UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The new global politics of responsible investment Balasubramanian, Priyanjali 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2018_november_balasubramanian_priyanjali.pdf [ 2.96MB ]
JSON: 24-1.0372962.json
JSON-LD: 24-1.0372962-ld.json
RDF/XML (Pretty): 24-1.0372962-rdf.xml
RDF/JSON: 24-1.0372962-rdf.json
Turtle: 24-1.0372962-turtle.txt
N-Triples: 24-1.0372962-rdf-ntriples.txt
Original Record: 24-1.0372962-source.json
Full Text

Full Text

THE NEW GLOBAL POLITICS OF RESPONSIBLE INVESTMENT by  Priyanjali Balasubramanian  B.A., The University of British Columbia, 2002 M.A., Royal Roads University, 2005 M.A. The University of British Columbia, 2009  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Political Science)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  October 2018  © Priyanjali Balasubramanian, 2018 ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled: The New Global Politics of Responsible Investment  submitted by Priyanjali Balasubramanian in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Political Science  Examining Committee: Dr. Lisa Sundstrom, Political Science Supervisor  Dr. Peter Dauvernge, Political Science Supervisory Committee Member  Dr. Tessa Hebb, School of Public Policy and Administration, Carleton University Supervisory Committee Member Dr. Yves Tiberghien, Political Science University Examiner Dr. Stepan Wood, Allard School of Law University Examiner  Additional Supervisory Committee Members:  Supervisory Committee Member  Supervisory Committee Member iii  Abstract This dissertation offers new perspectives on long-standing debates about private actors in global politics.  It does so via three journal-length papers on the role of institutional investors in advancing human rights compliance by multi-national firms. The thesis innovatively bridges international relations scholarship on private authority, human rights norms and transnational advocacy, with academic work on corporate governance, responsible investment and business ethics. These disparate academic themes are unified through an empirical focus on the institutions through which responsible investment activism occurs, and how these institutions respectively inform and challenge existing conceptions of shareholder power, as a form of transnational private authority.  Using comparative analysis, the first paper examines how Dutch and Norwegian pension funds responded to allegations that a Chinese state-owned firm in their portfolios was complicit in human rights violations in Sudan and Burma. In this paper, I argue that state-based institutional checks on shareholder power can affect the tactical strategies employed within investor-led human rights advocacy, and in some cases, can limit the scope for ethical deliberation on these strategic choices. The second paper maps thirty-three investor-driven governance networks to show how their institutional design choices vary significantly in the degree to which they allow for meaningful engagement with civil society. This paper argues that although the mainstreaming of responsible investment has relied heavily on a narrative of values alignment with stakeholders, governance mechanisms to incorporate stakeholders within investor activism on sustainability issues remain rare. The third paper analyses 72 shareholder proposals on the topic of global human rights filed in Canada from 1982-2017.  The analysis confirms that iv  contrary to dominant viewpoints, shareholder power is not ethically neutral. Rather, its use reinforces particular social hierarchies that do not advance global human rights, despite the appearance of doing so.  The dissertation’s key contribution highlights the need for institutional adaptations to enhance the democratic qualities of global investor activism in ways that move the global business and human rights agenda towards collective publics rather than individual and private solutions. This study represents the first systematic effort to theorize and empirically evaluate shareholder power in the context of global human rights.   v  Lay Summary Multi-national companies are obligated to respect global human rights standards. Developing effective mechanisms to ensure these obligations are met remains a persistent challenge. The role that institutional investors can play in addressing this challenge is under-explored, even though these investors have tremendous political and economic power.  By looking at the different ways through which investors try to improve human rights compliance by firms, this project’s research question asks whether investors leverage their influence in ways that shift power away from firms and towards other social actors, such as communities where human rights violations are occurring.  Reviewing data from multiple sources, the project concludes that investor activism for human rights mainly functions as a tool for the financial elite to reduce investment risks. However, investors can indeed play a transformative role towards more equitable social power relations, if the institutions that facilitate such activism on human rights are re-designed accordingly.   vi  Preface This dissertation is an independent and original intellectual product of the author, Priyanjali Balasubramanian.  Research conducted as part of this dissertation and reflections provided in the concluding chapter were used to develop the following book chapter: Bala-Miller P. (2018). Has the Mainstreaming of Responsible Investment Eroded Ethics? Insights from Investor Advocacy on Human Rights in Conflict Zones. In: Poff D., Michalos A. (eds.) Encyclopedia of Business and Professional Ethics. Springer International, Cham (Switzerland).  The interviews referenced in the dissertation were conducted as part of fieldwork undertaken in Canada, the United Kingdom, the United States, Norway and the Netherlands from October 2013 to October 2015. The interviews were conducted in accordance with the TriCouncil Ethics Board’s standards. The research was approved by the UBC Behavioural Research Ethics Board, as covered by certificate number H13-02632. Data in Chapter 5 pertaining to Canadian shareholder proposals filed between 1982 and 2017 was compiled by the Shareholder Association for Research and Education (SHARE) and is publically accessible. vii  Table of Contents Abstract ......................................................................................................................................... iii	Lay Summary .................................................................................................................................v	Preface ........................................................................................................................................... vi	Table of Contents ........................................................................................................................ vii	List of Tables ................................................................................................................................ xi	List of Figures .............................................................................................................................. xii	List of Abbreviations ................................................................................................................. xiii	Acknowledgements ................................................................................................................... xvii	Dedication ................................................................................................................................... xix	Chapter 1: Introduction ................................................................................................................1	1.1	 Empirical justification ..................................................................................................... 8	1.2	 Key arguments and theoretical significance ................................................................. 12	1.3	 Methods and data .......................................................................................................... 20	1.4	 Chapter overview .......................................................................................................... 23	Chapter 2: Theoretical Approaches to Shareholder Power .....................................................29	2.1	 Agency theory ............................................................................................................... 31	2.2	 Organizational theory .................................................................................................... 36	2.3	 Stakeholder theory ........................................................................................................ 42	2.4	 Socio-political approaches ............................................................................................ 49	2.5	 Broadening the scope of analysis on shareholder power .............................................. 53	Chapter 3: Explaining Tactical Variations in Investor Responses to Human Rights Issues 64	3.1	 Theoretical Framework ................................................................................................. 68	viii  	 Exit, voice and loyalty .............................................................................................. 68		 Exit and voice as a framework for shareholder activism .......................................... 72	3.2	 Empirical context .......................................................................................................... 84	3.3	 Analytic framework, data and measures ....................................................................... 88	3.4	 Findings ........................................................................................................................ 92		 Hypothesis 1: Firm responsiveness ........................................................................... 93		 Hypothesis 2: Financial risk ...................................................................................... 96		 Hypothesis 3: Reputational risk ................................................................................ 98	3.5	 Analysis and conclusion ............................................................................................. 104	Chapter 4: Power and Authority in Investor-driven Governance Networks .......................111	4.1	 Literature review ......................................................................................................... 117		 Institutional investors and the global sustainability agenda ................................... 117		 NGOs and shareholder activism on environmental and social issues ..................... 125	4.2	 Analytic framework .................................................................................................... 133		 Defining recognition ............................................................................................... 133		 Defining participation ............................................................................................. 134		 Data, research strategy and measures ..................................................................... 137		 Description of sample ............................................................................................. 142	4.3	 Findings ...................................................................................................................... 145		 Recognition ............................................................................................................. 145		 Participation ............................................................................................................ 148	4.4	 Conclusion .................................................................................................................. 157	ix  Chapter 5: Using or Abusing Shareholder Power? Reframing Shareholder Proposals as a Discourse Structure ...................................................................................................................160	5.1	 The mechanics of filing shareholder proposals: A comparative overview ................. 167		 United Kingdom ...................................................................................................... 168		 The United States .................................................................................................... 170		 Australia .................................................................................................................. 171		 Canada ..................................................................................................................... 172	5.2	 Literature review ......................................................................................................... 173	5.3	 Framing shareholder proposals as a discourse structure ............................................. 182		 Structured properties of shareholder proposals. ...................................................... 183		 Relational/emergent properties of shareholder proposals ....................................... 184		 Shareholder proposals as a conduit for social/shareholder power .......................... 190	5.4	 Analyzing shareholder proposals as a discourse structure .......................................... 193		 Discursive representations of corporate social responsibility (CSR) ..................... 197	5.5	 Data and Findings ....................................................................................................... 209		 General description of data and key trends ............................................................. 209		 Discursive representations of CSR in Canadian shareholder proposals on human rights ................................................................................................................................. 216		 Discursive representations of shareholder power in Canadian shareholder proposals on human rights ................................................................................................................... 232	Chapter 6: Conclusion ...............................................................................................................249	6.1	 Project scope ............................................................................................................... 252	6.2	 Theoretical contributions ............................................................................................ 254	x  6.3	 Key implications ......................................................................................................... 255		 Institutionalised patterns of exclusion .................................................................... 256		 The dominance of business-case logic .................................................................... 258		 The erosion of capacity for ethical deliberation ..................................................... 259	6.4	 Final reflections – Who wins and who loses in the mainstreaming of responsible investment? ............................................................................................................................. 263	References ...................................................................................................................................266	Appendices ..................................................................................................................................324	Appendix A Fund Profiles ...................................................................................................... 324	A.1	 Position in the domestic pension system ................................................................ 324	A.2	 Fund governance and management ......................................................................... 326	A.3	 Strategic investment mix ......................................................................................... 327	Appendix B Interview Information ......................................................................................... 333	Appendix C List of Shareholder Resolutions on Human Rights Issues, 1982-2017 .............. 335	 xi  List of Tables Table 3.1. Scope of corporate engagement, 2013 ......................................................................... 87	Table 3.2. Summary of analytic framework ................................................................................. 89	Table 4.1 List of IGNs studied .................................................................................................... 138	Table 5.1. Sources of authority in global governance and related observable implications (Avant, Finnemore and Sell, 2010, pp 11-14) .......................................................................................... 203	Table 5.2. CDA-based strategy for analysing shareholder proposals ......................................... 207	Table 5.3. Proposal filers by category ........................................................................................ 216	Table 5.4. Discursive representations of CSR by frequency ...................................................... 219	Table 5.5. Discursive approaches to framing a firm's responsibilities ....................................... 224	Table 5.6. A snapshot of Canadian shareholder proposals related to employees and labor rights..................................................................................................................................................... 226	Table 5.7. Discursive representations of CSR by frequency ...................................................... 229	Table 5.8. Geographic areas of concern [human rights] for Canadian institutional investors .... 242	Table 5.9. Types of demands made within Canadian shareholder proposals on human rights .. 244	Table A.1 ABP portfolio by asset class 2010-2012 .....................................................................327 Table A.2 Scope of corporate engagement, 2013 ........................................................................332 Table B.1 List of interview subjects ............................................................................................333 Table C.1 List of shareholder resolutions on human rights issues, 1982-2017 ...........................335 xii  List of Figures Figure 1.1. Number of SSE partner exchanges, 2012-2016 ....................................................... 3	Figure 1.2. Growth of PRI signatories and respective assets under management, 2006-2017 ....... 4	Figure 1.3. Geographic diffusion of investor stewardship codes .................................................... 6	Figure 1.4. Assets under management and allocation to public equity by different types of institutional investors (in trillion USD, 2011) ................................................................................ 9	Figure 1.5. Geographic spread of countries of concern in examined cases of investor led human rights activism ............................................................................................................................... 21	Figure 4.1. Growth of IGNs, 1970-2017 ..................................................................................... 143	Figure 4.2. Where are IGNs headquartered? .............................................................................. 144	Figure 4.3. IGNs studied by organisation type ........................................................................... 145	Figure 4.4. Recognition of civil society by IGN type ................................................................. 147	Figure 4.5. CSO participation in IGN agenda-setting ................................................................. 149	Figure 4.6. Do IGNs offer INGOs membership? ........................................................................ 151	Figure 4.7 Can INGOs participate in IGN governance? ............................................................. 153	Figure 4.8. Do INGOs implement IGN activities? ..................................................................... 155	Figure 4.9. Do INGOs participate in an IGN's M&E activities? ................................................ 156	Figure 5.1. Strategy for organising text data in CDA ................................................................. 197	Figure 5.2. Overview of sector distribution of proposals, 1982-2017 ........................................ 210	Figure 5.3. Types of demand by frequency ................................................................................ 247	 xiii  List of Abbreviations ABG Amnesty International Business Group AGM Annual General Meeting AO Asset Owners AOW Algemene Ouderdoms Wet Savings Fund ATCA United States Alien Tort Claims Act AUM Assets Under Management CalPERS California Public Employees Retirement Scheme CBCA Canada Business Corporations Act CDA Critical discourse analysis CEO Chief Executive Officer CERES Coalition for Environmentally Responsible Economies CII Council of Institutional Investors CME coordinated market economy CNPC China’s National Petroleum Corporation CRN Conflict Risk Network CSO Civil Society Organisation CSR corporate social responsibility DFID U.K. Department for International Development DNB De Nederlandsche Bank DTR U.K. Disclosure Guidance and Transparency Rules ENGO Environmental non-governmental organisation xiv  ERIN European Responsible Investment Network ESG Environmental, social and governance issues EVL Exit, Voice and Loyalty (Hirschman, 1970) FBC Free Burma Campaign FCA U.K. Financial Conduct Authority FOE Friends of the Earth FRC U.K. Financial Reporting Council GPFG Norwegian Government Pension Fund-Global GPFN Government Pension Funds – Norway ICCR Interfaith Centre for Corporate Responsibility ICGN International Corporate Governance Network IEHN Investor Environmental Health Network IFC International Finance Corporation IGN Investor-driven governance networks IIRC The International Integrated Reporting Council ILO International Labour Organisation IMF International Monetary Fund INGO International non-governmental organisation IO International organisation IR International relations M&E Monitoring and evaluation MNC multi-national corporation MOGE Myanmar Oil and Gas Enterprise xv  MOU Memorandum of understanding NBIM Norges Bank Investment Management NEI North West and Ethical Investments NGO Non-governmental Organisation NGPF Norwegian Government Pension Fund OECD Organisation for Economic Co-operation and Development PRI United Nations backed Principles on Responsible Investment RIAA Responsible Investment Association of Australasia S&P 500 Standard & Poor’s 500 Index SDG United Nations Sustainability Development Goals SEC United States Securities and Exchange Commission SHARE Shareholder Association for Research and Education SRI socially responsible investment SSE Social Shareholder Engagement SSE Sustainable Stock Exchanges initiative STAN shareholder transnational advocacy network TAN transnational advocacy network TCCR  Taskforce on Churches and Corporate Responsibility TWAIL Third World Approaches to International Law U.K. United Kingdom of Great Britain and Northern Ireland UN United Nations UNEP United Nations Environment Programme UNEP-FI United Nations Environment Program-Finance Initiative xvi  UNFCCC United Nations Framework Convention on Climate Change UNGP UN Guiding Principles on Business and Human Rights U.S. United States of America USD United States (US) dollars VOC varieties of capitalism WDI Workforce Disclosure Initiative WWF World Wide Fund for Nature  xvii  Acknowledgements In pursuing this thesis project, I am fortunate to have had the patient guidance of my committee, comprised of Lisa Sundstrom, Tessa Hebb and Peter Dauvergne.  Engaging with Lisa’s scholarship on transnational activism through her invitations to collaborate on related research projects, papers and conference presentations have been an enriching experience. Our efforts to facilitate community service learning opportunities for undergraduate students at UBC were especially meaningful, and continue to influence my commitments to transformative learning in my current work.   Lisa’s graceful navigation between a well-respected academic career and the demands of family life was a frequent inspiration to me. I am grateful to Tessa for bringing me into the fold of a talented network of academics and responsible investment professionals through her leadership of the Responsible Investing Initiative at Carleton University and the Principles for Responsible Investment’s  (PRI) Academic Network. The four months I spent at the Carleton Centre for Community Innovation in 2011 gave me a deep appreciation for the influence Tessa’s body of work has had on establishing responsible investment as a specialized academic field. Her network of mentees also led me to delightful collaborations with peers such as Omar Dominguez and Heather Hachigian. Peter Dauvergne’s impressive and accessible scholarship on sustainable consumption and production was a key factor that motivated me to pursue doctoral studies at UBC.  I appreciate Peter’s consistent encouragement that I probe power dynamics through this thesis project. Despite my own initial misguided misgivings to the contrary, following his steer ultimately made for a richer intellectual journey. Any remaining gaps in the dissertation are mine to own. My time at UBC was also enriched by the tutelage a number of dedicated professors including Yves xviii  Tiberghien, Erin Baines, Anjali Bohlken, Alan Jacobs, Katia Coleman, Maxwell Cameron and Richard Price. The financial pressures of graduate school were greatly eased through doctoral fellowships and stipends provided by the Social Sciences and Humanities Research Council (SSHRC), UBC’s Department of Political Science and Faculty of Graduate Studies. Bahja Alammari’s diligent efforts to collate data used in Chapter 5 was deeply appreciated.  Carolyn Hostinsky, Elaine Ryan, Merima Kostecki, Cherlyn McKay, Junie Desil, Agustin Goenaga, Serbulent Turan, Yana Gorokhovskaia, Jen Allan, Lalani DeSilva, AiTiang Ng and Domenika Nackiewicz are cherished friends. We have seen each other through the stuff of life, and built ties I hope will keep us close for years to come.  In addition, colleagues at SHARE and CIRDI accorded me many professional courtesies so that I could clear a number of academic hurdles along the way. Marie-Luise Ermisch showed up when I needed an angel most. Noushin Khushrushahi and Laura O’Neil’s cutting wit and good humour often put my “grad-school problems” in perspective. My family members are beacons of personal fortitude and I offer everyone in the Bala-Miller clan my deepest thanks for seeing me through a time that has challenged me in more ways than I can count. I owe a debt of gratitude for my brother Sudeep’s generosity, my brother Sid’s ability to de-mystify financial markets, and my mother Sundari’s nurturing care - emotional and otherwise. Through the thesis process, my son Vihaan showed me love and understanding beyond his years.  Blair Miller is the best human I know. With an uncanny ability to detect when tea is the ideal solution for a problem at hand, if I need to be talked off a ledge (especially those of my own making), and to decipher if enthusiastic encouragement or silent support is the order of the day, he has been my unfailing champion and true partner in all the ways that matter. For this and much more, thank you my love.      xix  Dedication For Blair.   1 Chapter 1: Introduction The key to the power of ethical investing is the recognition that corporations and institutions are owned and managed by people. Ethical investing insists on their humanity. It denies them the luxury of pretending to be impersonal economic forces. It requires a recognition of responsibility for one’s actions, and it demands that others assume responsibility for theirs. As an instrument for change, both personal and social, ethical investing is unequaled (Domini and Kinder, 1984,  p. 211). We've been doing this year after year, using this forum [shareholder proxy requests]  to allow people to have a voice in Canada and talk to the shareholders and directors directly…This year, they [Barrick Gold] rejected almost all the proxies, and there was no reason given. The first thing that came to my mind is, 'silence is violence'... This is how you silence people (Coumans, 2017, ¶ 8). Responsible investing – the alignment of investment decision-making with financial, social and environmental sustainability, and ethical values – has increased dramatically over the past three decades.1  What started out in the late 1970s as a niche effort by predominantly small groups of religious orders seeking to align their investments with religious values is now a global phenomenon (Sparkes and Cowton, 2004). Leading the charge for this transformative shift are financially powerful institutional investors who own the lion’s share of global capital.2 Echoing                                                 1 In this dissertation, the term responsible investing is used interchangeably with sustainable investing and ethical investing.  2 The term institutional investor generally refers to organisations that invest money on behalf of members or future beneficiaries (Brancato 1997). Typical institutional investors are pension funds, hedge funds, mutual funds, insurance firms, endowments and trusts for non-profit organisations, and sovereign wealth funds. These investors may have a diverse portfolio of investments across asset classes such as public equity, bonds, real estate and infrastructure.  They a separate class of investor from individual (retail) investors and may hold investments in foreign and domestic markets.    2 the sentiments expressed above by Amy Domini and Peter Kinder (1984), public discourse on responsible investment is replete with promises of systemic transformation that puts capital in the service of society rather than the reverse (Fink, 2018; Battilana, 2018; UNPRI, 2018). Amidst claims that responsible investment is now “mainstream” (Economist, 2017; Revelli, 2017; Cowton, 1999; Dunfee, 2003), is this movement meeting its mandate for humanising global capital markets? The answer will depend on who you ask – financial industry insiders or their critics.  In this dissertation, I suggest that competing appraisals about the impacts of the responsible investment movement are a window into broader debates about the nature of power and authority wielded by private actors in global politics in the arena of human rights (Hall and Biersteker, 2002; Cutler, 2003; Büthe, 2004; Hall, 2007). Key fault lines in these debates centre on the social responsibilities of private actors, the rules and institutions that are best placed to mediate power differentials between the global multi-national companies and local communities, and ultimately whether the imperative for profit maximization that drives global capitalism can be harnessed to deliver on social goods such as universal human rights.  The general thrust of my dissertation is to engage with and contribute new perspectives on these long-standing debates through an inter-disciplinary approach that bridges international relations scholarship on private authority, compliance with global human rights norms and transnational advocacy, with academic work on corporate governance, responsible investment and business ethics.  These disparate academic themes are unified through a common concern with how they respectively inform, reinforce and challenge existing conceptions of shareholder power, as a form of private authority in transnational politics.   3 Financial industry insiders are quick to point to the increasing technical sophistication with which environmental and social data is incorporated into asset allocation models or to the overhauls of investment and corporate governance systems to accommodate the consideration of extra-financial risks as signals of the transformative changes investors are making towards greater social and environmental sustainability. Exemplifying these trends are efforts like the Sustainable Stock Exchanges (SSE) initiative, a peer-to-peer learning platform for exploring how stock exchanges can enhance corporate transparency and performance on environmental, social and governance (ESG) issues and ultimately encourage sustainable investment (UNEP-FI, 2018).  As of 2016, 58 stock exchanges had joined the SSE initiative, representing over 70% of publicly listed equity markets (see Figure 1.1)3.  Source:  Sustainable Stock Exchanges Initiative, 2016 Report on Progress The staggering growth in the number of signatories to the UN-backed Principles of Responsible Investment (PRI) is another oft-cited indicator that approaches to sustainable finance                                                 3 Public equity is an investment asset class where individuals and/or organizations can buy ownership in shares/stock of a company through a public market such as a stock exchange.  0102030405060702012 2013 2014 2015 2016Figure 1.1. Number of SSE partner exchanges, 2012-2016  4 are being mainstreamed (see Figure 1.2). Since its launch in 2006, more than 1800 institutional investors, asset managers and investment service providers have committed to the initiative’s six voluntary principles for incorporating ESG issues into investment practice (PRI, 2018a). Indeed, in an effort to stress that signatories “can’t just sit there”, Managing Director Fiona Reynolds announced at the annual meeting in 2016 that the PRI would create a watch list of signatories that were not making sufficient progress in implementing the principles. She further explained that signatories who did not improve within two years of being placed on the list or who incurred fines or regulatory sanctions that violated the spirit of the PRI would be de-listed (Brooksbank, 2016, ¶2). The poignancy of such rhetoric in a post-2008 global financial crisis context is worth noting.   Source:  PRI Website (2018b), About the PRI.  0200400600800100012001400160018000. Apr-07 Apr-08 Apr-09 Apr-10 Apr-11 Apr-12 Apr-13 Apr-14 Apr-15 Apr-16 Apr-17AO	AUM	($	US	trillion) Assets	under	management	(US$	trillion)Number	of	AOs Number	of	SignatoriesFigure 1.2. Growth of PRI signatories and respective assets under management, 2006-2017  5 Observers and pundits argued that a key precipitating factor for the 2008 global financial crisis was that institutional investors were derelict in their duties to oversee corporate managers and in curbing risky and unsustainable corporate behaviour (for example, Della Croce et al., 2011). The global financial crisis also brought rising levels of income inequality into sharper relief, where the plight of ordinary people experiencing home foreclosures and financial ruin contrasted sharply with run-away levels of CEO compensation at the very financial institutions that ushered in the crisis (Bhagat and Bolton, 2014; Bebchuck et al., 2010).  