GREEN MANAGERS AND EXTERNALLY ORIENTED
MANAGERS AS NON-EXECUTIVE BOARD MEMBERS:
THE ROLE OF LEGITIMACY
by
Bernard Sybren Harm (Ben) Wijnstok S2364719
University of Groningen Faculty of Economics and Business
Strategic Innovation Management
Supervisor: prof. dr. J. Surroca Co-assessor: dr. J. Dong
Groningen, January 2018
2
ABSTRACT
In this research, it is hypothesized that U.S. firms with above average environmental controversies in their industry causing illegitimacy, are likely to appoint a specific manager. The objective of this research is to find if a particular green and externally oriented manager is preferred in boards of illegitimate firms.
A binary logistic regression analysis was conducted to test the hypotheses, followed by a logit regression analysis to check the robustness. Findings suggest if a firm is illegitimate, it is more likely to appoint a green manager as non-executive board member. Furthermore, if firm is illegitimate, it is more likely to appoint an externally oriented manager as board member.
This research contributes to the unexplored relationship between firms with above average environmental controversies in their industry and the appointment of CEOs for a director function as non-executive board members. The findings will increase understanding of the effects of legitimacy on the board appointment strategy of firms.
3
TABLE OF CONTENT
ABSTRACT ... 2
INTRODUCTION ... 4
THEORY AND HYPOTHESES DEVELOPMENT ... 8
Towards a legitimacy crisis in the firm ... 8
Legitimacy crisis and green managers as non-executive directors ... 11
Externally versus internally orientation ... 14
METHODOLOGY ... 16 Data collection ... 16 Sample ... 19 Measures ... 19 Analysis ... 20 RESULTS ... 21
Descriptive statistics and correlations ... 21
Regression results ... 22
Robustness ... 27
DISCUSSION ... 30
Theoretical implications ... 30
Practical implications ... 31
Limitations and further research ... 32
CONCLUSION ... 33
REFERENCES ... 34
4
INTRODUCTION
Business leaders of today are increasingly expected to engage in corporate social responsibility (CSR) next to managing a profitable organization. McWilliams & Siegel (2001) defines CSR as the actions beyond requirements by law and the interests of the organization. Thus, firms voluntarily integrate social and environmental concerns into the business. In recent years the concept of CSR is receiving tremendous attention both from academics and business managers that invest their time and effort (Moon, 2007). According to the UN Global Compact-Accenture Strategy CEO Study (2016), a study about the development of motivations to engage in environmental, social and governance issues, 89 percent of more than 1.000 interviewed CEOs believe that commitment to sustainability is translating into impact in their industries. These days, green managers engaged in social responsibility and sustainable business are more desired than ever. This study investigates why certain firms hire these managers.
Legitimacy can be achieved through CSR practices, by activities that reduce negative consequences of non-compliance to CSR and by practices that will maximize the benefits of legitimacy from positive evaluations (Zheng, Luo, & Maksimov, 2015). Suchman (1995) stated that legitimacy is “a generalized
perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, definitions.” (p.869). This implies that the harmony
between social values that is associated with the actions of the firm and what kind of behavior is acceptable in the society, determines if the firm is legitimate. In fact, legitimacy is a social license that allow firms to operate (Baba & Raufflet, 2014). Moreover, CSR is used as a mean to maintain legitimacy with the firms’ audiences. This legitimacy is considered by the institutional theory as the result of a social construction. Within the institutional theory, legitimacy is created when firms are able to conform to social expectations (Deephouse & Carter, 2005). This depends on the preservation of reciprocal relationship with stakeholders (Maignan & Ralston, 2002). When stakeholders make a comparison between norms of social acceptable behavior and the perceived activities of firms, expectations are created. If the firm meets these expectations it can result in social acceptance. Thus, firms are motivated to engage in CSR to protect the legitimacy (Duff, 2017; Georgallis, 2017).
5
this illegitimacy. When a firm does not have the capabilities for enhancing legitimacy, it must either create or purchase them in order to survive. In such a crisis a firm may go for purchasing, since creation would take too long. This time saving sustainable innovation of purchasing environmental CSR capabilities from externals that have a possible solution for gaining the social acceptance, especially in local markets, may be a perfect strategy.
By making use of networks, the firm can scout the managers with the needed knowledge, capabilities or legitimacy externally. These collaborative networks are important for the firm, since one of the most influencing entities is the board of directors (BoD). The BoD that can acquire resources and capabilities including the first step towards legitimacy after being involved in controversies. Boards can find these survival capabilities and legitimacy with a non-executive director (Scott, 1995). That is, non-executive directors can bring legitimacy to the board (Pfeffer & Salancik, 2003). According to a survey from The Spencer Stuart Board Index (2016) executives from public listed companies are desired candidates for firms searching for new independent non-executive directors. The percentage of CEOs of S&P 500 firms that serve in one or more boards next to the own board is the same as 2015, 43% in 2016 (Spencer Stuart Board Index, 2016). On average for every two CEOs one serves on another board. According to Ferris, Jagannathan & Pritchard (2003) there is no theoretical limit on the number of directorships held by an individual. Interesting is to see how firms appoint non-executive directors in the BoD when they face a CSR related illegitimacy and what influence capabilities or the prior experience of the non-executive director have on the firms’ choice of appointing. To elaborate on the board structure of firms, in public held firms directors are elected to join a board. The board consist of inside executive and outside non-executive directors. Inside directors have a specific function and are employed in the firm. The non-executive director is less involved in the firm and can have a different perspective, bring capabilities, or other experience to the board. They can also keep an eye on the inside directors and on the whole firm itself.
6
are assumed to be green1. The individual is responsible for environmental issues on a strategic level and defines, develops and integrates these issues into the fundaments of business. These green managers have built experience, capabilities or an individual reputation to the outside world. With this distinctiveness, in contrast to a non-green manager, a firm is perhaps more likely to appoint a green manager in the board of firms that needs legitimacy.
Next to studying if illegitimate firms are more likely to appoint green managers, it is interesting to investigate if a firm is likely to appoint a manager with specific capabilities, when the firm is illegitimate as result of a specific misdeed. Using the classification of Hawn & Ioannou (2016), these special managers can be identified that are preferred by the illegitimate firms. The firms with illegitimacy searching for a particular executive may appoint an individual who is externally oriented that can deal with illegitimacy. Another possibility is that a firm needs an internal oriented manager, who is inward looking. Managers can be placed in an internal or external way of acting, in order to investigate how the managers’ way of work is related to the appointing of these individuals. This could be desired, because of the specific situation in the firm. According to Hawn & Ioannou (2016) internal CSR orientation refers to actions that consist inward looking actions, where organizational capabilities are developed and the social actors are met. The same authors state that external actions are publicly oriented, visible initiatives and the aimed at gaining outside audiences’ legitimacy. Both CSR oriented managers could be asked to join other boards, when a firm needs a board member that has one of the two different orientations. This possible relation between the firms’ illegitimacy due to a specific problem in its industry and a managers CSR orientation may have an impact on the type of executives preferred by firms.
