Codes of conduct – social responsibility or window-dressing? A case study o

(1)

A case study of DONG Energy’s code of conduct:

What are the strengths and weaknesses for ensuring social responsible behaviour of coal suppliers?

Master’s Thesis

Master of Arts in

International Business Communication

Department of Cand.ling.merc Lis-Wenke Mero

Thesis advisor: Jette Steen Knudsen Department of Business & Politics

December 1st 2011 Pages: 79.4

Number of characters:180.791 Copenhagen Business School

(2)

Acknowledgements

I would like to express my deep gratitude to my thesis advisor, Jette Steen Knudsen, for guidance, inspiration and support throughout the process of developing this thesis.

I would also like to thank each and every respondent who despite their busy scheduals took the time to share their expertise and experience on the subject with me in a highly obliging manner:

Kristian Heydenreich, DONG Energy

Sanne Borges, Amnesty International Denmark Linna Palmqvist, PricewaterhouseCoopers

Peter Thagesen, Confederation of Danish Industry Jesper Nielsen, 3F

Carsten Hansen, 3F

Lis-Wenke Mero

(3)

Etiske regelsæt – social ansvarlighed eller marketing?

- et casestudie af styrker og svagheder ved DONG Energy’s etiske regelsæt

Siden globaliseringen tog fart i 1980‟erne og 90‟erne har politiske aktivister afsløret utallige sager, om hvordan virksomheders jagt på billig arbejdskraft og lave produktionsomkostninger i udviklingslande har ført til kritisable arbejdsvilkår og menneskerettighedskrænkelser.

Medieskandaler og politisk aktivisme har tvunget virksomheder til at implementere etiske regelsæt, som et konkret forsøg på at regulere egne aktiviteter i områder med svag lovgivning.

Etiske regelsæt beskrives teoretisk og af virksomheder selv som social ansvarlighed, hvor virksomheder bidrager til en ansvarlig leverandørkæde ved at stille specifikke krav til sine leverandører om bl.a. arbejdsforhold, helbred og sikkerhed, anti-korruption og miljø. Men på baggrund af det stadigt stigende antal skandaler i leverandørkæder trods etiske regelsæt, kritiseres etiske regelsæt samtidig for at blive brugt som et marketing-værktøj af virksomheder, der forsøger at minimere omdømmerisici, politisk aktivisme og lovgivning.

De mange positive og negative vurderinger af etiske regelsæt vækker en række spørgsmål: Hvad er styrker og svagheder ved etiske regelsæt? Kan etiske regelsæt anvendes til at øge leverandørers socialt ansvarlige adfærd? Og hvis ja, hvordan? Baseret på DONG Energy som et casestudie tager denne afhandling udgangspunkt i syv kvalitativt semi-strukturerede interviews kombineret med teori af Professor David Vogel og Virginia Haufler, som fokuserer på virksomheders etiske regelsæt på et samfundsmæssigt plan, samt teori af Professor Richard Locke og Dara O‟Rourke, som fokuserer på etiske regelsæt på organisatorisk plan, for at udforske styrker og svagheder ved DONG Energy‟s etiske regelsæt for kulleverandører.

Analysen viser, at flere af de teoretisk antagede styrker også gør sig gældende i forhold til DONG‟s etiske regelsæt. Sammenlignet med andre energivirksomheder i Danmark, har DONG et veludviklet og detaljeret regelsæt, der stiller specifikke krav til leverandørernes socialt ansvarlige adfærd. Gennem et monitoreringssystem bliver leverandørernes efterlevelse af regelsættet kontrolleret af eksterne auditorer, som gennemfører audit hos udvalgte leverandører.

Udover at audittenes varighed er 29 dage længere end teoretikere påpeger, undergår DONG‟s monitoreringssystem desuden en proces af kontinuerlige forbedringer. Analysen viser derudover,

(4)

at DONG‟s etiske regelsæt sætter i gang en proces af kontinuerlige forbedringer hos udvalgte leverandøerer samt bidrager til, at leverandører implementerer socialt ansvarlige tiltag.

Men samtidig viser analysen, at teoretiske antagelser om svagheder også gør sig gældende for DONG. Idet regelsættet fokuserer på principper fremhævet af politiske aktivister og medier og udelader principper af høj værdi for arbejdere og lokalområder, fremstår regelsættet som værende mere målrettet interessenter i Danmark end interessenter i leverandørkæden – hvilket peger i retning af en marketingstrategi. Analyser visen også, at flere af regelsættets principper ikke bliver overholdt. Derudover viser analysen, at selv om regelsættet generelt er accepteret som reguleringsgrundlag, øges presset på Finansministeriet, der er DONG‟s ejer, i takt med at kritikken om dårlige arbejdsforhold i DONG‟s leverandørkæde stiger – hvilket øger risikoen for yderligere lovgivning på området.

Det konkluderes derfor, at DONG‟s etiske regelsæt tilvejebringer styrker både på det organisatoriske og det samfundsmæssige plan og bidrager til social ansvarlig adfærd hos leverandører. På baggrund af væsentlige svagheder, der har negativ indflydelse på DONG‟s omdømme og sociale ansvarlighed, bør regelsættet dog optimeres for at opnå størst mulig effekt og for at undgå, at det fremstår som blot et led i en eventuel marketingstrategi. Et løsningsforslag er, at inkludere principper af væsentlig betydning for interessenter i leverandørkæden i regelsættet samt justere fremgangsmåden fra en traditionel model, hvor efterlevelse af regelsættet forsøges opnået med trusler, til en såkaldt ‟engageret-fokuseret‟ strategi, hvor efterlevelse sikres ved at involvere vigtige interessenter i en kontinuerlig proces med dialog, analyse af bagvedliggende årsager til brud på principper, informationsdeling m.m. for at styrke leverandørernes evne og vilje til at efterleve regelsættet. Det foreslås desuden, at DONG gennem det nyetablerede multi-stakeholder initiativ, Better Coal, som fokuserer på social ansvarlighed i kulleverandørkæden, lægger pres på relevante regeringer for at bidrage til øget social ansvarlighed. I sidste ende er social ansvarlighed nemlig et anliggende af national karakter, hvorfor DONG ikke på egen hånd kan sikre overholdelse af arbejdstager- og menneskerettigheder.