In this vein, reflections on the role of investors in the global financial crisis prompted the proliferation of so-called “stewardship codes” that aim to enhance investor engagement and transparency about how they define and discharge their ownership and governance responsibilities. These codes also provide a mechanism for investors to communicate their priorities on ESG issues such as human rights. In 2010, the United Kingdom became the first jurisdiction to adopt a stewardship code, following Sir David Walker’s 2009 post-crisis review of corporate governance in UK-based banks and other financial institutions (Walker, 2009). A number of other jurisdictions have followed suit, including Denmark, Hong Kong, Japan, Kenya, Malaysia, South Africa, Taiwan and Thailand (EY, 2017, p. 2). More recently, the European Union adopted a Shareholder Rights Directive that includes elements found in existing stewardship codes, and India and Kazakhstan are actively considering the adoption of similar initiatives (See Fig 1.3). Aside from these efforts, investor-led best practice guidance currently exists in Australia, Brazil, Canada, Italy, the Netherlands, Singapore, South Korea, Switzerland and the United States (EY, 2017, p. 2). In addition, the International Corporate Governance Network (ICGN), whose members represent governance professionals from over 47 countries, adopted the Global Stewardship Principles in 2016 which include a principled commitment that,  6 “investors should promote the long-term performance and sustainable success of companies and should integrate material environmental, social and governance (ESG) factors in stewardship activities.” (ICGN, 2016, p.11)  Figure 1.3. Geographic diffusion of investor stewardship codes Source: Ernst and Young © (EY, 2017). These positive trends in the uptake of responsible investment are tempered by critiques that the responsible investment movement neglects to account for the financial sector’s impact on local communities. For instance, the Sahrawi people in Western Sahara oppose the certification of energy infrastructure projects as “sustainable” by responsible investment advisory firms when these same projects are located in contested indigenous territories that are claimed by Morocco (Western Sahara Resource Watch, 2017).  Likewise, some civil society groups like Mining Watch remain skeptical of the ability of investors to advance corporate compliance with global human rights norms and standards, noting that shareholder-led activism on this issue is  7 prone to cooption or subversion by firms, as illustrated by a recent case in Canada detailed below (Coumans, 2011, 2012).   On 25 April 2017, Evelyn Guape and Jocelyn Mandi travelled from the Porgera Valley in Papua New Guinea to Toronto to speak at Barrick Gold’s annual general meeting (AGM) with shareholders (McSheffrey, 2017). Their testimony was intended to reveal personal stories of trauma and survival following horrific gang rapes perpetuated by security guards at Barrick’s Porgera mine nearly twenty years earlier, and to draw attention to patterns of inadequate access to justice and restitution experienced by more than one hundred women similarly brutalised by the company’s security forces on an on-going basis (McSheffrey, 2017). Guape and Mandi’s right to speak at the meeting was facilitated via a request for proxy access4 by Mining Watch, who held shares with the company. Citing a protocol mishap, the firm denied the women the opportunity to speak at the meeting and offered them a closed-door meeting with the firm’s managers instead.  At the AGM, Guape and Mandi’s statement was read by a Mining Watch representative, an act that according to the NGO, amounted to “taking away their voice” (Coumans in McSheffrey, 2017, ¶7).  In the remainder of this introductory chapter, I provide a brief empirical justification for the dissertation’s analytical focus on human rights advocacy by institutional investors (Section 1.1). This is followed by an overview of the key arguments and theoretical contributions of the                                                 4 One way that publicly traded companies report their activities to shareholders is through annual meetings. Before those meetings, shareholders receive information on pertinent information to be voted on in the meeting. Rather than physically attending the annual shareholder meeting, most investors delegate their voting rights (by proxy) to asset managers or other investment service providers (proxy voting). The person exercising this right must in line with the shareholder’s directions. Proxy holders attending a shareholder meeting may be permitted to speak to issues related to the firm’s business operations.     8 project as a whole (Section1.2). I then outline the general methodological orientation and data-collection methods used, reflecting on some of the challenges posed in the study of investor-led transnational activism (Section 1.3). Section 1.4 provides an overview of the structure of the thesis, noting it is comprised of three journal-length papers that respond to particular theoretical and empirical gaps in existing studies of shareholder power. These substantive chapters are preceded by a review of the relevant literature, and followed with a reflection on the main findings and directions for future research.    1.1 Empirical justification Two empirical trends underscore the significance of my analytical focus on institutional investors and their role in advancing human rights by corporate actors. First is their status as heavyweights in the global financial system. By virtue of the volume of capital under their control, institutional investors wield considerable material power and legitimacy in driving global financial markets at a systemic level (Hebb and Clark, 2005). Over the past twenty years, economic power has become highly concentrated among institutional investors who own an increasing share of public equity markets. For example by 2009, institutional investors collectively owned more than 70 percent of the largest 1,000 companies in the United States (Heineman and Davis, 2011).  According to Celik and Isaksson (2013), in the mid-1960s, individual persons held 84% of all publicly listed stocks in the United States compared to 40% currently; in Japan, only 18% of all public equity was held by physical persons (as of 2011) and in the United Kingdom, the portion of public equity held by physical persons has decreased from 54% to only 11% over the last fifty years (p. 96).   9 Figure 1.4 illustrates the total assets under management of different types of institutional investors and the portion of these assets that they have allocated to public equities (Celik and Isaksson, 2013, p. 97). The figure shows that as of 2011, the combined holdings of all represented institutional investors was USD 84.8 trillion. Out of this, 38% (USD 32 trillion) was held in the form of public equity. The largest institutions by far were investment funds, insurance companies and pension funds. Together they managed assets with a total value of USD 73.4 trillion, of which USD 28 trillion was held in public equity.   Figure 1.4. Assets under management and allocation to public equity by different types of institutional investors (in trillion USD, 2011)  Source:  Celik and Isaacson (2013, p. 9), with data drawn from the OECD Institutional Investors Database, Sovereign Wealth Fund Institute, International Monetary Fund, Preqin, BlackRock and McKinsey Global Institute.  This process of ownership concentration in public equities has empowered institutional investors, particularly in Anglo-American countries to take a very active role in corporate governance, broadly defined as the rules and practices governing power relations between various stakeholders in the modern corporation (Shleifer and Vishny, 1997; Soderberg, 2003; Deakin, 2005; Williams, 2016).  These rules and practices vary depending on the system of  10 corporate governance prevalent in a firm’s home state; nonetheless, all outline the key mechanisms by which investors can exercise their voice individually or collectively on issues that relate to the firm’s performance and efforts to preserve and enhance shareholder value.5  One reason that investors are taking a more active role in monitoring corporate sustainability efforts is the burgeoning amount of evidence showing clear financial benefits for doing so. For example, a recent comprehensive review of empirical studies of the financial results of corporate responsibility by Clark, Feiner and Viehs (2014) found that 90% of studies show that sound sustainability standards lower firms’ cost of capital6; 80% of studies show that the stock price performance of companies is positively influenced by good sustainability practices; and 88% of studies show that better environmental, social, or governance practices result in better operational performance.   Taken together, these trends confirm that institutional investors have a large stake in publicly-traded companies in major markets and that they are increasingly motivated to play an active role in overseeing the performance of firms when it comes of social and environmental sustainability. From this lens, the corporate governance arena becomes an important domain for the exercise of private authority by institutional investors. It is in this domain that institutional investors exercise legally-derived shareholder rights in order to shape and direct company policies and behaviour (Culpepper, 2011; Hebb and Clark, 2005).  Shareholder rights also provide investors with legitimacy in exercising private authority over firms, where authority is                                                 5A company’s shareholder value depends on strategic decisions made by senior management, including the ability to make wise investments and generate a healthy return on invested capital. If this value is created over the long term, the share price increases and the company can pay larger cash dividends to shareholders (Investopedia, 2018).  6 This refers to the cost of funds used for financing a business (i.e. the rate of return needed to make an investment worthwhile).   11 defined as “institutionalized forms or expressions of power” that are legitimate in the sense that “there is some form of normative, un-coerced consent or recognition of authority on the part of the regulated or governed (in this case firms)” (Büthe, 2004, p. 281-282; see also Haufler et al., 1999; Hall and Biersteker, 2002). In addition to the above drivers for expanding investor oversight of firms on social and environmental issues more generally, a second empirical justification for the project’s analytical focus on institutional investors is derived from critical developments in the past decade that have raised expectations for investors to play a more active role in advancing global human rights norms.  For example, institutional investors have a responsibility under the newly enshrined UN Guiding Principles on Business and Human Rights (2011) to respect human rights in their operations and their business relationships with the companies in which they invest. Similarly, revisions to the OECD Guidelines for Multi-national Enterprises in 2011 extended the culpability for violations of the Guidelines to institutional investors. These revisions explicitly drew on the earlier UN “Protect, Respect and Remedy” Framework for Business and Human Rights (2008).  In light of the power and influence institutional investors exert at the firm and systemic level, I suggest that calls for deeper engagement by these investors within the arena of global human rights compliance has the potential to shape and transform power relations between the private sector, states and civil society in a number of important ways. However, given this topic has not been widely explored to date, there is a clear need to assess the true prospects and direction of these power shifts. The theoretical significance for advancing this line of inquiry is elaborated next.    12 1.2 Key arguments and theoretical significance Few works are more ubiquitous in the literature on global human rights norms than Thomas Risse, Stephen Ropp, and Kathryn Sikkink’s (1999) The Power of Human Rights: International Norms and Domestic Change. The book’s analytically powerful “spiral model”7 describes the various socialization processes through which international norms on human rights were internalized by authoritarian states. When these scholars revisited their influential work more than a decade later, the scope of their analysis extended beyond the state to critically examine the role of private actors in protecting human rights (Risse, Ropp and Sikkink, 2013).  They argue that the persistent challenges of human rights compliance point to the limitations of unilateral state action, and that the traditional division of labour between states and private actors may not be sufficient to guarantee human rights within existing models of governance or market systems (see also: Brysk, 2002; Ruggie, 2013; Risse, Ropp and Sikkink, 2013).  While arguments marshalled in favor of this observation have overwhelmingly focused on INGOs, corporations, sub-national political groups or international organizations, very little empirical work has examined the validity of these arguments in the investment sphere. Addressing this gap, the study’s first contribution is to bring institutional investors into the frame of existing analysis on how human rights diffusion and compliance occurs at the transnational level.                                                 7 Risse and Sikkink (1999) use the “spiral model” to explain the steps that states must go through to change their norms and behavior regarding human rights. This model has five phases: repression, denial, tactical concessions, prescriptive status, and rule-consistent behavior. These five steps work in a natural progression to first expose a state’s human rights violations, and then change their norms to fit international standards. The authors argue that socialization is important to understand how the international society can promote norms to its members. They present three various processes that are vital to changing the human rights area: 1) "process of adaptation and strategic bargaining; 2) processes of moral consciousness-raising, ‘shaming,’ argumentation, dialogue, and persuasion; 3) processes of institutionalization and habitualization" (Risse and Sikkink, 1999, p. 11)  13 Choosing this new locus of investigation productively extends and refines our understanding of the role of private power and authority in the global governance of human rights. I suggest that observing the relationship between institutional investors and corporations in the context of human rights norms sheds light on why and how conventional divisions between the state, market and civil society are breaking down, with important implications for how power is exercised and authority is understood in the international system.  If governance is the exercise of political authority and the use of resources to manage society's affairs, human rights advocacy by institutional investors creates a veritable governance grey-zone, as the lines of accountability and authority between the owners of capital, those who use capital, those affected by how the capital is deployed and those with the power to control capital become increasingly blurred. Thus, the political importance of interrogating the role and function of institutional investors in global governance arises not only because their asset allocation decisions could affect the stability of the financial system, but also because they are crucial intermediaries between the market and society.   As a second contribution, the dissertation re-frames how the concept of shareholder power is understood. As noted above, corporate governance includes the mechanisms, processes and relations by which corporations are controlled and directed. These governance structures and principles identify the distribution of rights and responsibilities among different participants in the corporation (such as the board of directors, managers, shareholders, creditors, auditors, regulators, and other stakeholders). It also includes the rules, procedures and processes through which corporations' objectives are set and pursued in the context of the broader social, regulatory and market environment.  Much of the scholarship on corporate governance centres on a classic principal-agent problematique, where shareholder rights endow investors with the power to  14 ensure that a firm’s managers do not subvert profit maximization and investment returns in the pursuit of their own interests. As a consequence, within the corporate governance literature, shareholder power is framed in very narrow terms. It typically encompasses specific mechanisms of control used by investors to influence corporate policies and practices in line with their interests.  With intellectual roots in corporate governance, studies of responsible investing have inherited this narrow conceptualization of shareholder power, restricting discussions of its use and implications to the investor-firm dyad. However, as this dissertation highlights, while this conceptualization of shareholder power might work in the context of domestic or comparative corporate governance, it does not travel as easily to the context of global governance and transnational political action. That is, a more sophisticated theoretical understanding of shareholder power in a global context demands a far more multi-dimensional view of who leverages shareholder power, on whose behalf this power is exercised and resulting effects, well beyond the realm of the firm. Two features of transnationalism are theoretically relevant here. First is the notion that in the age of increasingly interconnected global financial markets, corporate governance itself is transnational in nature. Second, the efforts of investors to mainstream responsible investment mimic dynamics observed in transnational advocacy networks.  On the first point, the dissertation demonstrates that the rise of the responsible investment movement is one manifestation of how contested views are being negotiated around the role of firms in society and what counts as a legitimate subject in a new ‘global era’ of corporate governance.  As Peter Gourevitch and James Shinn (2005) argue in Political Power and Corporate Control: The New Global Politics of Corporate Governance, this new global era of  15 corporate governance is characterized by huge sums of investment capital flowing from developed to developing countries through institutional investors, such that the formation of political identities, issues and interests of financial actors are now global in scope.  On the second point, institutional investors functioning within the bounds of corporate governance who champion corporate compliance with human rights norms as part of their responsible investment initiatives are simultaneously engaged in aspects of what Sidney Tarrow (2005) refers to as “The New Transnational Activism”. As a form of transnational political action, the responsible investment movement shares a few notable characteristics in common with other forms of transnational activism (Piper and Uhlin, 2006) in that it occurs through organized networks of principled actors, poses a challenge to a state-centric international system and promotes a cosmopolitan view of citizenship. In this regard, Keck and Sikkink (1998) describe transnational advocacy networks as “voluntary, reciprocal and horizontal patterns of communication and exchange” (p. 8), while, Charli Carpenter (2007) refers to “networks of activists motivated by shared, principled discourse and aiming to affect political behaviour through moral argument” (p. 101). According to Tarrow (2005), the term “network” has both a structural and a purposive meaning (p. 164), and encompasses institutionalized links that may exist for other purposes, but can be appropriated for contentious collective actions (such as corporate engagement). Since the 1990s, the transnational sphere is populated with a significant number of investor networks organised with the specific purpose of addressing social and environmental issues. Initiatives such as the Asset Owners Disclosure Project, the Social Investment Organisation (Canada), The Social Investment Forum and its numerous national chapters, the Ceres Coalition, Interfaith Centre for Corporate Responsibility (ICCR), Conflict Risk Network, Investor Responsibility Research Network, EURESA Network, PRI National and  16 International Signatory Networks and provide empirical validation of the structural and purposive nature of transnational investor activism on sustainability.  These initiatives typically have defined membership, secretariat and governance structures, clear mechanisms through which to coordinate collective action on responsible investment topics, as well as defined mission statements, advocacy priorities and related programs of work. Aside from having a networked character in common with other transnational forms of activism, advocacy by investors on issues related to human rights similarly challenges the conventional view of a state-centric international system.  Investors do this by making corporations (not states) the subjects of human rights and environmental advocacy, and by emphasizing global standards, such as the Extractives Industry Transparency Initiative (EITI) and the OECD Guidelines for Multi-national Enterprises, over domestic legislation in framing and advancing advocacy claims, especially in weak-governance zones (Risse-Kappen, 1995; Keck and Sikkink, 1998; Khagram, Riker and Sikkink, 2002).   A final transnational feature of investor activism on human rights is that they implicitly advance a globalised notion of citizenship (Held, 1995; Held and McGrew, 2002).  The globalization of markets and the increasingly complex supply chains that straddle multiple jurisdictions are encouraging investors to embrace cosmopolitanism in order to carry out their duties of oversight more effectively. For instance, investors are aware that firms can no longer ignore the impacts of climate change on the access and availability of key resources such as water, and how such factors affect a firm’s business plan. Proactively addressing these problems also frequently requires investors to enter into sustained dialogue with foreign owned companies or firms with significant operations outside a ‘home’ state.  In the case of pension funds specifically, some scholars argue that by virtue of their remit to provide retirement security for  17 future beneficiaries over the long-term, the implicit responsibility of pension funds is to be concerned about the wellbeing of society and the natural environment as a whole. The duty of long-term investors to take a holistic and systemic view of investment decision-making is seen to be aligned with the principle of inter-generational equity. This principle posits that humans “hold the natural and cultural environment of the Earth in common both with other members of the present generation and with other generations, past and future” (Weiss, 1990, p. 8). Investors who take up the mantle of responsible investment through a sense of cosmopolitan citizenship, particularly in areas such as human rights are thus seen as “new capitalists” motivated to use corporate engagement as a tool for achieving a more “civil economy” (Berry, 2011; Lydenberg, 2007). However, there is also one important difference between global investor-led activism and other forms of transnational activism. Unlike some perspectives that treat transnational forms of activism as resistance to neo-liberal forms of globalization (see for example Higgot, 2002), through my analysis in this dissertation, I propose that although shareholder activism may be reacting to political realities generated by economic and financial globalization, investor activism on human rights is not fundamentally resisting the globalization per-se.  As Haigh and Hazleton (2004) observe of socially and environmentally-motivated advocacy by institutional investors, “The isolated instances of ‘success’ in achieving ‘outcomes’ at the firm-level mask the reality that systemic change to industrial practices, or engagement with issues at the industry-wide level has not been the focus of shareholder activists” (p. 60). For example, although investors may urge factories with weak occupational health and safety standards to improve their policies and practices in ways that improve employee well-being, these engagement efforts do not necessarily question or problematize the structural conditions that enable the migration of capital to  18 geographies with weak social and labour protections in the first place. Likewise, Cullis et.al (1992) point out that the ability of financial markets to transform the measure of moral commitments to pecuniary returns represents both an irony and a limitation of responsible investment. Furthermore, the fundamental governing principle of financial markets, namely the right to maximize profits and enjoy capital gains on investment, is not seriously challenged by shareholder activism (Cullis et al., 1992). So what are we to make of shareholder power that is leveraged by institutional investors in this new global context for corporate governance and transnational activism? Rather than a dyadic power relationship between firms and investors, this project argues that a truly transnational understanding of shareholder power must also account for interactions between investors and a variety of other actors who are simultaneously engaged in the global governance of human rights. These include fellow institutional investors, financial intermediaries such as asset managers and responsible investment service providers, the state, industry associations, communities, civil society organizations and inter-governmental agencies and international non-governmental organisations. Hence, the core objective of this dissertation is to advance a broader, relational understanding of shareholder power by exploring the institutions through which this unique form of private power is expressed. Within international relations, institutions are understood as the formal or informal principles, norms, rules and decision-making procedures and structures around which actors’ expectations converge in a given issue area (Krasner, 1982) or humanly-designed constraints8 that shape human interaction (North, 1990).  A wide range of international relations scholarship has been devoted to demonstrating how and                                                 8 Thanks to Lisa Sundstrom for drawing out this emphasis to distinguish my focus from environmentally-imposed constraints.  19 why international institutions matter in explaining particular global governance outcomes given their role in structuring choice and incentives, distributing power and defining identities and roles. These rules and practices are typically socially constructed, publicly known, anticipated and accepted (March and Olsen, 1996,p. 249). Hence, a robust understanding of shareholder power in the context of global governance requires close attention to the institutions through which this power is expressed. As Barnett and Duvall (2005) convey: Global governance involves formal and informal institutional contexts that dispose that action in directions that advantage some while disadvantaging others. Understanding power in this way makes it much more difficult to approach global governance purely in terms of cooperation, coordination, consensus and/or normative progress; governance is also a matter of institutional or systemic bias, privilege and unequal constraints of action. Conversely, if global governance itself is conceptualized in terms of production, reproduction, and/or transformation of such asymmetries, then theorization and analysis, must, by logical necessity, rest on a conception of power that sees power as interaction through diffuse social relations. (p. 17)   In a series of three papers that comprise the dissertation, I examine how institutional investors work through particular formal and informal institutions to call for greater corporate compliance with human rights. Towards this end, I bring together a previously disparate but mutually invigorating set of literature on corporate governance, transnational activism and responsible investment. This dissertation does not take an agency-centred approach, but focuses instead on structural determinants to explain particular outcomes stemming from investor advocacy on human rights.9 A key concern motivating this project is that the tactics, organizational forms and discourses used by institutional investors in their efforts to diffuse human rights norms reveal a number of tensions between the local and global, as illustrated by the contrasting assessments of responsible investment presented at the start of the chapter.                                                 9 This analytical choice is further explained in Chapter 2.   20 Focusing on the institutions that shape transnational investor advocacy on human rights not only sheds light on why these contrasting assessments arise, but also offer the promise of signalling what kinds of institutional adaptations may be needed in order to effectively bridge these divides in ways eventually align capitalism and long-term sustainability.   1.3 Methods and data I discuss the specific research design, data sources and methodological approach in separately in each of the three papers. However, here I provide a general indication of common aspects that lend cohesion to the project – namely the comparative outlook and largely qualitative approach. The geographic scope of the analysis focuses on investors and investor-driven governance networks that are primarily located in the Global North, though specific instances of human rights violations that trigger investor activism cover a much broader set of countries indicated by the darker shading in Figure 1.5 below.    21  Figure 1.5. Geographic spread of countries of concern in examined cases of investor led human rights activism The three papers are primarily qualitative and comparative in approach. In the first paper, I compare public pension funds in the Netherlands and Norway in order to probe why seemingly similar investors would respond to the same case of weak corporate compliance with human rights standards in dramatically different ways. In the second paper, I map thirty-two investor-driven governance networks across North America, Europe, the Asia-Pacific, Africa and Latin America. I qualitatively evaluate the governance rules that delimit these networks in order to identify the ways in which institutional choices may serve to consolidate or diffuse shareholder power, based on the degree to which civil society voices are accommodated therein. Although the final paper focuses on shareholder activists in one country – Canada – I comparatively assess shareholder proposals filed on sustainability issues over thirty-five-year period [1982-2017] in order to trace discursive shifts in investor activism on human rights, paying particular attention  22 to whether these shifts signal tangible transformations to power relations between private actors and other social actors such as communities where human rights violations are alleged to occur.  One challenge involved in the study of investor activism is that a bulk of this activity is conducted in private through confidential interactions between firms and investors. For this reason, most analysis of shareholder activism focuses on publicly available data such as shareholder resolutions and public announcements made by investors on particular responsible investment topics (Becht et al., 2008; Gifford, 2010). In an effort to extend the reach of my analysis, I made a deliberate effort to capture both public and private forms of investor activism. The first paper focuses on private engagement between Dutch and Norwegian investors and the Chinese state-owned oil and gas company Petro-China through letter-writing, press-releases and phone-calls. In addition to archival material pertaining to this case, I also conducted field-work in the Netherlands and Norway so as to include elite interviews as part of the data repository. It should be noted that the number of individuals involved in direct engagements between investors and firms tends to be very small (2-5). This poses a significant risk for studying direct corporate engagement as the failure to secure elite interviews and access to private correspondence between firms and investors poses a major constraint to data collection and triangulation. The limitations confronted with regard to data triangulation at both project sites is further discussed in Chapter 3. In addition to private, direct engagement, the third paper examines public and collective forms of investor activism on human rights. It focuses on the specific institution of shareholder proposals and examines instances where such proposals were filed by single institutional investors, as well as cases where groups of investors acted collectively. The breadth of cases analysed cover a wide variety of investor types: religious investors, mutual funds,  23 individual activists, non-governmental organisations and pension funds, thereby providing additional analytical leverage on the key themes of concern.   1.4 Chapter overview  As previously mentioned, the objective of this dissertation is to advance a broader, relational understanding of shareholder power by exploring the institutions through which this unique form of private power is expressed. To provide necessary background for this key concept, Chapter 2 provides a deep-dive into contrasting images of shareholder power across key theoretical traditions that have been applied to the study of shareholder activism. It identifies a number of gaps in extant understandings of shareholder power and explains how the three papers that comprise the thesis respond to the identified gaps.  In Chapter 3, the research question I explore is: When faced with a portfolio company that is a systematic violator of human rights norms, why would an institutional investor with strong human rights commitments continue engaging the company rather than divesting from it?  This question arises from examining the diverging responses by public pension funds in Norway and the Netherlands to allegations of complicity in human rights violations by their investee firm PetroChina (a state-owned Chinese oil and gas company) in Burma and Sudan. To probe this question, I provide a theoretical framework for understanding shareholder activism using Hirschman’s (1970) work on exit and voice that lays out the conditions under which divestment and engagement would be expected to occur. However, as my systematic analysis of observed outcomes in both cases  reveals that theoretical expectations do not hold. In particular, engagement with PetroChina by the Norwegian fund and divestment from the firm by the Dutch fund cannot be adequately explained by existing theoretical accounts for why voice and exit  24 options are used in shareholder activism, because the factors commonly identified as correlates of shareholder engagement (voice) in the former case, are also present for divestment (exit). I conclude that the case challenges the explanatory power of existing accounts for why investors may prefer engagement over divestment, and highlights the need for additional hypotheses in explaining variations in investor responses to human rights violations by firms.  One proposal to this effect includes considering how the presence of a state-based veto point on divestment and engagement decisions for public pension funds, and the institutional character of target firms (particularly state-owned enterprises) may affect an investor’s appetite for divestment. An expanded framework for evaluating tactical options available to investors advocating for corporate compliance with global human rights norms would advance not only the literature on shareholder activism specifically, but also theories of transnational advocacy more broadly by bringing in private actors from the periphery of this literature into the main frame. The chapter offers commentary on how, shareholder power, although leveraged in the context of transnational corporate governance, is still shaped and influenced by domestic institutions of the state. This finding provides support to the arguments of scholars who suggest that processes of transnational governance are shaped by an interaction between global and domestic forces, and that the authority of the state is not necessarily completely eroded as a result of these processes, though it is certainly contested (Lipschutz, 1997; Soderholm, 1997; Held and McGrew, 2002; Clapp and Fuchs, 2009; Robertson, 2012; Carbonnier, 2011; Jasanoff and Martello, 2004, Tarrow and Tilly, 2009).  