Disclosure of environmental matters - Galp Energy

There are a growing number of companies expressing social concerns where environmental worries are also included. Even before legal obligation of environmental disclosures, companies already used to adopt an active stance to safeguard their reputation. However, risk events may have negative impacts on companies’ legitimacy, making it necessary to implement strategies to repair / recover the damaged reputation. This study analyses the annual reports from 2001 to 2011 of the largest Portuguese oil company Galp Energia in the view of the occurrence of negative events, which, consequently, may have affected the company's reputation. Empirically, legitimacy theory can explain the implementation of strategies to repair reputation. Within this context, most of the existing research is based on common-law countries, with little literature grounded on code-law countries, such as Portugal. Findings have shown that Galp Energia has great concerns about the external perception of society. The company acts immediately after risk events, implementing strategies to minimise the risk impact, especially through the use of


Introduction
More and more companies have begun to integrate environmental and social concerns into their daily operations. In fact, this is not a new concept, even though standardisation in this area appeared only in 2003. Until then, manifestations in the area of social responsibility were very scarce, and many companies disclosed their social and environmental issues voluntarily (Gray et al., 1995;Wilmshurst and Frost, 2000;Deegan, 2002;Toms, 2002;Hossain et al., 2015).
However, if it is important for companies to assume their social responsibility and act accordingly, it is also important to describe and pass on such information to stakeholders. Companies' annual reports are identified in several studies as one of the best and most effective way of corporate communication with stakeholders and one of the best tools to create, protect, and improve their image or reputation. Consequently, many companies use annual reports to disclose environmental information voluntarily (Stanton and Stanton, 2002;Branco, 2006;Barros, 2008;Eugénio, 2010;Monteiro and Guzmán, 2010a). It is a powerful tool through which companies are able to influence the external perceptions of themselves (Unerman, 2008). To Ferreira (2004) environmental reports appeared to improve the public image of the company, withdrawing attention from other areas, bringing political benefits and, above all, giving them some experience, even before this kind of reporting became mandatory. However, the Portuguese companies, despite having a higher level of environmental disclosure than expected, they have a very descriptive, qualitative and self-complimentary type of reporting. They only report information that might contribute to a bigger and better reputation (Ferreira, 2004).
According to Bebbington et al. (2008), quoting Fombrun and Van (1997: 10), reputation is defined as "a collective representation of the past actions and results of a company, describing a company's ability to return valuable outcomes for the various interested parts." Reporting plays an important role in reputation risk management, assigning a direct relationship between the disclosure of social responsibility and the increase of the company's reputation. Branco (2006) concluded that companies with good reputation in social responsibility develop better relationships with its stakeholders and employees, which will lead to the construction of a positive corporate image and, therefore, a better financial performance. Corporate reputation is pointed out by the author as the most valuable intangible asset of an organisation, as well as one of the most difficult to create, maintain and / or recover, thus justifying its strategic value. Unerman (2008) stated that one of the main reasons for companies to disclose social / environmental information is the desire to minimise reputation risks. Reputation is a valuable asset that must be protected and developed. One of the key aspects to consider is the perception of how policies, practices and social responsibility results meet the values and the social and environmental expectations of its stakeholders. Roberts andDowling (2002: 1091) considered that "developing a good reputation takes a considerable time, and depends on the stable and consistent investments of the company over time." However, the company's reputation should not be reduced to the positive information disclosed by a company. The media or financial markets also play a key role. Therefore, the task of maintaining constant, good performance is quite difficult.
Some authors (Barros, 2008;Branco, 2006) suggested that the impact of social responsibility does not manifest itself equally in every company. It is important to consider the company's characteristics and its influence on society and environment. Environmentally sensitive companies will have a greater impact on their socially responsible activities. Because their stakeholders are far more sensitive to error, they should be aware of the difficult mission of creating a positive corporate reputation, making efforts to maintain it, and recover in case of an incident.
Despite the weight disclosure of social and environmental responsibility information may have in company's reputation, the costs and benefits of disclosure must be considered. According to Branco (2006), a sustainable long-term business is more important to companies than the impact of short-term disclosure.
According to Eugénio (2010), the motivations that lead to the disclosure of environmental matters are quite different. One of the most relevant factors is the desire to legitimise a company's activity. Even before reporting became mandatory, companies were already using environmental reports voluntarily as a way to obtain legitimacy (Monteiro and Guzmán 2010b).
