The influence of CEO characteristics on greenwashing: A quantitative research on the impact of CEO characteristics on environmental decoupling

(1)

The influence of CEO characteristics on greenwashing: A

quantitative research on the impact of CEO characteristics

on environmental decoupling

Combined thesis Accountancy & Control Date: 1/18/2020

Name: Menno Bouma Supervisor: Nazim Hussain Student number: S3853683 E-mail: m.bouma.13@student.rug.nl

Word count: 10.164

Abstract

Many corporate scandals occurred in recent years; these scandals lead to increasing importance of corporate social responsibility (CSR). This study analyses the CEO characteristics of age, gender, tenure, and financial education in relationship to environmental decoupling. The level of environmental decoupling is the gap between environmental performance and environmental disclosure. This study builds on the agency theory and upper echelon theory. The agency theory suggest that board independence should reduce information asymmetry, therefore the moderating effect of board independence is examined. Board independence serves its role in controlling and monitoring firms by a reducing effect on environmental decoupling. The upper echelon theory builds on the idea that the CEO is likely to impact organizational performance and disclosure of a company. A database of 16.880 firm-year observations from 4.404 organizations between 2006 and 2017 is being used. This study finds significant relationships between the CEO characteristics of age, tenure and, financial education with environmental decoupling. No relationship is found for gender. Furthermore, robustness checks were performed by focussing on the biggest outliers of greenwashing and brownwashing data in the sample.

Keywords: CSR decoupling, CEO characteristics, Agency theory, Upper echelon theory,

(2)

The influence of CEO characteristics on greenwashing: A quantitative research on the impact of CEO characteristics on environmental decoupling

2

Table of content

1. introduction ... 3

2. Literature review ... 7

2.1 Agency theory and the upper echelon theory ... 7

2.2 CSR decoupling and greenwashing ... 8

2.3 CEO characteristics ... 11

2.4 Hypotheses ... 15

2.4.1 CEO characteristics ... 15

2.4.2 Board independence. ... 18

3. Methodology ... 19

3.1 Data collection and sample selection ... 19

3.2 Variables ... 20

3.2.1 Environmental decoupling ... 20

3.2.2 Independent variable: CEO characteristics ... 20

3.2.3 Board independence ... 21 3.2.4 Control variables ... 21 4. Results ... 22 4.1 Descriptive statistics ... 22 4.2 Regression analyses ... 25 4.3 Robustness check ... 27 5. Discussion ... 28 6. Conclusion ... 31

6.1 Conclusion and contribution ... 31

6.2 Limitations and future research ... 32

(3)

3

1. introduction

One of the most famous cases of greenwashing is the Volkswagen ‘’clean’’ diesel engine scandal. With special software they made their cars appear more environmentally friendly than they actually were. To make this scandal even more painful, Volkswagen was just been named as the leader of the automotive industry in the Dow Jones Sustainability Index. It demonstrated the weaknesses and trustworthiness of Corporate Social Responsibility (CSR) performance. Just a few months later they were removed from all Sustainability Indexes because of this scandal (Wilburn & Wilburn, 2016). Another famous greenwashing case was Shell in 2007 when they used a rather controversial slogan: ‘’Don’t throw anything away, there is no away’’. The company was criticized for this misleading advertisement and was criticised for the differences in environmental communication and actual environmental performance. This case showed that many different actors interfered, such as the media, legal authorities, and environmental lobby groups. This is another example that greenwashing could lead to negative attention for an organization (Seele & Gatti, 2017). These are just two of the many greenwashing scandals that happened in different industries over the last decades. It becomes more common practice for various stakeholders that the success of a company should not be solely judged in economic terms. Stakeholders such as customers or investors are focussing increasingly on a company’s performance in terms of its social and environmental dimensions (Wiengarten et Al, 2017).

Scandals like the dieselgate of Volkswagen led to more external pressures to engage in environmentally friendly actions. The CEO plays a key role in the implementation of these environmental activities and strategies. CEO characteristics are important determinants in the CSR performance of companies (Walls & Berrone, 2017). Furthermore, scandals like these led to even more attention regarding the role of a Chief Executive Officer (CEO) and their role and influence on CSR activities. CEO activities can have a large impact on CSR activities and strategies, Velte (2019) assumes that the upper echelons theory can explain the impact of certain CEO characteristics on CSR performance. The managerial ability of CEOs has a direct and indirect impact on firm performance. The better the ability of managers, the more willing they are to disclose CSR information in favor of their stakeholders (García-Sánchez et

(4)

4

al, 2020). Lewis et al (2014) acknowledge that certain CEO characteristics play a role in the level of voluntary environmental disclosure of companies. Thus, CEOs do play an important role in the quality and extent of voluntary disclosure, and as a result they have an impact on the transparency of a company. There will always be information asymmetry between the company and its stakeholders as there are different interests. CEOs could be tempted to disclose only the positive information by removing or distracting the negative information in their disclosures (Lopes, 2013). Sauerwald & Su (2019) describe the gap between how firms communicate about CSR and what firms actually do in the area of CSR as CSR decoupling. In this paper we focus on the environmental aspects of CSR. When a firm discloses more than it actually does on environmental activities then it is called greenwashing. On the other hand, it could be the case that a firm environmentally performs more than it discloses. This is called brownwashing which is further elaborated in this paper. Both greenwashing and brownwashing are a form of CSR decoupling, more specifically decoupling on the environmental aspects of CSR. Therefore, we refer to it more specifically as environmental decoupling. Thus, we refer to both greenwashing and brownwashing as environmental decoupling in this paper.

More and more voluntary CSR reports are being issued by firms. The motivations for issuing these CSR reports are often not clear. Mahoney et al (2013) considers two explanations why firms voluntarily disclose these reports, namely out of signaling purposes to pose as a ‘’good’’ company, or out of greenwashing purposes. Greenwashing encompasses the communications which mislead people to adopt an overly positive belief about the environmental performance, practice, or products of a company (Lyon and Montgomery, 2015). Delmas and Burbano (2011) state that greenwashing arises from poor environmental performance and/or overly positive communication about environmental performance. More and more firms engage in greenwashing activities (Torelli et al, 2020). Walker and Wan (2012) investigated that greenwashing activities have a negative impact on the financial performance. An important side note in this research is that it is executed in the area of polluting industries, which are more subject to stakeholder scrutiny. More scrutiny and attention from the stakeholders will indirectly lead to a higher likelihood that they will identify environmental decoupling. Furthermore, information about the environmental

(5)

5

performance of polluting firms is publicly available due to stricter disclosure regulations, and thus it is easier to discover greenwashing activities compared to other industries. As the polluting industries are viewed as the leading cause of environmental damage, the expectations for firms in these industries to reduce their environmental damage are higher compared to other industries. Failure to comply with these expectations will result in a decrease of financial performance. Other literature describes other negative effects of greenwashing of firms. Delmas and Burbano (2011) describe that the incidence of greenwashing can have a negative influence on consumer confidence in (green) products and it could negatively influence investor confidence in environmentally friendly organizations. Therefore, Hinze & Sump (2019) describe that future research needs to clarify the difference between CSR disclosure and CSR performance, for a better understanding of the influence of CEO characteristics.

