a significant who was held responsible for a €4.9 billion trad- reputation penalty when they are seen to be ing loss for Société Générale, one of the major responsible for misconduct (Coombs & Holladay, European banks, in 2007 and 2008. The bank 2006), but individual members can also suffer was facing a €7 billion loss and many other con- adverse outcomes, even after disassociating troversies related to its involvement in the sub- themselves from the tainted organization (Pozner prime markets and its practices. The top & Harris, 2016; Wiesenfeld, Wurthmann, & management asked employees to stand together Hambrick, 2008). This context – in which both with them against the scapegoat. In this sense, the accused organization and its members are at scapegoating is a collective strategy for shifting risk of being blamed – is characterized by differ- blame from the focal collective actor to an indi- ent types of ambiguity: ambiguity regarding the vidual one through the creation of relevant attribution of responsibility (Crocker, Voelkl, discourses. Testa, & Major, 1991) and ambiguity regarding Individuals can also react to an accusation the moral judgement of wrongdoing (Shadnam & made against their organization and take a dis- Lawrence, 2011). Organizations and their mem- cursive position to limit contamination (Moore, bers are both likely to take this ambiguity into Stuart, & Pozner, 2011). Organizational mem- account when deciding to engage in discursive bers can engage in whistleblowing to blame strategies which aim to minimize the conse- wrongdoing on their organization as a whole quences of the blame they are exposed to. We use (Culiberg & Mihelič, 2017; Near & Miceli, the term ‘blame game’ to refer to this collective 1985, 1995). While scapegoating is an organi- and discursive phenomenon of social actors zational strategy for shifting blame to a member instrumentally positioning themselves to protect of the organization, whistleblowing can be initi- themselves and deflect blame. The Cambridge ated by individual members, for their own ben- dictionary defines a blame game as ‘a situation in efit, to strategically shift blame to the which people try to blame each other for some- organization (Choo, Grimm, Horváth, & Nitta, thing bad that has happened.’ We theorize blame 2019). In the context of the 2007–2008 finan- games as sequences of discursive strategies cial crisis, some bankers decided to turn their aimed at deflecting blame, through which organi- backs on their former employers and industry. zations and their members respond to accusations In March 2012, Greg Smith – executive director of misconduct. of the United States equity derivatives business Our blame game theory enables us to bridge at Goldman Sachs – published a resignation let- the gap between two well-studied phenomena in ter as an op-ed in the New York Times. He organization studies: scapegoating (Djabi & de pointed out a change in corporate culture, which 4 Organization Theory had switched from being client-oriented to a organizational misconduct, or the violation of ‘toxic and destructive’ environment where the laws, norms and rules by organizations (Pozner, norm was to ‘rip off’ the clients, often referred 2008). Within this body of research, many argue to in internal emails as ‘the muppets’. that organizational misconduct is socially con- In our model, we theorize different pathways structed (Greve et al., 2010). The line between through which blame games can unfold. To what is right and wrong is blurry and can be achieve this, we develop the concept of ambigu- manipulated by both audiences and perpetrators ity as the enabler and determinant of blame (Palmer, 2012, 2013). A pivotal triggering event games and explore the interactions of different for the social construction of misconduct is the organizational and individual blame game strate- accusation of misconduct, through which behav- gies. In doing so, we show how the blame game iour is identified as wrongful and is exposed to strategies of organizations and their members audience scrutiny. can coexist and succeed each other in a sequence The social construction of misconduct after of events. This enables us to integrate whistle- an accusation is the result of an interaction blowing and scapegoating, traditionally explored between organizations and audiences (Clemente in separate literatures, into a holistic model. We & Gabbioneta, 2017). This interaction relies on ask: what determines the discursive strategies a range of discursive practices (MacLean, employed by organizations and their members to 2008). Discourses are central to this process of deflect blame in the aftermath of accusations of social construction because, as mentioned, we misconduct? And when do blame games reach assume that morality is itself discursively con- an end or a settlement? We identify a variety of structed (Shadnam & Lawrence, 2011), as is blame game sequences and pathways through responsibility (Crocker et al., 1991; Jacquart & which blame is settled, depending on the level Antonakis, 2015). However, most of the and form of ambiguity that is present following research on the social construction of miscon- an accusation of misconduct. Overall, research duct has focused on the role of external audi- on misconduct has to date given little attention to ences rather than on the discourses produced by the role of ambiguity following accusations of the accused organization itself (Clemente & misconduct and the resulting potential for strate- Gabbioneta, 2017; Dewan & Jensen, 2019; gic responses by the accused actors. However, if Roulet, 2019). We therefore have limited these issues are not theorized, then our view of knowledge of how accused organizations and misconduct as being socially constructed will their members discursively react to accusations miss crucial pillars, meaning that the social con- of misconduct. struction of misconduct is theoretically underde- Some indications of this can be found in the veloped. Our model addresses this essential blind literature on crisis communication (Bundy, spot by adopting a discursive perspective of mis- Pfarrer, Short, & Coombs, 2017), particularly conduct and by accommodating prior research situated crisis communication theory (Coombs on crisis communication and whistleblowing. & Holladay, 2002). Organizations deal with The resulting blame game theory contributes accusations of misconduct through crisis com- specifically to the literature on organizational munication (Bundy & Pfarrer, 2015; Coombs, misconduct and more broadly to the literature on 2007b). When such accusations emerge, audi- social evaluations. ences often consider the crisis to have been avoidable, resulting in significant reputational A Discursive Perspective of threat for the organization (Coombs & Holladay, Organizational Misconduct 2006). Because of this threat, the organization will engage in discursive strategies, particularly Encouraged by high-profile empirical cases from scapegoating, to shift the blame (Coombs, Enron to Volkswagen, organization theorists have 2007a). Organizational members also suffer increasingly been looking at the phenomenon of adverse outcomes (Pozner & Harris, 2016) when Roulet and Pichler 5 they are associated with misbehaving organiza- (Radoynovska & King, 2019) where this sense- tions (Wiesenfeld et al., 2008). Therefore, organ- making is prompted by the disruptive and sur- izations and their members will attempt to avoid prising event of an accusation of misconduct. and deflect blame to avoid incurring social pen- The audiences that engage in sensemaking fol- alties. However, the literatures on crisis commu- lowing an accusation can be damaged parties, nication and organizational misconduct have customers or suppliers, governments, regula- both ignored the discursive strategies of organi- tors, the media or simply the general public zational members and how they interact with (Clemente, Durand & Porac, 2016; Palmer, organizational strategies to shift blame. When 2013). Some of these audiences are likely to be organizations are accused of misconduct, crises more invested than others in trying to attribute can trigger blaming processes between organiza- responsibility for the misconduct and thus are tional members (Gephart, 1993). We build on the more active in their sensemaking, especially crisis communication literature (Coombs, 2007a, those with a higher stake in the accusation of 2007b) to develop a discursive perspective of misconduct (e.g. if they were harmed by the organizational misconduct which can integrate misconduct or have ties to the accused organi- the discursive reactions of both organizations zation, etc.) (Palmer, 2012). Other audiences and their members. are motivated to make sense of organizational Because adverse outcomes trigger mecha- misconduct simply ‘to uphold [their] moral nisms of causal search (Weiner, 1986; Wong & standards’ (Barnett, 2014, p. 69). Weiner, 1981), crisis communication theorists Following an accusation of misconduct, argue that, in the aftermath of misconduct, audiences and accused actors are therefore organizations will try to shift responsibility to likely to engage in a struggle over the interpre- avoid reputational penalties (Coombs & tation of the misconduct, a struggle which arises Holladay, 2002). Major crises tend to generate from the ambiguity inherent in the context of sensemaking processes (Bail, 2015; Boudes & organizational misconduct. These discursive Laroche, 2009). Under such conditions, there efforts take place in a dedicated discursive may be a need to attribute blame for misconduct space (Grant & Hardy, 2004). The struggle to find a satisfactory explanation for it and to unfolds through the generation of discourse hold actors responsible (Bucher, 1957), particu- (Livesey, 2001), i.e. through the audiences larly as it is almost impossible for organizations involved and the actors accused ‘producing, to avoid being held responsible because mis- distributing, and consuming texts’ (Maguire & conduct is by nature deliberate (Coombs & Hardy, 2012, p. 234) related to the accusation, Holladay, 2002). Consequently, crises due to the alleged involvement of accused actors in the misconduct are commonly managed by com- misconduct, and the nature of the misconduct municating and producing discourses to attrib- itself. A discursive perspective acknowledges ute responsibility (Coombs, 2007a). The the importance of text and language in the purpose of these discursive reactions is defence social construction of organizational phenom- (Ashforth & Lee, 1990). We next define the ena, such that ‘language [is] constitutive and conceptual pillars of our discursive perspective constructive of reality’ (Hardy, Lawrence, & of organizational misconduct. Grant, 2005; Phillips & Oswick, 2012, p. 445). The discourse generated by audiences and the accused actors thus shapes the meaning of the The emergence of a discursive space misconduct. It is ‘embodied in sets of texts that to make sense of an accusation of come in a wide variety of forms, including writ- misconduct ten documents, speech acts, pictures, and sym- bols’ (Hardy et al., 2005, p. 59). The texts that We anchor our model in the view that audiences are produced within the discursive space com- actively make sense of their social context prise internal and external communications 6 Organization Theory

