The Factors Motivating Voluntary Disclosure of Carbon Information: Evidence Based on Italian Listed Companies

Abstract

The purpose of this research is to examine how environmental committees, institutional shareholdings, and board independence affect managerial carbon disclosure decisions, particularly those of firms belonging to highly polluting industries. We focus on Italian firms that operate in a code law environment but that have the option either to adopt the unitary corporate structure prevalent in common law countries or to retain the dual corporate structure used in code law countries. We use weighted and unweighted carbon disclosure indexes based on the Kyoto Protocol requirements. The findings show that all factors greatly affect voluntary carbon disclosure and that their impact is especially strong for firms in highly polluting industries. This study has important implications for managers and regulators.

Keywords

institutional shareholders’ power environmental committees board independence firm performance polluting sector voluntary disclosure Italy Kyoto Protocol carbon information

Introduction

The signing of the Kyoto Protocol (KP) in 1997 encouraged regulatory bodies in countries across the globe to consider ways and means to control carbon (greenhouse gas: GHG) emissions and develop guidelines and/or regulations for the disclosure of carbon information. Moreover, voluntary initiatives, such as the Carbon Disclosure Project (CDP), have put pressure on firms to promote GHG reporting (Lovell, Bebbington, Larrinaga, & de Aguiar, 2013). To incentivize firms in the European Union (EU) to improve their carbon performance and disclosure, the EU issued directives encouraging firms to set up environmental committees (ECs) on a voluntary basis. These committees could advise top management on how to reduce emissions and develop carbon disclosure policies that would meet the information needs of various stakeholders. Despite pressure from voluntary initiatives and encouragement from regulators, regulations are sparse, and voluntary carbon reporting has not yet received general recognition.

Although there is strong theoretical support for better carbon performance and disclosure by firms based on legitimacy theory (Patten, 2002) and stakeholder theory (Freeman, 1984), the existing empirical evidence shows that firms do not report full and/or reliable carbon information. Most of the studies are global in nature, meaning they are based on firms across different countries, whereas others are based on U.S., Canadian, and Australian firms. They evaluate the roles of different factors that influence managerial behaviour in reporting carbon information (e.g., Berthelot & Robert, 2011; Peters & Romi, 2014; Rankin, Windsor, & Wahyuni, 2011; Stanny, 2013; Wegener, Elayan, Felton, & Li, 2013; Yunus, Elijido-Ten, & Abhayawansa, 2016). Some research studies have also examined carbon performance and, to some extent, disclosures of firms in the United Kingdom, France, and the Netherlands (e.g., Apergis, Eleftheriou, & Payne, 2013; Liesen, Hoepner, Patten, & Figge, 2015; Scholtens & Kleinsmann, 2011). Overall, the evidence indicates that firm size, growth rate, and leverage are the important factors that stimulate managers to provide carbon information voluntarily, particularly managers of firms belonging to the polluting industries (Hahn, Reimsbach, & Schiemann, 2015).

Given the importance of the European capital markets in the global business and financial environment, carbon information provided by European firms, especially those in code law countries, is important for investors to evaluate risk associated with firms’ environmental and carbon performance. EU code law countries that are becoming increasingly market-oriented present an opportunity to examine whether the common law corporate governance mechanism will encourage the management of firms in these countries to report more carbon information.¹

In this study, we focus on Italy, which is a code law country where firms’ corporate governance is typically based on the two-tier corporate structure. The Italian Government has given firms the option to either adopt the unitary corporate structure used in common law countries or to retain the two-tier corporate structure. We focus on Italian firms because Italy has shown a strong commitment to meeting the KP requirements to reduce GHG emissions and has been encouraging firms to disclose carbon information through regulations, such as Decree Law No. 216/2006 and No. 30/3013. Moreover, Italian firms have shown their willingness to mitigate the problems associated with climate change, evidenced by the voluntary establishment of ECs.

Within the framework of the legitimacy and stakeholder theories, we specifically address the following research questions. First, we examine whether the creation of ECs results in increased carbon disclosure. This is especially important to gain a better understanding of the effectiveness of such committees in encouraging managers to be more transparent. Second, we analyse whether board independence, which has a significant influence on managerial behaviour in the Anglo-Saxon corporate structure because of the primary responsibility of boards to protect investor interests (e.g., Wagner, Hespenheide, & Pavlovsky, 2009), is also effective in code law countries moving toward the market-oriented system. Third, in view of the concentration of ownership of Italian firms in the hands of a few families (Jaggi, Allini, Manes Rossi, & Caldarelli, 2016), we examine whether the information needs of institutional investors, who generally have substantial shareholdings, influence carbon disclosures. Fourth, we reexamine whether the firms’ sensitivity to the issue of pollution encourages managers to provide more carbon information. Although firms operating in highly polluting industries are affected by constraints in the form of governmental policies (Chuang, Chen, & Chuang, 2013), the existing findings provide mixed evidence on carbon disclosures by firms in those industries (Hahn et al., 2015). Therefore, a reexamination of this issue will shed additional light on management’s strategic behaviour in disclosing carbon information.

Overall, the main findings show that the establishment of ECs, higher institutional shareholdings, and board independence are important determinants in stimulating voluntary carbon disclosures, particularly when firms belong to highly polluting industries. We conducted several additional tests, which also confirmed our results.

The results of this study contribute significantly to the debate on carbon disclosure in several distinct ways. First, the findings show that the creation of ECs is a step in the right direction to foster more transparency, even in code law countries. Consistent with the premise of legitimacy theory, the establishment of ECs is a good corporate governance mechanism through which firms reveal their commitment to climate change mitigation and maintain legitimacy. This is also consistent with stakeholder theory in that managers respond to the information needs of investors, particularly institutional investors with considerable shareholdings. Second, the study contributes to the ongoing academic debate on the usefulness of comprehensive and transparent reporting, including carbon information. Our results complement the findings of Cotter and Najah (2012), confirming that institutional shareholders’ demand for carbon information is important for their risk evaluation and investment decisions. Third, these findings may have validity for several other EU countries, as all EU countries are affected by the EU recommendations to manage climate change, and some have legal systems and corporate structures similar to those of Italy. Fourth, the results will be useful to policy makers and regulators in developing guidelines to control emissions.

The remainder of the article is organised as follows. The following section discusses the background, including an evaluation of the relevant existing literature on carbon disclosure. Hypotheses are also developed in this section. The research design is described further, and the results are presented following it. The final section consists of the conclusion, including a discussion of the limitations and the future direction of research.

Background

Historically, Italy has been a code law country, and its financial system is similar to that of most other code law countries. Businesses in Italy are organised on a relationship-based model, where ownership is concentrated in the hands of only a few families who also play dominant roles in managing those companies and where investor protection and legal enforcements are weak and private borrowing is strong (e.g., Di Pietra, Grambovas, Raonic, & Riccaboni, 2008; La Porta, Lopez-de-Silanes, Shleifer, & Vishny, 2000). The controlling families dominate the governance mechanism by appointing family members as directors on a regular basis (e.g., Jaggi et al., 2016).

Since 2004, the corporate governance of Italian firms has undergone a change, with the objective of enhancing the effectiveness of corporate governance mechanisms (referring mainly to Law No. 183/2011 and Law No. 208/2015). In accordance with the new rules, corporate boards consisting of seven members or more should have at least two independent members (Art. 147 ter, TUF). As of December 2013, the median board size was 10 members, and approximately 209 of 426 Italian listed companies had at least one third independent members (Assonime Report, 2016, IV Report on Corporate Governance). An interesting aspect of Italian corporate governance is that firms have the option to choose between the Anglo-Saxon unitary corporate governance structure and the German-style dual corporate governance model. Most of the listed firms prefer the two-tier corporate structure under the dual corporate governance model.

In terms of carbon policies, Italy ranks third (after the United Kingdom and Germany) among the countries participating in the EU Emission Trading Scheme (EU ETS) with the largest number of emission-reducing installations (EAA Report, No 10/2013, Trends and projections in Europe). Italy also ranks third with regard to average annual GHG targets relative to yearly emissions in Europe. As a KP ratifying country, Italy has shown a strong commitment to addressing the 2020 Kyoto targets, and climate change mitigation in particular has received renewed emphasis since 2010. With the issuance of Decree No. 216/2006, subsequently replaced with Decree No. 30/3013, the Italian Government has fully implemented various EU directives on GHG that recommend the monitoring and reporting of carbon emissions. It has also established penalties for noncompliance with the KP requirements. Moreover, in June 2013, the Italian Ministry of the Environment signed an agreement to encourage actions that promote sustainable growth and voluntary environmental and carbon reporting by firms. Recently, corrective and supplementary measures have been issued via Decree No. 111/2015, which established higher penalties for firms’ noncompliance.

Overall, carbon disclosure is becoming a strategic priority for Italian listed companies and a common agenda item for corporate board meetings. It is attracting greater attention from various stakeholders, particularly investors. A recent report by PricewaterhouseCoopers (2012) found there has been an increase in the number of Italian listed companies voluntarily providing information about their strategies on climate performance, environmental governance mechanisms, and corporate actions to promote energy efficiency and emission mitigation. In summary, Italian listed firms are recognising their environmental responsibilities and are exhibiting willingness to disclose carbon information.

