QUALITY OF SUSTAINABILITY DISCLOSURE AMONG THE ASEAN-5 COUNTRIES AND THE ROLE OF STAKEHOLDERS

This study was carried out to reveal the actual quality of sustainability disclosure, actuated by some recent studies that pointed out the lack of disclosure quality amid the growing trend of CSR and the tendency that CSR disclosure was dominantly constructed to manage the corporate image. This study also seeks to investigate the role of stakeholder groups (primary, secondary and regulatory stakeholders), by analyzing the sustainability disclosures of 224 primary sector companies among the five emerging markets in Southeast Asia: Indonesia, Malaysia, Singapore, Thailand, and Philippines in the year of 2016. The extensive disclosure index was also employed to assess each disclosure item under the GRI G4 Guidelines. The results revealed that the quality of sustainability disclosure is still low. In this regard, Thailand turns into the country with the highest score, followed by Malaysia and Indonesia. Labor practice aspect becomes the most expressed aspect by the companies, followed by environmental and social aspect. Based on results obtained, it is therefore argued that employees, auditors, mass media, and regulators have roles in encouraging companies to enhance the quality of sustainability disclosure. However, this study does not find a significant influence from the shareholders and international consumers. The contradictory result was found from creditors, conveying that they possess a negative influence on the quality of sustainability disclosure.


INTRODUCTION
Two decades have been passed since the concept of Triple Bottom Line was firstly mentioned by Elkington in 1997. The fundamental concept of TBL, also known as Triple P (Profit, People, Planet), emphasizes that the company should equally provide attention to its performance, both in economic, social and environment (Elkington 1997). Illustrated as three circles which have interconnection in each other, the Triple P addresses that the social and environmental responsibility of the company is as important as its profitability. To achieve success, a company is not only required to maintain its profitability but also to keep its business in a sustainable way (Perrini and Tencati 2006).
A company, as a nexus of contracts, receives considerable interest in a wide range of stakeholders (Jensen and Meckling 1976), such as shareholders, customers, employees, government, local community, and environment. In relation to this obvious issue, Clarkson (1995) who conducted a review of 10-years research related to CSR, prefers to view CSR as a concept of stakeholder management, by fulfilling the responsibility of the company to its stakeholders rather than its responsibility to the society only.
The capacity of company to produce long-term sustainable prosperity is determined from its management relation with critical stakeholders (Post et al. 2002;Perrini and Tencati 2006). Therefore, several previous studies have used stakeholder framework as a basis for companies in conducting and reporting their CSR activities, such as the study of Huang and Kung (2010), Dong et al. (2014), Chiu and Wang (2015), and Siregar and Rudyanto (2016). Those studies discovered that several stakeholders, such as employees (Huang and Kung 2010;Siregar and Rudyanto 2016), government (Lu and Abeysekara 2014), and consumers (Dong et al. 2014), have assigned influences in encouraging quality social and environmental disclosures.
Data from various CSR studies and surveys indicated that the development of sustainability reporting trend occurs globally, both in developed and developing countries (ASR 2010;Sharma 2013;KPMG 2015;Loh et al. 2016). Based on the report of Carrot and Stick: 2016 Edition, which was published by KPMG, GRI, UNEP, and the Center for Corporate Governance in Africa, the main factors causing such escalation is the existence of sustainability reporting regulations by the government. This is an encouraging fact that implies the increasing awareness about sustainability issue. Yet the remarkable question arises concerning the quality of such obvious disclosures. As well, De Villiers and Alexander (2014) convey that sometimes the company is just looking to "tick more the GRI boxes" by increasing the number of their reporting, regardless of the content and quality of information (De Villiers and Alexander 2014; Michelon et al. 2015). Some recent studies from Cho et al. (2012), Chiu and Wang (2014), Michelon et al. (2015) also pointed out a lack of significant association between CSR reporting practices and disclosure quality, as evidence a tendency that CSR reporting was used to manage the corporate image rather than to impact sustainable development (Gray 2010;Michelon et al. 2015) Therefore, the crucial matter is not only about the "quantity" but also the "quality" of the disclosure.
This study identified three research gaps in previous studies and tried to fill these gaps. Firstly, most of the previous studies only captured the quantity matter in sustainability reporting. Majority of previous studies using the traditional measure for content analysis, such as look at the percentage allocation of CSR disclosure in the annual report or using dichotomous scoring (score 1 for items that are reported, 0 if not reported) (Michelon et al. 2015). This measurement sometimes caused a less accurate result, because a company that making one sentence disclosure is treated as equal as a company which discloses 50 sentences (Hackston and Milne 1996;Dong et al. 2014). Therefore, this study aims to conduct a more detailed examination to see the actual quality of the disclosure of sustainability, by adapting disclosure scoring index used by Dong et al. (2014) and Vuontisjärvi (2006). Each of disclosure made by the company will be assessed from its accuracy (0-3 score range) and its substance (0-3 score range) for general aspects and its specificity level of performance indicator (0-6 score range) for specific/ performance aspects.
Second, most previous studies have examined the influence of each stakeholder group on the quality of CSR/sustainability report. Thus, this study uses a stakeholder framework to evaluate the role of stakeholder groups in encouraging companies to conduct quality reporting. This study employs stakeholder group classification of Buysse and Verbeke (2003), by dividing stakeholders into four groups: internal primary stakeholders, external primary stakeholders, secondary stakeholders, and regulatory stakeholders. Such classification is expected to facilitate us to draw a conclusion of varied results from previous studies because each group of stakeholders gives a different motivation towards companies to conduct quality disclosures.
Third, from the literature review conducted, majority of previous studies that examined determinants of CSR disclosure only evaluate one or two aspects of the CSR reporting, such as social disclosure only (Branco and Rodrigues 2008;Menassa 2010;Chiu and Wang 2015), environmental aspects only (Clarkson et al. 2008;Huang and Kung 2010;Thijssens et al. 2015), or social and environmental aspects (Tagesson et al. 2009;Gamerschlag et al. 2011;Lu and Abeysekara 2014). Therefore, this study attempts to provide a comprehensive picture of the quality of sustainability disclosure, by assessing the three dimensions in sustainability report under the GRI G4: economic, social, and environmental aspects.
This study is expected to contribute to the development of related research in de-veloping countries. Most of the prior researches has been conducted in developed countries, whereas there are different institutional backgrounds in developed and developing countries (Kuzey and Uyar 2017). A literature review was conducted by Ali et al. (2017) to compare the determinants of CSR disclosure in developed and developing countries. The concern of specific stake-holders in developed countries, for example, regulators (Chih et al. 2010;Shi et al. 2012), shareholders (Thorne et al. 2014), creditors (Roberts 1992), and the media (Deegan et al. 2002) possess important factor influencing CSR disclosure. Meanwhile, in developing countries, CSR reporting is influenced by the external forces or powerful stakeholders, for example, international buyers (Belal and Owen 2007), foreign investors (Chiu and Wang 2014), international media concerns (Islam and Deegan 2008), and international regulatory bodies such as the World Bank (Rahaman et al. 2004).
As well, Ali et al. (2017) emphasized that most of the CSR-research in developing countries is only conducted in one country (single-country case studies), thus the results cannot be generalized and depending on the contextual factors of the country. Therefore, this study acquires a wider sample, which consists of primary sector industries among five emerging markets in the ASEAN: Indonesia, Malaysia, Singapore, Thailand, and the Philippines (ASEAN 2015; Majid et al. 2008). The chosen ASEAN-5 countries as the object of study are also considering the Sustainable Development Goals (SDGs) as a new global set of development goals in the South East Asia region . This study is expected to give a contribution to strengthening the SDGs framework for regional integration by enriching SDGs-related regional research and intensifying knowledge-sharing between decision-makers and the research community (Olsen et al. 2015).
In addition, a study from Loh et al. (2016) which aimed at reviewing the progress of sustainability reporting in the year of 2014-2015 in Indonesia, Malaysia, Singapore, and Thailand, found that the overall progress of sustainability reporting is good enough, but additional effort is needed to produce a better quality of sustainability reporting. Hence, this study took 2016 as the observation year to extend the study of Loh et al. (2016) by investigating which stakeholder groups that have power to encourage companies to enhance its sustainability reporting quality. The selection of 2016 as observation year is also related to the mandatory regulation of Indonesia companies to report their CSR as regulated in the Financial Services Authority of Indonesia (OJK), effective as of year-end 2016, and the publication of Sustainability Reporting Guide by Malaysian Stock Exchange in October 2015.
Based on a literature review by Ali et al. (2017), the industry sector was found to be associated with CSR disclosure (Haniffa and Cooke 2005;Amran and Devi 2008;Huang and Kung 2010). Although it has no consistent effect on the ethical/social disclosure, it has an important effect on the amount of environmental disclosure (Adams et al. 1998), considering that environmental disclosure has greater relevance to natural resources related companies rather than the others. Hence, this study took primary sector companies as the research's sample, given that all three aspects of sustainability disclosure (economic, social, and environmental) are very crucial for this sector.
Through content analysis methods under the GRI G4 guideline, the assessment of sustainability disclosures was conducted upon 224 companies. The findings revealed that the level of quality of sustainability disclosure still tends to be marginally low. Thailand has the highest score, followed by Malaysia and Indonesia. The labor practices disclosures become the most frequent, followed by environmental and social aspects. Based on the regression results, the employee group as internal primary stakeholder, mass media and auditor as secondary stakeholder, and the regulatory stakeholder have a significant engagement in encouraging companies to conduct the quality sustainability disclosure. However, there is no significant influence from the shareholders and international consumers on the quality of sustainability disclosure. The study also found a negative influence of the creditors on the quality of sustainability disclosure.
This paper is organized as follows: Firstly, an introductory section is presented. A review of the literature and the development of the research hypotheses in the second section will be presented. The research methodology used will be discussed in the third section, followed by the results and analysis in the fourth section. Finally, in the fifth section, the conclusion along with the implications, limitations, and suggestions for subsequent research will be described.