(Frandsen & Johansen, 2011), including press Misconduct can be interpreted and assigned releases and official statements by spokesper- meaning in different ways (MacLean, 2008). sons, as well as interviews, news articles and Consequently, in the discursive struggle follow- investigation reports (Coombs, 2007b). ing an accusation of misconduct (Livesey, 2001), The struggle for meaning in the discursive accused actors will try to take discursive posi- space naturally revolves around a set of critical tions (Hardy & Phillips, 1999) to deflect blame. questions: what really happened, how bad was it, Actors suspected of misconduct can shift the dis- and who was responsible for it? (Livesey, 2001). courses by ‘reorganizing existing information Those questions are subject to significant discur- and conventions associated with the prior sive struggle. Given the possibility of alternative schema’ (Werner & Cornelissen, 2014, p. 1456). and competing accounts of the same event, In other words, they can attempt to assign new ‘[conduct] is not, of course, intrinsically or auto- meaning to the information given to audiences matically to be regarded as a violation, a trans- about the organizational misdeed to reduce or gression, or as reprehensible: It is constituted as escape blame. In this way, actors can attempt to such [. . .] through accounts of conduct and thus manipulate their discursive position through the in a general way through discourse’ (Drew 1998, strategic use of discourse. For example, a sus- p. 312). In this discursive struggle around mis- pected organization or its members may conduct (Livesey, 2001), audiences and accused acknowledge the evidence of misconduct identi- actors take ‘discursive positions’ (Hardy & fied by audiences but attribute the responsibility Phillips, 1999): they are discursively related to to another actor. Reusing existing elements of the accusation event through their own and oth- established discourses gives credibility to their ers’ discourse, for example, as a ‘victim’ or a blame game discourses (Hearit, 1995). ‘perpetrator’ of misconduct. This strategic production of discourse involves actors drawing on established dis- Discursive positions as part of the courses to frame the misconduct and their own role in it in a certain way (Cornelissen & blame game Werner, 2014). As actors frame the organiza- As we have pointed out, in a crisis situation, tional misconduct in different ways, their dis- actors will attempt to shift the blame to another cursive strategy can be a reaction to existing actor to avoid being held responsible themselves frames (Koopmans & Olzak, 2004). In other (Coombs, 2007b), particularly by generating dis- words, actors generate discourses as a function courses (Grant, Keenoy, & Oswick, 2001). Such of existing discursive dynamics to position discursive positions can be taken both by organi- themselves in a way that gives credibility to zations and their members. When organizations their perspective (Lawrence, Phillips, & Hardy, hold some of their own members responsible, 1999). In the case of misconduct, suspected attributions can be cascaded from one set of organizations and their members will generate actors to another (Bonardi & Keim, 2005). discourses that have a degree of compatibility Discourses produced in the aftermath of a crisis and credibility with existing discourses, and are thus often attribute responsibility to members of often cast as a reaction to those discourses. the organization itself (Coombs, 2007a). For example, in the case of the Volkswagen emission How ambiguity creates opportunities scandal, the interviews given by top executives for blame games in the aftermath of (which constitute discourses) often involved misconduct shifting the blame to lower-level employees. In response, organizational members may engage Evaluations of accusations of misconduct are in ‘dissenting discourses’, deviating at the indi- highly ambiguous because wrongdoing is socially vidual level from organizational communication constructed (Greve et al., 2010). Ambiguity refers (Teo & Caspersz, 2011). to a state where audiences struggle to establish Roulet and Pichler 7 meaningful links between events and objects in a with moral ambiguity (see also Green, 2004), social situation, and thus are unable to define the i.e. there are multiple ways of interpreting the situation (Ball-Rokeach, 1973). Ambiguity is morality of the behaviour that an accusation therefore the quality of a situation being open to of misconduct points out. Green (2004) more than one interpretation (Sennet, 2016). In examines cases of morally ambiguous organi- the context of accusations of misconduct and the zational misconduct and argues that there is resulting discursive struggle, we consider two moral ambiguity when audiences disagree types of ambiguity: attributional ambiguity about whether an action is morally wrong or (Crocker et al., 1991; Jacquart & Antonakis, not and are ready to debate this stance. The 2015), and moral ambiguity (Shadnam & debate around working conditions in Amazon Lawrence, 2011). warehouses in the United Kingdom is a typi- When attributing an outcome to a cause, cal case that demonstrates the difficulty audi- audiences may struggle to establish causal attri- ences can have in morally evaluating bution (Powell, Lovallo, & Caringal, 2006) and behaviour called out in an accusation. In this responsibility for misconduct cannot always be case, audiences disagree about whether clearly attributed to the organization or to indi- Amazon is acting in a morally condemnable vidual members. In complex organizations manner or is just putting a high level of pres- involving intricate processes, multiple actors sure on its employees to run an efficient oper- may be at fault and may not all be fully aware of ation. Amazon denies any misconduct, stating the consequences of their actions (Dahlin, that they ‘are proud of our safety record and Chuang, & Roulet, 2018). This creates attribu- thousands of Amazonians work hard every tional ambiguity (Jacquart & Antonakis, 2015), day innovating’.1 that is, there are multiple ways of attributing The range of moral ambiguity exemplified by responsibility for misconduct to actors follow- these two cases – a relatively unambiguous fraud ing an accusation of misconduct. The case of in the Volkswagen case and a more ambiguous the Volkswagen emission scandal shows how accusation of a harmful working environment in complex it can be for audiences to identify who the case of Amazon – illustrates the broad range is responsible (Rhodes, 2016). The ambiguity of accusations of misconduct that are within the of the situation created the conditions for a scope of our theorizing. We base our theorizing blame game where, for example, before the on the established definition of misconduct in the American House Committee for Energy and literature (Greve et al., 2010) and thus theorize Commerce, the US CEO Michael Horn accused about all instances where an organization is software engineers of being solely responsible accused of transgressing the line between right for the misconduct. and wrong (Palmer, 2012). In the most basic The second type of ambiguity stems from sense, this transgression is a behaviour that is the socially constructed nature of the line attributable to an actor and is perceived by audi- between right and wrong which demarcates ences as morally violating norms. The nature of misconduct (Greve et al., 2010); audiences the potential transgression – the specific norm contribute to and struggle over defining what violated and how it was supposedly violated – is right and wrong. While moral norms help along with the information available to audi- to determine whether an organizational ences at the time of the accusation will lead to behaviour can be considered as misconduct variation in the levels of moral and attributional (Roulet, 2019, 2020), these norms vary over ambiguity associated with different accusations time and between contexts (Palmer, 2012). At of misconduct. These varying levels of ambigu- the same time, there are differences in per- ity surrounding accusations of organizational spective and in the interpretation of norms misconduct create an opportunity for actors to (Shadnam & Lawrence, 2011). As a result, construct attributions discursively (Clemente & accusations of misconduct are associated Gabbioneta, 2017). 8 Organization Theory

Discursive Strategies others. Below, we theorize scapegoating and Constituting Blame Games whistleblowing as two key discursive strategies that an accused organization and its members Following an accusation of misconduct, moral can use to shift blame. and attributional ambiguity characterize the situ- ation in a way that leads to the use of blame games. The word ‘games’ is used as a metaphor Scapegoating as an organizational (Cornelissen, 2005) to signal the agency that discursive strategy accused individual and organizational actors have to discursively maneuvre a situation of Scapegoating is one of the key discursive strate- proven or alleged organizational misconduct. We gies used by organizations to shift blame fol- assume that those actors instrumentally adopt lowing a crisis (Coombs, 2007a). The term discursive positions in this context (Butler et al., ‘scapegoat’ was commonly popularized by 2020) and that they do so by assessing elements French social theorist René Girard (Girard, of the social context, in particular ambiguity, that 1982), who initially derived the term from the might make those positions credible. work of Burke (1940). The scapegoat mecha- We therefore expect that actors adapt their nism refers to a process of collectively blaming choice of discursive strategy depending on the an individual actor (Girard, 1982) which is moral and attributional ambiguity that is availa- inherently discursive in that it relies on the pro- ble in the discursive space following the accusa- duction of texts that support the blame-shifting tion of misconduct. Within this space, accused effort (Boeker, 1992). Gamson and Scotch actors can produce discourses that may deviate (1964) were the first in organization studies to from the reality of facts and take liberty with the use the term to explain why baseball team man- truth. This deviation is made possible by moral agers get blamed for their teams’ poor athletic and attributional ambiguity. Some accusations performance when they have very little respon- are more ambiguous than others. Actors attempt sibility. Scapegoating is employed by groups or to adjust to the structure of the discursive space organizations which shift responsibility to an by choosing discursive strategies that align with expendable member or members to save them and make use of the different levels of moral and from condemnation and to shift blame (Hargie, attributional ambiguity. In other words, ambigu- Stapleton, & Tourish, 2010) and in order to face ity enables discourse production strategies that up to the ‘demands of opposition groups’ serve instrumental purposes (Leitch & (Bonazzi, 1983, p. 4). When baseball team Davenport, 2007). Moral and attributional ambi- managers are scapegoated, the fans end up guity create the opportunity in the discursive being appeased and can regain hope of future space for actors to produce discourses strategi- success (Gamson & Scotch, 1964). A wide cally and, in this sense, they constitute a key range of work also shows that top executives determinant of actors’ blame game strategies. and CEOs can be targeted by scapegoating We theorize that actors employ their discur- (Hargie et al., 2010; Wiesenfeld et al., 2008). sive strategies through the systematic produc- Actors who can easily be discursively associ- tion of texts (van Dijk, 2006; Vaara & Tienari, ated with negative events, such as misconduct, 2008). The objective is to influence the settle- are good scapegoats for organizations; as CEOs ment of a misconduct in their favour and attain oversee the whole organization, their responsi- a discursive position that limits the reputational bility can easily be discursively established and other penalties that might result from being (Gangloff, Connelly, & Shook, 2016). judged responsible for the misconduct (Coombs Organizations that use scapegoating in the & Holladay, 2006). A discursive strategy con- discourses they produce put their integrity at risk sists of a set of texts that actors produce to attain and distance themselves from blame by identify- certain subject positions for themselves or ing one member as being responsible for the Roulet and Pichler 9 situation – the scapegoat (Hargie et al., 2010). Whistleblowing is a form of ‘dissenting dis- Building on the work of René Girard (1982), course’ (Teo & Caspersz, 2011) and involves an Grint (2010, p. 97) notes that scapegoating is ‘insider [going] public in their criticism of the ‘not an unfortunate consequence of individual policy and/or conduct of [a] powerful organiza- deviance or lack of control’. Instead, it is instru- tion’ (Perry, 1998, p. 235). Whistleblowers usu- mental (Ashforth & Lee, 1990; Bonazzi, 1983) ally target the organization they belong to, to in the sense that it is a rational discursive strategy make it accountable for a suspected or proven employed by groups when the group is at risk of wrongdoing (Near & Miceli, 2016). Lacking being held responsible. Gamson and Scotch power and status, whistleblowers rely on exter- (1964) explain that scapegoating is a ‘conveni- nal relays to trigger change (Near & Miceli, ent, anxiety-reducing act’. These scapegoating 1995; Summers-Effler, 2002; Weiskopf & efforts rely on the creation of discourses that Tobias-Miersch, 2016). While ethical and moral attribute the blame to identified scapegoats and judgements are often seen as a critical precursor provide supporting arguments. of whistleblowing (Chiu, 2003), we focus here The aim of scapegoating is to deflect blame on situations where the whistleblowing is trig- by appeasing or winning over the audiences in gered by strategic motives (Butler et al., 2020; which the blame originates (Gangloff et al., Choo et al., 2019). The positive outcomes from 2016). For an organization, the objective of this such an opportunistic move can be an active strategy is to deflect an accusation of misconduct driver and precursor of the act of whistleblow- by building the belief that it has changed satis- ing in the first place (Culiberg & Mihelič, factorily – or as Gangloff and colleagues put it 2017). Bonazzi (1983), for example, defines (p. 1617), by showing that the ‘fault for prior whistleblowing as the rational strategy of shift- wrongdoing resided squarely with [the scape- ing the responsibility to holders of power fol- goat] and did not permeate the rest of the organi- lowing an accusation of misconduct. In such a zation’. Gamson and Scotch (1964) also stress context, the whistleblower is not necessarily the ceremonial nature of scapegoating: the ritual pointing out an illegal or immoral action by the has minimal impact on material organizational collective or willing to change the situation, as outcomes and the discourses produced are pri- suggested by the founding literature on whistle- marily aimed at managing impressions (Bolino, blowing (Near & Miceli, 1985). Rather, whistle- Kacmar, Turnley, & Gilstrap, 2008). The accused blowers are motivated by the wish to avoid any organization can engage in the discursive strat- potential retaliation for their responsibility for egy of scapegoating through the production of poor performance or other adverse events press releases (Gangloff et al., 2016) and the (Westin, 1981, pp. 134–6). framing of the dismissal (Cornelissen & Werner, Assuming that whistleblowers have strategic 2014). The aim of this discursive framing is to motives in the context of blame games (Choo ‘rationaliz[e]’ the sacrifice (Grint, 2010, p. 97) et al., 2019), then the more they have to gain from and, more importantly, attribute responsibility positively distinguishing themselves from the and avoid blame through ‘dissociation’ in par- organization in order to limit contamination or ticular (Hargie et al., 2010, p. 721). blame (Moore et al., 2011), the more likely they are to blow the whistle. In the specific context of Whistleblowing as an individual a blame game, the goal of whistleblowers is to discursive strategy build positive distinctiveness by distancing them- selves from the organization being accused (Choo Individual members of organizations that are et al., 2019). Whistleblowing therefore relies on a suspected of misconduct can also participate in discourse that creates and supports this position. the blame game, particularly through whistle- This discourse emerges from the production of blowing (Culiberg & Mihelič, 2017), which is a texts – e.g. interviews, guest editorials in newspa- well-studied process in organization studies. pers – that draw a distinction between the 10 Organization Theory whistleblower and the organization being blamed. blame games to unfold in the first place. These texts cast the whistleblowers as individuals Organizational misconduct can remain uni- who are revealing evidence about the purported dentified and invisible for a long period wrongdoing, thus discursively distancing them- (Palmer, 2013; Pozner, 2008). It may take selves further from that wrongdoing. place but go unnoticed and unreported until Whistleblowing and scapegoating are, in stakeholders first spot it and take action (Greve some ways, two facets of a similar phenome- et al., 2010). For audiences to put pressure on non: whistleblowers scapegoat the organiza- an organization suspected of misconduct, the tion for wrongdoing in which they might have situation needs to be brought to their attention been involved (Near & Miceli, 1995). In our by constituents or the media through an accu- theoretical model, we therefore view scape- sation of misconduct (Faulkner, 2011; Greve goating and whistleblowing as archetypal indi- et al., 2010). Accusations of misconduct can vidual and organizational discursive strategies be triggered by different external stakeholders, of blame games. including the media, governments and con- sumers (Barnett, 2014). Whatever the source, How Blame Games Unfold in evidence of an adverse outcome needs to be visible and salient to potential accusers – con- the Aftermath of Misconduct sequently forcing them to make a negative Blame games involve a broad set of actors in causal attribution (Coombs, 2007a). An accu- interaction, each aiming to deflect the blame for sation of misconduct ‘is a publicly expressed an accusation of misconduct. Actors’ views con- and perspicuous statement of alleged wrong- front each other in a discursive space where dif- doing’ through which ‘the finger of blame is ferent strategies are adopted to deflect blame. We pointed at the culprit’ (Faulkner, 2011, p. 7). have seen that research has identified several phe- An accusation is characterized by its ‘in- nomena related to the attribution of responsibil- betweenness’, i.e. the fact that it goes beyond ity, such as whistleblowing (Culiberg & Mihelič, informal grievance in its public and accusatory 2017; Near & Miceli, 1985, 2016; Perry, 1998) nature but falls short of a formal charge of and scapegoating (Boeker, 1992; Bonazzi, 1983; wrongdoing by the state (Faulkner, 2011). It is Hargie et al., 2010). As essentially discursive thus inherently ambiguous. For example, audi- struggles, blame games can expose opposite posi- ences start to make sense of organizational tions with regard to the attribution of responsibil- misconduct following the occurrence of seem- ity (Hargie et al., 2010). Having identified the ingly anecdotal events (Boudes & Laroche, organizational and individual strategies for shift- 2009), those anecdotal events serving as accu- ing responsibility that can be used in blame mulating clues that a misconduct might have games, we now explore when they are used and happened. Crises, more generally, also begin how they can be concomitant and emerge sequen- when audiences start to attribute responsibility tially, depending on the conditions of ambiguity. for an adverse event (Bundy & Pfarrer, 2015), We integrate the concepts of whistleblowing and meaning that an accusation of misconduct may scapegoating to theorize different ideal-type emerge when there is only suspicion of organi- pathways through which blame games unfold as zational misconduct. Organizations are con- a function of moral and attributional ambiguity sidered to have agentic power which makes and in doing so conceptualize the starting and them more likely to be seen as villains rather endpoints of those pathways. than eliciting sympathy (Rai & Diermeier, 2015). Accusers usually have an interest in The starting point of blame games pointing their finger at an organization for misbehaving (Barnett, 2014), for example, Before theorizing these blame game pathways, when the suspected organization directly we must first define the conditions that lead impacts their activity. Roulet and Pichler 11