Literature Review

We present a literature review of both theoretical rationale and existing empirical evidence for voluntarily disclosure of carbon information. Several studies support the argument advanced by legitimacy theory that firms disclose environmental information, including carbon information, on a voluntary basis to convey a message to the market and to society in general that they are taking their responsibilities seriously (Brammer & Pavelin, 2008; Freedman & Jaggi, 2005, 2009; González-Benito & González-Benito, 2006).

Some studies have even pointed out that firms’ survival depends on their operating within the bounds of a value system (Gray, Kouhy, & Lavers, 1995) and conforming to the social contract and societal expectations (Patten, 2002). In a similar context, Deegan and Gordon (1996) have argued that voluntary reporting practices play an important role in justifying or legitimising firms’ continued existence in society. However, Dowling and Pfeffer (1975) question whether the practice of disclosing information could legitimize firms’ existence. They argue that firms make social and environmental disclosures voluntarily to achieve convergence between their financial objectives and rules based on the social contract.

In another stream of literature, researchers have explained firms’ voluntary disclosures of social, environmental, and carbon information from the perspective of stakeholder theory (e.g., Freeman, 1984; Roberts, 1992). The proponents of this theory argue that managers use disclosure as a basis for dialogue with various constituencies. In particular, firms have a strong incentive to respond to stakeholder demands for information on their pollution-related activities, which will demonstrate that they are complying with investors’ expectations. It is further argued that firms would especially meet investors’ information needs because of their power in the stock market (Deegan, 2006; Luo & Tang, 2014), which would have an impact on firm value.

Several studies have empirically examined which firm-specific factors would motivate managers to disclose social, environmental, and carbon information voluntarily (e.g., Brammer & Pavelin, 2008; Carroll & Shabana, 2010; Freedman & Jaggi, 2005, 2009; Luo & Tang, 2014; Stechemesser & Guenther, 2012; Tauringana & Chithambo, 2014). Because our focus is on carbon disclosures, we briefly review important studies that have evaluated managerial motivation to disclose carbon information.

Hahn et al. (2015) reviewed the current state of related literature that examines different factors influencing managerial behaviour vis-à-vis carbon disclosures and, to a lesser extent, the effects of such disclosures. Their review shows that most of the research is global in nature, that is, it is based on firms across different countries. Their analysis also includes studies based on firms in the United States, Australia, Canada, China, and three European countries (the United Kingdom, the Netherlands, and Germany). Overall, their findings reveal that the following firm-specific factors have a strong influence on management’s strategic carbon disclosure decisions: firm size, economic performance, and leverage. Freedman and Jaggi (2009), who have examined carbon disclosures by large firms across several countries, found that firms in countries that have signed the KP are associated with more detailed carbon disclosures, indicating that signing the KP plays an important role in stimulating firms’ carbon reporting.

The studies focusing on U.S. firms primarily investigated the frequency and nature of carbon disclosures. Doran and Quinn (2009) stated that 24% of U.S. S&P 500 firms report on climate change, and only 5.6% provide carbon information. Stanny (2013) found that the frequency of carbon disclosures increased from 2006 through 2008 and that firms revealed minimum information to avoid scrutiny. Another interesting motivation for providing as little information as possible is that, consistent with legitimacy theory, firms are interested in conveying to shareholders that they are taking their environmental responsibilities seriously.

As far as European firms are concerned, there are four important studies. First, Scholtens and Kleinsmann (2011) investigated incentives for subcontractors of a transport and logistics company to report on their carbon emissions. The study revealed that subcontractors in the Netherlands are extrinsically motivated to engage in carbon reporting and to evaluate reduction techniques, probably because of regulatory compliance and energy costs. It also found that British subcontractors are more intrinsically motivated to comply with regulations due to environmental awareness, relationships, and reputation building. Second, Dragomir (2012) evaluated sustainability reports of the top five largest European oil and gas companies and assessed their ability to provide environmental (GHG) disclosures. He found that the reports usually contain unexplained figures and are inconsistent in the methodologies used. He therefore concludes that companies failed in delivering high-quality information and in their efforts to mitigate global warming. Third, Apergis et al. (2013) examined the impact of research and development expenditures on emissions prior to and after mandatory adoption of International Financial Reporting Standards (IFRS) in the manufacturing sectors of Germany, France, and the United Kingdom. Their results suggest that in the post–IFRS mandatory adoption year, the research and development expenditures are negatively associated with emissions and that the reduction in GHG emissions is considered to be due to incentives provided by the new accounting disclosure regime of IFRS. Last, Depoers, Jeanjean, and Jérôme (2016) analysed the consistency of the GHG emissions information voluntarily disclosed by French firms through two different communication channels—corporate reports and the CDP—and found that information on GHG amounts was significantly lower in the corporate reports than in the CDP. In the present study, we extend the research on the determinants of the voluntary disclosure of carbon information by Italian firms that operate in a code law environment and that can adopt the unitary corporate structure prevalent in common law countries or retain the two-tiered corporate structure.

Development of Hypotheses

Several studies have supported the argument advanced by legitimacy theory that firms disclose environmental (including carbon) information with the objective of conveying a message to the market and society in general that they are taking their environmental responsibilities seriously (Brammer & Pavelin, 2008; Freedman & Jaggi, 2005, 2009; González-Benito & González-Benito, 2006). The legitimacy argument supporting the existence of firms is further strengthened when firms are perceived to be operating within the bounds of a value system (Gray et al., 1995) and to be conforming to the social contract (Patten, 2002). Moreover, voluntary reporting of environmental and carbon reduction activities to outsiders increases the justification and legitimisation of firms’ continued existence in society (e.g., Deegan & Gordon, 1996).

Another area of research that has also provided a theoretical rationale for the disclosure of environmental activities and carbon information in particular relates to stakeholder theory (e.g., Freeman, 1984; Roberts, 1992). The proponents of this theory argue that managers may use carbon disclosures as a basis for dialogue with various constituencies, with stockholders being the most important (Gray et al., 1995). Firms have a strong incentive to respond to the stockholders’ demand for information, including information on environment-related activities, because of investors’ power in the stock market (Deegan, 2006; Luo & Tang, 2014). Carbon disclosures help investors in assessing the impact of GHG emissions on firms’ investments, cost efficiency, and operating and economic performance (Solomon, Solomon, Norton, & Joseph, 2011).

Environmental Committees and Disclosure of Carbon Information

Responding to global environmental concerns, the EU recommended the voluntary establishment of ECs in EU countries. This recommendation is apparently based on the premise that a dedicated committee will be in a better position to address environmental issues from the perspective of opportunities and commitments to stakeholders (Elijido-Ten, 2007; Liao, Luo, & Tang, 2015). It also draws support from stakeholder theory because it would enable firms to meet their obligation of providing reliable environmental information, especially carbon information, to stockholders so that they can evaluate their investment risks, opportunities, and decisions. Scholars argue that ECs would serve as a proxy for firms’ concerns regarding environmental responsibility and indicate managers’ strategic posture in terms of providing environmental information to stockholders in response to their demand for such information (e.g., Luo & Tang, 2014). The establishment of ECs is also supported on the grounds that they provide expert opinions to managers to assess the pros and cons of their initiatives to mitigate climate risks, evaluate investments in abating pollution, and making strategic decisions on disclosing environmental information, including carbon information (Dietz, Hope, Sterv, & Zevghelis, 2007).

In fact, ECs would encourage even reluctant managers to recognize their environmental responsibility and persuade them to disclose carbon information to outsiders. ECs may achieve this goal through meetings with management and advising managers on different strategies to help them develop a comprehensive plan for sustainability and transparency. Thus, the establishment of ECs may create an expectation for higher transparency regarding firms’ environmental activities, including carbon activities. This higher expectation would provide an additional incentive to managers to disclose more carbon information to meet increasing information needs. ECs may also help in creating an effective monitoring mechanism to ensure the reliability of reported information. Overall, these arguments suggest that, consistent with stakeholder theory, ECs appear to be an important governance mechanism in advising managers on strategies that will meet investors’ carbon information needs effectively (e.g., Dietz et al., 2007).

The existing evidence on the effectiveness of ECs in ensuring greater carbon disclosure is, however, mixed. While some researchers find no association (i.e., Rankin et al., 2011), others find a positive correlation for U.K. firms (e.g., Liao et al., 2015) and Australian firms (Yunus et al., 2016). In view of strong evidence, based on stakeholder theory that ECs help managers formulate strategic policies on carbon performance and disclosure and are expected to play an important role in monitoring managers’ behaviours in meeting investors’ information needs, we hypothesize that ECs will encourage management to disclose detailed and reliable carbon information. Moreover, the EU recommendation is likely to serve as a strong signal to managers and outsiders that regulators are serious about reducing carbon emissions at the national level. We test the effectiveness of ECs in encouraging managers to disclose more carbon information using the following hypothesis:

Hypothesis 1: Firms that have voluntarily established ECs are associated with greater carbon disclosures compared with firms that have not established such committees.

