Companies operate in a broad social context and their activities affect various social groups including employees, suppliers, customers, government, and citizens in general, and the perpetuation of their existence depends on addressing the issues pertaining to this audience (Cormier et al., 2005). These groups are prominent in society and play a relevant role in relation to disclosure (Khlif et al., 2015), as they generate the need for companies to be responsive (Deegan, 2002; Vigneau et al., 2015), which makes the reports in SR to be influenced by the social values of the community (Deegan & Rankin, 1996).

The institutional forces of society are factors in companies’ commitment to the environment (Albertini, 2014). They generate social pressures around public demands and become objects of questioning from the scrutiny of pressure groups, media and authorities (O'Donovan, 2002), which makes companies act preventively to avoid negative reactions (Villiers & Alexander, 2014).

Most of the pressures trigger legitimizing responses of strategic importance (Khlif et al., 2015), and corporations try to change people’s perceptions, aiming to repair or defend their legitimacy (Suchman, 1995; O'Donovan, 2002), through omission (Leung et al., 2015), and expansion of disclosure (Vourvachis et al., 2016), considering that the disclosure of information draws the attention of different audiences around the company’s reputation (Longoni & Cagliano, 2018).

In this way, such institutional pressure factors contribute to the expansion and omission of environmental disclosure, exposing companies to the public (Etter et al., 2018); (Longoni & Cagliano, 2018), and increasing corporate vulnerability (Roberts, 2003).

The expansion of social media in people’s daily lives has given autonomy to ordinary citizens who manifest themselves autonomously in public arenas, where the activities of organizations are continually judged and evaluated (Etter et al., 2017).

The media is the most prominent source of perceived legitimacy (Cormier & Magnan, 2015). Published news influences the public’s perception of businesses, and ordinary citizens use this discourse as a source of information to assess the adequacy of organizational actions. Positive evaluations and judgments can be considered as legitimizing organizations, however, negative evaluations delegitimize companies. Therefore, news broadcasts in the media constitutes the most extensively explored source of organizational legitimacy (Etter et al., 2018). They address issues that generate concern among opinion-forming groups (Christensen et al., 2013), and the more the media reacts to a company’s activities, the greater its susceptibility to pressure (Dienes et al., 2016).

Research has paid special attention to the influence on the environmental disclosure of media publications, whether favorable or neutral or unfavorable (Cormier & Magnan, 2015). They shape the public agenda and the greater prominence and attention they give to specific themes results in increased community concerns impacting public attitudes. Furthermore, approaching issues from a negative perspective is recognized by society as being of utmost importance. For this reason, a negative media agenda has a greater effect on public opinion in relation to an issue with a positive or favorable focus (Deegan et al., 2002).

A study in Australian companies identified that executives used annual reports (AR) to respond to inquiries made by the public. Which caused media reports, particularly negative or unfavorable ones, to result in responses in corporate reports, by the fact that executives consider publications made by newspapers when preparing the AR (O'Donovan, 1999).

In this way, the media influences society's perception of the image of companies, because news reports subject companies to pressure, leading them to expand positive disclosure in response to negative news.

The regulatory environment and the adoption of global reporting standards create demands that encourage the conformation of institutions, influencing environmental disclosure (Barbu et al., 2014), being recognized as one of the main factors for expanding and encouraging disclosure (Villiers & Alexander, 2014). In this sense, the efforts of companies to follow standards such as the GRI for the preparation of SR, has as its main objective to cultivate a positive image of the companies (Beck et al., 2010).

Specific legislation often obliges companies to assume responsibility for the remediation of hazardous waste and more recently has led to the recognition of significant environmental liabilities (Cho et al., 2015). What has made organizations ensure their legitimacy by responding to the growing institutional pressure they are subjected to through environmental norms and standards (Christensen et al., 2013). Regulatory stimuli have led executives to voluntarily report environmental information in accordance with community expectations. In this way, they respond to the negative repercussions that arise in the media, in order to avoid the introduction of new restrictive regulations to the operations of companies (Deegan, 2002).

In this sense, there is greater voluntary disclosure in firms that operate in environments with weak regulation and the same level of globalization. The regulatory environment also creates demands for voluntary disclosure by multinational companies, more prone to informational asymmetry due to the size and complexity of operations (Webb et al., 2008). As a result, large companies disclose more than smaller companies and the level of environmental disclosure is greater in countries whose governments have chosen to establish stricter environmental laws. Thus, the level of environmental disclosure depends on the legal system, as well as the social and financial contexts of the country in which companies operate (Barbu et al., 2014).