Delmas and Burbano (2011) describe that greenwashing arises from external, organizational, and individual levels of the organization. First, the external level consists out of non-market actors such as regulators, non-governmental organizations (NGO’s), and media pressure. On the other hand, there are market actors such as customers, investors, and competitors. Berrone and Gomez-Meija (2009) state that ‘’environmental issues are a major social concern, companies in polluting industries face tight governmental regulations, increased media attention, and strong environmental activism’’. On the other hand, according to Delmas and Burbano (2011) industries do not face tight governmental regulations and they suggest that the current weak regulations are the key driver of greenwashing. Regulation on CSR disclosure is limited in most countries, and regulation between countries differs a lot as well. These differences in regulations make it difficult for multinationals to follow the different regulations across countries, which leaves them more space to engage in environmental decoupling. Stricter regulations would serve the best to reduce environmental decoupling. However, it is hard to define environmental decoupling as there is not one single explanation. This confusion and clarity around environmental decoupling make the implementation of more stringent regulations more challenging. Second, environmental decoupling arises from organizational-level drivers, such as firm characteristics, incentive structure, ethical climate, effectiveness of intra-firm

(6)

6

communication, and organizational inertia. It has been shown that incentive structures and ethical climates within firms can be determinants of (un)ethical behavior. Unethical behavior is the behavior that negatively affects others and is either illegal or morally unacceptable. For example, managers are more likely to engage in unethical behavior, such as environmental decoupling, when they receive rewards for attaining certain financial goals. These incentives could drive managers to take shortcuts in validating the correctness of communication messages or cause managers to ‘’look the other way’’. Third, individual psychological drivers can affect environmental decoupling of firms. Delmas and Burbano (2011) describe different psychological drivers, such as the optimistic biases where managers may over-estimate the positive results of greenwashing and underestimate the possible negative events out of the media or governmental regulatory institutions. However, the influence of CEO characteristics on environmental decoupling remains unexplored. As Delmas and Burbano (2011) described how challenging it is to improve regulations, they consider the role of the manager as another important factor to reduce greenwashing activities. Hinze & Sump (2019) state that the role of individual characteristics might play an important role in information disclosures. They assume that combining financial and sustainability reports in one integrated report can lead to agency problems, and therefore may not lead to valuable information. Research is limited whether these individual characteristics affect environmental decoupling of firms. Therefore, in paper adds to former research to investigate whether CEO characteristics influence environmental decoupling. Therefore, the main research question of this study is: Is environmental decoupling of firms affected by the characteristics of the CEO?

Kim and Lyon (2015) extended the environmental decoupling literature with a focus on brownwashing. It is a relatively new subject in research, and there is only some empirical research done. Brownwashing could be described as the opposite of greenwashing, where firms understate their environmental performances in their disclosure. Thus, they walk more than the actual talk. There are different reasons why companies are struggling to implement and disclose environmental activities. Kim & Lyon (2015) argue that companies may do it on purpose, especially when they face economic challenges. Brownwashing could be an incentive during challenging periods to alleviate the concerns of shareholders who are worried about environmental costs.

(7)

7

This paper adds to the empirical research in executing robustness checks on the influence of CEO characteristics and brownwashing activities.

This review is structured as follows. First, the literature review presents the used theories for this research and the hypothesises. Chapter three describe the methodology and the sample selection, followed by the results in chapter four. Chapter five discusses the results. Lastly, chapter six presents the conclusions, practical implications, and limitations of this research.

2. Literature review

2.1 Agency theory and the upper echelon theory

The agency theory suggest that the interests of managers and owners may diverge from each other and that firms’ managers do not always make the best decisions for society, or do not make the best decisions for their own firm, but act in their own interest (Nyberg et al, 2010). CSR activities are often considered as a possible agency problem between the stakeholders and shareholders. Although, firms with a good corporate governance structure have less agency problems and accomplish better CSR scores (Ferrel et al, 2016).

Information asymmetry arise when the sender communicates certain information and the receiver interprets the signal of the sender differently. Differences in information arises as the sender has access to more or other information than the receiver. Environmental information is often complex and subject for discussion, which make it difficult for outsiders to assess the quality of information. Every firm can choose what information to signal, and what information not to signal to outsiders (Seele & Gatti, 2017).

The dissemination of sustainability reports contributes to greater awareness of sustainable development. Individuals and companies respond to the demands of stakeholders to change sustainable behavior in line with their CSR related responsibilities such as social and environmental activities. The purpose of the dissemination of voluntary disclosures is to mitigate the information asymmetry between the firm and its stakeholders. On the other hand, the assumptions of the agency theory could lead to gaps in the voluntary disclosure of the company and the expectations of the stakeholders about the extent of this voluntary disclosure.

(8)

8

Companies tend to select the voluntary disclosed information and to detract or remove the negative information (Lopes, 2013). In this regard, managers may not, or not entirely include sustainability into their business strategies because it is not in their favor of interests (Garcìa-Sánchez & Martìnez-Ferrero, 2019).

The paper of Hambrick & Mason (1984) describe how managerial characteristics could be used to predict future firm performance, it is called the upper echelon perspective. It describes that the organization is a reflection of its top managers, or the so-called upper echelons, who act out of personal interpretations such as values, experiences, and personalities. The theory states that individual top managers influence organizational outcomes, which are based on different sorts of personal characteristics. The characteristics of these top managers can help to explain strategic choices and firm performance. According to Velte (2019), the upper echelon theory helps to explain the impact of different characteristics on CSR performance. The theory relies on the assumption that a CEO has strong influence in the top management team to influence the CSR activities and strategies. Hambrick & Mason (1984) draw on the upper echelon’s perspective because of 3 benefits. First, it may offer better measures to predict future firm performance. Second, it could benefit the selecting and developing methods of top management executives. And third, a strategist may be possible to predict competitor’s moves and strategy. They concluded that top management characteristics do influence firm performance. The upper echelons theory describe that the characteristics of the top management team could influence strategic outcomes of a firm (Peters et al, 2019).