(5)

1. Introduction

Due to globalisation during the 1980s and 90s, multinational companies‟ (MNCs) economic power increased while developing country governments‟ capabilities to regulate MNCs decreased. Companies were accused of skipping lightly from country to country in search of the most accommodating political environment (Haufler 2001:7). Many MNCs outsourced to countries with low or no governmental regulation where taxes were low and labour cheap, which led to major corporate scandals such as Nike‟s sweatshops in Vietnam, IKEA‟s child labour in India, and Shell‟s operations in Nigeria. In response to heavy negative public criticism and pressure, MNCs have established codes of conduct as a concrete effort to regulate themselves and control working conditions in markets wherein they operate. Codes of conduct are in other words not enforced by any state, but a part of corporate self-regulation, which occurs when companies design and enforce rules that govern their behaviour, either going beyond current regulatory requirements or establishing new standards (Haufler 2001:8). Monitoring mechanisms are developed together with codes of conduct to measure suppliers‟ compliance with the codes (Locke et al. 2009). Today, codes of conduct have grown immensely in number and scope;

however, corporate misbehaviour continues to hit the surface. Consequently, academics, businessmen, activists, and theory on codes of conduct discuss whether codes of conduct are a genuine effort to improve working conditions in the supply chain or a marketing ploy to avoid negative public attention and protect reputation.

The Danish energy company DONG Energy (from hereon DONG), which is state-owned, has been heavily critisiced in the media in relation to violations of labour and human rights in the coal supply chain. In 2006, DONG was subjected to negative critisicm for sourcing coal from the mine company Drummond in Colombia because Drummond was on trial for the killing of three unionists. After negative attention and stakeholder pressure, DONG terminated the business relationship. In 2007, DONG established a code of conduct for suppliers to ensure compliance with labour and human rights principles. However, in January 2010, DONG was heavily criticised for not complying with its own code of conduct due to a report published by the corporate watchdog DanWatch. The report revealed poor working conditions, labour persecution, forcefully relocation and isolation of local inhabitants, and corportate manipulation by DONG‟s supplier Cerrejón in Colombia. In December 2010, DONG was again frontpage news with accusations about poor working conditions in Russian coal mines where workers got

(10)

disabled and killed while producing coal to Danish consumers (Hvilsom 26.12.10a). Again, DONG was publicly accussed of not living up to its own code of conduct. In other words, DONG has established a code of conduct to respect labour and human rights and ensure proper working conditions in its supply chain, but is yet subjected to negative stakeholder attention in relation to this. Therefore, DONG‟s code of conduct for coal suppliers sets out the framework for the research set out in this thesis as I wonder: Can DONG‟s code of conduct really enhance social responsible behaviour of suppliers and thereby improve working conditions and prevent violations of human rights in the supply chain? Or, are critics right when they dismiss DONG‟s code as a marketing ploy? Why is DONG‟s code not complied with? Can DONG claim to be socially responsible when the company sources coal from mines in high-risk areas where corruption, threats, persecution, hazardous working conditions, and murder are everyday occurrences?

Theory on codes of conduct

Theory on codes of conduct highlights theoretical assumptions of strengths and weaknesses.

Proponents claim that codes of conduct have become a new dimension of corporate governance which contributes positively to social and environmental issues in developing countries wherein companies operate, as well as gaining competitive advantages for companies. For instance, Professor David Vogel argues that codes of conduct have “become a highly visible and increasingly legitimate dimension of global economic governance” (2010:68). Vogel states that from a business perspective “the adoption of codes of conduct has been effective. Reports of abuses continue to surface periodically, but the improvements companies have made or have assumed responsibility for making appear to have been sufficient to defuse or reduce most public criticism” (2005:109).

Likewise, Professor Virginia Haufler argues that codes of conduct respond to societal pressure, avoid rigid governmental interference, and balance interests of business and society without expanding government intervention in the economy (2001:3-4). She documents the changing role of business from private towards public due to the growth of codes of conduct as corporate self-regulation, which she refers to as: “a new source of global governance, that is, mechanisms to reach collective decisions about transnational problems with or without government participation” (Haufler 2001:1). Nevertheless, Haufler is at the same time skeptical towards the

(11)

real effect of codes due to weaknesses; problems addressed in codes of conduct are often problems of national governance wherein most of the responsibility rests (2001:121). She also argues that codes of conduct work best within political systems that encourage it and poorly when political systems work against it (ibid). Professor Dara O‟Rourke and Professor Richard Locke are much more sceptical towards codes of conduct as they find that results have been limited and monitoring methods significantly flawed. In his report “Monitoring the monitors”

(2000), O‟Rourke produces evidence of monitoring methods overlooking major violations and codes of conduct not being complied with. As a result, he concludes that; “independent monitoring can play a positive role in improving factory conditions, but only if it is much more transparent and accountable, includes workers more fully, and can be verified by local NGOs and workers themselves” (ibid:1). Locke et al also argue that codes of conduct and monitoring efforts have produced modest improvements in working conditions in the supply chain due to misguided assumptions about the power of MNCs in supply chains, the role of information from audits, and suitable incentives required to improve the conditions (2009).

The United Nation‟s Special Representative for Business and Human Rights, Professor John Ruggie, has recently set forth a framework that addresses problems in connection with codes of conduct. The „Protect, Respect and Remedy‟ Framework proposes measures to strengthen the human rights performance of the business sector around the world (2011). The Framework addresses consequences and issues of the dramatic expansion of the private sector and the challenges of bringing human rights violations to an end in a world with 192 UN Member States, 80,000 MNCs, 10 times as many subsidiaries, and countless millions of national firms (ibid).

What is especially interesting in this Framework is the emphasis on states‟ duty to protect human rights, and that additional steps must be taken when companies are controlled or owned by states.

Clearly, codes of conduct are of great importance for contemporary business but are called into question as non-compliance and corporate misbehaviour continue to persist. Little is known about the real effect of codes of conduct, and at the same time the aspects of transparency and credibility is questioned. Therefore, the research in this thesis is set out to investigate strengths and weaknesses of DONG‟s code of conduct through an explorative research in order to gain an in-depth understanding of codes of conduct in the supply chain.