Industries that operate in high environmental impact sectors have become hot spots and face more stringent regulatory environments (Hassan & Guo, 2017), which helps them to increase environmental disclosure (Albertini, 2014; Clarkson et al., 2011; Al-Shaer et al., 2017; Arora & Lodhia, 2016). The environmental impacts of their activities expose them and threaten their legitimacy, leading them to act defensively with ostensive actions to repair and increase environmental disclosure (Cormier & Magnan, 2015).

However, the quality of disclosure is lower in these firms (Cormier & Magnan, 2015), characterized by being of the more extensive type (Braam et al., 2016), showing environmental information with a lower degree of specificity (Albertini, 2014). Such polluting industries reduce specific disclosure, preferring to disclose more general information, fearing that this will be harmful to their reputation (Villiers & Staden, 2006).

In response to the institutional pressures they receive, companies operating in potentially polluting sectors are more inclined to selectively disclose environmental information (Leung et al., 2015). They expand environmental disclosure after an accident (Patten, 1992), emphasizing corrective actions taken because of this (Arora & Lodhia, 2016), without, however, specifying the impacts caused.

Thus, the environmental disclosure of potentially polluting sectors is directly related to pressures from society, the media, and the regulatory environment. It may be associated with periods when sectors became targets of criticism from conservationists for damage to the environment (Guthrie & Parker, 1989). This indicates reactive approaches by companies, in a negative relationship between environmental performance and disclosure (Hassan & Guo, 2017).

This reaction was verified in research investigating the 1989 Alaskan oil spill, which aimed to detect the impact of the accident on environmental disclosure. The study found that the oil company Exxon gave special coverage to the spill and the clean-up work in the SRs of that period, devoting 3.5 pages to the accident and another 2.5 pages of environmental disclosure which was unrelated to the accident. In contrast, only 0.6 pages of environmental information was included in the company’s SR in the year before the disaster (Patten, 1992).

Research in RA and SR found that the level and nature of disclosure was related to environmental performance, because the firms that pollute the most publish the most environmental information, relying also on hard disclosure (Clarkson et al., 2011). Another study showed that environmental disclosure was significantly higher in the years where companies had been sued, in order to neutralize the effects of litigation (Deegan & Rankin, 1996), confirming that the level of disclosure is a function of the company’s exposure to pressures. Therefore, companies with poorer performance have greater environmental disclosure (Cho & Patten, 2007), Table 1.

Table 1
Environmental disclosure in sectors with high environmental impact

Source: Authors

It appears, then, that research on environmental disclosure in high-impact sectors indicates a trend of information expansion (Albertini, 2014; Ahmadi & Bouri, 2017). Companies operating in these sectors are subject to greater public scrutiny (Delgado-Márquez et al., 2015), to tighter controls (Hassan & Guo, 2017), although some have lower environmental performance (Braam et al., 2016; Hassan & Guo, 2017). Industries in these sectors expand environmental disclosure after the occurrence of adverse events (Guthrie & Parker, 1989), as a result of processes (Deegan & Rankin, 1996), and tend to report negative information superficially (Albertini, 2014). They omit specific content (Villiers & Staden, 2006), expanding positive disclosure to shift the focus from adverse information (Arora & Lodhia, 2016).

The rupture of the containment structure at the Fundão dam released 43.7 million m³ of tailings (GRI, 2016). Enormous damage was caused to indigenous peoples, surrounding populations and distant cities within a radius of up to 663 km from the dam (Demajorovic et al., 2019), extending across two states of the federation. The disaster forced the company to halt its operating activities causing unemployment (GRI, 2016), and loss of income for families. There was a sudden stoppage of essential services with surprising demand for rescue and rescue services, health care, as well as social and psychological assistance to the affected population (IBAMA, 2015).

Due to the wide repercussion of the accident in the media, only news published on websites of newspapers, magazines with large national circulation, and the communication agency of the federal government, were verified. The survey, which covered the period of one year from the date of occurrence of the disaster, identified that 86.6% of the news were predominantly of a negative nature and only 9.8% were positive. It is also noteworthy that due to the damage caused, only 3.7% of the news published were neutral and without positive or negative bias, such was the polarization that the disaster generated, Figure 1.

Figure 1.
Nature of disclosed news
Source: AuthorsNote: Research data.

The Fundão dam broke at 4:20 pm on November 5, 2015 and the fact was quickly reported by the media. In the days that followed the disaster, the reports were reported in the newspapers. The printed newspaper O Estado de São Paulo had a headline the following morning: “Flood of mud buries and kills in Mariana” and on November 7th, it already pointed to the company’s accountability, showing on its front page that “Study warned in 2013 of risk of rupture” (Amorim, 2019). The repercussion of the accident in the media and the consequences for the communities and the environment led the authorities to determine the responsibilities. The investigations led to the indictment of professionals involved in the disaster and directly implicated the company, compromising its reputation, as Samarco was the owner, and responsible for the management and operation of the Fundão dam.