Institutional Shareholdings and Disclosure of Carbon Information

Both the legitimacy and stakeholder theories suggest that investors are expected to have a significant influence on managerial behaviour and decisions to reduce carbon emissions and to make reliable carbon disclosures. According to legitimacy theory, firms are encouraged to provide carbon information to convey a positive signal to investors that their firm is behaving as a good citizen and is doing its best to maintain and conserve the environment. Consistent with stakeholder theory, managers are motivated to meet investors’ information needs so they can evaluate their risk in terms of investment decisions. In the absence of adequate and reliable carbon information, there will be information asymmetry, which will result in greater uncertainty and will push investors away from investing in the firm. Consequently, managers are under pressure to meet investors’ information needs, including carbon information.

Some researchers have particularly emphasised that institutional shareholders have significantly greater power compared with individual investors because of their higher percentage of ownership (e.g., Cotter & Najah, 2012; Verstegen & Schneider, 2003). Some have argued that institutional investors with significant shareholdings and long-term interests in a firm will have a strong influence on that firm’s information disclosure policies, including carbon information (e.g., Ryan & Schneider, 2002). Smith, Morreale, and Miriani (2008) posited that institutional shareholders may even have a stronger influence than regulatory authorities. However, individual investors with a limited number of shares and/or a short-term horizon are less likely to exert any pressure on management. It is also possible that individual investors may not even be interested in carbon information (Cotter & Najah, 2012).

The above discussion suggests there is strong theoretical support for managers to pay special attention to the information needs of investors, especially institutional shareholders, because they have a strong impact on a firm’s stock price, which in turn affects firm value and managers’ personal wealth (Graham, Harvey, & Rajgopal, 2005). Therefore, managers cannot ignore institutional shareholders’ demand for information, including carbon information. We formulate the following hypothesis to test this:

Hypothesis 2: Firms with higher institutional shareholdings are associated with greater carbon disclosures.

Board Independence and Disclosure of Carbon Information

Corporate board independence is an important characteristic in common law countries to ensure that managers do not ignore investors’ interests, and it also guarantees effective monitoring of managerial decisions and activities to achieve greater transparency through comprehensive and reliable disclosure (Ghomi & Leung, 2013; Wang & Dewhirst, 1992). From the stakeholder theory perspective, independent outside directors have no or limited financial stake in the company, and are expected to perform the key function of ensuring managers act in the best interest of stakeholders (Haniffa & Cooke, 2005; Michelon & Parbonetti, 2012). Johnson and Greening (1999) have argued that independent directors are inclined to focus on long-term interests and may support the development of plans based on activities ensuring sustainable growth in the firm. The existing empirical evidence reveals that effective monitoring by corporate boards enhances the quality of reported information (Wagner et al., 2009). All this suggests that independent directors, who are more prone to higher transparency, would encourage firms to disseminate more information and ensure that managerial decisions are consistent with legitimacy theory (Liao et al., 2015).

Although independent directors’ roles in protecting investors’ interests, providing guidance to management, and monitoring managerial activities to ensure greater transparency are generally observed in common law countries, we conjecture that the changing business environments in code law countries will play a similar role. However, the existing evidence based on different countries, including common law countries, is mixed. Some researchers have found that board structure (including board composition) has a positive influence on social and environmental disclosures (e.g., Amran, Periasamy, & Zulkafli, 2014; Yunus et al., 2016), whereas others have found no evidence in this regard (e.g., Liao et al., 2015; Michelon & Parbonetti, 2012).

Consistent with stakeholder theory, we expect that a higher percentage of independent directors on the corporate boards of Italian companies would encourage managers to recognize the importance of investors’ carbon information needs and thus encourage the boards to report more carbon information. We test this expectation using the following hypothesis:

Hypothesis 3: Firms with a higher proportion of independent directors on their corporate boards are associated with greater carbon disclosures compared with firms with a lower proportion of independent directors.

Highly Polluting Industries and Disclosure of Carbon Information

Within the framework of stakeholder theory, regulatory agencies can be considered powerful stakeholders because they have the right to issue regulations to protect societal interests and enforce societal norms (Groening & Kanuri, 2013). Watts and Zimmerman (1978) posited that firms are willing to adopt socially responsible strategies to reduce risks related to governmental intrusions that may reduce firm value. The extent to which regulators can represent themselves as effective stakeholders depends on the nature of the industry sector to which a firm belongs. Firms in highly polluting industries are more likely to be concerned about rigorous regulations because regulators may consider issuing stringent mandatory rules to ensure proper managerial behaviour vis-à-vis the environment and greater disclosure of environmental information, including carbon information. Therefore, firms operating in industries that are more sensitive to carbon emissions are more likely to adopt transparent disclosure policies as this may reduce the potential of greater scrutiny by regulatory agencies and may reduce governmental sanctions (Deegan & Gordon, 1996; Elijido-Ten, 2007). In the EU market, the EU ETS program defines these industries and they include firms involved in the production of iron, aluminium, metals, cement, lime, glass, and ceramics; firms involved in power generation; energy-intensive firms; oil refineries; steel works; pulp and paper firms; firms producing organic chemicals; and civil aviation firms. Firms involved in other industries are considered as nonpolluting (or less polluting).

As a member of the EU ETS program, Italy regulates firms’ emission policies. Firms belonging to polluting industries are required to report their GHG emissions each year to the National Authority and to the EU Commission. Given the fear that regulatory authorities may impose mandatory rules, we argue that firms belonging to the sectors covered by the EU ETS voluntarily disclose more carbon information compared with firms outside those sectors (e.g., Rankin et al., 2011).

The existing research findings confirm that firms in highly polluting industries are inclined to provide greater environmental information to reduce the potential for sanctions and to respond to investors’ information needs (e.g., Jaffar, Iskandar, & Muhamad, 2002). In this study, we test whether Italian firms belonging to the EU ETS industries disclose more carbon information compared with firms in nonpolluting industries. We use the following hypothesis to test this:

Hypothesis 4: Firms belonging to highly polluting industries, as defined by EU ETS, are associated with greater carbon disclosures compared with firms that do not belong to these industries.

Research Design

Sample

We conducted a longitudinal analysis on a sample of firms for the period from 2010 to 2013 (financial data beyond 2013 were not available when we conducted this study). We based the selection of this specific period on the development of national policies focused on reducing GHG emissions and promoting energy efficiency based on the KP requirements (including tax incentives and trading mechanisms). The study used annual sustainability reports to obtain carbon disclosures. Looking at the largest 100 companies in the EU, in 2011, Italy ranked second in producing the highest percentage of sustainability reports (64%) after Spain and Denmark (which were both at 65%). Italy also ranked second in issuing formal assurance statements (64%; Bebbington, Unerman, & O’Dwyer, 2014). Based on the assumption that the information contained in sustainability reports is more reliable than that in CDP reports, we decided to use the carbon data provided in sustainability reports. Moreover, the number of Italian firms preparing sustainability reports is much higher compared with the number of firms that filed reports with the CDP. In particular, 24 listed firms provided information to the CDP in 2013, 20 provided information to the CDP in 2012, 18 provided information to the CDP in 2011, and 17 provided information to the CDP in 2010 (CDP, 2013).

Because financial institutions are governed by special regulations, we dropped financial firms from the analysis and focussed on nonfinancial Italian firms listed on the Italian Stock Exchange. The total number of firm-year observations over the 4-year period included in the original sample was 780. We dropped 109 firm-year observations from the sample due to the unavailability of data or because of extreme values; therefore, the final sample consisted of 671 firm-year observations.

Information on panel data is provided in Panel A of Table 1. Panel B shows the number of firms belonging to sectors covered by the EU ETS regulations (polluting industries). Panel C provides the number and percentage of firms from different sectors. The highest percentage of firms belong to the consumer and industrial sectors (38.60% and 28.91%, respectively), while the lowest percentage of firms belong to the health care (4.47%) and energy (5.36%) sectors.

Table 1.

Sample Selection.

Panel A: Sample size
Initial sample observations						780
No. of firm-year observations excluded						(109)
Total sample firm-year observations						671
Year firm distribution
Year 2010						169
Year 2011						164
Year 2012						171
Year 2013						167
Panel B: Subsample of firms under EU ETS
Firms not covered by EU ETS						342
Firms covered by EU ETS						329
Panel C: No. of firms by industry sector (source: GIC code classification)
Sectors	2010	2011	2012	2013	Total	% Total
Energy	9	9	10	8	36	5.36
Materials	12	12	13	13	50	7.45
Industrial	50	49	48	47	194	28.91
Consumer	67	65	64	63	259	38.60
Health care	7	7	8	8	30	4.47
IT and utilities	24	22	28	28	102	15.21
Total Italian firms	169	164	171	167	671	100

Note. EU ETS = EU Emissions Trading Scheme.

Measurement of Dependent and Independent Variables

Dependent Variable

Prior carbon disclosure studies adopted a variety of approaches to develop a disclosure index. Some studies used a survey (e.g., Stanny, 2013), while others built the disclosure indexes based on ISO 14061 (i.e., Rankin et al., 2011) or on the CDP (e.g., Depoers et al., 2016; Peters & Romi, 2014). Because only a very small number of Italian firms have been reporting information to the CDP, an index created on CDP data would not be reliable. Moreover, doubts have been raised with regard to the quality of information firms file with the CDP (e.g., Bebbington et al., 2014; Benn, Dunphy, & Griffiths, 2006). Therefore, we developed an index based on items associated with the KP requirements and assume it offers more relevant information regarding the behaviour of Italian firms in terms of reporting climate change mitigation.