(12)

1.1 Motivation

The motivation for writing this thesis is based on a theoretical wonder that arose from readings on codes of conduct. Based on theoretical assumptions that codes can improve working conditions in the supply chain, I wonder how this is compatible with reality as scandalous revelations about labour and human rights abuses continue to hit the surface. Theory describes that companiess have acknowledged their corporate social responsibility in the supply chain and therefore establish codes of conduct to improve conditions. But since abuses prevail: are companies genuinely interested in improving conditions? Or is it a marketing ploy for glossy CSR reports and window-dressing to prevent external pressure and protect reputation? Or is the real reason behind abuses suppliers‟ lack of willingness to comply with code requirements? How can companies prevent suppliers from covering up poor working conditions – and how can they reveal them when they take place? In fact; is it at all social responsible behaviour to source from a country where labour and human rights abuses are everyday occurrences? And what about states; can they continue to hide behind „decreased capability to control corporate behaviour‟?

In order to further examine these thoughts about codes of conduct, I had to find a company that has acknowledged the need for a responsible supply chain and is operating with a code of conduct to achieve this. To explore the question of „genuinely interested vs. marketing ploy‟ I found it relevant to find a company struggling with this issue. Consequently, I found DONG to be a perfect match for a case study: Despite its good intentions and code of conduct DONG is struggling with public accusations of ignoring poor working conditions in coal mines in Colombia and Russia.

My wondering contains many questions; however, in order to study codes in depth, the main focus of this thesis is to explore strengths and weaknesses of DONG‟s code of conduct in light of theoretical assumptions. Hence, the aim of this thesis is to conduct an analysis of DONG‟s code of conduct. As the critique of DONG‟s code has been targeted coal suppliers, the analysis will focus only on DONG‟s code of conduct for suppliers in the coal supply chain. If relevant, empirical findings may be used to develop a proposal of recommendation. Further details on the reasons for my choice of company will be elaborated on in section 2.1. Next, the research question will be established.

(13)

1.2 Research Question

The aim of the research set out in this thesis is to explore the following research question:

What are the strengths and weaknesses of DONG Energy’s code of conduct for ensuring social responsible behaviour of its coal suppliers? Can the code of conduct contribute to improve social responsible behaviour in DONG’s coal supply chain¹, and, if so, how?

1.2.1 Investigative questions

To answer the research question, some underlying factors must be examined. Thus, the following investigative questions have been developed with the purpose of ensuring an answer to the overall research question.

1.) What are the strengths and weaknesses of DONG‟s:

a) Code of conduct?

b) Monitoring efforts?

2.) How is DONG Energy practicing its code of conduct:

Traditional compliance model or commitment-oriented approach?

1.2.2 Clarification of questions

Each of the investigative questions seeks to explore different factors emphasised in the chosen theory that influence codes of conduct effectiveness. Sub-question 1a seeks to explore strengths and weaknesses of the code and analyse its characteristics according to theory by Rob van Tulder et al., and Ruth Pearson and Gill Seyfang. In addition, the code‟s strengths and weaknesses in relation to a societal level will be analysed based on theory by David Vogel and Virginia Haufler. Sub question 1b explores strengths and weaknesses of monitoring for compliance based on theory by Dara O‟Rourke and Richard Locke. Lastly, sub-question 2 is developed based on theory by Richard Locke with the purpose of exploring how DONG is using information, power, and incentives in relation to compliance.

The questions are developed based on theory and seek to explore DONG‟s code of conduct to achieve a nuanced and in-depth understanding of empirical problems in connection with DONG‟s code within its reality and context. Thus, the questions seek to reveal contradictory

1 Based on BSR‟s definition, supplier in this thesis refers to a company that sells goods, including raw materials, to DONG and supply chain refers to the network of suppliers that sells goods to DONG (Sisco et al. 2010). First-tier supplier is a supplier who sells directly to DONG, while second-tier supplier refers to a supplier‟s sub-supplier, and

(14)

aspects and perceptions of the code to gain a genuine picture of DONG‟s code in order to answer the overall research question. Strengths are viewed as factors contributing to enhance social responsible behaviour in the supply chain: compliance, transparency, credibility, information, incentives, and power. Weaknesses are viewed as factors limiting and challenging the effort to enhance social responsible behaviour in the supply chain: non-compliance, lack of transparency and credibility, falsified information, incentives, and power.

If relevant, research findings will be used to develop a proposal of recommendation on how DONG‟s code can contribute to improve social responsible behaviour of its coal suppliers. The empirical findings will be based on qualitative semi-structured interviews and relevant documents.

1.3 Delimitation

The subject „corporate self-regulation‟ contains a broad range of different voluntary codes, standards, certification, and labelling systems; however, this thesis will only focus on codes of conduct in relation to the research question.

DONG has a broad range of business activities in different markets, such as oil, gas, energy, and mining. Nevertheless, this thesis will focus on DONG‟s code of conduct for suppliers in the coal supply chain, more precicely in Colombia and Russia since revelations about poor working conditions have been found here. It could be relevant to analyse DONG‟s organisational culture as well as country-specific culture in relation to Colombia and Russia to gain a deeper understanding of human rights violations and how to tackle this. However, this will not be included here.

1.4 Disposition and Structure

In order to give the reader an overview of the thesis‟ disposition and structure, this will briefly be described in the following.

In the chapter to come, DONG is introduced (chapter 2). The chapter elaborates on the reason for selecting DONG as a case study and on factors relevant for the research, including the code of conduct. Chapter 3, which is centered on methodology and research method, describes considerations in relation to theory of science and the reasons behind the chosen methodology

(15)

and methods used to collect empirical data and answer the research question. The chapter is completed with a quality assessment of the thesis. Chapter 4 outlines the theory that forms the foundation for exploring the empirical data in the analysis in chapter 5. The empirical data consists of semi-structured interviews and relevant documents that are analysed based on theoretical assumptions about codes of conduct to explore strengths and weaknesses of DONG‟s code of conduct and whether the theoretical assumptions align with reality. Chapter 6, discussion, focuses on whether, and if so, how DONG‟s code of conduct can contribute to enhance social responsible behaviour of its coal suppliers. After the discussion, chapter 7 forms the overall conclusion of the thesis, and in chapter 8 the research is put into perspective together with future research proposals. Chapter 9 consists of a bibliography, while appendices are gathered in chapter 10.