Stakeholder Theory
The stakeholder theory is one of the most widely used theories in the CSRstudies (Clarkson 1995;Harrison and Freeman 1999;Sharma and Henriques 2005;Huang and Kung 2010;Dong et al. 2014;Beckman et al. 2016). This theory starts from a change of business paradigm that was originally focused on maximizing the welfare of the stockholder only, become wider and considering many parties related to the company (Clarkson 1995;Andriof et al. 2002) which became known as 'stakeholder' terms.
In this respect, Freeman (1983) explained that there are two definitions of stakeholders, in the narrow sense and the wide sense. Under narrow sense definition, stakeholders are simply defined as individuals or groups to which the organization depends on, to survive in operating the business. On the other hand, in the wide sense definition, stakeholder is defined as groups that might affect the achievement of corporate goals and groups that are affected by the existence of the company. However, in formalizing and managing corporate strategy in the long term, stakeholders must be viewed in a wide sense (Freeman 1983). Clarkson (1995) tried to distinguish stakeholders into two major premises; primary stakeholders and secondary stakeholders. Primary stakeholders are defined as the main stakeholder groups that determine the going concern of the company. Thus, without a good relationship with these groups, the company cannot survive and run the business properly. The secondary stakeholder is groups that can affect or be affected by the company, but not as important as the primary stakeholder in determining the business sustainability of the company. In accordance with the argument of Freeman (1994) in the wide sense definition of stakeholder, management should pay attention both to the primary and secondary stakeholder groups.
As well, some contention has emerged around the development of stakeholder theory. There are two major branches in stakeholder theory; ethical (moral) branch and the managerial (positive) branch (Gray et al. 1996;Deegan 2000;Belal 2008). The ethical branch of stakeholder theory asserts that all stakeholders have certain fair treatment rights that should be protected by the organization . Companies should engage in activities that meet the expectations of all stakeholders (Deegan 2000;Fernando and Lawrence 2014). On the other hand, the managerial branch of stakeholder theory assumes that management takes into account the interests of a limited number of stakeholders, who have significant power to influence the success of the business (Roberts 1992).
A study from Mitchell et al. (1997) also gives critical thought about the theory of stakeholder salience, by explicating how and under what circumstances managers can and should respond to various stakeholder types. Based on the study, the salience of stakeholder is determined by three attributes: (1) power; the ability of those who possess power to bring about the outcomes they desire, (2) legitimacy; a generalized perception that the actions of an entity are appropriate within some socially constructed system of norms, and (3) urgency; the degree to which stakeholder claims call for immediate attention. Thus, stakeholders who have power attribute only do not mean that they will have a high degree of salience. Power needs legitimacy to gain authority, and it gains exercise with urgency. As a note that urgency is not a steady-state attribute but can vary across the relationships of stakeholder and manager or within a single relationship across time (Salancik and Pfeffer 1974, Suchman 1995, Mitchell et al. 1997.