Chapter 4 moves on from investor activism by single institutional investors to collaborative platforms for investor advocacy. MacLeod and Park (2011) refer to these coalitions or alliances led by investors formed around a specific public goods issue or issue-areas as  25 investor-driven governance networks (IGNs). MacLeod and Park’s definition of IGN’s is conceptually valuable. However, their claim that non-governmental organisations (NGOs) and social activists also participate in IGNs, either formally or informally is worth probing, and offers a unique vantage point from which to unpack intra-network relations in transnational advocacy politics. In particular, Deborah Avant’s (2006) work on private authority reminds us that despite similar rhetorical commitments to sustainability, INGOs and private actors draw from vastly different sources of authority when engaging in transnational politics. Avant notes that the private sector’s profit imperative offers a very different materially-derived source of authority held by investors, from the moral or expert authority possessed by INGOs such as Greenpeace and the WWF. Hence, investigating how these two very different types of advocacy actors interact through the auspices of IGNs could provide a fascinating account of how diverse forms of authority compete and complement each other at the global level.  To address this issue, the chapter’s main research question is:  How do IGN governance arrangements accommodate or exclude specific forms of authority that are held by INGOs? The chapter first provides a brief review of existing work on NGOs in the context of shareholder activism, whereby my assessment highlights a very low level of engagement with issues of power and authority. Having justified the theoretical need to advance work on this subject, I present an analytical framework for assessing INGO-IGN relations. I suggest the degree to which an IGN is open or closed to civil society input affects the range of sources it can draw from to legitimize its authority. I then extend the empirical reach of previous work on IGNs by mapping civil society integration in thirty-two such networks, across a range of advocacy themes and geographies. I measure civil society integration in IGNs by focusing on two attributes: the degree to which IGNs a) recognize civil society as key to their mandate and b) enable the participation  26 of INGOs in delivering that mandate, either formally through board governance structures and mandate-implementing activities or informally through ad-hoc collaboration.  Overall, Chapter 4 demonstrates how choices around institutional design (formal and informal) limit whose voices and whose norms shape global investor advocacy on human rights. Specifically, IGNs are usually designed to act as “industry-associations” that advance the interests of investors, rather than as “aggregators of multi-stakeholder consensus” demanding corporate compliance with global human rights norms. Chapter 4 also shows that studying governance arrangements within IGNs matters against claims that activism by institutional investors is poised to be a crucial and potentially transformative pathway for aligning capitalism with social and environmental sustainability (Monks and Minnow, 1995; Hawley and Williams, 2000; Clark and Hebb, 2004). If investors are to live up to this expectation, unpacking their interactions with the very organisations tasked with articulating environmental and social demands within global governance can offer insights into whether shareholder power in this context serves to reinforce private interests, or whether IGNs can indeed generate a better balance between the interests of society and profit maximization. The chapter concludes with the observation that evaluating INGO-IGO relations from a governance lens presents the opportunity to re-design or reform IGNs in ways that can catalyze this transformative potential of shareholder activism. The study is exploratory and descriptive in nature, i.e. it seeks primarily to map variations in the degree to which INGOs include civil society participation. It does not take on the onus of theorising factors that produce the observed variations, and stops short of evaluating whether IGNs that access more diverse sources of authority are more successful or impactful in delivering their mandate.   27 Chapter 5 conceptualizes shareholder advocacy not simply as a mechanism by which investors address principal-agent dilemmas that arise from inefficiencies and imperfect information in global markets, but rather as a discourse structure that has the potential to enact, confirm, reproduce or challenge relations of private power (and its abuse) in society (Fairclough, 2015; Wodack and Meyer, 2016). To illustrate the validity of my “socially-embedded” conceptualisation of shareholder power, I conduct a content analysis of 72 Canadian shareholder proposals on human rights from 1982-2017 using a critical discourse approach. The chapter draws attention to how images of corporate social responsibility and shareholder power conveyed by investor discourses have evolved against changing global norms. It also demonstrates how the global “mainstreaming” of responsible investment has impacted the frames that Canadian investors use to improve firm compliance with global human rights standards. It further sheds light on how the formal rules around the filing of shareholder proposals has resulted in discursive choices that advance private authority and interests, over those of communities actually being impacted by corporate human rights violations. In closing, Chapter 6 summarises the key conclusions of the dissertation project by reflecting on the question: Who wins and who loses in the New Global Politics of Responsible Investment? The dissertation’s central finding is that a closer inspection of responsible investment reveals that the movement is yet to deliver on meaningful transformations that redress existing social power hierarchies that advantage private actors over other social actors, especially for communities where corporations may be complicit in or directly contributing to human rights violations. This finding troubles the robustness of claims by scholars and practitioners alike that responsible investment is indeed capable of aligning the public and private interests.   28 Across these three studies, we see that the exercise of shareholder power, as a form of private authority, is highly contingent on institutions that mediate how this power is leveraged, by whom and for whom. Cumulatively, the dissertation establishes a series of important findings. First, I contend that state-based institutional checks on shareholder power can affect the tactical strategies employed within human rights advocacy by investors, and in some cases limit the scope for ethical deliberation on these strategic choices. Moreover, the mainstreaming of responsible investment has relied heavily on a narrative of values-alignment with stakeholders, yet governance mechanisms to incorporate stakeholders within human rights advocacy by investors remain rare, raising questions around the ethical representation of human rights claims.  Finally, the exercise of shareholder power, contrary to dominant viewpoints, is not ethically neutral. Rather it conveys and reinforces particular relations of social power that do not necessarily advance global human rights, despite the appearance of doing so. In this regard, the dissertation confirms that shareholder engagement and responsible investment are tools for the financial elite and not the powerless. The implications of these findings are that if capital is truly to serve society, and not the other way around, a number of institutional adaptations will be necessary to enhance the democratic qualities of global investor activism in ways that increase spaces for beneficiaries, civil society and affected communities to participate in and inform investment decision-making.   29 Chapter 2: Theoretical Approaches to Shareholder Power The study of shareholder activism straddles a number of disciplines ranging from law, psychology, business ethics, comparative corporate governance, sociology and organizational studies. As these different approaches can create conceptual fuzziness, I follow Goranova and Ryan’s (2014) definition of shareholder activism as “actions taken by shareholders with the explicit intention of influencing corporations’ policies and practices” (p. 1232). The relationship between shareholder activism or advocacy and shareholder power is rarely made explicit. In a general sense, the ability of investors to engage in activism in the first place is seen as being derived from certain sources of power: legal (shareholder rights), material (financial) or normative (value-based or religious moral authority). However, shareholder power itself can be difficult to define and conceptualize, depending on whether one focuses on the sources, function or mechanisms by which this power is exercised. While I am interested in all these aspects of shareholder power, this dissertation approaches the concept in a novel way by considering the types of power relations at play when shareholders engage in transnational advocacy. I suggest that within the domain of human rights, investor-driven activism animates four sets of power relations between investors and a) firms, b) states, c) advocacy coalitions working on similar issues and d) affected communities on whose behalf the advocacy is initiated.   In this chapter, I review the existing literature on shareholder activism with three objectives in mind: a) to examine the contrasting images of shareholder power used, b) to evaluate the extent to which existing theoretical approaches account for all types of power relations noted above (i.e. the extent to which multiple stakeholders are accommodated in conceptualizations of shareholder power), and c) to consider the different scales at which  30 shareholder power is conceptualized – local, national or global (i.e. whether the way shareholder power is conceptualized travels reasonably well to the transnational domain). As this chapter outlines, reviewing the literature from these dimensions suggests that approaches to shareholder power that tend to dominate the business ethics and finance literature confront some limitations. First by viewing shareholder power as a pre-requisite for shareholder activism, the literature tends to feature shareholder power as an independent variable. That is, shareholder power is studied in terms of how it may influence a particular advocacy outcome, but not how activism itself may subsequently affect shareholder power. Existing works do not examine how the use of shareholder power in given contexts and over certain issues may serve to further legitimize this particular form of private power over other complementary or competing sources of power in the international political system. A second limitation is that the work on shareholder advocacy so far has tended to frame shareholder power mainly as a relationship between investors and firms, such that the implications of shareholder power as it relates to other actors operating at different political levels are rarely considered.  By systematically reviewing the existing literature in this chapter, I also aim to illustrate that a main theoretical contribution of this thesis lies in its effort to bridge disciplinary silos between business ethics and international relations. Specifically, as the three papers which comprise the dissertation will illustrate, existing analysis of shareholder activism can be fruitfully extended by considering perspectives on power that are offered by scholars of international relations, particularly when investor-led advocacy is viewed through the lens of global governance and transnational political action. The remainder of this chapter analyses the contributions and limitations of approaches used within the business and finance literature on shareholder activism, namely agency theory (Section 2.1), organizational theory (Section 2.2),  31 stakeholder theory (Section 2.3) and socio-political approaches (Section 2.4) and how these inform extant understandings of shareholder power. I then explain in Section 2.5 why socio-political approaches can serve as an entry-point to introduce a more transnational conception of shareholder power that may enable the concept to migrate more easily from a narrow field of investor-firm relations to a broader space of global political action, involving relations between multiple social actors and impacts at different levels of governance.   2.1 Agency theory Stemming from an interest by economists to broaden existing literature on financial risks and risk distribution, agency theory sought to understand issues that arise when one party delegates work to another by means of a contractual relationship (Eisenhardt, 1989). The theory has been especially relevant to the study of corporate governance – the system of rules, practices and processes by which a company is directed and controlled. Here, the relationship between investors and firms is explained within the frame of the principal-agent problem, particularly within Anglo-American systems of corporate governance10 (Eisenhardt, 1989; Fama, 1980; Jensen and Meckling, 1976).                                                  10  The “Varieties of Capitalism” (VOC) research program distinguishes primarily between liberal market economies and coordinated market economies based on how firms coordinate their activities with other key actors. As Aguilera and Jackson (2003) summarise: in most comparisons researchers contrast two dichotomous models of Anglo-American and Continental European corporate governance (Becht & Röel, 1999; Berglöf, 1991; Hall & Soskice, 2001; La Porta, Lopez-de-Silanes, Shleifer, & Vishny, 1998). They stylize the former in terms of financing through equity, dispersed ownership, active markets for corporate control, and flexible labor markets, and the latter in terms of longterm debt finance, ownership by large blockholders, weak markets for corporate control, and rigid labor markets. Yet this classification only partially fits Japan and other East Asian countries (Dore, 2000; Gerlach, 1992; Khan, 2001; Orrú, Biggart, & Hamilton, 1997; Whitley, 1992), the variations within Continental Europe (Barca & Becht, 2001; Rhodes & van Apeldoorn, 1998; Weimer & Pape, 1999; Whittington & Mayer, 2000), Eastern Europe (Martin, 1999; Wright, Filatotchev,& Buck, 1997), and multinational firms (Fukao, 1995). The caution here is that despite the rich description found in the VOC research literature, the challenge remains to conceptualize cross-national diversity in corporate governance systems and to identify the key factors explaining these differences.   32 Financial markets in the United States and United Kingdom are organized such that owners of capital who are shareholders of listed companies, delegate the running of the company to the firm’s management (Solomon, 2007, p. 17). This separation of ownership and control (Berle and Means, 1932) leads to a notorious agency problem where firm managers (agents) do not necessarily act in the best interests of shareholders (principals) (Shliefer and Vishny, 1997). For example, as the recent global financial crisis strikingly demonstrated, managers may pursue short-term profit maximization or egoistic personal goals over maximizing shareholder value11 over the long-term (Boatright, 1999; Sykes 1994). Hence, the challenge for shareholders is how to align their interests with those of corporate management given information asymmetries and incomplete contracts (Shleifer and Vishny, 1997). The reduction in shareholder value that can result from the agency problem is known as residual loss. In addition to residual loss, Solomon (2007) reiterates that an important and basic assumption of agency theory is that it is costly for principals to verify an agent’s actions are in line with the former’s interests. Likewise, agents expend resources to adequately demonstrate they are maximizing shareholder value (such as through added accounting and reporting procedures). Among the strategies to mitigate this particular problem, the allocation of stronger voting rights that endow shareholders with more power to influence managerial discretion has been explored in significant detail (Davis and Thompson, 1994; Shleifer and Vishny, 1997; Useem, 1996). Shleifer and Vishny (1997), for example, argue that ownership rights would improve existing information systems and increase transparency, thus minimizing agency costs. Other                                                 11 A company’s shareholder value deepens on strategic decisions made by senior management, including the ability to make wise investments and generate a healthy return on invested capital. If this value is created over the long term, the share price increases and the company can pay larger cash dividends to shareholders (Investopedia, 2018).  33 mechanisms include the threat of takeover as a disciplining force over managers, divestment of capital, the passing of shareholder resolutions where a group on shareholders collectively lobby the company on issues they are dissatisfied with, or one-to-one dialogue and engagement with firms (Solomon, 2007). As applied to the context of shareholder advocacy, shareholder power itself is under-specified within agency theory. For example, although Jensen and Meckling’s (1976) foundational work is fundamentally driven by concerns with the abuse of power by corporations, defining shareholder power as a countervailing force is not addressed. Power is defined very narrowly as the decision-making abilities of firms, where power abuse primarily entails deviation from the expectations of investors (325). In general, agency theory is heavily reliant on assumptions of self-interest, bounded rationality and risk aversion. The strength of the theory is that it allows room to understand situations where the principal and agent engage in cooperative behavior despite differing goals and attitudes towards risk. However, it almost exclusively emphasizes the avoidance of responsibilities by agents as self-interested behavior and sees the principals as being at the mercy of agents (Bond, 1992).  Subsequent work on the rise of investor capitalism has challenged the direction in which power is exercised. Rather than being at the mercy of agents, studies by Useem (1996) and Hawley and Williams (1997) for example, demonstrate that investors can and do exert tremendous influence over firms due to a number of developments in the market such as the concentration of share ownership among institutional investors. This concentration serves to incentivize active monitoring of firms over divestment and provides a material basis through which to understand shareholder power, though this is not made explicit by the scholars themselves. Hawley and Williams’ (1997) work widened the scope of discussions on power  34 relations by asking, for instance, whether developments toward the concentration of ownership among institutional investors are compatible with a shared sense of the common good (Hawley and Williams, 1997). Despite these promising openings however, the locus of power itself is still primarily discussed within the confines of a bilateral relationship between firms and owners of capital. As a result, I am inclined to echo Eisenhardt’s (1989) view that “agency theory presents a partial view of the world that, although it is valid, also ignores a good bit of the complexity of organizations” (p. 57-74), as well as key critiques of agency theory offered by Aguilera and Jackson (2003) in their influential work on comparative corporate governance. With these scholars, I find that agency theory, as applied to studies of shareholder activism, tend to overlook the diverse identities of stakeholders within the principal-relationship and echo their call that comparative research must address this “social construction” of these diverse interests (Maurice, Sellier, & Silvestre, 1986 in Aguilera and Jackson, 2003). Jackson and Aguilera (2003) marshal a range of compelling arguments in this regard. They suggest that agency theory overlooks important interdependencies among other stakeholders in the firm because of its exclusive focus on the bilateral contracts between principals and agents—a type of “dyadic reductionism” (Emirbayer & Goodwin, 1994, in Jackson and Aguilera, 2003, p. 449). In addition, agency theory retains a thin view of the institutional environment influencing corporate governance. Within the literature, the theory’s focus on  shareholder rights does not capture the entire complexity of institutional domains by limiting actors’ financial behavior to the effects of law (La Porta, Lopez de Silanes, & Shleifer, 1999). These shortcomings lead Jackson and Aguilera (2003) to assert that “the unmet theoretical challenge, in comparative studies, remains to conceptualize corporate governance in terms of its embeddedness in different social contexts” (Dacin, Ventresca, & Beal, 1999; Granovetter, 1985 in Aguileria and Jackson 2003, p. 449).   35 Recognizing the limits of agency theory, this dissertation departs from a focus on exploring shareholder power purely from the lens of the principal-agent problem in order to shed light two on neglected properties of shareholder power – its relationship to a broader set of actors beyond the firm, and the idea that shareholder power itself is socially constructed through an interplay of agents (responsible investors, the state, communities, NGOs etc.) and structures (institutions, norms etc.). Nonetheless, although I acknowledge this interplay, in order to keep the scope of the dissertation feasible, I focus mainly on the structural aspects, rather than individualism and micro-foundations as explanatory mechanisms.12  My emphasis on structure throughout the dissertation is informed by Margaret Archer’s (1995) morphogenetic approach and her subsequent work on structural properties and powers in the context of social action (Archer, 2000). This research strategy aims to avoid the conflation of agency and structure in analytical work. Although Archer does not dispute the co-constitution of agents and structures on philosophical terms, as argued by scholars such as Anthony Giddens (1984), she does argue these are analytically distinct phenomena that operate at different timescales. The framework she uses to make this argument is the morphogenetic sequence (Archer, 1995, pp. 165-194). At any particular moment, pre-existing structures constrain and enable agents, whose interactions produce intended and unintended consequences, which lead to structural elaboration and the reproduction or transformation of the initial structure. The resulting structure then provides a similar context of action for future agents. Likewise, the initial pre-                                                12 Many scholars have stressed the need to make the micro-foundations of institutional theory more explicit (DiMaggio and Powell, 1991; Zucker, 1991). Scholars seeking to focus on these aspects draw on work by Goffman on interaction ritual to Wieck on sensemaking (Powell and Colyvas, 2008). Powell and Colyvas (2008) chapter the Microfoundations of Institutional Theory in Greenwood et al’s (2008) edited volume on organizational institutionalism offers a very helpful overview.   36 existing structure was itself the outcome of structural elaboration resulting from the action of prior agents (Archer, 1995, pp. 165-194). So, while structure and agency are interdependent, by isolating structural and/or cultural factors which provide a context of action for agents, it is possible to investigate how those factors shape the subsequent interactions of agents and how those interactions in turn reproduce or transform the initial context.  For example, Chapters 3-5 reveal a number of ways in which communities, as sites of human rights violations, have very limited institutional pathways to engage in responsible investment advocacy on human rights, even when such advocacy is initiated on their behalf. This can be explained in part by the rules and norms at the disposal of institutional investors that channel shareholder power in ways that reinforce the materially-derived authority of investors in their role as providers of capital, by excluding or limiting alternative non-material sources of authority in the context of responsible investment practices.13 Here Weber’s (1978) work is an important reminder that the concept of  “embeddedness” as applied to responsible investment is helpful because he stresses that economic action is also social action oriented toward others and may be constrained by noneconomic objectives or supported by noneconomic social ties (Streeck, 2002 in Aguilera and Jackson, 2003, p. 449).   2.2 Organizational theory  Another widely-used theoretical frame applied to the study of shareholder advocacy is Hirschman’s (1970) theory of the firm. Hirschman argues that members in diverse organizational contexts can express their degree of satisfaction or dissatisfaction either through exit, by voicing                                                 13 I elaborate on these connections to authority in Section 2.5.   37 complaints with the objective of inducing changes, or remaining loyal. The theory is frequently applied on its own, or in combination with other theories to understand how investors navigate the menu of monitoring options available to them in order to alleviate agency problems, and to what effect. In Exit, Voice and Loyalty (EVL), Hirschman (1970) discusses political power and market power primarily with regard to how voice or exit may affect both. He observes, “[p]olitical power is very much like market power in that it permits the powerholder to indulge either his brutality or his flaccidity” (p. 58). A key argument in this seminal work is that although exit may be a viable option, there is considerable risk that it may entrench the status quo, reinforce a cycle of decline and entail costs (Schumpeter, 2012). For instance, if you have invested heavily in a company that starts performing badly, then you may be better off agitating for a change in management (voice) rather than selling your shares at a loss (exit). Just as Latin American powerholders have long encouraged their political enemies and potential critics to remove themselves from the scene through voluntary exile (Hirschman, 1970, p. 60), we can imagine situations where the exit of vocal critics from a firm may actually serve to reinforce corporate malfeasance. The main point here is not that exit is bad but that exit (or the threat of exit) and “voice” work best together. Taken together these assumptions have gained purchase in studies of shareholder advocacy.  Typically, Hirschman’s work has been used to ground empirical studies that unpack processes of exit and voice involved in shareholder advocacy. Work in this tradition includes Beunza and Ferraro’s (2013) study of shareholder advocacy by religious investors in the United States. They use Hirschman’s framework, in combination with insights from cognitive organizational theories to reveal how shareholders’ identity as company “insiders” permits firms  38 to engage in learning and face-saving in light of corporate governance failings that are traditionally obscured in contentious and public interactions with activists who are external to the firm. In particular, Beunza and Ferraro (2013) refocus Hirschman’s concept of voice as a mechanism of influence in organizations through two routes: formulating emerging issues and their business implications and sensitizing corporations by influencing internal corporate debates (p. 43). They argue these mechanisms give shareholders a “moral voice” that complements the more contentious approaches used by other activists.  In a similar vein as Beunza and Ferraro (2013), a number of empirical studies on shareholder advocacy reference Hirschman’s voice and exit framework in order to categorize voice and exit options available to investors, while relying on alternative theoretical approaches to advance particular insights on the mechanics of the engagement process.14 Hirschman’s framework can also be found in Goodman et al’s (2014) analysis of how and why the voice and exit options are used in social shareholder engagement (SSE)15 through the analysis of seven engagements by three religious organizations in the United States and United Kingdom. Here, Hirschman’s frame is coupled with specific assumptions about how an investor’s identity affects the engagement process. Their analysis identifies distinct stages in the engagement process (issue raising, information seeking, change-seeking and outcomes) by religious investors. In contrast to much of the corporate governance literature, they find that religious organizations base exit and voice decisions on political rather than economic motivations, often using voice to further their beliefs and mission in society (196). In addition, they argue that Hirschman’s voice and exit                                                 14 See for example: Allen, Letourneau and Hebb (2012) 15 The term social shareholder engagement refers to the process by which institutional investors engage firms on social sustainability topics such as human rights.   39 options are dynamic, mutually reinforcing and not necessarily sequential. Divestment does not close the door to continuing external engagement with a company, nor is exit always the consequence of an unsatisfactory voice strategy (Goodman et al., 2014, 208).  In general, the application of Hirschman’s framework has added some texture to the concept of shareholder power. For some, the material source of shareholder power provided by agency theorists persists as the dominant point of departure. For instance, using the EVL frame to explain the growth of shareholder activism, Jahnke (2017) argues that market characteristics such as rising ownership concentration16 and the growth in passive assets17 means that the institutional investor community is both more powerful and more compelled to be involved in corporate affairs. Alternatively, Beunza and Ferraro (2013) and Goodman et al. (2014) offer perceived moral legitimacy as a different source of shareholder power than the more material conception derived from extensive share ownership. For example, Goodman et al (2014) note: “Shareholders with large holdings are able to apply economic pressure and can affect management decision making through the threat of exit and the divestment threat can be considered a form of voice” (p. 196). Research into the salience of shareholders has demonstrated that this traditional source of power (emphasis added) is not usually available to shareholders undertaking SSE (Gifford, 2010; Lee and Lounsbury, 2011). Furthermore,                                                 16 The author argues that the increased concentration of ownership and centralization of decision-making resulting from the rise of “fiduciary capitalism”—a “capital market in which large institutional investors purchase shares on behalf of ultimate owners to whom they have fiduciary responsibilities,” with these intermediaries exercising power on behalf of the owners—has increased the ability of shareholders to exercise power. Concentration of ownership in the United States. is substantial, with the top 5 investors controlling 47% of assets at year-end 2016, up from 36% in 2005, while the share of assets managed by the top 10 firms has increased from 47% in 2005 to 58% at year-end 2016. 17 Passive investing is an investment strategy that aims to maximize returns over the long run by keeping the amount of buying and selling to a minimum. The idea is to avoid the fees and the drag on performance that potentially occur from frequent trading. This incentivizes shareholders to engage in more active monitoring of firms through engagement, rather than selling shares in companies they are dissatisfied with.   40 Goodman et al., (2014, p. 196) explain that in SSE voice can be interpreted as social action, and can be used for the pursuit of justice or for wider social change, with financial impact being relatively inconsequential for smaller holdings typical of SSE (Hollenbach, 1973; Clark et al., 2008, Proffitt and Spicer, 2006; Sjostrom, 2010 in Goodman et al., 2014, p. 196-198).  However, given that religious investors are a small portion of the global institutional investor landscape, this view of shareholder power is highly contextual, with limited generalizability. One troubling aspect in Goodman et al’s (2014) reference to traditional vs. non-traditional sources of shareholder power as discussed in the quote above is the elision between two very distinct phenomena involved in the exercise of shareholder power: a) investor motivations for engaging in social shareholder activism (namely wanting to enhance justice, affect social change, and an emphasis on values-alignment over financial impacts), and b) the institutional characteristics of investors that may prompt desired changes in firm behavior (size of ownership stake). This elision is evident in the argument that for religious organizations, exit can sometimes function as a form of voice: “According to Hirschman, exit is an economic argument and entails revenue loss. However, the shareholdings of religious organizations tend to be small and therefore cannot rely on traditional shareholder power to have a financial impact on investee companies” (Goodman et al, 2014, p. 205-206). Hence, exit as a private and anonymous economic act was not an option. Rather by combining divestment with public statements, it was used as a political argument more in line with a voice strategy, described by Hirschman (1970) as “various types of actions and protests, including those that are meant to mobilize public opinion (p. 30). As noted, the application of Hirschman’s frame to scholarship on shareholder advocacy has added to our understanding of non-material sources of shareholder power, but all too frequently these arguments under-developed and implied, rather than explicitly addressed.  41 Furthermore, owing to the emphasis on the EVL framework as a tool to understand how investors work around certain types of coordination problems, organizational theory too confronts some of the limitations noted with agency theory. Although the EVL framework offers more room than agency theory to consider multiple social influences on shareholder advocacy, discussions of shareholder power typically focus on this concept as an explanatory variable for particular advocacy tactics or outcomes. Both approaches also tend to discuss shareholder power strictly within the context of investors and firms, rather than a more expansive view of how the state and civil society my affect shareholder power, or in whose interests this power is exercised. The latter point is important because responsible investment practices are meant to be aligned with long-term social and environmental sustainability. Rather than assuming this is always the case, the examination of shareholder power in this dissertation critically investigates whether this assumption holds in practice. In the three chapters that follow, I explore which corporate social responsibility and sustainability norms are advanced by responsible investors, and in particular evaluate how institutions mediating the exercise of shareholder power (such as shareholder engagement, investor-driven governance networks and shareholder resolutions) shape these norms. I do so in Chapters 3-5 by considering the implications of shareholder advocacy on human rights for the beneficiaries on whose behalf institutional investors are entrusted to serve as financial stewards, as well as the communities whose human rights are the subject of investor advocacy. A further observation from my review of this literature is that most studies adopting the EVL framework have focused on single-case studies within a specific jurisdiction or limited comparative work primarily within the Anglo-American context (typical cases are the Interfaith-Centre for Corporate Responsibility in the United States and the Ecumenical Council for  42 Corporate Responsibility in the United Kingdom). These studies mainly measure the impacts of investor-led human rights advocacy in terms of pro-CSR shifts in a firm’s policies and the preservation of financial value in the period following investor activism on a human rights issue.  