1.4.1 Formalities

DONG Energy is referred to as DONG in order to lighten the reading and make best possible use of the pages available. In addition, „code‟ or „codes‟ refers to code(s) of conduct in order to make the text more coherent. All citations taken from DONG‟s notes of explanation to the Ministry of Finance, the Finance Minister‟s comments (with the reference „Svar på‟), and selected quotations by the respondents interviewed (Appendix 5) are translated directly from Danish to English.

Consequently, in case of any discrepancy the Danish version prevails. A list of the respondents‟

names is provided in appendix 3, including the different stakeholders interviewed by DR in connection with the two included radio programmes by DR P1 Dokumentar. Each interview is transcribed into Danish text in appendix 5. The transcription of the interviews has been carefully read and marked with specific colours; hence, a description of this coding is provided in appendix 4. The interviews will be available as audio files on the enclosed CD-ROM as well. All references can be found in the bibliography.

(16)

2. Case study: DONG Energy

In this section I will elaborate on the reasons for selecting DONG as a single case study before I move on to describe DONG as a CSR forerunner, the code of conduct, and how contact to DONG was established.

2.1 Reason for choice of company

DONG was chosen as a case study based on the criteria of relevance and accessability.

2.1.1 Relevance

First and foremost I searched for a company engaged in CSR² with a code of conduct for suppliers. DONG has a broad CSR strategy that is integrated into its values and overall business operations. In 2007, DONG developed a code of conduct for its suppliers to ensure social responsible behaviour in the supply chain. Even though DONG is concerned with its social and environmental impacts on society and stakeholders (Appenix 6), DONG is struggling with accusations of poor working conditions at suppliers‟ in Colombia and Russia. Danish newspapers accuse DONG of not living up to its own code of conduct and social responsibility with headlines stating: “DONG buys coal from Russian death mines”³, “Work in explosive mines to make coal to DONG”⁴, “Revelation: You get the heat – they get their arms ripped off”⁵,

“DONG closed their eyes”⁶, “DONG buys coal from perilously mines^”7 – just to mention a few.

Consequently, despite the code of conduct, DONG is subjected to critisicm and is thereby facing the contradicting aspects that theory on code of conduct describes. Therefore, I find that DONG fulfils the criteria of relevance.

2.1.2 Accessability

Many companies accussed of corporate misbehaviour are multinational companies with headquarters in the US and are therefore difficult for me to access. DONG, on the other hand, is

2 “A concept whereby companies integrate social and environmental concerns in their business operations and in their interaction with their stakeholders on a voluntary basis" (Comminucation from…25.10.11).

3 Ekstra Bladet December 25th, 2010

4 Ekstra Bladet December 27th, 2010

5 Ekstra Bladet December, 26th, 2010

6 Politiken December 26th, 2010

7 TV2 Nyhederne December 26th, 2010

(17)

a Danish company with headquarter in Denmark. Hence, DONG fulfils the criteria of accessibility. Moreover, since DONG is a state-owned company and also aims to be open and transparent, information about the company and its code of conduct is available online.

Therefore, I find DONG to provide interesting and contradicting elements necessary to conduct an explorative research on strengths and weaknesses of codes of conduct.

2.2 DONG Energy

This section describes elements of DONG that I find relevant for the research.

DONG Energy was founded as Dansk Naturgas A/S by the Kingdom of Denmark on March 27^th, 1972. It was created as a vehicle to develop Danish energy activities, and as a state-owned company DONG has established Denmark‟s natural gas infrastructure. The company has expanded significantly through organic growth and acquisitions both in Denmark and throughout Europe, and in 2006 the merger of sixDanish energy companies (DONG, Elsam, ENERGI E2, Nesa, Copenhagen Energy, and Frederiksberg Forsyning) led to the foundation of DONG Energy. Today, DONG is one of the leading energy groups in Northern Europe and the business is based on procuring, producing, distributing, and trading in energy and related products in Northern Europe.

2.2.1 DONG Energy as a CSR forerunner

CSR is an integrated part of DONG‟s business strategy as the company endeavours to act responsibly and live up to society‟s expectations every day by conducting business based on its core values: result-oriented, responsible and responsive (Appendix 5). These values are also evident in DONG‟s long-term vision: to provide clean and reliable energy “while at the same time acting responsibly towards our employees, the local population, the environment, and the markets in which we operate” (ibid). During the last couple of years, DONG has engaged in numerous initiatives to conduct and encourage social responsible behaviour. In 2006, DONG signed the UN Global Compact, and the Compact‟s ten principles form an integral part of DONG‟s strategy, operations, and culture (ibid). DONG also reports on these principles in their

(18)

annual report using the Global Reporting Initiative⁸ (GRI Report 2010 N/A). In 2007, DONG established its code of conduct for its suppliers, and in 2008, DONG won the Petroleum Economist Cleaner Energy Initiative Award in the category “Cleaner Energy initiative of the Year 2007” which placed DONG among the world‟s leading companies in developing clean plant technology (DONG Energy wins…N/A). In 2010, The Financial Times nominated DONG for the Environment Award based on “having a green vision but for also having gone beyond offsetting its carbon footprint to considering sustainability on a broader front” (DONG Energy nominated…25.02.10). In 2011, DONG established „Better Coal‟ consisting of a group of major players in the European energy market seeking to create a common European effort to ensure continuous improvement of social and environmental responsibility in the coal supply chain (Better Coal 2011). The aim is to develop a standard for auditing in collaboration with relevant NGOs and coal producers before the end of 2012 (ibid). In 2009 and 2010, DONG‟s total import of coal amounted to approximately 4 million tonnes sourced from different countries (Appendix 7). DONG‟s strategy is to continuously reduce its consumption of coal at Danish power stations to two million tonnes by 2015⁹ (ibid).

2.2.2 DONG Energy’s code of conduct

DONG‟s code of conduct (Appendix 8) forms the motivation behind this thesis and is therefore the target of research. As the code will be analysed in depth in chapter 5, the code‟s purpose and content will briefly be introduced in this sub-section.

