• No results found

The Adoption and Diffusion of Sustainability Practices: The Role of the Interlocking Directorate Network

N/A
N/A
Protected

Academic year: 2021

Share "The Adoption and Diffusion of Sustainability Practices: The Role of the Interlocking Directorate Network"

Copied!
43
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Guus Denteneer

Research Master Thesis Faculty of Economics and Business University of Groningen August 2017

Supervisors

Dr. Rian Drogendrijk Dr. Kees van Veen

The Adoption and Diffusion of

(2)

The Adoption and Diffusion of Sustainability Practices:

The Role of the Interlocking Directorate Network

University of Groningen Faculty of Economics & Business Duisenberg Building, Nettelbosje 2

9747 AE Groningen Author: Guus Denteneer - S2749386 g.t.denteneer@student.rug.nl Supervisors: Dr. Rian Drogendijk

Dr. Kees van Veen

Word count: 11,709

14th of August 2017

(3)

This study examines the role of the European interlocking directorate network on the diffusion of sustainability practices issued by the Global Reporting Initiative (GRI). Various characteristics of the interlocking directorate network gathered from 292 organisations of 14 countries for the years of 2007-2013 are assessed to this end. With the use of an event history methodology, the influence of both the national as well as the cross-national interlocking directorate network on the diffusion of the GRI practices is examined. The results support the majority of the hypotheses and show that there are interesting differences between the national and cross-national interlocking directorate network. According to the evidence, highly internationally connected organisations are important for successful adoption between countries, while central organisations within a country are important for diffusion within a country. Besides, the prevalence of practice adoption within a country is also an important determinant of the effectiveness of the interlocking directorate network and results indicate that interorganisational ties could potentially be a liability once the national standard of practice adoption is high. Overall, the outcomes are consistent with prior research on interlocked directors and show that interorganisational ties are important to take into account when investigating the diffusion of sustainable practices.

Key words: Interlocking directorate network, sustainability practices, practice adoption, diffusion.

(4)

ACKNOWLEDGEMENTS ... 5

1 INTRODUCTION ... 6

1.1. Research question ... 8

2 LITERATURE REVIEW ... 9

2.1. Development of Sustainability Reporting and the GRI ... 9

2.2. The Diffusion of a Voluntary Practice ... 10

2.3. The Interlocking Directorate Network ... 12

3 HYPOTHESES DEVELOPMENT ... 15 3.1. Interconnectedness ... 15 3.2. Ties to adopters ... 16 3.3. International interlocks ... 16 3.4. Organisational environment ... 17 4 METHODOLOGY ... 18

4.1. Sample and Data Collection ... 18

4.2. Variables ... 19

4.2.1. Dependent variable ... 19

4.2.2. Independent variables ... 19

4.2.3. Control variables ... 20

4.3. Statistical method ... 21

4.3.1. Functional form of the baseline hazard function ... 21

5 RESULTS ... 23

5.1. The European Interlocking Directorate Network ... 23

5.2. Descriptive statistics ... 23

5.3. Cox Discrete-Time Proportional Hazard Model ... 25

5.3.1. The GRI Adoption on a national level ... 26

5.3.2. The GRI Adoption on a cross-national level ... 26

5.3.3. Additional analyses ... 30 6 CONCLUSION ... 33 6.1. Managerial Implications ... 35 6.2. Limitations ... 36 REFERENCES ... 37 APPENDICES ... 39

Appendix A – Diffusion pattern of GRI practices by Industry and Country ... 40

(5)

ACKNOWLEMENTS

“We often take for granted the very things that most deserve our gratitude.”

- Cynthia Ozick

Before providing information on the research I conducted as part of my Research Master thesis, I first would like to sincerely thank my supervisors dr. Rian Drogendijk and dr. Kees van Veen for their guidance throughout the process of writing my thesis and their valuable comments during our brainstorm sessions.

Furthermore, I would like to thank prof. Craig Crossland for his useful comments and feedback in the presentation session in which I presented my preliminary results.

Last, I gratefully thank my family, girlfriend, and friends that supported me up until completion of the thesis and helped me think critically about my research by exchanging thoughts with me about my thesis topic.

(6)

1 INTRODUCTION

In recent decades, organisations worldwide have increasingly started to disclose environmental sustainability information in their corporate reports. Due to globalization, the strength, number, and variety of stakeholder pressures have increased and forced organisations to be socially responsible (Surroca, Tribó, & Zahra, 2013; Campbell, 2007). However, meeting both domestic as well as international stakeholder’s incremental expectations on sustainability issues is a tough job for organisations, particularly because it is unusual for organisations and almost no attention was devoted to the sustainability topic in the past (Byun & Kim, 2017). To be able to meet the expectations of stakeholders and respond to their pressures, organisations started to communicate how they justified their actions and contributed towards a more sustainable future by disclosing information about their corporate social responsibility (CSR) behaviour in their corporate reports. As a consequence, an increasing number of organisations have adopted sustainability practices to report their environmental and social performance alongside their economic performance (Fuente, García-Sánchez, & Lozano, 2017).

While organisations make attempts to meet stakeholder expectations by disclosing information with reference to their sustainability issues, concerns are raised about the effectiveness and quality of sustainability reporting (Hess, 2008). In order to provide guidance and establish standards for this sustainability reporting trend, third-party institutions emerged that guaranteed a certain level of quality and were able to provide legitimacy by including organisations in a community (Junior, Best, & Cotter, 2014). A third-party institution that provides the most-used sustainability practices by organisations is the Global Reporting Initiative (GRI) (Brown, de Jong, & Levy, 2009). The mission of the GRI is to enable organisations to consider their impacts on a wide range of sustainability issues, enabling them to be more transparent about the risks and opportunities they face (Global Reporting Initiative, 2017). The GRI developed a framework for companies to write sustainable reports and launched their first version in 1999. Since the introduction of the first global practices for sustainability reporting by the GRI, the number of companies adopting these GRI practices has risen extensively. Started with a handful of organisations that adopted the GRI practices in 1999, nowadays over 4,000 organisations worldwide have adopted the GRI practices, turning the GRI practices into a global standard for sustainability reporting.

(7)

practices by organisations, such as the GRI practices, can be a starting point for exploring the widespread engagement in sustainability reporting. Prior research suggested that one of the key reasons for organisations to engage in sustainability reporting are pressures from stakeholders (Byun & Kim, 2017; Chan, Watson, & Woodliff, 2014). Furthermore, by conforming to criteria set by the external environment, organisations obtain legitimacy that is a way for organisations to justify their actions (Meyer & Rowan, 1977). Another important reason why organisations disclose information about their sustainability actions might not be driven by external pressures but might come from internal sources; the board of directors. As the decisions made by organisations are influenced by the choices, motives, and values of the board of directors (Hambrick & Mason, 1984; Hambrick, 2007), the adoption of practices might be contingent on who is represented on the board of the organisation (Fiss & Zajac, 2004). Directors make corporate decisions and manage the organisation based on experiences and perceptions they obtain from the business environment and advice that is shared with them. Since their experiences are often collected from positions held at other boards, prior research on the diffusion of practices (e.g. Davis, 1991; Shipilov, Greve, & Rowley, 2010) suggests that the interorganisational network, composed of directors that sit on multiple boards, is an important source for explaining an organisation’s likelihood and rate of practice adoption. Not only is an organisation able to signal to stakeholders that the organisation is a legitimate organisation by creating director ties with important organisations (Mizruchi, 1996), an organisation can also benefit from directors that are represented in multiple boards by “business scans” that can be performed (Useem, 1984). A board director that serves on the board of multiple organisations is often referred to as an interlocking directorate (Haunschild, 1993). The interlocking directorate network is often used to investigate the diffusion of practices among organisations and is thus considered an important reason of why organisations adopt practices. However, research on the European interlocking directorate network, and international interlocks in particular, has thus far received almost no academic attention (Lamb & Roundy, 2016). Staples (2007) and Van Veen & Marsman (2008) investigated the international character of boards of directors and found that international directors are represented in the majority of large organisational boards, which suggests that international interorganisational relations may be important for the diffusion and adoption of practices.