An accusation of misconduct by external The endpoint of blame games stakeholders is thus the starting point for the triggering of blame games in our model. We Although we have defined the ingredients for conceptualize the accusation as a type of dis- the blame game to take place – an accusation of course in it is own right which opens the discur- misconduct and the resulting ambiguity – we sive space for accused actors to challenge this have not discussed the pathways involved and accusation. The aim of such accusation dis- possible endpoints. In the pathways that we courses is to allocate an actor to the discursive theorize below, we explain how the beginning position of a perpetrator of wrongdoing. As and the continuation of the blame game depend such, the accusation is: on the level of existing ambiguity. Indeed, as we will see, the discursive positions taken by suspected organizations and their members are an event expressed through catchphrases and not only affected by but also affect the ensuing keywords [which] chronicle, capture, and classify ‘signature elements’, framing and promoting ambiguity. We specifically assume that, as the definitions of ‘what happened’, ‘who was discursive struggle over misconduct unfolds involved’, and ‘what went wrong’, shaping the and accused actors take discursive positions to story and providing a theme. It is a symbolic prevent blame, ambiguity will decrease in the packaging [. . .], a virtuoso exercise in the longer term. ‘redescription of behavior in order to transform We understand ambiguity in the discursive its moral significance’. (Faulkner, 2011, p. 16) struggle to exist from the point of view of audi- ences. Ambiguity arises following an accusation The accusation bears the seeds of the blame of misconduct because audiences find it difficult game. The discursive space that has been opened to make sense of the misconduct (Clemente & by this accusation is structured by attributional Gabbioneta, 2017). The ambiguity decreases as (Crocker et al., 1991) and moral ambiguity audiences make sense of the accusation, influ- (Shadnam & Lawrence, 2011). After an accusa- enced by their consumption of the discourse tion, attributional ambiguity emerges when the produced by accused actors. Because ambiguity suspected organizational misconduct cannot be ‘arises as an information problem’ and is caused directly and clearly attributed to the organiza- by a lack of information (Ball-Rokeach 1973, p. tion (Coombs & Holladay, 2002). Moral ambi- 379), by definition, any new information will guity materializes when audiences struggle to help to reduce the ambiguity (Green, 2004; Lee evaluate whether an organizational behaviour & Gailey, 2007; Leitch & Davenport, 2007). To can be clearly categorized as right or wrong interpret blame game discourses, audiences take (Greve et al., 2010; Shadnam & Lawrence, account of the information they obtain about the 2011). Transgressions can be judged to be mor- misconduct itself and the positions that actors ally ambiguous when they appear excusable, or attempt to take in relation to the misconduct borderline, from a normative point of view (Lee with their discourses. In the short run, ambiguity & Gailey, 2007). Moral ambiguity enables ‘the may increase as a result of the blame game, as construction of vocabularies of evasion – ways audiences have to interpret the information pro- to say that the rules do not apply or do not apply vided by the discursive strategies of the accused to you’ (Reichman, 1993, p. 82). Ambiguity can actors. However, as time goes by, we expect trigger a ‘short-circuited logic’ – cognitive audiences to gradually reduce such ambiguity shortcuts that lead audiences to precipitate for two reasons. causal attribution (Smelser, 1963). Most accusa- First, with each discursive move that the tions of organizational misconduct are ambigu- accused actors make, they introduce more infor- ous, meaning that the guilt of the organization mation about the misconduct that audiences will and its members is uncertain and unclear for use as part of their sensemaking process (Green, some time (Greve et al., 2010). 2004). For example, the discursive strategies of 12 Organization Theory scapegoating and whistleblowing will provide actors to discursively challenge an accusa- the audiences with additional information tion of misconduct. (Johansen et al., 2012). This information will help them to make sense of the moral status of a Blame game pathways behaviour (thus reducing the moral ambiguity) and to attribute responsibility to actors (thus Following an accusation of misconduct, moral reducing attributional ambiguity). Second, the and attributional ambiguity are not necessarily longer the blame game goes on, the more present to the same extent. For example, urgently audiences will wish to make sense of although a behaviour can be judged as morally the misconduct and settle on an interpretation wrong with relative certainty (low moral ambi- that resolves the ambiguity (Ball-Rokeach, guity) (Green, 2004), it can be hard to attribute 1973). Audiences’ sensemaking will be driven responsibility for this behaviour to a specific by ‘plausibility rather than accuracy’ and by the actor (high attributional ambiguity). Dutton, need to process the accusation of misconduct in Dukerich and Harquail (1994) cite the example a way that enables them to move on and take of the lack of support provided to the homeless action in response (Weick, 1995, p. 55). by the Port Authority of New York and New Therefore, even if the available information is Jersey. While not helping individuals in need not sufficient for reaching factual certainty, can clearly be considered wrong, audiences audiences will eventually settle on interpreta- were unclear about whether the Port Authority tions of the misconduct that are plausible, and was responsible for the situation in the first this becomes increasingly likely the longer place. On the other hand, although audiences blame game goes on. Whether the plausible might struggle to establish the moral status of a interpretation that audiences settle on at the end behaviour (high moral ambiguity), they may of the blame game is factually correct lies out- easily be able to identify and point to whoever side the scope of our blame game theory: it is is responsible (low attributional ambiguity). In likely that the strategic discourse of accused the context of the financial crisis, investment actors will sometimes be effective in influencing bankers were found to engage in behaviours audiences’ sensemaking to a significant extent, such as excessive bonuses, extreme risk-taking leading to plausible interpretations that in some or lobbying, and, while major banks were instances do not match the reality of the facts. clearly identified as the actors responsible for Overall, as the blame game unfolds, attribu- these behaviours, it was unclear whether their tional and moral ambiguity gradually decline practices were morally wrong. From the per- (Green, 2004) and the discursive space for spective of investment bankers, those practices accused actors shrinks. As there is less ambigu- were aligned with the norm of their field, while ity for the accused actors to exploit (Reichman, from the perspective of the broader society, 1993), they are prevented from taking discur- they were seen as misconduct (Roulet, 2019, sive positions to depict their role in the sus- 2020). This contradiction was due to a clash pected misconduct in a favourable way. Accused between institutional prescriptions at the indus- actors will realize that there is little for them to try level, and broader norms (Roulet, 2015). gain if they continue with the blame game and By additionally considering situations where attempt to avoid taking responsibility. They are moral and attributional ambiguity are either both therefore less likely to use these discursive high or both low following an accusation of mis- strategies. We thus start by formulating the fol- conduct, we can thus theorize four starting points lowing proposition about the end point of the from which blame games unfold into four ideal- blame game. type pathways. Table 1 provides an overview of the four starting points, each of which typically Proposition 1: The blame game continues triggers a different first move in the blame game, until ambiguity is too limited for accused and leads in turn to a different pathway. Roulet and Pichler 13

Table 1. Configurations of the Discursive Space after an Accusation of Organizational Misconduct.

Low moral ambiguity High moral ambiguity High attributional Pathway B Pathway C ambiguity •• The action identified is broadly •• There is no consensus around the recognized as morally wrong moral nature of the action identified •• There is no consensus on who is •• There is no consensus on who is responsible for the action identified responsible for the action identified Low attributional Pathway D Pathway A ambiguity •• The action identified is broadly •• There is no consensus around the recognized as morally wrong moral nature of the action identified •• The party responsible for the action •• The party responsible for the action identified is clearly singled out identified is clearly singled out

Figure 1 details the different pathways we strategies that attempt to shift blame away from theorize below. It identifies the different condi- the accused actor (Leitch & Davenport, 2007). tions of ambiguity that cause a blame game to Such strategies will be ineffective because the emerge and how the blame game unfolds as a accuser has been clearly identified and linked to consequence. It elaborates on the processes the misbehaviour in the accusing discourses. through which accusations of misconduct can This could even direct audience attention to the be settled: as ambiguity decreases, the discur- fact that there is relatively little doubt about who sive space that is available for actors to engage the culprit is, meaning that the route of discur- in the blame games shrinks. The blame game sively exploiting attributional ambiguity is itself, by providing information that supports closed off to the accused actor. audiences’ sensemaking of the accusation of Assessing the available discursive strategies, misconduct, may contribute to reducing ambi- the accused actor is therefore likely to discount guity. But as previously noted, it may also con- strategies that focus on manipulating conditions tinue for as long as audiences are unable to of attributional ambiguity. Instead, we might evaluate whether the misconduct is morally expect the accused actor to create discourses wrong and who is responsible for it. that exploit the relatively high level of moral ambiguity. The most promising discursive strat- egy will therefore be to counter the accusation Pathway A: Blame game in a situation by denying that the called-out behaviour was of high moral ambiguity and low wrongful in the first place. The accused actors, attributional ambiguity we argue, will produce and formulate argu- ments denying that the identified behaviour is The first pathway we theorize starts when the morally condemnable. discursive space, following an accusation of mis- The discursive strategy of denying wrong- conduct, is structured by high moral ambiguity fulness consists of producing texts which situ- and low attributional ambiguity. This situation is ate the behaviour in question away from the line characterized by the difficulty audiences have in that separates right and wrong, attempting to establishing whether the targeted action is mor- draw a clear distinction between that behaviour ally wrong (Green, 2004) although there is a vis- and truly wrongful behaviours. This strategy ible organizational culprit. In such a situation, might involve emphasizing the legality, the the accusation of misconduct can best be chal- prevalence and the normalcy of the behaviour lenged by the accused organization on normative in the wider institutional context (Pozner, grounds. The low level of attributional ambigu- 2008), as well as explaining the benefits of the ity is not conducive to engaging in discursive behaviour and promising to address some of its 14 Organization Theory