As mentioned, DONG established its code of conduct for suppliers in 2007. The code is an integrated part of all contractual relations above DKK 50,000 and is currently adopted by approximately 11,000 suppliers (Redegøresle 06.01.11) wherein 20-25 suppliers are coal suppliers. The code seeks to “encourage dialogue with suppliers in order to promote their commitment to improving their focus on social, environmental and ethical business aspects”

(Responsible supplier management N/A). Independent, third-party on-site audits are conducted at selected suppliers to assess performance and compliance with the code. In cases of non-

8 Global Reporting Initiative is the organisation behind one of the world‟s most recognised sustainability reporting frameworks called GRI Sustainability Reporting Guidelines; www.globalreporting.com

9 Denmark‟s CO₂ emission from coal/peat was 15.7 million tonnes CO2 in 2009 which is higher than Colombia‟s 11.7 million tonnes and 13.5 million tonnes higer than Norway‟s emission from coal/peat (Appendix 6).

(19)

compliance, DONG enters into a dialogue with the supplier to clarify the circumstances and to demand that the conditions are improved (ibid) – this is not stated in the code.

The code is made up of three sections: introduction, general requirements and specific requirements. Introduction briefly presents DONG, its values, and its purpose with the code, which is “to outline in greater detail the standards we expect our suppliers to adhere to”

(Appendix 8). It states that DONG is signatory to the UN Global Compact and expects suppliers to share and adhere to the fundamental principles which are an integral part of the business relationship with individual suppliers. If adhered to, both DONG and suppliers may achieve competitive advantage. DONG will promote ethical, social, and environmental standards through continuous engagement and dialogue. However, in cases of gross and repeated violations, DONG reserves the right to terminate the business relationship.

General requirements state that suppliers are expected to comply with national laws and regulations as well, and suppliers working in DONG operated areas must meet specific quality, health and safety, and environment requirements as well. Furthermore, suppliers are expected to ensure that sub-suppliers comply with the code.

Specific requirements set forth ten specific principles within the area of human rights, labour, environment, and anti-corruption:

1. Remuneration and employment conditions 2. Working hours

3. Freedom of association and the right to collective bargaining 4. Discrimination

5. Harassment and disciplinary measures 6. Child labour

7. Forced labour 8. Health and safety 9. Corruption and bribery 10. Environment

(20)

2.2.3 Establishing contact to DONG Energy

The contact to DONG was established through my network wherein an acquaintance working in the financial department forwarded my question for whom to contact regarding my research to one of her colleagues. The colleague advised me to contact Kristian Heydenreich, Manager of Corporate Responsibility, and provided his email address. As advised, I sent an email to Heydenreich May 25^th (Appendix 9). To gain his attention, I briefly explained that I had received his contact information from one of his colleagues, and that if he was interested I could contribute with information on codes of conduct through my research. I then put forward my enquiry; that I would like to learn about his experience and knowledge on DONG‟s code of conduct. I described in more detail exactly what the research was set out to explore and why I found DONG to be of great interest. To gain his trust, I stated that a confidentiality agreement could be made and I offered to send an interview guide before the meeting. I did so for two reasons: so he could feel confident that my interest in DONG was genuine and so he could reflect upon the questions in advance. Heydenreich replied five days later and we agreed on a meeting June 9^that DONG, Nesa Allé 1, Gentofte. Two days before the meeting, I sent the interview guide (Appendix 10) and explained that it was not meant as a list of questions to be strictly followed but rather as a guideline of themes that I wished to explore. He could also talk about other aspects that he found relevant to the research. I also stated that I wished to obtain a genuine picture of reality; therefore, I was not seeking specific answers. The meeting itself will be described in section 3.6.3.4.

(21)

METHODOLOGY Critical realism & Subjectivism

APPROACH Deductive & inductive

STRATEGY Single case study

CHOICE Mono-method

TIME HORIZON Cross-sectional

TECHNIQUES & PROCEDURES Semi-structured interviews and

document analysis

3. Methodology and research method

Saunders et al. refers to methodology as “the theory of how research should be undertaken” and includes theoretical and philosophical assumptions that research is based upon, whereas methods refer to “the techniques and procedures used to obtain and analyse data” (2009:3). In other words, methodology is the way the researcher views the world (ontology, epistemology, etc.) which underpins the way research is conducted (quantitative or qualitative data collection techniques) and how the collected data is reviewed and utilised.

The following model has been adopted from Saunders et al. (2009:108) and modified to give an overview of the research elements and stages this thesis is based upon. Each stage will be elaborated on in the following sections.

Figure 1: The research stages Adopted from Saunders et al. 2009 and modified

(22)

3.1 Methodology

Saunders et al. argue that methodology is the most important stage in research as it “can have an important impact on the research we decide to pursue and the way in which we pursue it”

(2009:107). Therefore, it is necessary to choose the methodology most appropriate for the research question. As mentioned before, this thesis is based upon a wonder about strengths and weaknesses of codes of conduct that spurred from theory on the subject. The aim of this thesis is to explore DONG‟s code of conduct by pursuing critical realism in order to find the „truth‟

about the code and explore this in light of theoretical assumptions highlighted in the theory chapter. Thus, the aim of this thesis is not to generalise but rather gain an in-depth understanding of this specific code of conduct in this specific company at this specific period of time within the specific context. Critical realism claims that there are two steps to experience the world; the

„thing‟ itself and the sensations it conveys, and the mental processing that goes on after the sensations meets our senses; thus, our knowledge of reality is a result of social conditioning (ibid:115). It could be argued that codes of conduct exist independent of social actors as an objective entity that states the rules suppliers have to adhere to and that codes are more or less similar in all companies, i.e. an objectivistic worldview (ibid:110). However, as the research set forth in this thesis is based on the perspective that codes of conduct emerged due to perceptions and consequent actions of social actors, a subjectivist worldview will be adopted (ibid:111). The subjectivist methodology makes it possible to study details of the situation, which is necessary to understand the reality or the reality working behind the details (ibid). Thus, it is necessary to explore subjective meanings motivating the actions of social actors, also called „social constructivism‟ (ibid, Darmer et al. 2010:142). Through interviews based on subjectivism, the researcher seeks to understand how and what each respondent feels about certain situations, phenomenons, or the alike; consequently, the interviewer must analyse each respondent‟s interpretation (Kvale and Brinkmann 2009:47). As the interviewer cannot become the respondent and get to know the whole truth, the interviewer must gain an insight into the respondent‟s worldview (perception of the matter in question) in order to interpret this and analyse it in relation to theory (Darmer et al. 2010:226). Therefore, the interviewer asks the questions and records the answers but does not discuss them during the interview (Kvale 1997). Instead, clarifying and probing questions may be asked to ensure a correct understanding and analysis of the respondent‟s perception of the matter in question. In this thesis, qualitative interviewes will be conducted to gain an in-depth understanding of perceptions about DONG‟s code and uncover