To summarize, in this research the illegitimacy of firms with above average misdeeds in their industry is the result of the misfit between the firm’s practices and the expectations of stakeholders. The board does not have the capabilities to deal with the problem itself. This board appoints non-executive directors with the desired capabilities to re-establish legitimacy. It is interesting to see what kind of executives, with what CSR backgrounds are likely to get appointed by firms to gain illegitimacy. This individual could have been chosen to join the board due to the efforts made in the current firm, where the individual is CEO. This green or non-green and internally or externally CSR orientation can have a different effect on the appointees ability to help firms by serving on the BoD. This research investigates if this is also the case with green non-executive directors, when the firm faces illegitimacy. Thus, the main effect in
1 Since current literature does not provide a solid definition of a green manager, this research uses a distinction
7
this research is, what impact illegitimacy has on the appointment of non-executive green board members and board members with an externally environmental CSR orientation. The central research question is:
Are firms involved in above average environmental controversies in their industry more likely to appoint green managers, instead of non-green managers and externally oriented managers, instead of
internally oriented managers as external board members?
8
THEORY AND HYPOTHESES DEVELOPMENT
The following section it is theorized how a firm can become illegitimate due to a misfit between the firms’ practices and the expectations of stakeholders, as a result of above average environmental controversies in the firms’ industry. The second section explains the effect of appointing non-executive directors. It also elaborates on green managers’ ability to establish legitimacy. This adds up to the first hypothesis. In the third section the appointment of specific green managers is theorized, followed by the second hypothesis.
Towards a legitimacy crisis in the firm
As being stated in the introduction, integration of social and environmental concerns, or CSR is a form of corporate self-regulation that is integrated into a business model, monitored and ensured by the firm for its compliance to law, ethical standards and national and international norms (Rasche, Morsing, & Moon, 2017). The firm voluntarily follows these CSR policy functions beyond the interests of the organization and what is minimally required by law (McWilliams & Siegel, 2001). The concept of CSR have included both internal and external stakeholders into its scope over the years (Capriotti & Moreno, 2007). According to Freeman (2010) “A stakeholder in an organization is any group or individual who
can affect or is affected by the achievements of the organization’s objectives.” (p. 46). The external
stakeholders are from outside the firm, like community partners, creditors, customers, government, shareholders, society and suppliers. The internal stakeholders are from inside the firm, who already committed to serve the firm, like employees, staff and volunteers. CSR policies can create expectations with these stakeholders, that play an important role on tolerance towards the firm (Dawkins & Lewis, 2003). When these expectations are not met by the firm, stakeholders may become intolerant.
9
Consequently, the stakeholders can make a significant impact when they do not conform with the firm. As in research of Roome & Wijen (2006) in the context of organizational learning, external stakeholders have power as problem setters, but not much as contributors for solving issues. This implies that firms must solve the problems as good as on their own. On the other hand, as being found in the same research, internal stakeholders have more impact when looking at a solution. They can have a direct influence on the firms’ behavior. This implies, they can help to solve the problem and by socially accepting and supporting a solution. Then, the expectations of these stakeholders and the actions of the firm are met again and legitimacy could be re-established. However, next to that they can give more input, the internal stakeholders are dependent on the firm, since they are employed by it. This could imply that the misfit will rather lie between the external stakeholders and the firm.
Since there is a constant interaction between the environment and firm, the legitimacy of the firm is continuously tested and redefined (Baum & Oliver, 1991). Thus, the legitimacy also depends on the environment and this context will judge the firm according to the different legitimacy standards operated (Kostova & Zaheer, 1999). When a single firm faces a bad event and there are no similar events at other firms in the same industry, the firm will perceive relatively more negative attention from the environment, because of its increased visibility towards the public. This reasoning where a bad event is novel or salient in the industry of a firm and is negatively reflected by stakeholders is known as the safety in numbers effect. It is based on social psychological research on attention and is used in research of Zavyalova et al. (2012). Thomson Reuters (2017) included controversies in their database of ESG data. The controversies are included if they are publicly known and thus can be checked. When a firm engage in environmental controversies, the public such as media and stakeholders will search for new information about the firm (Planalp & Fitness, 1999). They might find industry standards and compare them to the focal firm. Then, they will reshape their ideas about the firm (Taylor, 1991). This violation of expectations of stakeholders could eventually lead to illegitimacy of the particular firm. For illustration, from research of Fiske & Taylor (2013) “If a person is unpleasant, being a solo will cause
disproportionate condemnation” (p. 69) and other persons will react negatively. The same effect occurs
when generalizing it to industry level, the attraction of a disproportionate share of negative attention can have impact on the firms’ legitimacy. In contrast to this, the effect of relative more negative attention will decrease when other firms in the industry also engage in similar misdeeds, because the event will become less novel (Ahmadjian & Robinson, 2001).
10
towards legitimacy of a single firm. According to Brønn & Vidaver-Cohen (2009) the motives of a firm to engage in CSR differs per industry. This could also imply that the external expectations of the stakeholder towards a firm also differ depending on the industry the firm operates in. The stakeholders can perceive these violations as misdeeds, as Thomson Reuters (2017) uses these violations in their ESG database as controversies. These measures are benchmarked against firms in the same industry group.
The public expect firms to engage in social initiatives (Margolis & Walsh, 2003). Stakeholders are intolerant of those firms that break CSR standards (Dawkins & Lewis, 2003). Regardless if the expectations are breached by accident or on purpose by the firm, the stakeholders will have a negative attitude towards it. Perhaps small shareholder do not have the strength to pressurize the firm when they disagree, but organized stakeholders, like institutional investors or other organized pressure groups can have significant power towards the firm. As in research of Georgallis (2017) about the effect of social movements on firms, they state that these groups push firms to engage in social initiatives. This ideological pressure groups, that can have specific ideas on what role business should have, can target firms directly that have irresponsible behaviour. When these groups disagree, then the intolerance, or mismatch between practices of the firm and the expectations of powerful stakeholders escalate in the losing of legitimacy of the firm.
According to Bansal & Clelland (2004) legitimacy is crucial for survival of the organization. In other words, the intolerance of stakeholders, because of their unmet expectations and a firms’ actions may cause the illegitimacy for the firm, what can result in a threat to firm survival. As Claasen & Roloff (2012) state, the stakeholders’ comparison of the firms’ actions with what they assume the firm can do and what the firm actually does, is a challenge for legitimacy management. Legitimacy is emphasized by the widely accepted institutional theory. The institutional theory considers the processes where structures become established as guidelines for social behavior. These structures of norms, routines, rules and schemes will be seen as authoritative. A firm is expected to follow them. This continuous process of reaction to external expectations by adapting it, is often unconscious (Palazzo & Scherer, 2006). Furthermore, the components of the institutional theory are subjected to change. This means that what is legitimate now, could be illegitimate over time. Firms need to continuously interact with these changes.