(8)

influence of the interlocking directorate network on GRI adoption across countries, more insight will be provided into the importance of international interlocks. Furthermore, this study will contribute to the interlocking literature on diffusion by linking the interlocking directorate network to the diffusion and adoption of voluntary, sustainability practices (Ortiz de Mandojana & Aragon-Correa, 2015).

1.1. Research question

The limited amount of research that looks into the international interlocking directorate network and increasing attention from both organisations and governments devoted to sustainability reporting form the principal motives to further investigate this topic. The main objective of this study is to provide insight into the importance of the (inter)national interlocking directorate network in European countries and its influence on the diffusion and adoption of voluntary practices. I aim to investigate whether, and to what extent, the European interlocking directorate network is diversified or centralized in order to identify future practical implication for policy makers. In doing so, I aim to answer the following research question:

What explains the role of the interlocking directorate network in the diffusion of sustainability practices, and what aspects of the organisation’s interlocking directorate network are most important for the successful diffusion of a voluntary practice?

To answer this research question and bring a certain structure to the study, a number of sub questions are formulated:

• How has the field of sustainability reporting developed over time? • What is diffusion and how does a voluntary practice diffuse?

• What important interlocking directorate network characteristics can be derived from existing literature?

(9)

2 LITERATURE REVIEW

In this section, literature on sustainability reporting, the diffusion of practices and the interlocking directorate network will be reviewed. Additionally, more insight will be provided into the development of the GRI practices and the diffusion of voluntary practices. First, I will devote attention to the development of sustainability practices in the last decades. Second, literature on diffusion and voluntary practices will be discussed in order to clarify the stages that diffusion normally consists of and to highlight the potential differences between diffusion of compulsory and voluntary practices. Last, background information on the interlocking directorate network will be discussed and prior findings of studies that investigated the role of the interlocking directorate network in the diffusion of practices will be presented.

2.1. Development of Sustainability Reporting and the GRI

Concerns about the impact of organisations’ activities on the natural, economic, and social environment have resulted in a worldwide call for action towards organisations to increase their accountability, transparency and stakeholder engagement (Daub, 2007). As a direct response to these changing needs, organisations started to disclose information about their sustainability to transparently communicate their values, actions, and performance to stakeholders. This responsibility that organisations took towards stakeholders and society as a whole is often referred to as ‘corporate social responsibility’ or ‘corporate citizenship’ (Ackerman & Bauer, 1976; Carroll, 1989). Being socially responsible as an organisation is derived from business ethical behaviour and implies that an organisation should not only create economic value to be used solely by the organisation itself, but rather should create value that can be given back to society (Daub, 2007). In the process of deriving value from its operations, an organisation also generates negative external effects that possibly harm the environment or society. The larger the organisation becomes, the more value and negative effects it will generate, the more visible it will become to the public eye and the more obliged the organisation becomes to justify its actions and presence in society. In order to legitimise their actions and respond to stakeholders’ pressures, organisations publish sustainability reports.

(10)

in the 1990s prompted organisations to disclose environmental information in sustainability reports. Today, sustainability reports are used as a management tool that provides the possibility for organisations to communicate their sustainability and CSR initiatives.

Nevertheless, organisations used their own formats to publish their sustainability reports. To make these reports more comparable and to ensure the quality and validity of the sustainability reports published by organisations, a reliable set of standards was desirable for organisations worldwide (Lozano & Huisingh, 2011). As an answer to this need, the Global Reporting Initiative was founded in 1997 under the initiative of the North American Coalition for Environmentally Responsible Economies (CERES) of Boston and the Tellus Institute. Although the first GRI practices that were issued in 1999 were adopted by only a handful of organisations, nowadays over 4,000 organisations worldwide have adopted the GRI practices (Global Reporting Initiative, 2017), making GRI the most-used global standard for sustainability reporting worldwide (Skouloudis, Evangelinos, & Kourmousis, 2009). The success of the GRI practices as a global reference can be shown in Figure 1. Here, the cumulative diffusion pattern of the GRI practices by publicly traded companies of 15 European countries is displayed since the first version of the GRI practices was issued1. GRI practices successfully aligned with the CSR strategies of

organisations have helped organisations with the improvement of human rights, labour rights, and both environmental and anticorruption practices (Perez-Batres, Miller, & Pisan, 2010). Additionally, the GRI requires that organisations report about information on economic, environmental and social issues disregarding whether the reported information has negative consequences for the organisation (Prado-Lorenzo, Gallego-Alvarez, & Garcia-Sanchez, 2009). The latter makes the GRI an important communication tool for mitigating the information asymmetry problem between an organisation and its stakeholders and provides investors and analysts with a more accurate valuation of the organisation, while simultaneously stimulating organisations to become more sustainable (Lozano & Huisingh, 2011).

2.2. The Diffusion of a Voluntary Practice

Studies on diffusion often investigate an innovation which is communicated through channels within a social system over time (Rogers, 1983). The term innovation applied by Rogers (1983) implies that the idea or practice that is being diffused is novel to the members of the social system, which creates a degree of uncertainty. This results in the iconic S-shaped curve that is frequently used in research on diffusion and represents the different stages that a diffusion

1Countries included are Austria, Belgium, Denmark, Finland, France, Germany, Greece, Italy, Ireland, Luxembourg,

(11)

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Pe rc en ta ge A do pt in g G R I St an da rd s Year

Diffusion of GRI Practices in Europe between 1999 and 2015

process goes through; from innovators and early adopters to majority adopters to finally end the diffusion process with laggards.

Innovators and early adopters are characterized by adoption at an early stage, while laggards are characterized by adoption after the greater majority adopted. The social system in which the diffusion occurs is defined as a “set of interrelated units that are engaged in joint problem solving to accomplish a common goal” (Rogers, 1983, p. 24).

(12)

information demand from outsiders (e.g. business analysts, accounting firms, stakeholders), implementation of voluntary disclosing practices by organisations might be pressured which weakens the contagion of these practices through the social system. Lastly, differences in the costs and benefits of implementing voluntary practices among organisations may affect how information and experiences diffuse via the interorganisational network (Cai et al., 2014).

2.3. The Interlocking Directorate Network

In his book, Rogers (1983) argued that the diffusion of an innovation can be facilitated by personal networks within a social system. Such a network consists of a set of individuals that interact with each other and is also referred to as an interlocking personal network. Numerous scholars (e.g. Davis, 1991; Davis & Greve, 1997; Fich & Shivdasani, 2006; Shipilov et al., 2010 among others) have investigated the influence of personal networks on the diffusion of practices within organisations by looking at the interlocks that exist between directors of organisational boards. These board interlocks are constituted when a director is represented on multiple boards simultaneously (Useem, 1984; Haunschild & Beckman, 1998). Board interlocks form interorganisational linkages that can facilitate interaction and communication between organisations (Pfeffer & Salancik, 1978). As these linkages are perceived by organisations as a reliable, valuable and often inimitable information resource, information and practices communicated through the interlocking directorate network can influence corporate decisions (Galaskiewicz & Wasserman, 1989; Haunschild, 1993). Galaskiewicz and Wasserman (1989) argue that interorganisational ties act as pipes for diffusing ideas and are reliable information sources that organisations trust. Due to the fact that decision-makers have confidence in the interlocking directorate network, the uncertainty that often is accompanied by an innovation will be mitigated and it is more likely that mimetic behaviour will take place between organisations. This decreases the environmental uncertainty to which the organisation is exposed to and provides legitimacy (Shropshire, 2010).

