Hypocrisy and legitimacy in the aftermath of a scandal: an experimental study of stakeholder perceptions of nonfinancial disclosure

Purpose – This study aims to investigate how stakeholders perceive the company’s nonfinancial disclosure after a scandal has occurred. More specifically, the authors examine whether and how sustainability reporting practices in the aftermath of a scandal can influence the perceptions of stakeholders in terms of hypocrisy and legitimacy. Design/methodology/approach – The present research represents a companion paper to another study in this issue that investigates the adaptation of companies’ reporting behaviors after a scandal. The results of the initial qualitative study informed the subsequent quantitative study developed in this article. The authors build on the evidence of the main paper and perform a 2 3 2 between-subjects experiment to examine how stakeholders perceive the actions of companies that aim to restore their eroded legitimacy through social, environmental and sustainability (SES) reporting. Findings – The results suggest that when companies take responsibility and develop remedial, socially responsible corporate activities are perceived as less hypocritical and more legitimate. Moreover, we show an interaction effect between taking responsibility and developing remedial socially responsible actions on hypocrisy and legitimacy perception. Hypocrisy and legitimacy in the aftermath of a scandal

This paper is organized as follows. Section 2 reviews the literature and develops hypotheses. Section 3 illustrates the experimental design. Section 4 presents the results of the study. Section 5 highlights our main conclusions, limitations and further research.
This study is a companion paper to another article in this issue. Hence, it represents a development of the analysis conducted in the first paper, where we explore how companies adapt their sustainability reports after a scandal. In this paper, we provide evidence of how stakeholders perceive companies' responses to a scandal in terms of reporting, thus offering a wider picture of the consequences of SES scandals for both the companies and stakeholders involved.
2. Literature 2.1 Stakeholder perceptions of corporate legitimacy and hypocrisy Hypocrisy occurs when a gap between assertions and actions begins to appear (Shklar, 1984). The term refers to the "motivation to appear moral yet, if possible, avoid the cost of actually being moral" (Batson et al., 2006, p. 321). Hampered accountability risks are engendering accusations of hypocrisy. This latter response occurs in a world in which values, ideas or people are in conflict, and it is a means for both individuals and organizations to address such conflict (Brunsson, 2007). In contrast, sincerity is interpreted as a "degree of congruence" that "does not relate to how ethical a company is but how true that company is to its mission statement, value declarations or corporate charter" (Fassin and Buelens, 2011, p. 587).
Recent research argues that firms use hypocritical talk, decisions and actions to manage divergent stakeholder interests and hence maintain legitimacy (Cho et al., 2015). The authors refer to "talk" as written or spoken words presenting organizations' commitments and policies to interact with the general and competitive environment and, in particular, with external stakeholders. While "decisions" are a special type of talk that indicates a future intention and an increased probability of corresponding actions (Brunsson, 2007), "actions" represent the execution of previous talk and decisions (Brunsson, 1993). Every "talk" contributes to building organizational façades, namely, symbolic appearance used to manage organizational legitimacy (Abrahamson and Baumard, 2008;Cho et al., 2015). More specifically, a "progressive façade" aims to show the organization's progress toward strategic goals, while a "reputational façade" illustrates the organization's positive images for stakeholders (Abrahamson and Baumard, 2008). In other words, organizations build façades to influence stakeholders' assessment of their social and institutional practices in an attempt to improve perceptions of the organization, and thereby managing its legitimacy (She and Michelon, 2019, p. 55).
This latter element, in turn, is a particularly valuable intangible asset for business success and even simply for the survival of an organization. Mitchell et al. (1997) argue that "legitimacy is a social good, it is something larger and more shared than a mere self-perception, and it may be defined and negotiated differently at various levels of social organization" (p. 867). In explaining their concept of legitimacy, the authors accept Weber's (1947) proposition that legitimacy and power are distinct attributes that can combine to create authority but that can exist independently as well. Mitchell et al. (1997) make use of Suchman's (1995) definition of organizational legitimacy, which is established when actions are perceived to be desirable according to social norms. Lee et al. (2018) support the idea that legitimacy is determined by the way members of a society perceive, think about and feel toward an organization (see also : Hatch and Schultz, 1997). Even though companies may attempt to create legitimacy (Sonpar et al., 2010), if consumers, and stakeholders in general, do not accredit it, then their corporate social responsibility (CSR) efforts could be in vain (Palazzo and Scherer, 2006). Hypocrisy and legitimacy in the aftermath of a scandal After a scandal, the corporate image is naturally compromised, but the methods of reaction to the scandal are decisive for rebuilding or definitively losing the organization's legitimacy and credibility: if stakeholders do not trust the corporation, then its reputation is damaged, as is its license to operate in society (Petty and Guthrie, 2000). Moreover, if reputation is damaged, relational capital is compromised (Casonato et al., 2019, p. 147;see also: de Castro et al., 2004). Once the license to operate is breached by a scandal, even shareholders and funders often sell their sharescausing share prices to plummetor withdraw their credit lines. Therefore, a good reputation helps to maintain value, whereas a bad reputation tends to destroy it (Gatzert, 2015).