11
in de U.S. (DeTienne & Lewis, 2005). The public did not accept to buy from Nike when it outsourced to unethical subcontractors. They saw it as Nike’s responsibility to know every detail of their products and business, and this production facility was obviously not monitored enough. This firm suddenly had a problem with its stakeholders. This illegitimacy not only threatened Nike’s operations, but also the survival of the entire firm.
According to the institutional theory, when firms conform to social expectations, legitimacy is created (Deephouse & Carter, 2005). This implies that when a firm is illegitimate, it can become legitimate again by satisfying the expectations of stakeholders. The illegitimate firms need to bring CSR capabilities in practice in order to meet these stakeholders. Moreover, in research of Duff (2017) and Georgallis (2017) both authors state that firms are motivated to engage in CSR in order to protect legitimacy. Since the solving of illegitimacy also depends on having the preservation of the reciprocal relationship between firm and stakeholders (Maignan & Ralston, 2002), illegitimacy should be solved through this channel. Thus, when a misfit occurs because the firm is involved in environmental controversies, the firm can implement environmental CSR related actions to gain organizational legitimacy.
Legitimacy crisis and green managers as non-executive directors
As being theorized in the previous section, a situation may occur where a firm becomes illegitimate. The efforts to gain back social acceptance for the firm after an expectation misfit, caused by an accident or misdeed could go beyond a firms own capabilities. Here, an external source of expertise, that may improve legitimacy could be preferred to deal with the problems on a strategic level. With approval of stakeholders, a firm can search the capabilities externally by new non-executive board appointments. According to Pfeffer & Salancik (2003), one of the benefits non-executive directors can bring to boards is legitimacy. Moreover, the survival of the firm partly depends on the legitimacy given by stakeholders, so the non-executive director can be the solution for the problem.
12
reasoning, the illegitimacy of a firm, caused by above average environmental controversies can be turned into legitimacy by integrating capabilities that are specially designed for dealing with these issues. The green CEO that possesses these environmental capabilities, can help a firm by implementing practices that meet the expectations of stakeholders. The cause of the illegitimacy might influence the choice to appoint the green CEO as non-executive director. The environment in which the firm operates could have an effect on appointing a non-executive director with a green orientation. This relationship can be investigated in the context of environmental controversies among industries. It would imply that a green manager as a non-executive director is a key for dealing with this illegitimacy.
A board searching for a new member due to illegitimacy may take notice of a green director. Note that the firm searches a non-executive director after the firm have become illegitimate. Boards can decide to ask this individual to join, because they recognize or appreciate the efforts this individual has made with CSR issues, or they want to amplify the board with this type of capability that is otherwise missing. On the other hand, one could argue that firms also have the possibility to copy other organizations with a desired corporate status in the form of isomorphism, that could have a low-cost impact on legitimacy. But the real impact of this approach will be difficult to foresee and that is not convenient during illegitimacy, where the firm has to put everything in order to solve it. This intangible asset could also be hard to copy. Either way, according to Staw & Epstein (2000), creating new practices by appointing non-executive directors is a difficult process and can even be seen as a form of innovation, where other firms in the field follow common trends. However, the impact of appointing non-executive directors with a proven track record related to CSR is better to foresee. These non-executive directors with their capabilities can bring the desired legitimacy.
13
2006). It also implies that non-green managers are less able to bring the needed legitimacy than green managers, since they do not have a proven track record in the form of experience or a personal reputation in being extendedly involved in environmental specific CSR. Since the firm can deal with illegitimacy by using CSR practices, it seems logical for the firm appoint green non-executive board members that know how to handle the problems. When firms appoint non-executive green board members they also show an act of good will to the stakeholders by offering a direct solution to the problem. Stakeholders may be sympathetic by this move and it probably lead to an improvement in legitimacy.
To summarize, when stakeholders’ expectation are not met, as a result of any environmental controversy where the firm is involved in that is above the average in its industry, it can result in organizational illegitimacy. The firm can appoint a green manager as non-executive board member, for the capabilities and experience, to deal with the illegitimacy. This individual is assumed to meet the demands of the stakeholder and gain legitimacy for the firm. Assuming that firms appoint green managers, since they are able to change the firms’ illegitimacy in legitimacy, it could imply that due to non-executive board appointments this supply and demand will meet. Thus, based in the argumentation presented above the assumption can be made that firms involved in above average environmental controversies in their industry are more likely to appoint green managers in the BoD as non-executive directors. This leads to the following first hypothesis:
14
Externally versus internally orientation
In the previous section it is theorized that firms involved in above industry average environmental controversies, leading to illegitimacy are likely to appoint a non-executive green manager in their board.
As being illustrated by using Zavyalova et al. (2012) controversies could cause illegitimacy for a single firm its industry. Moreover, the firm may response with a manager that fits the specific environmental issues faced by the firm. Thus, a non-executive director that can gain the desired legitimacy for the specific problem. Such a manager was identified in research of Hawn & Ioannou (2016), where they have made a theoretical distinguish between internally CSR and externally CSR orientation. A manager who is engaged in environmental CSR can belong to one of the two CSR orientated groups. This managers’ typical way of acting can be interesting for firms searching for a particular manager. Internal CSR consist of practices that are inward looking. It involves actions where organizational capabilities are developed and the expectations of social actors, that gives important resources where the firm depends on, are met (Hawn & Ioannou, 2016). Here, the internally focused actions are to achieve structural change. External actions consist of visible practices, public oriented initiatives and the gaining of legitimacy for the outside audiences (Hawn & Ioannou, 2016). Both orientations can be measured on the environmental level.
Firms that face illegitimacy may be interested and could have a preference for a new board member with a certain orientation of CSR. A manager with an external CSR orientation is assumed to be able to take care of the illegitimacy in the firm. This manager is acting in a more visible and public oriented way and have experience by means of a high involvement in externally visible activities. This individual has initiatives for the gaining of legitimacy from outside audiences. This individual can do this on environmental level, since the misdeeds are also environmental related. In terms of the stakeholder theory, the external stakeholders of the firm. In firms that are legitimate, an externally oriented CSR approach may not be so desired, since there is no need for trying to bend public opinion. The manager with an internal CSR orientation is assumed to have the ability to meet the expectations of social actors, by performing internal actions. The individual has experience by means of the efforts made in developing internally CSR. This individual can bring structural change within the organization. According to the stakeholder theory, this individual deals with the internal stakeholders. While the illegitimacy faced after a negative situation has occurred may need external CSR to deal with it. This manager is assumed to be able to deal with these issues more properly than the internal manager.