We carefully examined the sustainability reports to see if firms mention carbon information in any form (e.g., Yunus et al., 2016). Consistent with the existing studies (Amran et al., 2014; Freedman & Jaggi, 2005, 2011), we used content analysis to capture KP-related information on the following five items:

Mention of global warming or of the KP

A firm’s plans to deal with global warming and the objective to control global warming

Information on potential costs to achieve the global-warming objectives

Information on current costs to reduce GHG emissions

Information on the amount of GHG emissions

We created two disclosure indexes, the Equally Weighted Disclosure Index (EWDI) and the Unequally Weighted Disclosure Index (UWDI). In the EWDI, we assigned equal weight to each item (i.e., one point). The maximum value is five and the minimum value is zero. In the UWDI, we weighted each item on the assumption that the information conveyed by these items differs in importance. There is a compelling argument for adopting differential weights for individual items because their importance differs in developing this index. To be consistent with the existing literature (Freedman & Jaggi, 2005), we used the following scheme:

Item	Weight
Mention global warming	1
Firm’s plans	2
Potential costs	3
Current costs	3
Amount of emissions	3

The maximum score on the UWDI is 12. It is possible that the weights used in the scheme may not truly reflect the relevance of the items in the index. However, a relative weighting system enables us to differentiate the importance of each item in relation to the others. To overcome certain deficiencies in these indexes, we conducted regressions on both. If the results for both were the same, we assumed that they were robust and at least reflected the direction of the association between the dependent variables and their determinants.

Independent Variables

We selected four test variables to address our hypotheses. A dummy variable was created for the presence of ECs; the variable equals one if the firm has established an EC and zero otherwise (Elijido-Ten, 2007; Liao et al., 2015; Rankin et al., 2011). Institutional shareholdings (INST) reflect the percentage of stocks held by institutions in relation to total stocks (e.g., Smith et al., 2008). Board independence (B_IND) is proxied by the ratio of independent directors in relation to the total number of directors on the board (e.g., Liao et al., 2015; Prado-Lorenzo & Garcia-Sanchez, 2010; Yunus et al., 2016). The dichotomous variable (EU ETS) was given the value of one when firms belong to the highly polluting industries according to the EU_ETS program; otherwise their value was zero (e.g., Rankin et al., 2011).

Control Variables

Control variables were included based on the existing carbon disclosure studies (e.g., Freedman & Jaggi, 2011; Liao et al., 2015; Rankin et al., 2011; Yunus et al., 2016). Scholars have argued that creditors play an important role in financing Italian firms. Because the Italian financial market is less developed than the Anglo-Saxon markets, Italian firms (and firms in other code law countries) depend highly on the credit-oriented financing system. The average of the leverage ratio for the listed Italian companies is reported to be approximately 47% of the bank debt (Jaggi et al., 2016). Based on the arguments presented in the literature, we expected that creditors may also be interested in carbon information to evaluate their risk associated with a firm’s economic performance (Clarkson, Li, & Richardson, 2004). We include the debt-equity ratio (DE) for leverage. Porter and Van der Linde (1995) have posited that firms doing a good job environmentally are likely to perform better economically, suggesting that such firms are inclined to disclose carbon information voluntarily. To alleviate investors’ risk concerns, the firms with good economic performance are expected to report more carbon information. We include both return on assets (ROA) and market value on book value (MV/BV) as proxies for economic performance.

Firm size is another important control variable because larger firms are expected to provide voluntary disclosures as they undertake activities that may affect the environment and attract greater scrutiny (e.g., Prado-Lorenzo, Rodríguez, Gallego, & García, 2009; Stanny & Ely, 2008). Larger firms also have a higher capability of absorbing environmental costs associated with reducing GHG emissions, whereas smaller firms may not be able to bear such costs. Furthermore, larger firms have the resources and are financially strong enough to disclose carbon information compared with smaller ones (e.g., Stanny & Ely, 2008). We use natural logarithm of total assets (L_ASS) as a proxy for size.

We obtained financial data from the Compustat and Datastream databases. If financial information for any firm was not available, we obtained it from the company’s website.

Regression Model

The following ordinary least square (OLS) regression model based on the pooled time series and cross-sectional data are used for the 2010-2013 period, and unbalanced panel data are adopted to test our hypotheses:

\begin{array}{l} {EWDI / UWDI}_{jt} = α + β_{1} {EC}_{jt} + β_{2} {INST}_{jt} + β_{3}_B_{IND}_{jt} + β_{4} {EU ETS}_{jt} + β_{5} {DE}_{jt} + \\ β_{6} {ROA}_{jt} + β_{7} {MV / BV}_{jt} + β_{8} {L_ASS}_{jt} + ε, \end{array}

where EWDI = equally weighted disclosure index, UWDI = unequally weighted disclosure index, EC = environmental committee, INST = institutional shareholdings, B_IND = board independence, EU_ETS = highly polluting industries, DE = debt-equity ratio, ROA = return on assets, MV/BV = market value on book value, L_ASS = log of total assets.

We also examine whether the combined effect of ECs and larger institutional shareholdings would result in comparatively higher carbon disclosures. The objective of this test is to evaluate whether there is any incremental effect of ECs on carbon disclosures influenced by INST. The incremental effect is reflected by an interaction variable between EC and INST in Equation (2). Similarly, we test whether ECs would have a stronger role in disclosing carbon information when firms belong to the highly polluting industries. To evaluate the joint effect, we include an interaction variable between EC and EU ETS in Equation (2):

\begin{array}{l} {EWDI}_{jt} / {UWDI}_{jt} = α + β_{1} {EC}_{jt} + β_{2} {INST}_{jt} + β_{3} {B_IND}_{jt} + β_{4} {EU_ETS}_{jt} + β_{5} {DE}_{jt} + \\ β_{6} {ROA}_{jt} + β_{7} {MV / BV}_{jt} + β_{8} {L_ASS}_{jt} + β_{8 - 10} ({EC}_{jt} {x Factor}_{it}) + ε, \end{array}

where factor value = INST or EU ETS. Other variables have been discussed earlier.

Results

Descriptive Statistics

Descriptive statistics for the total sample are provided in Panel A of Table 2. The mean of EWDI is 2.05 (minimum and maximum 0 and 5, respectively), whereas the mean of UWDI is 3.55 (minimum and maximum 0 and 12, respectively). The mean of EC is 0.39, revealing that about 40% of the sample firms have established ECs. The mean of INST is almost 14%, with a range from 0% to about 99%, suggesting that shares held by institutional investors vary significantly among sample firms. The average ratio of independent directors to the total number of directors on the boards is 0.37, which means that more than one third of all directors are independent. The mean of the indicator variable of firms belonging to polluting industries (EU_ETS) is 0.49, indicating that slightly less than 50% of the sample belongs to the pollution-sensitive sectors. The variable of DE has a mean of 3.28. The mean of ROA for the sample is 2.10. The mean of MV/BV is 2.53, and the range varies between −21.31 and 24.35. The mean of L_ASS is 7.15, and the minimum and maximum are 6.06 and 8.95, respectively. The Pearson correlation is shown in Panel B of Table 2. The results show that INST, EC, EU_ETS, B_IND, and L_ASS are positively associated with carbon disclosure indexes and the coefficients are statistically significant, while there is no significant association of DE, ROA, and MV/BV with carbon disclosures indexes.

Table 2.

Descriptive Statistics.

Panel A: Total sample (N = 671)
			Min		Max		M		SD
EWDI			0		5		2.05		1.62
UWDI			0		12		3.55		1.10
EC							0.39		0.49
INST			0		0.99		0.14		0.22
B_IND			0		0.90		0.37		0.22
EU ETS							0.49		0.50
DE			−19.48		21.74		3.28		1.16
ROA			−22.92		23.18		2.10		1.67
MV/BV			−21.31		24.35		2.53		1.37
L_ASS			6.06		8.95		7.15		0.83
Panel B: Pearson correlation for total sample (N = 671)
	EWDI	UWDI	INST	EC	EU_ETS	B_IND	DE	ROA	MV/BV	L_ASS
EWDI	1
UWDI	0.98***	1
INST	0.83***	0.83***	1
EC	0.62***	0.60***	0.49***	1
EU ETS	0.36***	0.38***	0.29***	0.17***	1
B_IND	0.23***	0.20***	0.24***	0.12***	0.10**	1
DE	−0.02	−0.01	−0.04	0.00	0.07**	−0.07*	1
ROA	0.02	0.03	0.07**	0.01	−0.05	0.08**	−0.03	1
MV/BV	−0.00	−0.01	0.07*	−0.01	−0.08**	0.06	0.03	0.65***	1
L_ASS	0.19***	0.17***	0.16***	0.13***	0.05	0.05	−0.04	−0.00	−0.04	1

Note. EWDI = Equally Weighted Disclosure Index; UWDI = Unequally Weighted Disclosure Index; EC = environmental committee; INST = institutional shareholdings; B_IND = board independence; EU ETS = highly polluting industries; DE = debt-equity ratio; ROA = return on assets; MV/BV = market value/book value; L_ASS = log of assets.

p < .10. **p < .05. ***p < .01.