2.2 CSR decoupling and greenwashing

CSR disclosure is the response of a company to the external world where it describes its actions and policies for being responsible and transparent. In these disclosures the organization communicates its CSR initiatives to enlarge the visibility and transparency. Companies disclose these CSR reports to reduce the information asymmetry of different stakeholders and to accomplish the expectations of these stakeholders to gain legitimacy and gather social support from the external world (Sauerwald and Su, 2019). CSR performance is the actual internal CSR actions undertaken by a firm. The gap between CSR disclosure and CSR performance is described as CSR decoupling. García-Sánchez et al (2020) state that a wider CSR

(9)

9

decoupling leads to higher analyst’ forecasts, a greater cost of capital, and reduced access to finance. On the other hand, more voluntary disclosure of CSR activities leads to a reduction in the cost of equity capital and lower forecast errors. This is an incentive for companies to engage in voluntary CSR disclosure, but also create incentives to engage in CSR decoupling activities (Dhaliwal et al, 2011).

The term ‘’greenwashing’’ originates from 1986 when an activist named Jay Westerveld said that hotels claimed to be environmentally engaged by asking the customers to re-use their towels. Actually, they asked this out of cost saving measures, not directly to be environmentally engaged. Furthermore, hotels were using this statement to mask other, more serious environmental impacts of the industry (Pearson 2010). Nowadays, greenwashing is still a broadly used term and is increasingly important for different stakeholders. Greenwashing draws on the environmental aspects of CSR decoupling. Greenwashing becomes a more and more important topic for academic literature, but despite this growing interest, there is still no generally accepted definition of greenwashing. This is because of variance in different perspectives of different researchers about greenwashing. Torelli et al (2020) state different definitions of greenwashing. They define greenwashing as ‘’disinformation disseminated by an organization to present an environmentally responsible public image’’. Another definition of greenwashing is ‘’the act of misleading consumers regarding the environmental practices of a company or the environmental benefits of a product or service’’. Greenwashing could be defined as the gap between what a firm communicates it does in regard to environmental performances versus what it actually does. The paper of Walker & Wan (2012) describe this gap respectively as the green ‘’talk’’ versus the green ‘’walk’’. The paper of Cho et al (2015) argues that this gap exists because of contradictory societal and institutional pressures. The consequence of this is that companies are tempted to engage in hypocrisy and to develop facades. As result, sustainable reports will never evolve into strong reliable disclosures.

If greenwashing has so many disadvantages, you can question why businesses engage in environmental decoupling. Businesses want to improve the reputation of the firm, get access to more resources, a stronger workforce, and get access to more or better business partners. Normally, in the end these improvements will result in improved financial performance of firms (Seele & Gatti, 2017). Shortly, there are a lot

(10)

10

of reasons to ‘’talk’’ more than the actual ‘’walk’’, and thus reason to engage in greenwashing. Although, you have to be aware that firms do not always engage in environmental decoupling on purpose. Bowen and Aragon-Correa (2014) describe this and have some additional concerns regarding four assumptions in academic literature about the definition of greenwashing. First, the academic literature describe that greenwashing is only about information disclosure. Second, greenwashing is always deliberate. Third, greenwashing is dominantly initiated by companies. And last, greenwashing is beneficial for firms and costly to society (Lyon and Montgomery, 2015).

At last, this research will execute additional empirical analyses on the relatively new subject of brownwashing. It can be described as the opposite of greenwashing. It is the understatement of environmental performance in the firm’s disclosure (Kim & Lyon, 2015). The paper makes clear that brownwash focuses on the environmental aspects of CSR, but the term also applies to the other CSR aspects of social and governance disclosures. Testa et al (2018) describe brownwashing as the non-communication about environmental activities of a firm to its external users. When firms don’t disclose their environmental activities, they will not exploit any reputational benefits out of these environmentally friendly actions. Furthermore, the results of Testa et al (2018) show that a lack of green communications leads to information asymmetry between the firm and its external investors. These investors have a negative view of the environmental actions of this firm and don’t see the substantial environmental actions of these firms. Due to this information asymmetry, the external investors can’t fully judge the firm’s environmental performance. This information asymmetry leads eventually to a higher risk premium, which increases the firm’s expenditures and will finally result in a lower bottom line performance. Investors are increasingly searching to invest in more green sectors these days. These investors could decide to take fewer shares in brownwashing firms in their portfolio, which will negatively affect market value and liquidity.

Kim & Lyon (2015) describe several reasons why firms would brownwash by issuing disclosure which understate their environmental performance. They suggest that managers may understate their social activities to disguise the related costs. This could be the case when shareholders suspect that managers pursuing CSR programs

(11)

11

which negatively affects other programs where shareholders are more interested in. Other studies suggest that CSR activities negatively affect firm performance in the short term, but are beneficial on the long term. Thus, short term oriented shareholders could create incentives for managers to engage in brownwashing. These findings are supported by empirical research of Jacobs et al (2010), which shows significant evidence that the announcement of voluntary emission reductions results in a negative return of -0,95%. The paper of Lyon et al (2013) shows that firms who are winning Green Company Awards in China did experience on average a lower firm value, especially the private ownership reacted more negatively to winning Green Company Awards. We will execute robustness checks whether CEO characteristics do influence brownwashing activities of companies. There is no research done yet in this area and thus this research will give empirical evidence, which can be used as a starting point in future research. Table 1 provides a literature overview of earlier research in the area of greenwashing and brownwashing.

2.3 CEO characteristics

The earlier described upper echelons theory state that the characteristics of the top management team could affect strategic outcomes of a firm (Peters et al, 2019). The CEO is as decision maker responsible for the firm’s corporate CSR strategy and is in the prime position to influence the CSR practices (Huang, 2013). Therefore, we will investigate whether different types of CEO characteristics have an impact on the firm’s environmental performances. We argue that CEO characteristics have an influence on environmental decoupling of companies. Environmental decoupling is executed in disclosures about environmental performance, these disclosures are often on a voluntary basis and therefore subject for discussion and managerial discretion (Clarkson et al, 2008). This paper will investigate the characteristics of gender, age, tenure, and education.

(12)

Table 1: Overview of prior studies on CSR decoupling

Study Independent variable Dependent variable

CSR measure Theory Sample Country

Cho et al. (2015)

Contradictory societal and institutional pressures (-) CSR decoupling Examination of CSR disclosure and CSR performance Signaling theory and legitimacy theory

2 oil and gas corporations US Crilly et al. (2012) Competing stakeholder expectations (+) CSR decoupling Stakeholder interviews Institutional theory 190 interviews International Delmas and Burbano (2011)

Nonmarket ext. drivers (+), Market external drivers (+), Organizational drivers (+), Individual psychological drivers (+).

Greenwashing Literature review Institutional theory n/a (literature review) n/a (literature review) Garcia-Sánchez et al. (2020) Analysts’ coverage (-) CSR decoupling Gap between CSR disclosure and CSR performance scores. Stakeholder theory 7681 firm-year observations, for the years 2006 -2015 US Graafland and Smid (2016) CSR responsibility (-) CSR decoupling Divergence among CSR policies and programs Institutional theory 1000 firm observations International Kim and Lyon (2015) Growth in output (+), Deregulation (-), Low profits under deregulation (-).