15
Furthermore the greatest part of the outside audiences consist of stakeholders outside the firm. For illustration, some firms particularly invest in more visible green projects, as in the research of Eichholtz, Kok, & Quigley (2009), oil companies rent environmental friendly office space, while other firms are investing in a less noticeable way. As being discussed previously, illegitimacy could be caused after regulations or standards are violated. So, to deal with the illegitimacy the firm can appoint a non-executive board member with an external orientation. This manager might make effort to change certain norms of acceptable behavior and the taken-for-granted assumptions about what is legitimate behavior within the industry group, or other strategies for gaining outsiders’ legitimacy, for the firms’ involvement in a misdeed. In this research the environmental controversies of firms are investigated. When a firm has a higher environmental controversies, in comparison with its industry peers, a manager that is external environmental oriented is likely to deal with this problem.
To summarize, when a controversy of a firm is above the average in its industry group, it will attract relatively more negative attention. The firm’s visibility increases disproportionately. When such a notorious event happens, a manager that is able to deal with this issue could be preferred. The external CSR orientation of Hawn & Ioannou (2016) will be specially important. External orientation is aimed at gaining legitimacy from outside audiences. This can help solve the relatively more negative attention of the firm. By aiming the visibility on other (positive) parts of the firm or by damage control. The external oriented environmental CSR manager is more able to deal with outside audiences and solve the disproportional effect due to the individuals’ proven track record, such as capability and experience. Now the assumption can be made that when the firm’s environmental controversies are higher than the average in their industry, the firm is assumed to appoint a non-executive director that is externally oriented. This individual is likely to solve the illegitimacy. Thus, firms involved in controversies, committed by the firm itself or by accident, that is above the industry group average are more likely to appoint a manager that is externally oriented with environmental CSR as board members. This relationship has not yet been studied empirically. This leads to the second hypothesis:
16
METHODOLOGY
In this section the methodological choices are described and justified. First, the type, number and collection of data and the sample are described. Second, the dependent, independent and the control variables used in this study are presented.
Data collection
In this research a quantitative approach was conducted, since it focuses on the results from data of a large number of constituents. In doing so, it summarizes the characteristics across the relationships and to recognize patterns statistical techniques were required. The CSR data was obtained via Datastream from Thomson Reuters ASSET4. The benefit of using ASSET4 is that all relevant internal corporate governance mechanisms are already included in the development of the data. ASSET4 can provide data about environmental, social and governance data (ESG) performance scores. It is primarily firm-level data from various sources, such as annual reports, media and the firm itself to verify quality and accuracy. Thomson Reuters ESG score based on reported data is calculated per firm per fiscal year (Thomson Reuters ESG Scores Structure, 2017). This source offers relevant, objective and systematic CSR data to professionals. ASSET4 also offers the CSR scores2 and ESG controversies scores. The analysts of Thomson Reuters gathered evaluation points that are publicly available per firm. Then this information is transformed into categories of environmental, social and governance. Now for this research, a quantitative analysis can be made. In order to systematically test the two hypothesis, the following steps were followed.
First, in this research the list of public limited firms of the United States, available in Datastream was used as the basic data. Only firms active in the years 2013, 2014 and 2015 were included. This was done for the one year lag measures of the possible illegitimacy in firms, the change in the board comparison and the fact that not all data was available yet when using later years. With including these years, all data can be checked if it meets minimum required basic data points.
Second, the environmental pillar score (ENVSCORE) variable was used to identify which CEOs of firms could be classified as green or non-green (and at the third step externally or internally oriented). This balanced rating is measured by the firms’ performance on the area of environmental from the ESG framework. With the environmental pillar score of 50 or higher on a scale of 0-100, the firm was labeled
1 for green firms and 0 for non-green firms. The score per firm is given by ASSET4, but the classification
of whether a firm is a truly green firm is calculated from 5 measures per firm. Since current literature does not provide a solid definition of a green manager, this research uses a distinction that can be
2 According to Farooq, Ullah, & Kimani (2015) the CSR score is: “A company's capacity to generate trust and
17
calculated. This is measured by taking the environmental performance score of firms and linking it to their CEOs. This assumption is made since the CSR orientation of the CEO reflects the firm, as being theorized in the previous chapter. A person is labeled CEO if the function title contains ‘CEO’ (e.g. interim CEO or CEO-Chairman). Thus, firms with a higher than 50 environmental pillar score are assumed to have a green CEO, whereas firms with a lower score are assumed to have non-green CEO.
Third, to find internally and externally oriented managers, the glossary of Thomson Reuters ASSET4 was used to manually categorize the 18 ESG indicator scores, values and data points. These points were available via DataStream and are part of the Thomson Reuters ASSET4. The same classification is used in research of Hawn & Ioannou (2016). This classification helps to test the second hypothesis, if the managers are externally or internally oriented. 11 points represent external CSR and 7 internal CSR. Within the ASSET4 Glossary, Yes and No values were labeled as 1 and 0, respectively. When values measured the same variable as a score, they were removed. Then, the sum of values of every point in the distinction of externally and internally environmental CSR was divided by the number of points in that category. Now, with calculating if the average of external or the average of internal has the overhand. The highest value is taken for the dominant orientation. A clear distinction between every two scores was visible, when there was data available for these categorization. When firms did not have any data on these categories, they were removed from the sample. The external oriented CSR firms were labeled as 0 and the internal oriented CSR firms were labeled as 1. Here, the same reasoning is applied that a firm reflects its CEO. This means that if a firm has a higher external score, the CEO of the firm is assumed to be externally oriented.
18
group3 by ASSET4. If a firm was illegitimate in 2013 and legitimate in 2014, it was labeled as legitimate, since the stakeholders’ expectations were eventually met in 2014. Legitimate firms in 2013 that were illegitmate in 2014 were labeled illegitimate, where the expectations were not met. When the firm was illegitimate, or legitimate in both years, the change in expectations was negative and positive, respectively. In practice, only checking the 2014 score was enough to make the same conclusions. This performance was relative to external expectations, so when the external expectations are negatively changed, a firm is in a crisis and thus illegitimate.
Fifth, the change in the composition of the board was measured between 2014 and 2015. A one year lag was used between CEOs and an eventual non-executive directorship. When the focus (t) is on one year, the measure of illegitimacy and changes in board comparison can be measured in the year (t) and following year (t + 1)4, respectively. This identifies if the CEOs is green or non-green and the eventual externally or internally orientation at the appointee’s firm in 2013 and if the CEO accepted a directorship at a target firm in 2014. Thus, the director in question has been CEO up to at least 2013 and has been appointed as non-executive director at the target firm upward of 2014. When the CEO of the firm also is a non-executive director at another firm the person is labeled 1. All other non-external directors (CEOs) are labeled 0. This measure of changes in boards is between 2014 and 2015, because the illegitimacy check is earlier, between 2013 and 2014. This is to make sure the CEO has nothing to do with the illegitimacy of the firm (between 2014 and 2015).