Regression Results

The models in Equations (1) and (2) are estimated based on panel data and by taking into account the time and industry effects of cross-sectional data. Because our sample contains observations over 4 years, we first tested whether a fixed or random effect model would be more appropriate for our data set. A fixed effects model examines the group differences in intercepts, assuming the same slopes and constant variance across firms. A random effects model estimates variance components for groups or times and errors, assuming the same intercepts and slopes. To evaluate the significance level between the two estimators, we used the Stata system and conducted a Hausman specification test, which supported the validity of fixed-effects regression. We performed a Breusch–Pagan test to find out whether there was a heteroscedasticity problem in our sample data. The (untabulated) results indicated that this problem existed, and we corrected it by employing the fixed-effects model with Driscoll–Kraay robust standard errors.

Regression Results Based on OLS Fixed-Effect Model

First, we used an OLS fixed-effect model, and the results are presented in Table 3. The R² of the models is between 0.51% and 0.75%, and F values show that all models are statistically significant at 1%. The coefficients in Models 1 and 3 show that the variables of ECs (coefficients = 0.79 and 1.28 for EDWI and UWDI, respectively), INST (coefficients = 4.09 and 9.67 for EWDI and UWDI, respectively), and EU_ETS (0.29 and 0.63 for EWDI and UWDI) are positive and statistically significant at the 1% level. The coefficients for B_IND (0.03 and 0.16 for EWDI and UWDI) are positive but statistically significant at 5% and 10% only for UWDI. These results are consistent with our Hypotheses 1 (EC), 2 (INST), and 4 (EU_ETS). They are in the expected direction for board independence (Hypothesis 3) but do not fully support the hypothesis. Next, we examined whether these results were influenced by the endogeneity problem.

Table 3.

Regression Results Based on OLS Fixed Effects for the Total Sample (N = 671).

Variables	Model 1: EWDI	VIF	Model 2: EWDI	VIF	Model 3: UWDI	VIF	Model 4: UWDI	VIF
Constant	0.25 (5.19)***		0.23 (4.29)***		0.21 (1.89)**		0.19 (1.87)**
EC	0.79 (3.37)***	1.37	0.41 (10.75)***	2.41	1.28 (8.62)***	1.37	0.61 (9.90)***	2.46
INST	4.09 (9.31)***	1.52	4.41 (3.00)***	7.17	9.67 (5.32)***	1.52	8.11 (3.26)***	7.22
B_IND	0.03 (0.87)	1.13	0.02 (0.54)	1.13	0.16 (1.94)*	1.13	0.14 (2.13)**	1.08
EU_ETS	0.29 (10.77)***	1.15	0.06 (3.31)***	1.68	0.63 (9.69)***	1.15	0.13 (2.69)***	1.73
(EC × INST)			0.02 (1.69)*	7.70			0.05 (1.26)	7.75
(EC × EU ETS)			0.57 (10.21)***	3.31			2.04 (8.33)***	3.36
Control variables
DE	0.02 (1.87)*	1.07	0.01 (1.11)	1.07	0.01 (0.63)	1.07	0.00 (0.37)	1.02
ROA	−0.01 (−3.43)***	1.35	−0.01 (−2.29)**	1.35	−0.01 (−0.80)	1.35	−0.00 (−0.27)	1.30
MV/BV	−0.01 (−3.45)***	1.34	−0.00 (−2.18)**	1.34	−0.01 (−4.47)***	1.34	−0.02 (−3.07)***	1.29
L_ASS	0.02 (3.17)***	1.31	0.03 (3.50)***	1.31	0.01 (0.21)	1.31	0.02 (1.59)***	1.26
Adjusted R²	0.51		0.75		0.71		0.57
F test	8.17***		548.97***		546.57***		6.66***
Hausman test χ²	250.64***		276.13***		270.48***		325.97***

Note. OLS = ordinary least squares; VIF = variance inflation factor; EWDI = Equally Weighted Disclosure Index; UWDI = Unequally Weighted Disclosure Index; EC = environmental committee; INST = institutional shareholdings; B_IND = board independence; EU ETS = highly polluting industries; DE = debt-equity ratio; ROA = return on assets; MV/BV = market value/book value; L_ASS = log of assets. t values are given in parentheses; time and industry effects included in the panel analysis.

p < .10. **p < .05. ***p < .01.

Endogeneity Problem and Two-Stage Least Squares (2-SLS) Regression Results

It is possible that ECs, institutional shareholdings, and carbon disclosures are endogenously determined, which might bias the regression results. We addressed endogeneity by using an instrumental variable in the two-stage least squares regression. Assuming that the voluntary establishment of ECs might be influenced by board independence, firm size, economic performance, or firm financial structure, we first regressed ECs on the dummy variables of B_IND (where firms with independent directors are coded as 1 and as 0 otherwise), L_ASS, ROA, and DE and obtained the predicted value of EC. In the second stage, we used the predicted value of EC in place of the actual value in the regression test. Similarly, we obtained the predicted value of INST and used it in the second stage of the regression. We did not estimate the predicted value of B_IND because most of the firms have at least two independent members on their corporate boards, as required under the regulations. We conducted 2-SLS regression tests with the predicted values of EC and INST; the regression results are presented in Table 4.²

Table 4.

Regression Results Based on 2-SLS With Estimated Values for EC and INST for the Total Sample (N = 671).

Variables	Model 1: EWDI	VIF	Model 2: EWDI	VIF	Model 3: UWDI	VIF	Model 4: UWDI	VIF
Constant	0.95 (4.82)***		−0.59 (−3.33)***		−1.22 (−3.05)***		0.50 (−3.06)***
EC	3.79 (2.94)***	1.33	3.62 (2.66)***	2.51	4.07 (2.58)**	1.33	1.88 (2.78)**	2.51
INST	3.76 (2.65)**	1.51	6.54 (3.71)***	7.38	2.75 (2.61)**	1.51	1.72 (2.42)**	7.38
B_IND	0.79 (10.63)***	1.07	0.79 (11.22)***	1.08	1.28 (8.50)***	1.08	1.77 (7.56)***	1.08
EU_ETS	0.90 (3.63)***	1.12	0.61 (3.08)***	1.74	2.58 (8.20)***	1.12	0.41 (2.01)**	1.74
(EC × INST)			0.82 (2.30)*	7.38			0.43 (1.98)*	8.96
(EC × EU ETS)			4.18 (7.47)***	3.49			7.59 (7.50)***	3.49
Control variables
DE	0.02 (2.58)***	1.02	0.02 (3.15)***	1.04	0.01 (1.02)	1.02	0.01 (0.69)	1.04
ROA	−0.01 (−1.52)	1.77	−0.01 (−1.77)	1.78	−0.01 (−0.57)	1.77	−0.01 (−1.09)	1.78
MV/BV	−0.02 (−2.52)**	1.78	−0.02 (−2.85)**	1.80	−0.02 (−1.95)*	1.78	−0.03 (−1.98)*	1.80
L_ASS	0.03 (2.45)**	1.04	0.04 (2.44)**	1.04	0.01 (2.35)*	1.04	0.03 (2.29)*	1.04
Adjusted R²	0.84		0.73		0.82		0.88
F test	589.85***		238.63***		524.60***		660.46***

Note. 2-SLS = two-stage least squares; VIF = variance inflation factor; EC = environmental committee; INST = institutional shareholdings; B_IND = board independence; EU ETS = highly polluting industries; DE = debt-equity ratio; ROA = return on assets; MV/BV = market value/book value; L_ASS = log of assets. t values are given in parentheses; time and industry effects included in the panel analysis.

p < .10. **p < .05. ***p < .01.

The results contained in Models 1 and 3 of Table 4 show that the coefficients for EC (3.79 and 4.07 for EWDI and UWDI, respectively), INST (3.76 and 2.75 for EWDI and UWDI, respectively), B_IND (0.79 and 1.28 for EWDI and UWDI, respectively), and EU_ETS (0.90 and 2.58 for EWDI and UWDI, respectively) are positive. They are statistically significant either at the 1% or 5% level. Overall, these results suggest that carbon disclosures reflected by both indexes are positively associated with EC, INST, B_IND, and EU_ETS.

Discussion of the Results

Association Between Environmental Committees and Carbon Disclosures

The results for OLS and 2-SLS reveal that coefficients for ECs are positive and statistically significant at the 1% level. These findings are consistent with our Hypothesis 1 that Italian firms with ECs are associated with higher carbon disclosures compared with firms that have not established ECs and confirm the argument that ECs stimulate management to develop a disclosure policy that includes more detailed carbon information. In other words, the establishment of ECs encourages managers to implement a reporting policy that enhances the extent and transparency of carbon disclosures. Our finding is also consistent with the existing studies based on firms in common law countries (i.e., U.K. and Australian firms), which shows the positive impact of the environmental management system (e.g., Liao et al., 2015; Yunus et al., 2016). Overall, this study shows that in an age of heightened societal concerns about climate change, the creation of ECs is likely to play a crucial role in meeting the challenge of reporting detailed carbon information, even in code law countries.