With reference to reporting as a legitimation tool, Kuruppu et al. (2019Kuruppu et al. ( , p. 2081) sustain that the choice regarding whether direct action, reporting or a combination of both, should be used in organizational attempts to gain, maintain and manage legitimacy is affected by the visibility of the issue, the salience of the stakeholders involved and the stakeholder networks affected. For example, if an issue becomes public, a firm may need to take action and report to offer explanations or excuses.
This study contributes to understanding possible remedial actions that are effective not only for restoring the organizational façade and corporate image, thereby avoiding stakeholder perceptions of a gap between assertions and actions (hypocrisy in the sense of Batson et al., 2006), but also for rebuilding or reinforcing legitimacy through reporting action that is perceived to be desirable according to social norms (Suchman, 1995).

Hypothesis development on remedial actions and admission of responsibility
According to the literature on SES disclosure after a scandal, companies facing crises often develop CSR initiatives as a crisis strategy management response. However, this reactive approach necessitates caution, as when a corporation engages in CSR after a crisis, its CSR efforts may be viewed as an attempt to rectify the perception of corporate legitimacy and clear an eroded reputation.
Some companies decide to offer compensation to stakeholders affected by the damage through remedial CSR activities, whereas others disclose CSR activities that are not directly aimed at remedying the damage caused. Thus, companies implement CSR strategies that either do or do not fit with the damage they caused. The implementation of remedial or nonremedial CSR activities can affect stakeholders' responses. Indeed, how stakeholders interpret the relationship between the CSR initiative and the cause of the firm's hampered accountability could define their perceptions (Garc ıa-Jim enez et al., 2017). This effect can be explained by stakeholders' attribution of motives that guide companies' behavior. According to attribution theory, CSR initiatives can have two different CSR motivations for implementation: intrinsic and extrinsic. When companies engage in CSR for intrinsic motives, stakeholders perceive the company as sincere, as they trust the "benevolent" character of the organization and the true values of the firm. In contrast, extrinsic CSR practices are those "perceived to be done to get something back or to avoid some kind of punishment from the community in general" (Story and Neves, 2015, p. 113). Therefore, if companies implement remedial CSR activities that attempt to remedy the damage they caused, they are perceived as sincere; as such activities reflect genuine concern. Conversely, the incongruence between CSR activity and the damage caused raises stakeholder skepticism over a company and its social projects, signaling the existence of an ulterior motive to engage in a CSR activity, such as appearing moral while avoiding the cost of being moral (Sen et al., 2006). Hence, if a company launches a CSR initiative in a different domain from the crisis issue, the CSR may not align with stakeholders' expectations that the company concentrate on addressing the crisis. By contrast, a CSR initiative in the same domain as the crisis issue would be considered consistent with the expectations of stakeholders, who recognize the company's genuine care about crisis problem solving (Kim and Choi, 2018).
We hypothesize that companies are perceived by stakeholders more positively when implementing a CSR initiative that is congruent with the cause of the specific disaster or scandal that eroded that company's legitimacy as follows: H1. Companies are perceived as less hypocritical and more legitimate if they develop a remedial CSR activity.
Both impression management theory and account episodes theory identify the admission of responsibility-as well as offers of compensation-as a relevant component that enables the communication strategies of scandal-ridden companies to appear sincere, convincing and effective.
As information is necessary to change perceptions (Adams and Zutshi, 2004), a strategy that aims at improving a company's legitimacy and affecting external parties must be accompanied by disclosure.
In the field of psychology, hypocritical behavior is closely associated with the fundamental error of attribution (Polman and Ruttan, 2012). Individuals often try to explain and justify their behavior by attributing it to environmental and/or alien causes, while they often attribute the actions of others to innate characteristics. Some companies declare their responsibility for causing certain damage, whereas other companies prefer not to mention the violation they committed or not to assume their responsibility to avoid the risk of showing that their activities were unacceptable and inappropriate and that their social contract is not legitimate (Parmar et al., 2010). Thus, some companies decide not to share their responsibility for a negative event with stakeholders because doing so highlights failures and mistakes. However, recent studies have shown that disclosing bad news may be perceived as more credible than disclosing good news because management has less incentive to publish it when untrue, revealing weaker perceived hypocrisy of negative than positive information (Jahn and Br€ uhl, 2019).