BoardEx is was used to find the board interlocks. In the BoardEx North American Director Network data purchased by the University of Groningen, the CEOs were found that have non-executive directorships. The CEOs that do not have an executive directorship are ignored, since they do not have added value in the sample. The ASSET4 firm names that were identified in the first step, were checked in BoardEx if they have a CEO that is also a director at another firm from within the sample. Then, to measure if a CEO entered a board of other firms, every individual board interlock was analyzed. The names and personal identification numbers of the CEOs from the labeled firms were included. This was done by manually picking every firm by name. When one CEO have multiple directorships at different firms, it is also counted as the number of positions filled. When two CEOs of one firm have non-executive directorships it is also separately counted (e.g. an interim CEO in the same year as the CEO). If one CEO have multiple directorships at one firm and the function title is the same (e.g. independent director) it is counted as one. This double count in BoardEx is probably caused by a change in the function title of the CEO.
3 This was done for the environmental score with the Thomson Reuters Business Classifications (TRBC)
Industry Group (Thomson Reuters, 2017).
19
Sample
U.S. public limited corporations based on the Thomson Reuters ASSET4 available firms were included in the sample. Only U.S. firms were included, because of the great amount of research, Orbis and ESG data available about firms from this country (Bureau van Dijk, 2018; Thomson Reuters, 2017). Also with identifying listed companies around the world that qualify to the requirements of the sample, then the work may be overwhelming and unnecessarily large. 248 of the 1203 firms were excluded after the minimum required variables for the CSR, illegitimacy and internally or externally orientation tests were checked. From this basic sample of 955 firms, 77 CEOs were found that have 87 non-executive board positions in one in the firms of this sample. From this point the final sample consisted of 87 different non-executive directorships.
Measures
In the section below the independent, dependent and several control variables are presented.
The independent variable (IV) is the above their industry average environmental controversy score of the appointing firm. This is measured by all data in the ASSET4 Glossary that contained an environmental controversy description. The variables are all benchmarked at data point level against industry and country. The dependent variable (DV) for the first hypothesis is the appointment of green managers as non-executive board member (ENVSCORE). This dummy variable is 1 if the board of the connected firm appoints a new green manager that is a green, otherwise 0. For the second hypothesis this variable is 1 if a firm appoints a manager as non-executive board member that is externally environmental CSR oriented, otherwise in case of an internally environmental oriented director 0. The dependent variable is a function of the independent variable; the controversy score of the firm described above and the control variables as being explained below.
Finally several control variables that influence the dependent variable will be included in this research. The set of independent control variables that can potentially influence a firms decision to appoint a certain non-executive director includes: governance score, firm performance, industry type, firm size, board size and CEO experience. The list of Datastream firms was uploaded to Orbis to download the controls. The corporate governance score pillar (CGVSCORE) from ASSET4 was used as control. The pillar was part of the ESG framework and benchmarked against other firms5. Governance literature state that the actual impact a director can make on strategic decision making variates (Carpenter & Westphal, 2001). This could imply that the governance in a firm can affect the decision to appoint a non-executive director. The control for firm performance was return on assets (ROA). Interests of dealing with environmental issues might be affected by a firms’ profitability (Aguilera-Caracuel & Ortiz-de-Mandojana, 2013). As being done in existing literature, firm performance is measured as the ROA of
5 The governance score was benchmarked against Thomson Reuters Business Classifications (TRBC) Industry
20
2014. The control industry type was used since differences in appointing non-executive board members can be explained by factors that are industry related. It was checked in what industry the firm belongs to, as in the study of Peng (2004). As been done in previous research, the US Standard Industrial Classification (SIC) primary codes (2 digit) were used (Barnett & King, 2008). The next control variable is firm size. Meyer & Rowan (1977) state that larger firms, are likely to be under more pressure to maintain legitimacy. These firms are being watched and checked more by institutions as government, simply because they are standing out in terms of visibility. In research of Fich (2005) it is used as control for CEO appointments as non-executive board members. Firm size can be measured with the proxy of the natural logarithm of number of employees, as used by Rettab, Brik, & Mellahi (2009). Previous research argues that access to external resources is increases with larger boards (Siciliano, 1996). In this study the firm needs certain resources, so the control variable of board size was included. Lastly, existing studies argue that experience of a board member can have an influence on the appointment of this individual (Carpenter & Westphal, 2001) and can be measured by the age of the CEO (Guthrie Datta & Deepak, 1997). This research controls for CEO experience, measured by the CEOs age.
Analysis
With the above data, the two hypotheses can be tested. The firm was the unit of analysis. The reason why a firm appoints a new non-executive green and/or external oriented board member is because it is legitimate or illegitimate. The independent variable and dependent variable consists of non-continuous data. The DV was dichotomous thus, a binary logistic regression analysis was used to test the two hypotheses. This regression analysis predicts values of the DV by means of values of the IV. However, multicollinearity must be eliminated with this test so a Variance Infliction Factors (VIF) test was conducted. In this research it was tested if x can predict a unique variance in y. The likelihood of a firm appointing a non-executive that is green and/or externally oriented is modeled in the following logistic model (1):
(1) 𝑦 = 𝛼 + 𝛽𝑥 + 𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠 1, 2, 3, 4, 5, 6 + 𝜖
Where y is the dependent variable, a dummy variable of appointments of managers in the appointing firm, taking the value of 1 if a firm appoints a new green non-executive board member (H1) and if the new green non-executive board member is externally environmental oriented (H2), otherwise 0. The y is evaluated as (t+1). The 𝛽 controversy score dummy is indicating whether or not the firm is illegitimate and if the firm has above average controversies in its industry. The variable x is the independent variable and is evaluated as (t). Note that the firm here already is labeled legitimate or illegitimate for the IV.
Controls are the independent control variables of each hypothesis of 1 governance score, 2 firm
21
RESULTS
In this section the results of this research are presented. First, the descriptive statistics and correlations are given. Second, results of the regression analyses that are used to test the two hypotheses are shown. Third, a probit regression analysis is conducted to check the robustness of the results.
Descriptive statistics and correlations
Table 1 shows the descriptive statistics of the correlations, mean and standard deviation (Std. dev.). There are no high correlation values between the variables above the 0,7. To test for multicollinearity the Variance Infliction Factors (VIF) scores are analyzed. According to Robinson & Schumacker (2009), when a VIF exceeds 3, there is a probability of multicollinearity, when it is higher than 5, there is a very good chance of it. A VIF score should not exceed 10, or the VIF mean considerably larger than 1. The variable natural logarithm of firm size has the largest VIF of 1,271 and the mean VIF of all variables is 1,156. Thus, there is no potential for multicollinearity. From the total of 955 CEOs of firms found, 77 are connected with one other firm in this sample. These 77 CEOs have a total of 87 different non-executive board positions in different firms. The sample used to test the first hypothesis is 87. As can be seen in table 2, 17 CEOs from the sample have no externally or internally orientation, so the DV of second hypothesis consists of 70 CEOs. The maximum number of non-executive board positions a CEO has is four.