Association Between Institutional Shareholdings and Carbon Disclosures

The coefficients for INST are positive and statistically significant at 1% for both OLS and 2-SLS tests and support our Hypothesis 2 that large institutional shareholdings have a significant influence on managerial behaviour in terms of disclosing carbon information to meet their information needs. This finding is similar to that reported by Cotter and Najah (2012) and reveals that the demand for carbon information is strong when a substantial percentage of shareholdings are held by institutions that generally have long-term interest in the firm. Another interpretation suggests that managers meet the information needs of institutional investors because the magnitude of their shareholdings gives them significant market power. Therefore, managers take into consideration the carbon information needs of institutional shareholders that enable those shareholders to evaluate risk associated with their investment in the firm. Consistent with stakeholder theory, the findings suggest that even Italian firms, which are more family-dominated and operate in a code law environment, recognize the power of institutional shareholders and are willing to meet their carbon information needs.

Association Between Board Independence and Carbon Disclosures

The OLS results for board independence show that coefficients are positive but marginally significant for UWDI and insignificant for EWDI. The results of the 2-SLS regression tests reveal that coefficients are positive and statistically significant for both disclosure indexes. Therefore, the test based on the 2-SLS supports our Hypothesis 3 that Italian firms with a higher percentage of independent directors on their corporate boards are associated with greater carbon disclosures when compared with firms with a lower percentage of independent directors. The OLS results do not fully support these results. The 2-SLS tests, which are consistent with the results of an existing study by Yunus et al. (2016), are also consistent with the stakeholder theoretical perspective, which suggests that independent directors are more sensitive to the social and environmental demands of investors and therefore should ensure the disclosure of environmental information, especially carbon information, for use by investors. An important aspect of this finding is that in today’s changed environment independent directors, even those in code law countries, are likely to perform the useful functions of providing advice to management and monitoring their activities to ensure that strategic plans regarding environmental performance and disclosure are congruent with societal expectations.³

Association Between Highly Polluting Industries and Carbon Disclosures

The results in Table 3 based on the OLS regression and the results in Table 4 based on the 2-SLS are all statistically significant at the 1% level. They are consistent with the argument that firms in highly polluting industries are under greater pressure to disclose carbon information compared with others and have greater incentive to be transparent to support their legitimacy. Firms disclosing carbon information convey a message to investors and society that they are taking their environmental responsibility to reduce carbon emissions and mitigate environmental risks seriously. These results are consistent with various previous studies (e.g., Bae Choi, Lee, & Psaros, 2013; Chu, Chatterjee, & Brown, 2013; Elijido-Ten, Kloot, & Clarkson, 2010; Hrasky, 2012; Liu & Anbumozhi, 2009).

The Joint Effect of Environmental Committees, Institutional Shareholdings, and Highly Polluting Industries

Models 2 and 4 in Tables 3 and 4 contain the results based on the joint effect of ECs and INST. The findings based on the OLS regression tests show that the coefficients of interaction (coefficients = 0.02 and 0.05 for EWDI and UWDI, respectively) are positive in both disclosure indexes but are marginally statistically significant (10% level) for EWDI. Similarly, the coefficients of the interaction term in the 2-SLS regression (0.82 and 0.43 for EWDI and UWDI, respectively) are positive and marginally statistically significant at the 10% level. The coefficients of ECs and INST remain statistically significant in both tests. These findings suggest that ECs have an incremental effect on the disclosure of carbon information when firms have high institutional shareholdings, but this incremental effect is not significant.

We also examined the joint effect of ECs for firms in highly polluting industries (EU_ETS). The findings reveal that the coefficients of the interaction term for OLS tests (0.57 and 2.04 for EWDI and UWDI, respectively) and for 2-SLS tests (4.18 and 7.59 for EWDI and UWDI, respectively) are positive and statistically significant at the 1% level. This indicates that carbon disclosures are especially high when firms belong to polluting sectors and at the same time have established ECs. Therefore, the establishment of ECs in firms in polluting sectors plays an especially important role in encouraging managers to disclose more detailed carbon information.

Results for Control Variables

The coefficients of DE are positive and in the expected direction for all models, but they are only statistically significant for Models 1 and 2 of Table 4, under EWDI, at the 1% level and are insignificant under UWDI. These results do not fully support the expectation that carbon disclosures are higher when firms are highly leveraged.

The coefficients of MV/BV are negative and not in the expected direction. Hahn et al. (2015) found that one of the two studies that used this ratio as a control variable found an insignificant association (Gallego-Álvarez, Rodríguez-Domínguez, & García-Sánchez, 2011), whereas the other found a positive association (Prado-Lorenzo et al., 2009). The coefficients of ROA are also negative and not in the expected direction, but they are insignificant. Hahn et al. (2015) found that several studies examined the impact of operating performance; only one documented a positive association (Jira & Toffel, 2013), while all others found an insignificant association. A possible explanation for the negative coefficients may be that firms that perform better economically focus more on their activities and do not pay enough attention to environmental plans and/or voluntary environmental disclosures. Finally, the coefficients for L_ASS are positive and in the expected direction. They are mostly statistically significant, except for the UWDI model in Table 4, where it is marginally significant. This result is mainly consistent with the argument that larger firms have stronger motivation to disclose environmental information, especially carbon information.

Additional Tests

We conducted additional tests to check the model’s robustness. We selected the reduced subsample of firms that disclose carbon information. We dropped firms with zero value for disclosure indexes (i.e., firms that have never reported carbon information) from the total sample. We report the results in Table 5. The findings based on 2-SLS tests therefore illustrate that the coefficients for EC, INST, B_IND, and EU_ETS are still positive and statistically significant for both indexes.⁴

Table 5.

Regression Results Based on 2-SLS With Estimated Values for EC and INST for the Subsample Disclosing Carbon Information (N = 281).

Variables	Model 1: EWDI	VIF	Model 2: EWDI	VIF	Model 3: UWDI	VIF	Model 4: UWDI	VIF
Constant	−1.38 (−3.37)***		−1.60 (−2.91)***		0.52 (−2.14)*		−0.32 (−2.69)**
EC	11.72 (4.13)***	1.01	13.05 (5.53)***	5.01	6.66 (1.95)*	1.01	8.16 (3.72)***	5.01
INST	16.90 (3.79)***	1.02	5.87 (5.15)***	6.02	5.55 (1.98)*	1.02	2.80 (2.16)**	6.02
B_IND	1.12 (5.48)***	1.01	1.05 (4.85)***	1.01	1.09 (4.22)***	1.01	0.63 (2.48)***	1.01
EU_ETS	0.69 (6.50)***	1.03	1.52 (4.02)***	1.03	2.45 (6.41)***	1.03	1.16 (2.03)**	1.03
(EC × INST)			5.88 (6.08)***	7.52			6.81 (5.74)***	7.52
(EC × EU ETS)			4.40 (2.30)**	6.88			2.83 (3.72)***	6.88
Control variables
DE	0.09 (5.25)***	1.01	0.09 (6.90)***	1.01	0.17 (4.09)***	1.01	0.22 (5.84)***	1.01
ROA	−0.02 (−1.84)*	1.07	−0.03 (−2.90)***	1.07	−0.02 (−0.72)	1.07	−0.05 (−2.01)**	1.07
MV/BV	0.03 (2.45)**	1.07	0.02 (2.65)***	1.07	0.02 (0.63)	1.07	0.01 (2.85)***	1.07
L_ASS	0.97 (3.66)***	1.05	1.02 (4.02)***	1.05	1.03 (2.69)***	1.05	1.09 (3.89)***	1.05
Adjusted R²	0.62		0.65		0.78		0.85
F test	75.66***		76.88***		102.20***		211.56***

p < .10. **p < .05. ***p < .01.

These findings are similar to those in Table 4 and confirm that the creation of ECs, substantial institutional shareholdings, board independence, and highly polluting industries play relevant roles in encouraging managers to report carbon information.

We also conducted an additional test on the subsample belonging to the highly polluting industries that is covered under the EU_ETS program to examine whether the positive association between the motivating factors is stronger. The results based on 2-SLS are shown in Table 6.

Table 6.

Regression Results Based on 2-SLS With Estimated Values for EC and INST for the Subsample Under the EU ETS (N = 329).