(23)

its strengths and weaknesses for ensuring social responsible behaviour of its coal suppliers, and to develop a proposal of recommendation on how DONG‟s code can contribute to enhance social responsible behaviour of the coal suppliers. Due to the subjectivistic methodology, other researchers conducting a similar research will possibly reach different conclusions as the analysis is based on a subjectivistic and context based interpretation of respondents‟ perceptions.

This will be explained further in section 3.6.5.

3.2 Research approach

The research approach taken in this thesis is both deductive and inductive. It is deductive in the sense that the research question and the investigative questions are developed based on theory and explore theoretical assumptions in a specific context (Darmer et al. 2010:164). At the same time, an inductive approach is taken to collect empirical data focusing on qualitative aspects in order to gain relevant and in-depth data with reference to explore theory with empirical findings (Darmer et al. 2010:165, Saunders et al. 2009:126). Therefore, the chosen theory has been operationalised in the research question and the interview guides to target the research.

To explain the work process in this mixed approach, a model adopted from Darmer et al. (2010) and modified is provided below. Although this model looks somewhat linear, the work process has not been linear, but more of a process where I had to go back and forth and go through some of the steps several times.

Figure 2: The mixed approach work process Adopted from Darmer et al. 2010 and modified Choose

theory

Outline research question

Select case study and

respondents

Collect

data Analyse data

Theore- tical reflec-

tions

Results &

conclusion

Adjust research questions Find

additional respondents

(24)

3.3 Research strategy

According to Saunders et al. (2009:147) the case study strategy is an effective way to explore existing theory. Likewise, Yin argues that a case study benefits “from the prior development of theoretical propositions to guide data collection and analysis” (2009:18). The purpose of this research is to explore the strengths and weaknesses of codes of conduct; therefore, a single case study of DONG‟s code will be conducted to gain a rich understanding of the code. It can be argued that a multiple case study could strengthen the results due to multiple sources of evidence. However, DONG provides a critical case study because DONG is a CSR forerunner and one of the few energy companies in Denmark that has developed a code of conduct for suppliers. In addition, DONG controls supplier compliance with the code. At the same time, DONG is also subjected to critisicm about not complying with its own code of conduct which is an issue often mentioned in the theory as well as reality. Therefore, I argue that a single case study of DONG will provide the information necessary to conduct valid and reliable research.

Also, as the purpose of this thesis is not to generalise but gain a rich understanding of the subject in question, this single case study will provide detailed and contradicting information about codes of conduct that can be used to problematise the theory (Darmer et al. 2009:151,153). To collect empirical data, different sources will be used giving a nuanced and clear picture of DONG‟s code. The code will therefore be analysed based on qualitative interviews and relevant documents. Herein, two radio programmes by DR P1 Dokumentar retrieved online will be included because they both consist of a qualitative interview with a DONG representative and provide highly relevant information for this research. This will be elaborated on in more detail in section 3.6.3.1. In addition, also to gain an in-depth understanding of DONG‟s code, it will be analysed compared to three other Danish energy companies‟ codes of conduct as well;

Vattenfall, Energi Danmark, and Nordjysk Elhandel. The companies are selected based on comparability; they are the only companies among the ten biggest Danish energy companies that have a code of conduct for suppliers, besides DONG. The companies will not be introduced or described in detail since the case study is DONG.

3.4 Research choice

In connection with data, qualitative is a synonym for any data collection technique or data analysis procedure that generates or use non-numerical data, like words and pictures, whereas quantitative data collection techniques generate or use numerical data (numbers) (Saunders et al.

(25)

2009:151). The research set out in this thesis is based on a mono-method qualitative data collection technique based on qualitative semi-structured interviews.

3.5 Time horizon

Like the research question indicates, this research is a cross-sectional study, i.e. it is conducted over a short period of time producing a „snapshot‟ of the current state of the subject in question (Saunders et al. 2009:155). This also means that the respondents‟ statements reflect perceptions at this given point in time. The empirical data was retrieved from June 9^th – October 11^th_, 2011. I expected this period to be much shorter, but the summer holiday made it difficult to get in contact with respondents.

3.6 Research techniques and procedures

In this sub-section, techniques and procedures in connection with the research will be described.

3.6.1 Target audience

The target audience for this thesis is primarily Copenhagen Business School; however, the analysis and conclusions will be of interest to DONG as well. In addition, DONG‟s stakeholders and other companies within the coal mining industry may find it interesting as well.

3.6.2 Qualitative interviews

The empirical data is based on seven semi-structured interviews. The strength of qualitative interviews is their openness (Kvale 1997:92), which provides the opportunity to gain in-depth knowledge about a specific phenomenon in an organisation described through the eyes of the respondents themselves (Darmer et al. 2010:213, Kvale 1997:109). A semi-structured interview with DONG‟s CR Manager, Heydenreich, was carried out to explore specific aspects and characteristics of DONG‟s code of conduct in practise and as percepted by Heydenreich. My intention was to interview several individuals across different departments within DONG, and compare their perceptions with Heydenreich‟s to gain a deeper understanding of the code‟s

„reality‟. Unfortunately, this was not possible, as the respondents contacted were not interested in participating in the research. I also carried out semi-structured interviews with other respondents external to DONG in order to explore stakeholders‟ perceptions of codes of conduct and social responsible behaviour in the suplly chain. See appendix 3 for a list of respondents.