(13)

other organisations, which results in “an inexorable push towards homogenization” (DiMaggio & Powell, 1983, p. 148).

A complementary stream of literature that focuses more on the effect of successful contagion between objects in the network itself and is important when researching the influence of the interlocking directorate network was stimulated by Grannovetter’s (1973) work “The Strength of Weak Ties”. In this paper, Granovetter (1973) argues that weak ties often are the important links that connect otherwise distant ‘cliques’ with each other. Weak ties provide the interconnectivity in a social system, such that it becomes a total system. Removing these weak ties would result in an unconnected network that consists of separate bundles of nodes. As a result, the ties that are pivotal for an organisation in order to obtain new and valuable information are the weak ties, in contrast to the strong ties an organisation has within its ‘clique’. This also infers that the ties that an organisation has with other organisations determine its behaviour (i.e. the decisions made by the organisation that are based on information obtained from interlock ties). Hence, the behaviour of an organisation is socially embedded and is likely to be affected by the organisation’s relations with other firms (Granovetter, 1985). Social network theorists acknowledge that organisational behaviour is embedded in social networks and the extent to which organisations are connected with each other influences their behaviour (Shropshire, 2010). Prior research found support for the impact of social embeddedness on organisational behaviour and show that the interlocking directorate network influences the process of embeddedness (Davis, 1991; Mizruchi, 1996; Davis & Greve, 1997).

(14)

However, while it is reasonable to assume that information will be transferred to organisations via board interlocks, we cannot always assume that the focal board will decide to implement practices or innovations associated with that information (Shipilov et al., 2010). This could particularly hold for voluntary practices that might be costly to implement, yet bring no immediate benefits for the organisation.

(15)

3 HYPOTHESES DEVELOPMENT 3.1. Interconnectedness

(16)

adapt to. As one example, Fligstein (1990) investigated the diffusion of management models across organisations and found that among the innovators and early adopters, the majority were the more central organisations. Once the management models were proven to be useful for responding to new politico-economic conditions, the management models diffused to the larger, less central business community. Therefore, I hypothesize:

Hypothesis 1: Interlock network centrality will increase a firm’s rate of GRI practice adoption.

3.2. Ties to adopters

An innovation or practice that is diffused within a social system is new to the members of the social system and encompasses a certain degree of uncertainty (Rogers, 1983). Organisations need to overcome this uncertainty before they will adopt the practice (e.g. Galaskiewicz & Burt, 1991). Interlock ties will reduce the uncertainty by organisations, as the interlocking directorate network is perceived as a reliable information source through which information and experiences are communicated. However, as the costs and benefits for implementing voluntary sustainability practices might be different among organisations, having direct connections with organisations that adopted the practices will further mitigate the ambiguity surrounding the practice (Cai et al., 2014). In the case of GRI practice adoption, interlocked directors’ first-hand experience with organisations that adopted GRI practices can help the focal organisation by sharing their experiences and promoting the adoption of GRI practices. This effect will be reinforced when more organisational ties of the focal organisation adopt, which could lead to the point that the adoption of the practice becomes normatively appropriate (DiMaggio & Powell, 1983). Consequently,

Hypothesis 2: Interlocks with other firms that have already adopted GRI practices will increase a firm’s rate of GRI practice adoption.

3.3. International interlocks

(17)

organisation with a large international interlock network will be more visible to the public eye as it is exposed to a variety of stakeholders and will be expected to act as a role model for other organisations (Daub, 2007). Such an organisation is more willing to respond to pressures from stakeholders by adopting sustainability practices to legitimise its actions. Another important aspect of organisations with an international interlock network is that these organisations connect large parts of the interlocking directorate network. Organisations with international interlocks are able to connect the interlock networks of countries and could serve as crucial ‘weak ties’ that contain valuable and new information, particularly when the country standard for sustainability practice adoption differs between countries. Thus, due to the high visibility and the important information that is likely to be obtained by organisations with an international interlock network, the likelihood of practice adoption will be higher. Therefore, I hypothesize:

Hypothesis 3: Interlocks with international firms will increase a firm’s rate of GRI practice adoption.

3.4. Organisational environment

As can be inferred from the process of homogenization that is established by the pressures from the environment, organisations often base their decisions on what their peers are doing (DiMaggio & Powell, 1983). By participating in the process of mimetic isomorphism, organisations focus more on what their competitors are doing and will imitate the behaviour from the ones that are most similar to themselves (Burt, 1987; Meyer & Rowan, 1977). This might particularly hold for organisations that operate in the same industry. These organisations often have similar suppliers and buyers or produce a comparable product or service, which makes them more or less equivalent. As a result, organisations that operate within the same industry are likely to adopt the same practices even though the costs and benefits of implementing the practices differ between organisations. The pressures to adopt the practices from similar organisations that are operating in the same industry increase as the industry standard becomes higher and the voluntary sustainability practice turns into a social norm. This would suggest that the industry standard with regards to practice adoption is likely to influence the rate of practice adoption. Put differently, the more common it is to adopt the practice in the industry, the higher the rate of practice adoption will be among organisations and the higher the pressures will be to conform to the industry standard. Hence,

(18)

4 METHODOLOGY 4.1. Sample and Data Collection

An initial sample was used that included all publicly traded organisations available in the European Top Management Project (ETMP) database. The ETMP database is constituted by the Institute for Governance and Organisational Responsibility (iGOR) of the University of Groningen. In this database, the board composition and characteristics of individual directors of corporate boards are recorded from the top-20 largest publicly traded organisations in 15 European countries for the years 2005-2015. The latter yielded a total sample of 358 European organisations. As the country Luxembourg that was included in the ETMP database only consisted of four publicly traded firms, this country and its firms were excluded from the sample. Furthermore, because a timeframe of ten years is relatively long, it could be the case that organisations changed names, got delisted or were acquired by other organisations. Consequently, I performed a background check on all organisations and excluded the ones that got delisted, went bankrupt or were taken over by other companies, leaving 317 companies in the sample. After excluding 25 companies with missing board or financial information, the effective sample size consisted of 292 European organisations. Unfortunately, the financial information used as control variables was only available in the Orbis database for the period 2007-2013. As a result, the timeframe used for the analyses is reduced to the period 2007-2013 in order to be able to have complete information for each organisation in each year. To construct the national and international interlocking directorate network of the European countries present in the sample, board information is used from the ETMP database. The ETMP database includes board information from over 10,000 directors of European companies and includes director characteristics, such as tenure, nationality, gender and age that were hand-collected from the companies’ annual reports and their corporate websites. Additionally, each director is given a personal ID, which allows me to track directors over time and identify directors with multiple directorships.

To obtain information on the diffusion pattern of the GRI practices in the 14 European firms, I obtained data from GRI’s Sustainability Disclosure Database. This database is constructed by the GRI and includes sustainability and integrated reports that GRI is aware of published from 1999 till present and classifies whether corporate reports are constructed following the GRI practices. It contains information on the country, industry sector and region in which the company is active and indicates the adherence level of the report to the GRI practices.