Media Agenda Setting Theory
Media agenda setting theory was first introduced by McCombs and Shaw (1972) who found how political news coverage from the US press determines public opinion in the 1968 US presidential election (Brown and Deegan 1998). They provide evidence that there is a strong influence from the news media on people's consideration of major issues in a presidential campaign (Elijido 2011).
The main idea is the ability of the media to influence the degree to which the public prioritizes certain topics (McCombs and Reynolds 2002;Liao et al. 2018). Media are not seen as mirroring public priorities, rather, they are seen as shaping them (Brown and Deegan 1998). Hence, the more the news media report about particular issues, the more prominence these issues will gain among the general public (McCombs et al. 1997;Pollach 2014).
The role of the media is especially relevant for business and CSR behaviors (Zucker 1978). As the development of the application of media agenda setting theory in communication, political, and business research, various studies also used this theory to explain the corporate behavior regarding its CSR activities. The disclosure of CSR information may be used as a strategy by companies to meet the expectations of society (Gray et al. 1995) and as a mechanism to manage their exposure to media and public pressures (Patten 1991;Garcia-Sanchez et al. 2014). Through the media power in shaping the public opinion, firms could reasonably expect that engaging in CSR activities to be discussed in the media will improve their organization's image (Zyglidopoulos et al. 2012).
Along with the increased media concern of the community's social and environmental issue, many studies found that the firms responded by increasing the extent of their disclosures through either annual reports (Deegan 2000;Patten 2002). Even though the media categorized as secondary stakeholder for a company, but it has a crucial role in determining public opinion about topics for society in general (Ader 1995;Brown and Deegan 1998). Firms with high visibility also receiving high levels of media attention. Thus, engaging in CSR is one of their powerful tools to improve their image in public (Zyglidopoulos et al. 2012).

Sustainability Disclosure from Stakeholder Perspective
In the past two decades, CSR and sustainability reporting has attracted worldwide attention by the increasing number of its adoption and being the subject of substantial academic research (Haniffa and Cooke 2005;Jamali and Mirshak 2007). One of the purposes of sustainability reporting is to communicate a company's efforts and sustainability progress to its stakeholders (Dalal-Clayton and Bass 2002, Lozano and Huisingh 2011). Sustainability disclosures may evaluate the performance of the company based on three different parameters: economic, social and environmental (Dutta et al. 2012). It provides the overall performance of the company to satisfy the needs of various groups of stakeholders, rather than financial statements alone (Siregar and Bachtiar 2010). Different from a financial report which provides quantitative information, sustainability disclosure contains both quantitative and qualitative information related to firm's economic, social, and environmental performance in a balanced manner (KPMG/WIMM 2002), to accommodate the information needs of each stakeholder.
The development of CSR concept cannot be detached from stakeholder theory. Coase (1988) explained that the term 'CSR' should be reserved for the process by which companies identify and voluntarily neutralize the harmful effects their operations have on society (Johnston 2011). However, as the development of sustainability concept from stakeholder perspective, CSR is not only seen as a philanthropic activity, because it tends to be interpreted as reciprocity by the company for its irresponsible activities (Hörisch et al. 2014).
Under the ethical branch of stakeholder theory, the fundamental idea of CSR has been shifted as an obligation for business corporations to work towards meeting the needs of a wider array of stakeholders (Clarkson 1995;Waddock et al. 2002, Jamali andMirshak 2007). Hence, the ethical branch of stakeholder theory broadens the scope of CSR by changing its purpose to create value for all stakeholders (Freeman et al. 2010). Based on the concept of the social contract, as the part of a society with various stakeholder relationship, the company exists at the will of its society to the extent that it continues to maintain society with benefits , as its effort to be socially responsible.
However, from the managerial branch of stakeholder theory that assumes management focuses mainly on managing the relationship between an organization and its critical stakeholders, CSR information is disclosed to comply with the expectations of powerful stakeholders rather than all the stakeholders of the firm (Ali and Rizwan 2013). Furthermore, based on the concept of the salient stakeholder theory, only stakeholder who has power and legitimacy will become the salient stakeholder for the company (Mitchell et al. 1997). In this context, the stakeholder needs to gain authority by its power and legitimacy to encourage the company to put the sustainability issue as a serious matter.
Thus, through the examination on this matter, this study desires to point out the stakeholder groups which are considered to be the salient stakeholder by the companies, so that it can push them to reveal the quality sustainability disclosure as the way to meet the stakeholders' expectation about the sustainability matters.

Hypotheses Development
Several previous studies use stakeholder framework as a basis for evaluating the firm's performance. Most of these studies (Huang and Kung 2010;Lu and Abeysekara 2014;Chiu and Wang 2015) figure out that some stakeholder groups have power to encourage the companies to do the CSR reporting, especially the primary stakeholder groups. However, to understand the role of various groups in CSR reporting, it becomes important for us to not only understand the role of primary groups, but also the role of the other group of stakeholders (Clarkson 1995). Therefore, this study used stakeholder's framework developed by Buysse and Verbeke (2003) by dividing stakeholders into four types: internal primary, external primary, secondary, and regulatory stakeholders.