Since hypocritical behaviors derive from a gap between assertions and actions, companies that do not recognize their responsibility could be perceived as hypocritical by definition. For this reason, we expect companies to be perceived as hypocritical and less legitimate if they did not accept responsibility for their actions and as less hypocritical and more legitimate if they took responsibility for their actions.
Accordingly, we formally frame the following hypothesis regarding companies' behavior in relation to a crisis as follows: H2. Companies are perceived as less hypocritical and more legitimate if they take responsibility.
Classical SES disclosure theories assume coherence among intentions, decisions and actions (Michelon et al., 2015). Discrepancies between SES disclosure and action can be viewed as a way for companies to attempt to ward off criticism and give false impressions of actual CSR practices (Schoeneborn et al., 2020). If companies do not recognize their responsibilities but admit to developing remedial CSR activities, stakeholders might charge them with duplicity if they perceive the information as incoherent (Higgins et al., 2020). Thus, we expect that an interaction effect between the disclosure of remedial actions and the admission of fault exists. Hence, we formulate the hypothesis as follows: H3. When companies develop a remedial CSR activity, they are perceived as more hypocritical and less legitimate if they do not admit their responsibility.
Hypocrisy and legitimacy in the aftermath of a scandal

Methodology
We implement an experimental design to test our hypotheses on stakeholder perceptions of sustainability reports. Previous studies have used experiments to assess users' reactions to disclosures in SES reports (e.g. Kuruppu and Milne, 2010). We draw on prior literature to design the different cases related to remedial actions and admission of responsibility.
We developed an online 2 3 2 between-subjects experiment to investigate the effects of CSR actions (remedial CSR vs nonremedial CSR) and admission of responsibility (own attribution vs attribution to others) on perceptions of companies' hypocrisy and legitimacy by potential users of nonfinancial disclosure. Specifically, regarding the first variable, we considered the relationship between the CSR activities adopted by the company and the damage caused by the company, identifying "remedial" and "nonremedial" CSR activities. The second variable-the admission of responsibility-refers to the company's admission of its responsibility for the damage or its attribution of responsibility to an external cause.
The hypotheses were tested using the data collected through a questionnaire that was randomly distributed through the Prolific Academic (ProA) online crowdsourcing platform. ProA, which was launched in 2014, was selected because the literature claims that its participants produce higher-quality data than participants on other platforms, such as CrowdFlower and Amazon Mechanical Turk (Peer et al., 2017).
Because the study aimed to examine the influence of companies' admission of responsibility and CSR activities on hypocrisy and legitimacy perceptions, different stimuli were used to create four different scenarios. All scenarios involved a partial SES report page from a fictional company. We opted not to use scenarios referring to real companies because the resulting associations could have produced distortions that undermined our findings (McCarthy and Norris, 1999). However, we built on the existing literature on corporate scandals and on our content analysis of the 11 cases to present a realistic scenario. We manipulated the two variables via the contents of the SES report. The fictitious brand was presented in the cover story as follows: TechAdvance is a German-based multinational information technology company founded in 1990 that currently has 121,000 employees. It produces a wide variety of consumer electronics products and appliances (including electronic kitchen appliances, personal computers, telephones, MP3 players, audio equipment, etc.).
Then, we specified that the company misused chemical substances, which gave 14 employees leukemia. A fictitious page of the corporate responsibility report that TechAdvance published at the end of 2017 was presented to respondents.