"Legitimacy is a perception or a widespread assumption that the actions of an entity are desirable, proper or appropriate within some socially constructed system of norms, values, beliefs and definitions" (Suchman 1995: 574). Legitimacy theory is the explanation for the increase in environmental disclosure since the early 1980s and became the most quoted theory in environmental and social accounting (O'Donovan, 2002;Linsley and Kajüter, 2008).
This theory argues that companies only exist because society considers them legitimate and it is this same society that defines the right of companies to use resources and subsist (Hossain et al., 2015).
According to Eugénio (2010: 113), "the disclosure of social and environmental information can be seen as a technique to legitimize the company's activity in order to make known what the company does for environmental protection and to support the society". Thus, this is the theory that best explains the motivation of managers to disclose environmental and social information in the annual reports, providing many benefits for the organisation.
To Monteiro and Guzmán (2010a), legitimacy theory suggests that the disclosure of environmental information by companies results from the daily social pressures. Thus, they choose to disclose more and more environmental information to preserve their image of a legitimate company and avoid negative consequences that could result in legitimacy crises. O'Donovan (2002) examined to what extent the disclosures are related to attempts to gain, maintain or recover legitimacy as well as tactics / methods to legitimise the company. Branco and Rodrigues (2008) found that the dissemination of good social and environmental behaviour influences external perception, which will benefit the company's legitimacy. In turn, Wilmshurst and Frost (2000) report that, in the context of disclosures, legitimacy theory can be seen as a reaction to external problems. Linsley and Kajüter's (2008) intended to analyse the legitimacy theory versus legitimacy repair strategies through the interpretation of a risk event in the Allied Irish Bank. The study corroborates that strategic actions for the recovery of legitimacy were implemented. However, it was also verified that the repair attempt was unsuccessful. According to Tilling (2004), four strategies can be found in this theory, namely establishing, maintaining (most difficult stage), increasing and protecting / recovering / restoring, and each company can adopt each of these depending on the "state" of its reputation.
Most of the existing studies on social and environmental issues focus on the latter and before environmental incidents, and they try to support the idea that the discourse may be different after the loss of reputation in an attempt to repair or recover it. The different types of strategies depend on the objectives to be achieved regarding legitimacy. Their implementation and success depend on a good interpretation of the origin of the event.
According to Linsley and Kajüter (2008), the strategies consist of the following: writing a report (trying to excuse or justify the event), being in institutional compliance (to emphasise the company's processes and actions with an institutional basis), carrying out a strategic restructuring (distance the entity from the event), and remaining calm (since panic is often misunderstood). Bebbington et al. (2008) also conducted a survey of the possible strategies / actions in the face of a negative event, based on the study carried out by Benoit (1995). However, the approach is slightly different. For this author we can have three actions: denial, evading responsibility and attempting to reduce the impact caused. If a denial of the event occurs, then we may be facing a simple denial or denial with guilt transfer. On the one hand, if the entity intends to evade responsibility, they might do it through justifications. Finally, if their intention is to take responsibility, trying to minimise the impact on reputation, they might adopt strategies such as support, differentiation, compensation, implementation of corrective actions and even humiliation / shame. Taking responsibility for the event can often be a driving force to minimise the impact caused, and it is more valued by the interested parties, as far as the reasons presented are justified in an appropriate and consistent manner. Aerts (2005) also refers this kind of strategies. For this author, there is always a tendency in entities for a selfish pattern, i.e., as a rule, those events with positive impacts derive from their own actions or have social origins, and the events with negative impacts derive from external actions (taking the "good" as something internal and the "bad" as something external / non imputable to the organisation).
The author believes that in the occurrence of results that might affect the company's image three different strategies can be adopted: the excuse (the entity assumes the occurrence of the event but denies responsibility), justification (resulting from the assumption of responsibility, aimed at reducing the impact), and denial (of responsibility and / or internal causality -though it might simply be the case). Thus, this author supports two models of management: the information model and the impression model, which often relies on self-proclamation.
The study of Bebbington et al. (2008) suggested the study of pre and after reporting of adverse events and their impact on the company's reputation through a longitudinal analysis of one or more companies. As a result of their findings, Linsley and Kajüter (2008) also suggested that useful future studies will have to be based on the analysis of companies that were faced with problems of legitimacy after the occurrence of risk events, which might have resulted in the reduction / loss of legitimacy.