Internal Primary Stakeholder
In this obvious issues, two groups of stakeholders would simply be tested as internal primary stakeholders for the company: the shareholders and employees. Creating value for stakeholders creates value for shareholders (Freeman 2004), given that the capital owners are one of the salient stakeholders for the company. They obtain power with their capital for business continuity and legitimacy with their ownership in the company. Sustainability reporting mechanism provides information needed by the shareholder groups as the main stakeholder of a company, both quantitative and qualitative information about firm's economic, social, and environmental performance (KPMG/ WIMM 2002). Thus, the degree of shareholder concentration will affect the disclosure policy undertaken by the company (Cormier and Magnan 2015). Evidence from previous studies has suggested that dispersed ownership across many investors has contributed to increasing the need for voluntary disclosure (Cullen and Christopher 2002;Brammer and Pavelin 2005). Distributed company ownership, especially when owned by several investors concerned about social responsibility issue, will put pressure on management to increase the quality of the company's CSR report (Chiu and Wang 2015). As outlined in the stakeholder theory, the wider company's responsibility scope towards the stakeholder groups, the larger its incentive to reveal more information in sustainability reporting, and vice versa (Lu and Abeysekara 2014). Therefore, the hypothesis is proposed as follows. H1a: The diffusion rate of shareholder has a positive effect on the quality of sustainability disclosure.
Employees, as the intellectual capitals and the executors of corporate strategy, are the main and influential stakeholders of the company. As described by the salience stakeholder theory, corporate liability for employees can become a tool to gain power and legitimacy, so they can be considered as potential salient stakeholder (Mitchell et al. 1997). Considering the surge in CSR awareness, Fernandez-Feijoo et al. (2014) found that employees became one of the most influential stakeholders for transparent sustainability reporting. Through the interviews with companies' executives, Jamali and Mirshak (2007) also found that CSR communication with employees is essential and has positive spillover effects on their morale and motivation. Since the rights and interests of the employees are closely connected to the firm prospects, the employees are particularly concerned about the firm's attitudes toward social responsibility. The presence of extensive reporting is also one of the management tools to inform the company's social performance and to maintain an ongoing good relationship with its employees (MMSD 2002;Siregar and Rudyanto 2016). Consequently, the larger the number of employees, the greater the influence they have to demand CSR/sustainability information from the company (Huang and Kung 2010). Therefore, the next hypothesis is stated as follows. H1b: Labor intensity rate has a positive effect on the quality of sustainability disclosure.

External Primary Stakeholder
Not only the internal primary groups, but the external primary groups also become key stakeholders who have a major influence on companies to conduct the quality of sustainability reporting (Huang and Kung 2010;Dong et al. 2014). There are two groups of stakeholders to be tested as external primary stakeholders in this study: international consumers and creditors.
International consumers, as stakeholders in foreign countries, also exert pressure on companies to be socially responsible (Andrew et al. 1989;Haniffa and Cooke 2005). As one of the important stakeholders, consumers have certain expectations for the company regarding its social responsibility (Podnar and Golob 2007). Explained by the managerial branch of stakeholder theory and the salient stakeholder theory, international consumers can be salient stakeholder for the company through their power and legitimacy in the market mechanism with their buying decision and their rating to a product in many aspects, including the assessment of CSR issue. Thus, international buyers can often influence corporate behavior through the exercise of bargaining/buying power in contractual relationships (Chiu and Wang 2015). The consumers that concerns with global CSR issues will make companies operating in international markets, particularly those from developing countries, motivated to conduct quality sustainability disclosures (Branco and Rodrigues 2008) to fulfill the needs of CSR information of their consumers. As well, a study from Belal and Owen (2007) indicated that domestic companies in Bangladesh, particularly those who are competing in export markets, provided considerably particular attention to pressure emanating from international buyers regarding their CSR reporting. Thus, the hypothesis is as follows. H2a: International consumers have a positive effect on the quality of sustainability disclosure.
The creditor as a provider of funds for the business operations is one of the stakeholders that can affect the activity and disclosure of the company (Hossain et al. 1994;Lu and Abeysekara 2014). Applying the concept of salient stakeholder theory, the creditor obtains its power and legitimate claim to the company through its loan agreement with the company. Moreover, Huang and Kung (2010) argued that the creditors of a firm with higher financial leverage become more influential, and managers will step up the response to their demands for corporate social activities information. Consistent with Lu and Abeysekara (2014), this study has also predicted a positive relationship between creditors and quality disclosures, considering the growing concept of sustainable finance and social banking in the business world. The concept suggested that company's social and environmental aspects turned into one of the considerable factors for funding decision, as a commitment to safeguarding the environment and better social life (Weber and Remer 2011;OJK, 2014). It is important to the creditor since they will face default risk if the firm is penalized when generating a negative impact on the social/environment (Huang and Kung 2010). Thus, the next hypothesis is stated as follows.

H2b: Creditors have a positive
influence on the quality of sustainability disclosure.