CSR decoupling Voluntary Reporting of Greenhouse Gases Program Legitimacy theory 396 firm-year observations, for the years 1995 -2003 US Lyon and Montgomery (2015) Articles mentioning

(13)

13

Peters et al. (2019)

CSO appointments (+) _{CSR performance CSO appointments} _{Upper Echelon} theory 1768 firm-year observations between 2002 -2008 US Sauerwald (2019)

CEO overconfidence (+) CSR decoupling Issued CSR reports Agency theory Firm-years between 2006 -2014

US

Seele and Gatti (2017)

Effect on legitimacy (-) _Greenwashing _{Literature review} _Signaling theory and legitimacy theory n/a (literature review) n/a (literature review) Shabab et al. (2019) CEO research background (+), CEO financial background (+), CEO foreign exposure (+), CEO age (-). Environmental performance RKS sustainability rankings, Upper echelon theory 16324 firm-year observations between 2010 -2017 China Tashman et al. (2019)

Home country institutional

voids (-) CSR decoupling CSR performance _{from IVA database,} score between 1 to 10 Neo-institutional theory 333 firm-year observations between 2005 -2012 International Testa et al. (2018)

Firm performance (-) _{CSR decoupling} _{Green practice} index (GPI), green communication index (GCI) and discrepancy index (DI) Stakeholder theory and Signaling theory 3.490 firm-year observations between 2002 - 2014 International

(14)

14

Torelli et al. (2020)

Stakeholders’ responses

to greenwashing (-) Greenwashing (Non-) _{Environmental} sensitive industry (ESI and non-ESI)

Legitimacy and signaling theory 128 surveys’ n/a Walker and Wan. (2011)

Financial performance (-) Greenwashing Substantive action versus symbolic action (Score 1 to 7) Legitimacy and signaling theory 103 firm observations in 2008. Canada

(15)

2.4 Hypotheses

2.4.1 CEO characteristics

Sustainability could be associated with renewal and innovation which is often associated with young and ambitious people. The paper of Hambrick & Mason (1984) summarizes that the age of TMT managers have an effect on corporate growth. Young managers appear to create higher growth rates compared to older managers, implement more novel ideas, and take more risks. Yasser et al (2020) conclude that directors’ age has a significant impact on CSR performance, but this research focuses only on emerging economies and it is difficult to align these results with other economies. Kish-Gephart et al (2010) suggest that age is related to unethical choices. Their results shows that elder CEOs operate at higher levels of moral reasoning. Further research will add to these weak findings whether CEO age does matter in unethical behavior, such as environmental decoupling.

There is not much research conducted about influence of CEOs age and their influence on CSR performance, neither on their influence on environmental decoupling. Thus, this hypothesis will just give some empirical evidence whether CEOs age influences environmental decoupling. We hypothesize that younger CEOs engage less in environmental decoupling because we argue that younger CEOs are more engaged with environmental performance compared to older CEOs, and therefore

engage less in environmental decoupling.

H1a: Environmental decoupling at firms reduces by a younger age of the CEO.

The results of Setó-Pamies (2015) shows a significant positive relation between women in the TMT and CSR performance, which suggest that women play an important role in improving CSR activities. These results show that the presence of women in the TMT influences CSR performance, but not specified on certain board positions such as the CEO. The paper of Huang (2013) adds to this that CEO gender does affect CSR performance. Peni & Vähämaa (2010) showed that female CFO’s are associated with income-decreasing discretionary accruals, which suggest that female CFO’s are more conservative. Croson and Greezy (2009) state that a large amount of literature documents differences in risk taking, women are more risk averse compared to men. They describe three reasons for these differences. First, women differ in their emotional reaction to uncertain situations, which results in differences in risk taking.

(16)

16

Second, men are more confident, which results in a different perception of risks. And last, women see risks as threats where men see challenges. These differences lead to variations in risk tolerances.

This research adds to this former literature to investigate whether a female CEO influences environmental decoupling of firms. Former research showed that gender influences CSR performance, but not whether gender has an effect on environmental decoupling. The paper of Peni & Vähämaa (2010) suggest that female CFO’s are more conservative, thus we expect this for female CEOs as well. Therefore, we argue that female CEOs are more conservative in voluntary environmental disclosure, and thus engage less in environmental decoupling. Furthermore, if firms engage in environmental decoupling they take risks for possible damage to firms reputation and performance. As women are more risk averse than men, we expect that female CEOs

will reduce environmental decoupling of firms.

H1b: Environmental decoupling at firms reduces by a female CEO on the board.

It is known that long tenured CEOs rely on routines and habits and grow in ‘’stale in the saddle’’ position which make it difficult to make changes in a firm. Most often these CEOs fear to lose money or reputation (Hambrick 2007). Newly appointed executives have been shown to experiment more and take more innovative decisions, while long tenured CEOs are more resistant to experiment and innovative changes in the strategy. Almost all major actions taken by CEOs seems to take place in the first who and a half year of their tenure (Lewis et al, 2010). It is clear that CEO tenure affects firm performance. The results of Henderson et Al (2006) show an inverted relationship between CEO tenure and profitability which means that the effectiveness of a new CEO increases over the first years of his tenure but will decrease after a longer period of time. But other research finds the opposite, Musteen et al (2010) finds that a longer tenure of CEOs appears to be associated with a more active approach regarding to innovation.

Lewis et al, (2014) find that newly appointed CEOs are more willing to voluntarily disclose environmental information, but whether they use it to engage in environmental decoupling remains unclear. We argue that a newly appointed CEO has to gain his reputation at the beginning of his tenure. The CEO can gain a good reputation with his stakeholders in the beginning period of his tenure by disclosing more environmental

(17)

17

information. Although, he has got more incentives to talk more than the actual walk, and thus to engage in greenwashing activities. A long tenured CEO has gained his reputation already and therefore has less incentives to engage in environmental decoupling. Thus, we argue that a short tenured CEO does engage more in environmental decoupling compared to long tenured CEOs. H1c: Environmental decoupling at firms increase by a lower tenure of the CEO.