Starting Late stage and Initial move Blame game dynamics P1 point settlement

HighHigh moralmoral Low aambiguitymbiguity = ambiguityambiguity PropositionProposition 2 andand lowlow BlBlameame gamegame is DenyDeny wrwrongfulnessongfulness A attributionalattributional settledsettled ambiguityambiguity

AmbiguityAmbiguity PropositionProposition 33aa PropositionProposition 44aa reremainsmains = BlBlameame LowLow moralmoral SScapegoatingcapegoating WWhistleblowinghistleblowing gamegame continuecontinues andand highhigh attributionalattributional B Low aambiguitymbiguity = ambiguityambiguguity PropositionProposition 3b PropositionProposition 44bb BlBlameame gamegame is WWhistleblowinghistleblowing ScapegoatingScapegoating sesettlettled

HighHigh moralmoral Low ambiguity = andand highh PropositionProposition 5 Blame game is attributionalattribuutional DenyD wrongfulnessf l C ambiguityambiguity settled

LowLow moralmoral anandd low PropositionProposition 6 LowLow aambiguitymbiguity = Accusation of organizational misconduct attributionalattributional TaTakeke thethe blblamamee BlBlameame gamegame is D ambiguityambiguity sesettlettledd

Figure 1. Summary of the Theoretical Model: Pathways of Blame Games. drawbacks (Roulet, 2019, 2020). The discursive media (Roulet, 2019). Banks justified large position taken here thus relegates the issue of bonuses and risk-taking by showing that those responsibility to the background, blurring the practices were common in the industry and relationship between the accused actor and the important for recruiting the best and the bright- potential misconduct. This discursive strategy est. They also stressed how their practices con- has the function of making full use of the exist- formed with the law. In 2010, Goldman Sachs ing moral ambiguity and avoiding getting and one of its employees were accused of securi- caught up in the attribution of responsibility ties fraud because the firm had designed a prod- which, due to the low level of attributional uct, called Abacus, which was considered to ambiguity, provides the accused actors with mislead investors. However, although it settled relatively little discursive leeway. We can there- in court, Goldman Sachs always denied wrong- fore formulate the following proposition: doing, taking advantage of the moral gray zone with regards to investor and client relationships. Proposition 2: When moral ambiguity is high In this first scenario, we can identify two pos- and attributional ambiguity is low following sible next steps. Following proposition 1, we an accusation of misconduct, the accused might expect ambiguity to decrease as audiences organization is likely to deny the wrongful- accept the suspected actor’s arguments that the ness of the behaviour in question. identified behaviour was not morally wrong (Green, 2004). In this case, the discursive We previously discussed how many of the response of denying wrongfulness will be effec- practices in the investment banking industry tive in discarding the accusation by reducing were not seen as morally wrong by field-level moral ambiguity. The ambiguity will be settled, actors although they were condemned by the and the accusation will fizzle out as audiences Roulet and Pichler 15 fail to be convinced of the wrongfulness of the impossible in this situation. Instead, the situa- action. The other possibility is that the accused tion is conducive to exploiting the relatively actor’s discursive positioning will not convince high level of attributional ambiguity that char- the audiences (Leitch & Davenport, 2007). acterizes the discursive space. To deflect blame, Accusers might further refine their arguments accused actors will therefore opt for discursive and reinforce their discursive position to counter strategies that rely on attributional ambiguity, the denial of the accused organization. For exam- producing discourses that deny their responsi- ple, the accusers’ discourses might re-emphasize bility and assign the position of being the perpe- and reinforce the wrongfulness of the behaviour, trator or chief architect of the misconduct to e.g. by highlighting distinctions between the law someone else. and morality or by calling out the harm that the Based on previous work on crisis manage- behaviour has caused. If sufficient moral ambi- ment (Coombs, 2007a), we define scapegoating guity remains at this point, the accused organiza- as a critical organizational discursive strategy tion could, in turn, again deny wrongfulness. for shifting blame following an accusation of This discursive struggle will continue as moral misconduct with low moral but high attribu- ambiguity remains sufficiently high, unless audi- tional ambiguity.2 In this situation, a clear ences settle in between on an interpretation of the wrongdoing is made visible but it is still unclear behaviour as wrongful or not wrongful. which actors exactly are responsible. This may then trigger efforts by audiences to identify the cause and the source of the organizational Pathway B: Blame game in a situation wrongdoing, bearing the risk that the finger of of low moral ambiguity and high blame is pointed at the organization as a whole attributional ambiguity (Bonazzi, 1983; Devers, Dewett, Mishina, & Belsito, 2009). This vulnerability triggers a dis- In this second pathway, we theorize what hap- cursive reaction from the organization (Girard, pens after an accusation of misconduct when 1982). The accused organization will attempt to moral ambiguity is low (the behaviour identified discursively shift blame, for which the high by the accuser is perceived as morally wrong) level of attributional ambiguity provides the and attributional ambiguity is high (audiences necessary discursive opportunity. To do so, the are unclear about who is responsible for the organization will generate discourses that blame wrongdoing). In this context, accused actors individual organizational members in order to have the opportunity to engage in discursive shift the blame away from the organization. strategies that shift and deflect blame, taking Individual employees with limited retaliatory advantage of the difficulty audiences have in power, especially at lower levels of the organi- attributing responsibility. zation, can become easy targets for this scape- The low level of moral ambiguity in this sit- goating (Djabi & de Longueval, 2020). But uation means that discursive strategies which high-level executives and even CEOs can also exploit moral ambiguity are unlikely to be become the target of scapegoating discourses, effective, as there is already a relatively strong mainly because they are the most visible to out- consensus among audiences that the accused siders and naturally seem to hold most of the organization’s behaviour transgresses the line responsibility (Gangloff et al., 2016), which between right and wrong (Palmer, 2012). makes it easier for audiences to associate them Denying the wrongfulness of the behaviour in with the misconduct. Thus, we formulate the this situation could even backfire and exacer- following proposition: bate negative audience evaluations by making the accused actor appear indifferent to conven- Proposition 3a: When moral ambiguity is tional moral standards. Therefore, the discur- low and attributional ambiguity is high fol- sive exploitation of moral ambiguity is lowing an accusation of misconduct, the 16 Organization Theory

accused organization is likely to engage in organization instead, taking advantage of the scapegoating. high level of attributional ambiguity. Therefore, following an accusation of misconduct which The discursive strategy of scapegoating relies might result in scapegoating attempts by the on creating a distinction between the accused organization, organizational members may pre- organization as a whole and some of its mem- empt such a move by whistleblowing. bers. The discourse produced attributes respon- Whistleblowing involves organizational mem- sibility for the misconduct to specific individual bers producing text that shifts the responsibility members, thus distancing the organization from to the organization as a whole and offers infor- the accusation. Going back to the case of the mation which is only available to insiders and Volkswagen emissions scandal, the relatively which in many instances substantiates this attri- high attributional ambiguity surrounding the bution of responsibility (Bonazzi, 1983). This allegation of fraudulent behaviour enabled discursive strategy enables members to avoid Volkswagen to scapegoat software engineers as being blamed by the organization for their ‘rogue coders’.3 Because it was hard for external responsibility in the wrongdoing (Westin, 1981) audiences to understand the chain of command, and to avoid future reputational penalties for the targeted organization was able to use that themselves (Coombs, 2007b). We theorize that attributional ambiguity to its advantage to scapegoating can be pre-empted by members of deflect blame through scapegoating. the organization who attempt to produce whistle- While scapegoating is a top-down process blowing discourses to exonerate themselves and for shifting blame from the organization to indi- shift blame. vidual members, we can also expect a similar bottom-up effort by organizational members to Proposition 3b: When moral ambiguity is low shift blame to the organization (Frandsen & and attributional ambiguity is high following Johansen, 2011). Indeed, individual organiza- an accusation of misconduct, members of the tional members can be in an advantageous posi- accused organization – particularly those that tion to exploit attributional ambiguity and to could be held responsible for the misconduct generate their own discourses attributing – are likely to blow the whistle. responsibility for misconduct to the organiza- tion. Information about organizational miscon- There are many examples where whistle- duct will be available within the organization, blowers who revealed misconduct might have and the discursive struggle that follows the been blamed if they had not intervened. In accusation of misconduct can involve individ- 2013, Laurence do Rego, the chief executive for ual employees introducing this information into risk and finance for Ecobank – one of the major the discourse and making it available to outside banking conglomerates in Africa – revealed the audiences. Organizational members may have wrongdoing of the chairman and the incoming direct information and witness the wrongdoing executive director, pre-empting potential inves- (Frandsen & Johansen, 2011) and they are thus tigation of the entire company by the Nigeria in a better position to suggest causal attributions Securities & Exchange Commission. Similarly, of misconduct. in 2010, Cheryl Eckard, quality manager at As previously stressed, whistleblowing is an pharma giant GlaxoSmithKline, exposed sev- individual discursive strategy that can shift eral wrongdoings related to the quality of medi- blame away from the individual. Through cines sold by the firm after the Federal Drug whistleblowing, members within the accused Agency had issued relevant warnings. Given organization who are at risk of being blamed can their positions within those firms, both execu- position themselves in the discursive space. They tives would have been considered responsible can produce arguments that will exonerate them if the wrongdoing had become visible to of responsibility and shift the blame to the stakeholders. Roulet and Pichler 17