This dissertation widens the lens on how impact is conceptualized. For example, in Chapter 5, my study of human-rights related shareholder proposals filed in Canada over a thirty-five-year period reveal a shift in investor discourses around human rights, whereby a steady erosion of purely ethical justifications for corporate responsibility for human rights based on notions of harm to particular communities has given way to predominantly hybrid justifications based on the idea that human rights violations pose risks to shareholder value. This discursive shift, I argue, heightens ethical ambiguity around the social responsibilities of investors (not just firms). By repeatedly legitimizing this form of persuasion in their responsible investment advocacy, investors may constrain themselves from speaking out against corporate-related human rights violations in perverse situations where the violations of such rights may actually be financially profitable.18   2.3 Stakeholder theory Stakeholder theory can be traced to the work of Freeman (1984) who proposes extending the focus of managers beyond traditional interest group of shareholders in order to understand the needs, expectations and values of groups previously perceived to be external to the firm.  Stakeholder theory has both normative (moral/ethical) and instrumental (profit/wealth                                                 18 On this point, I acknowledge Raj Thamotheram (Preventable Surprises) and Rob Lake’s(Authentic Investor) for calling attention to the idea that investors are not passive bystanders in how markets (and related dysfunctions) are created and sustained, as well as the need for greater personal accountability on the part of these investors for maintaining perverse incentives (interviews with author, November 2013).   43 enhancing) implications, since dealing with stakeholders can be regarded as a responsibility to meet the legitimate claims of all stakeholders, and/or as a means to maximize organizational wealth (Donaldson and Preston, 1995; Jones 1995; Jones and Wicks, 1999 in Ayuso et al., 2007).  The development of stakeholder theory ushered in one of the deepest fault lines in the corporate governance literature regarding the purpose of the firm – the shareholder/stakeholder debate.19 From shareholder perspective corporations have no duty to exercise “social responsibility” (Friedman, 1962), where social responsibility means the expending of corporate resources for socially beneficial purposes, regardless of whether those expenditures are designed to help achieve the corporation’s financial ends (Schaefer, 2008, p. 298). This view holds that the purpose of the corporation is to realize the specified ends of shareholders, with the caveat that those ends are legal and basically non-deceptive. In practice, those ends are nearly always to maximize the corporation’s profits (Schaefer, 2008). By contrast, stakeholder theory holds that the corporation should be run for the benefit of all stakeholders regardless of whether doing so maximizes the corporation’s profits (Freeman, 1984). The stakeholders of a corporation can be characterized as those persons vital to the continued survival and success of the corporation, and this is usually thought to include owners, management suppliers, employees and customers (Schaefer, 2008, p. 297). Stakeholder theory has been widely applied to issues related to corporate sustainability and social responsibility since it provides a suitable framework for understanding the relationship                                                 19 Within the political science/political economy literature on Varieties of Capitalism, scholars have distinguished Continental European/Japanese stakeholder systems of corporate governance, where labour is incorporated into corporate decision making through works’ councils for example as distinct from Anglo-American shareholder systems. These are highly stylized portraits primarily used to explain differences in finance, ownership, labor relations and the role of the market in corporate control or issues of institutional isomorphism in corporate governance.  44 between business and society (Clarkson, 1995; Donaldson and Preston, 1995; Waddock and Graves, 1997). By broadening the scope of influences on corporate decision-making beyond shareholders and the state, a major thrust of scholarship has focused on how firms decide ‘who or what really counts’ (Freeman 1994; Mitchell et al., 1997). Addressing this issue of “stakeholder salience” has meant that stakeholder theory has made a more focused effort to address the concept of power than the theories previously discussed.  In this respect, Mitchell, Agle and Wood’s (1997) theory of stakeholder identification and salience has proved to be highly influential in studies of shareholder activism, as it offers a framework to understand firm responsiveness (or lack thereof) to the demands of institutional investors.20 Mitchell et al.’s (1997) central thesis is that the salience of particular stakeholders can be identified by looking at three attributes: (1) the power of stakeholders to influence the firm, (2) the legitimacy of the stakeholder’s relationship with the firm and (3) the urgency of the stakeholder’s claim. Here, the authors make an important distinction about how stakeholders are defined in the first place. They suggest that on order to clarify the term "stake," we need to differentiate between groups that have a legal, moral, or presumed claim on the firm and groups that have an ability to influence the firm's behavior, direction, process, or outcome (Mitchell et al., 1997, p. 859). Influencers have power over the firm, whether or not they have valid claims or any claims at all and whether or not they wish to press their claims. Claimants may have                                                 20 I present Mitchell et al.’s (1997) framework in substantial detail here. I do this primarily because their framework has been highly influential in studies of shareholder engagement and has provided the theoretical underpinning for a number of studies on the topic. It should be noted however, I am not adopting their definitions of concepts like authority and legitimacy wholesale. These concepts are discussed at great length and with a high degree of contestation across the various theoretical traditions in international relations (particularly Realism, Liberalism, Constructivism and a range of critical theories).  In the interest of brevity, I have not added a comparative analysis of how these concepts are defined and understood differently by stakeholder theorists and IR scholars. I have chosen instead, to focus my discussion on areas where these concepts inhibit a relational, multi-scalar and socially-embedded understanding of shareholder power.   45 legitimate claims or illegitimate ones, and they may or may not have any power to influence the firm. Thus, power and legitimacy are different, sometimes overlapping dimensions, and each can exist without the other (Mitchell et al., 1997, p. 859). Within Mitchell et al.’s (1997) framework, power is defined in simple terms as the ability of one actor to influence another. The authors draw on the early Weberian idea that power is "the probability that one actor within a social relationship would be in a position to carry out his own will despite resistance" (Weber, 1947 in Mitchell et al., 1997) and echoed later by Dahl (1981) as "a relationship among social actors in which one social actor, A, can get another social actor, B, to do something that B would not otherwise have done" (p. 3). The emphasis that this power relationship is deeply tied to a misalignment of interests is noteworthy. To elaborate the mechanisms through which power is exercised, Mitchell et al. dismiss French and Raven’s (1960) typology of power for failing to incorporate a sorting logic to create mutually exclusive and exhaustive categories. Rather, they turn to Etzioni’s (1964) categorization based on the type of resource used to exercise power: coercive power, based on the physical resources of force, violence, or restraint; utilitarian power, based on material or financial resources; and normative power, based on symbolic resources (Mitchell et al., 1997, 865). Applied to the context of shareholder advocacy, investors are seen to possess power to the extent they possess or can gain access to coercive, utilitarian, or normative means to impose their will in a given set of relationships. However, this “access to means” is not a steady state, which is one reason why power is transitory: it can be acquired as well as lost (Mitchell et al., 1997, p. 865-866). With regard to the second pillar of their framework, Mitchell et al. (1997) observe that the notion of "legitimacy," loosely referring to socially accepted and expected structures or behaviors, is often coupled implicitly with that of power when people attempt to evaluate the  46 nature of relationships in society. Davis (1973), for example, distinguishes legitimate from illegitimate use of power by declaring, "in the long run, those who do not use power in a manner which society considers responsible will tend to lose it" (p. 314). Seeking to define a firm's stakeholders narrowly, some scholars also make an implicit assumption that legitimate stakeholders are necessarily powerful, when this is not always the case (e.g., minority stockholders in a closely held company), and that powerful stakeholders are necessarily legitimate. Despite this common linkage, Mitchell et al.’s stakeholder salience framework accepts Weber's (1947) proposal that legitimacy and power are distinct attributes that can combine to create authority (defined by Weber as the legitimate use of power), but that can exist independently as well. An entity may have legitimate standing in society, or it may have a legitimate claim on the firm, but unless it has either power to enforce its will in the relationship or a perception that its claim is urgent, it will not achieve salience for the firm's managers. This understanding leads Mitchell et al., to follow Suchman (1995) in defining legitimacy as "a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions" (in Mitchell et al., 1997, p. 867). The third and final pillar of the stakeholder salience framework is urgency. This is defined by Mitchell et al., (1997) as the degree to which stakeholder claims call for immediate attention. This definition involves two key components:  time sensitivity –  the degree to which managerial delay in attending to the claim or relationship is unacceptable to the stakeholder, and criticality –  the importance of the claim or the relationship to the stakeholder.   The stakeholder salience framework has proved quite valuable in allowing scholars to explore which attributes are likely to contribute to successful outcomes in shareholder advocacy campaigns, and the original work has benefitted from a number of useful extensions. For  47 example, combining resource dependency theory and stakeholder theory, Frooman (1999) argues that a resource relationship exists between a firm and its stakeholders, and an over-emphasis on power may obscure how dependencies may shape impressions of stakeholder salience. Frooman’s early work has been expanded upon in subsequent studies (Hendry, 2005; Gifford, 2009, 2010; Mattingly and Greening, 2002; Hamilton and Erickson, 2011). For example, Gifford (2010) further subdivides Mitchell’s three attributes of urgency, salience and power so as to get a richer picture of shareholder advocacy practices. Within Gifford’s work, legitimacy is enhanced to include ‘what a shareholder says, not just who they are’ and urgency is modified to include ‘a sense of intensity of engagement’ (Gifford 2010, p. 28). Three moderating factors are also added in Gifford’s model: coalition-building, relative size of the investor and the company, and the values of the managers. In bridging normative and instrumental aspects, Gifford finds that tying shareholder advocacy to the business case is seen as a key criterion which increases the likelihood of a positive response by corporate management (Gifford, 2010; Guay, Doh, and Sinclair 2004). While such research has confirmed the merits of the Frooman model, it has also found it to be too simplistic to explain the true variety of stakeholder–firm relationships encountered. Going beyond formal rules and the investor-firm dyad, Hendry (2005) adds factors such as a stakeholder’s communicative ability and the potential to actually form an alliance, and type of influencing strategy adopted as key variables that may explain stakeholder salience through an analysis of four environmental non-governmental organizations (ENGOs).  In general, the application of stakeholder theory to gain analytical leverage on shareholder power encounters some difficulties. For instance, what Gifford terms as sources of shareholder power might more accurately be described as mechanisms through which this power is exercised. In addition, conceptually differentiating issue salience from actor salience is needed  48 for further clarity. Furthermore, the criticism that much traditional managerial stakeholder analysis fails to capture broad structural societal patterns of power also persists in analyses of shareholder advocacy to date (Jones and Fleming, 2003). For instance, the role of the state and public policy was not included in revised iterations of stakeholder analysis proposed by Frooman and Gifford (Reed, 2002), despite the fact that the state is a crucial intermediary in corporate governance.  Taken together, the stakeholder salience framework offers a higher degree of dynamism than agency theory and the EVL framework. The assumptions that stakeholder attributes are variable, socially constructed and may not always be consciously exercised, potentially create more room to advance a broader understanding of shareholder power. Unfortunately, while the framework has expanded the scope of who counts as a stakeholder from the view of firms, this is not the case when viewed from the investor perspective. As such, stakeholder theories do not fulfil their promise of advancing a more holistic understanding of shareholder power because existing empirical work has not actively sought to understand how shareholder power affects stakeholder relations beyond the firm, and at multiple levels of governance. In this respect, stakeholder theories that are applied to understanding shareholder power have not truly escaped the investor-firm dyad and have remained rooted in the Mitchell et al.’s (1997) proposition of the centrality of managers in stakeholder theory, illustrated when they note that “although groups can be identified reliably as stakeholders based on their possession of power, legitimacy, and urgency in relationship to the firm, it is the firm's managers who determine which stakeholders are salient and therefore will receive management attention” (p. 871).    49 2.4 Socio-political approaches A substantive critique of agency, organizational and stakeholder theories relates to the concept of embeddedness. According to Granovetter (1985), embeddedness is a matter of the degree to which economic behavior is affected by or submerged in social relations. Critical, sociological and constructivist perspectives, though prevalent in the literature on corporate social responsibility (CSR)21, are only recently being taken up in studies of responsible investment and shareholder advocacy. Key themes that have been pursued within this “social-studies of finance” research agenda include cultural ethics and the relationship between ethical behavior and cognition (sensemaking). By emphasizing the social rather than the economic (and thus, reductionist) view of organizational activities scholars have followed arguments advanced by interpretive sociology (Garfinkel, 1967; Schütz, 1967), the sociology of knowledge (Berger and Luckmann, 1966), social movements (Tarrow and Tilley, 2009) and social psychology (Weick, 1979). It also includes and builds on the work of constructivist scholars (Wendt, 1992; Adler, 1997; Checkel 1998) to include collective ideas and identity, which may be analytically prior to material interests.  As pointed out by Basu and Palazzo (2008), a sensemaking perspective allows “studying those factors that might trigger or shape CSR activities in the first place” (p. 123) and thus, reveal the critical differences among organizations in terms of how they enact their notion of responsibility (in Zarbafi, 2011, p. 16). Influence emerges, in a retrospective sense, out of the reciprocal process of how both parties involved, the investor and the corporation, make sense of                                                 21 See for example Gond and Moon (2012); Gond, J.P. and Piani, V. (2013); Gond, J.P. and Boxenbaum, E. (2013). Cabantous, L.,; Gond, J.P. and Johnson-Cramer, M. (2009).   50 the actions that have been performed by the counterpart.22 This new view on influence or power brings into play what Schelling (1960) has called the credibility of enforcement mechanisms. “Only if an enforcement mechanism such as filing shareholder resolutions etc. is credible, or framed differently, makes sense both internally and externally for the targeted corporation, it succeeds in provoking the desired change in behavior” (Zarbafi, 2011, p. 97). As Zarbafi (2011) insightfully explains this approach advances the agency model in that it looks at managerial discretion and active ownership as both necessary for transforming equivocal data into sensible information. In this context, a main challenge for the principal-agent problem confronting firms and investors was historically attributed to information asymmetries between the two parties. However, applying a sensemaking perspective to the principal-agent problem suggests that we are no longer predominantly confronted with the problem of information asymmetry that has to be reduced via governance structures that align the interests of passive (and powerless) investors with those of active (and powerful) managers. Rather, we face the problem that when information asymmetry exists, participants tend to perceive different cues and attach different meanings to their observations (Vlaar et al., 2006 in Zarbafi, 2011, p. 98).   We are thus confronted with an information plurality resulting from the different ways of how participants make sense of equivocal data which has to be reduced via collective sensemaking among both parties involved (Zarbafi, 2011, p. 98).                                                  22 A recent example of this type of scholarship includes Sabina DuRietz’s (2014) work on how accounting practices are interpreted as information in the context of responsible investment. DuRietz’s work is influenced by work on epistemic practice (Kalthoff, 2005; Knorr-Cetina, 2001; Rheinberger, 1992a). Instead of focusing on frames for interpretation, their origin in economics and how these shapes the behaviour of their users, which the social studies of finance already do (e.g., Beunza & Garud, 2007; Callon, 1998; MacKenzie & Millo, 2003), DuRietz focuses on the interaction between analysts and accounts that become identified as information and input to such models (DuRietz, 2014, p. 397).    51 This shift has significant implications for how shareholder power is understood and underscores the idea that shareholder influence should not be solely viewed as a function of power derived from access to resources or resulting from a central network position. Rather, the question “who counts is largely a function of the identity orientation of the organization doing the counting” (Zarbafi, 2011, p. 109). As a result, Zarbafi (2011) concludes that researchers in the field of responsible investment should thus become more creative and explore phenomena such as investor responsibility or investor influence from multiple perspectives.   Scholars drawing on social movement theories explore the agency problem and corporate control from a sociological perspective. Scholars such as Davis and Thompson (1994) argue “corporate control is inherently political, and politics is accomplished by coalitions of mutually acquainted actors that recognize or construct a common interest … Social movement theory adds insight into the process by which actors translate shared interests into collective action” (p. 152, in Zarbafi, 2011, p. 75).23 For example, Reid and Toffel (2009) and Lee and Lounsbury (2011) consider the impact of environmental shareholder resolution filing on the environmental performance of corporations from a social movement perspective, while Jackie Cook’s (2012) work examines the role of shareholder resolution filing in advancing a public policy agenda.  Collectively, social movement theories as applied to responsible investment have focused on identity, framing and network politics as key variables in explaining particular shareholder advocacy outcomes and processes (Proffitt and Spicer, 2006; Macleod, 2009; Macleod and Park, 2011). Social movement theory presents shareholder power as a function of political struggles                                                 23 Research building upon this argument includes the work of den Hond and de Bakker (2007), Derry and Waikar (2008), King (2008), King and Soule (2007), and Rao et al., (2000).  52 between investors and managers as well as social structures going beyond the organization (Davis and Thompson, 1994). Nonetheless, a shortcoming of this approach is that the embeddedness of corporate governance in social processes both within and between organizations is neglected (Zarbafi, 2011, p. 75), a gap that is taken up by social constructivists. For social constructivists, shareholder advocacy is viewed as a process of “ongoing social constructions, emerging from the skillful accomplishments through which organizational members impose themselves upon their world to create meaningful and sensible structure” (Morgan, 1980, p. 617). While the social constructivist research agenda in international relations sets out to “problematize the identities and interests of states, to open up the historical constitution of the states-system and to reflect on issues of systemic change” (Bieler and Morton, 2008, p. 107), applied in the context of globalized market relations, the unit of analysis here shifts to private actors. Nonetheless, both traditions share the view that “intersubjective frameworks of meaning are attached to social norms, which in turn are not taken as simple descriptive categories, but as components of generative structures that shape, condition and constrain action, (Ruggie, 1982, 1983, 1993 in Bieler and Morton, 2008, p. 107).  Bieler and Morton (2008) argue that a challenge with social constructivism is that it is grounded in an idealist understanding of transformations in social relations because of the dis-embedding of intersubjective ideas, norms and values from the social relations in which they cohere. For these scholars:  The recurring questions are therefore: Whose values and beliefs have constituted or embodied state identities and interests and the relevant constitutional structure of the international society of states? Which agents shape the core intersubjective beliefs of underlying social and world orders? Why does a particular set of ideas become part of the structure and not another? As it stands, there exists an under-theorized notion of power across social constructivist perspectives that fail to ascertain whose interpretations come to constitute the social world and why they do so (Bieler and Morton, 2008, p. 109).   53 Building in a more critical lens on constructivist approaches to shareholder power could offer a viable route to address this concern, as illustrated by DeGoede’s (2001) analysis of legitimacy deficits in finance. DeGoede (2001) does not concentrate on the material structure of financial markets and argues instead that “in order to criticize the legitimacy deficit in finance and to broaden financial debates, it is imperative to understand how financial science became a historical possibility and how financial decision-making became depoliticized in the first place” (DeGoede, 2001, p. 151 in Bieler and Morton, 2008, p. 111). These phenomena are traced through the historical constitution of financial speculation as a technical issue, which depoliticizes the circumstances surrounding notions of risk. Once discourses of financial speculation as well as the discursive constitution of modern finance are broadly deconstructed, the process of thinking about possible alternatives can then begin (DeGoede, 2005, xxvi; DeGoede 2004 in Bieler and Morton, 2008, p. 112).  In summary, socio-political theories provide analytical leverage on issues that are neglected by alternative theories. They open up spaces to consider non-financial sources of power and authority, reveal the ways in which shareholder power is relational and socially-constructed. Bridging critical approaches here also permits going beyond the question of whether ideas, norms and social structures matter, but to also interrogate power hierarchies within the context of responsible investment that may reinforce the perspectives, needs and interests of some social actors over others, as I elaborate in greater detail below.   2.5 Broadening the scope of analysis on shareholder power  As the literature review has sought to demonstrate, existing theoretical approaches are limited in their ability to move discussions of shareholder power beyond the investor-firm dyad,  54 vary in their genuine engagement with the influences of a broader set of actors (including the state and civil society) and generally do not consider how the exercise of shareholder power via shareholder advocacy may entail particular implications for the issues and actors that are the subject of these efforts, and that these implications may occur at varying geographic scales: local, national and international. In this dissertation, I propose that moving away from materially and instrumentally-rooted conceptions of power towards a more socially-contingent conception is essential for moving towards a more holistic understanding shareholder power that overcomes the aforementioned limitations. A first step towards this more holistic understanding is adopting a framework of power that can transcend the investor-firm dyad. Constructivist scholars working in the domain of global governance have a number of valuable insights to offer here. Alice Ba and Mathew Hoffman’s (2018) book is a much-needed synthesis on how social constructivism makes sense of global governance. Their work outlines the social processes in order to shed light on how different actors come to be authoritative rule makers, how diverse norms and practices come to define global rules and how ideas and identities shape global rule-making across various arrangements of global governance. In light of the growth of influence of international institutions and non-state actors, “international rule-making has become the domain of multiple overlapping actors and regimes, rather than the clear-cut leadership by one (hegemonic) state or multilateral conformity with a small and homogenous set of shared rules backed by enforcement mechanisms” (Haas, 2002, p. 73).  In this respect, Anne-Marie Slaughter offers a socially-grounded definition of global governance as “the formal and informal bundles of rules, roles and relationships that define and regulate the social practices of states and non-state actors in international affairs” (Slaughter, 1998 in Haas, 2002, p. 73).   55 For constructivist scholars therefore, the institutional, discursive and intersubjective procedures by which international governance develop have been paramount (Haas, 1999, 2002; Ruggie, 1998; Held et al. 1999; Price and Reus-Smit, 1998; Finnemore, 1996; Barnett and Finnemore, 2004; Reus-Smit, 2002). For example, Lisa Sundstrom (2005), provides a norms-based explanation for varying successes that foreign development assistance donors have had in strengthening Russian social movements. Her analysis finds that,  where foreign assistance is employed to promote norms that are universally embraced, it is highly likely to lead to a successful NGO movement. In contrast, when foreign assistance pursues norms that are specific to other societal contexts, it will fail to develop an NGO movement, regardless of the amount of funding foreign donors devote (Sundstrom, 2005, p. 419).  According to Sundstrom (2005),  NGOs and foreign donors have succeeded by articulating a universal norm against physical harm in the cases of soldiers' rights and domestic violence, but have failed by voicing specifically Western norms of gender equality and feminism in the case of women's rights” (p. 419).   My project aligns with this broader scholarly community of constructivists through its focus on distinctive processes of institutionalization, socialization, persuasion, discourse and norm diffusion to understand the ways in which global governance develops. A key contribution of note is that I engage with these processes within the previously unexplored empirical terrain of shareholder advocacy on human rights by institutional investors. For instance, in Chapter 3, I offer an explanation for tactical variations adopted by institutional investors in their efforts to diffuse emerging norms on the responsibilities of multi-national corporations to respect human rights and to avoid complicity in human rights violations through due diligence measures. In Chapter 4, I examine how the various forms of authority at the disposal of transnational investor-driven governance networks can involve complex institutional arrangements that can shape international advocacy agendas. Likewise, in Chapter 5, I employ critical discourse analysis to  56 study how investor discourses in Canada are shaped by developments in transnational human rights governance, as well as to show how these discourses themselves are emergent, socially constructed, and serve as vehicles through which certain power hierarchies may be maintained and even reinforced. As noted earlier, the unifying force across the three papers that comprise the dissertation is the concept of shareholder power. Although constructivist scholarship on power is quite dense, my socially-grounded analysis of shareholder power is primarily shaped by Michael Barnett and Raymond Duvall’s (2005) introductory chapter in their edited volume Power in Global Governance. These scholars characterise power as the production, in and through social relations, of effects that shape the capacities of actors to determine their circumstances and fate.24 This definition entails two dimensions: 1) the kinds of social relations through which power works in relations of interaction or in social relations of constitution; 2) the specificity of social relations through which effects on actors’ capacities are produced. These two dimensions generate a four-fold taxonomy of power: compulsory, institutional, structural, and productive (Barnett and Duvall, 2005, p. 48) that is especially useful to consider how shareholder power manifests at the global level. The taxonomy is elaborated in greater detail below.25                                                  24 Barnett and Duvall (2005) criticise existing international relations literature on power for its attachment to realism, namely how one state uses its material resources to compel another state to do something it does not want to do. However, they also note that in steering away from realism, both neoliberals and constructivists have neglected to develop how power is conceptualised within their own frameworks. 25 An alternative to Barnett and Duvall’s (2005) framework is available in Doris Fuch’s (2005, 2007 and 2009 with Jennifer Clapp) typology of corporate power which has three forms: instrumental, structural and discursive. I thank Peter Dauvergne and Jane Lister for drawing my attention to this framework via their 2011 book Timber.  Because I was especially interested in how institutions mediate shareholder power as a form of private authority, Barnett and Duvall’s (2005) inclusion of this factor, in addition to those in Fuch’s (2005) framework, offered greater analytical leverage for my project. In addition, Barnett and Duvall’s framework is less contingent on having corporations as the primary unit of analysis. Future iterations of my this analysis may offer more granular comparison of the two frameworks.  57 According to Barnett and Duvall (2006), compulsory power can entail both material and normative resources. This is an important departure from the realist conception of power, which focuses only on the former, and one that matters in bridging concepts of power and authority. For example, transnational activists who may not be on a level-footing with states or firms when it comes to financial clout, can nonetheless use naming and shaming tactics to influence these actors to comply with the values and norms they advance (see for example, Hendrix and Wong, 2012; Murdie and Davis, 2012; Schmitz, 2013; and Hafner-Burton, 2008 for a counter-viewpoint). However, naming and shaming is effective only to the extent that these activists have legitimacy—hence moral authority (Price, 1998). To illustrate this link between power and authority, Barnett and Duvall (2005) discuss how international organisations use their expert, moral, delegated, and rational-legal authority as a resource to compel (influence or exercise power over) state and non-state actors to alter their behaviour (p. 50). Whereas compulsory power entails the direct control of one actor over another, institutional power is the ability to control others in indirect ways. Formal and informal institutions allow one party to work through rules and procedures that define those institutions, and guide and constrain the actions and conditions of socially distant others. This description is particularly apt for exploring shareholder power in the domain of human rights because institutional investors are usually geographically disconnected from the sites of alleged human rights concerns. Hence, what I find particularly compelling about Barnett and Duvall’s (2005) discussion on institutional power is that they highlight both spatial and temporal dimensions within particular mechanisms of influence.  