Prior research suggest that CEO characteristics do have a significant effect on firm’s performance, but not much research is done whether the education of a CEO does affect environmental decoupling of firms. Koyuncu et al (2010) examined the role of the educational background of a CEO on firm performance. This research found some significant differences between different sorts of education. They showed that CEOs with an engineering educational background had better firm performances compared to CEOs with a finance educational background. The paper of Manner (2010) draws on the upper echelon theory of Hambrick and Mason (1984) which suggest that education contains a lot of information about an individual. Many universities around the world introduced CSR education to make students more aware of social development. But there is no clear evidence that this CSR education changes the behaviors of students in their working places. There are studies which show that environmental education does not increase the environmental awareness of students (Huang, 2013). Shabab et al. (2020) found that CEOs with financial expertise in China is positively related with increased sustainable performance and increased environmental reporting, this suggest that financial expertise won’t affect environmental decoupling if both sustainable performance and environmental reporting increases. Former research shows differences in firm performances and educational backgrounds of CEOs. The role of a financial degree on environmental decoupling remains under investigated.

Therefore, we argue that CEOs with a financial degree are more likely to engage in environmental decoupling than CEOs without a financial degree. We expect that CEOs with a financial degree perceive voluntary disclosure as a chance to improve financial performance, and thus see more chances to greenwash. H1d: Environmental decoupling at firms increase with the presence of a financial degree of the CEO.

(18)

18

2.4.2 Board independence.

Scandals in the recent decade and loss of moral values of firms lead to questions around the governance structure of a variety of firms. Monitoring needs and disclosure of CSR performance is more questioned than ever. The role of an independent board should increase voluntary CSR disclosures and their reliability (Dah & Jizi, 2018). For instance, Peasnell et al (2005) found that board independence reduces earnings management practices by reducing its abnormal accruals. It is widely accepted that independent directors are better able to control and monitor the management, which is in line with the agency theory that an independent board can control and monitor the actions of the agent more effectively. An independent board symbolizes higher transparency (Hussain et al, 2018). The board should include external auditors who control and monitor the disclosure of environmental performance, so no individual or small group of directors are able to engage in environmental decoupling. Independent directors holding a non-official position in the organization and therefore provide a more objective view on firm performance. Furthermore, CEOs can express less power over independent directors compared to dependent internal directors. As a result, the independent board should have less incentives to collude with the internal directors (Liao et al, 2015).

Fama & Jensen (1983) hypothesis that independent directors have incentives to develop a reputation. The own human capital value of the CEOs relies on their performance as internal decision manager in other firms. They are outside directors at other firms to signal that they are decision experts and that they understand the importance of separate control within a company. This suggest that independent directors will look more critical to the performance and disclosure of internal management on environmental reporting. Therefore, we hypothesize that board independence has a negative moderating effect on environmental decoupling. H2: Board independence has a negative effect on the level of environmental decoupling.

(19)

19

3. Methodology

This chapter elaborates on the research method used in this research. We elaborate on the sample that is being used. After that we explain the dependent variable of environmental decoupling, the different independent variables, the control variables, and the moderator.

3.1 Data collection and sample selection

To test the proposed hypothesis, we extracted data from multiple databases to measure environmental decoupling and multiple CEO characteristics. We matched and merged data from 3 different datasets. First, we use the data of Bloomberg which provides ESG disclosure data from different publicly listed companies around the world. Second, Thomson Reuters Asset4 database which provides ESG performance data from different companies from all over the world. Third, the data of CEO characteristics will be provided by BoardEX. We collected a sample of 16.880 firm-year observations from 4.404 organizations from these databases, between 2006 and 2017.

Because we focus on environmental decoupling in this research, we will focus on the environmental aspects of the ESG scores. The Bloomberg data provides the ESG disclosure scores of firms. Bloomberg transforms the ESG disclosures of firms into scores from 0,1 to 100 and gives an overall ESG score. This score is based on 3 components: environmental, social, and governance (ESG) disclosure scores. In this research we use the environmental scores. The higher the environmental score, the more environmental information is disclosed by the company. The Bloomberg data is available between 2006 and 2017, therefore we use this time range in this research. Thomson Reuters Asset4 provides a database with the overall ESG performance score of firms. These overall scores are based on 4 individual pillars, namely the corporate governance score, economic score, environmental score, and social score. We subtracted only the environmental performance score of the Thomson Reuters Asset4 database. The environmental pillar is based on a lot of different components of firms such as the use of resources, emissions of companies, and environmental innovations. The environmental performance scores differ from 0,1 to 100, which are the same

(20)

20

variances in scores as the Bloomberg database for environmental disclosure. (ResearchFinancial.wordpress.com, 2018)

And last, BoardEX provided the data regarding certain CEO characteristics. BoardEX is a database with biographical data of board members of different sorts of organizations. We extracted data from different CEOs of companies around the world, to extract the CEO characteristics of age, gender, tenure, and education to use in this research and to merge them with our environmental decoupling data.

3.2 Variables

3.2.1 Environmental decoupling

Environmental decoupling is the dependent variable in this research. We use the measurement approach of Tashman et al (2019), which measures the decoupling as the absolute difference between ESG disclosure scores and ESG performance scores. Therefore, we will measure environmental decoupling as the absolute difference between environmental disclosure scores and actual environmental performance scores from the Bloomberg database and the Thomson Reuters Asset4 database. The bigger the negative score between the environmental disclosure score and the environmental performance score, the bigger the degree of greenwashing is. In this case, a company talks more than it actually does. With a positive gap, the walk is bigger than the talk and refers to brownwashing.

3.2.2 Independent variable: CEO characteristics

The independent variables in this research are the CEO characteristics age, gender, tenure, and education. This research builds on the upper echelon theory of Hambrick and Mason (1984) which describe characteristics of TMT members do influence firm performance, we want to investigate whether the characteristics of CEOs affects environmental decoupling. The data of these characteristics are extracted from the BoardEX database which provides datasets of people profiles with biographical information of TMT managers.

We made some changes in the lay out of the data from BoardEX. First, we filtered on the board roles, thus we only selected CEOs. The age of the CEO is the age at the registered report year. For gender we created a dummy variable, ‘’0’’ is male and ‘’1’’ is female. Tenure is measured with the use of a logarithm and the number of

(21)

21

working years that a CEO has worked for an organization. To measure CEO tenure, we used the natural logarithm of the number of years of the tenure of the CEO. If the tenure of a CEO is 10 years or longer, then the value is equal to or higher than ‘’1’’. When the tenure of the CEO is 10 years or shorter, then the value is equal to or lower than ‘’1’’. At last, for the characteristic of education we created a dummy variable for CEOs with a financial degree as ‘’1’’, and without a financial degree as ‘’0’’.

3.2.3 Board independence

Board independence could have a significant impact on environmental activities of firms. An independent board can control and monitor the actions of the agents and would take action when the agent engages in environmental decoupling (Hussain et al, 2018). A more independent board is associated with higher levels of social disclosures, and transparency levels. Board independence does not improve CSR disclosures, but it also positively influences CSR performance of firms (Dah & Jizi, 2018). Therefore, board independence is a variable that could affect the relationship between a CEO and environmental decoupling. Board independence is measured with the percentage of independent board members on the board. Thus, the board independence scores do vary between 0 and 100.