(19)

4.2. Variables

4.2.1. Dependent variable

To trace the diffusion and adoption of GRI practices, a dependent variable is constructed that was assigned with the value equal to one if an organisation in a certain year adopted the GRI practices and zero otherwise. Once a firm had adopted the GRI practices, it no longer generated an observation and was removed from the sample in the subsequent year. Hence, the data structure consists of organisation-year observations for those organisations that had the possibility of adopting the GRI practices.

4.2.2. Independent variables

(20)

also highly skewed and measured by applying a logarithmic transformation. Prevalence of adoption is measured by the percentage of organisations that adopted the GRI practices in the industry in which the focal organisation is active. Industries in which a high percentage of firms already adopted GRI practices will make it more likely for the focal firm to also adopt the GRI practices.

Figure 2 Distribution of the number of interlocks among the organisations in the sample.

4.2.3. Control variables

Since organisations that are larger and more profitable are likely to be involved in sustainability reporting as they will have more room to pursue extensive sustainability programs, I control for the size of the organisation and financial performance. The size of the organisation is measured by the total Number of employees on a fulltime basis. The financial performance of the organisation is measured by the Return on Assets (ROA). Additionally, as governments can be important stakeholders of organisations that can pressure organisations to behave socially responsible, I control for governmental influence. Governmental influence is measured as a dummy variable and is coded one if the government is a shareholder of the company with more than 5% of the shares and zero otherwise. In a similar vein, companies that have more foreign shareholders might be more visible to the public eye and, as a result, be exposed to a broader range of pressures from stakeholders. This could accelerate the adoption rate of sustainability practices. To control for Foreign ownership, I included a dummy variable that is coded one if a foreign shareholder holds more than 5% of the shares of the company and zero otherwise.

(21)

4.3. Statistical method

To be able to test the hypotheses, a Cox discrete-time proportional hazard model with time-varying covariates will be used (Cox, 1972; Allison, 1982). Here, a discrete proportional hazard model is used as usual Cox models for analysing event histories assume that time is measured on a continuous scale. Since the event of a company adopting the GRI practices and the specifications of interlocking directorate network are observed on an annual basis only, a time model is more appropriate than a continuous-time model. Additionally, applying a time model allows me to easily incorporate time-varying explanatory variables into a discrete-time analysis. Since I collected data on boards of directors and firm characteristics over several years, these variables are likely to change over time. An advantage of a Cox hazard model over other cross-sectional approaches, such as logit or probit, is the explicit incorporation of the timing of changes in the dependent variable.

In practice, a firm could adopt a GRI practice or change the composition of its board at any time during the year, which assumes that there exists a continuous hazard that is observed only in certain periods in time. With the use of a complementary log-log specification (cloglog command in Stata), I am able to measure the discrete representation of a continuous time proportional hazards model.

4.3.1. Functional form of the baseline hazard function

(22)

one that is increasing over time and the functional form that best fits this perspective is an exponential functional form. To evaluate what statistically is the most appropriate functional form, I look at the data and estimate multiple models with each different functional forms of the baseline hazard function. From the Akaike’s information criterion (AIC) and Bayesian information criterion (BIC) output and the Chi-square values that are shown in Table 1, the functional form that showed the best model fit is the model with the log(time) baseline hazard function. As a result, the Cox discrete-time proportional hazard model will be estimated with a log(time) baseline hazard function.

Table 1 Model fit statistics of baseline functional forms

Functional Form Observations AIC BIC !"

(23)

5 RESULTS

5.1. The European Interlocking Directorate Network

With the use of the information from the ETMP database, the interlocking directorate network in Europe can be constructed and can be tracked over time. Information on the directors with multiple positions on organisational boards can be used to create a multimodal network, one where organisations are connected by virtue of an interlocking director. In Figure 3, the European interlocking directorate network is displayed for the years 2007 and 2013, based on the director information available from the 292 organisations located in 14 European countries. The organisations are represented by coloured nodes, connected by shared directors that form the edges. Each node colour represents a county (e.g. light blue is a UK-based organisation) and the size of the node represents the centrality of the of the organisation in the international network. The thickness of the edges indicates the number of directors that have the same tie between organisations; the thicker the edge line, the more directors sit on the boards of the same two organisations. From Figure 3, one can observe that the organisations in Germany and France are most central in the network, although over time organisations in the countries Sweden and Finland have become more central as well. In contrast, organisations in Greece and Portugal have few connections with other organisations, both within the country as well as cross-country. As we look at the specifications of the interlocking directorate network in 2007 and 2013, certain things can be noticed. First, the average number of interlocks of an organisation decreased over time from 5.08 to 3.68. This trend of diminishing interlocks is also observed by Heemskerk (2010) and Davis and Greve (1997). As can be inferred by the two interlock networks, the number of thick edge lines has decreased as well, suggesting that the diminished number of interlocks is due to the fact that fewer directors of one organisation are now sitting on the same board of another organisation.

5.2. Descriptive statistics

To be able to distinguish between the adoption process within a country and between countries, variables are measured both at the national level as well as the cross-national level. Tables 2 and 3 present the descriptive statistics of the variables and their correlations. Table 3 also includes the variable International interlocks, as this variable can only be measured at the cross-national level.

(24)

Figure 3 The European Interlocking Directorate Network in 2007 (left) and in 2013 (right).

Table 2 Descriptive Statistics and Correlation Matrix (National level)†

Mean S.d. 1 2 3 4 5 6 7 8

1 GRI Adoption 0.25 0.43 --

2 Network centrality 0.34 0.31 .160** --

3 Interlocks (log) 1.09 0.71 .179** .784** --

4 Ties to adopters (log) 0.74 0.69 .201** .652** .834** --

5 Industry adoption 0.43 0.32 .519** .019 .152** .331** --

6 Number of employees (log) 9.31 1.84 .300** .214** .340** .443** .230** --

7 Return on Assets (ROA) 5.58 9.13 .083* .058 .062 .026 .054 .083* --

8 Government ownership 0.41 0.49 .006 -.005 .126** .137** .036 .114** -.091** --

9 Foreign ownership 0.67 0.47 .016 -.087** -.121** -.074* .017 .039 -.065* -.095**

The size of the sample declined over time as follows: 2007 (291); 2008 (166); 2009 (130); 2010 (120); 2011 (100); 2012 (79); 2013 (66).

*Correlation is significant at the 0.05 level (two-tailed). **Correlation is significant at the 0.01 level (two-tailed).

(25)

Table 3 Descriptive Statistics and Correlation Matrix (Cross-national level)†

Mean S.d. 1 2 3 4 5 6 7 8 9

1 GRI Adoption 0.25 0.43 --

2 Network centrality 0.10 0.17 .194** --

3 Interlocks (log) 1.30 0.77 .230** .699** --

4 Ties to adopters (log) 0.93 0.77 .233** .729** .867** --

5 International interlocks (log) 0.49 0.63 .247** .609** .691** .677** --

6 Industry adoption 0.56 0.15 -.046 -.081* -.145** -.034 -.135** --

7 Number of employees (log) 9.31 1.84 .300** .350** .410** .431** .301** -.122** --

8 Return on Assets (ROA) 5.58 9.13 .083* -.008 .041 .013 -.002 -.073* .083* --

9 Government ownership 0.41 0.49 .006 .134** .084** .112** .010 .041 .114** -.091** --

10 Foreign ownership 0.67 0.47 .016 -.047 -.071* -.024 .035 .034 .039 -.065* -.095**

The size of the sample declined over time as follows: 2007 (291); 2008 (166); 2009 (130); 2010 (120); 2011 (100); 2012 (79); 2013 (66).