Given that Samarco’s environmental disaster was classified by IBAMA as a “very large disaster”, an attempt was made to compare the form of SR reporting with IBAMA’s preliminary technical report, which aimed to support the proposition of a public civil action for the accountability of Samarco. The report found the “direct destruction of ecosystems, damage to fauna, flora and socioeconomic”, in addition to the loss of balance in the Rio Doce hydrographic basin. The impacts were divided into five types covering areas of permanent preservation, fauna, ichthyofauna, water quality, in addition to socioeconomic impacts, Table 3 (IBAMA, 2015, p. 2).

The comparison of the IBAMA report with the content published in RS shows the divergence of reports. The SR briefly reports that all impacts were identified, while the IBAMA report mentions the devastation of riparian forests, burial, suppression, and uprooting of trees. On the one hand, IBAMA reports the interruption of fishing for an indefinite period (IBAMA, 2015), but the company’s SR cites an “abundance of fish” (GRI, 2016, p. 77).

Table 3
Impacts of dam failure

Source: Authors

In the item related to water, the IBAMA report shows a profound and perverse impact. However, SR minimizes the destructive impacts with the narrative that the tailings plume only temporarily affected the water uptake of the Doce River. The report mentions that populations of small animals were probably decimated, making it impossible to estimate the return of fauna. About this, SR only reports that a large amount of information about the health of animals has been accumulated.

The socio-economic impacts in the report mention the destruction of buildings, bridges, roads and equipment; damage to medical care, public health and emergency medical services; compromise of the rainwater and sewage system. However, such impacts are omitted in SR, in which, on the contrary, there are highlights for actions performed by the company, with the supply of medicines, medical equipment, health agents, and actions to clean public spaces. The company expands the reports of civil works, such as the rescue of architectural assets, but omits that the tailings mud invaded churches in cities of recognized historical value (IBAMA, 2015; GRI, 2016).

Although the company has dedicated special coverage to the accident with a focus on repair and containment work, the survey found a marginal increase in the number of pages in SR, which went from an average of 95 pages in the period prior to the accident, to an average of 99 pages after the crash. On the other hand, an inverse trend was verified in the number of photographs published in SR, going from 40 photographs before the dam burst to 37 after the breakup. This reduction implied a drop in the disclosure of photographs, which went from an average of 14.58 pages in the period prior to the accident, to an average of 11.21 pages in SR 2015/16, Figure 2.

Figure 2.
Environmental Disclosure of Photographs
Caption: Environmental Disclosure of Photographs: dimensions of photographs in relation to the page.Source: Authors

It should be noted that the SR published by the company after the disaster in November 2015 comprises two years, 2015 and 2016, and unlike the others, it reports an atypical situation of the company, which had its operations halted by determination of government authorities, which made the longitudinal analysis of the reports difficult.

The analysis of textual disclosure identified changes in the content of SR published before the disaster in relation to the content published after the disaster. In the latter, the environment continued to be highlighted, however, issues related to society started to receive greater attention as a result of the image repair strategy, primarily aimed at the external public.

The content of the 2015/16 SR was predominantly related to environmental disaster. This theme was distributed throughout the sections of SR with explanations of a justifying nature, works of remediation and compensation, but, above all, actions to assist communities affected by the disaster.

In the sections with content related to the environmental disaster, which corresponded to 44 pages of the SR, we found that 40 pages (90.9%) had repair actions, damage containment, and assistance to communities. On the other hand, only four pages (9.1%) contained negative information about the disaster. In the section About the Burst of the Fundão Dam, the counting of paragraphs showed similar numbers, with six paragraphs of negative disclosure (5.7%), and 99 paragraphs (94.3%), reporting repair and assistance actions performed by the company, Figure 3.

Figure 3.
Textual Disclosure of the Environmental Disaster in SR 2015/16
Caption: number of pages and paragraphs in relation to the total number of pages in SR.Source: Authors

Research numbers point to the expansion of positive disclosure, reinforced by the low environmental disclosure of negative information related to the disaster. In its SR, the company expands the disclosure of the positive actions it takes, while omitting the consequences of the dam failure. This increase in disclosure is a response by the company, which seeks to counteract the expansion of information of a negative nature widely publicized about it by the media after the tragedy occurred.

fceamorim@gmail.commtereza@fei.edu.br