Secondary Stakeholder
Although the secondary stakeholder group does not exert great influence on the going concern of the company such as the primary group, they also have effects on the CSR reporting. The sustainability of a firm depends on the sustainability of its stakeholder relationships, not only to its shareholders, employees, and customeras primary stakeholders − but also public authorities, civil society in general, etc (Perrini and Tencati 2006). There are two groups of stakeholders to be tested as secondary stakeholders in this study: the mass media and auditor.
As the secondary stakeholder, the mass media play an important role in shaping the firm's reputation in society. The influence of media is explained by the media setting agenda theory. The visibility of a firm depends on the quantity of media coverage on the firm (Gamerschlag et al. 2011). Companies with high visibility will receive greater media attention, so they tend to be careful with their corporate image (Lindgreen et al. 2008). Empirical studies from Chiu and Wang (2015) and Gamerschlag et al. (2011) found that the mass media became one of consideration by firms in conducting disclosure related to social aspects. Through the media power in shaping the public opinion, the companies may be used their CSR disclosure as an effective way to improve their positive image (Zyglidopoulos et al. 2012) and to manage their exposure to media and public pressures (Patten, 1991;Garcia-Sanchez et al., 2014). This is evidence by some studies, such as Deegan (2000), Reverte (2009), and Garcia-Sanchez et al. (2014) who found an increase in the amount of CSR information disclosed in consequence of media pressure. Therefore, the hypothesis is as follows. H3a: Media exposure has a positive influence on the quality of sustainability disclosure.
Auditor, as an independent and professional party, can provide a role in influencing and directing its clients to initiate the evolving accounting practices, such as the concept of CSR (Lu and Abeysekara 2014). Even though it acts as an external-intermediary party that does not have a major influence on the company's operational activities, auditor take its role by providing assurance services to the credibility of financial statements (Huang and Kung 2010). As one of the company's stakeholders, auditor can encourage the company to enact full disclosure about firm performances, both in economic, social, and environmental aspects (Ahmad et al. 2003). A study from Wallace et al. (1994) found that large public accounting firms have more expertise and experience in influencing firms to disclose broader information. Similar research result obtains by Ahmad et al. (2003) who found that Malaysian firms which are audited by Big 5 tend to report higher environmental disclosure. Therefore, the next hypothesis is stated as follows. H3b: Firms that are audited by the Big 4 tend to enhance their quality of sustainability disclosure.

Regulatory Stakeholder
As the regulatory stakeholder, government or the other regulatory bodies related to CSR/sustainability reporting also become the salient stakeholder, by influencing the companies on such reporting practices. Based on the result of KPMG survey, the main factor causing the increasing number of sustainability reporting is related to regulation (KPMG 2015) stipulated by the government or related institutions, such as the stock exchange. As described in the managerial branch of stakeholder theory and the salience stakeholder theory, the government and the other regulatory bodies have power attribute by setting the rules which must be obeyed by the company and legitimacy attribute by setting the norms and standards of sustainability reporting practice and its guidance (Mitchell et al. 1997, Dong andXu 2016). A research conducted by Dam and Scholtens (2008) found that there is a significant relationship between the strength of environmental regulations on CSR activities conducted by multinational companies. A similar study from Dong and Xu (2016) also found that the government, through its CSR regulation, became one of the influential stakeholders in increasing the extent of disclosure of environmental aspects in mining companies in China. Thus, the last hypothesis is as follows.

H4:
The strength of CSR/ sustainability regulation has a positive influence on the quality of sustainability disclosure.

RESEARCH METHODOLOGY
This study uses a quantitative approach with Ordinary Least Square (OLS) regression for testing the hypothesis. Data required in this study were taken from the firm's website, the stock exchange's website, and Thomson Reuters Database. The object was chosen based on primary industry groupings on the Global Industry Classification Standards, provided by Thomson Reuters database. In this method, there were 255 companies listed as primary sector companies in the observation year, 2016. A total of 31 companies were eliminated as research samples, because there are 20 companies which not publish its annual report/sustainability report, 6 companies which publish the report in national language only (not in English), and 5 companies with incomplete data. In the end, 224 companies were selected as the final samples, which consist of mining (111), agricultural (58), oil and gas (40), and husbandry companies (15).
The dependent variable in this study is the quality of the company's sustainability disclosure. This study uses sustainability reporting guidelines published by GRI G4 as a basis for conducting content analysis on each aspect, both disclosed in the annual report or sustainability report. Under the GRI G4 guidelines, there are 149 disclosure items, which consists of 58 general disclosures and 91 specific disclosures consisting of three categories: economics (9 items), social (48 items), and environment (34 items). As for the social category, there are 4 sub-categories that should be reported by the company: labor practice (16 items), human right (12 items), society (11 items), and product responsibility disclosure (9 items).
This study employs a disclosure quality measurement model adapted from Dong et al. (2014) and Vuontisjärvi (2006). Considering the differences in the type of information, general disclosure and specific disclosure will be assessed from a distinct dimension. For general disclosure which contains general company related information, strategies and risks encountered, and corporate governance the disclosure will be assessed from accuracy dimension (score 0-3) and substance dimension (score 0-3). For accuracy dimension assessment, we observed how each item was disclosed, whether it is not disclosed (0), expressed in descriptive/narrative form only (1), expressed in quantitative form (2), or expressed in monetary unit (3). For substance dimension assessment, we observed whether it is not disclosed (0), expressed as a normative value/aims to be achieved only (1), as the action practiced (2), or as quantitative measure reflecting the company's achievement of the practice (3). Therefore, the maximum score for general disclosure is 348 ((58× 3) + (58×3)).
The specific disclosure which contains company's performance on three aspects of sustainability: economics, social and environmental, will be assessed from its performance indicator disclosed on a 0-6 scale (Clarkson et al. 2008). Thus, the maximum score for specific disclosure is 546 (91 × 6). As a result, the maximum score that can be achieved by a company for its general and specific disclosure is 894. The detail scoring index can be seen in Table 1.
There are five independent variables in the research model. The first variable is shareholder's diffusion rate (SOD), measured by the number of share of public ownership, counted by dividing free float (shares outstanding -treasury shares - Not disclosed 0 Disclosures that provide information on stated aims/value 1 Disclosures that covers how the company addresses a given issue by describing the action/practice adopted 2 Disclosures that reflects actual achievements by providing quantitative measures 3

Performance Indicator (PI)
Performance data is not presented 0 Performance data is presented 1 Performance data is presented relative to peers/rivals or industry 2 Performance data is presented relative to previous periods (trend analysis) 3 Performance data is presented both in absolute and normalized form 4 Performance data is presented relative to target 5 Performance data is presented at disaggregate level (i.e., plant, business unit, geographic 6 Source: Dong et al. (2014) and Vuontisjärvi (2006) Source: Adapted from Waverman and Koutroumpis (2011)