Spatially, one actor can influence others through institutional arrangements, such as decisional rules, lines of responsibility, divisions of labour, and structures of dispersed  58 dependence, hence power works through socially extended, institutionally diffuse relations (Barnett and Duvall, 2005, p. 51). Mirroring the point made earlier in reference to Archer’s morphogenetic sequence, temporally, institutions can have ongoing and unintended effects at a later point (Barnett and Duvall, 2005, p. 51). Institutions themselves are the result of strategic interactions in different domains, generating shared beliefs that, in turn, impact those interactions in a self-sustaining manner (Aoki, 2001). Hence, the value of applying Barnett and Duvall’s framework to analyze shareholder power is that it also permits researchers to specify how the role of each stakeholder toward the firm is shaped by different institutional domains and thereby generates different types of conflicts and coalitions in shareholder activism (a related point is made by Aguilera and Jackson, 2003). This is a key step in allowing analysis of shareholder activism to also migrate across different geographic scales from the local to the global. Exploring the institutional dimensions of shareholder power therefore offers the opportunity to expose how governance biases embedded within the institutions of responsible investment can create winners and losers at different levels of governance to the extent that the ability to use these institutions and benefit from collective rewards are unevenly distributed or serve to reinforce unjust power hierarchies (Barnett and Duvall, 2005, p. 58). However, institutional power also appears moot if the institution is not deemed legitimate at least to some extent. Here, fairness, or the perception thereof, becomes crucial in lending fundamental legitimacy to institutions. Insofar that an institution is deemed legitimate, the institution can then shape the strategic environment of actors by setting the rules to bargaining, affecting policy bias, and mobilising power. In sum, the concept of institutional power enables researchers to unpack the idea that institutions shape the social and political processes of how stakeholders’ interests  59 are defined (“socially constructed”), aggregated, and represented with respect to each other (Aguilera and Jackson, 2003). Within Barnett and Duvall’s (2005) framework, structural power concerns the constitution of subjects’ capacities in direct structural relation to each another. The focus is on social relations of constitution and an actors’ self-understanding and subjective interests. To the degree that actors do not recognise their own domination, they serve to reproduce, rather than resist the differential capacities and privileges of structure (Barnett and Duvall, 2005, p. 53). Historical materialists view the structure of global capitalism as substantially determining the capacities and resources of actors, shaping their ideology so that it serves the interests of capitalists. The arena of responsible investment advocacy offers a fascinating arena within which to test the proposition that the actions of those in subordinate positions are disposed toward reproduction, rather than transformation of the structural relations of domination (Barnett and Duvall, 2005, p. 54).  The final form of power discussed in Barnett and Duvall’s (2005) scheme is productive power, which involves the constitution of social subjects with varying social powers through systems of knowledge and discursive practices. From this vantage point, the discourses and institutions of transnational investor-led advocacy on human rights prioritize certain types of actors with associated specific social powers (in this case over firms), self-understandings and performative practices. Whereas structural power envisions binary and hierarchical relations, productive power concerns the boundaries of all social identity, and the capacity and inclination for action for the socially advantaged and disadvantaged alike, and the myriad social subjects that are not constituted in binary hierarchical relationships (Barnett and Duvall, 2005, p. 56). It is this feature of Barnett and Duvall’s (2005) framework that allows for my analysis of shareholder  60 power to transcend the investor-firm dyad and consider a much broader set of social relations. To illustrate the concept of productive power, the authors note that categories such as “civilised,” “Western,” “rogue,” “democratic” represent productive power, in that they generate asymmetries of social capacities (Barnett and Duvall, 2005, p. 56). Here, productive power seeks to establish legitimacy, which in turn becomes intrinsic in the discourse itself, as diffuse and contingent social processes produce certain subjects, fix meanings, and create a structure of intersubjective public beliefs. An example is the Global Compact, whose goal is to engage “the private sector to work with the UN, in partnership with international labour and nongovernmental organisations (NGOs), to identity, disseminate, and promote good corporate practices based on nine universal principles” (Barnett and Duvall, 2005, p. 60). Architects of the agenda envision two mechanisms to realise this development: 1) as MNCs sign onto a set of evolving standards, a discursive space is formed in which actors are produced as subjects empowered legitimately to comment on their performance; and 2) the creation of the socially responsible corporation with a new self-understanding that embraces best practices (Barnett and Duvall, 2005, p. 61). Power and authority in this case are difficult to distinguish, as legitimacy and authority become inherent in the continuing production and maintenance of identity and understanding. As a whole, Barnett and Duvall’s (2005) framework of power complements the under-socialized view of shareholder power that has persisted in the literature on shareholder activism thus far. It does so by capturing both material and normative dimensions of power, but also expanding this focus to include the embeddedness of investors in a nexus of formal and informal rules (North, 1990), their role as co-creators of social systems and as agents who are actively shaping and legitimizing a particular discourse around the social responsibilities of firms through their shareholder advocacy. Applying Barnett and Duvall’s (2005) framework to existing studies  61 of shareholder activism also sheds light on some gaps that need to be filled. For the most part existing studies have focused almost exhaustively on compulsory forms of power (material and normative), but less so on institutional, structural and productive forms. Each of these aspects is investigated in inter-related ways in the papers that follow.  Throughout this chapter, I have referenced authority in relation to shareholder power on a number of occasions, and this relationship merits some elaboration. In this dissertation, I address shareholder power as it is exercised in the context of investor advocacy aimed at enhancing corporate compliance with global human rights norms. As has been discussed thus far, shareholder power can only be effective in influencing firms, if the authority of investors (as private actors) to leverage this form of power is viewed as legitimate.  Hence, this project’s focus on relational and socially-embedded aspects of shareholder power also helps us to probe the nature of private authority in global governance. As Hall (2005) suggests, “authority is a social commodity that cannot be usurped or successfully claimed unilaterally. It must be publicly claimed and publicly acknowledged by the subjects of the exercise of power” (¶4).  Scholarship on the role of international organisations (IOs) have shown that particular forms of authority serve as mechanisms by which institutions can shape state identities and interests. For example, IOs influence social processes through authority derived from their status as technical experts and as beneficiaries of delegated state authority in specific issue areas (Barnett and Finnemore, 2004, p. 20).  Hall’s work with Thomas Biersteker (2002) observes that the globalisation of economic liberalism has led the state to yield previously sovereign domains to private and non-state actors. This in turn has given way to a rise in private authority that is challenging public authority in a number of ways. In the sphere of the market, private authority stems from two sources. The first  62 regards the capacity to establish manufacturing, productive, regulatory and reporting standards that become recognized by others and are subsequently adhered to, otherwise known as “institutional market authority” (p. 214). The second form regards the increasing acceptance, particularly in advanced industrialized economies, of market-based decision-making over politically-based decision making, often on efficiency and normative grounds, and generally referred to as “normative market authority” (p. 214).  Other scholars such as Claire Cutler, Tony Porter and Virginia Haufler (1999) refer to this concept as “private international regimes” (p. 21) which can include a complex array of actors, industry norms and practices, coordination service firms, production alliances and business associations. In her later work, Cutler (2003) argues that the system of global governance is undergoing a fundamental transformation with the creation of a new private transnational legal order that has historical antecedents in prior international systems (the law merchant). She suggests three trends are driving this transformation, namely an increase in the juridification of international social life, an increase in heterogeneity among forms of governance (pluralization), and an increase in the significance of governance in the private sphere (privatization) (pp. 17-32). As a consequence of these processes, Cutler (2003) argues that we are witnessing three transformations that also shed some light on why this study’s empirical focus on institutional investors is warranted. First, the welfare state is being replaced by the “competition state” as the globalization of liberalism induces states to adjust to a neo-liberal global environment (p. 29). Second, the world is moving from national to transnational patterns of capital accumulation characterized by a third transformation (pp. 29-30). This third transformation, in turn, is the emerging dominance of “flexible accumulation” constituted by a combination of capital mobility and a high degree of flexibility in transnational production arrangements (pp. 30-31). Modern  63 commercial law (the law-merchant) abets these transformations, according to Cutler, by generating publicly binding sources of private international commercial law through networks of private actors (pp. 39-41). Bringing together this wider scholarship on private power and authority in global governance, my study examines the ways in which institutional investors draw on both forms of private market authority to persuade firms to improve their performance when it comes to aligning with global human rights norms and standards. I do not assume that their identity as market actors necessarily means that institutional investors draw on market authority alone. For instance, Chapter 4 examines the ways in which collaboration with INGOs through investor-driven governance networks may offer investors the opportunity to draw on the moral authority of their civil society counterparts. Similarly, Chapter 5 sheds light on how investors may (deliberately or inadvertently) coopt moral authority by citing INGO reports on human rights violations without having to give up any institutional space or power to these actors. In this chapter, I have reviewed the key theoretical schools that frame existing analyses on shareholder activism in order to identify the merits and gaps of each approach in theorising shareholder power. The result of this review suggests that sociologically-grounded theories and critical perspectives can offer productive avenues through which to interrogate shareholder power, as existing accounts tend to overlook power relations between investors and other social actors beyond the firm. Previous studies have also avoided examining shareholder power a dependent variable, thereby limiting our understandings of how investors exercising this form of private authority may do so in ways that affect a broad range of stakeholders, as well as the governance of human rights at different geographic scales ranging from the local to the global. The three chapters that follow are a modest attempt to address these gaps.   64 Chapter 3: Explaining Tactical Variations in Investor Responses to Human Rights Issues Over the past two decades, political economy scholars have convincingly argued that conflict is understood not only in terms of the political fractures, demands and interests of warring groups, but also through the resources at their disposal to wage and sustain violence (Banfield et al., 2005; LeBillon, 2001). As suppliers of political, material and financial resources, the role of transnational corporations in fueling conflict dynamics and related human rights violations has been investigated at length (for example: Cilliers and Dietrich, 2000; Banfield et al., 2005; Renner, 2002; Avant, 2005; Watts, 2005; LeBillon, 2001 and 2012; Global Witness, 2010a; 2010b; Goodhand, 2002). Cumulatively, these studies have documented a deeply concerning accountability deficit for corporate complicity in human rights abuses (most notably by Ruggie, 2007). Responses to this intractable challenge have predominantly focused on regulatory (Koenig-Archibugi, 2004; Vogel, 2010), judicial (Ratner, 2001; Deva, 2004; Weissbrodt and Kruger, 2003), non-judicial mechanisms (Lukas et al., 2016) and voluntary approaches (Mares, 2010, 2011; Vogel, 2008). However, the role of institutional investors26 as a mechanism for advancing corporate compliance with human rights has been largely ignored by international relations scholars. As elaborated in Chapter 1, this neglect is surprising given the many ways in which institutional investors are already engaging with the global politics of human rights. For example, in 2012, thirty-nine institutional investors and investment management firms asked                                                 26 Defined in Chapter 1.  65 member states of the United Nations to adopt a strong, legally binding and comprehensive arms trade treaty (Bâtirente, 2012). Together, this group represented USD$3 trillion in assets under management. These investors suggested that without a stronger system to control international trade in conventional arms, companies operating in the defense industry face regulatory and reputational risks, as they may be perceived to be complicit in human rights violations committed by irresponsible end-users (Batirente, 2012). Similar global investor advocacy efforts have been evident across a whole range of issues such as human trafficking (ICCR, 2011) and labour rights in global supply chains (PRI, 2017a) to the trade in minerals from conflict zones (Compere, 2010), and indigenous rights in mining and agro-forestry dependent communities (PRI, 2017b). Broadly speaking, when faced with information that companies they have invested in are committing or complicit in human rights violations, investors react in one of three ways. First, they may remain passive, choosing to remain silent and continue to invest in these firms. Alternatively, they may engage in some form of activism to express their preferences around human rights compliance by investee companies through positive engagement such as letter-writing, dialogue, media stories and participation in multi-stakeholder or industry forums seeking to address a particular issue. As a third option, investors may choose to “cut-and-run” by divesting their shares from the company. However, investor responses to human rights issues may not follow predictable patterns.   66 For instance, the widely-condemned human rights violations associated with the trade in conflict minerals from the Democratic Republic of Congo27 prompted investors to engage in public-policy advocacy, dialogue with technology firms whose supply chains included conflict minerals, and even divestment from firms likely to source conflict minerals (Khushrushahi, 2012), whereas the trade in other conflict-commodities such as jade in Myanmar or gold in Colombia did not trigger comparable investor responses. Likewise, in response to the deadly collapse of the Rana Plaza garment factory in Bangladesh in 2013, the Bangladesh Accord – a legally-binding agreement between retail firms and trade unions that was designed improve labour rights in the Bangladeshi Ready-Made Garment Industry was roundly endorsed by investors (SHARE, 2013; ICCR, 2016; PRI, 2014), while no comparable investor activism is evident in light of almost-identical labour rights violations documented in other garment-producing countries like Turkey in a similar time-frame (FairWear Foundation, 2017; Indonesia (Clean Clothes Campaign, 2014) and Cambodia (Human Rights Watch, 2015).  How do we explain these diverse advocacy responses to strikingly similar human rights concerns? Given that institutional investors have not been systematically taken up as a subject of analysis within international relations scholarship on transnational advocacy, it is necessary to pivot to other disciplines in search of possible answers. Although explicit attention to tactical approaches within investor advocacy is fairly limited, it is possible to distill a number of implicit                                                 27 Global Witness (2010a) defines conflict minerals as “natural resources whose systematic exploitation and trade in a context of conflict contribute to, benefit from or result in the commission of serious violations of human rights, violations of international humanitarian law or violations amounting to crimes under international law.” (p. 8) The term typically refers to cassiterite, coltan, gold ore and wolframite (for tungsten), diamonds, timber and petroleum. Conflict minerals was first addressed by the UN General Assembly in 2001 in relation to diamond mining in Liberia, the scale of conflict-mineral extraction and related human rights abuses in the DRC have made it a notorious and widely-discussed case.   67 hypotheses from existing work on shareholder activism by corporate governance and business ethics scholars. By engaging with this literature, I argue that existing explanations for tactical variations in investor advocacy are insufficient, and a more holistic approach to unpacking mechanisms by which investors exercise power in the domain of global human rights is warranted. Undertaking this analysis may provide insights on how to assess competing appraisals about the impacts of the responsible investment at different levels of governance related to human rights, as well as the degree to which different stakeholders occupy or are absent from these governance spaces.  To develop this argument, the chapter is structured as follows: I first provide a theoretical framework for understanding shareholder activism using Hirschman’s (1970) work on exit and voice in Section 3.1. Next, I lay out the empirical context that motivates the chapter’s core research question: When faced with a portfolio company that is a systematic violator of human rights norms, why would an institutional investor with strong human rights commitments continue engaging the company rather than divesting from it?  This question arises from examining the diverging responses by public pension funds in Norway and the Netherlands to allegations of complicity in human rights violations by their investee firm PetroChina (a state-owned Chinese oil and gas company) in Burma and Sudan (Section 3.2).  The data and methods for analyzing these investor responses are explained in Section 3.3, followed by a systematic analysis of observed outcomes in both cases against hypotheses articulated in the theoretical framework. As evidenced through the findings elaborated in Section 3.4, engagement with PetroChina by Norwegian fund and divestment from the firm by the Dutch fund cannot be adequately explained by existing theoretical accounts  of why voice and exit options are used in shareholder activism because the factors commonly identified as correlates of shareholder  68 engagement (voice) in the former case, are also present for divestment (exit). I conclude in Section 3.5 by noting that the case challenges the explanatory power of existing accounts for why investors may prefer engagement over divestment, and highlights the need for additional hypotheses in explaining variations in investor responses to human rights violations by firms. Such a framework would advance not only the literature on shareholder activism specifically, but also theories of transnational advocacy more broadly by bringing in private actors from the periphery of this literature into the main frame. Finally, because the concept of shareholder power serves as the thesis’ connective tissue, the chapter closes by reflecting on the implications of the case analysis regarding the question of who leverages power in global human rights politics, and the mechanisms by which they do so within and outside the state.  3.1 Theoretical Framework   Exit, voice and loyalty Hirschman’s (1970) theory of the firm presented in Exit, Voice and Loyalty is both simple and far-reaching in its explanatory power. The framework distinguishes between likely responses of stakeholders28 their dissatisfaction with particular institutions – be they social, political or economic. In Hirschman’s view, stakeholders signal their concerns through exit (such as consumers ceasing to buy a firm’s product or service) or voice (via complaints to management or a regulatory body). Both options could serve as a trigger for an organization’s leadership to chart a course to recovery from decline. Unlike exit, voice involves any attempt to change, rather than                                                 28 Note: Hirschman’s terminology mainly refers to employees, consumers or customers of a firm and citizens or members of a political or social organization. I use the term stakeholder to capture a broader set of relationships than those referenced by Hirschman.   69 escape from an objectionable state of affairs (p. 30). Hirschman’s work primarily sought to address the following questions: “Under what conditions will the exit option prevail over the voice option and vice versa? What is the comparative efficiency of the two options as mechanisms of recuperation? In what situations do both options come into play jointly?” (Hirschman, 1970, p. 5).  Within Hirschman’s (1970) framework, for either exit or voice to function as a trigger for recovery within an organization, there are two important pre-conditions: elasticity of demand and differentiation among stakeholders. First, demand for the product or service provided by the organization is assumed to be elastic in response to changes in quality, such that any decline in demand results in revenue losses (p. 23).29 The second assumption requires a mixture of “alert” and “inert” stakeholders (pp. 24-32). The alert stakeholders provide the organization with a feedback mechanism about decline, whereas the inert stakeholders maintain the flow of capital, legitimacy or time needed by the organization’s leadership to make necessary changes needed to address the decline in quality (pp. 24-32). Furthermore, the effectiveness of exit and voice are subject to threshold effects. For instance, mass exit may accelerate institutional decline to such a degree that recovery becomes impossible. Likewise, if the voice option is too distracting or does not leave enough time for recovery, then it can trigger negative returns (p. 33).   In considering the interplay of both responses, Hirschman (1970) explains that voice can be viewed as subordinate, alternate or complementary to exit. The subordinate view is that voice is exercised only when the exit option is unavailable (such as in monopolistic conditions).                                                 29 As part of this assumption, price and costs are assumed to remain constant, and the decline in quality is not initiated as a deliberate strategy on the part of the firm. These conditions are expected under perfect competition, which also assumes perfect knowledge on the part of consumers.   70 However, voice can also act as an alternative to exit. For example, if stakeholders are sufficiently convinced that voice will be effective, then they may well postpone exit (Hirschman, 1970, p. 37). Hence, exit can also be viewed as dependent on the ability and willingness of stakeholders to take up the voice option. If organizational decline plays out over a length of time, Hirschman’s framework assumes that,  the voice option is more likely to be taken at an early stage. Once you have exited, you have lost the opportunity to use voice, and vice versa; in some situations it will therefore be a reaction of last resort (emphasis original) after voice has failed (Hirschman, 1970, p. 37).   In Hirschman’s (1970) view, the decision to stick with a deteriorating firm or organization – that is preferring voice to exit – is based on a) the willingness to take the chances of the voice option as against the certainty of the exit option, and on (b) the probability with which a stakeholder expects improvements to occur as a result of actions to be taken individually, with others, or just by others (p. 39). Therefore, in comparison to exit, voice is costly and conditioned on the influence and bargaining power members can bring to bear on the organization (p. 40). Because of these characteristics, voice becomes more expensive in direct proportion to the number of organizations of relevance to a particular stakeholder. Moreover, voice is more likely to function as an important mechanism in markets with few buyers, or where a few buyers account for an important proportion of total sales, both because it is easier for a few buyers to combine for collective action, and because each one may have much at stake and wield considerable power, even in isolation (p. 41).30 For these reasons, Hirschman expects that voice is more salient for members of an organization rather than individual customers of firms, whereas the wide availability of substitutes in advanced economies favours exit over voice.                                                  30 Here, Hirschman relies on Olson’s (1965) Logic of Collective Action.   71 Nonetheless, Hirschman concedes that,  the domain in which the voice option is likely to be deployed, especially as a substitute for exit, the territory left to it remains both considerable and somewhat ill-defined. Moreover, once voice is recognized as a mechanism with considerable usefulness for maintaining performance, institutions can be designed in such a way that the cost of individual and collective action would be decreased. Or, in some situations, the rewards for successful action might be increased for those who initiated it (Hirschman, 1970, p. 42).  Thus,  he argues that while the exit option involves a clear-cut either-or decision, the voice option “is essentially an art constantly evolving in new directions” (Hirschman, 1970, p. 42). This situation makes for an important bias in favour of exit when both options are present, where members of an organization are: likely to base their decision on past experience with cost and effectiveness of voice, even though the possible discovery of lower cost and greater effectiveness is the very essence of voice. The presence of the exit alternative can therefore tend to atrophy the development of the art of voice. (Hirschman, 1970, p. 43, emphasis original).   Within Hirschman’s (1970)  model, loyalty is a key concept in the battle between exit and voice (p. 82). Loyalty applies to groups or organizations where exit is unlikely or impossible, such as the family, tribe, church, and the state, making voice the principal way for an individual member to register their dissatisfaction (Hirschman, 1970, pp. 76-78). Loyalty is especially powerful because its very existence is predicated on the possibility of exit. As such, the chances for voice to function effectively as a mechanism for recovery from deterioration is strengthened “if voice is backed up by the threat of exit, whether it is made openly or whether the possibility of exit is merely well understood to be an element of the situation by all concerned” (Hirschman, 1970, p. 82).  Throughout Exit, Voice and Loyalty Hirschman draws on examples from particular market conditions and political structures. Writing at a time when economic thought had gained  72 wide purchase in political science, Hirschman’s work was deliberately inter-disciplinary. He writes:  Exit and voice, that is market and non-market forces, that is economic and political mechanisms have been introduced as two principal actors of strictly equal rank and importance. In developing my play on that basis, I hope to demonstrate to political scientists the usefulness of economic concepts and to economists the usefulness of political concepts (Hirschman, 1970, p. 19, emphasis original).  Though Hirschman’s eloquence and theoretical reach are difficult to replicate, this thesis project is largely inspired by his candid efforts to bridge economic and political studies for mutual benefit. This chapter in particular seeks to make a modest contribution towards that effort.     Exit and voice as a framework for shareholder activism As discussed in Chapter 2, Hirschman’s (1970) rubric of “Exit and Voice” dominates the corporate governance literature as an explanatory framework for how and why institutional investors exercise shareholder power over the firms they own31 (Becht et al., 2009; Berle and Means, 1932; Fama and Jensen, 1983; Gifford, 2010; Goodman et al., 2014). From this lens, institutional investors may choose to champion firm compliance with human rights standards through a punitive approach (exit), where deviant companies are deprived of capital through exclusion from investment portfolios (Divestment). Alternatively, these investors may opt to voice support for human rights compliance through a variety of engagement mechanisms such as letter-writing, active dialogue, media campaigns, proxy voting and the filing of shareholder proposals that raise human rights concerns (Engagement) (March and Olsen, 2008; Sjöström 2008, p. 142).                                                  31 Within corporate governance, shareholder activism is also referred to as ‘monitoring’ of firm behaviour.   73 Before moving onto reviewing when exit or voice are more likely within shareholder activism, it is first necessary to address how the two key preconditions for Hirschman’s (1970) framework are met in this context. Recall that the first precondition is a variation on the typical demand function (i.e. declining demand signalled through voice or exit harms a firm’s revenue). This means that the issue being taken up through shareholder activism needs to have a tangible impact on the firm’s revenue in order to effectively function as a signal of decline. At minimum, this impact is generated because firms allocate resources away from operational tasks for addressing investors’ concerns through various mechanisms (quarterly investor calls, investor relations teams, coordination between sustainability teams, operations staff and management etc.). Aside from these costs, in practice, most investor activism on ESG is issues are predicated on a clear rationale for how the particular issue at stake affects firm performance, and by extension detracts from shareholder value (Khan et al., 2016).32 The second pre-condition that would allow Hirschman’s model to apply in the context of shareholder activism is a mix of active and inert shareholders. Across most advanced economies, institutional investors typically own shares in a very large number of companies. This poses a practical difficulty in monitoring all firms at the same time. As a result, at any given time, some investors will be more active in certain firms over others, creating a mix of active and passive investors (Black, 1990; Huppé and Bala-Miller, 2014).                                                  32 “Shareholder value is the value delivered to shareholders because of a firm’s ability to grow sales, earnings and free cash flow over time. A company’s shareholder value depends on strategic decisions made by senior management, including the ability to generate a healthy return on invested capital. If this value is created over the long term, the share price increases and the company can pay larger cash dividends to shareholders” (Investopedia, 2018b, ¶ 1).   74 To articulate likely conditions for exit and voice in the context of investor advocacy, I now turn to a brief overview of the literature on shareholder activism – a growing topic of study for scholars interested in corporate governance and business ethics. Over the last decade, this set of literature has specifically focused on understanding drivers and motivations of shareholder activism on environmental, social and governance issues (ESG). Previously these issues usually externalised as drivers of firm performance and profit maximization. However as firms seek to gain competitive advantage through sustainability initiatives (Dauvergne, and Lister, 2013), these topics are increasingly being taken up by activist shareholders (Khan et al., 2016). Although studies that specifically address tactical variations between engagement and divestment are very limited, it is possible to infer a set of probabilistic hypotheses that may explain why institutional investors may prefer exit over voice (and vice versa).   Likely conditions for divestment Within Hirschman’s model, exit is usually chosen when institutionalized mechanisms for voice do not exist or because it is typically the cheaper option. Within the context of shareholder activism, the literature on comparative corporate governance provides a strong indication that certain structural factors and institutional designs may alleviate the burden of costs associated with corporate engagement, thereby making exit cost-effective in some markets, but not others. For example (as discussed in Chapter 2), the Anglo-American model of corporate governance is often contrasted with the Continental European corporate governance model (Becht & Röell, 1999; La Porta et al., 1998). Wen (2009) argues that the fixation on shareholder value and the highly diffused ownership structure of corporations in the Anglo-American corporate governance model minimizes incentives for investors to actively engage corporations, whereas the social- 75 dialogue model of corporate governance prevalent in Continental Europe and Scandinavia forces firms to be responsive to issues raised by social partners involved in the firm’s governance. From this literature, it is possible to infer the first hypothesis to explain tactical variations in shareholder activism: H1 [Corporate governance model]: Institutionalised social dialogue structures lower the costs of engagement, making divestment a less likely option for investors operating in European/Scandinavian corporate governance models.  Likely conditions for engagement Hirschman’s model also suggests that voice is more likely when stakeholders are confident of their ability to influence the firm, and determine the voice option is more likely to yield the desired results. In the context of shareholder activism, results are mainly assessed to the degree that a firm’s response ensures it preserves shareholder value by maximising returns and minimising risks to a firm’s profitability.  The primary argument in favour of corporate engagement by institutional investors is that maintaining an equity stake in a firm that is under-performing in relation to human rights compliance enables investors to act as a source of pressure on the company to improve its practices, whereas divestment amounts to a “cut and run” strategy that does not address or mitigate alleged human rights violations. Here, investors appear to align with Hirschman’s assertion that “voice is necessary because an exclusive reliance on exit precludes an organization from learning from its most articulate critics” (Hirschman, 1970, p. 4). Also in line with Hirschman (1970), it has been widely assumed that voice and exit are alternative options, and that a failed voice strategy will lead to exit (Withey and Cooper, 1989). Interview data with institutional investors in the United Kingdom, United States and Netherlands, that were asked to  76 explain their preferred mechanism for addressing human rights abuses such as forced labour in corporate supply chains, exemplify this general preference for engagement over divestment (Bala-Miller, 2010): Direct company engagement has to be productive. Companies may not want to talk with shareholders about these issues and it can be very difficult to get the right company representatives to have the discussion with. Sometimes, going straight to top executives and company boards is the only way to get directly to the heart of an issue and take concrete steps. Filing shareholder proposals can put additional pressure for this dialogue to take place. However, a proposal should be a starting point, and not an end unto itself. For Teamsters [General Fund], divestment is a measure of last resort. Most long-term shareholders should try to push for reform rather than take a “cut and run” approach because maintaining shares in a company gives voice to labour rights issues. (p. 8)  At All Pensions Group Netherlands, “We feel that shareholder proposals and engagement can be quite effective because they generate public pressure on a company to improve its performance. On the other hand, divestment is difficult and can be counter-productive. We base our strategy on a realistic assessment of what we can change through playing our role as a large investor with some influence. For this reason, our research on sustainability or corporate governance provides the basis for our decisions on whether or not to make an investment, to sell an investment, or to reduce or increase the size of an investment” (p. 12).   