Moreover, Stanton and Stanton (2002), showed that if an annual report has the power to enhance corporate image, it can also reveal conflicts. For example, what are annual reports like in the face of an environmental incident? On the other hand, according to the study of Eugénio (2010), the legitimacy theory is the explanatory basis for the disclosure of environmental matters. However, if an environmental incident occurs, resulting in a reputation loss, will this theory continue to support the choice of specific communication strategies to recover company's reputation?
Consistently, our study tries to analyse only the annual reports of Galp Energia within a time frame of 10 years (2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011). Our main research objective is to determine whether the legitimacy theory can be used to explain the adoption of legitimacy restoration strategies in "code-law" countries like Portugal. More specifically, grounded on legitimacy theory analyze in order to verify which strategies are implemented by a Portuguese environmentally sensitive company (such as Galp Energia) to repair / recover the affected legitimacy when faced with an environmental negative event.
The existing literature defines legitimacy theory as the founding base for topics related to environmental and social issues (O'Donovan, 2002;Barros, 2008;Bebbington et al., 2008;Patty and Penn, 2011). On the one hand, this theory provides an explanatory framework for environmental and social disclosures. On the other hand, the issues that have been raised regarding the explanatory power of this theory must also be considered, since there are other internal and external factors that should be studied. Thus, this theory addresses four strategies: establishing legitimacy, maintaining legitimacy, increasing legitimacy and defending/recovering/restoring legitimacy. Existing studies focus primarily on this last phase, and on environmental incidents, trying to corroborate the idea that discourse may be different after reputation loss in an attempt to repair or recover it (Deegan. 2002;Aerts, 2005;White, 2006;Bebbington et al., 2008).
The research setting in this study is interesting because it has the ability to contribute to extend prior literature in several ways. First, the literature on legitimacy restoration strategies of reputation losses, particularly in environmentally sensitive companies, focused on common-lawbased countries of Anglo-Saxon origin (Wilmshurst and Frost 2000), excluding other nations, where a different legal framework exists. Among them, Portugal is considered one of the least developed countries in the Euro zone, with a very specific capital market, where the main stakeholders are investors who fund the banking sector. Second, Second, Monteiro and Guzmán (2010b) admit that research in this area is not common in Portugal, meaning that the existing literature is rather scarce. Third, according to Barros (2008) and Branco and Rodrigues (2008), compared to the English-speaking countries (common-law), the determinants of social responsibility information disclosure in Continental Europe (Code-law), particularly in Portugal, are still unknown. However, Monteiro and Guzmán (2010a) suggested a content analysis of reports and long-term accounts, according to the type and quality of information that is being disclosed by the Portuguese companies (and not by the determinants that lead to this disclosure). Finally, the Portuguese company Galp Energia considers that social responsibility is a determining vector in the company's reputation. The company believes that an appropriate social responsibility status is, without doubt, very difficult to achieve, but very easy to affect negatively (Branco, 2006;Barros, 2008;Bebbington et al., 2008;Kamal and Deegan, 2013). Galp Energia is a environmentally sensitive company. Consequently, it is often confronted with events that have a negative impact on the environment, and for which it must be prepared under the penalty of losing its legitimacy.
In the following section, we explain our research method. Thereafter, we report results and present conclusions.

Case study description (Galp Energia)
The Portuguese business sector is mainly composed of small enterprises, to which social concerns and environmental protection issues are not so relevant. Most studies conducted in this area are based on Anglo-Saxon countries. Therefore, there is very little literature on code-law countries, such as Portugal. In that sense, we choose a Portuguese company, environmentally sensitive (with an environmentally aggressive activity and with concerns in this regard), certified in the field of social and environment responsibility, with great responsibility towards the stakeholders, and concerned about its reputation. The company chosen for this study was Galp Energia.
Galp Energia is the only oil refining enterprise in Portugal and it is a leader in refined petroleum products (distribution in the retail and wholesale sectors). It is also one of the most environmentally sensitive Portuguese companies, with greater environmental risk. Galp Energia has always been one of the Portuguese companies with the strongest awareness of environmental issues. Over the years, the company has committed itself to disclose valuable environmental information, not only qualitatively but also quantitatively with regards to provisions of costs and investments.