Table 2 provides an overview of the number of non-executive directorships and the orientation of directors appointed in illegitimate and legitimate firms. Of the 87 firms where a CEO is appointed as non-executive board member, 52 firms were illegitimate due to controversies between 2013 and 2014 and 35 were not. Illegitimate firms appointed 22 green managers and 30 non-green managers. Legitimate firms appointed 16 green managers and 19 non-green managers. Of the CEOs that are non-executive board members of other firms in the sample, a total of 70 are labeled with an internally or externally environmental CSR orientation. 20 have an externally environmental CSR orientation and 50 an internally environmental CSR orientation., For a clear interpretation of table 2, the combinations of number of appointments as a percentage of the total sample are also shown in parentheses.
22
Regression results
A binary logistic regression analysis is performed to test the hypotheses. The results are shown in table 3. The DV for H1, with the 6 control variables is included in model 1 and the DV for H2, with the 6 controls in model 2. The relation between the explanatory variables and the response variable is analyzed below. By making use of a variation of tests, this analysis offers a more comprehensive view of the results. For testing the predictor model in terms of goodness of fit the following test can be used. The Omnibus tests of model coefficients for both models 1 and 2 shows the predictor model does fit better to the data than the null model, with no predictors. This means a significant improvement in fit for both models 1 (Chi-square of 18,032, df = 7, p < 0,05) and model 2 (Chi-square of 12,713, df = 7, p < 0,10).
When looking at what the predictive accuracy and classification would be in the intercept only model (null model) for the first hypothesis, it would be 64,4 percent for the overall classification accuracy. In this intercept only model, 0 non-green managers were predicted to be non-green managers (classification accuracy rate of 0%). 56 of the observed green managers were predicted to be green (classification accuracy rate of 100%). Below the model with multiple variables is described. The Cox & Snell R Square of 0,187 and the Nagelkerke R Square of 0,257 the model with predictors are explaining roughly between 18,7 and 25,7 percent of the variation. When looking at the overall fit of model in the classification table, the overall classification accuracy is 72,4 percent. This table is about the number of cases that are predicted to fall in that category in relation to the actual number of managers in that category. As with the intercept only model in total 31 cases were observed to be non-green and in total 56 cases green. Of these non-green cases, 14 predicted non-green managers were actually non-green, the other 17 are misses in terms of classification accuracy (specificity of 45,2%). 49 of the predicted green managers were actually green and there are 7 misses (sensitivity of 87,5%). The 17 false positives and 7 false negatives were predicted by the model to be green and non-green, respectively but were actually the opposite (false positive rate of 12,5% and false negative rate of 54,84%). Thus, the sum of 14 and 49 divided by the sample of 87 makes the accuracy of 72,4 percent. The model does a better job in determining green appointments than non-green, because of the higher accuracy rate.
23
In the intercept only model (null model) for the second hypothesis, all the 50 internal oriented managers were predicted to be internal oriented (classification accuracy rate of 100%). However, 20 of the observed externally oriented managers were predicted to be externally oriented (classification accuracy rate of 0%). The predictive accuracy here is 71,4 percent. Next, the model with multiple predictor variables is described. When looking at the Cox & Snell R Square of 0,166 and the Nagelkerke R Square of 0,238 the model with predictors are explaining between 15,1 and 21,7 percent of the variation. The classification table shows an overall accuracy of the model 72,9 percent. A total of 50 cases were observed as internal oriented and a total of 20 externally oriented. 45 of the 50 internal oriented were predicted to be in this category. The other five are classification misses, what gives an accuracy of 90 percent. (Thus, specificity of 90,0% and a false negative ratio of 10,0%.) Of the observed 20 externally oriented managers 6 were predicted to be in this category. There are 14 misses, what gives an accuracy of 30 percent. (So, the sensitivity is 30,0% and the false positive rate is 30,0%.) This model determines the externally orientation better than the intercept only model, in terms of accuracy. When looking at the global fir of the model, it appears that it does predict slightly better than the null model.
As can be seen in model 3 of table 3, the intention of having environmental controversies increases the logit of the estimated log odds of appointing an externally oriented manager by 5,605. So, when this positive score increases, the likelihood to fall in the externally oriented manager group is increasing. The environmental controversies have a significant impact on the appointment of managers (B = 1,724,
p < 0,05). When a firm is illegitimate, it is 5,605 times as likely that the firm appoint an externally
oriented manager. Model 2 indicates a significant difference in the population in the logarithm of the odds of appointing externally oriented managers as illegitimate firms in comparison with legitimate firms. The second hypothesis is supported (B = 1,724, p < 0,05) and there is a relationship between illegitimacy of a firm and the appointment of externally oriented managers. Next, the control variables are described.
24
Variables Mean Std. Dev. 1 2 3 4 5 6 7 8
1 Appointment green manager 0,64 0,482
2 Appointment of externally oriented manager 0,29 0,455 0,276*
3 Governance score 77,8856 14,0114 0,303** 0,067
4 Firm performance (ROA) 6,9009 8,3324 0,041 0,001 0,143
5 Industry (SIC) 45,11 15,014 -0,187 -0,116 0,155 0,066
6 Firm size (log) 10,0386 1,4726 0,178 0,184 0,324** 0,280** 0,055
7 Board size 10,80 1,879 0,089 0,164 0,279** 0,155 -0,024 0,278**
8 CEO experience (age) 57,86 6,726 0,182 0,001 0,291** 0,018 0,128 0,241* 0,097
9 Environmental controversies 0,60 0,493 0,124 0,241* 0,011 -0,128 0,196 -0,043 -0,048 -0,059
Number of observations: 87 (variable 1, 3-9) and 70 (variable 2)
25 TABLE 2
Count of non-executive directorships 2015
Total
Illegitimate firms Legitimate firms
Total Internally oriented Externally oriented No orientation Total Internally oriented Externally oriented No orientation Green managers 56 (64,37) 36 (41,38) 20 (22,99) 52 (59,77) 27 (31,03) 16 (18,39) 9 (10,34) 35 (40,23) 23 (26,44) 4 (4,60) 8 (9,20) Non-green managers 31 (35,63) 16 (18,39) 15 (17,23) Subtotal 87 87 Number of observations: 87.