The admission of responsibility was manipulated by adopting the following sentences: M1a. "As stated publicly several times by our CEO, the company is clear on its responsibilities related to the damages caused to its 14 employees." M1b. "As stated publicly several times by our CEO, the company declines all responsibility and attributes the damages caused to its 14 employees to external causes." To manipulate the CSR activity, we adopted the following sentences: M2a. "During the year, we continued to meet our commitments to those affected. All legitimate claims for compensation have been and will continue to be met. By the end of 2011, US$5m had been provided to the families of the employees who were the victims of this accident." M2b. "TechAdvance has been working hard to improve its social responsibility and is helping the communities in Sri Lanka, where a part of the production is located, with the construction of a new school, investing US$5m in this project." The manipulation check of the cover story's contents was assessed by undertaking a preliminary test with a sample of 20 respondents that confirmed the perception of the gravity of the damage caused by the company. We asked respondents to rank the gravity of the violation on a seven-point Likert scale: the results show a mean of 6.10, which is significantly different from 4, the central value of the Likert scale (p < 0.01; SD 5 1.16). We also checked manipulations of the two independent variables by randomly providing the four scenarios developed for the experiment to a sample of 20 people through ProA. The admission of responsibility was tested by asking respondents to indicate the degree of agreement with the following two statements (with a seven-point Likert scale): "TechAdvance assumed its responsibility regarding the damage"; "TechAdvance communicated its responsibility for the damages caused to its 14 employees." The T-test of means confirmed the validity of the manipulation effect (p < 0.01 for both the statements). To test the CSR activity manipulation, we asked respondents to indicate the degree of agreement with the following two statements (with a seven-point Likert scale): "The remedy that TechAdvance implemented is related to the damage caused to its employees" and "The social remedy implemented by TechAdvance is related to the health issue caused to its employees." The T-test of means confirmed the validity of the manipulation effect (p 5 0.02 and p 5 0.004, respectively). Moreover, we consulted five other academic experts regarding this methodological approach to check our experimental design.  Table 1 for measures and properties of the constructs.
The experiment was conducted in July 2018 and involved a sample of 281 participants, who were randomly assigned to scenarios. Respondents were pre-screened according to the following three criteria. The first was age: the questionnaire was sent only to people aged between 22 and 65. This criterion allowed us to orient the study to the age range of people familiar with the industry in which the fictitious company operated and potential users of its nonfinancial disclosure. Second, only participants with English as their native language were included; this ensured that the respondents understood the text of the questionnaire and the

Constructs Source Items
Cronbach alpha

Perceived hypocrisy
Wagner et al.
(1) TechAdvance acts hypocritically (2) What TechAdvance says and does are two different things (3) TechAdvance pretends to be something that it is not (4) TechAdvance does exactly what it says a (5) TechAdvance keeps its promises a (6) TechAdvance puts its words into action a 0.88 Organizational legitimacy Bachmann and Ingenhoff (2016) (1) It seems to me that TechAdvance acts consistently with socially accepted norms and values (2) In my opinion, TechAdvance seems to be a legitimate company Hypocrisy and legitimacy in the aftermath of a scandal stimuli. Third, because of the influence of culture on individuals' behavior, we selected only British respondents. The sample included 100 men (36%) and 181 women (64%). Among them, 52% were between 22 and 34 years of age, 25% were between 35 and 44 years of age, 13% were between 45 and 54 years of age and 10% were between 55 and 64 years of age.

Results
After the data collection, we tested our hypotheses by performing an analysis of variance (ANOVA) using IBM SPSS software.
As illustrated in Figures 1 and 2, the effect of CSR activities on perceived hypocrisy (F 5 4.584; p 5 0.033) and legitimacy (F 5 8.417; p 5 0.004) is significant. Subjects exposed to the "remedial CSR" scenario perceived a lower level of hypocrisy (M 5 4.163; SE 5 0.096) and a higher level of legitimacy (M 5 4.459; SE 5 0.100) than subjects exposed to the "nonremedial CSR" scenario (hypocrisy: M 5 4.066; SE 5 0.113 and legitimacy: M 5 3.590; SE 5 0.119). The results validate our first hypothesis that companies are perceived as less hypocritical and more legitimate if they develop a remedial CSR activity (H1).  The effect of admission of responsibility on perceived hypocrisy (F 5 19.704; p < 0.01) and legitimacy (F 5 18.082; p < 0.01) is significant. Subjects exposed to the "admission of responsibility" scenario perceived a lower level of hypocrisy (M 5 4.004; SE 5 0.094) and a higher level of legitimacy (M 5 4.177; SE 5 0.112) than subjects exposed to the "attribution of responsibility to external causes" scenario (hypocrisy: M 5 4.618; SE 5 0.101 and legitimacy: M 5 3.479; SE 5 0.120). Hence, the results provide support for our second hypothesis that companies are perceived as less hypocritical and more legitimate if they take responsibility (H2).