We have identified two potential first moves whistleblowers who react to scapegoating will in blame games in situations of high attribu- aim to clarify the link between those whistle- tional ambiguity and low moral ambiguity blowers and the behaviours attributed to them which originate from either the organization through scapegoating (Kenny, 2019). The aim (scapegoating) or its members (whistleblow- of those discourses will be to answer scapegoat- ing). Both of these discursive strategies can ing claims and to provide an alternative account initiate a blame game in this pathway, as the of who is responsible for the misconduct, e.g. discursive space with high attributional and by using the insider knowledge of the whistle- low moral ambiguity is conducive to either blower to reframe the misconduct as a systemic, paths. However, because the discursive space organizational problem that goes beyond indi- for blame games shrinks in the longer term vidual culpability (Keil, Tiwani, Sainsbury, & (proposition 1), the actor who makes the first Sneha, 2010). We thus theorize that, as a move has a certain advantage. On the other response to scapegoating, organizational mem- hand, as a strategy, whistleblowing can be very bers will blow the whistle, shifting the blame costly to individual members; e.g. it may back- back to the organization as a whole: fire or require them to leave the organization. Therefore, individual members will only make Proposition 4a: In a continued situation of a first move following an accusation of mis- low moral and high attributional ambiguity, conduct if they perceive the risk of organiza- and once they have been targeted by scape- tional scapegoating discourses to be particularly goating discourses, members of the accused high. We do not formally theorize this mecha- organization are likely to blow the whistle to nism but expect it will depend on the nature of deflect blame. the misconduct accusation: if this accusation initially appears to point relatively more to One example of the discursive reaction of a individual than to organizational responsibility, scapegoat is that of John Schnatter, founder of organizational members are more likely to fear the American pizza franchise Papa John. He that they will soon be blamed for the potential had to resign from his position as chairman in misconduct by their organization through July 2018 after making a racist remark in a con- scapegoating, and are thus more likely to make ference call and was scapegoated by the firm’s the first move. top executives. In the meantime, Schnatter We can expect the discursive strategies of pointed out the company’s problematic organi- scapegoating and whistleblowing to be met by zational culture. In an interview in August, discursive reactions that target the discourses Schnatter talked about ‘rot at the top’ and which started the blame game. Up until now, we blamed the company’s problems on the new have conceptualized whistleblowing as a pre- CEO and the ‘vindictive and controlling’ lead- emptive strategy (proposition 3b): individual ership style of the top executives.4 This body of actors at risk of being blamed by the organiza- discourses illustrates how scapegoated actors tion pre-empt such scapegoating by blowing the respond to blame when attribution is difficult. whistle (Butler et al., 2020). However, if scape- In a similar case, Jerome Kerviel, after being goating happens first, whistleblowing will condemned in the rogue trading affair in 2008, become an individual-level discursive reaction started to denounce the culture of his organiza- to scapegoating discourses. Scapegoating dis- tion, Société Générale, as a significant driver of courses are generated by the accused organiza- misconduct. tion and aim to shift the blame to the individual. Whistleblowing, whether as a pre-emptive However, members of an organization who are discursive strategy or as a response to scape- made scapegoats can engage in whistleblowing goating, provides the accusers with more infor- as a way of producing discourses that respond mation, potentially reducing attributional to this blame-shifting: the texts provided by ambiguity by enabling them to home in on the 18 Organization Theory responsible party. Whistleblowing discourses revenge on the organization.5 This discursive thus trigger discursive reactions from the position aimed to deflect the whistleblowing accused organization: the organization has to claim by delegitimizing its source. However, address the new information introduced by when institutional shareholders and other whistleblowing and its implications for the stakeholders started to investigate the matter, attribution of responsibility. The organizational the bank was forced to fire the controversial discourse in response to whistleblowing will, chief executive and reinstate the whistle- we argue, focus on deflecting the blame, justi- blower in her post.6 fying the behaviour of the organization, and Scapegoating and whistleblowing both delegitimizing the discursive position of the contribute to furthering the misconduct whistleblower. The discursive reaction of the inquiry as the discourses generated provide organization is therefore another element of a new information. This can increase ambiguity scapegoating strategy which aims to distance in the short run while audiences engage with the organization as a whole from the miscon- this new information. However, in the long duct by continuing to shift the blame to indi- run, as the blame game unfolds, we can expect vidual members (Djabi & de Longueval, 2020). this ambiguity to decrease with the emergence This strategy may rely, for example, on attribut- of new information. Audiences will use this ing blame for the misconduct to the whistle- new information to form plausible links blower and highlighting it as an isolated between the misconduct event and accused instance of individual misconduct. Alternatively, actors, converging on a definition of the situa- it may attack the whistleblower on moral tion that reduces attributional ambiguity grounds and leverage another, unrelated accu- (Ball-Rokeach, 1973). The public availability sation of misconduct against the whistleblower of this information puts external audiences in order to delegitimize the original whistle- and internal organizational actors on an equal blowing claim. Such a discursive reaction will footing to make causal attributions. At the only be effective for the accused organization if same time, the scapegoating and whistleblow- the level of attributional ambiguity remains ing discourses generated by the suspected high enough, despite the information that came organization and its members become less to light through the whistleblowing. A continu- credible (Grant et al., 2001): the validity of ing high level of attributional ambiguity makes the discourses which attempt to shift blame is it difficult for accusers to clearly identify an evaluated as they are checked against emerg- individual or organizational culprit and, as a ing information. We expect the blame games consequence, audiences will continue to strug- in this pathway to die out as the discursive gle to evaluate the validity of organizational space shrinks because of decreasing ambigu- and individual discourses. ity, thereby giving accused actors fewer opportunities to adopt discursive blame-shift- Proposition 4b: In a situation of continuing ing strategies through additional whistleblow- low moral and high attributional ambiguity, ing or scapegoating moves. after organizational members have blown When Volkswagen tried to blame the emis- the whistle, the accused organization is sions scandal on ‘rogue’ software engineers, likely to engage in scapegoating its members political and legal stakeholders questioned this to deflect blame. accusation. As the scandal unfolded, new infor- mation emerged, suggesting that top executives In the 2013 whistleblowing case involving had also known about and concealed the soft- Laurence do Rego, Ecobank top executives ware manipulation.7 The Volkswagen example countered the accusation of the whistleblower illustrates how blame game discourses lose by accusing her of not having the qualifica- credibility as more information emerges, thereby tions she claimed to have and of taking reducing ambiguity. Roulet and Pichler 19

Pathway C: Blame game in a situation (Coombs & Holladay, 2002), actors push the of high moral ambiguity and high question of responsibility to the background but attributional ambiguity do not necessarily take responsibility for the behaviour. Accused actors can deny wrongdo- This third pathway unfolds if both moral and ing while keeping attributions of responsibility attributional ambiguity are high following an deliberately vague in order to retain discursive accusation of misconduct. In this case, the dis- leeway in the next step of the blame game, cursive space appears to offer the broadest range thereby preserving the strategic option of gen- of opportunities for the accused actors to deflect erating scapegoating discourses. In the case of the blame. But the discursive space, while pre- the Abacus scandal, Goldman Sachs denied the senting a variety of strategic opportunities, is wrongfulness of its behaviour as much as it also uniquely complex. Actors accused of mis- could but ultimately opted to make one of its conduct have two options: they can exploit the vice presidents a scapegoat on the path to set- high moral ambiguity by engaging in a discur- tling the situation with the Securities & sive strategy of denying wrongfulness, or they Exchange Commission. Thus, we can conclude can exploit the high attributional ambiguity by that, due to the specific interdependency of high engaging in discursively shifting blame. moral and attributional ambiguity in this par- However, these choices are not independent of ticular situation, the most rational course of each other, and actors have to consider their action is for actors to start the blame game by interdependency when choosing their first move denying wrongfulness, keeping their discursive in this particular blame game scenario. options open further down the line. On the one hand, if actors try to exploit attri- butional ambiguity and engage in blame-shift- Proposition 5: When moral and attributional ing by attributing responsibility to another actor ambiguity are both high following an accu- (through scapegoating or whistleblowing), they sation of misconduct, the accused organiza- will forego the option of denying wrongfulness. tion is likely to first deny the wrongfulness This problem arises because shifting the blame of the behaviour in question. for a behaviour to another actor, through scape- goating or whistleblowing, implicitly acknowl- This denial of wrongfulness will set off a edges the wrongfulness of the said behaviour discursive struggle about the moral status of the (Kent & Boatwright, 2018). If the behaviour behaviour in question, which is eventually were not wrongful, it would not be necessary to likely to settle at a low level of moral ambiguity point the finger at another actor. Therefore, the with audiences interpreting the behaviour either discursive exploitation of high attributional as a misconduct or as a morally acceptable ambiguity through blame-shifting simultane- behaviour. In the latter case, the denial of ously reduces the existing moral ambiguity and wrongfulness has been successful, and the thus restructures the discursive space to pre- accusation is neutralized in a similar way to clude subsequent denials of wrongdoing. pathway A. However, in the former case, the Therefore, if, in this scenario, accused actors behaviour is now clearly seen by audiences as start a blame game with a whistleblowing or misconduct, and the suspected organization still scapegoating move, the situation will transition faces the accusation. This new situation is char- into the blame-shifting dynamics of pathway B. acterized by low moral ambiguity but still by On the other hand, if actors choose to exploit high attributional ambiguity, leaving some moral ambiguity and deny wrongfulness, they room for an attribution of responsibility. will leave space for subsequent scapegoating or Therefore, the accused actors have retained the whistleblowing in case the denial of wrongful- option of engaging in blame-shifting discursive ness is unsuccessful. The rationale for such strategies, and the dynamics of the blame game action is that, when denying wrongfulness transition into those of pathway B. 20 Organization Theory

Pathway D: Blame game in a In July 2020, Deutsche Bank’s links with the situation of low moral ambiguity and criminal Jeffrey Epstein were exposed and the low attributional ambiguity bank was accused of having enabled fraudulent transactions despite knowing about Epstein’s The final blame game pathway unfolds in a criminal history. They immediately issued an situation where both moral and attributional apology, saying that they ‘deeply regret’ their ambiguity are relatively low following an association with Epstein.8 Another example of accusation of misconduct. Here, the discursive this situation is the case of Fuji TV and Sankei space available for a blame game is limited, Shimbun in January 2020 when the two making it short-lived or even unlikely. There is Japanese media firms were caught red-handed not enough ambiguity for the accused actors to using partly fabricated polls in their pro- exploit to generate discourses that can shift grammes. Although a subcontractor was blame. Therefore, shifting the blame is difficult involved, it was clear that the media companies because the accused organization is clearly were responsible for not carefully checking the associated with the misconduct (low attribu- content being shared and how it was produced. tional ambiguity). Denying the wrongfulness The wrongfulness of their action and their of the called-out behaviour also has little responsibility in the misconduct were clear and chance of success (because of the low moral offered no discursive opportunities other than ambiguity). Indeed, denying wrongdoing is accepting the blame. likely to backfire in a context of low attribu- tional and moral ambiguity as the accused actor is already perceived as being responsible Discussion and Conclusion (Coombs, Frandsen, Holladay, & Johansen, In our blame game model, we explored how 2010) and the behaviour clearly appears to be accused organizations and their members pro- wrongful. This discursive strategy will only duce strategic discourses as a reaction to accusa- lead to further reputational damage for the tions of misconduct, thus attempting to shift accused actor (Coombs, 2007a). blame and influence the social construction of In this case, the accused actor is cornered misconduct. Building on the crisis management by the lack of moral and attributional ambigu- literature and developing a discursive perspec- ity. Here, the most rational discursive strat- tive of misconduct to integrate whistleblowing egy, apart from remaining silent and not and scapegoating, we detailed the determinants reacting, is to take responsibility for the of blame game strategies and their sequential wrongdoing pointed out by the accusers. nature. In the process, we identified the critical Actors are therefore most likely to generate role of moral and attributional ambiguity in ena- discourses in which they take the blame for bling and animating blame games. We developed the situation to minimize damage to their rep- four blame game scenarios following an accusa- utation and try to repair it (Coombs & tion of misconduct, as a function of the levels of Holladay, 2006). That is, they will, for exam- moral and attributional ambiguity. ple, engage in apology, reduce the extent of In the first scenario (pathway A), we their responsibility by highlighting extenuat- explained how a high level of moral ambiguity ing factors while acknowledging their fault (i.e. the behaviour pointed out cannot be clearly (Coombs et al., 2010), and signal their will- labelled as morally wrong by audiences) and a ingness to make amends (Coombs, 2007b). low level of attributional ambiguity (i.e. a cul- prit is clearly identifiable by audiences) is likely Proposition 6: When moral and attributional to trigger efforts by the accused organization to ambiguity are low following an accusation create discourses that deny wrongdoing by of misconduct, accused actors are likely to challenging the claim that an action was wrong- take the blame for misconduct in response. ful (proposition 2). In the second scenario Roulet and Pichler 21