26 TABLE 3
Binary logistic regression analysis results (H1, model 1; H2, model 2)
Model 1
Appointment of green managers
Model 2` Appointment of externally oriented managers Independent variables B SE OR B SE OR Constant -5,310 † 2,799 0,005 -7,638 * 3,829 0,000 Governance score 0,047 * 0,021 1,011 0,015 0,025 1,015
Firm performance (ROA) 0,011 0,032 0,956 -0,028 0,042 0,973
Industry (SIC) -0,045 * 0,018 1,137 -0,033 0,023 0,968
Firm size (log) 0,128 0,184 0,958 0,426 † 0,255 1,531
Board size -0,043 0,143 1,051 0,233 0,185 1,262
CEO experience (age) 0,050 0,041 2,520 -0,016 0,046 0,984
Environmental controversies 0,924 † 0,526 1,001 1,724 * 0,707 5,605
-2 Log likehood 95,289 71,045
Cox & Snell R Square 0,187 0,166
Nagelkerke R Square 0,257 0,238
Classification accuracy (%) 72,4 72,9
Number of observations 87 70
27
Only for the DV of the first hypothesis it is practically possible when coding the variables, to change requirements for falling into a category, to check the robustness. It is checked for a green or non-green manager classification score of 50 or higher (ENVSCORE above 50 means the manager is green) and 75 or higher (ENVSCORE above 75 means the manager is green). The classification table of the 50 model is better able to explain the non-green managers and the 75 model explain the green managers more accurate. However, the overall fit is better for the 50 model. Moreover, using the above 75 score resulted in a less significant model. Eventually, the 50 model is used.
Next, to check for robustness of the test results, a probit regression is conducted. When looking at the findings, the following statements regarding the first hypothesis can be made. The -2 Log Likehood of the final model is improved upon the intercept only model from 90,288 to 108,352. The -2 Log Likelihood of the binary regression is with 95,289 close to the probit regression. The Likelihood Ratio (LR) Chi-Square is (108,352 minus 90,288) 18,064 (df = 7, p < 0,05). A probit regression (and a binary log regression) does not have an ordinary least squares where the predictor variables explains the proportion of variance of the response variable, so the Pseudo R-Square should be interpreted with caution (Hoetker, 2007). As shown in table 4, the Cox and Snell Pseudo R-Square is 0,187 and the Nagelkerke 0,257. Both values are exactly the same as with the R-Squares of the binary logistic regression. Roughly, both models explains the same amount of variation. When looking at the Regression Coefficients Estimate (E), the control variable governance score is significant (E = 0,028, df = 1, p < 0,05), the control industry (E = -0,027, df = 1, p < 0,05) and the independent variable (E = 0,564, df = 1, p < 0,10). When making a comparison of the significant variables of table 3 and 4, both statistic tests show the same variables to be positive or negative, and significant. Only industry type is significant at the 0,05 alpha level in the binary logistic regression and significant at the 0,01 alpha level in the probit model. The Wald test statistic for the constant term in the model (in table 4, the threshold dependent variable (H1)) is 3,637 with p-value of 0,057. When using an alpha level of 0,10, the null hypothesis can be rejected and this model intercept is actually marginal statistically different from zero (p < 0,10). When looking at the Wald for the governance score of 5,491 and the alpha level on 0,05, the null hypothesis can be rejected. The regression coefficient of governance score is statistically different from zero, given that all variables are in the model. When using the same alpha, the industry and IV are also statistically different from zero, given that all variables are in the model.
28
binary logistic regression model of the second hypothesis. The differences are just over one percent point, so the models roughly explain the same variation. Table 3 and 4 show the same significant variables being positive or negative. As can be seen in table 4, the control variable natural logarithm of employees for the second hypothesis (E = 0,271, df = 1, p < 0,10) and the independent variable (E = 1,035, df = 1, p < 0,05) are both at the same significance level as with these variables in the binary logistic regression. The Wald test statistic for the constant term in the model (in table 4, the threshold dependent variable (H2)) is 4,546 with p-value of 0,033. When using an alpha level of 0,05, the null hypothesis can be rejected and this model intercept is actually marginal statistically different from zero (p < 0,05), given all variables are in the model. The natural log of employees and the IV are both marginally significant (Wald = 3,399, p < 0,10) and significant (Wald = 6,588, p < 0,05), respectively.
29
(H1, model 1; H2, model 2)
Model 1
Appointment of green managers
Model 2
Appointment of externally oriented managers
Variables E SE Wald E SE Wald
Threshold dependent variable (H1) 3,143 † 1,648 3,637
Threshold dependent variable (H2) 4,733 * 2,220 4,546
Governance score 0,028 * 0,012 5,491 0,010 0,015 0,418
Firm performance (ROA) 0,007 0,019 0,127 -0,017 0,025 0,450
Industry (SIC) -0,027 ** 0,010 6,857 -0,020 0,013 2,234
Firm size (log) 0,076 0,111 0,474 0,271 † 0,147 3,399
Board size -0,025 0,085 0,085 0,134 0,108 1,535
CEO experience (age) 0,030 0,024 1,527 -0,009 0,027 0,111
Environmental controversies 0,564 † 0,312 3,273 1,035 * 0,403 6,588
-2 Log likehood (final) 90,288 70,648
Likelihood Ratio Chi-Square 18,064 * 13,110 †
Pseudo R-Square (Cox & Snell) 0,187 0,171
Pseudo R-Square (Nagelkerke) 0,257 0,245
Number of observations 87 70
30
In this section the results of the previous parts are discussed. First, the theoretical implications are discussed. Second, the practical implications about what this research outcome can mean for firms, managers and what can be drawn from this at practical level. After this, the limitations of this study are given and further research recommendations are made. The main research question is: Are firms involved
in above average environmental controversies in their industry more likely to appoint green managers, instead of non-green managers and externally oriented managers, instead of internally oriented managers as external board members? Both managers of the first and second hypothesis are found to
be significantly more likely to get appointed as external board members in illegitimate firms.
Theoretical implications
The Omnibus test model coefficients show a significant improvement in fit for the predictor models. This means the goodness of fit is improved, when including the predictors to the model. When looking at the model 1 and 2 in table 3 with the Cox & Snell R Square and the Nagelkerke R Square, the first model predicts slightly more of the variation than model 2 (model 1, roughly between 18,7 and 25,7% and model 2, between 16,6 and 23,8%). Both models 1 and 2 do a better job determining 72,4 percent for green managers and 72,9 percent, respectively for external oriented managers, by means of the overall accuracy rate. Since both models have higher accuracy rate than the predictor only model, it can be assumed that with including all variables, the overall model fit is improved.
31
Secondly, it was tested whether firms with above average environmental controversies in their industry are more likely to appoint externally oriented managers as board members. This is significantly supported. A possible explanation offered by Carpenter & Westphal (2001) is that director ties to other boards can provide the firm with the desired capabilities. In contrast, all control variables are not significant according to the lower and upper limits confidence interval. Once again, the binary logistic regression shows that the IV explains the choice between appointing an externally or internally oriented environmental CSR manager. The firms’ above average controversies in the industry seems to have effect on the likeliness of firms to appoint externally oriented managers. Because of the visibility by means of negative media attention, the firm can use the manager that is specialized to handle it, according to the theory development section. This can be in the form of appointing an externally oriented manager, as being supported by the findings.