Data collection and method of analysis
We have defined a ten year time frame to analyse the financial reports of the Galp Energia: from 2001 to 2011. Ferreira (2004) used a sample of Portuguese companies in which he included Galp Energia, and analysed their annual reports within a time frame of five years (1997)(1998)(1999)(2000)(2001). Our study aims to go beyond Ferreira (2004) trying to analyse the annual reports of the company referred above, which was selected for our study, within a time frame of 10 years (2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011).
Our study is based on the analysis of these annual reports, since annual reports are considered companies' elected tool to disclose social responsibility. Moreover, they are easy to access. When we started collecting data for our study, it was important to determine the type of data to be taken into consideration. First, we decided to only include non-financial data related to our theme. In addition, we limited our research to the occurrence of events with a negative impact on the company's reputation, the strategies used to handle such impacts, and the search for evidence regarding the company's attitude towards those events.
Based on previous studies we conducted a survey on the strategies mentioned in each study to repair/recover legitimacy (Aerts 2005;Linsley and Kajüter 2008;Bebbington et al. 2008). Starting from the point of view of these three authors, we created a summary table, classifying those strategies into three possible types of events by associating the objective to be achieved regarding legitimacy and the different strategies / sub-strategies to be implemented to achieve the intended goals (Table 1).
(Insert Table 1 here) When we characterised the type of event, we considered the impact that the event would have on corporate image / reputation, and we associated it with the goal the company wanted to reach. Thus, when dealing with a positive event, the company should take advantage of the situation to gain or increase its legitimacy. The same should happen in the occurrence of neutral events, after which the company should try to maintain or increase its reputation. When a negative event occurs, affecting the company's image, the company should adopt strategies to repair / recover the affected/lost legitimacy.
Regarding the strategies adopted, we have assumed that the company uses selfproclamation as a method to improve its image, whenever a positive or neutral event occurs, attributing the positive aspects to internal actions, either their own actions or of social origin. On the other hand, if a negative event occurs, the company will rely on its defence, either through denial or assumption.
However, regarding the latter two strategies, we have listed several sub-strategies through which the company seeks to defend itself. Among those listed by the three authors, we selected the most common ones for our study. They were later used as a standard / guide for the data collection of our sample. Table 2 shows the sub-strategies chosen, and provides a brief explanation of each of them (Aerts 2005;Linsley and Kajüter 2008;Bebbington et al. 2008): (Insert Table 2 here) According to the 2011 annual report "the social responsibility of Galp Energia covers four areas: education, environment and energy efficiency, health and wellness, prevention and road safety "(Galp Energia 2011: 65). A content analysis was carried out to analyse all annual reports, identifying all the events with environmental impact in a first stage. Then, a second careful reading was done to identify the sections / chapters of the Annual Report where these events were mentioned. A third reading was performed, focusing on the key chapters only. Each chapter was carefully analysed in order to detect the legitimisation strategies used by the company to repair, recover, increase, or maintain legitimacy. The annual reports were analysed individually, bearing in mind if the report illustrated a neutral year or with incidents, causing a negative impact on corporate reputation / image.

Presentation and Interpretation of the Results
We will present and interpret the results in two ways: first we analyze the company's posture before the negative events and relate such events with previous studies. Then, we analyze the strategies followed by the company to restore its legitimacy.

Results
In some years (2001, 2002, 2003, 2005, 2006 and 2007), there were no incidents with negative environmental impacts, but there was a great concern with social and environmental issues, since reporting on social responsibility was observed for every year. Although no incidents were reported over the years, serious incidents occurred in the following years, which, therefore, deserved our attention. Table 3 presents a summary of those incidents.
(Insert Table 3 here) Over these 10 years, Galp Energia always had a great concern for the society, keeping its voluntary disclosure levels constant, and developing innovative measures in the context of social responsibility in order to fulfil its responsibilities towards the community and the environment. However, this analysis did not allow us to draw specific conclusions on the application of the legitimacy recovery strategies after risk events. No significant differences were found that allow us to assign a direct relationship between the negative event and the application of legitimisation strategies.