(pathway B), we argued that if audiences are we suggest that accused actors take advantage unsure of the identity of the culprit, this opens a of high moral ambiguity to deny the wrongful- discursive space for suspected actors to deflect ness of the identified behaviour (proposition 5), blame. With moral ambiguity being low (i.e. the while leaving open the option of engaging in a behaviour pointed out is clearly morally wrong) blame game of scapegoating and whistleblow- and attributional ambiguity being high (i.e. ing if that denial does not convince audiences. audiences are unsure about who should be held Finally, in pathway D, we conceptualized a situ- responsible), the organization can exploit the ation in which moral and attributional ambigu- attributional ambiguity and shift the blame by ity are low, leaving the accused organization scapegoating some of its members who can be with no option other than taking the discursive associated with the misconduct – from the position of accepting the blame (proposition 6). CEOs to lower-level employees (proposition 3a). We then expect in turn scapegoats to retali- Contributions and implications for ate and generate whistleblowing discourses future research aimed at shifting the blame to the organization (proposition 4a). Our blame game theory contributes to the litera- As an alternative first move in this scenario, ture on organizational misconduct. First, by potential whistleblowers can also take advan- developing a discursive perspective on miscon- tage of attributional ambiguity to kick off a duct accusations, we advance our understand- blame game themselves when they anticipate ing of the social construction of misconduct. that they are likely to be scapegoated in the Importantly, our discursive approach differs future. The organizational members who from more material approaches to misconduct already fear being blamed for the wrongdoing and situations in which problematic behaviours assess whether it is in their interest to positively are transparently established (Mohliver, 2019), distance themselves from the organization and thus acknowledging the socially constructed may pre-emptively do so through whistleblow- nature of wrongdoing (Greve et al., 2010). We ing (proposition 3b). In response to such do not consider instances of misconduct as whistleblowing, the organization may then in something objectively given, to which stake- turn engage in further discursive strategies to holders react and for which organizations initi- shift the blame once more, scapegoating the ate corrective action (Hersel, Helmuth, Zorn, whistleblowers or other organizational mem- Shropshire, & Ridge, 2019; Shymko & Roulet, bers (proposition 4b). This back-and-forth 2017). Rather, our discursive perspective con- movement between organizational and individ- ceptualizes misconduct as being constructed ual discursive blame game strategies can con- through a struggle for meaning between audi- tinue until moral and attributional ambiguity ences and accused actors in a discursive space. generally decrease and external audiences are Here, accused actors strategically produce able to attribute responsibility (proposition 1). blame game discourses, and audiences actively In other words, yet further moves in the blame make sense of potential misconduct by consum- game become at that point less credible as audi- ing and producing discourses. ences are able to assess their validity with more Second, we shed light on a critical period of information at their disposal. In addition, the time in the social construction of misconduct: longer blame games go on, the more likely i.e. when audiences are making sense of organi- audiences are to simply settle for one of the zational misconduct following an accusation more plausible accounts that has been con- (Dewan & Jensen, 2019). Our theoretical frame- structed through their interactions with the work unpacks this understudied point in the accused parties and other stakeholders. organizational misconduct literature. We argued In a third scenario, in which moral and attri- that this period is characterized by potentially butional ambiguity are both high (pathway C), high levels of moral (Shadnam & Lawrence, 22 Organization Theory

2011) and attributional ambiguity (Jacquart & the same conceptual umbrella. Where previ- Antonakis, 2015), which open a discursive ously, the differences in levels of analysis, ini- space that accused actors can exploit and in tiating actors and a focus on non-instrumental which audiences struggle to interpret the behav- motivations (Djabi & de Longueval, 2020; iour of accused actors. We thus show how this Kenny, 2019) may have prevented the integra- definitional turmoil is particularly well suited to tion of those two literatures, we show how studying the social construction of misconduct those streams of work can be integrated. because meanings are in flux, which makes the Based on our theory, future misconduct discursive struggle more intense and exposes research could further explore the role of ambi- the social construction processes. guity in the process through which organiza- Third, we contribute to the literature by devel- tional misconduct is settled. For example, oping our understanding of the strategic behav- scholars could bring in more of the existing iour of accused actors, and thus of their role in work on causal attribution (Powell et al., 2006) the social construction of misconduct. Previous to explore how audiences process strategic dis- work on misconduct has focused primarily on courses by accused actors in making moral the role of external audiences (media, regulators, judgements and attribution of responsibility. governments) in the social construction of mis- Furthermore, more detailed examination of the conduct (Clemente & Gabbioneta, 2017; Greve content of scapegoating, whistleblowing and et al., 2010). In contrast, we focus on how other strategic blame game discourses could accused actors themselves feed into the discur- enable future research to develop a better sive construction of misconduct and blame. In understanding of how discursive strategies particular, we suggest that given the difficulty in depend on the nature of the misconduct and morally evaluating and attributing responsibility audiences. Other aspects could be considered for misconduct in complex organizations and to understand how actors decide to engage in settings, accused actors will attempt to influence whistleblowing, as we know from previous audiences’ sensemaking by employing discur- research that organizational position or power sive strategies aimed at avoiding blame. Rather are crucial determinants, as is material evi- than focusing on blame games as a sensemaking dence (Kenny, 2019). process that generates explanatory content In addition to our contributions to the mis- (Boudes & Laroche, 2009), we argue that blame conduct research, our blame game theory has games potentially disturb the sensemaking pro- key implications for the literature on social cesses in the aftermath of an accusation of mis- evaluations, particularly negative social evalu- conduct (Daudigeos, Roulet, & Valiorgue, 2020). ations (Pollock, Lashley, Rindova, & Han, Our theory highlights the agency that accused 2019). We theorize how individual and collec- actors have to actively shape the social construc- tive actors can strategically manipulate how tion of misconduct, thereby revealing the poten- they are socially evaluated in order to avoid tial for manipulation by these actors. being held responsible for misconduct, and the Finally, we bridge the gap between the con- presumed cost and benefits associated with cepts of whistleblowing and scapegoating. these strategic manipulations. In this sense, we Whistleblowing and scapegoating have so far bring together the literatures on social evalua- been studied in separate streams of work in tion and misconduct, linking more explicitly crisis communications (Coombs, 2007a; how evaluation affects the consequences of Coombs & Holladay, 2006) and in organiza- misconduct (Dewan & Jensen, 2019) and, con- tion theory (Culiberg & Mihelič, 2017; Kenny, versely, how misconduct affects evaluation 2019; Near & Miceli, 2016). We show how (Roulet, 2020). Because misconduct is an act those two phenomena are intertwined in con- of deviance – a key precursor of negative social texts of organizational wrongdoing and inte- evaluation – those two bodies of work can grate them into the misconduct literature under complement each other. Roulet and Pichler 23

We also enrich our understanding of social Extensions of our theory and practical evaluations by developing a discursive perspec- implications tive (Phillips & Oswick, 2012; Werner & Cornelissen, 2014). Social evaluations are Overall, our theoretical framework is concerned effectively framing devices and there would be with organizational misconduct and clearly benefits in further exploring their discursive focuses on blame games at the organizational nature (Roulet, 2019; Roulet, Paolella, level. However, we believe our model is extend- Gabbioneta, & Muzio, 2019; Ruebottom & able to a more macro level of analysis if we Toubiana, 2020). By examining how actors consider trickling down and trickling up mech- instrumentally act to make other actors illegiti- anisms (Roulet et al., 2019) and evaluative mate through scapegoating or whistleblowing, spillovers outside of an incriminated field we offer a strategic discursive perspective on (Aranda, Conti, & Wezel, 2020). In particular, it social evaluations (Suchman, 1995). Actors not could also be applied to fields or groups of only affect the social evaluations of others in organizations that interact in the same institu- the process, they also discursively manipulate tional arena (Zietsma, Groenewegen, Logue, & their own evaluation, for example, when Hinings, 2017). Thus, we can imagine blame whistleblowers distance themselves from a games taking place at the field or society level, blamed organization to avoid harmful contami- involving a wide range of different agents nation (Moore et al., 2011). However, we could (groups, institutions, communities, organiza- argue that there is a decreasing marginal return tions, individuals) (Bettman & Weitz, 1983). In in doing so: more actors opting out and adding the case of the 2008 financial crisis, the entire their voice to the public criticism will result in investment banking industry was accused of any additional whistleblowers standing out less triggering the financial crisis (Roulet, 2019), from the crowd. Insiders are likely to question and some actors within this sector shifted the the behaviour of the whistleblowers and expose blame to the field rather than the organizational their true motives. Thus, discursive blame strat- level. One example is the position taken by egies may have decreasing returns as the blame M&T Bank and its CEO Robert Wilmers, who game unfolds. accused the ‘big banks’ of tarnishing the reputa- Our theoretical framework also fleshes out tion of the whole industry. During the 2008 cri- the processes of attributing social evaluations sis, some actors tried to differentiate themselves and shows why they cannot be studied in isola- by engaging in what we might call ‘field-level tion. The processes through which social evalu- whistleblowing’. In this way, we see how blame ations are attributed work in ‘cascades’ (Bonardi games can target fields and groups of organiza- & Keim, 2005): evaluations are successively tional actors. More generally, accused organiza- triggered by each other as defensive reactions. tions or CEOs may blame external actors such While most of the research on social evalua- as law makers or governments, or point out tions (Pollock et al., 2019) focuses on one level their competitors. Future theorization could of analysis and on a particular point in time. We consider an external locus of blame game strat- distinguish between collectively and individu- egies and how they might interact with blame ally attributed social evaluations in line with the games within organizations. emerging literature which links categorization We also believe that our model could be and social evaluations (Devers et al., 2009; applied to a broader set of contexts in which Roulet, 2020) and do so while being attentive to audiences attribute responsibility for failure changing evaluations over time as well. Future (Dahlin et al., 2018). In cases of major industrial research could explore whether judgements of accidents (e.g. oil spills, nuclear hazards, build- status, reputation or legitimacy emerge at dif- ings collapsing, etc.), the major organizational ferent levels and how they can result from col- players may be accused of negligence or irre- lective and individual processes. sponsibility (Moura, Beer, Patelli, & Lewis, 24 Organization Theory