3.2.4 Control variables

There are other factors besides the CEO characteristics which can affect the level of environmental decoupling. These are included as control variables to avoid biased results. First, multiple studies argue that firm size and firm’s profitability affect CSR activities and thus can be associated with environmental decoupling (Kang, 2017). Tashman et al (2019) describe that larger firms face more public scrutiny over their environmental impacts and affect environmental performance. Second, firm’s profitability could affect the results, therefore we control for the effect of profitability with the ratio of return on assets (ROA). Third, we control for capital intensity, the results of Hussain et al (2018) suggest that capital intensity affects social performance and it may affect environmental decoupling as well. Capital intensity is measured as the ratio of the assets in place to total sales. Fourth, Padgett & Galan (2010) show that manufacturing firms with a higher R&D intensity invest more in CSR. Thus, we control for the R&D intensity of firms. The R&D intensity is measured as the ratio of R&D expenses to total sales. Fifth, we control for the presence of a BIG4 as an auditor. We

(22)

22

measure the presence of a BIG4 with a dummy variable of ‘’1’’, and ‘’0’’ if there is no BIG4 presence. We argue that a BIG4 company delivers a different level of quality of disclosure and thus could affect environmental decoupling. Sixth, we control for the leverage of firms. The leverage is measured as the ratio of total debt to total assets. Leverage may impact the risk-taking level of a firm and may influence the risk taken in CSR decoupling. Seventh, the size of the board could affect CSR activities. Hussain et al (2018) state that different studies show fragmented results between board size and sustainability activities of firms. Therefore, we control for board size. Eight, we control for CEO duality which means that the CEO holds also the position as the chairperson on the board, this may lead to high influence power of the CEO and may result in weak monitoring. The more power the more the CEO can influence the implementation of certain strategies. Thus, we control for CEO duality as it may affect CSR decoupling (Hussain et al, 2018). Last, we control for CSR related compensation within a company. When a CEO receives a CSR related compensation this could create incentives to engage in CSR decoupling.

Additionally, we controlled for industry and country effects in the model. As incentives for environmental decoupling can differ between industries (Halme & Huse, 1995), we expect that industry and country do have an effect on environmental decoupling. Furthermore, we controlled for fixed year effects in the random effects model.

4. Results

4.1 Descriptive statistics

Table 2 provides the descriptive statistics. By merging the data from BoardEX, Bloomberg, and Thomson Reuters Asset4 we have a final sample of 16.880 firm-year observations from 4.404 organizations. Environmental decoupling is measured as environmental performance from Asset4 minus the environmental disclosure from Bloomberg. The scores of environmental decoupling differ from -82,17(which indicates greenwashing) to 98,51(which indicates brownwashing), with a mean of 0,82. The mean of the variable age is 56,12, which means the overall CEO is quite old. The age varies a lot between an age of 27 and 87. We created a dummy variable for gender with males indicated as ‘’0’’ and females as ‘’1’’, the majority of the sample consists

(23)

23

out of male CEOs and only 610 female CEOs. The low amount of diversity in gender could lead to insignificant results. This is to a lesser extent also the case with the sample of financial education of CEOs, the low mean of 0,22 shows that the majority of the CEOs in the sample doesn’t have a financial degree. The overall board independence in the sample is high with a mean of 68,65 percent.

Table 2: Descriptive statistics

Variable observations Mean Min Max

Environmental Decoupling 16.880 .824492 -82.17 98.51 Age 16.880 56.12405 27 87 Gender 16.880 .0361374 0 1 LNTenure 16.880 .7306109 .0413927 1.700704 Fin Education 16.880 .2155213 0 1 Board Independence 16.880 68.65026 0 100 Size 16.880 15.67814 0 27.81105 ROA 16.880 4.937898 -417.73 188.19 Capital intensity 16.880 37.90797 0 182915.9 R&D Intensity 16.880 .01095575 0 13.55593 BIG4 16.880 .8274289 0 1 Leverage 16.880 25.20406 0 378.15 Board Size 16.880 10.3465 1 35 CEO duality 16.880 .5022512 0 1 CSR related compensation 16.880 .2805095 0 1

N = 16.880 firm-year observations of 4.404 organizations

Table 3 presents the correlation of the different variables. Out of the different independent variables, we find positive significant correlations with tenure and financial education in correlation with environmental decoupling. The outcomes of age and gender are insignificant. The moderating effect of Board independence is significant. Out of the control variables we see positive significant correlations with firm size and the presence of a BIG4 company in relation to environmental decoupling. ROA, board size, CEO duality, and CSR related compensations showing a negative significant correlation with environmental decoupling. Capital intensity, R&D intensity, and leverage do not show any significant relationship with environmental decoupling.

(24)

Table 3: Correlation matrix 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1.Env. decoupling 1.0000 2.Age 0.0049 (0.5258) 1.0000 3.Gender 0.0063 (0.4136) -0.0573*** (0.0000) 1.0000 4.LNTenure 0.2520*** (0.0011) 0.2092*** (0.0000) -0.0252*** (0.0010) 1.0000 5.Fin. Education 0.0383*** (0.0000) 0.0143* (0.0631) -0.0081 (0.2938) -0.0398*** (0.0000) 1.0000 6.Board Independence 0.1294*** (0.0000) 0.0495*** (0.0000) -0.0058 (0.4502) -0.0016 (0.8309) 0.2315*** (0.0000) 1.0000 7.Size 0.3276*** (0.0000) 0.0146* (0.0584) -0.0018 (0.8110) -0.0141* (0.0666) 0.0179** (0.0198) 0.0359*** (0.0000) 1.0000 8.ROA -0.0365*** (0.0000) -0.0050 (0.5154) -0.0110 (0.1514) 0.0458*** (0.0000) 0.0235*** (0.0023) -0.0276*** (0.0003) -0.0056 (0.4680) 1.0000 9.Capital intensity 0.0026 (0.7360) -0.0075 (0.3286) -0.0024 (0.7542) 0.0096 (0.2119) -0.0052 (0.4999) -0.0164** (0.0330) 0.0006 (0.9395) -0.0300*** (0.0001) 1.0000 10.RD intensity -0.0064 (0.4045) -0.0419*** (0.0000) -0.0046 (0.5476) 0.0072 (0.3505) -0.0490*** (0.0000) -0.1013*** (0.0000) -0.0283*** (0.0002) -0.0384*** (0.0000) 0.0665*** (0.0000) 1.0000 11.BIG4 0.1181*** (0.0000) -0.0036 (0.6393) 0.0002 (0.9766) -0.0039 (0.6138) 0.0114 (0.1397) 0.0131* (0.0889) 0.0914*** (0.0000) 0.0074 (0.3356) 0.0012 (0.8742) -0.0120 (0.1189) 1.0000 12.Leverage 0.0018 (0.8141) -0.0078 (0.3122) 0.0001 (0.9946) -0.0036 (0.6355) 0.0636*** (0.0000) 0.0090 (0.2406) 0.0274*** (0.0004) 0.0062 (0.4241) 0.0025 (0.7480) 0.0125 (0.1034) 0.0191** (0.0130) 1.0000 13.Board size -0.1917*** (0.0000) 0.0736*** (0.0000) -0.0102 (0.1871) -0.0187** (0.0153) 0.0231*** (0.0027) -0.1651*** (0.0000) -0.0039 (0.6156) 0.0313*** (0.0000) -0.0330*** (0.0000) -0.0531*** (0.0000) 0.0007 (0.9315) -0.0003 (0.9731) 1.0000 14.CEO duality -0.0677*** (0.0000) -0.1793*** (0.0000) 0.0442*** (0.0000) -0.0490*** (0.0000) -0.1015*** (0.0000) -0.1364*** (0.0000) -0.0091 (0.2375) -0.0372*** (0.0000) 0.0147* (0.0557) 0.0180** (0.0194) 0.0126 (0.1027) -0.0154** (0.0448) -0.0431*** (0.0000) 1.0000 15.CSR comp. -0.1436*** (0.0000) 0.0002 (0.9836) 0.0494*** (0.0008) -0.0281*** (0.0003) 0.0778*** (0.0000) 0.1482*** (0.0000) 0.0030 (0.6965) -0.0038 (0.6182) -0.0017 (0.8232) -0.0114 (0.1391) 0.102 (0.1867) 0.0257*** (0.0008) 0.0347*** (0.0000) 0.0340*** (0.0000) 1.0000 Statistically significance: *** P < 0.01, ** P < 0.05, P* < 0.1