*Correlation is significant at the 0.05 level (two-tailed). **Correlation is significant at the 0.01 level (two-tailed).

From outcomes of the tables, it can be inferred that the network variables are significantly correlated, some with high correlations. Overall, larger organisations are better connected to other firms in the network, have more interlocks and ties to adopters and correlate positively with GRI adoption. Government influence and influence by large foreign shareholders do not seem to be correlated substantially with the independent variables, suggesting that their influence is limited.

As one can infer from Table 2 and 3, the independent variables – interlocks, ties to adopters, and international interlocks – are highly correlated. This can easily be explained by the fact that these variables are highly interdependent or the computation of one variable depends on the other. The number of international interlocks is a component of the total amount of interlocks of an organisation. Additionally, as the number of firms that adopt GRI practices increases over time, the number of ties to adopters that companies have to prior adopters becomes roughly the same as the number of total interlocks. This leads to a high correlation between the total number of interlocks and the number of ties to adopters. As a consequence, each of the independent variables will be introduced separately in the model, in order to mitigate the problems of multicollinearity. Please note that the final models in which all variables are included in the Cox-model need to be interpreted with caution, as multicollinearity might play a role here.

5.3. Cox Discrete-Time Proportional Hazard Model

(26)

be discussed first, after which the results will be discussed of diffusion of the GRI practices on a cross-national level.

5.3.1. The GRI Adoption on a national level

In Table 4, the Cox regression results are displayed for the GRI adoption on the national level. In this model, measures of the variables are used that are based solely on the national interlock network of the countries, disregarding the international interlocks that exist between organisations of countries. The model (Model 1) that only includes the control variables shows that the hazard rate significantly increases with the number of employees and that the baseline hazard decreases with elapsed survival time. This suggests that larger firms are more inclined to adopt GRI practices and that the probability that firms adopt GRI practices decreases over time. Model 2-5 each add one of the independent variables to Model 1, corresponding to the Hypotheses. Model 6 presents the results from including all independent variables simultaneously. Model 2 integrates the firm centrality within the country network in which the firm is established and shows that the hazard rate of adopting GRI practices increases as firms are more central in a network. Although Model 6 has to be interpreted with caution due to a high correlation between several variables, the effect remains stable and increases in magnitude in the last Cox regression in which all independent variables are included. Model 3 includes the number of interlocks of the organisation, however, although the hazard rate is positive it is not significant (p = 0.159). This effect remains insignificant in the last model as well. Model 4 includes the number of ties to prior adopters and although the hazard rate slightly increases with the inclusion of the number of ties to adopters, this effect is highly insignificant as well (p = 0.879). Model 5 integrates the industry GRI adoption prevalence and shows that the hazard rate significantly increases when more organisations within the same industry have adopted GRI practices. This effect also remains stable in Model 6.

5.3.2. The GRI Adoption on a cross-national level

(27)

within the international network, which shows a positive hazard rate that is significant (p < 0.05). This suggests that organisations with more connections are more likely to adopt GRI practices. The hazard rate in Model 3 increases with the inclusion of all independent variables in Model 7, although it becomes less significant (p < 0.10). Model 4 includes the number of ties to adopters an organisation has within the international network; however, the increase of this hazard rate is insignificant. The inclusion of the number of international interlocks in Model 5 indicates that organisations with international connections have a higher probability to adopt GRI practices and is highly significant (p < 0.01). This might be due to the fact that these firms are more exposed to a variety of stakeholders and more pressured to adopt GRI practices. The effect of international interlocks also remains stable in Model 7. Similar to the national industry GRI adoption prevalence, the integration of the international industry GRI adoption prevalence significantly increases the hazard rate of GRI practice adoption (p < 0.01). The positive hazard rate also remains stable in Model 7.

Table 4 Cox regression analysis of GRI Adoption (National level; 2007-2013)

(1) (2) (3) (4) (5) (6) Network centrality 0.570** (0.239) 1.909*** (0.575) Interlocks (log) 0.191 (0.136) 0.176 (0.354)

Ties to adopters (log) 0.032

(0.139) -0.982*** (0.271) Industry adopters 6.063*** (0.544) 6.478*** (0.577) Number of employees (log) 0.360***

(0.063) 0.346*** (0.063) 0.349*** (0.063) 0.358*** (0.063) 0.334*** (0.069) 0.342*** (0.071) Return on Assets (ROA) 0.013

(0.009) 0.012 (0.010) 0.013 (0.010) 0.012 (0.010) 0.008 (0.010) 0.006 (0.011) Government ownership 0.138 (0.177) 0.165 (0.178) 0.152 (0.177) 0.140 (0.178) 0.112 (0.200) 0.171 (0.202) Foreign ownership 0.001 (0.160) 0.006 (0.160) 0.004 (0.160) 0.001 (0.160) 0.080 (0.179) 0.082 (0.182)

Industry dummies Yes Yes Yes Yes Yes Yes

Country dummies Yes Yes Yes Yes Yes Yes

Constant -4.413*** (0.775) -4.435*** (0.775) -4.461*** (0.776) -4.410*** (0.775) -7.895*** (0.935) -8.515*** (0.963) c2 205.02 210.51 207.04 205.07 436.43 461.59 D.f. 24 25 25 25 25 28

(28)

Table 5 Cox regression analysis of GRI Adoption (Cross-national level; 2007-2013) (1) (2) (3) (4) (5) (6) (7) Network centrality 0.666 (0.502) -0.484 (0.694) Interlocks (log) 0.328** (0.130) 0.425* (0.257)

Ties to adopters (log) 0.184

(0.128)

-0.391* (0.232)

International interlocks (log) 0.359***

(0.122) 0.355** (0.171) Industry adopters 4.235*** (0.782) 4.267*** (0.796) Number of employees (log) 0.360***

(0.063) 0.346*** (0.063) 0.326*** (0.063) 0.341*** (0.064) 0.315*** (0.063) 0.376*** (0.063) 0.339*** (0.064) Return on Assets (ROA) 0.013

(0.009) (0.010) 0.012 (0.010) 0.013 (0.010) 0.013 (0.009) 0.012 (0.010) 0.012 (0.010) 0.013 Government ownership 0.138 (0.177) 0.168 (0.178) 0.165 (0.177) 0.151 (0.177) 0.157 (0.176) 0.048 (0.184) 0.055 (0.185) Foreign ownership 0.001 (0.160) -0.008 (0.160) 0.010 (0.160) 0.001 (0.160) 0.011 (0.160) -0.008 (0.162) 0.016 (0.163)

Industry dummies Yes Yes Yes Yes Yes Yes Yes

Country dummies Yes Yes Yes Yes Yes Yes Yes

Constant -4.413*** (0.775) -4.255*** (0.782) -4.458*** (0.769) -4.354*** (0.771) -4.143*** (0.762) -6.754*** (0.897) -6.847*** (0.921) c2 205.02 206.75 211.70 207.11 213.63 232.42 243.88 D.f. 24 25 25 25 25 25 29

Standard errors in parentheses. *p<0.10 **p<0.05 *** p<0.01

(29)

practices first. Hypothesis 2, which suggests the number of ties a focal organisation has to prior adopters increases the rate of GRI practice adoption, is not supported. On both the national and cross-national level, ties to prior GRI adopters had an insignificant effect. This insignificance is mainly due to the country dummies that are controlled for. The latter might indicate that the diffusion of practices is not similar for all countries. In section 5.3.3, I will look at the country effects into more detail. Hypothesis 3 that hypothesized that the number of international interlocks will increase a firm’s rate of GRI practice adoption is supported, as the result of Table 5 showed increasing and significant hazard rates. This result indicates that, within the international network, firm connections across countries are important links for the diffusion of practices. The last Hypothesis, which argues that the adoption prevalence of an organisation’s industry increases the GRI practice adoption of the focal firm, is supported as well. In both analyses, the inclusion of the percentage of organisations in the industry that already adopted the GRI practices significantly increased the hazard rate of adopting GRI practices. This might also suggest that firms do not solely depend on information obtained from their network but are also observing their environment and respond to best practices by adopting the practices that are implemented by their peers in the environment.