Figure 1 4-Quadrant Model of Regulatory Stakeholder shares held by strategic investors) to the
shares held by strategic investors) to the share outstanding (Chiu and Wang 2015). The next variable, employees (EMPL) is proxied by labor intensity ratio, by dividing total labor-related costs to total fixed assets owned by the company (Schmenner 1986). International consumer (CONS) is proxied by the export proportion of the total sales (total export / total sales of the company) and the creditor (CRED) is proxied by the financial leverage ratio of the company, by dividing total long-term debt to total assets. Auditor (AUD) is proxied by dummy variables (Lu and Abeysekara 2014). A value of 1 is for firms audited by Big 4, while a value of 0 for firms audited by non-  (2016), Loh et al. (2016) Big 4. Media exposure (MEDX) is proxied by the number of news related to the firm on Google search engine in the reporting year. This proxy refers to Garcia-Sanchez et al. (2014) to determine the visibility of firms by mass media. The regulatory group (REG) is proxied by analyzing the strength of regulation related to sustainability reporting in each ASEAN-5 countries as for year ended 2016. This study adopts the 4quadrant model formulated by Waverman and Koutroumpis (2011). There are two aspects which are considered in determining the strength of the regulation: the sustainability reporting enforcement level and the presence of sustainability reporting guideline provided by the stock exchange. The 4-quadrant model is presented in Figure 1.
Five sample countries in this study would be placed according to the 4quadrant model of regulatory stakeholder. Countries in Zone 4 would get score 4, countries in Zone 3 would get score 3, and so on. Information about sustainability reporting regulation in each country is obtained from the United Nations -Sustainable Stock Exchange (SSE) Initiative and study of Loh et al. (2016), who conducted an analysis of sustainability reporting in the ASEAN region. The regulation summary and assessment are presented in Table 2.
Besides the variables, this study also used two control variables, the size of the firm (SIZE) which proxied by the natural logarithm of total sales and firm's profitability (ROA) which proxied by return on assets (Huang and Kung 2010). Both control variables are included to obtain a better research model. We test the research hypotheses by estimating the following regression: The research model refers to Huang and Kung (2010), with several adjustments to the stakeholders' framework of Buysse and Verbeke (2003) used in this study.

The Scoring Result: Quality of Sustainability Disclosure in ASEAN-5 Countries
Before turning to the regression result to analyze the influence of stakeholder groups, here is the scoring result which is expected to reflect the quality of sustainability disclosure in ASEAN-5  Table 3 present the detail of sustainability disclosure assessment conducted in this study. As summarized in Table 3, there is a similar trend among the ASEAN-5 countries. Based on the accuracy dimensions, majority items are descriptively disclosed, given that disclosure in general aspects is dominated by qualitative information, such as organizational profiles, corporate strategy, stakeholder engagement, and information related to corporate governance practices. Whilst, based on the substance dimension, majority items are expressed by describing the action/practice adopted by the company. A small number of disclosures (5-7 items) are disclosed in quantitative measure of sustainability practice. However, many items in general aspect are not disclosed by the company. These results indicate that the level of disclosure in general aspects is still quite low and needs to be improved because this aspect consists of fundamental information related to the company's profile and corporate governance practice. Meanwhile, the disclosure of specific aspects is also low. Of the total 91 items in specific aspects, there are 8-29 items disclosed by companies. The disclosures were also dominated by descriptive disclosure. A small amount of disclosures is presented with trend analysis and presented in detailed level (break down per geographic area, per division, etc). Therefore, from the average score of quality sustainability reporting, we can see that the labor practice aspect has the highest average quality score. The employment issues become one of frequently noticedissues in ASEAN. The attention is also encouraged by the cooperation between ASEAN and ILO (International Labour Organization) in disseminating issues related to labor practices in the ASEAN region (ASEAN Service Employees Trade Union Council 2013). This result also relates to the characteristics of primary sector industry, where labor-related issues and occupational safety rules become the main issues in this sector.
Much attention has been devoted to Thailand, it has the highest average disclosure score, in all aspects of sustainability, followed by Malaysia and Indonesia. Indeed, based on the review of sustainability/CSR reporting regulation in each country, Thailand has strong regulation related to CSR. This is evidenced by a strong commitment from the Thai government and stock exchange authorities in Thailand, by requiring CSR reporting for public listed companies and intensively socializing and training the corporate executives regarding the sustainability reporting guidelines, such as GRI Guidelines (Sharma 2013; The Stock Exchange of Thailand 2017). Graph 1 and 2 illustrate these findings.  QSD = quality of sustainability disclosure; SOD = shareholder diffusion rate; EMPL = labor intensity; CONS = international consumer ratio; CRED = financial leverage ratio; MEDX = number of news related to the firm on Google News; AUD = dummy variable, value of 1 if the firm is audited by Big 4, 0 if others; REG = assessment of sustainability reporting in each country (1-4) ; SIZE = natural logarithm of total sales; ROA = Net Income / Total Assets Table 4 and 5 shows descriptive statistics and Pearson correlation of all research variables. The data presented is the final data with normalization from outliers, by doing the winsorization technique. The data that is normalized with winsorization is not more than 5% of the total sample. Table 4 provides descriptive statistic result for variables used in this study. The sustainability disclosure's quality of the sample companies is quite low, indicated by the low mean value of the QSD variable. The low value of QSD standard deviation means that the companies' quality of sustainability disclosure scores range near to the QSD mean value. While, the mean value of the diffusion rate of share ownership (SOD) implies that on average, the proportion of total public shares of the total sample is 38.9%. The mean value of the labor intensity (EMPL) means that the average ratio of the employee's expenses per unit of fixed assets is 0.114. The high standard deviation value compared to the average indicates that the variation of the labor intensity ratio is quite high among the entire sample.
From the mean value of international consumer (CONS), it is known that the export value made by the sample  00 QSD = quality of sustainability disclosure; SOD = shareholder diffusion rate; EMPL = labor intensity; CONS = international consumer ratio; CRED = financial leverage ratio; MEDX = number of news related to the firm on Google News; AUD = dummy variable, value of 1 if the firm is audited by Big 4, 0 if others; REG = assessment of sustainability reporting in each country (1-4) ; SIZE = natural logarithm of total sales; ROA = Net Income / Total Assets **significant at α = 1% (two tailed test) * significant at α = 5% (two tailed test) companies is 19.9% of its total sales. The creditor pressure variable (CRED) is proxied with the company's financial leverage rate. On average, the company has a longterm debt ratio of 13.5% of its total assets. The mean value of media exposure variable (MEDX) implies that the average number of news related to the company is 41 times per year. The mean score of the auditor variable (AUD) indicates that majority of the sample companies were audited by Big 4. Whilst, as the result of global price declining of oil-gas and mine commodities in 2015 and early 2016 (Stocker et al. 2018), return on assets (ROA) variable has negative mean value, which means that on average, the overall sample suffered losses 1.9% of its total assets.
Based on Pearson correlation presented in Table 5, in general, there is a positive correlation between QSD with all independent variables, except with employee intensity variable (EMPL). There are significant relationships between several independent variables. It can be seen in the interaction of firm size (SIZE) with all other independent variables, except for shareholder ownership diffusion (SOD) variables. However, SIZE has a negative correlation with the level of employee intensity (EMPL). It indicates large companies are capital-intensive companies. In addition, there is also a negative correlation of EMPL to the creditor variable (CRED) and media exposure (MEDX). This indicates that companies with a high level of employee intensity tend to have a lower level of leverage and less attention from media. The MEDX also has a positive correlation with the auditor variable (AUD), which indicates that high visibility companies tend to be audited by the Big-4. The CRED and MEDX also have a positive correlation that indicates companies with high ratio of long-term debt tend to get high attention from media.