When discussing the control variables, a higher governance score shows to positively affect the likelihood of appointing green managers. Interestingly, previous research found that corporate governance is weakened as a result of network ties between non-executive board members and the firms’ CEO (Fracassi & Tate, 2012). This indicates that there is a negative backwards effect. This is possibly because the negative effect of a larger board on the ability of a firm to monitor it. However, according to the results in the previous section, board size has no significant impact. Siciliano (1996) argues that firms have better access to resources when the board size increases. This could support the statement made in the theory section, that managers can bring special capabilities to the board. However, it is not found in the results. On the other hand, literature state that the size of boards decreases in times of high uncertainty, as with illegitimacy of the firm (Shropshire, 2010). The fact that four of the six control variables of the first hypothesis and five of the second hypothesis does not have a significant impact, is also analyzed with a probit model. This robustness check gives roughly the same results.
To summarize, illegitimate firms (in comparison with legitimate firms), with a higher governance score are more likely to appoint green managers as board members, and belonging to the same industry group are less likely to appoint green managers as board members. Further, illegitimate firms (in comparison with legitimate firms), with a higher number of employees are more likely to appoint externally oriented managers as board members.
Practical implications
32
research. Although, firms may not yet pay much attention to classifications of where their non-executive board members belong to, it can expose important possibilities. With knowing a manager is more experienced as with a green or externally oriented environmental CSR manager, firms have capabilities in stock in case of a threat to organizational legitimacy. The only thing left is to test the real long term impact of such a manager, a certain time after the appointment as non-executive board member.
Limitations and further research
The following most important limitations are found that affect the generalizability of this research. It can provide directions for future research.
Non-profit or foundations were not included in the sample, only U.S. public held firms, because of the great amount of data available. This also means that with using U.S. firms, it is not taken into account that there might be differences between the firms’ host country legitimacy standards and environment of a multinational corporation (Kostova & Zaheer, 1999). Scholars could include firms with different legal forms from multiple countries to see if it has any effect on the type of directors appointed. Only CEOs are used to identify green and non-green managers. Measuring if other executives are green managers cannot be done with firm level measurements, that only reflects the end responsible CEO. Furthermore, the eventual differences of newly appointed male and female directors in boardrooms is not taken into account (Singh, Terjesen, & Vinnicombe, 2008).
33
CONCLUSION
34
REFERENCES
Aguilera-Caracuel, J., & Ortiz-de-Mandojana, N. (2013). Green innovation and financial performance: An institutional approach. Organization & Environment, 26(4), 365-385.
Ahmadjian, C. L., & Robinson, P. (2001). Safety in numbers: Downsizing and the
deinstitutionalization of permanent employment in Japan. Administrative Science Quarterly,
46(4), 622-654.
Archel, P., Husillos, J., Larrinaga, C., & Spence, C. (2009). Social disclosure, legitimacy theory and the role of the state. Accounting, auditing & accountability journal, 22(8), 1284-1307.
Baba, S., & Raufflet, E. (2014). Managing relational legacies: lessons from British Columbia, Canada.
Administrative Sciences, 4(1), 15-34.
Bansal, P., & Clelland, I. (2004). Talking Trash: Legitimacy, Impression Management, and
Unsystematic Risk in the Context of the Natural Environment. The Academy of Management
Journal, 47(1), 93-103.
Barnett, M. L., & King, A. A. (2008). Good fences make good neighbors: A longitudinal analysis of an industry self-regulatory institution. Academy of Management Journal, 51(6), 1150-1170.
Baum, J. A., & Oliver, C. (1991). Institutional linkages and organizational mortality. Administrative
science quarterly , 187-218.
Ben Barka, H., & Dardour, A. (2015). Investigating the relationship between director’s profile, board interlocks and corporate social responsibility. Management Decision, 53(3), 553-570.
Booth, J. R., & Deli, D. N. (1996). Factors affecting the number of outside directorships held by CEOs. Journal of Financial Economics, 40(1), 81-104.
Brønn, P. S., & Vidaver-Cohen, D. (2009). Corporate motives for social initiative: legitimacy, sustainability, or the bottom line? Journal of Business Ethics, 87, 91-109.
Bureau van Dijk. (2018). Corporate ownership structures. Retrieved from Bureau van Dijk: https://www.bvdinfo.com/en-us/our-products/our-expertise/find-out-how-we-add-value-to-company-information/corporate-ownership-structures
Bureau van Dijk. (2018). Orbis company information worldwide. Retrieved from Bureau van Dijk:
35
Bureau van Dijk. (2018). Osiris. Retrieved from Bureau van Dijk: http://www.otago.ac.nz/library/pdf/OSIRISDataGuide.pdf
Capriotti, P., & Moreno, A. (2007). Corporate citizenship and public relations: The importance and interactivity of social responsibility issues on corporate websites. Public relations review,
33(1), 84-91.
Carpenter, M. A., & Westphal, J. D. (2001). The strategic context of external network ties: Examining the impact of director appointments on board involvement in strategic decision making.
Academy of Management journal, 44(4), 639-660.
Claasen, C., & Roloff, J. (2012). The link between responsibility and legitimacy: The case of De Beers in Namibia. Journal of Business Ethics, 107(3), 379-398.
Dawkins, J., & Lewis, S. (2003). CSR in stakeholde expectations: And their implication for company strategy. Journal of Business Ethics, 44(2), 185-193.
Deephouse, D. L., & Carter, S. M. (2005). An examination of differences between organizational legitimacy and organizational reputation. Journal of management Studies, 42(2), 329-360.
Del Vecchio, N. (2010). Réseaux de conseils d'administration et adoption de pratiques de gouvernance d'entreprise. Revue française de gestion, (3), 145-161.
DeTienne, K. B., & Lewis, L. W. (2005). The pragmatic and ethical barriers to corporate social responsibility disclosure: The Nike case. Journal of Business Ethics, 60(4), 359-376.
Duff, A. (2017). Corporate social responsibility as a legitimacy maintenance strategy in the professional accountancy firm. British Accounting Review, 49(6), 513-531.
Eichholtz, P., Kok, N., & Quigley, J. M. (2009). Why companies rent green: CSR and the role of real estate. Academy of Management Proceedings, Vol. 2009(1), 1-6.
El Ghoul, S., Guedhami, O., Kwok, C. C., & Mishra, D. R. (2011). Does corporate social
responsibility affect the cost of capital? Journal of Banking & Finance, 35(9), 2388-2406.
Farooq, S. U., Ullah, S., & Kimani, D. (2015). The relationship between corporate governance and corporate social responsibility (CSR) disclosure: evidence from the USA. Abasyn Journal of
Social Sciences, 8(2), 197-212.