Consequently, a broader analysis of other type of information was conducted: the President's message, relevant facts occurred during that period, accidents, safety, health and environment. Regarding the negative events occurred in a given year and subsequent years (Appendix), the data collected on the occurrence of incidents, the applied strategy to recover reputation, and the examples of quotes from the annual reports that directly identify the strategy applied are the following:  Year 2004 -an accident at a construction site in the Leixões Terminal occurred, resulting in two different assumptions of the event. One of them consisted in a simple assumption, in which less positive financial results in exportation were reported; the other one included the corrective actions implemented to ensure that the incident would not repeat, in order to minimise the impact on reputation;  Year 2005 -although an increase was observed in the dissemination of the social responsibility measures implemented, it is worth noticing that the accident which took placed at the Leixões terminal is mentioned in topics such as provisions and claims;  Year 2008 -a new incident occurred, this time involving a collaborator. The incident is mentioned in the annual report, as well as the application of the measure to recover legitimacy and the implementation of corrective actions.  Year 2009 -there were five incidents. Three of them with personal injuries and two with property damage, leading to the implementation of two types of strategies, both of defensive type (one of denial and another of assumption with implementation of corrective measures).  Year 2010 -there were three serious accidents involving personal injuries and one accident that has resulted into death. Three of the accidents occurred in conditions not imputable to the company, and the other resulted in the assumption and implementation of measures to prevent the occurrence of further incidents.  Year 2011 -incidents occurred and strategies to recover reputation were again reported, namely, denial and assumption through corrective actions. However, in these two years there was no attempt to handle the loss of legitimacy caused by the incidents of the preceding years.  All the evidence found is a response to the events of the same year. Table 4 shows a summary of the results found.
(Insert Table 4 here) The results show a constant dedication of Galp Energia in their annual reports to social responsibility. In fact, this has been a subject of a great concern to the company. Over the years, the company started to introduce a chapter (either a few lines, or a complete page) on this topic in their annual report disclosing information about all the measures developed and applied. Ferreira (2004) found that, until 2001, 90% of the sampled companies showed constant environmental concerns. Galp Energia was included in that group. Thus, in this study, we carried out an analysis of the next ten years (2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011) and we also noticed strong constant concerns.

Discussion
The web page of Galp Energia also corroborates the results of Monteiro and Guzmán (2010b). There are some companies that have been disclosing information voluntarily on social responsibility for several years now. Therefore, we have concluded that Galp Energia is one of the companies that did not wait for the legal requirements to start disclosing information on this matter. The authors mentioned have found that "the disclosure practices of companies are influenced by two distinct forces: social pressures (caused by the demands of their stakeholders) and legal requirements" (Monteiro and Guzmán 2010b: 415). This statement can be corroborated by the existence of voluntary disclosure practices, leading us to infer that companies disclose information due to social pressure, that is, to satisfy their stakeholders. Our study corroborates the study conducted by Branco and Rodrigues (2008) and Bebbington et al. (2008). Galp Energia has a strong public visibility and, as such, it is greatly concerned about improving its image. Simultaneously, our findings also allow us to corroborate the study carried out by Branco (2006), since Galp Energia has a good reputation in terms of social responsibility and good relationships with external stakeholders (clients, investors, suppliers and competitors).
Like Gray et al. (1995), we also found a direct link between voluntary disclosure and corporate reputation. Galp Energia is a leading company in the adoption of voluntary disclosure on social responsibility, which also allows us to believe that the main reason for such behaviour is to maintain their reputation.
According to Linsley and Kajüter (2008: 68), reacting in a calm and neutral manner when handling an incident, avoids misunderstanding of the impact caused. Galp Energia is a good example to support the previous statement. In the occurrence of such events, the company provides a report of the incident, explaining the measures taken without resorting to panic, which could lead to further loss of legitimacy. In the years when the company is affected by incidents, and in the following years, we have identified some of the strategies used by Galp Energia to recover legitimacy, corroborating the study of Wilmshurst and Frost (2000). Galp is likely to assume the error and take responsibility for it, if it is their responsibility, showing an attitude of assumption. In the present study, we found that in order to minimise the impact on their reputation, Galp Energia uses (mostly) the implementation of corrective actions. This measure is mentioned by Bebbington et al. (2008) as an effective measure to reducing the "offense" caused by the incident.
Over the ten years of our study, we found that whenever an incident with a negative impact occurred, Galp Energia opened investigation proceedings, not only to correctly detect the cause, but also to attribute responsibility. On the one hand, they assume their own responsibilities. However, they also resort to denial, as in case there is no evidence of being held responsible for the incident. According to Aerts (2005), Bebbington et al. (2008) and Linsley and Kajüter (2008), denying the event is a very common attitude, and it can either be a simple denial (such as with Galp Energia), or a denial with guilt transfer.