Variables	Model 1: EWDI	VIF	Model 2: EWDI	VIF	Model 3: UWDI	VIF	Model 4: UWDI	VIF
Constant	−1.90 (−2.68)***		−0.45 (−1.92)*		−0.07 (−1.74)*		−1.40 (−1.68)*
EC	9.48 (2.65)***	1.01	4.69 (2.45)**	4.48	0.87 (2.10)**	1.01	9.10 (2.61)**	4.48
INST	9.00 (2.05)**	1.03	7.24 (2.69)***	1.07	6.90 (2.69)**	1.03	1.51 (2.30)**	1.07
B_IND	1.69 (6.61)***	1.01	1.64 (6.68)***	1.07	4.34 (6.11)***	1.01	4.40 (6.80)***	1.07
(EC × INST)			3.70 (2.39)**	6.03			4.60 (1.40)*	6.03
Control variables
DE	0.03 (1.99)*	1.01	0.03 (1.55)	1.01	0.05 (0.90)	1.01	0.03 (0.58)	1.01
ROA	−0.02 (−1.50)	1.07	−0.01 (−0.99)	1.07	0.02 (1.06)	1.07	0.01 (1.06)	1.07
MV/BV	0.01 (0.55)	1.07	0.03 (0.25)	1.07	0.02 (0.09)	1.07	0.01 (0.11)	1.07
L_ASS	2.15 (3.00)***	1.05	1.69 (3.25)***	1.05	1.36 (2.89)***	1.05	1.33 (3.02)***	1.05
Adjusted R²	0.35		0.35		0.35		0.41
F test	35.90***		31.66***		37.41***		40.22***

Note. 2-SLS = two-stage least squares; VIF = variance inflation factor; EC = environmental committee; INST = institutional shareholdings; B_IND = board independence; DE = debt-equity ratio; ROA = return on assets; MV/BV = market value/book value; L_ASS = log of assets; t values are given in parentheses; time and industry effects included in the panel analysis.

p < .10. **p < .05. ***p < .01.

The coefficients of EC, institutional shareholdings, and board independence continue to be positive and statistically significant at the 1% and 5% levels for all models (for both EWDI and UWDI). The coefficients of the interaction term between EC and INST are also positive and statistically significant at the 5% and 10% levels for EWDI and UWDI, respectively. These results corroborate the argument that the motivating factors have a significantly stronger impact on carbon information from firms belonging to highly polluting industries than on carbon information from firms in other sectors.

Conclusion

This study extends the research on carbon disclosure, something that has been receiving increased attention for some time because of regulatory changes, worldwide nongovernmental voluntary initiatives, and the dangers associated with carbon emissions, which are considered the major culprit in global warming. In the absence of regulatory requirements, there is a general expectation that firms reduce their emissions voluntarily and disclose sufficient carbon information that would be useful to stakeholders to evaluate firms’ carbon performance and associated risks.

The results of this study reveal that Italian listed firms with voluntarily established ECs are more inclined to disclose carbon information than those without. Consistent with the premise of the stakeholder and legitimacy theories, one can reasonably conclude that having an EC is a step in the right direction by which firms can show their commitment to climate change mitigation and maintain legitimacy. Additionally, in accordance with stakeholder theory, we find that substantial institutional shareholdings, which have the power to influence stock prices, provide a strong motivation for management to disclose higher levels of information so that investors can evaluate risk in making their investment decisions. The study also indicates that board independence results in greater carbon disclosures, suggesting that having a larger percentage of independent directors improves firms’ strategic disclosure policies, even in code law countries. Moreover, firms belonging to highly polluting industries have stronger incentives to report carbon information compared with firms in other sectors.

The findings support the premise of the stakeholder theory that the stakeholders’ power and the firms’ strategic postures have significant influence on the managerial behaviour of Italian firms. The findings show that code law countries are now also feeling institutional shareholder pressure. Managers should pay special attention to the information needs of institutional shareholders because they have the power to influence firms’ stock prices, which in turn affect firms’ value and managers’ personal wealth (Meng, Zeng, Shi, Qi, & Zhang, 2014). Furthermore, consistent with legitimacy theory, managers recognize their environmental responsibilities to keep the stakeholders, particularly society in general, informed about their efforts to maintain and protect the environment. Managers should not forget that transparent corporate strategies and being “green” pay in the long run, and the reward can be a firm’s continued existence.

The findings provide useful information for regulators and confirm that the establishment of ECs is a step in the right direction as this might be helpful in improving firms’ carbon performance and associated disclosures. It is important that regulators keep a close watch over firms to ensure that disclosures are comprehensive and reliable. If voluntary disclosures fail to meet the transparency or reliability tests, regulators should consider developing appropriate mandatory rules and regulations to ensure that firms disclose detailed and reliable carbon information.

Given that the EU directives are applicable to all EU countries, the findings based on Italian firms should have validity for firms in other countries whose financial markets and legal environments are similar to those of Italy. Additional research across EU countries is needed to obtain better insight into the importance of ECs and board independence in other code law countries. This would enable managers to develop better policies and help regulators ensure the disclosure of high-quality information for stakeholders.

The findings of this study must be interpreted with caution because of the following limitations. First, the reliability of the results depends on the reliability of the carbon disclosures included in the sustainability reports issued by the Italian firms. Second, we may not have used all the control variables that might influence carbon disclosures. Third, the development of our disclosure index may not have fully captured different aspects of information that are necessary for carbon information to be of high quality. Our index is also not comprehensive; it primarily examines disclosures from the KP perspective. Furthermore, the carbon disclosure index may not necessarily reflect GHG emissions.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Bikki Jaggi is full professor of Accounting. He has extensively published in several business and accounting journals, including Journal of Accounting and Economics, Accounting Review, Journal of Accounting and Public Policy, Journal of Accounting, Auditing and Finance, and the International Journal of Accounting. Jaggi’s areas of expertise include financial accounting, environmental social accounting, international accounting, and corporate governance.

Alessandra Allini is associate professor of Accounting. She has published in several national and international accounting journal, including articles and book chapters published by referred international editors. Allini’s main areas of expertise include financial accounting, environmental and social disclosure, risk reporting, value relevance, and performance measurement in cultural organizations.

Riccardo Macchioni is full professor of Accounting. He published articles in referred international journals, national journals, monographs and books’ chapter published by referred editors. His research interests cover the areas of social and environmental disclosure, international financial accounting, and management accounting.

Claudia Zagaria is Research Fellow in Accounting. She has publications in referred international journals, national journals, and books’ chapter published by referred international editors. Her research interests cover the areas of social and environmental disclosure, risk reporting and risk management systems, management accounting, and business ethics.

References

Amran

Periasamy

Zulkafli

A. H.

(2014). Determinants of climate change disclosure by developed and emerging countries in Asia Pacific. Sustainable Development, 22, 188-204.

Apergis

Eleftheriou

Payne

J. E.

(2013). The relationship between International Financial Reporting Standards, carbon emissions, and R&D expenditures: Evidence from European manufacturing firms. Ecological Economics, 88, 57-66.

Assonime Report. (2016). IV Report on Corporate Governance.

Bae Choi

Lee

Psaros

(2013). An analysis of Australian company carbon emission disclosures. Pacific Accounting Review, 25, 58-79.

Bebbington

Unerman

O’Dwyer

(2014). Sustainability accounting and accountability. New York, NY: Routledge.

Benn

Dunphy

Griffiths

(2006). Enabling change for corporate sustainability: An integrated perspective. Australasian Journal of Environmental Management, 13, 156-165.

Berthelot

Robert

A. M.

(2011). Climate change disclosures: An examination of Canadian oil and gas firms. Issues in Social and Environmental Accounting, 5, 106-123.

Brammer

Pavelin

(2008). Factors influencing the quality of corporate environmental disclosure. Business Strategy and the Environment, 17, 120-136.

Carbon Disclosure Project. (2013). The Italy climate change report. Retrieved from https://www.cdp.net/en/reports/archive

10.

Carroll

Shabana

(2010). The business case for corporate social responsibility: A review of concepts, research and practice. International Journal of Management Reviews, 12, 85-105.

11.

Chu

C. I.

Chatterjee

Brown

(2013). The current status of greenhouse gas reporting by Chinese companies: A test of legitimacy theory. Managerial Auditing Journal, 28, 114-139.

12.

Chuang

C. H.

Chen

S. J.

Chuang

C. W.

(2013). Human resource management practices and organizational social capital: The role of industrial characteristics. Journal of Business Research, 66, 678-687.

13.

Clarkson

Richardson

(2004). The market valuation of environmental capital expenditures by pulp and paper companies. Accounting Review, 79, 329-353.

14.

Cotter

Najah

M. M.

(2012). Institutional investor influence on global climate change disclosure practices. Australian Journal of Management, 37, 169-187.

15.

Deegan

Gordon

(1996). A study of the environmental disclosure practices of Australian corporations. Accounting and Business Research, 26, 187-199.

16.

Deegan

C. M.

(2006). Financial accounting theory (2nd ed.). North Ryde, New South Wales, Australia: McGraw-Hill.

17.

Depoers

Jeanjean

Jérôme

(2016). Voluntary disclosure of greenhouse gas emissions: Contrasting the carbon disclosure project and corporate reports. Journal of Business Ethics, 134, 445-461.

18.

Dietz

Hope

Sterv

Zevghelis

(2007). Reflections on the stern review: A robust case for strong action to reduce the risks of climate change. World Economics, 8(1), 121-168.

19.

Di Pietra

Grambovas

C. A.

Raonic

Riccaboni

(2008). The effects of board size and “busy” directors on the market value of Italian companies. Journal of Management & Governance, 12, 73-91.

20.

Doran

K. L.

Quinn

E. L.

(2009). Climate change risk disclosure: A sector by sector analysis of SEC 10-K filings from 1995-2008. North Carolina Journal of International Law and Commercial Regulation, 34, 721-766.

21.

Dowling

Pfeffer

(1975). Organisational legitimacy: Societal values and organizational behavior. Pacific Sociological Review, 18, 122-136.