2017). In such a case, moral ambiguity is likely or an organization? Returning to the finance to be high because it is difficult to assess the example, the fact that a broad range of actors intentionality of causing a failure or the existence were accusing each other turned public opinion of gross negligence (Castro, Phillips, & Ansari, against them (Ho, 2009). There was no consen- 2020). Attributional ambiguity will also be high sus about who was responsible, and the sur- because of the complexity of the processes and rounding cacophony required public opinion to the multiplicity of actors involved in industrial make radical categorization against the finance operations (Dahlin et al., 2018; Palmer, 2012: industry as it was impossible to identify a clear chapter10). Such a situation is therefore likely culprit at a lower level (Roulet, 2019). Another characterized by a discursive space similar to the promising area would be to look at the outcomes one we described in pathway C and may be fer- of blame games, particularly in terms of learn- tile ground for blame games to unfold. ing processes (Boudes & Laroche, 2009). How Our theory could be further extended by do the agents who remain benefit from the blame empirically exploring the blame game process game? Milliken and Lam (2008), for example, and answering the research questions that suggested that voicing concerns contributes to emerge from our model. We can imagine situa- organizational learning. However, an organiza- tions in which the blame game does not reduce tion engaged in a blame game may likely lose ambiguity if audiences are unable to assess the some of its members as a result of them voicing validity of new information brought to their concern through whistleblowing, thus jeopard- scrutiny. In such cases, the blame game may izing the learning from such incidents. never be resolved. For example, the blame game Furthermore, future research could explore may never end if the institutional environment how variations in audiences’ sensemaking pro- creates a greater level of opacity, preventing the cess influence the blame game strategies of evidence to surface (Rodner, Roulet, Kerrigan, accused actors. While we assume that audiences & Vom Lehn, 2020). Another example would be actively attempt to make sense of wrongdoing, a situation where accused actors deliberately try some audiences may instead follow agenda set- to increase the ambiguity through their discur- ters such as the media and social control agents sive responses in the hope of leading audience (Clemente et al., 2016), be ambivalent about sus- sensemaking astray. It would be interesting for pected actors (Roulet, 2020), and be influenced future research to account for actors voluntarily by the social and economic context or by heuris- increasing ambiguity. Further research could tics (Bianchi & Mohliver, 2016). It would be explore a broader range of accused parties’ com- interesting for future research to consider how munication strategies. A corollary question is these variations influence the reaction of accused whether can we consider alternative triggers of actors. Specifically, accused actors might follow blame games. Using the concepts of moral and blame game strategies that try to exploit these attributional ambiguity, we selected precursors variations to disturb audience sensemaking. For that fit with our discursive approach in this example, accused actors might try to manipulate model and a strategic and interest-driven per- agenda setters, or trigger conflict between audi- spective on actors’ behaviours. However, we ences with ambivalent stances. can imagine other determinants that also affect Finally, our discursive approach to miscon- actors’ choices of blame game strategies – such duct could be extended by developing the link as the severity of the misconduct or the retalia- with material evidence and observable elements tory power of incriminatory audiences (Palmer to understand how the material and the symbolic et al., 2016). Second, is a blame game a process are intertwined. For example, in our blame game of redistributing social capital (e.g. legitimacy is theorization, the layer of material evidence could transferred from one agent to another), or does it be considered to be linked to the discursive space alter the general level of social capital of a field through the production of text, whereby facts and Roulet and Pichler 25 evidence are given meaning in the discursive Notes struggle and become discourse objects (Hardy & 1. https://www.theguardian.com/business/2018/ Phillips, 1999) that are used to construct notions may/31/amazon-accused-of-treating-uk-ware- of responsibility and morality. house-staff-like-robots By theorizing how accused actors deal with 2. We conceptualize scapegoating as an organi- blame, our model also has practical implica- zational discursive strategy. Texts associated tions. It could help stakeholders, such as the with this discursive strategy can be produced media, regulators, or governments (Greve et al., by individuals that are part of the organization, 2010), to critically assess the responses of sus- such as top managers, press officers and oth- pected organizations and their members. The ers. However, in the case of an organizational discursive strategy, these individuals represent validity of claims made by scapegoating or the organization as an entity and speak in its whistleblowing could be examined, and the name. This approach differs from an individ- level of caution to be exercised would depend ual discursive strategy such as whistleblowing, on interpreting the available evidence. When a where individuals produce discourses in their scandal makes organizational misconduct visi- name. ble (Daudigeos et al., 2020), stakeholders will 3. https://www.forbes.com/sites/francescoppola be tempted to take existing discourses at face /2015/10/15/attack-of-the-rogue-coders/ value, and follow the majority point of view 4. https://money.cnn.com/2018/08/28/news/com- (Adut, 2005; Clemente & Roulet, 2015). Our panies/john-schnatter-interview/index.html framework serves as a useful reminder that the 5. https://www.ft.com/content/d253340e-10bb discursive positions taken by actors suspected -11e3-b291-00144feabdc0 6. https://www.ft.com/content/a51ad338-a93c- of misconduct serve as a strategic tool to influ- 11e3-b87c-00144feab7de#axzz3LPFCy4u4 ence the meaning-making of audiences. The 7. https://www.nytimes.com/2019/03/15/business/ discursive positions taken by those playing volkswagen-winterkorn-sec-fraud.html blame games can therefore also guide stake- 8. https://twitter.com/DeutscheBank/status/128 holders’ search for reliable information. In other 0490499007283201 words, blame game discourses can carry the seeds of their de-legitimization: they provide crucial pointers for audiences, helping them to References dig for and interpret further evidence. Adut, A. (2005). A theory of scandal: Victorians, homosexuality, and the fall of Oscar Wilde. Acknowledgments American Journal of Sociology, 111, 213–248. Allport, G. (1979). The nature of prejudice. New We thank the editor Joep Cornelissen and the two York: Doubleday Anchor. (First published in reviewers for their support and guidance in develop- 1954.) ing this paper. Aranda, A. M., Conti, R., & Wezel, F. C. (2020). Distinct but not apart? Stigma reduction and Declaration of Conflicting Interests cross-industry evaluative spillovers: The case The author(s) declared no potential conflicts of inter- of medical marijuana legalization. Academy of est with respect to the research, authorship, and/or Management Journal. publication of this article. Ashforth, B. E., & Lee, R. T. (1990). Defensive behavior in organizations: A preliminary model. Funding Human Relations, 43, 621–648. Bail, C. A. (2015). The public life of secrets: The author(s) received no financial support for the Deception, disclosure, and discursive fram- research, authorship, and/or publication of this article. ing in the policy process. Sociological Theory, 33(2), 97–124. ORCID iD Ball-Rokeach, S. J. (1973). From pervasive ambigu- Thomas J. Roulet https://orcid.org/0000-0002 ity to a definition of the situation. Sociometry, -9148-3743 36(3), 378–389. 26 Organization Theory

Barnett, M. L. (2014). Why stakeholders ignore An experiment. European Economic Review, firm misconduct: A cognitive view. Journal of 119, 287–301. Management, 40, 676–702. Clemente, M., Durand, R., & Porac, J. (2016). Bettman, J. R., & Weitz, B. A. (1983). Attribution Organizational wrongdoing and media bias. in the Board Room: Causal reasoning in cor- Organizational wrongdoing: Key perspectives porate annual reports. Administrative Science and new directions (pp. 435–473). Cambridge: Quarterly, 28, 165–183. Cambridge University Press. Bianchi, E. C., & Mohliver, A. (2016). Do good times Clemente, M., & Gabbioneta, C. (2017). How does breed cheats? Prosperous times have immediate the media frame corporate scandals? The case and lasting implications for CEO misconduct. of German newspapers and the Volkswagen Organization Science, 27, 1488–1503. diesel scandal. Journal of Management Inquiry, Boeker, W. (1992). Power and managerial dismissal: 26, 287–302. Scapegoating at the top. Administrative Science Clemente, M., & Roulet, T. J. (2015). Public opinion Quarterly, 37, 400–421. as a source of deinstitutionalization: A ‘spiral Bolino, M., Kacmar, M., Turnley, W., & Gilstrap, J of silence’ approach. Academy of Management B. (2008). A multi-level review of impression Review, 40(1), 96–114. management motives and behaviors. Journal of Coombs, W. T. (2007a). Ongoing crisis communica- Management, 34, 1080–1109. tion: Planning, managing, and responding (2nd Bonardi, J. P., & Keim, G. D. (2005). Corporate ed.). Los Angeles: SAGE Publications. political strategies for widely salient issues. Coombs, W. T. (2007b). Protecting organization rep- Academy of Management Review, 30, 555–576. utations during a crisis: The development and Bonazzi, G. (1983). Scapegoating in complex organi- application of situational crisis communication zations: The results of a comparative study of theory. Corporate Reputation Review, 10(3), symbolic blame-giving in Italian and French pub- 163–176. lic administration. Organization Studies, 4, 1–18. Coombs, W. T., Frandsen, F., Holladay, S. J., & Boudes, T., & Laroche, H. (2009). Taking off the heat: Johansen, W. (2010). Why a concern for apo- Narrative sensemaking in post-crisis inquiry logia and crisis communication? Corporate reports. Organization Studies, 30, 377–396. Communications: An International Journal, Bucher, R. (1957). Blame and hostility in disaster. 15(4). American Journal of Sociology, 62, 467–475. Coombs, W. T., & Holladay, S. J. (2002). Helping Bundy, J., & Pfarrer, M. D. (2015). A burden of crisis managers protect reputational assets: responsibility: The role of social approval at Initial tests of the situational crisis communi- the onset of a crisis. Academy of Management cation theory. Management Communication Review, 40, 345–369. Quarterly, 16, 165–186. Bundy, J., Pfarrer, M. D., Short, C. E., & Coombs, Coombs, W. T., & Holladay, S. J. (2006). Halo W. T. (2017). Crises and crisis management: or reputational capital: Reputation and cri- Integration, interpretation, and research develop- sis management. Journal of Communication ment. Journal of Management, 43, 1661–1692. Management, 10, 123–137. Burke, K. (1940/1969). A rhetoric of motives. Cornelissen, J. P. (2005). Beyond compare: Berkeley, CA: University of California Press. Metaphor in organization theory. Academy of Butler, J. V., Serra, D., & Spagnolo, G. (2020). Management Review, 30, 751–764. Motivating whistleblowers. Management Cornelissen, J. P., & Werner, M. D. (2014). Putting Science, 66(2), 605–621. framing in perspective: A review of fram- Castro, A., Phillips, N., & Ansari, S. (2020). ing and frame analysis across the manage- Corporate corruption: A review and an agenda ment and organizational literature. Academy of for future research. Academy of Management Management Annals, 8(1), 181–235. Annals, 14(2), 935–968. Crocker, J., Voelkl, K., Testa, M., & Major, B. Chiu, K. (2003). Ethical judgment and whistleblow- (1991). Social stigma: The affective conse- ing intention: Examining the moderating role of quences of attributional ambiguity. Journal locus of control. Journal of Business Ethics, 43, of Personality and Social Psychology, 60, 65–74. 218–228. Choo, L., Grimm, V., Horváth, G., & Nitta, K. (2019). Culiberg, B., & Mihelič, K. K. (2017). The evolution Whistleblowing and diffusion of responsibility: of whistleblowing studies: A critical review and Roulet and Pichler 27

research agenda. Journal of Business Ethics, Grant, D., Keenoy, T., & Oswick, C. (2001). 146, 787–803. Organizational discourse: Key contribu- Dahlin, K. B., Chuang, Y. T., & Roulet, T. J. (2018). tions and challenges. International Studies of Opportunity, motivation, and ability to learn Management & Organization, 31(3), 5–24. from failures and errors: Review, synthe- Green, S. P. (2004). Moral ambiguity in white col- sis, and ways to move forward. Academy of lar criminal law. Notre Dame Journal of Law, Management Annals, 12(1), 252–277. Ethics & Public Policy, 18, 501. Daudigeos, T., Roulet, T., & Valiorgue, B. (2020). Greve, H. R., Palmer, D., & Pozner, J. E. (2010). How scandals act as catalysts of fringe stakehold- Organizations gone wild: The causes, pro- ers’ contentious actions against multinational cesses, and consequences of organizational corporations. Business & Society, 59, 387–418. misconduct. Academy of Management Annals, Dewan, Y., & Jensen, M. (2019). Catching the big 4, 53–107. fish: The role of scandals in making status a Grint, K. (2010). The sacred in leadership: liability. Academy of Management Journal. Separation, sacrifice and silence. Organization https://doi.org/10.5465/amj.2018.0685 Studies, 31, 89–107. Devers, C. E., Dewett, T., Mishina, Y., & Belsito, Hardy, C., & Phillips, N. (1999). No joking matter: C. A. (2009). A general theory of organizational Discursive struggle in the Canadian refugee stigma. Organization Science, 20, 154–171. system. Organization Studies, 20, 1–24. Djabi, M., & de Longueval, O. S. (2020). Hardy, C., Lawrence, T. B., & Grant, D. (2005). Scapegoating in the organization: Which regu- Discourse and collaboration: The role of con- lation modes? Management. versations and collective identity. Academy of Drew, P. (1998). Complaints about transgressions Management Review, 30, 58–77. and misconduct. Research on Language and Hargie, O., Stapleton, K., & Tourish, D. (2010). Social Interaction, 31, 295–325. Interpretations of CEO public apologies for Dutton, J. E., Dukerich, J. M., & Harquail, C. V. the banking crisis: Attributions of blame and (1994). Organizational images and mem- avoidance of responsibility. Organization, 17, ber identification. Administrative Science 721–742. Quarterly, 39, 239–263. Hearit, K. M. (1995). ‘Mistakes were made’: ‘’Faulkner, R. (2011). Corporate wrongdoing and the Organizations, apologia, and crises of social art of the accusation. London: Anthem Press. legitimacy. Communication Studies, 46, 1–17. Frandsen, F., & Johansen, W. (2011). The study of Hersel, M. C., Helmuth, C. A., Zorn, M. L., Shropshire, internal crisis communication: Towards an inte- C., & Ridge, J. W. (2019). The corrective actions grative framework. Corporate Communications: organizations pursue following misconduct: An International Journal, 16(4). A review and research agenda. Academy of Gabriel, Y. (2012). Organizations in a state of Management Annals, 13(2), 547–585. darkness: Towards a theory of organizational Ho, K. (2009). Liquidated: An ethnography of Wall miasma. Organization Studies, 33, 1137–1152. Street. Durham, NC: Duke University Press. Gamson, W., & Scotch, N. (1964). Scapegoating in Jacquart, P., & Antonakis, J. (2015). When does baseball. American Journal of Sociology, 70, charisma matter for top-level leaders? Effect 69–76. of attributional ambiguity. Academy of Gangloff, K. A., Connelly, B. L., & Shook, C. L. (2016). Management Journal, 58, 1051–1074. Of scapegoats and signals: Investor reactions to Johansen, W., Aggerholm, H. K., & Frandsen, F. CEO succession in the aftermath of wrongdoing. (2012). Entering new territory: A study of inter- Journal of Management, 42, 1614–1634. nal crisis management and crisis communica- Gephart, R. P. (1993). The textual approach: Risk tion in organizations. Public Relations Review, and blame in disaster sensemaking. Academy of 38, 270–279. Management Journal, 36, 1465–1514. Keil, M., Tiwana, A., Sainsbury, R., & Sneha, S. Girard, R. (1982) The scapegoat. Baltimore, MD: (2010). Toward a theory of whistleblowing Johns Hopkins University Press. intentions: A benefit-to-cost differential per- Grant, D., & Hardy, C. (2004). Introduction: spective. Decision Sciences, 41, 787–812. Struggles with organizational discourse. Kenny, K. (2019). Whistleblowing: Toward a new Organization Studies, 25(1), 5–13. theory. Boston, MA: Harvard University Press. 28 Organization Theory