Regression Results
Regression test results can be seen in the following Table 6. Based on classical assumption test, there is a heteroscedasticity problem in the research model, but it had been solved by using the robust treatment in the regression model. Therefore, it can be concluded that the research model has met the classical assumption test, both normality, multicollinearity, and heteroscedasticity test.
Hypothesis 1 examines the influence of internal primary stakeholder groups on Becoming one of the largest oil-price shocks in modern history (Stocker et al. 2018), no wonder that it can affect the result, given that most of the samples (49.5%) are oil & gas and mining companies. Under these conditions, investors may be tempted to focus on the company's financial performance to stabilize the company's performance after the decline. While the H1b (EMPL), can be accepted at the level of significance of 10%. The result shows that firms with higher level of employee intensity tend to perform quality sustainability disclosure compared to firms with lower level of employee intensity (capital-intensive company). In line with salience stakeholder theory and the results of previous studies (Huang and Kung 2010;Fernandez-Feijoo et al. 2014), the company discloses more sustainability information as a manifestation of its responsibilities to employees, given that they are one of the salient stakeholders for the company. Thus, hypothesis 1 is partially accepted, only for employee groups (H1a). Furthermore, hypothesis 2 examines the external primary stakeholder groups: international consumer and creditor. The result shows that H2a (CONS) is rejected. Based on the review of the export activities of the sample companies, only 36% of companies doing export trading activities. In addition, the majority of this activity, with an average export value of 85.2%, is only conducted in intra-ASEAN countries and some countries in Asia, such as China, Hongkong, Japan, or South Korea. Only a few companies that export their product to Europe and American countries, which tend to have higher social and environmental awareness than Asian countries in the last 5 years (KPMG 2015). The contrary result appears from H2b. There is a significant negative impact of the creditor group (CRED) on the quality of sustainability disclosure. Although this result contradicts some previous studies, this result are in line with Cormier and Magnan (2003), which suggest that firms with a better financial condition (indicated by low leverage ratio) encouraged to perform social and environmental disclosures than firms with higher debt ratios. Hence, hypothesis 2 is not accepted.
The results showed that the hypotheses which examine the influence of primary stakeholders are rejected, both internal and external primary group, except for H1b. On the contrary, based on the managerial branch of stakeholder theory and the salience stakeholder theory, the primary stakeholder -who have significant influence to the continued viability and success of the business (e.g. shareholder, international consumer, creditor)possess prominent power and legitimacy to influence the companies to meet their expectations, including those related to CSR. Putting aside some possible reasons explained before, these findings show that the primary stakeholders, in the ASEAN context, do not seem to put pressure on management to perform quality disclosures.
For the secondary stakeholder group (hypothesis 3), the result shows that the media exposure (MEDX) significantly impacts the quality of sustainability disclosure, thus H3a is accepted. This result is in line with the media agenda setting theory and some previous studies (Zyglidopoulos et al. 2012;Pollach 2014) who found that firms with high media attention tend to reveal more information about their CSR activities. Through the power in shaping public opinion, mass media can play an important role in encouraging firms to conduct an extensive sustainability disclosure. H3b (AUD) is also accepted. This indicates that firms audited by Big 4 tend to perform quality-sustainability disclosures compared to firms audited by Non-Big 4. Big-4 auditors tend to encourage their client to develop CSR activities and ask them to disclose more information, in order to maintain their reputation and avoid future litigation opportunities (Huang and Kung 2010). Thus, we can conclude that hypothesis 3 is fully acceptable.
Contrary to the result of primary stakeholders, significant positive results were obtained from the secondary stakeholders. These results are quite interesting since the secondary stakeholders can influence companies to enhance their sustainability disclosure instead of the primary stakeholders. Explained by the media agenda setting theory, this result indicates that the mass media can be the salient stakeholder for the company, regarding its important role in shaping the firm's reputation in society. Based on the correlation analysis in Table 5, there are interrelated positive correlations between the media exposure (MEDX), size of the company (SIZE), and auditor (AUD) variable. In addition, there is a significant positive effect of SIZE, as a control variable, on the quality of sustainability disclosure. This result indicates that large companies tend to have high visibility and receive high media attention. Since there are significant positive correlations of MEDX, SIZE, and AUD, it is possible that the H3b (AUD) is accepted because the majority of companies that are audited by Big 4 are the companies with high visibility in the society. Therefore, the influence of secondary stakeholder groups in this study is prominently explained by the media agenda setting theory.
The last hypothesis, H4 (REG) is accepted at the 10% significance level. However, the company's return on assets does not affect the quality of sustainability disclosure. In line with the salience stake-holder theory, this finding (H4) reveals that the regulator possesses power to encourage companies to contrive quality sustainability disclosures. The regulator plays its role as a salient stakeholder who has the authority to coerce companies to enhance their sustainability reporting through its mandatory and supplementary rules. The company certainly seeks to meet these rules to fulfill its responsibility and to avoid penalties thatmay be imposed by regulator. This instrument seems to be quite effective in ASEAN countries context, as exemplified by Thailand. Looking back to the sustainability disclosure scoring result presented in Table 3, the regulatory instruments stipulated by the Thai government have succeeded in making this country achieve the highest score among the ASEAN-5 countries. Besides requiring CSR disclosure as a part of listing rule since 2014, the Thai government also set up the Sustainability Reporting Guidance and its training for corporate executives. Hence, the regulation set by the regulator groups can be an effective way to intensify the companies' concern about sustainability reporting.
From the result and analysis discussed, we find that the managerial branch of stakeholder theory and the salience stakeholder theory is relevant in explaining the sustainability reporting practice in the ASEAN context, proven by the existence of the limited number of salient stakeholder (the employees, mass media, auditor, and regulator) who have significant influence to encourage companies to conduct quality sustainability disclosure. The companies' rationale to conduct sustainability reporting also explained by the media agenda setting theory. It is evidenced by the positive influence of media exposure on the quality of sustainability disclosure.