However, the sample also includes years with no evidence of events with a negative impact. We have defined those years as Neutral. The analysis showed that even though the level of disclosure remained constant, a self-proclamation type of reporting was also found, as a way to increase / maintain Galp's reputation. These results support the conclusion drawn by Ferreira (2004: 572): "so far, the information disclosed voluntarily by Portuguese companies is mostly descriptive, qualitative and self-proclamation."

Results
Regarding the possibility of using the legitimacy theory to explain the adoption of legitimacy restoration strategies in the disclosure of environmental issues in Portugal, all the 10 annual reports analysed showed evidence of concerns with corporate legitimacy (Figure 1):  10 reports showed concerns about increasing / maintaining legitimacy through selfproclamation, disclosing social responsibility measures;  5 reports showed concerns about repairing / restoring the legitimacy affected by negative events, through the implementation of strategies such as denial, assumption and corrective actions;  None of the reports showed evidence of carelessness or negligence regarding the company's reputation.
(Insert Figure 1 here) The present study focused mainly on the negative events, as these were the events which caused the greatest impact on the company's reputation. As previously mentioned, this happened because the general public, and especially the stakeholders, are much more sensitive to a negative incident than to a positive event. For those interested, the main expectations were good results and good company's actions. So, when the opposite happened the reaction was quite controversial and difficult to forget.
Galp Energia has been working on its own legitimacy for several years now. The voluntary disclosure of social and environmental information is a clear sign of their work. As a big company they have a great reputation to maintain, besides being environmentally sensitive, and are more affected by the negative impacts. Whenever an incident occurs, there is evidence of their concern to repair / recover the affected reputation.
This is exactly what the legitimacy theory conveys: the constant concern to legitimise the company before the society and all the stakeholders. Without legitimacy the company cannot operate in the same way, since the perception held by society is crucial in the formation of a favourable opinion. Bebbington et al. (2008: 354) concluded in their study that "reputation and the importance of Corporate Social Responsibility and sustainability are clearly related, particularly in more developed markets." However, we also found some evidence of this relationship in our study, supporting the idea that in less developed markets, such as the case of Portugal, the company's reputation is also related to their social responsibility concerns.

Discussion
"Corporate reputation is directly linked as a driver of customer satisfaction thus establishing a direct link between CSR and customer satisfaction." (Bebbington et al. 2008: 354). Throughout our study, we cannot claim to have found concrete evidence that disclosures were directly linked to stakeholders' satisfaction. However, we are able to infer that if the company's annual report is the primary means of communication between the company and its stakeholders and if the company's reputation is directly linked to these, then there is indeed a link between reputation, social responsibility and stakeholder satisfaction.
In short, we can conclude that legitimacy theory is in fact a good framework to explain the reasons for the disclosure of social and environmental information, and in the present case it is a good support to explain the implementation of communication strategies after events with negative impacts on the environment and society. Galp Energia is representative of other Portuguese environmentally sensitive companies. This allows us to conclude that in Portuguesespeaking countries (code-law) the legitimacy theory is likely to explain the implementation of legitimacy strategies after the occurrence of risk events.
In the course of our study, we also found some limitations. For example, the nonoccurrence of incidents in each of the years included in our sample, the lack of chapters where accidents were addressed, very brief descriptions of the measures implemented to manage legitimacy (the company may use other means to disclose this type of information that we did not analyze). The present study focused in only one company which reduced the possibility to draw general conclusions. Additionally, the annual reports of those ten years were the only selected mean of communication.

Conclusion
Over the years there has been a significant increase in the disclosure of environmental issues. The environment is a public good that should be a concern for companies, under the penalty of being judged by society. The inclusion of environmental and social practices in the day-to-day business is noticeable, not only in terms of disclosure, but also in the development of new strategies in the area of social responsibility. This level of concern is largely related to the maintenance of a good reputation. However, keeping it is not an easy task, given that businesses are daily faced with events / incidents that may jeopardise reputation.
This study analysed the annual reports from 2001 to 2011 of the largest Portuguese oil company -Galp Energia -taking into account the occurrence of negative events that consequently affected the company's reputation. The main aim was to understand if Legitimacy Theory explains, somehow, the communication method / strategy for environmental matters, and to investigate the application of legitimacy strategies after the occurrence of environmental incidents that have resulted in reputation loss for the company.