(26)

A quantitative approach could have provided me with the opportunity to investigate additional and broader aspects of DONG‟s code and stakeholders‟ perceptions. But since my intention is not to gain a broad picture of DONG‟s code, I find the qualitative approach more suitable. The weakness of the qualitative approach is that results are very much influenced by the interaction between the interviewer and the respondent, and the interviewer‟s ability to ask good questions and be a good listener at the same time (Darmer et al. 2010:213). Also, the circumstances wherein the interviews take place can influence the interview or the respondent and thereby affect the quality of the reserach.

3.6.3 Data collection

In this sub-section, I will elaborate on different aspects in relation to the interviews and on the individual respondents.

3.6.3.1 Respondents

Two groups of experts were selected: internal and external. The internal group consists of:

Kristian Heydenreich, Manager of Corporate Responsibility, and Louise Münter, (former) Director of Communication and Public Relations. I did however not interview her myself.

Heydenreich is responsible for DONG‟s code and has worked with it since its origin. He developed the code in 2006 and established it in 2007 and has furthermore contributed to develop DONG‟s monitoring for supplier compliance wherein he has participated in audits.

Consequently, his expertise and experience with the code is of immense value to the research set out in this thesis.

Two different interviews with Louise Münter carried out in the radio programme DR P1 Dokumentar is included as empirical findings because: 1) She is interviewed about the criticism put forth by the media and DanWatch in January 2010 regarding poor working conditions in Colombia. 2) She describes factors of DONG‟s code and social responsible behaviour highly relevant for this research. 3) She was responsible for DONG‟s CSR when the two cases hit the media; thus, she sent the notes of explanation to the Ministry of Finance about DONG‟s social

(27)

responsibility in relation to Cerrejón and SUEK, which are included as empirical findings as well¹⁰. 4) DR is a valid source of information and the program is based on critical journalism.

The second group of experts consists of respondents external to DONG. As I wanted to explore different stakeholders‟ perceptions of and expectations to codes of conduct and companies‟

social responsibility, I selected respondents from diffent sectors. Linna Palmqvist, Senior CSR Consultant at PwC, Sanne Borges, Human Rights Advisor at Amnesty International Danmark and Representative of the 92-Group in the CSR Council, and Jesper Nielsen, International advisor at 3F and board member of Danish Ethical Trading Initiative, were selected based on their expertise. To my advantage, Jesper Nielsen contacted Carsten Hansen, General Secretary at 3F and coordinator for Fagligt Colombia Fokus, and brought him along for the interview. Both Jesper Nielsen and Carsten Hansen have expertise in labour rights and trade union conditions in Colombia. Lastly, I had a meeting with Peter Thagesen, Director of International Market Policy at Confederation of Danish Industry.

This group of experts proved to be very interesting as they contribute with different and contradicting perceptions. For instance, Peter Thagesen, and Linna Palmqvist view codes of conduct and CSR from a business perspective. In addition, Palmqvist has personal experience with audits and monitoring. Sanne Borges, Jesper Nielsen, and Carsten Hansen view codes of conduct and CSR from a social perspective, which provides an interesting contrast to the business perspective.

The beforementioned radio programmes also include statements by different stakeholders in Colombia, such as mine workers, trade union representatives, and management representatives.

These stakeholders are highly interesting as they experience the strengths and weaknesses of DONG‟s code and therefore provide information and perceptions I would not otherwise have access to. See appendix 3 for a list of these respondents.

I also contacted DanWatch, but it was not possible to get an interview. Neither was it possible to get an interview with SUEK. Cerrejón would however like to be of assistance by email, but did not respond to my questions after all.

(28)

3.6.3.2 Preparations before interviews

The interviews could be regarded as elite interviews (Kvale and Brinkmann 2009:167), i.e. the respondents hold the expert knowledge, thus the power to control the interview. Therefore, I prepared for ways to politely get back to the issues I wished to cover, if the respondents should steer in another direction. However, this turned out not to be necessary at all. Each respondent was very nice and obliging and gave me control of the situation. To prepare for the interviews, I read Kvale and Brinkmann‟s book InterView (2009). Aspects of particular relevance to the interviews will be elaborated on in the following sub-sections.

3.6.3.3 Semi-structured interviews and interview guides

Before the interviews took place, interview guides were drawn up: one for Heydenreich (Appendix 10) and one for the external group (Appendix 11). These were developed based on the research question, investigative questions, and the chosen theory to operationalise the theoretical assumptions I wanted to explore. I tried to make the questions short, open, and apprehendable encouraging honest and detailed answers. The interview guides were developed in order to conduct semi-structured interviews, which were chosen to explore specific issues in a less structured way than closed interviews. This way, I was able to ask follow-up questions and let the respondents focus on relevant matters that I might not have included in the interview guide. Thus, the respondents were asked questions about the same issues with some adjustments according to their expertise and the respective situations. I also had the opportunity to ask additional questions and elaborate on questions as well as change the order of the questions when necessary, as suggested by Kvale and Brinkmann (2009:155). Leading questions were avoided to gain a genuine picture of the respondents‟ own perceptions without being influenced by me.

3.6.3.4 Conducting the interviews

All interviews were held at the respondents‟ choice of location, which were at the company/organisation. This way, the respondents could feel at home. The atmosphere was informal and positive during all the interviews. DONG‟s impressive building at Nesa Allé gave the feeling of being outside, as the interview was conducted in an open coffee place with spaciousness, brightness, green trees, and people walking by. Before each interview began, a short briefing about background and purpose of the thesis was described as well as the purpose of the interview, like recommended by Kvale and Brinkmann (2009:149). Permission to record

(29)

the interview was asked and given. Each interview began with a question about the respondents‟

background and current position in order to make the respondent feel at ease. The interview guides were followed to some extent, although some questions were asked randomly to make it a more fluent conversation. Additional and probing questions were asked in connection with information I was not aware of in order to learn more about specific issues and/or perceptions.

Also, follow-up questions were asked to clarify meanings where these were found unambiguous.