The third hypothesis (H3: when companies develop a remedial CSR activity, they are perceived as more hypocritical and less legitimate if they do not recognize their responsibility) is confirmed by a pairwise comparison that is significant for hypocrisy (p < 0.01; SE 5 0.191) and for legitimacy (p 5 0.009; SE 5 0.227). Indeed, when the company adopts a remedial CSR, it is perceived as more hypocritical (M 5 4.559; SE 5 0.134) and less legitimate (M 5 3.766; SE 5 0.159) if it does not recognize its responsibility compared with the scenario in which it recognizes its responsibility (hypocrisy: M 5 3.768; SE 5 0.137 and legitimacy: M 5 4.366; SE 5 0.162).

Conclusions
In this paper, we investigate how stakeholders perceive companies' nonfinancial disclosure after a scandal and whether and how sustainability reporting practices in such an aftermath can influence stakeholder perceptions in terms of hypocrisy and legitimacy. Our methodology aims at understanding how stakeholders perceive the actions of companies that aim to restore their eroded legitimacy after a scandal through SES reporting. In this sense, the paper contributes to the understanding of possible remedial actions that are effective for avoiding stakeholder perceptions of a gap between assertions and actions (Batson et al., 2006). This paper has value for investigating the adaptation of SES reporting from the integrated perspective of both organizations and their stakeholders. Given the emphasis on accountability and CSR, it is increasingly relevant for business research to contribute to the understanding of the connection between SES reporting and stakeholder perceptions. Even though accounts are one of the few managerial tools to mitigate negative outcomes during a crisis, little research has successfully investigated the effects of different crisis accounts on stakeholder reactions (McDonald et al., 2010). As responses become more accommodating to stakeholders, they become more expensive (Stockmyer, 1996): the success of less-accommodative accounts would provide practitioners with lower-cost communication options, but the results of our experiment support the view that companies are perceived as less hypocritical when they develop a remedial CSR activity and when they take responsibility. Given that potential users of nonfinancial disclosure can implement boycotting behaviors if they perceive a loss of corporate legitimacy, research investigating their legitimacy perceptions is important (Kates, 2004;Kozinets, 2002;Hu et al., 2018). Attribution theory predicts that voluntarily reporting negative information is likely to be attributed to a company's honesty (Jahn and Br€ uhl, 2019). In this light, our experiment suggests that avoiding clear responsibility, omitting negative information and developing CSR activities unlinked to the issues of the core business appear to violate the implicit terms of the social contract and lead to a perception of hypocrisy.
We show that stakeholders perceive companies that admit their responsibility as less hypocritical and more legitimate. This evidence suggests to managers that stakeholders can appreciate a disclosure that highlights failures or mistakes, especially for voluntary disclosures. In contrast, stakeholders negatively evaluate a statement that avoids or denies responsibility; thus, managers should use their sustainability reporting to inform Hypocrisy and legitimacy in the aftermath of a scandal stakeholders about particular events that occur even if they seem to negatively affect a company's reputation. This transparency is important because the consequent perception of companies' legitimacy is related to other crucial factors. For instance, companies with a high level of legitimacy can improve their relationships with and gain the support of their stakeholders. In contrast, low levels of legitimacy can increase the risk of losing market share, and thus, damage corporate profitability (Li et al., 2017). A recommendation that emerges from this research is related to CSR remedial activities. As illustrated by the results of the experiment, remedial and other CSR activities are appreciated when they are considered coherent with the specific issue but are considered illegitimate and hypocritical otherwise. After a scandal, the corporate reputation is inevitably compromised but how organizations react to the scandal is crucial in determining their legitimacy. This finding implies the need for organizations to design CSR activities that are genuinely linked with improvements in the managerial system that caused or failed to prevent the scandal. The design of CSR activities should also be coherent with the admission of responsibility. In this vein, managers should strictly integrate their CSR actions and disclosure choices. Indeed, a gap between actions and assertions defines, and thus causes stakeholders' perception of hypocrisy and, at the same time, delegitimizes their rights to operate.
This study has some limitations with regard to the research design for examining stakeholder perceptions. We could not distinguish among the opinions of different stakeholders, and we focused on a general category of "potential users of nonfinancial disclosure." Further research could focus on the different expectations of different categories of internal and external stakeholders. Stakeholders' perceptions can change depending on the group to which they belong and the impact of the damage caused by the company on such a group. Stakeholders who are directly affected by the damage caused by the company may be more engaged and interested in the company disclosure and more sensitive in the evaluation of companies' remedial action and, thus in the detection of any sign of hypocrisy.
Moreover, although this research focuses on the role of companies' communication strategies, we recognize the impact that counter accounts can have on stakeholder perceptions. Therefore, we encourage further research to include counter accounts in new experimental designs and observe how they affect stakeholders' reactions.