Kent, M. L., & Boatwright, B. C. (2018). Ritualistic Near, J. P., & Miceli, M. P. (1985). Organizational sacrifice in crisis communication: A case for dissidence: The case of whistleblowing. Journal eliminating scapegoating from the crisis/apo- of Business Ethics, 4, 1–16. logia lexicon. Public Relations Review, 44, Near, J. P., & Miceli, M. P. (1995). Whistle-blowing: 514–522. Myths and reality. Journal of Management, 22, Koopmans, R., & Olzak, S. (2004). Discursive 507–526. opportunities and the evolution of right wing Near, J. P., & Miceli, M. P. (2016). After the wrongdo- violence in Germany. American Journal of ing: What managers should know about whistle- Sociology, 110, 198–230. blowing. Business Horizons, 59, 105–114. Lawrence, T. B., Phillips, N., & Hardy, C. (1999). Palmer, D. (2012). Normal organizational wrongdo- Watching whale watching: Exploring the dis- ing: A critical analysis of theories of miscon- cursive foundations of collaborative relation- duct in and by organizations. Oxford: Oxford ships. Journal of Applied Behavioral Science, University Press. 35, 479–502. Palmer, D. A. (2013). The new perspective on organ- Lee, M. T., & Gailey, J. A. (2007). Attributing izational wrongdoing. California Management responsibility for organizational wrongdoing. In Review, 56(1), 5–23. International handbook of white-collar and cor- Palmer, D. (2014). Robert R. Faulkner: Corporate porate crime (pp. 50–77). Boston, MA: Springer. wrongdoing and the art of the accusation. Leitch, S., & Davenport, S. (2007). Strategic ambigu- Administrative Science Quarterly, 59, 370–373. ity as a discourse practice: The role of keywords Palmer, D., Greenwood, R., & Smith-Crowe, K. in the discourse on ‘sustainable’ biotechnology. (Eds.) (2016). Organizational wrongdoing: Key Discourse Studies, 9(1), 43–61. perspectives and new directions. Cambridge: Livesey, S. M. (2001). Eco-identity as discursive Cambridge University Press. struggle: Royal Dutch/Shell, Brent Spar, and Perry, N. (1998). Indecent exposures: Theorizing whistle- Nigeria. Journal of Business Communication blowing. Organization Studies, 19, 235–257 (1973), 38(1), 58–91. Phillips, N., & Oswick, C. (2012). Organizational MacLean, T. L. (2008). Framing and organizational discourse: Domains, debates, and directions. misconduct: A symbolic interactionist study. Academy of Management Annals, 6(1), 435–481. Journal of Business Ethics, 78, 3–16. Pollock, T. G., Lashley, K., Rindova, V. P., & Han, J. Maguire, S., & Hardy, C. (2012). Organizing pro- H. (2019). Which of these things are not like the cesses and the construction of risk: A discursive others? Comparing the rational, emotional and approach. Academy of Management Journal, moral aspects of reputation, status, celebrity and 56, 231–255. stigma. Academy of Management Annals. Milliken, F. J., & Lam, N. (2008). Making the deci- Powell, T. C., Lovallo, D., & Caringal, C. (2006). sion to speak up or not: Implications for organi- Causal ambiguity, management perception, and zational learning. In J. Greenberg, M. Edwards, firm performance. Academy of Management & C. Brinsfeld (Eds.), Voice and silence in Review, 31, 175–196. organizations. Thousand Oaks, CA: SAGE Pozner, J. E. (2008). Stigma and settling up: An inte- Publications. grated approach to the consequences of organi- Mohliver, A. (2019). How misconduct spreads: zational misconduct for organizational elites. Auditors’ role in the diffusion of stock-option Journal of Business Ethics, 80, 141–150. backdating. Administrative Science Quarterly, Pozner, J. E., & Harris, D. J. (2016). Who bears the 64, 310–336. brunt? A review and research agenda for the Moore, C., Stuart, H. C., & Pozner, J. E. (2011). consequences of organizational wrongdoing for Avoiding the consequences of misconduct: individuals. In D. Palmer, R. Greenwood, & K. Becoming licensed by and insulated from stigma. Smith-Crowe (Eds.), Organizational wrongdoing: Working Paper. Harvard Business School. Key perspectives and new directions (pp. 404– Moura, R., Beer, M., Patelli, E., & Lewis, J. (2017). 434). Cambridge: Cambridge University Press. Learning from major accidents: Graphical Radoynovska, N., & King, B. G. (2019). To whom representation and analysis of multi-attribute are you true? Audience perceptions of authen- events to enhance risk communication. Safety ticity in nascent crowdfunding ventures. Science, 99, 58–70. Organization Science, 30, 781–802. Roulet and Pichler 29

Rai, T. S., & Diermeier, D. (2015). Corporations effect of extraneous stakeholder relationships are cyborgs: Organizations elicit anger but not in the field of cultural production. Academy of sympathy when they can think but cannot feel. Management Journal, 60, 1307–1338. Organizational Behavior and Human Decision Smelser, N. (1963). Theory of collective behavior. Processes, 126, 18–26. New York: Simon & Schuster. Reichman, N. (1993). Insider trading. Crime and Suchman, M. (1995). Managing legitimacy: Justice, 18, 55–96. Strategic and institutional approaches. Academy Rhodes, C. (2016). Democratic business ethics: of Management Review, 20, 571–610. Volkswagen’s emissions scandal and the dis- Summers-Effler, E. (2002). The micro potential ruption of corporate sovereignty. Organization for social change: Emotion, consciousness, Studies, 37, 1501–1518. and social movement formation. Sociological Rodner, V., Roulet, T. J., Kerrigan, F., & Vom Lehn, Theory, 20(1), 41–60. D. (2020). Making space for art: A spatial per- Teo, H., & Caspersz, D. (2011). Dissenting dis- spective of disruptive and defensive institu- course: Exploring alternatives to the whistle- tional work in Venezuela’s art world. Academy blowing/silence dichotomy. Journal of Business of Management Journal. Ethics, 104, 237–249. Roulet, T. (2015). ‘What good is Wall Street?’ Vaara, E., & Tienari, J. (2008). A discursive perspec- Institutional contradiction and the diffusion of tive on legitimation strategies in multinational the stigma over the finance industry. Journal of corporations. Academy of Management Review, Business Ethics, 130, 389–402. 33, 985–993. Roulet, T. J. (2019). Sins for some, virtues for others: Van Dijk, T. A. (2006). Discourse and manipulation. Media coverage of investment banks’ misconduct Discourse & Society, 17, 359–383. and adherence to professional norms during the Weick, K. E. (1995). Sensemaking in organizations. financial crisis. Human Relations, 72, 1436–1463. Thousand Oaks, CA: SAGE Publications. Roulet, T. J. (2020). The power of being divisive: Weiner, B. (1986). An attributional model of moti- Understanding negative social evaluations. vation and emotion. New York: Springer- Standford, CA: Stanford University Press. Verlag. Roulet, T. J., & Clemente, M. (2018). Let’s open the Weiskopf, R., & Tobias-Miersch, Y. (2016). media’s black box: The media as a set of hetero- Whistleblowing, parrhesia and the contestation geneous actors and not only as a homogenous of truth in the workplace. Organization Studies, ensemble. Academy of Management Review, 37, 1621–1640. 43, 327–329. Werner, M. D., & Cornelissen, J. P. (2014). Framing Roulet, T. J., Paolella, L., Gabbioneta, C., & Muzio, D. the change: Switching and blending frames and (2019). Microfoundations of institutional change their role in instigating institutional change. in the career structure of UK elite law firms. Organization Studies, 35, 1449–1472. Research in the Sociology of Organizations. In Westin, A. F. (1981). Whistle-blowing: Loyalty and Microfoundations of institutions. dissent in the corporation. New York: McGraw- Ruebottom, T., & Toubiana, M. (2020). Constraints Hill. and opportunities of stigma: Entrepreneurial Wiesenfeld, B. M., Wurthmann, K., & Hambrick, D. emancipation in the sex industry. Academy of C. (2008). The stigmatization and devaluation of Management Journal. elites associated with corporate failures: A pro- Sennet, A. (2016). Ambiguity. In E. N. Zalta (Ed.), cess model. Academy of Management Review, The Stanford encyclopedia of philosophy 33, 231–251. (Spring 2016). Metaphysics Research Lab, Wong, P. T. P., & Weiner, B. (1981). When peo- Stanford University. https://plato.stanford.edu/ ple ask ‘why’ questions, and the heuristics of archives/spr2016/entries/ambiguity/ attributional search. Journal of Personality and Shadnam, M., & Lawrence, T. B. (2011). Social Psychology, 40, 650–663. Understanding widespread misconduct in organ- Zietsma, C., Groenewegen, P., Logue, D. M., & izations: An institutional theory of moral col- Hinings, C. R. (2017). Field or fields? Building lapse. Business Ethics Quarterly, 21, 379–407. the scaffolding for cumulation of research on Shymko, Y., & Roulet, T. J. (2017). When does institutional fields. Academy of Management Medici hurt da Vinci? Mitigating the signaling Annals, 11(1), 391–450. 30 Organization Theory

Author biographies Annals, and featured in outlets such as the Economist, Bloomberg, the Washington Post and Le Monde. Thomas Roulet is a faculty in organisation theory at the Cambridge Judge Business School and Fellow in Rasmus Pichler is a PhD candidate in Strategic Sociology at Girton College, both at the University Management at the University of Cambridge, Judge of Cambridge. His work on negative social evalua- Business School. His research focuses on corporate tions, misconduct and institutions has been published wrongdoing, the social evaluation of corporations, in the Academy of Management Journal, Academy of and the interplay of top management with corporate Management Review and Academy of Management governance.