(30)

represent organisations that have adopted the GRI practices. From this figure can be inferred that in the year in which I started to construct the European interlocking directorate network, the greater part of the organisations that are most central in the international network adopted the GRI practices. Over time, the GRI practices have diffused to the majority of the organisations in the network, including the organisations that are more remote or share fewer connections with other organisations. Of course, this is only an interpretation from an observation but it is consistent with the findings of the Cox regressions.

5.3.3. Additional analyses

As briefly touched upon earlier in section 5.3.2, there might exist differences between countries in their prevalence of GRI adoption. As becomes apparent from Figures 5 and 6 of Appendix A, there do exist differences in the prevalence of GRI adoption among countries. In some countries (e.g. Spain, the Netherlands, and Germany) the rate of GRI practice adoption is higher than in other countries (e.g. Greece and Ireland). The prevalence of GRI adoption in an organisation’s country might also influence the impact of the interlocking directorate network. One could argue that the interlocking directorate network might be more valuable for organisations in countries in which the adoption rate is below the international average rate of adoption than in countries

2007 2013

(31)

in which the adoption rate is above the international average. In fact, the interlocking directorate network of organisations in countries in which the prevalence of GRI adoption is high might even become a liability for other organisations that are more isolated and have not adopted practices yet. In order to test whether the prevalence of GRI adoption in countries has an influence on the interlocking directorate network, I constructed an average adoption percentage for each country per year and an average adoption percentage for all the countries in the social system. Based on the yearly average adoption percentage of the entire network, I classified countries as being below or above the international average rate of adoption. With these two groups, I reran the Cox regressions in order to analyse whether there are differences between countries that are classified below or above the international standard of practice adoption. The results of these Cox regressions are displayed in Table 6 below. Each outcome is a result of a separate Cox regression, no model is used that included multiple independent variables simultaneously. For the sake of convenience, only the hazard rates of the independent variables are displayed. The Cox regressions displayed in Table 6 are performed with the independent variables measured at the cross-national level.

Table 6 Cox regression analyses for countries below and above the international standard of adoption prevalence.

Countries below international adoption standard Countries above international adoption standard Network centrality 3.038** (1.359) -0.059 (0.402) Interlocks 0.645*** (0.193) -0.077 (0.127) Ties to adopters 0.421** (0.192) -0.028 (0.129) International interlocks 0.448** (0.196) (0.132) 0.177 Standard errors in parentheses. *p<0.10 **p<0.05 *** p<0.01

(32)

international standard are more valuable than interlocks originated from countries in which the adoption prevalence is below the international standard.

Because the diffusion of the GRI practices already started in 1999 and the timeframe I investigate only runs from 2007 to 2013, it is highly likely that at the starting point of my data, not all countries are in the same diffusion phase. In some countries, organisations may still be perceived as innovators by adopting GRI practices, whereas in other countries organisations that adopt the GRI practices belong to the majority adopters. As a robustness check and in order to control for the fact that some organisations might have already adopted GRI practices before I started to measure the practice adoption among organisations, I investigate whether the outcomes from Table 4 and 5 remain stable when only firm-observations are included for countries in which the GRI adoption prevalence is above the threshold of 20%. This means that firm-observations in countries in which the GRI adoption percentage is below the threshold of 20% are excluded from the sample. By doing so, the sample will be better represented by organisations in which the GRI practice adoption follows a similar pattern. The Cox regressions are performed on the new sample that is obtained and the outcomes of the Cox regressions are shown in Table 7 and 8 of Appendix B.

(33)

6 CONCLUSION

The primary aim of this paper is to increase our understanding of how the interlocking directorate network influences the adoption and diffusion of sustainability practices. To this end, the diffusion and adoption of the GRI practices are investigated. It is argued that the board of directors discloses information to respond to pressures from stakeholders and legitimize their actions. Due to the popularity of disclosing sustainability information in corporate documents, the need to be able to compare sustainability disclosures of companies and frameworks to assess the quality of the disclosed information emerged. Independent third-party institutions played an important role to ensure the quality of sustainability reporting and could provide organisations legitimacy by membership in a community. An institution that provides the most-used practices worldwide is the Global Reporting Initiative. It issued its first version of the GRI practice in 1999 and now have almost 4,000 organisations that make use of the GRI practices. There is limited research in the adoption of sustainability practices and the diffusion of voluntary practices. Furthermore, research on the diffusion of practices through international interlocks is scarce. By examining the GRI practice adoption within and between 14 European countries, the diffusion of a voluntary sustainability practice is investigated on a national and cross-national level. By looking at 292 European companies for the years 2007-2013, it can be concluded that several characteristics of the interlocking directorate network significantly influence an organisation’s rate of GRI practice adoption.

Overall, central organisations in the interlocking network of the country have an increased probability of adopting GRI practices. However, this does not hold when the international interlocking network is considered. This outcome indicates that, as an organisation, being central in a country network is more important than being central in the international network. Since rules and regulations are country-specific and one country might place greater importance on pursuing the environmental agenda than the other, this could explain the difference in the analyses. By being a central organisation in a country network, information on new rules, practices or future expectations will likely to be communicated to the focal organisation earlier than to other organisations more distant in the country network. This provides the central organisations with an informational advantage and high chance to adopt new practices first. Since this is much less the case in the international network, being central as an organisation in the international network does not have to result in informational advantages per se.

(34)

As the international interorganisational ties increase the interconnectedness of the total network and are proven to be important links that connect the multiple ‘cliques’ of country networks (Granovetter, 1973), international interlocks are important weak ties that have a profound influence on the adoption of voluntary practices.

Firms with ties to prior adopters of the GRI practices do not have a higher probability of adopting the same practices, both in the country network as well as in the international network. This is an interesting observation as prior studies showed significant effects (e.g. Davis, 1991; Byun & Kim, 2017). Multiple reasons might explain why I found an insignificant effect. First, as the GRI practices have become a global standard over time and have become well-known, many of the public organisations are likely to be aware of the GRI’s presence which makes the need to know prior adopters of the GRI practices in the interlocking network less influential. Second, as the GRI practices are voluntary practices and the benefits and costs to implement the practices are often unknown by organisations and case-specific, organisations might be more reluctant to adopt the GRI practices even if directors’ first-hand experiences about the GRI practices are known. Third, I was unable to incorporate data on the beginning of the diffusion of the GRI practices in my database and investigated the latter part of ‘the diffusion cycle of the GRI’, in which ties to prior adopters might be less important. As the newness of the GRI practice is larger in the early years of the introduction of the GRI practices, it is likely that direct ties to GRI practice adopters are more valuable in this stage than in later stages in which the greater part of the organisations acknowledge the GRI’s presence.