(25)

4.2 Regression analyses

Hypothesis 1(1a, 1b, 1c, and 1d) was about the influence of CEO characteristics on environmental decoupling. The four characteristics of these hypotheses are respectively age, gender, tenure, and financial education. The results of the random effects model regression are shown in table 4. Model 1 shows the relationship between environmental decoupling and the different control variables. Out of the control variables we see significant relationships with firm size, profitability(ROA), the presence of a BIG4 company, board size, CEO duality, and CSR compensation in relation to environmental decoupling.

Model 2 in table 4 shows the relationship between environmental decoupling and the different independent variables of the four CEO characteristics. The independent variables of age and financial education shows significant outcomes. The characteristic of age is negatively significant, this result indicates that age has a decreasing effect on environmental decoupling. This is the opposite of the expected outcome in hypothesis 1a. The characteristic of financial education shows a positive significant relationship, which supports hypothesis 1d. This indicates that a financial education degree of the CEO increases the gap between environmental performance and environmental disclosure. This finding supports hypothesis 1d. Unexpectedly, we did not find any support for the outcomes of gender and LNTenure. Thus, Hypothesis 1b and 1c are rejected.

Model 3 in table 4 represents the results of hypothesis 2, which predicts the moderating effect of board independence on the relationship between the CEO characteristics and environmental decoupling. The interactions for age and financial education are again significant. In addition to that, the tenure of the CEO is significantly related to environmental decoupling with the interaction effect of board independence. The model shows again significant outcomes with the relationships of age and financial education. This supports hypothesis 2.

(26)

26

Table 4: Random effects model Environmental decoupling Model (1) (2) (3) Age -.1011739** (0.025) -.0884583* (0.052) Gender 1.516962 (0.384) 1.440484 (0.405) LNTenure 1.015298 (0.132) 1.019674 (0.130) Fin Education 3.179869*** (0.000) 2.598809*** (0.004)

Age * Board Independence .0052301***

(0.003)

Gender * Board Independence .0163439

(0.835)

LNTenure * Board Independence -.0615248**

(0.045)

Fin Education * Board Independence -.00959666**

(0.022) Size 3.990462*** (0.000) 3.984845*** (0.000) 3.971103*** (0.000) ROA -.0780253*** (0.000) -.0790036*** (0.000) -.0777445*** (0.000) Capital intensity .0000128 (0.923) .0000113 (0.932) .0000264 (0.843) R&D Intensity .1545233 (0.838) .1737103 (0.818) .3931438 (0.602) BIG4 5.807456*** (0.000) 5.817902*** (0.000) 5.811162*** (0.000) Leverage -.0036512 (0.797) -.0067336 (0.635) -.0066531 (0.637) Board Size -1.571871*** (0.000) -1.557203*** (0.000) -1.444227*** (0.000) CEO duality -2.892217*** (0.000) -2.873957*** (0.000) -2.516748*** (0.000) CSR compensation 5.359945*** (0.000) 5.517354*** (0.000) 6.131893*** (0.000) Constant -30.02156 (0.000) -25.94122 (0.000) -26.54542 (0.000)

Industry control Yes Yes Yes

Country control Yes Yes Yes

R2 _0.2020 _0.2042 _0.2114

Observations 16.880 16.880 16.880

(27)

27

4.3 Robustness check

For the robustness checks we will run a probit regression on the sample. We created dummy variables for greenwashing and brownwashing. We constructed the dummy variables in the same manner as the paper of Testa et al (2018). A dummy variable of ‘’1’’ for the bottom 25th _{percentile, and a variable of ‘’0’’ for the other values.}

In this manner, we measure the lowest 25% of the environmental decoupling scores of the sample, which refers to greenwashing. The greenwashing scores in this percentile variate between -18,125 and -82,17. We did the same for the highest 75th_{percentile to}

measure brownwashing. We created a dummy variable of ‘’1’’ for the scores in the highest percentile, and a ‘’0’’ for the rest of the scores. In this manner, we measure the highest 25% of the environmental decoupling scores of the sample, which refers to brownwashing. The brownwashing scores in this percentile variate between 13,44 and 98,51.

Brownwashing is part of environmental decoupling, the other part consists out of greenwashing. Therefore, we did the robustness checks on both brownwashing and greenwashing. Table 5 presents the greenwashing outcomes of the probit regression on the lowest environmental decoupling scores, which are the biggest outliers of greenwashing scores. We see a significant relationship between age and tenure and the dummy for greenwashing. Furthermore, the table presents the brownwashing outcomes on the highest environmental decoupling scores, which are the biggest outliers of brownwashing scores. We only see a significant relationship between financial education and the dummy for brownwashing scores. What stands out in these results are the differences in significant outcomes between greenwashing and brownwashing.