Hermes Equity Ownership Services in the UK “sought to build an investor coalition through the UNPRI with significant assets to obtain adequate leverage for approaching the companies. We succeeded in constructing a coalition of nine separate asset managers. For Hermes, divestment was not an option because we wanted to achieve positive change in the management of labour rights in the Brazilian pig iron supply chain, and use our ownership of company shares to act as a lever.” (p. 15).  The efficacy of engagement by institutional investors in achieving their demands is also shown in other thematic areas. For example, firms targeted by shareholder resolutions are more likely to disclose their greenhouse gas emissions (Reid & Toffel, 2009). In this context, it is likely that investors with previous experience of successfully achieving their demands through engagement strategies are more likely to replicate this approach than opt for divestment when confronted with similar issues and firms.   77 Ferraro and Beunza (2013) argue that because shareholder rights provide an institutionalized relationship between firms and investors, the voice option offers a much better avenue for communication and corporate learning than extra-institutional contentious tactics that may be deployed by stakeholders such as INGOs. They note,  [a]s a channel of activist voice, shareholder engagement avoids the limitations of contentious tactics by building on the institutionalized relationship between the equity investors and the publicly-listed company: it is as part-owners, not as concerned citizens, that engaged shareholders address managers (Ferraro and Beunza, 2013, p. 4).  However, it should be noted that this assumption is contested in the literature. For example, Eesley and Lenox (2006) found that more confrontational tactics such as boycotts, protests and lawsuits, typically favored by external stakeholders were more effective in achieving changes to a firm’s policies and practices, than letter-writing campaigns and shareholder resolutions - the tools-of-the-trade of shareholder activists. Vasi and King (2012), on the other hand, found that shareholder activism through shareholder resolutions had a stronger effect on environmental risk than more confrontational tactics. They suggest that the different results might stem from the fact that these two tactics target different audiences. Shareholder engagement is closely followed by risk analysts, leading to a greater impact on their perceptions of corporate risk, while protests and boycotts are more public, and thus might represent a more imminent threat to corporate reputation. As a result, the latter is more effect in triggering changes to corporate policies and practices for firms that are sensitive to consumer pressure. Another crucial point to note is that these studies did not include attention to shareholder dialogue which is a preferred engagement strategy by investors, and hence they do not provide a complete account of whether external tactics are indeed more effective than shareholder engagement (Ferraro and Beunza, 2013). While the literature on tracing the actual impacts of engagement on corporate policies suffers from an attribution problem, especially when multiple actors are simultaneously putting  78 pressure on the firm through various means at their disposal, from an investor perspective, there does appear to be a bias in favor of engagement over divestment due to the perceived efficacy in articulating demands to the firm’s management (Hebb, Hachigian and Allen, 2011; Gond and Piani, 2013). One of the reasons for this bias is that shareholders enjoy a reasonably higher degree of salience with a firm’s management compared to other stakeholders. Empirical studies on this theme are largely inspired by the stakeholder saliency model developed by Mitchell et al. (1997), and generally investigate how managers identify their primary stakeholders and how stakeholders gain the attention of managers (Gifford, 2010; Hebb, Hachigian and Allen, 2011; Goodman and Arenas, 2015). Taken together, the literature suggests that the expectation by investors that engagement is likely to yield a response from firms in favour of shareholder demands, generates the second hypothesis that could explain tactical variations in approaches to shareholder activism: H2 [Firm responsiveness]: Low firm responsiveness to shareholder engagement is more likely to result in divestment Aside from a belief in the effectiveness of engagement as a change-inducing strategy, investors may also prefer engagement because of emerging evidence linking this strategy with better financial returns and shareholder value. This link is most famously associated with the ‘CalPERS effect’ in the United States or the ‘Hermes effect’ in the United Kingdom. Brad Barber (2007) found that the California Public Employees Retirement Scheme’s (CalPERS) pursuit of reforms aimed at increasing shareholder rights in a set of ‘focus-list’ firms resulted in over USD$3 billion in wealth-creation.33 Becht et al. (2008) conducted an analysis of                                                 33 CalPERS selects a small number of companies as part of the Focus List Program. These companies are engaged on sustainability issues. In September 2014, Wilshire (a consultant to the CalPERS Board) published an analysis that measured the stock performance of the 188 companies targeted by the Focus List Program from 1999 through the fall of 2013. The report showed engaged companies performed better (CalPERS, 2018, ¶1-4).   79 shareholder activism based on an analysis of the private engagements by the U.K’s BT Pension Scheme and found that the fund substantially outperforms benchmarks and these abnormally high returns are largely associated with the achievement of engagement objectives rather than stock picking (Hermes effect). Similar findings are reported by Buchanan et al., (2009), Carleton et al., (1998) and Gillan and Stark (1998 and 2007).  Barnett and Salomon (2003) found that institutional investors which rigorously screen investments based on social factors “may effectively weed out bad firms from their portfolios, thereby improving financial performance” (p. 384). Similarly, environmental shareholder resolutions have been shown to positively affect corporate environmental performance (Lee & Lounsbury, 2011). In a review of thirty-six academic studies on the relationship between environmental, social and governance (ESG) factors and financial performance, twenty studies showed evidence of a positive relationship, while only three show evidence of a negative relationship (Mercer, 2009). Dimson et al. (2015) examined highly intensive ESG corporate engagement activities at U.S. public companies between 1999 and 2009. The findings revealed that successful engagements generate cumulative abnormal returns of +7.1%. Aggarwal et al. (2011) analyzed the portfolio holdings of institutional investors across 23 countries over the period 2003–2008. They concluded that the greater presence of institutional ownership (and their related engagement activities) translates into higher likelihood of dismissing poorly performing CEOs, leading to improvements in valuation over time.    Moderating factors for engagement The size of an institutional investor can also determine its propensity for engagement. Proponents of corporate engagement use the logic of markets to explain that since institutional  80 investors have grown so large in recent decades that they now jointly own the majority of all financial assets worldwide, and have in effect become “universal owners” (Hawley and Williams, 2002, p. 286). As such, increased global diversification within institutional investors’ investment portfolios creates greater exposure to investment risks that requires greater oversight of companies. The considerable size of holdings and use of passive index funds34 mean that it is difficult for investors to exercise control by simply selling firms (exit) that they are dissatisfied with (Hirschman, 1970). The resulting lack of liquidity35 therefore requires active corporate engagement (voice) in firm level decision-making to ensure share value is maintained (Hirschman, 1970). Hence, universal owners have an incentive to integrate any environmental, social or governance criteria which affects the world economy into their investment processes instead of just considering those criteria that individual corporations cannot externalize (Hawley and Williams, 2000, 2002a, 2002b, 2007; Thamotheram and Wildsmith, 2007).  From this view, corporate engagement on issues like climate change or human rights represents a credible investment strategy that, in some cases, delivers attractive risk adjusted returns (Clark and Hebb, 2005; Kiernan, 2007; Sethi, 2005; Monks, 2001; Camejo., 2002; Davis et al., 2006, Ivanova, 2017). Generally, empirical evidence also shows that investors with bigger stakes in their investee companies have higher monitoring incentives as their influence and resulting benefits from activism increase (Shleifer & Vishny, 1986). The links between shareholder engagement with positive financial returns, as well as the moderating influence of                                                 34 A passive index fund is an investment vehicle wherein the investment portfolio is constructed to match or track the components of a market index, such as the Standard & Poor's 500 Index (S&P 500). This investment strategy is said to provide broad market exposure (diversification), low operating expenses and low portfolio turnover. This typically characterizes the approach taken by mutual funds.  35 “Liquidity describes the degree to which an asset or security can be quickly bought or sold in the market without affecting the asset's price” (Investopedia, 2018a, ¶ 1).   81 the size and characteristics of institutional investors as universal owners generates the third hypothesis that could explain tactical variations in shareholder activism. H3 [Financial risk]: The greater the exposure to financial risk for the investor stemming from the issue being raised, the more likely it is for shareholder value to diminish, making divestment more likely.  However, financial measures are not the only indicators of costs incurred in the context of shareholder activism. As alluded to earlier, while a significant body of literature has explained how and why firms respond to reputational threats, a similar emphasis on reputational effects on investor behaviour is lacking. However, there is a small but emerging set of scholarship that examines the motivations of financial actors in adopting responsible investment policies and approaches where reputational issues are addressed. Adherence to these policies typically commits investors to align investment principles with long-term sustainability and to raise related, material environmental and social concerns with investee firms. For example in their study of why investors join the PRI, Majoch, Hoepner and Hebb (2017) evaluate different forms of power as explanatory factors. They note that as a voluntary and aspirational framework the PRI itself does not use the means of threat or coercion to influence an investor’s decision to sign up to the Principles. Coercive power as a means of increasing adoption of the PRI would have to come from other sources such as regulatory bodies or investment fund trustees. A utilitarian, financial incentive is most likely to come from an investor’s clients. The desire to attract new clients or to satisfy the demand from existing ones by adopting the PRI would be typical examples of utilitarian drivers. The reputational benefits and the signaling of capacity to address environmental and social issues by virtue of becoming a PRI signatory are examples of normative power, as non-material, symbolic incentives to adopt the principles. Their study found the strongest evidence for normative power driving decisions to adopt the PRI (Majoch, Hoepner  82 and Hebb, 2017). Although understanding reputational issues for institutional investors is at an early stage, it is possible to infer a further potential explanation for variations in investor advocacy strategies based on this factor: H4 [Reputational risks]: The greater the exposure to reputational risk for the investor stemming from the issue being raised, costs of engagement increase, making divestment more likely.   Likely conditions when engagement is combined with divestment In Hirschman’s model, engagement is likely used in combination with divestment when stakeholders have long-term interests in the organization. Empirical analysis based on the temporal horizons of particular investors and their propensity for engagement as a strategy for shareholder activism has also yielded inconsistent results. On the one hand, it is believed that institutional investors, as long-term shareholders have an incentive to actively monitor and engage the corporation as they are more interested in long-term profitability and are likely to still hold the shares when the corresponding benefits arise (Chen, Harford, & Li, 2007; Neubaum & Zahra, 2006). Similarly, Callen and Fang (2013 in Ivanova, 2017) provide empirical evidence that institutional investors act as monitors and reduce future stock price crash risk, because their dialogue and oversight prevent management from hoarding bad news. On the other hand, short-term, or transient investors such as hedge funds may have greater incentives to intervene to correct firm behaviour, and may do so more often in pursuit of short-term profits at the expense of long-term firm value (Bratton & Wachter, 2010 in Ivanova, 2017).  Studies of when engagement and divestment are used in combination are very rare. Instead, most studies are case specific and focus on specific engagement strategies (dialogue vs. filing of resolutions etc.). For example, Hebb et al.’s (2014) study on shareholder power and  83 salience drew on a large sample of minority shareholder resolutions submitted at U.S. companies. Contrary to the assumption that minority shareholder resolutions are put forward when engagement between shareholders and company managers breaks down, they found that in most cases where shareholder engagement is combined with the filing of a shareholder proposal, the minority resolution preceded the engagement and may well have been required in order to gain attention from the company. These temporal dimensions require further empirical investigation and for the moment can be used to inform analysis of tactical variations in shareholder activism on a case-by-case basis. To summarise, a review of the existing literature on shareholder activism on the mechanics of exit and voice suggests that factors that are most likely correlated with shareholder engagement as a tactical preference by institutional investors when raising concerns with investee firms include a European/Scandinavian corporate governance model and relatively low financial and reputational risks stemming from the issue being raised. These factors are moderated by specific investor characteristics such as their relative size and confidence in their salience, and ability to have their demands met through engagement rather than divestment. However, as the case detailed in the following section suggests, these characteristics may be theoretically insufficient, prompting the need for more a more holistic framework. Recall in Chapter 2, I also pointed out that most analyses that have applied Hirschman’s Framework to the study of shareholder activism do not necessarily consider how power relations may shape tactical decisions to pursue exit or voice. To address this gap in the literature, I am particularly interested to explore whether the three hypotheses investigated in the case study that follows confirms or challenges the need to consider the influence of power relations beyond the investor-firm dyad.  84 3.2 Empirical context The Dutch Pension Fund ABP and the Norwegian Government Pension Fund-Global (GPFG) are among the world’s biggest pension funds by assets under management (Towers Watson, 2012). They have similar frameworks and approaches to implementing responsible investment, in addition to being signatories to the PRI. Both funds were also invested in PetroChina, a subsidiary of China’s National Petroleum Corporation (CNPC). Despite these similarities, when the company’s systematically weak compliance with human rights norms became apparent, each fund took a very different approach to addressing these violations.  PetroChina Company Limited, the largest oil and gas producer and distributor in China, was created in July 1998 in accordance with Plan for the Organizations Structure Reform of the State Council (PetroChina, 2013). It was established as a joint stock company with limited liabilities by China National Petroleum Corporation (CNPC) under the Company Law and the Special Regulations on the Overseas Offering and Listing of Shares by Joint Stock Limited Companies in 1999 (PetroChina, 2013). Company shares have been listed on the New York Stock Exchange and Hong Kong Stock Exchange since 2000, and on the Shanghai stock exchange since 2007 (PetroChina, 2013). CNPC is the sole sponsor and controlling shareholder of PetroChina.  PetroChina’s corporate strategy focusses exclusively on aggressive growth, with strategic planning goals set to 2022 in light of China’s need for resources to sustain its economic development trajectory.  In tandem with these ambitious plans, PetroChina’s human rights record has been plagued with criticism on multiple accounts (Human Rights Watch, 2003; UN Global Compact Critics, 2012; EarthRights International, 2008).  In December 2011, Norway’s Ministry of Finance announced that the GPFG would  85 continue with investments in Burma (Myanmar) despite warnings from the Norwegian Council of Ethics (the body responsible for ensuring the fund follows its’ own code of ethics) that these investments may contribute to violations of human rights. The fund’s exposure through these violations occurred via its investments in PetroChina’s oil and gas pipeline projects in Burma.    The Council of Ethics report in May 2010 recommended divestment after finding that PetroChina and CNPC operate as a single unit in Burma and that the “fund’s investment in PetroChina entails an unacceptable risk of contributing to ongoing and future, serious or systematic violations of human rights” (Norwegian Council of Ethics, 2011, p. 49). Construction of the pipeline involved human rights abuses including land confiscation, forced relocation, forced labour, arbitrary arrest and physical abuse at the hands of police and military personnel. In addition, the pipes were being laid in areas of active conflict in northern Burma, where an estimated 30,000 were displaced by fighting in Shan State alone in 2011 (Norwegian Council of Ethics, 2011, p. 49).  Just a month later, the Dutch pension fund ABP came to a very different conclusion about maintaining its investments in the Chinese oil company. In January 2012, ABP fully divested its interests in PetroChina and blacklisted the company for non-compliance with the United Nations Global Compact Principles.36 In explaining this decision, ABP confirmed that it had repeatedly urged CNPC, to improve its behavior to prevent complicity in human rights violations in conflict-ridden Sudan through its extensive financial links to the country’s repressive regime (ABP, 2012, p. 6).                                                  36 Launched in 2000, “The UN Global Compact is a strategic policy initiative for businesses that are committed to aligning their operations and strategies with ten universally accepted principles in the areas of human rights, labour, environment and anti-corruption” (UN Global Compact, 2013, ¶1).  86 These divergent outcomes across the funds are puzzling for two reasons. The first relates to the criteria for exclusions or divestment used by each fund. Both funds have product and conduct-based criteria, though a stronger commitment to humanitarian and human rights principles is evident in the Norwegian context. Typically, this stronger commitment is likely to translate into a lower level of tolerance for human rights violations and greater likelihood of divestment from norm violators. The GPFG’s assets are prohibited from being invested in companies that are engaged in the following activities (either directly or through subsidiaries): the production of weapons that violate fundamental humanitarian principles through their normal use (chemical weapons, biological weapons, anti-personnel mines, undetectable fragmentation weapons, incendiary weapons, blinding laser weapons, cluster munitions and nuclear arms) (Norwegian Government Pension Fund Global, 2004). The GPFG is also not permitted to be invested in companies that develop or produce key components for these types of weapons. Investments in tobacco companies and bond investments in countries restricted due to sanctions are also prohibited (currently Iran, North Korea and Syria) (Norway Ministry of Finance, 2014a).  Product-based exclusions are similar for ABP, where companies involved in the production of land mines, cluster bombs, or chemical or biological weapons are excluded from the investment universe (ABP, 2014b). In cases where a company may become exposed to these sectors through acquisitions or suppliers, the fund is expected to exclude the company (ABP, 2014b). However, the criteria for conduct-based exclusions are somewhat stricter (with a distinct governance process for deliberating these issues) in the Norwegian setting than the Dutch.  Under this set of criteria, a company may be excluded from the GPFG if there is an unacceptable risk that the company contributes to serious human rights issues outlined in the Council of Ethics’ mandate, including aspects such as murder, torture, deprivation of liberty,  87 forced child labour, severe environmental damage and corruption (Norway Ministry of Finance, 2014a; 2014c; GPFG, 2004). These types of explicit commitments are not made by ABP, whose policy simply states that conduct-based exclusions are based on company violations of the UN Global Compact (ABP, 2012 and APG, 2014) which cover a much broader set of issues beyond human rights. Aside from the policy, the GPFG is more likely to divest from a company than the ABP when divestment decisions taken by the Ethics Council and NBIM are aggregated (see Table 3.1).  Table 3.1. Scope of corporate engagement, 2013 Institutional Investor ABP (Dutch) GPFG (Norwegian) Engagement leader APG NBIM Ethics Council Companies engaged37  196 2,304 40 Companies divested  2 2738 2139 Given that these two influential institutional investors made their respective decisions within weeks of each other, what is PetroChina reasonably expected to glean about the consequences of failing to comply with human rights norms in light of these mixed signals? Are some human rights abuses and some repressive regimes more palatable to investors than others? How are these tactical variations across seemingly similar cases to be explained? This empirical puzzle motivates the core research question for the paper: When faced with a portfolio company that is a systematic violator of human rights norms, why would an institutional investor with strong human rights commitments continue engaging the company rather than divesting from it?  To                                                 37 These figures refer to direct contact through letters, phone calls and meetings. Proxy voting-related engagements are disclosed separately.  38 11 coal-mining and 16 gold-mining companies 3913 for environmental damage, 3 for human rights violations, 2 ethical norms, 3 for violations related to war or conflict. For observations: 1 was kept for observation on corruption and 1 for confidential observation  88 answer this question, I conduct an in-depth comparative analysis to investigate whether existing theoretical expectations for this variation hold in this particular case. The research design used for this investigation is elaborated next.   3.3 Analytic framework, data and measures Given the focus of the literature on traditional corporate governance issues such as mergers and acquisitions or executive remuneration, it is not clear whether similar exit and voice dynamics hold true for social shareholder engagements (SSE’s) on issues such as human rights, as is the focus for this paper. As pointed out by Goodman et al. (2014), the concept of SSE attempts to reconcile the array of definitions used in the SRI and engagement literature whereby shareholders voice issues of concern to companies and where the nature of the issues tend to focus on the social, environmental and ethical impacts of corporate behaviour. However, unlike Goodman et al (2014), and drawing from the previously alluded to body of literature on shareholder value, I do not assume that profit maximisation and principled action are mutually exclusive drivers for corporate engagement on human rights by institutional investors. The expectation that SSE will gain prominence within the responsible investment movement as a whole (Gond and Piani, 2013), makes it important to unpack the dynamics of exit and voice more fully in thematic areas such as human rights.  Thus, in an effort to test whether current explanations for voice and exit options by institutional investors holds true in SSE, the remainder of the paper evaluates the explanatory power of three likely independent variables: firm responsiveness, financial risk and reputational risks as per the probabilistic hypotheses inferred from the literature (see Table 3.2).   89 Table 3.2. Summary of analytic framework Dependent Variable Independent Variables Hypotheses Indicators Divestment from systematic human rights norm violator  Firm responsiveness [IV1]  [H1] Low firm responsiveness is more likely to result in divestment [Assumption: divestment follows engagement] • Quality and quantity of communication between firm and investor Financial risk profile   [IV2] [H2] The greater the exposure to financial risk stemming from the norm violation, the more likely divestment will occur  • Size of holding • Direct financial costs attributed to the type of rights violation by the company  Reputational risk profile   [IV3]  [H3] The greater the exposure to reputational risk stemming from the norm violation, the more likely divestment will occur. • Urgency (direct civil society pressure, media pressure and peer pressure (through other investor networks and global standards). • Self-view as ‘responsible investor’    The analytic framework is applied using a qualitative case study approach (Eisenhardt, 1989) to enable the examination of exit and voice processes within their institutional context.   Of note is that both pension funds (NGPF and ABP) analysed show a high degree of complementarity on their relative size, commitment to responsible investment and the types of approaches used to implement this commitment. Likewise, the pair of chosen cases falls under  90 the category of “coordinated market economies” (CMEs) based on the classification system provided by Hall and Soskice’s (2001) seminal work on varieties of capitalism. Cumulatively, these constitute a vector of variables that are held constant across the two cases. As such, the cases are selected using a most similar systems design. The features of this design are elaborated in greater detail in Appendix 1 which covers the positon of each fund within the domestic pension system, relative size, governance structure, strategic investment mix and respective responsible investment approach.  The chosen design is appropriate to the current state of the literature and the research question because it enables the study to move from a hypothesis testing mode of analysis to hypothesis generation (based on the degree of confirmation achieved) (Gerring, 2007, p. 131). In the first instance, the cases are selected because they differ on the outcome of theoretical interest (engagement vs. divestment) but are similar on various factors that may have contributed to that outcome. The expectation of the analysis is that intensive study of the cases will reveal the key factor (s) that differ across the cases and provide an indication of potential causes (Gerring, 2007, p. 131). If a particular strong hypothesis emerges from the analysis, future analysis can move on to hypothesis testing across cases that exhibit different scores on the independent variable, but similar scores on all other possible causal factors. Confirmation of the hypothesis could also enable a deeper examination of potential causal mechanisms (Gerring, 2007, p. 131). The design also provides the advantage of not having to measure control variables in great detail, where factors assumed in common can simply be asserted as such (Gerring, 2007, p. 133).  Data was collected through 16 interviews with expert representatives of the institutional investors in the two cases of interest (ABP, GPFG), human rights NGOs, responsible investment service providers, finance-sector specialists and representatives of relevant global governance  91 initiatives. These interviews took place between November 2013 and October 2015, and summary breakdown of interviewees is included in Appendix 2. Additional data was gathered through desk research covering government publications, press releases, policy statements and publicly-available NGO reports. With regard to measures for each independent variable, firm responsiveness is evaluated qualitatively based on communications between firm and investor in the three-year period prior to the decision to continue engagement/divestment. Typically large-N studies on the impacts of engagement and divestment focus on changes in share price as a measure of financial risks. However, this measure would not be appropriate for the present study because a post-hoc share-price increase or decrease does not necessarily convey whether perceptions or calculations of these changes factored into a tactical decision about how an investor should respond to a firm’s human rights practices. For this reason, this study qualitatively evaluates whether direct financial costs were attributed to the type of rights violation by the company in their interactions with investors, and the size of the investor’s holding in the firm.   Within the corporate governance literature, reputational risk is operationalised in a variety of ways. In some instances, the resources diverted from a company in order to mitigate reputational damage in response to negative publicity or consumer boycotts is seen as a direct cost, and thereby considered a direct financial risk (Hebb and Wojick, 2005). In other instances, these financial effects may be more difficult to quantify, and reputational risk is addressed apart from direct costs (Gifford, 2010). This paper aligns more closely with the latter approach, particularly because reputational risks for an investor differ from those faced by a company. For investors reputational risks matter far more in a normative sense, and as a basis for legitimacy (Gifford, 2010). Hence, reputational risk is measured along two dimensions:   92 1) The degree of civil society, media or peer pressure which shape norms of ‘appropriate behaviour’ expected by a fund, and  2) The extent to which divestment or engagement is justified based on whether or not the action aligns to the fund’s view of itself as a ‘responsible investor’ (failure to align will mean that it cannot be taken ‘seriously’ as a responsible investor).   3.4 Findings Based on the analytical framework, theoretical expectations drawing on Hirschman (1970) would suggest that ABP’s decision to divest from China and the GPFG’s decision to continue engagement could be explained if the following conditions are met: a) PetroChina was more responsive to the GPFG’s engagement efforts than ABP’s, prompting the former to continue engagement through dialogue and observation. b) Exposure to financial risks stemming from ownership of PetroChina and the company’s alleged human rights violations were deemed to be greater in ABP’s case than the GPFG’s, prompting the former to divest.   c) Exposure to reputational risks stemming from ownership of PetroChina and the company’s alleged human rights violations were deemed to be greater in ABP’s case than GPFGs, prompting the former to divest.  Each of these conditions were qualitatively investigated in the study through a review of documentary evidence, media reports and in-depth expert interviews, and the findings are summarized next.    93  Hypothesis 1: Firm responsiveness Low firm responsiveness to shareholder pressure is more likely to result in divestment. Firm responsiveness is measured by examining the quantity and content of interactions between the company and respective pension funds. Both the GPFG and ABP had been engaging with CNPC for approximately three years regarding the company’s operations and related complicity in human rights violations being perpetuated in Sudan and Burma (Brooks, Lake, DenUyl, Mestad, interviews with author, 2013; APG, 2009, p. 13).  According to ABP Investments, the pension fund engaged with the local management of CNPC in Sudan and requested that company to fulfill several demands related to compliance with human rights norms (Global Compact Critics, 2012; ABP 2012; APG, 2010). These included a request for management to execute an assessment of the impact of CNPC’s activities on the local community; transparency on the working methods of their security staff; the initiation of accountable dialogue with company stakeholders and a public statement on the company’s role in the democratic development processes in Sudan (Global Compact Critics, 2012; ABP, 2012; APG, 2009-2010). However, CNPC responded that they do not comment on political situations, staying silent on the alleged human rights violations and their potential role (Lake, interview with author, 2013). ABP’s investment policy is based on the ten principles of the Global Compact40, and that CNPC’s lack of concrete policies to prevent involvement in human rights violations in Sudan violated Principle 2, which states that businesses should make                                                 40 Founded in 2000, the UN Global Compact is an initiative that aims to encourage business to adopt sustainability-related and socially-responsible policies and to report on their implementation. Firms sign up to ten principles related to human rights, labour rights, the environment and anti-corruption.   94 sure that they are not complicit in human rights abuses (ABP, 2012; Lake, DenUyl, Brooks, interviews with author, 2013).  