(35)

perception on which interlock ties are important for the diffusion of a practice. Additionally, even though the importance of international interlocks has been observed in this study, more research on the subject of international interlocks is needed in order to know what consequences the intensification of the international interlocking network might have (e.g. Heemskerk, 2010). Besides, studies taking a worldwide perspective of the interlocking directorate network might be able to provide information on what impact intercontinental interlocks might have in the diffusion and adoption of practices among organisations and countries. Last, although the GRI practices are increasingly being adopted by organisations in countries over time, it might be the case that certain organisations or countries decide to reject the adoption of practices or fall behind due to unknown reasons. As mentioned in the work by Etzion (2014), it might be interesting to examine whether a classification can be made of the members of a social system. This would further increase our knowledge about diffusion and allows us to observe what characteristics benefit or impair successful practice adoption.

6.1. Managerial Implications

(36)

6.2. Limitations

(37)

REFERENCES

Ackerman, R., & Bauer, R. (1976). Corporate Social Responsiveness: The Modern Dilemma. Cambridge: Harvard University Press.

Allison, P. (1982). Discrete-Time Methods for the Analysis of Event Histories. Sociological Methodology, 13, 61-98.

Alonso-Almeida, M. (2009). Use of ICTs to Generate Trust in Shareholders and Investors in Companies. Argos, 26(51), 6-26.

Barzuza, M., & Curtis, Q. (2015). Board Interlocks and Corporate Governance. Delaware Journal of Corporate Law, 39, 669-701.

Brown, H., de Jong, M., & Levy, D. (2009). Building Institutions Based on Information Disclosure: Lessons from GRI's Sustainability Reporting. Journal of Cleaner Production, 17(6), 571-580.

Burt, R. (1987). Social Contagion and Innovation: Cohesion versus Structural Equivalence. American Journal of Sociology, 92, 1287-1335.

Bushee, B., Matsumoto, D., & Miller, G. (2003). Open Versus Closed Conference Calls: The Determinants and Effects of Broadening Access to Disclosure. Journal of Accounting and Economics, 34, 140-180.

Byun, H., & Kim, T.-H. (2017). Identity Claims and Diffusion of Sustainability Report: Evidence from Korean Listed Companies, 2003-2010. Journal of Business Ethics, 140, 551-565. Cai, Y., Dhaliwal, D., Kim, Y., & Pan, C. (2014). Board Interlocks and the Diffusion of

Disclosure Policy. Review of Accounting Studies, 19, 1086-1119.

Campbell, J. (2007). Why Would Corporations Behave in Socially Responsible Ways? An Institutional Theory of Corporate Social Responsibility. The Academy of Management Review, 32(3), 946-967.

Carroll, A. (1989). Business and Society: Ethics and Stakeholder Management. Cincinnati: South-Western College Publishing Company.

Chan, M., Watson, J., & Woodliff, D. (2014). Corporate Governance Quality and CSR Disclosures. Journal of Business Ethics, 125(1), 59-73.

Cox, D. (1972). Regression Models and Life Tables. Journal of the Royal Statistical Society, 34, 187-220.

Daub, C.-H. (2007). Assessing the Quality of Sustainability Reporting: An Alternative Methodological Approach. Journal of Cleaner Production, 15, 75-85.

Davis, G. (1991). Agents Without Principles? The Spread of the Poison Pill through the Intercorporate Network. Administrative Science Quarterly, 36(4), 583-613.

Davis, G., & Greve, H. (1997). Corporate Elite Networks and Governance Changes in the 1980s. American Journal of Sociology, 103(1), 1-37.

DiMaggio, P., & Powell, W. (1983). The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organisational Fields. American Sociological Review, 48(2), 147-160.

Etzion, D. (2014). Diffusion as Classification. Organisation Science, 25(2), 420-437.

Fiss, P., & Zajac, E. (2004). The Diffusion of Ideas over Contested Terrain: The (Non)adoption of a Shareholder Value Orientation Among German Firms. Administrative Science Quarterly, 49, 501-534.

(38)

Fligstein, N. (1990). The Transformation of Corporate Control. Cambridge: Harvard University Press.

Fuente, J., García-Sánchez, I., & Lozano, M. (2017). The Role of the Board of Directors in the Adoption of GRI Guidelines for the Disclosure of CSR Information. Journal of Cleaner Production, 141, 737-750.

Galaskiewicz, J., & Burt, R. (1991). Interorganisational Contagion in Corporate Philanthropy. Administrative Science Quarterly, 36, 88-105.

Galaskiewicz, J., & Wasserman, S. (1989). Mimetic Processes Within an Interorganisational Field: An Empirical Test. Administrative Science Quarterly, 34, 454-479.

Global Reporting Initiative. (2017, May 15). About GRI. Retrieved May 15, 2017, from Empowering Sustainable Decisions: www.globalreporting.org/aboutGRI

Global Reporting Initiative. (2017). Sustainability Disclosure Database. Amsterdam: Global Reporting Initiative.

Granovetter, M. (1973). The Strength of Weak Ties. American Journal of Sociology, 78, 1360-1380.

Granovetter, M. (1985). Economic Action and Social Structure: The Problem of Embeddedness. American Journal of Sociology, 91, 481-510.

Hambrick, D. (2007). Upper Echelons Theory: An Update. Academy of Management Journal, 32(2), 334-343.

Hambrick, D., & Mason, P. (1984). Upper Echelons: The Organisation as a Reflection of Its Top Managers. Academy of Management Journal, 9(2), 193-206.

Haunschild, P. (1993). Interorganisational Imitation: The Impact of Interlocks on Corporate Acquisition Activity. Administrative Science Quarterly, 38, 564-592.

Haunschild, P., & Beckman, C. (1998). When Do Interlocks Matter? Alternative Sources of Information and Interlock Influence. Administrative Science Quarterly, 43, 815-844. Heemskerk, E. (2010). The Social Field of the European Corporate Elite: A Network Analysis of

Interlocking Directorates Among Europe's Largest Corporate Boards. Global Networks, 0-27.

Hess, D. (2008). The Three Pillars of Corporate Social Reporting as New Governance Regulations: Disclosure, Dialogue, and Development. Business Ethics Quarterly, 18, 447-482.

Junior, R., Best, P., & Cotter, J. (2014). Sustainability Reporting and Assurance: A Historical Analysis on a World-Wide Phenomenon. Journal of Business Ethics, 120, 1-11.

Lamb, N., & Roundy, P. (2016). The "Ties that Bind" Board Interlocks Research: A Systematic Review. Management Research Review, 39(11), 1516-1542.

Larcker, D., So, E., & Wang, C. (2013). Boardroom Centrality and Firm Performance. Journal of Accounting & Economics, 55(2), 225-250.

Lozano, R., & Huisingh, D. (2011). Inter-linking Issues and Dimensions in Sustainability Reporting. Journal of Cleaner Production, 19, 99-107.

Meyer, J., & Rowan, B. (1977). Institutionalized Organisations: Formal Structure as Myth and Ceremony. American Journal of Sociology, 83(2), 340-363.

(39)

Ortiz de Mandojana, N., & Aragon-Correa, J. (2015). Boards and Sustainability: The Contingent Influence of Director Interlocks on Corporate Environmental Performance. Business Strategy and the Environment, 24, 499-517.