In the ten-year period under analysis, we found evidence that Galp Energia discloses information on social responsibility in its annual reports, which includes environmental issues. Despite the introduction of legal requirements in 2003 to make disclosure mandatory, the company had already adopted a strategy of voluntary disclosure. This has always been a way of communication with society, allowing the company to build a great reputation over the years. Our findings confirm that the legitimacy theory can explain the communication strategies used by Galp Energia after the occurrence of negative events.
Over the ten years under review, 14 accidents were detected. Based on legitimacy theory we found that after an accident, Galp Energia immediately implemented legitimacy strategies to minimise the negative impacts on their reputation. Our findings also shown that those years classified as neutral self-proclamation strategies were implemented to increase, or at least maintain, the company's legitimacy level. We have also found evidence that after an accident Galp Energia uses strategies to repair/recover reputation.
As previously mentioned, the concerns of companies about environmental and social issues derive from the desire to have a good social image. The legitimacy theory is indeed an explanatory basis for these concerns. However, in future studies, it would be interesting to analyse other environmentally sensitive companies of Portuguese origin in future studies, as well as other means of communication and other time frames so that we can generalize the conclusions and make comparisons. Aside from this, it would also be of interest to address other theories, such as the stakeholder theory, in order to understand if the concerns of companies are equally supported by these theories.

Leixões Terminal
"In view of the event, the Executive Committee decided to involve the whole Company in a very ambitious security plan in order to make us a European reference in the oil and gas industry." (Galp Energia, 2004: 10) Assumption -Minimising the impact -Corrective actions "The increase in the Iberian distribution market (+ 7%) was not sufficient to sustain a break of 9% in exports and contracts for these products, caused by the accident at the Leixões Terminal, which affected sales to the foreign market." (Galp Energia, 2004: 35) Simple Assumption, through the recognition of negative results.

2008
Fatal accident in the maintenance of Sines refinery "During the the maintenance works at the Sines refinery, an employee of a service provider had a fatal accident. An investigation was immediately initiated in order to identify the causes of the accident. This research led to the identification of a set of corrective measures that are currently being implemented. " (Galp Energia, 2008: 104) Assumption -Minimising the impact -Corrective actions 2009 2 Accidents with fatalities and 1 with multiple casualties; 1 Accident with loss of product; "... three serious accidents with personal injury involving third parties, two fatal accidents and one with multiple casualties. All accidents were properly investigated and occurred under conditions not controlled by Galp. " (Galp Energia, 2009: 123) Simple Denial "In 2009 there was a material accident with loss of product. This accident was investigated internally and was communicated and monitored by the competent authorities. The causes found led to recommendations and an action plan that the company promptly implemented to prevent its recurrence. " (Galp Energia, 2009: 123) Assumption -Minimising the impact -Corrective actions "We also highlight the accident at the Sines refinery, after which there was a fire with significant damage in the central utilities. This accident did not cause personal or environmental damage, but forced the emergency stop of the system having some production units restarted work after three months, resulting in economic losses for the company. " (Galp Energia, 2009: 123) Simple Assumption 2010 3 Serious accidents involving personal injury; "There were, however, three serious accidents with personal injuries involving third parties. All accidents were properly investigated and occurred under conditions not controlled by Galp Energia. "(Galp Energia, 2010: 101) Simple Denial "In 2010, we regret a fatality that occurred during maintenance work in the Sines refinery. As for the remaining accidents, the investigation and the subsequent identification of the causes at the root of this accident led to the formulation of a set of recommendations and an action plan in order to prevent its recurrence. "(Galp Energia, 2010: 101) "In 2011, there is, however, to regret the occurrence of four serious accidents, which were duly investigated. Of these, three involved products sold by Galp Energia, but in unrelated conditions to its activity, and in conditions that cannot be controlled by the Company, namely a road accident, which involved a transportation vehicle of hazardous materials, and two accidents in client homes with GPL. " (Galp Energia, 2011: 69) Simple Denial "There was a serious accident, namely a fire on the premises of Galp Energia, which resulted in material damage, with no personal injuries or environmental losses. As for the remaining accidents, the investigation and subsequent identification of the causes of this accident led to the identification of a set of recommendations and a plan of action to prevent its recurrence. " (Galp Energia, 2011: 69) Assumption -Minimising the impact -Corrective actions