The goal of this divestment decision is explained in terms of a signaling effect, particularly as an outcome of poor firm responsiveness (Lake, 2013, interview with author). According to the ABP’s annual report:  By excluding PetroChina, it is our intention to signal to CNPC that we do not agree with the manner in which it has developed its activities in Sudan and Burma. We spent three years conducting an intensive dialogue with PetroChina and its parent company CNPC, but the latter refuses to discuss the political and human rights risks associated with its activities. Furthermore, the company is implementing insufficient structural measures to identify the impact of its operations on the environment and local communities and to limit the damage caused by them. (ABP, 2012, p. 6) Thus, ABP had urged CNPC repeatedly to improve its behavior to prevent complicity in human rights violations and the CNPC failed to adequately address the concerns, leading to the divestment decision (ABP, 2012; Lake, DenUyl, interviews with author, 2013) The Norwegian Council on Ethics reports a similar lack of engagement on the part of the company. The Council’s recommendation on divestment from PetroChina explained that in accordance with the ethical guidelines, the Council sent a letter to PetroChina through Norges Bank41 on 20 November 2007 requesting information about the company’s role in the planned gas pipeline from the Shwe field on the west coast of Burma (Norwegian Council on Ethics, 2010, 2). The letter was sent in response to media reports that PetroChina had signed a Memorandum of Understanding (MoU) with Myanmar Oil and Gas Enterprise (MOGE), a Burmese stated owned company, on the sale of gas from the Shwe field in Burma (Norwegian Council on Ethics, 2010, 17). The Council requested information on whether the agreement                                                 41 Norges Bank is responsible for the fund’s asset management. Further details in Appendix 1.   95 included the construction of the pipeline, as well as PetroChina’s role and responsibilities in the pipeline project, as the Council was previously aware of serious human rights violations in the Burmese Oil and Gas sector due to previous engagement with the French oil company Total, regarding their operations in the region (Mestad, interview with author, 2013; Norwegian Council on Ethics, 2010, p. 1; Norwegian Council on Ethics, 2005).  The Council recommendation reports that in a letter, dated 12 December 2007, PetroChina responded that it “does not have any direct business contacts with, ties to, or associations with Burma; it has no exploration, production or operations in Burma; nor does it maintain any joint ventures, offices or employees, sell any products or provide any services in Burma” (Norwegian Council on Ethics, 2010, 17). PetroChina also noted that that the agreement with MOGE was entered into by CNPC, the controlling shareholder of PetroChina and stressed that PetroChina and CNPC are two separate entities and PetroChina has no control over CNPC’s business activities (Norwegian Council on Ethics, 2010, p. 17).  However, having been suitably convinced through their own due diligence efforts that the CNPC and PetroChina de-facto operated as a single corporate unit, on 26 March 2010 the Council sent a draft recommendation to PetroChina in order to give the company the option of providing comments to the Council’s findings and assessments 15 April (Norwegian Council on Ethics, 2010, p. 1; Mestad, interview with author, 2013). The company was subsequently contacted on two occasions. PetroChina responded in a message dated 27 April 2010 that it wished to address the Council’s letter, but failed to meet the subsequent deadline (Norwegian Council on Ethics, 2010, p. 1; Mestad, interview with author, 2013). Other than denying direct involvement in Burma, the company did not provide any further insights into its processes for addressing human-rights related issues and conducting appropriate due diligence. Based on the  96 evidence of the strong likelihood of human rights violations and in accordance with the Council on Ethics’ mandate, divestment from PetroChina was recommended (See Norwegian Council on Ethics, 2010 for a full discussion). In this regard, then Chair of the Norwegian Council on Ethics Ole Mestad (interview with author, 2013) draws an important distinction between correspondence and dialogue, aptly stating: “They [PetroChina] did not respond. They only formally replied with a letter. There was no dialogue.” Overall no significant variation in the CNPC’s reaction can be detected between the two cases of interest, suggesting that firm responsiveness is not a distinguishing factor in accounting for the divergent responses take by ABP and the GPFG in this instance.      Hypothesis 2: Financial risk  The greater the exposure to financial risk stemming from the norm violation, the more likely divestment will occur. Each fund’s exposure to financial risks stemming from PetroChina’s complicity in human rights violations are measured by the size of the holding, as well as whether direct financial costs were attributed to the norm-violating behavior of the company, either in the long-term or the short term.  At the time of the Council of Ethics’ recommendation on the divestment from PetroChina, the GPFG’s shares in the company were worth approximately USD$92 million, corresponding to an ownership of 0.03 percent (CNPC owned 86.71 percent of PetroChina’s shares) (Norwegian Council on Ethics, 2010, p. 1). The specific value of ABP’s shares in PetroChina was not disclosed in media reports and figures were not accessible via interview data  97 regarding the divestment decision. Direct relationships to financial risk were not identified as a material factor in accounting for the decision to divest from PetroChina in ABP’s case either in media releases or interviews with the fund and its asset manager (Lake, DenUyl, interviews with author, 2013; ABP, 2012; Brooksbank, 2012; Global Compact Critics, 2012).  Likewise, in the Norwegian case, financial risks were not mentioned at all in the text of the Council of Ethics recommendation (Norwegian Council of Ethics, 2010). This is perhaps not surprising given the structural bi-furcation between deliberations related to human rights and financial value between the Council of Ethics and Norges Bank. However, it is noteworthy that financial risks were also not referenced in the Ministry of Finance’s response to the Council’s recommendation (Norway Ministry of Finance, 2011). This is a curious outcome because, as the institutional body delegated with financial oversight of the GPFG, it would be more likely to see their language reflect concerns with the financial implications of divestment.42   Thus, in this instance, immediate concerns with financial risks do not appear to have strong explanatory power in account for the variation in engagement outcomes across the two cases. While some interviewees discussed human rights in light of long-term value creation (Lake, DenUyl, Piani, interviews with the author, 2013; Thomas, interview with author, 2014), the substance of the comments that related directly to the Dutch case mainly referred to the fund’s overall strategic orientation towards responsible investment, rather than immediate                                                 42 I am not sure how to explain this outcome. As explained later, the Council’s recommendation for divestment from PetroChina was based on the fact that its parent company, the CNPC, was involved with alleged human rights abuses in China. The Council also made the point that there was a very high degree of overlap between the persons tasked with corporate governance at both firms. However, the Ministry disputed its responsibility for human rights risks committed by CNPC because corporate responsibility does not run upward from a subsidiary to the parent company (but rather the converse). Hence, the substance of their disagreement was on the route of exposure to the human rights violations being committed, rather than on the financial risks these violations may pose for their investments in PetroChina.   98 concerns about share-price declines due to PetroChina’s involvement in with potential or actual human rights violations.    Hypothesis 3: Reputational risk  The greater the exposure to reputational risk stemming from the norm violation, the more likely divestment will occur. Based on the analytical framework, theoretical expectations would suggest that the overall reputational risk to ABP (through different pressure points and internalisation of responsibilities for human rights) would be qualitatively higher than in the case of GPFG. The lower thresholds for these risks and lack of pressure for divestment could have motivated the latter fund to continue engagement without fear of damaging its own reputation. Findings as they relate to each indicator are elaborated next.   Pressure points (civil society, media, peers) Neither Burma nor Sudan were exempt from sustained civil society and media coverage in relation to human rights violations (Frankental, Howard, Tripathi, Dunnebacke, Williams, interview with author, 2013; Conflict Risk Network, 2010; Global Compact Critics, 2012).  Trade unions in both Sweden and Netherlands had expressed long-standing concerns about the use of forced labour in Burma (Simard, interview with author, 2014), and particularly in the Dutch case, had raised these issues with employee trustees associated with the governance of the  99 fund (Den Uyl, interview with author, 2013).43 However, based on public statements by the funds and interview data, there is no indication that one fund was under greater pressure than the other to divest from PetroChina due to media or civil society pressure. The press being highly free in both countries suggests that it is reasonable to expect both funds faced similar levels access to information about civil society pressures for divestment (Freedom House, 2010).44 A case in point is the formal complaint against PetroChina filed with the UN Global Compact in January 2009 (prior to each fund’s respective decision on PetroChina). Given that both funds benchmark portfolio companies against the UN Global Compact, and are well-integrated within related informational networks (Piani, Mestad, Lake, Brooks and DenUyl, interviews with author, 2013) it would be surprising if this significant event went unnoticed by either fund. The initiative garnered support from over 80 civil society organizations from 25 countries (Investors Against Genocide and SOMO, 2008).  According to the complainants, the company was uniquely positioned to influence the government of Sudan, but failed to act and denied involvement in the humanitarian crisis in Darfur. However, to their disappointment, the Global Compact Board dismissed the complaint and maintained that the CNPC (PetroChina’s parent company) had been active in supporting sustainable development in the country Sudan. The Board also took note that CNPC "had engaged in Global Compact learning and dialogue activities on conflict-sensitive business practices" (Global Compact Critics, 2012).                                                  43 One challenge empirically is the issue of temporality. External pressure may have prompted engagement, but the role of these pressures in sustaining engagement is less clear.  44 The relatively high flow of media across borders in Europe may make it difficult to also sharply separate such pressure across the two cases. The one caveat to this may be the linguistic differences involved, where national-language press may have a higher viewership domestically.   100 There is limited evidence that peer pressure may have been a contributing factor in the Netherlands, as ABP was not the first Dutch pension fund divested from PetroChina (Lake, interview with author, 2013). In 2008 PGGM, another Dutch pension investment giant sold its PetroChina shares after their own engagement strategy failed to produce desired results on human rights issues. At the time, PGGM indicated that the company had not taken adequate steps to avoid involvement in human rights violations or to contribute to resolving human rights issues in Sudan (Responsible Investor, 2009). Moreover, in January 2009 the U.S. pension group TIAA-CREF also sold its PetroChina shares in protest against PetroChina’s business links to the Sudanese government presiding over genocide in Darfur (Global Compact Critics, 2012).  Data limitations in accessing personal communications of fund staff with peers and peer networks prevented a thorough investigation of the extent to which this factor may have shaped the engagement outcome. Interviewees did not independently identify peer-pressure from other pension funds (other domestically or regionally) as a factor that influenced decision-making on tactical options for how to respond to PetroChina, though it should be noted that interview prompts did not address this factor specifically.   Internalization of the fund’s obligations for human rights Dutch interviewees placed a strong emphasis in explaining ABP’s decision to divest from PetroChina in light of the fund’s identity as a responsible investor and having a strong responsible investment policy. For example, Rob Lake states “With an [ABP] office in Hong Kong, the fund wanted to be seen as an ‘appropriate investor’ by China…but nonetheless there were strong signals at the Board level to demonstrate that the responsible investment policy had teeth, particularly related to exclusions under the complex area of human rights. These policies  101 were a matter of the fund’s own reputation and beliefs” (interview with author, 2013). Similarly, from the board perspective, Den Uyl observes: [within ABP] there has always been a strong movement that we don’t want to invest in a company that’s doing bad things. The other is the idea that you can use RI [responsible investment] as a means for change or influencing and making a better world. I am really convinced that you could and should use RI to make it a better world…	You see that by having a fundamental policy you can have and play a role in improving workers’ rights, not everywhere, but in certain cases, engagement really helps to change policies of companies…and when you don’t as in Sudan and Burma [referring to PetroChina], then you should move on. (interview with author, 2013) These comments are illustrative of a deep-level of internalization of responsibility for human rights as part of ABP’s approach that are also captured across organizational documents noting that the fund is “aware of the far-reaching influence of [its] investments and the social responsibility this implies” (ABP, 2014b). The fund’s annual reports are replete with similar sentiments. For instance, the 2010 annual report highlights: “through its policy of responsible investment, the Board of Trustees is reflecting the growing realisation of the importance of social responsibility as a pension fund” (ABP, 2010, 45).  Another internal document explains that sustainable economic growth, as well as information on environmental, social and corporate governance issues are all important factors in the fund’s investment analysis, and although issues like human rights are not always covered in a company’s financial statements, they are particularly relevant for a long-term investor such as ABP (ABP, 2012). As noted earlier, the observed outcome in the PetroChina case is also puzzling because the GPFG’s policy is far more explicitly grounded in respect for human rights than APB’s policy. GPFG’s policy prohibits investments in companies which produce weapons that violate fundamental humanitarian principles and recommends exclusion if there is an unacceptable risk that the company contributes to or is responsible for the following serious and systematic human  102 rights violations, particularly in situations of war or conflict (GPFG, 2004). In assessing whether a company shall be excluded, the Ministry may among other things consider: the probability of future norm violations; the severity and extent of the violations; the connection between the norm violations and the company in which the Fund is invested; whether the company is doing what can reasonably be expected to reduce the risk of future norm violations within a reasonable time frame; the company’s guidelines for, and work on, safeguarding good corporate governance, the environment and social conditions; and whether the company is making a positive contribution for those affected, presently or in the past, by the company’s behaviour (GPFG, 2004). This commitment was highly evident in the justifications provided by the Council of Ethics to the Norwegian Ministry of Finance. Ola Mestad (interview with author, 2013) explains “we felt quite confident from our analysis that in this [PetroChina] pipeline project, that there was a higher risk of violations of human rights on that pipeline project than on the Total pipeline [analysed earlier] because North-East in Burma, where this pipeline is, has several minorities and active conflicts, so when the pipeline reaches those areas, then conflicts are going to escalate.”  Similarly, the Council of Ethics recommendation states:  In the Council’s opinion, nothing indicates that the present situation in Burma is any different from 20 years ago, when preparations for the Yadana pipeline started. The reports of the UN Special Rapporteur and other sources confirm that the Burmese military’s systematic abuses against the civilian population are extensive and follow the same pattern as before. The Council therefore assumes that the construction of CNPC’s pipelines will entail severe and systematic human rights violations (Norwegian Council on Ethics, 2010, p. 5).  It goes on to stress:  the construction of onshore oil and gas pipelines in Burma will entail severe and systematic human rights violations, including forced labour and extensive abuses against the civilian population. Even though it is the Burmese authorities and not the company who in principle will commit the violations, there is a link between the violations and the company’s operations in the sense that the violations take place to facilitate the company’s future operations (p. 19).  It is also important to note that although the Norwegian Ministry of Finance ultimately rejected the Council on Ethics’ recommendation to divest from PetroChina, care was taken to  103 clarify that the decision was made on the basis of divergent opinions regarding company law, and not on the likelihood of human rights violations.  Excerpts from the Ministry of Finance’s (Norway Ministry of Finance, 2010) response to the Council on Ethics’ Recommendation illustrate their rationale: The company in which the GPFG owns shares, PetroChina, is not directly involved in the activities on which the Council on Ethics based its recommendation. It is Southeast Asia Crude Pipeline Company, another subsidiary of PetroChina’s parent company, China National Petroleum Corporation (CNPC), which is constructing the pipelines in cooperation with, among others, the state-owned Burmese company Myanmar Oil and Gas Enterprise (MOGE). The main issue in this case is therefore whether PetroChina can be held accountable for human rights violations to which its parent company, CNPC, may be involved through its subsidiary Southeast Asia Crude Pipeline Company. CNPC, which is 100 percent state-owned, owns 87 percent of PetroChina” (¶ 5). In its assessment, the Ministry of Finance takes as a premise that large international corporate structures differ greatly in terms of their degree of integration, and that organisational, and thus staff-related, overlap of corporate functions arises relatively frequently. Moreover, it is not unusual for leading employees of a majority shareholder in a company to occupy leading positions in the governing bodies of the subsidiary. PetroChina and its majority shareholder CNPC appear to be closely linked. Nevertheless, the Ministry of Finance has concluded that PetroChina’s links with CNPC are not such that the two companies should in practice be regarded as a single entity. (¶ 6).  The Ministry’s response to the Council’s recommendation stressed that the human rights situation in Burma remained very worrying. The response also confirmed that Ministry had placed other restrictions applicable to the Fund's portfolio investments related to Burma, investments in Burmese government, and  in companies that sell weapons to the Burmese regime (Norway Ministry of Finance, 2010). This justification for continued investment in PetroChina was seen as a weak response by a number of civil society organizations concerned about human rights violations in Burma. For example, Paul Donowitz, the International Campaign’s Director for EarthRights International said “Norway’s decision to continue investing in PetroChina was based on its belief that CNPC and PetroChina should be regarded as separate companies…and  104 EarthRights International disagrees… (I)t is important to note that the Ministry did not deny that the pipelines posed serious human rights concerns to civilians in Burma. Their stated concern was in the corporate structures.” (Earth Rights International, 2011, ¶4).  The interim case evidence suggests that both funds had strongly internalized a sense of obligation towards human rights, and that the expected pattern of variation does not hold. Indeed, given the stronger emphasis on human rights and the particular attention to conflict dynamics in the GPFG’s responsible investment policy, the expectation would be that it would be more likely than ABP to divest PetroChina in light of the violations uncovered.   3.5 Analysis and conclusion A close examination of the cases reveals that expected variations across all three hypotheses are not realized and this points to some limitations in agency, organizational and stakeholder theories to accounting for tactical variations in shareholder advocacy on human rights. Across the two cases PetroChina was equally unresponsive to engagement efforts of both funds. Immediate concerns with financial risks due to PetroChina’s involvement in with human rights violations were also not significant in explaining each fund’s choices with regard to continuing engagement and divestment. Finally, it is not apparent that ABP was under any greater pressure to divest than GPFG, based on the threats to its own reputation. There is however some evidence to suggest that peer pressure may have been a stronger factor in the Netherlands than in the Norwegian case. Cumulatively however, variations in firm responsiveness, exposure to financial risk and reputational risks do not appear to have strong explanatory power in accounting for why ABP was compelled to divest from PetroChina but  105 GPFG was not. The case analysis reveals that the likely explanatory variables for tactical variations in shareholder advocacy may need further interrogation to be theoretically robust.  One promising possibility for developing this more robust theoretical account is to examine the political institutions that shape exit and voice decisions.  One significant difference between the Norwegian and Dutch systems is the existence of a state-based veto point on divestment decisions – in this case the Norwegian Ministry of Finance for the GPFG. In the GPFGs case, it is important to note that the Fund was invested in PetroChina and while the company did not have direct investments in Burma, its parent company CNPC did. The Council on Ethics made the case that PetroChina and the CNPC effectively operated as one organisation, and that the CNPC had a history of securing state-to-state infrastructure projects and subsequently passing them onto PetroChina (Norwegian Council on Ethics, 2010).  As a result, the likelihood that the fund would hold a company that was complicit in human rights violations was a highly significant risk. The Ministry counter-argued that lines of accountability can only run from a parent company to its subsidiary, and not the other way around, thereby precluding the fund from divesting (Norway Ministry of Finance, 2010).  The motivations and political context for this decision have not been investigated thoroughly as yet, although both documentary and interview evidence run counter to the Minister of Finance Siv Jensen’s recent assertions that the GPFG is not a foreign policy tool (Norway Ministry of Finance, 2014b). At the time of the Norwegian Ministry of Finance’s decision on PetroChina, diplomatic relations between China and Norway were under significant strain as the Norwegian Nobel Committee had just awarded the Nobel Peace Prize to Chinese dissident Liu Xiaobo (Anonymous, interview with author, 2013). China reacted swiftly and negatively to the announcement of the award, initially even censoring news about the announcement of the award  106 in China. Foreign news broadcasters including CNN and the BBC were immediately blocked, while heavy censorship was applied to personal communications (Anonymous, interview with author, 2013; Baculinao and Gu, 2010).  The Chinese Foreign Ministry denounced the award to Liu Xiaobo both in the media and through formal diplomatic channels – lodging an official complaint with the Norwegian Ambassador to China (Vehaskari, 2010; Xinhua News Agency, 2010).  Following the announcement of the Nobel Peace Prize, celebrations in China were either stopped or curtailed and prominent intellectuals and other dissidents were detained, harassed or put under surveillance (Feng, 2010).  Sixty-five countries with missions in Norway were all invited to the Nobel Prize ceremony, but fifteen declined, in some cases due to heavy lobbying by China (Norwegian Nobel Institute, 2010).45  It is possible that the divestment from PetroChina would have been seen as a further condemnation of China’s human rights policies (given the dominance of state-ownership in the company) and the diplomatic gamble generated by announcing divestment from PetroChina was perhaps deemed too risky by the Norwegian Government. The possibility that diplomatic relations may have influenced the Norwegian Ministry of Finance’s decision suggests that power relations between states, and not just investors and firms may shape how responsible investment is carried out. From a comparative perspective, the privatization of ABP and its de-linkage from a state-based institution suggests that a similar foreign policy quandary in the Dutch scenario                                                 45 Non-attendees included Russia, Kazakhstan, Tunisia, Saudi Arabia, Pakistan, Iraq, Iran, Vietnam, Venezuela, Egypt, Sudan, Cuba and Morocco.  107 may not have necessarily impacted its decision to divest from PetroChina (though this is speculative).46  Without over-reaching the argument at this stage, the case study at minimum suggests that existing theories of shareholder engagement that have been developed in the context of primarily financial issues and traditional corporate governance concerns may not translate easily into the context of social shareholder activism on issues such as human rights. Specifically, the case study suggests that the transnational power relations as well as particular institutional arrangements at the domestic level could have a causal relationship to particular outcomes (divestment or engagement).  For instance, differences between public and private pension funds with regard to the variation in their sources of authority (public versus private) to diffuse human rights norms through divestment and engagement strategies should be explored more systematically. The case also suggests that firm characteristics may also matter such that a public pension fund may be less likely to divest from a foreign-owned state enterprise.   Theoretical grounding for research into local and global political factors shaping engagement and divestment decisions by institutional investors may need to come from outside the business ethics and corporate governance scholarship.  Specifically, there appears to be adequate cause to search for explanatory narratives that go beyond the current literature’s emphasis on shareholder salience and Hirschman’s (1970) model. Although institutional investors have been neglected as a unit of analysis in studies of transnational activism by political scientists, this body of literature confirms the importance of unpacking advocacy tactics                                                 46 As noted in the ABP’s fund profile in Appendix 1, the Dutch pension fund was originally a government controlled entity that fell under the authority of the Minister of Home Affairs in The Hague (Lutgens, 1999). In January 1996, ABP was privatized although its primary function remains unchanged (Lutgens, 1999).   108 in achieving better compliance with global human rights norms and standards. Prominent examples include Charli Carpenter’s (2007) probing analysis of why children born of rape in conflict zones (as a subset of issues related to sexual violence in conflict zones) received very little attention by human rights advocacy groups, despite meeting expectations for issue salience uncovered through previous analysis. Carpenter’s (2007, 2014) scholarship draws special attention to network politics – the variable degree of linkage INGOs have with issue gatekeepers and issue entrepreneurs as an important explanatory variable in accounting for particular advocacy outcomes. Like Carpenter, Murdie (2014) suggests relations between INGOs within a particular advocacy network are affected by the structural characteristics of organizations, and outcomes such as ‘free-riding’ can limit the benefits organizations receive from the network. Turning away from external relationships, other scholars probe the internal dynamics of advocacy organizations for insights on tactical variations.  As Wendy Wong (2012) argues, the preferences, ideals, altruism, or self-interest of an INGO’s own managers, employees, and volunteers shape the organization’s mission, vision, values, structure, and tactics. Aside from organizational attributes and network linkages, institutional factors have also been explored at length as potential explanations for tactical variations in transnational advocacy. Sarah Stroup’s (2012) Borders Among Activists contends that INGOs are deeply tied to the legal and cultural environments of their place of origin. Institutional factors including regulations political opportunity structures, resources and social networks leads Stroup to hypothesize about “varieties of activism.” Specifically, she anticipates that the institutional environment of the home country explains differences in INGO behavior, including issue selection, advocacy and research agendas, and fundraising strategies. Studying the effects of these factors on shareholder-led advocacy presents a number of rich avenues to add to the current theories that are usually  109 applied by business ethics and corporate governance scholars, while also providing an expanded empirical focus for political scientists.  This mutual benefit across disciplines is perhaps an outcome Hirschman himself would condone.   Implications for shareholder power Expectations of corporate responsibility for human rights, particularly in zones of weak governance continue to evolve within the realm of transnational politics. In some cases overt political violence can erupt and die down just as quickly, such as in countries like Bahrain during the Arab Spring (Bellin, 2012), whereas in other cases historical political cleavages can span decades, such as in the Democratic Republic of Congo. The different possibilities for action and exposure to human rights violations across these temporal horizons, as well as the complexity of political violence mean that it can be difficult for investors to have a one-size-fits-all approach to human rights. In particular, the issue of how to deal with companies operating in environments with questionable governance regimes and complicity in on-going human rights abuses is symptomatic of the larger problem that responsible investors face in aligning investment practice with long-term sustainability, while still reacting to immediate core-business challenges and crises. One implication of these different possibilities for action relates to how issues of representation and accountability are connected to the exercise of shareholder power. It should also be noted that very little documentary and interview evidence related to ABP’s and the GPFG’s respective decisions regarding PetroChina referenced how these decisions might impact accountability to the fund’s beneficiaries or indeed the communities where the alleged human rights violations were occurring. The representative from the Norwegian Council on Ethics  110 stated that their due diligence procedures involved in-country research and the recruitment of local consultants to investigate human rights claims (Mestad, interview with author, 2013). However beyond this statement, the institutional mechanisms and ethical deliberation of these issues did not arise.  Through their expanded corporate engagement activities, Clark and Hebb (2004) argue that in effect institutional investors are redefining the power relationships within the firm, and that rather than simply reshuffling the players within the existing framework of the shareholder-dominated financial system, corporate engagement holds new possibilities for humanizing capital in the global arena. Some scholars go so far as to identify an “evolutionary shift” in institutional investment that has sharpened the focus of the financial sector on corporate social responsibility (Sparkes and Cowton, 2004, p. 49), with analogous propositions advanced by Monks (2001) and Hawley and Williams (2007).  The important debates such scholarship has generated about the relationship between financial returns and corporate engagement by institutional investors certainly sheds light on why investors are increasingly motivated to exercise active oversight over investee companies. Nonetheless as this chapter demonstrates, a significant shortcoming is that political factors and power relations tend to receive scant attention in accounting for the tactical trade-offs investors make in choosing how to ensure investee companies adhere to particular global norms. As I have argued here, extant theoretical expectations about the conditions under which divestment or engagement should occur when investors are faced with systematic human rights violators do not always hold. More importantly it has shown that significant scope remains for fruitful research programs that set out to unpack the internal dynamics of shareholder engagement through attention to political variables.    111 Chapter 4: Power and Authority in Investor-driven Governance Networks The previous chapter examined how human rights issues related to investee firms are addressed by single institutional investors through diverse tactics ranging from direct dialogue to divestment. An alternate transnational advocacy model involves investors acting collectively across borders with their peers or other stakeholders to address particular environmental and social issues with one firm or a group of companies. Collaborative platforms for investor activism include the Interfaith Centre for Corporate Responsibility (ICCR), the International Corporate Governance Network (ICGN), CERES coalition and others. MacLeod and Park (2011) refer to these coalitions or alliances led by investors formed around a specific public goods issue or issue-areas as investor-driven governance networks (IGNs). The intent of these networks is “to purposively steer, i.e. govern, the behavior of market actors through the broad range of tools at their disposal, including the legally defined rights they have as shareowners, but also their power to shape and define the obligations of the business community at large” (MacLeod and Park, 2011, p. 54). Focusing specifically on the thematic area of climate change and environmental governance, MacLeod and Park view the governance function of these networks as “an innovative form of public governance that is created and managed by private organizations for specific purposes, which in this case are investors and their collective actions” (p. 55). At the core of these networks are pension funds and mutual funds who increasingly engage corporations to advance more sustainable forms of capitalism by including non-traditional considerations into their investments. MacLeod and Park’s definition of IGN’s is clear and compelling. However, their claim that non-governmental organisations (NGOs) and social activists also participate in IGNs, either  112 formally or informally and “whose presence is often critical to the functioning of collective investor activities” (p. 54) is worth probing. This relationship between IGNs and international NGOs (INGOs) offers an interesting vantage point from which to investigate intra-network relations in transnational advocacy politics. Although the literature on transnational advocacy has proliferated dramatically over past two decades, existing scholarship has tended to focus on relations between NGOs and states (see for example Brysk, 1993; Keck and Sikkink, 1998; Price, 1998; Sundstrom, 2003a; 2003b) or the inter-relationships between NGOs within a particular advocacy network (for example Carpenter, 2007a, 2007b, 2007c). However, very little work has focused on the relationships between private actors, especially institutional investors and international NGOs who may share advocacy priorities related to the environment or human rights.  