Perez-Batres, L., Miller, V., & Pisan, J. (2010). CSR, Sustainability and the Meaning of Global Reporting for Latin American Corporations. Journal of Business Ethics, 91, 193-209. Pfeffer, J., & Salancik, G. (1978). The External Control of Organisations: A Resource

Dependence Perspective. New York: Harper & Row.

Prado-Lorenzo, J., Gallego-Alvarez, I., & Garcia-Sanchez, I. (2009). Stakeholder Engagement and Corporate Social Responsibility Reporting: The Ownership Structure Effect. Corporate Social Responsibility and Environmental Management, 16, 94-107.

Rogers, E. (1983). Diffusion of Innovations. New York: The Free Press.

Shipilov, A., Greve, H., & Rowley, T. (2010). When Do Interlocks Matter? Institutional Logics and the Diffusion of Multiple Corporate Governance Practices. Academy of Management Journal, 53(4), 846-864.

Shropshire, C. (2010). The Role of the Interlocking Director and Board Receptivity in the Diffusion of Practices. Academy of Management Review, 35(2), 246-264.

Skouloudis, A., Evangelinos, K., & Kourmousis, F. (2009). Development of an Evaluation Methodology for Triple Bottom Line Reports Using International Standards on Reporting. Environmental Management, 44, 298-311.

Staples, C. (2007). Board Globalization in the World's Largest TNCs 1993-2005. Corporate Governance: An International Review, 15(2), 311-321.

Strang, D., & Soule, S. (1998). Diffusion in Organisations and Social Movements: From Hybrid Corn to Poison Pills. Annual Review of Sociology, 24, 265-290.

Surroca, J., Tribó, J., & Zahra, S. (2013). Stakeholder Pressure on MNEs and the Transfer of Socially Irresponsible Practices to Subsidiaries. Academy of Management Journal, 56(2), 549-572.

Useem, M. (1984). The Inner Circle: Large Corporations and the Rise of Business Political Activity in the U.S. and U.K. New York: Oxford University Press.

Van Veen, K., & Marsman, I. (2008). How international are executive boards of European MNCs? Nationality diversity in 15 European countries. European Management Journal, 26, 188-198.

Westphal, J., & Zajac, E. (1997). Defections From the Inner Circle: Social Exchange, Reciprocity, and the Diffusion of Board Independence in U.S. Corporations. Administrative Science Quarterly, 42, 161-183.

(40)

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 GR I A do pt io n United Kingdom The Netherlands Germany Belgium France Italy Luxembourg Denmark Ireland Greece Portugal Spain Austria Finland Sweden 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 GR I A do pt io n Construction (n=9) Energy (n=24) Financial Services (n=58) Information & Communication (n=35) Manufacturing (n=106) Mining (n=11) Other (n=18) Transport (n=11)

Wholesale & Retail (n=20)

Appendix A – Diffusion pattern of GRI practices by Industry and Country

The diffusion pattern of GRI practice adoption among large publicly traded organisations by industry classification and country.

Figure 5Diffusion pattern of GRI practice adoption among large publicly traded organisations by industry.

Figure 6Diffusion pattern of GRI practice adoption among large publicly traded organisations by country.

(41)

Appendix B – Additional Cox regression analyses

Table 7 Cox regression analysis of GRI Adoption (National level – threshold of country adoption > 20%)

(1) (2) (3) (4) (5) (6) Network centrality 0.528** (0.250) 2.029*** (0.620) Interlocks (log) 0.156 (0.139) 0.218 (0.370)

Ties to adopters (log) -0.021

(0.139) -1.111*** (0.298) Industry adopters 5.949*** (0.573) 6.238*** (0.594) Number of employees (log) 0.360***

(0.064) 0.345*** (0.064) 0.350*** (0.064) 0.362*** (0.064) 0.329*** (0.071) 0.342*** (0.072) Return on Assets (ROA) 0.011

(0.010) 0.010 (0.010) 0.011 (0.010) 0.011 (0.010) 0.005 (0.010) 0.004 (0.011) Government ownership 0.079 (0.180) 0.107 (0.181) 0.093 (0.181) 0.077 (0.180) 0.024 (0.206) 0.086 (0.210) Foreign ownership -0.039 (0.163) -0.041 (0.164) -0.036 (0.163) -0.39 (0.163) 0.038 (0.184) 0.039 (0.187)

Industry dummies Yes Yes Yes Yes Yes Yes

Country dummies Yes Yes Yes Yes Yes Yes

Constant -4.291*** (0.782) -4.286*** (0.781) -4.313*** (0.782) -4.295*** (0.783) -7.677*** (0.951) -8.271*** (0.972) c2 176.87 181.28 178.16 176.90 387.10 410.75 D.f. 24 25 25 25 25 28

(42)

Table 8 Cox regression analysis of GRI Adoption (Cross-national level – threshold of country adoption > 20%) (1) (2) (3) (4) (5) (6) (7) Network centrality 0.584 (0.509) -0.468 (0.705) Interlocks (log) 0.292** (0.131) 0.460* (0.274)

Ties to adopters (log) 0.138

(0.128)

-0.452* (0.241)

International interlocks (log) 0.337***

(0.124) 0.334** (0.176) Industry adopters 4.480*** (0.811) 4.489*** (0.819) Number of employees (log) 0.360***

(0.064) 0.347*** (0.065) 0.326*** (0.064) 0.345*** (0.065) 0.315*** (0.064) 0.376*** (0.063) 0.340*** (0.065) Return on Assets (ROA) 0.011

(0.010) (0.010) 0.010 (0.010) 0.011 (0.010) 0.010 (0.010) 0.010 (0.010) 0.011 (0.010) 0.012 Government ownership 0.079 (0.180) 0.105 (0.181) 0.109 (0.181) 0.090 (0.180) 0.101 (0.180) -0.030 (0.188) -0.012 (0.189) Foreign ownership -0.039 (0.163) -0.047 (0.164) -0.029 (0.163) -0.038 (0.163) -0.026 (0.164) -0.066 (0.167) -0.035 (0.167)

Industry dummies Yes Yes Yes Yes Yes Yes Yes

Country dummies Yes Yes Yes Yes Yes Yes Yes

Constant -4.291*** (0.782) -4.142*** (0.790) -4.295*** (0.775) -4.240*** (0.780) -4.015*** (0.770) -6.747*** (0.913) -6.835*** (0.939) c2 176.87 178.18 181.99 178.04 184.29 205.81 216.34 D.f. 24 25 25 25 25 25 29

(43)

Referenties

GERELATEERDE DOCUMENTEN

Whether foreign currency denominated public debt is associated with an increase in the Value at Risk of debt sustainability thus depends on whether a positive interest rate

Concerning the results of the second objective - to evaluate a frame-of-reference training programme for assessors at assessment centres – the evaluation and assessment

The AREA % values from the GC traces were divided by the response factors listed above and then normalized.. These mass percentages were then divided by the

relative bad road safety situation and for cooperation in research for an effective road safety in Europe the national road safety research insti- tutes took in

Handiger lijkt aan te sluiten bij terminologie die ook in recente beleidsnota's (Vijno, Natuur voor mensen, mensen voor natuur) is terug te vinden: monofunctionele landbouw

This survey-based study examines the relationship between the sustainability practices and financial performance for Small Medium Enterprises (SMEs).. Literature proposed that for

When buying firms implement competitive incentives to motivate suppliers for social sustainability performance improvement, these firms usually promise suppliers future business

Sustainability performance indicator Impact on the environment the energy usage per unit of production the amount of raw materials required per unit of production