(28)

28

Table 5: robustness checks greenwashing and brownwashing

Greenwashing Brownwashing Age .005479*** (0.001) -.0007094 (0.669) Gender -.0575445 (0.347) -.1076572* (0.074) LNTenure -.1613824*** (0.000) -.0531207 (0.132) Fin Education -.0555241* (0.057) -.00638429** (0.024) Board Independence -.0063346*** (0.000) -.0043974*** (0.000) Size -.0733999*** (0.000) -.1761295*** (0.000) ROA .0058142*** (0.000) .001606* (0.069) Capital intensity -.0003857** (0.029) -.0000056 (0.653) R&D Intensity .0243038 (0.545) .0348037 (0.364) BIG4 -1576448*** (0.000) -5525329*** (0.000) Leverage -.0002604 (0.678) -.0001296 (0.820) Board Size .1004369*** (0.000) .0251857*** (0.000) CEO duality 0.1306286*** (0.000) 0.1388527*** (0.000) CSR compensation .4903025*** (0.000) .1762003*** (0.000) Constant .3199954*** (0.000) 1.680106*** (0.000)

Industry control Yes Yes

R2 _0,1246 _0,1420

Observations 16.880 16.880

Statistically significance: *** P < 0.01, ** P < 0.05, P* < 0.1

5. Discussion

This study examined the impact of CEO characteristics on environmental decoupling. By building on the upper echelon theory and the agency theory it was expected that the characteristics of a CEO impact environmental decoupling. Therefore, the first hypothesis was about the relationship between environmental decoupling and the CEO characteristics of age, gender, tenure, and financial education. We expected that these

(29)

29

characteristics would influence environmental decoupling, but against this expectation we did not found evidence for every single characteristic.

It was expected that CEOs age (H1a) would influence environmental decoupling, as we expected that younger CEOs are more engaged with environmental performance and therefore less likely to engage in environmental decoupling. Surprisingly the results showed the opposite and suggest that elder CEOs do engage less in environmental decoupling compared to younger CEOs. Thus, the results show evidence that younger CEOs do engage more in environmental decoupling. This research adds empirical evidence that age is associated with higher levels of moral reasoning and as a result lower levels of environmental decoupling. In hypothesis 1b we suggested that women play an important role in CSR activities, therefore it was expected that CEOs gender would affect environmental decoupling. We expected a decreasing level of environmental decoupling with a female CEO, as females are overall more conservative and more risk averse. We found no evidence in this study that gender is related to environmental decoupling, and thus that women affect environmental decoupling. There is an important side note to make regarding the sample. Only 3,6% of the whole sample are women. This small amount could have led to insignificant results, further research is necessary to investigate whether the role of women do impact environmental decoupling or not. Furthermore, in hypotheses 1c we expected that CEO tenure would affect environmental decoupling, but against this expectation we found no significant evidence in the relationship between environmental decoupling and CEO tenure. The last hypothesis was whether the presence of a financial degree (H1d) affects environmental decoupling. We found significant evidence that financial education does have an increasing effect on the level of environmental decoupling. There is not much research done on the effects of a financial degree on CSR related performances. Thus, this research provides only empirical evidence, but the strong significant results do suggest that the presence of a financial degree does lead to more environmental decoupling. CEOs could perceive the environmental voluntary disclosure as an opportunity to improve environmental performance. Further research is needed to get more evidence on the under investigated role of the presence of a financial degree. These outcomes could be used

(30)

30

for controlling and monitoring purposes. Firms and external parties could control and monitor more on environmental decoupling when the CEO has a financial degree.

The second hypothesis tested the moderating effect of board independence on the relationship of CEO characteristics and the level of environmental decoupling. We found strong significant evidence that the moderating effect of board independence is related to environmental decoupling. The results showed again significant relationships for age and financial education. In addition to that, the results showed a significant effect of CEO tenure on environmental decoupling. Thus, we conclude that board independence serves its controlling and monitoring role by the decrease in the level of environmental decoupling. This is in line with the arguments of other papers such as Fadli et al (2009) which state that board independence has a positive effect on the level of CSR reporting. Adding to that these findings are in line with the agency theory and strengthens earlier research that an independent board is able to control and monitor the actions of the CEO.

Environmental decoupling is the gap between environmental performance and environmental disclosure. Environmental decoupling could be divided into two sorts, namely greenwashing or brownwashing. In this manner you can see which characteristics have influence on both greenwashing and brownwashing separately. Brownwashing is a relatively new theoretical subject in research and there is not much research done yet. Therefore, this additional research could be a point of departure for future research on brownwashing. Earlier studies conducted research especially on financial drivers of brownwashing. This study focuses on the non-financial drivers of CEO characteristics.

By analyzing the robustness checks on greenwashing, we found a significant relationship with CEO age, tenure, and financial education. In relation to brownwashing we found significant evidence for CEO gender and financial education. Surprisingly both have different relationships with different CEO characteristics. This suggest that both greenwashing and brownwashing are two different subjects and should be investigated separately in future research. Although, as it is a robustness check, these results provide only empirical evidence and need further research. The way of measurement could have led to distortions in the results because only the biggest outliers of greenwashing and brownwashing were measured in this analysis and the

(31)

31

smaller scores were excluded from the sample. Again, further research in greenwashing and brownwashing is needed to clarify these results. The results of the robustness checks suggest that there are different relationships between CEO characteristics and greenwashing or brownwashing. This paper could function as a good starting point for future research on brownwashing, as there is not much research done yet. Adding to that, these results suggest that research on environmental decoupling needs to be split into greenwashing and brownwashing studies, as the results suggest that both are affected by a different set of CEO characteristics.

6. Conclusion

6.1 Conclusion and contribution

This study provides new insights into the influence of CEO characteristics on environmental decoupling. This study examined whether CEO characteristics would influence environmental decoupling by building on the agency theory and the upper echelon theory. The agency theory builds on the idea that information asymmetry arises due to diverging interests between managers and owners. The upper echelon theory describe that the organization is a reflection of its management, and therefore a CEO is likely to impact (environmental) performances and disclosures. This research made several contributions to the existing literature. First, the relationship between CEO characteristics on environmental decoupling is under investigated. Earlier research focussed more on overall CSR topics, but not solely on the environmental aspects of CSR. Second, this research contributes to corporate governance literature by adding the moderating effect of board independence. The results show that board independence serves its role in controlling and monitoring CEOs.

This study included a sample of 16.880 firm-year observations from 4.404 organizations with environmental performance and disclosure scores, and databases with CEO characteristics. We conducted a random effects regression and found significant evidence that CEOs age and a financial education degree affect environmental decoupling. A younger CEO increases the gap between environmental performance and environmental disclosure. This suggest that younger CEOs take more risks to achieve better disclosed results and take more unethical choices compared to older CEOs. This result suggests that elder CEOs operate at higher levels of moral reasoning. Furthermore, we found a positive significant effect on the