22.

Dragomir

V. D.

(2012). The disclosure of industrial greenhouse gas emissions: A critical assessment of corporate sustainability reports. Journal of Cleaner Production, 29, 222-237.

23.

EAA Report, No 10/2013. Trends and projections in Europe. 09 Oct 2013. Retrieved from http://www.eea.europa.eu/publications/trends-and-projections-2013

24.

Elijido-Ten

(2007). Applying stakeholder theory to analyze corporate environmental performance: Evidence from Australian listed companies. Asian Review of Accounting, 15, 164-184.

25.

Elijido-Ten

Kloot

Clarkson

(2010). Extending the application of stakeholder influence strategies to environmental disclosures: An exploratory study from a developing country. Accounting, Auditing & Accountability Journal, 23, 1032-1059.

26.

Freedman

Jaggi

(2005). Global warming, commitment to the Kyoto Protocol, and accounting disclosures made by the largest public firms from polluting industries. International Journal of Accounting, 40, 215-232.

27.

Freedman

Jaggi

(2009). Global warming and corporate disclosures: A comparative analysis of companies from the European Union, Japan and Canada. In Freedman

Jaggi

(Eds.), Sustainability, environmental performance and disclosures (pp. 129-160). Bingley, England: Emerald.

28.

Freedman

Jaggi

(2011). Global warming disclosures: Impact of Kyoto Protocol across countries. Journal of International Financial Management and Accounting, 22, 46-90.

29.

Freeman

(1984). Strategic management: A stakeholder approach. London, England: Pitman, Marshall.

30.

Gallego-Álvarez

Rodríguez-Domínguez

García-Sánchez

I. M.

(2011). Study of some explanatory factors in the opportunities arising from climate change. Journal of Cleaner Production, 19, 912-926.

31.

Ghomi

Z. B.

Leung

(2013). An empirical analysis of the determinants of greenhouse gas voluntary disclosure in Australia. Accounting and Finance Research, 2(1), 110-127.

32.

González-Benito

(2006). The role of stakeholder pressure and managerial values in the implementation of environmental logistics practices. International Journal of Production Research, 44, 1353-1373.

33.

Graham

J. R.

Harvey

C. R.

Rajgopal

(2005). The economic implications of corporate financial reporting. Journal of Accounting and Economics, 40, 3-73.

34.

Gray

Kouhy

Lavers

(1995). Corporate social and environmental reporting: A review of the literature and a longitudinal study of UK disclosure. Accounting, Auditing & Accountability Journal, 8(2), 47-77.

35.

Groening

Kanuri

V. K.

(2013). Investor reaction to positive and negative corporate societal events. Journal of Business Research, 66, 1852-1860.

36.

Hahn

Reimsbach

Schiemann

(2015). Organizations, climate change, and transparency reviewing the literature on carbon disclosure. Organization & Environment, 28, 80-102.

37.

Haniffa

R. M.

Cooke

T. E.

(2005). The impact of culture and governance on corporate social reporting. Journal of Accounting and Public Policy, 24, 391-430.

38.

Hrasky

(2012). Carbon footprints and legitimation strategies: Symbolism or action? Accounting, Auditing & Accountability Journal, 25, 174-198.

39.

Jaffar

Iskandar

T. M.

Muhamad

(2002). Investigation of environmental disclosures: Evidence from selected industries in Malaysia. International Journal of Business and Society, 3, 55-68.

40.

Jaggi

Allini

Rossi

F. M.

Caldarelli

(2016). Impact of accounting traditions, ownership and governance structures on financial reporting by Italian firms. Review of Pacific Basin Financial Markets and Policies, 19(1). doi:10.1142/S0219091516500016

41.

Jira

Toffel

M. W.

(2013). Engaging supply chains in climate change. Manufacturing & Service Operations Management, 15, 559-577.

42.

Johnson

R. A.

Greening

D. W.

(1999). The effects of corporate governance and institutional ownership types on corporate social performance. Academy of Management Journal, 42, 564-576.

43.

La Porta

Lopez-de-Silanes

Shleifer

Vishny

(2000). Investor protection and corporate governance. Journal of Financial Economics, 58, 3-27.

44.

Liao

Luo

Tang

(2015). Gender diversity, board independence, environmental committee and greenhouse gas disclosure. British Accounting Review, 47, 409-424.

45.

Liesen

Hoepner

A. G.

Patten

D. M.

Figge

(2015). Does stakeholder pressure influence corporate GHG emissions reporting? Empirical evidence from Europe. Accounting, Auditing & Accountability Journal, 28, 1047-1074.

46.

Liu

Anbumozhi

(2009). Determinant factors of corporate environmental information disclosure: An empirical study of Chinese listed companies. Journal of Cleaner Production, 17, 593-600.

47.

Lovell

Bebbington

Larrinaga

de Aguiar

T. R. S.

(2013). Putting carbon markets into practice: A case study of financial accounting in Europe. Environment and Planning C: Government and Policy, 31, 741-757.

48.

Luo

Tang

(2014). Does voluntary carbon disclosure reflect underlying carbon performance? Journal of Contemporary Accounting & Economics, 10, 191-205.

49.

Meng

X. H.

Zeng

S. X.

Shi

J. J.

G. Y.

Zhang

Z. B.

(2014). The relationship between corporate environmental performance and environmental disclosure: An empirical study in China. Journal of Environmental Management, 145, 357-367.

50.

Michelon

Parbonetti

(2012). The effect of corporate governance on sustainability disclosure. Journal of Management & Governance, 16, 477-509.

51.

Patten

D. M.

(2002). The relation between environmental performance and environmental disclosure: A research note. Accounting, Organizations and Society, 27, 763-773.

52.

Peters

G. F.

Romi

A. M.

(2014). Does the voluntary adoption of corporate governance mechanisms improve environmental risk disclosures? Evidence from greenhouse gas emission accounting. Journal of Business Ethics, 125, 637-666.

53.

Porter

Van der Linde

(1995). Toward a new conception of the environment–competitiveness relationship. Journal of Economic Perspectives, 9(4), 97-118.

54.

Prado-Lorenzo

J. M.

Garcia-Sanchez

I. M.

(2010). The role of the board of directors in disseminating relevant information on greenhouse gases. Journal of Business Ethics, 97, 391-424.

55.

Prado-Lorenzo

J. M. J.

Rodríguez

Gallego

García

(2009). Factors influencing the disclosure of greenhouse gas emissions in companies world-wide. Management Decision, 47, 1133-1157.

56.

PricewaterhouseCoopers. (2012). Climate change addresses the Stock Exchange. Milan, Italy.

57.

Rankin

Windsor

Wahyuni

(2011). An investigation of voluntary corporate greenhouse gas emissions reporting in a market governance system: Australian evidence. Accounting, Auditing & Accountability Journal, 24, 1037-1070.

58.

Roberts

R. W.

(1992). Determinants of corporate social responsibility disclosure: An application of stakeholder theory. Accounting, Organizations and Society, 17, 595-612.

59.

Ryan

L. V.

Schneider

(2002). The antecedents of institutional investor activism. Academy of Management Review, 27, 554-573.

60.

Scholtens

Kleinsmann

(2011). Incentives for subcontractors to adopt CO₂ emission reporting and reduction techniques. Energy Policy, 39, 1877-1883.

61.

Smith

J. A.

Morreale

Miriani

M. E.

(2008). Climate change disclosure: Moving towards a brave new world. Capital Markets Law Journal, 3, 469-485.

62.

Solomon

J. F.

Solomon

Norton

S. D.

Joseph

N. L.

(2011). Private climate change reporting: An emerging discourse of risk and opportunity? Accounting, Auditing & Accountability Journal, 24, 1119-1148.

63.

Stanny

(2013). Voluntary disclosures of emissions by US firms. Business Strategy and the Environment, 22, 145-158.

64.

Stanny

Ely

(2008). Corporate environmental disclosures about the effects of climate change. Corporate Social Responsibility and Environmental Management, 15, 338-348.

65.

Stechemesser

Guenther

(2012). Carbon accounting: A systematic literature review. Journal of Cleaner Production, 36, 17-38.

66.

Tauringana

Chithambo

(2014). The effect of DEFRA guidance on greenhouse gas diclosure. British Accounting Review, 1, 1-20.

67.

Verstegen

R. L.

Schneider

(2003). Institutional investor power and heterogeneity implications for agency and stakeholder theories. Business & Society, 42, 398-429.

68.

Wagner

Hespenheide

Pavlovsky

(2009). The responsible and sustainable board. Deloitte Review, 7, 53-64.

69.

Wang

Dewhirst

H. D.

(1992). Boards of directors and stakeholder orientation. Journal of Business Ethics, 11, 115-123.

70.

Watts

I. L.

Zimmerman

(1978). Towards a positive theory of the determination of accounting standards. Accounting Review, 1, 112-134.

71.

Wegener

Elayan

F. A.

Felton

(2013). Factors influencing corporate environmental disclosures. Accounting Perspectives, 12, 53-73.

72.

Yunus

Elijido-Ten

Abhayawansa

(2016). Determinants of carbon management strategy adoption. Managerial Auditing Journal, 31, 156-179.