A few times, when I was not sure what the respondent meant, I rephrased the answer to gain clarification. Some of the questions from the interview guides were never explicitly asked, as they had already been elaborated on in connection with other questions, which I allowed as this provided a more fluent dialogue and comprehensive answers. I made an effort to listen carefully and not ask leading questions to ensure that respondents gave a genuine description of his/hers subjective perception. To show my interest and encourage further details, I responded to statements with „yes‟, „no‟, a nod with my head, or a pause to stimulate continuing response, as well as keeping eye contact. I was also aware of my body language, as I did not want to appear too eager nor uninterested or offhand. During the interview with Heydenreich, I was also aware that the criticism DONG had been subjected to could be a delicate issue, so I was careful not to appear negative or judging. Each interview ended with a debriefing wherein an open question was asked about whether the respondent would like to add anything or whether I had forgotten to ask about something, like recommended by Kvale and Brinkmann (2009:149).

I find that the semi-structured interview was very suitable for the research set out. A lot of information would have been excluded if the interview had been closed, as the respondents described things I did not expect nor had any knowledge of. An open interview might have provided more information, however, it would be more time consuming and much of the information would possibly be irrelevant for answering the research question.

3.6.4 Transcription of interviews

All the interviews were transcribed. This process provided new insight into the empirical data as I got deeper into the conversation and became more aware of the content and was able to find relevant elements and characteristics. The transcriptions made it possible to read through the conversations again and again and discover new relevant elements. The interviews were carried out in Danish for the convenience of the respondents, thus the transcriptions are in Danish as

(30)

well. Small talk, words like „ehm‟ and „altså‟ and interviewer‟s responses, like „ja‟, „ok‟ and

„nej‟, are left out to ease the transcription process and to make the interviews more coherent.

Also, I do not consider this important to the analysis, and I still find the transcriptions objective, reliable, and true to the respondents (Kvale & Brinkmann 2009:209). The transcribed material has been read carefully, and common features have been marked with specific colours to indicate which factor or issue respondents are talking about (See appendix 4). The transcription of the interviews is added as appendix 5. Audio files of the interviews can be found on the enclosed CD-ROM.

3.6.5 Quality assessment

It is important to reflect and be aware of the choices that describe the objectivity of the research.

In other words, how the knowledge retrieved from the empirical findings can be verified. Here, the knowledge retrieved from the empirical findings is verified by the terms „validity‟ and

„reliability‟.

In relation to research, validity refers to whether the research in fact investigates the research question set forth (Kvale & Brinkmann 2009:272, Saunders et al. 2009:157). The empirical data collection consists of seven qualitative interviews wherein respondents‟ subjective perceptions have been analysed. According to critical realism, phenomenons create sensations that are open to misinterpretation (Saunders et al. 2009:119); therefore, I planned the interviews carefully in order to appear as objective and professional as possible to avoid leading respondents in any direction. When I was uncertain about a respondent‟s perception of a certain matter, I asked clarifying or probing questions. After having transcribed the interviews, respondents were also given the opportunity to approve or disapprove of their quotations to avoid misinterpretation. I am aware that my presence in the interviews might influence the respondents, and that my interpretation of the theory as well as empirical data is not necessarily the same as other researchers‟. Consequently, if the research was carried out again it might not yield precisely the same conclusions. I have also strived to ensure validity by reflecting and describing each stage of the research method throughout the thesis. This can also contribute to enhance reliability.

Reliabiliy refers to whether the data collection techniques or analysis process yield consistent findings and whether the result can be re-produced at another point in time (Kvale & Brinkmann 2009:271, Saunders et al. 2009:156). The questions were prepared before the interviews in order

(31)

to ensure consistent, open, and non-leading questions. The selected respondents were all experts within their field to ensure data collection of high quality. In addition, respondents were selected to ensure different perceptions and avoid bias. To ensure transparency in how data is interpreted the transcriptions are coded with specific colours. The selected methodology and research methods are decribed in detail to provide the opportunity to carry out the same research or test the results at another point in time. However, as codes of conduct are constantly evolving, and stakeholders‟ perceptions might change, a similar research might not lead to the same conclusions nor is it likely that respondents will answer in the same manner even if asked the same questions.

As the data collection technique is qualitative and seeks to gain specific and in-depth knowledge of DONG‟s code and not to generalise, generalisability is not relevant here. Nevertheless, according to Yin, case studies are generalisable to theoretical propositions and not populations;

thus, an analytic generalisation could be relevant (instead of a statistical generalisation) (2009:15). Thus, based on an analytic generalisation the research findings may apply to other companies that either have or plan to establish a code of conduct for suppliers in high-risk countries, especially in Colombia.

(32)

4. Theory

In this section, I will begin by setting out the theory chosen for this thesis. Thereafter, I will move on to expound on the chosen theory.

As stated in the introduction, theory and debates on codes of conduct focus on whether they are ment for self-interesting purposes or a social obligation to society. As DONG states: “we endeavour to act responsibly and live up to society‟s expectations every single day” (Appendix 6), the research question is viewed from the normative perspective, i.e. companies seek to do well by doing good driven by an obligation to society (Morsing et al. 2007:88). Nevertheless, in order to conduct a thorough analysis, the instrumental perspective will be included as well, i.e.

that companies establish codes of conduct as self-interesting mechanisms to prevent crisis and protect reputation (ibid).

The main theorists used in this thesis are David Vogel, Virginia Haufler, Richard Locke, and Dara O‟Rourke. They are chosen, because they are acknowledged within the field and focus on codes of conduct from different point of views and therefore supplement each other well in relation to this research. The theorists are all Professors and have published several books and/or articles. I will however only focus on those relevant to the research question. As Professor John Ruggie‟s Framework was recently published and deals with factors relevant to this research, this will be included were relevant. Furthermore, other sources from academic articles on the subject are included where relevant as well. All articles are academic articles retrieved from peer- previewed journals in order to ensure high quality. To highlight the theorists‟ relevance to this thesis, they will be presented in relation to the subject areas they are used within.

4.1 Introducing the main theorists by subject area

4.1.1 Stakeholder theory

A brief description of stakeholder theory will be presented based on literature by Professor Richard Edward Freeman as he is the main contributor to this subject and his theory is highly acknowledged. In addition, Freeman‟s stakeholder theory formed the modern perspective of CSR, which codes of conduct are a part of and, which also challenged Milton Friedman‟s shareholder perspective. Stakeholder theory is included in this research because stakeholders are