CONCLUSION
This study was carried out to examine the quality of sustainability disclosures and investigates the influence of stakeholder groups on the quality of sustainability disclosure upon primary sector companies among the ASEAN-5 countries. To investigate the matter, this study divided stakeholder groups into internal primary stakeholders, external primary stakeholders, secondary stakeholders, and regulatory stakeholders. Through content analysis methods under the GRI G4 guideline and the scoring index developed by Dong et al. (2014) and Vuontisjärvi (2006), the assessment of sustainability disclosures was conducted upon 224 sample companies in 2016.
Based on the results, it revealed that the quality of sustainability reporting among the ASEAN-5 countries is still low and needs to be improved. The employees as internal primary stakeholders, mass media and auditors as secondary stakeholders, and regulator have a significant role in encouraging companies to conduct the quality sustainability disclosure. There was no significant pressure from the shareholders and international consumers. This study also figured out a negative influence of the creditor group on the quality of sustainability disclosure.
This study may have important implications for the future development of concepts and practices of sustainability reporting. First, regarding the theoretical implication, the findings support the managerial branch of stakeholder theory, the salience stakeholder theory, and media agenda setting theory in explaining the motivation of companies to conduct the quality sustainability disclosures. These results place the employees, mass media, auditor, and regulator to be the salient stakeholders who have power to encourage companies to conduct quality sustainability disclosure. The media agenda setting theory also support this study, evidenced by the results that companies with high visibility and high media attention tend to be active in social and environmental activities, as an effort to improve their image in the public.
The second is about the practical implications of this study. The role of stakeholders in encouraging companies to enhance the quality of sustainability disclosure is not only possessed by primary stakeholders as the most influential stakeholder, but can be owned by secondary stakeholders, and regulatory stakeholders. Therefore, given that the quality of sustainability disclosure is still low, global awareness and public attention to sustainability issues are highly required to encourage companies to pay more attention to their sustainability business practice.
Third, the findings also demonstrate that regulation related to sustainability reporting and corporate CSR activities play a major role in encouraging companies to conduct quality disclosures. Based on the result, countries with stronger regulation of CSR reporting (such as Thailand) have a higher average quality score. The mandatory regulation is more effective if supported by sustainability reporting guidelines and other facilities initiated by related regulators, such as seminars and training in the preparation of sustainability reports for company executives, as well as giving rewards to companies for their sustainability performance. It seems that these instruments can effectively build a strong commitment from the top management regarding the sustainability issue of the company.
This study has several limitations. Firstly, is about the small number of the samples within this research, which only covers primary sector companies in the ASEAN-5 countries. Each type of industry has unique characteristics and a different emphasis on each aspect of sustainability. Business characteristics difference might cause varying results regarding the quality of each aspect of sustainability disclosures, depend on which aspects that important by each industry. Corcerning the fact that some stakeholder groups may be influential for companies in the primary sector, yet it does not surely affect the service sector companies, and vice versa. The single-year data used in this study may also restrict the generalization of the results. Thus, it is therefore argued that further efforts should be made in order to expand sample coverage, by taking samples from other industrial sectors or expand the observation year.
Secondly, most samples are oil-gas and mining companies. Meanwhile, in 2015 there was a global price decline of oil-gas and mine commodities. It potentially affects the results because in 2016 the financial conditions of these companies were not fully stable. Further study may conduct an event study to compare the quality of sustainability disclosures between normal and abnormal/crises period, to examine the impact of macroeconomics factors on sustainability disclosures.
By way of conclusion, we would simply like to draw attention to the fact concerning the nature of content analysis method to assess the quality of sustainability disclosure. Although this study employs extensive disclosure index that might present a more reliable score of quality of sustainability disclosure rather than dichotomous scoring, it cannot be denied that there is an opportunity for the subjectivity in carrying out the assessment. However, this has been minimized by designing a detailed and comprehensive rule to facilitate consistent assessment and structured working papers that require page references and a brief justification for each disclosure assessment.