Consider the striking similarity in the rhetoric of INGOs Greenpeace and the WWF, alongside the Investor Initiative on Sustainable Forests, a relatively new IGN: For the benefit of the world's forests – their peoples, biodiversity, carbon stock and other ecological values – and honoring the rights of Indigenous Peoples and other local communities living in forested landscapes, Greenpeace will work as part of a broader global movement to deliver the following goals: Halt deforestation globally by 2020, massively reduce forest degradation globally (prioritizing intact forest landscapes) and restore 500 million hectares of native forests by 2030. (Greenpeace Australia, 2018, ¶ 3) Deforestation can disrupt the lives of local communities, sometimes with devastating consequences. Forests provide a vast array of resources to all of us, including food, wood, medicine, fresh water, and the air we breathe. Without the trees, the ecosystem that supports the human population can fall apart… To eliminate one of the largest drivers of deforestation—the irresponsible expansion of agricultural operations—WWF is focused on ensuring that agribusinesses, governments and others meet their commitments to help conserve the world’s forests. (WWF, 2018) The Investor Initiative for Sustainable Forests “supports institutional investors on their engagement with companies to eliminate deforestation and also address other environmental, social and governance issues related to soft commodity production, such as poor working conditions, land rights, and impact on indigenous peoples…The objectives of this collaborative engagement includes achieving a full commitment by companies to no deforestation and no human rights violations throughout entire supply chain and to encourage company participation in collaborative forums to develop  113 standards, policies, certifications, and/or tools to facilitate deforestation-free supply chains. (CERES, 2017, ¶ 2-8)  How might these shared philosophies and commitments play out in the global governance of sustainability issues such as environmental sustainability and human rights? The dissertation’s concern with how shareholder power may be leveraged in ways that affect power relations in transnational politics, particularly with actors beyond the firm, is one route by which this question may be explored further. Addressing this question is important in light of Avant et al’s (2010) persuasive arguments that the profit imperative which drives private actors offers a very different materially-derived source of authority in global governance compared to the moral or expert authority possessed by INGOs such as Greenpeace and the WWF.  Hence, investigating how these two very different types of advocacy actors interact through the auspices of IGNs could shed light on whether these relatively new institutional forms are indeed reshaping power relations in ways that flatten existing power hierarchies, and increase the space for marginalised voices to come to the fore in shaping the sustainable finance agenda.  In this chapter, INGO is used as an umbrella term that refers to a broad range of organizational types, from operational to advocacy organizations (Jordan, Van Tuijl and Edwards, 2006). As Collingwood (2006) notes, “a variety of structures - whether loosely or hierarchically-organized, networked, member-driven or privately funded, independent or linked to government - exist side by side within the non-governmental sector. In addition, the sector is characterized by a medley of political viewpoints and outlooks, from cosmopolitan to fascist, secular humanist to religious fundamentalist” (p. 441). Following this lead, INGO is used in this chapter to denote an ideal category that contrasts to the equally idealized private (market-oriented) and governmental sectors operating transnationally. Since the late 1980s, there has been a remarkable change in the scale and significance of INGOs as they move to center-stage in  114 areas such as poverty alleviation, sustainable development, human rights and women’s emancipation (Wils, 1995; Meyer, 1997; Meyer and Boli,1999; Sundstrom, 2003a). One of the main defining features of INGOs is that they can work simultaneously at different levels of the global system. In principle, therefore, INGOs are able to link micro-level experience with macro-level policy (Madon, 1999). According to Meyer and Boli (1999), INGOs are more or less authoritative transnational bodies employing limited resources to make rules, set standards, propagate principles, and broadly represent "humanity" vis-a'-vis states and other (market) actors. When it comes to including these “representatives of humanity” IGNs appear to vary widely. Some follow a more inclusive, multi-stakeholder organisational model, with strong institutionalised commitments to working with INGOs. An example of this model is the Coalition for Environmentally Responsible Economies (Ceres), which was founded by a small group of investors in 1989 in response to the Exxon Valdez oil spill. The membership base of 130 organisations includes environmental and social non-profit groups such as the Natural Resources Defense Council, Union of Concerned Scientists and Oxfam International, institutional investors such as the California and New York public pension funds, labor unions and other key stakeholders (Ceres, 2015).  Other IGNs working on sustainability issues are less inclusive, and tend to have fairly restricted membership, along the lines of a professional or trade association. A typical example here is the UK-based International Corporate Governance Network (ICGN) - a corporate-governance focused investor network with 500 members based in over 50 countries. The origins of the ICGN stem from the convergence of corporate governance initiatives in North America and Europe 4 during the mid-1980s (ICGN, 2015). Formally founded in 1995 in Washington DC, the ICGN does not refer to the potential role civil society  115 organisations could play in advancing its mission “to promote effective standards of corporate governance to advance efficient markets and economies world-wide” (ICGN, 2015). Both CERES and the ICGN have a global membership base, conduct research, lobby for sustainable investment policies, organise conferences, provide practical guidance to investors on responsible investment issues, and yet the observed model of governance with regard to the inclusion of other stakeholders is very different.   Do these differences represent extreme poles in the governance models of IGNs or does the reality of IGN governance reveal a far more complex set of arrangements for accommodating various forms of authority that are held by investors and INGOs? Answering this research question is the primary objective of this chapter. A number of scholars have argued that activism by institutional investors can serve as a transformative pathway for aligning capitalism with social and environmental sustainability (Monks and Minnow, 1995; Hawley and Williams, 2000; Clark and Hebb, 2004).  If investors are to live up to this expectation, surveying their interactions with the very organisations tasked with articulating environmental and social demands within global governance can offer insights into whether shareholder power in this context serves to reinforce private, market interests, or whether IGNs can indeed generate a better balance between the interests of society and profit maximization. In this respect, I view transnational shareholder advocacy for sustainability as being aligned with a vision of global governance articulated by Avant, Finnemore, and Sell (2010), namely that [global], “Governance involves the creation of new issues, new interests, and new modes of action by creative agents.  Governors can gain acceptance from those they seek to lead by offering attractive new ideas, formulating new strategies, and persuading people of the importance of new social goals” (Avant et al 2010, p. 9).   116 Here, I also echo Stevens and Kanie’s (2016) idea that  [g]lobal collective action does not end when decisions are reached, but these decisions introduce new practices in a complex political process that can bring in new actors, new ideas, and new action for sustainability. To unpack the practices of global governance that can contribute to a transformation toward sustainability, it is crucial to analyze the decision-making processes and the transformative ideas that are captured in those decisions. While neither processes nor ideas are a guarantee of transformation, they help to understand the potential impacts (p. 393).  In order to address this chapter’s research objective, I analyse INGO-IGN relations as a means to gain analytical leverage on how we understand shareholder power and authority in the context of global governance. The remainder of the chapter is structured as follows. I begin with a brief review of existing work NGOs in the context of shareholder activism in Section 4.1. My assessment of the existing scholarship on this topic shows a very low level of engagement with issues of power and authority. Next, in Section 4.2, I present an analytical framework for assessing INGO-IGN relations. I suggest the degree to which an IGN is open or closed to civil society input affects the nature of authority it can draw, with important implications for how an IGN shapes existing power relations as it carries out its mandate in the realm of global governance In Section 4.3, I extend the empirical reach of previous work on IGNs by mapping civil society integration in thirty-two such networks, across a range of advocacy themes and geographies. I measure civil society integration in IGNs by focusing on two attributes: the degree to which IGNs a) recognize civil society as key to their mandate and b) enable the participation of INGOs in delivering that mandate, either formally through board governance structures or informally through ad-hoc activities and collaboration. The chapter concludes in Section 4.4 with the observation that evaluating INGO-IGO relations from a governance lens presents the opportunity to re-design or reform IGNs in ways that can catalyze the transformative potential of shareholder activism. The study is exploratory and descriptive in nature, i.e. it seeks primarily to  117 map variations in the degree to which INGOs include civil society participation. It does not provide explanations for the observed variations, nor does it evaluate whether IGNs that access more diverse sources of authority are more successful or impactful in delivering their mandate. These are however fruitful next steps in this emerging research program.    4.1 Literature review  The objective of this literature review is to provide some background for this chapter’s focus on the role of institutional investors in advancing a sustainable financial system in the post-2008 Global Financial Crisis context. I find that although INGOs and investors may be aligned in their vision of a sustainable financial system, their cooperation towards this goal via investor-driven governance networks has not been addressed in great detail. I review the literature related to INGOs and shareholder activism, with the specific objective of learning how the concepts of power and authority have been tackled thus far, and find that attention to these issues has been minimal.    Institutional investors and the global sustainability agenda  Although states retain a pivotal role in world politics, there is an increasing acceptance that governance functions are exercised through a variety of mechanisms and that in certain contexts, governments are not necessarily the most important actors (Koenig-Archibugi and Zurn, 2006; Held and McGrew, 2007; Duffield, 2001). This evolution has motivated a broader  118 shift in international relations from a ‘cooperation paradigm’47 to a ‘governance paradigm’ (Koenig-Archibugi, 2006, p. 2). Governance in this sense, not only includes "the processes and institutions, both formal and informal, that guide and restrain collective activities" (Keohane and Nye 2000, p. 12) but also the mechanisms and relationships between states, markets, citizens and organizations through which collective interests on a global plane are articulated; duties, obligations and privileges are established; and differences are mediated (Weiss and Thakur, 2010). Global governance is a multi-faceted subject and there is little agreement about the extent to which political authority and governance capacity have migrated to transnational level. While there is evidence that major parts of the world are indeed moving in a post-national direction, this process is neither linear, nor consistent across sectors and regions, nor irreversible” (Koenig-Archibugi, 2006, p. 1). Perhaps as a consequence of this “messy center” political action that occurs transnationally and through processes of global governance engenders its fair share of criticism, especially in relation to economic globalization.48  In Globalization and its Discontents for instance, Joseph Stiglitz (2002) argues that rather than working for equity and poverty alleviation, global financial institutions are spokespersons for the financial sector and have widened the gap between developed and developing countries through undemocratic paternalism, the lack of accountability and weak transparency. In Stiglitz’ estimation, in the absence of a) effective institutional and ideological checks and balances on power and b) champions for global social justice, global governance will not be able to manage threats to financial, environmental and social sustainability, peace and security.  Such calls for                                                 47 A traditional concern in international relations has been the study and causes of cooperation between states (Krasner, 1983; Keohane, 1984; Wendt, 1992; Keohane and Martin, 1995; Lake and Powell, 1999). 48 For broad debates see: Scholte, 2004; Griffin, 2004; O’Brien, 2000; Soros, 2002.  119 fundamental reforms to better align the global financial system with social justice and environmental sustainability have been echoed by other scholars (Barnett and Finnemore, 2004; Kapur, 1998), and are even more pronounced in the post-Global Financial Crisis era (Lybeck, 2011; Calhoun and Derluguian, 2011)49. Citing slow political progress and a lack of political will on this front, Christine Lagarde, Managing Director of the IMF, describes global financial sector-reform as an “unfinished agenda” with many vested interests actively working against change (Lagarde, 2012). She cautions, “a robust recovery will only last if supported by resilient institutions and markets…Moreover, there is simply no alternative: the costs of crisis are so much higher than the costs of building a more stable system” (Lagarde, 2012, ¶ 46-48).  In light of the political challenges associated with ushering in a more sustainable global financial system, it is important to highlight that focusing on multi-lateral financial institutions such as the World Bank, World Trade Organization or International Monetary Fund offers only a narrow range of possibilities for change. Beyond these institutions, there are innumerable alternatives in the way the international community does the world's business (Forman and Segar, 2006). Various arrangements for allocating capital to global sustainable development have emerged with non-state and private actors often occupying crucial roles, including in hybrid financing mechanisms such as the Global Environment Facility, public-private partnerships like the Global Alliance for Vaccines and Immunization and numerous other “coalitions of the willing” (Forman and Segaar, 200650).                                                 49 Both books point to an alarming concern: despite the staggering scale of the crisis, much of the financial sector has gone back to “business as usual”.  50 Forman and Segaar (2006) are mainly concerned with the security sector and offer some good examples in this regard.   120 In this vein, the United Nations Environment Programme’s Inquiry into the Design of a Sustainable Financial System (the Inquiry) recently made the case for increased private sector involvement, (alongside public sector leadership) in driving sustainable finance forward.51 The Inquiry was launched in 2014 on the basis that the global financial system needed reshaping to finance an inclusive, prosperous and environmentally-sound future (UN Environment, 2016). Through an in-depth analysis of policy developments across national governments and international bodies,52 the Inquiry concluded that a “quiet revolution” is underway in aligning the financial system with sustainable development. It acknowledged that private-sector partnerships will be critical in ensuring the trillions of dollars needed to meet global targets set out in the Sustainable Development Goals53 and the Paris Climate Accord.54  Within the academic literature investigating these evolving forms of global financial governance, Eric Helleiner (2011) notes that the large flows of money associated with private global financial markets have been largely neglected. In agreement with Helleiner (2011), I view this neglect as unfortunate since the scale of private international financial flows particularly                                                 51 The Inquiry into the Design of a Sustainable Financial System was initiated by the United Nations Environment Programme (UNEP) to advance policy options to improve the financial system’s effectiveness in mobilizing capital towards a green and inclusive economy — in other words, sustainable development.  52 Results of the Inquiry were underpinned by over 80 published reports and technical papers, prepared together with more than 60 collaborating institutions including banking associations, research institutions, central banks and financial regulators, finance ministries, civil society and international organizations. Experience was drawn from 15 diverse country-level contexts and engagements and included policy and technical reviews across banking, bond markets, insurance, institutional investment and stock exchanges, with focused assessments in areas as diverse as human rights, social banking, fiduciary duty and the changing roles of central banks (UN Environment Inquiry, 2016). 53 The Sustainable Development Goals (SDGs) or the 2030 Agenda for Sustainable Development is a set of 17 Global Goals. They are measured against a set of targets and indicators that UN Member states are expected to use as a framework in setting development agendas and policies over the next fifteen years (UN Department of Economic and Social Affairs, 2018).  54 The Paris Climate Accord is an agreement within the United Nations Framework Convention on Climate Change (UNFCCC) dealing with greenhouse gas emissions mitigation, adaptation and finance starting in the year 2020 (UN Climate Change, 2018). The language of the agreement was negotiated at the 21st Conference of the Parties of the UNFCCC in Paris and adopted by consensus on 12 December 2015 (UN Climate Change, 2018). The Accord aims to keep global temperature rise this century well below 2 degrees Celsius above pre-industrial levels, and to pursue efforts to limit the temperature increase even further to 1.5 degrees Celsius (UN Climate Change, 2018).  121 through institutional investors dwarfs not just public international lending but also volume of international trade and foreign direct investment.55 If the global economy is to be made more environmentally sustainable, this powerful "electronic herd" of global money will need to be steered in greener directions, argues Helleiner (2011, p. 51).   Helleiner’s observations hold true for the social aspects of global sustainability as well. I suggest that leveraging private finance for sustainability will be shaped by the character of the global sustainability agreements in place where, at least on paper, human rights occupy a prominent space. According to the Danish Institute for Human Rights (2017a): “The 2030 Agenda for Sustainable Development is explicitly grounded in human rights. Consequently, the SDGs reflect key international human rights and labour standards and affirm that they seek to realize the human rights of all” (p. 6). More than 90% of SDG targets are linked to international human rights and labour standards, and the pledge of the 2030 Agenda to “leave no one behind” reflects the fundamental human rights principles of non-discrimination and equality (Danish Institute for Human Rights, 2017a). The Agenda also stipulates that “Follow-up and Review (FUR) processes to reach the SDGs should be inclusive, transparent, respect human rights, enhance accountability and have a particular focus on those furthest behind” (Danish Institute for Human Rights, 2017b, p. 7).  Similarly, the preamble to the Paris Agreement includes an acknowledgement “that climate change is a common concern of humankind” and that “Parties should, when taking action to address climate change, respect, promote and consider their respective obligations on human rights” (UNFCC, 2015, ¶ 11).  The latter statement, Benoit Meyer (2016) confirms, has no equivalent in the UNFCCC or in the Kyoto Protocol, and that its                                                 55 As discussed also in Chapter 1.  122 insertion in the Paris Agreement is, largely, the outcome of years of advocacy by civil society organizations. Also adopted were provisions to promote gender equality, and participation, sustainable development, and poverty eradication as a context for climate action (Meyer, 2016). However, despite the importance of human rights for the future of sustainable financial markets, empirical analysis on the politics of these “steering” efforts have almost exclusively addressed environmental issues.  An example of work that examines linkages between institutional investors and sustainability includes MacLeod and Park’s (2011) work on ‘investor-driven governance networks’ (IGNs). Focusing on climate politics, their analysis shows that many IGNs are involved in more than just the development of voluntary standards for information disclosure at the international level. They have also been engaged in more direct shareholder activism56  to influence corporate managers, and the lobbying of official institutions such as the United States Securities and Exchange Commission to require mandatory reporting of material risks associated with global warming. Likewise, Harmes (2011) argues that “investor environmentalism” through carbon disclosure initiatives overlooks the fact that these efforts address imperfect information rather than the externalization of environmental costs by firms. As a result, investors are not really addressing the core challenge of climate change, as they will continue to invest in “dirty” firms. Harmes (2011) concludes that investor-driven carbon disclosure initiatives are likely supported for reputation reasons, rather than for the utility of the data. Both works address an important question in transnational politics in the extent to which actors’ perceptions of self-                                                56 Responsible investment and shareholder activism function as strategies through which the steering of finance towards sustainable development goals occurs.    123 interest may be transformed (Helleiner, 2011). While Harmes assumes the profit-maximizing goal of investors are fixed, MacLeod and Park argue that IGNs may be able to shape the identities of investors by transforming their conceptions of economic "rationality" through processes of socialization by drawing on the work of scholars like Keck and Sikkink (1998). Together, these studies offer an initial foray into the study of investor-driven networks in the context of global governance and sustainable development. They also open up productive conversations with the now burgeoning research agenda on transnational advocacy networks and intra-network politics. It is from this point of departure that the chapter justifies its focus on unpacking how IGNs engage with other non-state actors, namely INGOs57 in steering finance towards sustainability goals within the domain of human rights. Furthermore, accepting IGNs as exemplars of private authority offers an additional site from which to interrogate how investors are leveraging shareholder power as a form of private authority within the realm of global governance.58  In the pursuit of a more sustainable global financial system, Steve Waygood (2013) makes a convincing call for more advocacy partnerships between international financial institutions (including institutional investors) and international NGOs. His arguments draw on his experience of working within both types of organisations, and particularly the lessons learned when he founded and coordinated the lobbying efforts of the Corporate Sustainability Reporting Coalition (with both NGO and financial-sector partners) at the 2012 UN Conference on                                                 57 Van Tuijl (1999) distinguishes between "operational" and "advocacy" NGOs. Operational NGOs provide social services such as education, health, or human relief, whereas advocacy NGOs lobby governments, corporations, and international organizations. In this paper, I focus on the latter only. 58 Power and authority have been discussed at length in Chapter 2. References to these concepts within this chapter follow from the earlier discussion.   124 Sustainable Development (Rio+20). Waygood (2013) essentially argues that “NGOs and financial institutions could improve the impact and effectiveness of their sustainable capital markets advocacy by working together “(p. 69). Such partnerships could help design and advocate reforms to all of the following:  1. capital market regulation;  2. transparency, accountability, and reporting provisions;  3. investment training standards, and  4. market incentive structures.  Waygood insightfully notes that to date, capital market campaigning by INGOs has mainly targeted investors as a way of influencing corporations rather than attempting to change the structure of capital markets themselves. In spite of initiatives such as the Collevecchio Declaration59, there are very few examples of NGO-led campaigns to change government policy on capital markets. In the post-Global Financial Crisis era, it is expected that there will be more attempts to generate structural reforms, including increased calls for government to reduce financial market short-termism,60 an increased emphasis on the chain of incentives that motivate companies and investors, and more focused attempts to correct the long-term market failures                                                 59 This declaration calls on financial institutions to embrace six main principles which reflect civil society's expectations of the role and responsibilities of the financial services sector in fostering sustainability. They cover commitments to sustainability; do no harm; responsibility; accountability; transparency and sustainable markets/governance (BankTrack, 2018). Following the launch of the Declaration in 2003, INGOs formed a network to carry out join direct action, lobbying and research on banks and sustainability (BankTrack, 2018). This network now consists of 40 organizations that coordinate their activities through BankTrack. The Declaration was the first civil society statement on finance and sustainability and it was signed by over 100 organisations (BankTrack, 2018).   60 Lynn Dallas (2012) explains short-termism, as the excessive focus of corporate managers, asset managers, investors and analysts on short-term results, whether quarterly earnings or short-term portfolio returns, and a repudiation of concern for long-term value creation and the fundamental value of firms, which is increasingly linked to systemic sustainability issues such as climate change, resource scarcity, migration etc. Short-termism was seen as a key driver of the 2008-09 global financial crisis.  125 arising from the capital market’s promotion of unsustainable patterns of growth (UN Environment, 2016; Lybeck, 2011).  Within this context, Waygood (2013) asserts that there is a critical need for forums where NGOs and financial institutions can come together to learn from each other in a spirit of mutual trust and respect. He further stresses that advocacy partnerships between financial actors and NGOs would be very different from the partnerships between companies and NGOs that exist today. According to Waygood (2013), contemporary business-NGO partnerships tend to involve companies taking voluntary action in the absence of government intervention, such as the marine or forest stewardship councils. The new “sustainable capital market advocacy partnerships” between NGOs and financial actors would lobby for greater government intervention to correct market failures. (Waygood, 2013, 68). Probing the dynamics of these coalitions through the auspices of IGNs, is the main task in the remainder of this chapter.    NGOs and shareholder activism on environmental and social issues  Kenneth Rodman’s (1998, 2001, 2002) scholarship on this topic examines civil society campaigns for corporate social responsibility via case-studies pertaining to human rights issues in South Africa, Burma and Nigeria. Through this body of work, Rodman argues that if global civil society can replicate the experience of the anti-apartheid movement in securing corporate withdrawal from repressive regimes, then economic sanctions will no longer be the exclusive preserve of nation states (Rodman, 2002).  For instance, in analyzing the Free-Burma Campaign (FBC), Rodman observes that consumer pressure, shareholder pressure and municipal procurement power were vital strategies – a mix of economic coercion and public shaming – for persuading multi-national corporations that their self-interest coincides with withdrawal from  126 Burma (Rodman, 2002, p. 33).   He concludes that in Burma, this strategic mix astonishingly “persuaded roughly 60 firms either to disinvest or sever their licensing and franchising relationships in Burma” (Rodman, 2002, p. 33).  One shortcoming of this study is that Rodman does not clearly distinguish between investors as shareholders, and corporate investments in Burma (such as through subsidiary relationships). For example, he notes that the aforementioned tactics were  less effective with the most strategically significant investors. Some tried to deflect societal pressures by implementing social responsibility practices, the most significant example of which was Premier Oil’s decision to provide human rights training to the security forces designated to protect its operations (Rodman 2002, p. 33).   In this context, Rodman is actually referring to firms as foreign direct investors, rather than institutional investors that own shares in firms. He also notes that unlike calls for divestment which dominated the Apartheid campaign, the Burma campaign focused more on filing shareholder proposals calling for firms to disclose the impacts of doing business in Burma (p. 16). What remains unclear from this analysis however, is the nature of the relationship between the FCB and the filers of these proposals – mainly socially responsible investment firms, religious investors working through the Interfaith Centre for Corporate Responsibility and labour-managed pension funds. As such, Rodman’s analysis treats NGOs and investor activists as a homogenous “NGO” group leaving little room to explore competing or diverging capacities to exert authority over firms. This conflation makes his conclusion that “NGOs cannot command a corporation to withdraw when private calculations of risk and reward point in the other direction” (Rodman, 2002, p. 34) problematic, as this view underemphasizes the instrumental leverage institutional investors have over firms in terms of the threat of divestment or their relative size and influence as major providers of capital in global markets.   127 Emily McAteer and Simone Pulver (2009) investigate the dynamics of transnational advocacy networks (TANs) that target not states, but multi-national corporations.  Extending Keck and Sikkink’s (1998) framework, they develop a corporate boomerang model to describe how TANs exert leverage over corporate targets. They use this model to analyze two case-studies of corporate-focused TANs, where corporate shareholders play a central role in the network – shareholder transnational advocacy networks (STANs). They explain differential goal attainment across two STANs (aimed at the Chevron Corporation and Burlington Resources respectively) by the relative cohesiveness of the two networks and the respective vulnerability of their corporate targets. The Chevron STAN was focused on getting the company to commit to environmental remediation in Ecuador a decade prior, while the Burlington STAN focused on preventing oil extraction in the Ecuadorian Amazon and to change its policy on indigenous rights.  In McAteer and Pulver’s (2009) account, corporate-focused TANs emerge when local communities living at points of production or extraction are blocked in their efforts to influence the operating practices of a corporate subsidiary. Appeals to local corporate managers may fail because they are protected by local governments loath to challenge corporate practices or have weak regulatory and enforcement capacity. In such situations, local communities engage in the strategy of creating external linkages to other groups in order to drive change via top-down pressure at a corporation’s headquarters. This pattern of influence is called the corporate boomerang model, and differs from NGO boomerang models in terms of the strategies they employ, the determinants of network effectiveness and assessments of goal achievement (McAteer and Pulver, 2009, p. 3).  Unlike Rodman’s analysis, McAteer and Pulver (2009) pay much closer attention to mapping diverse actors collaborating via STANS, that networks in both  128 cases are “complex strucures with actors operating at four levels: local nodes (including indigenous and settler communities); domestic actors (mainly domestic NGOs); international NGOs and corporate shareholders (p. 10). Across both cases, an international NGO (Amazon Watch and Amnesty International) served as a crucial link to a socially-responsible investor (Boston Common Asset Management and Trillium Asset Management), which then invited other institutional investors to co-file a shareholder proposal on the advocacy topic at hand.  On reflection, McAteer and Pulver’s (2009) study offers a much more granular understanding of network relationships at different levels, when the target is a firm and the network involves a mix of civil society and shareholder activists. For example, they allude to an investor “fact finding” mission to Ecuador whose goal was to promote increased understanding and coordination between network members and to familiarize shareholders with the governance structures of the indigenous communities in Burlington’s concessions (McAteer and Pulver, 2009, p. 15). Likewise, the Chevron STAN also facilitated direct contact between shareholders and indigenous communities. Their study offers rich insights into understanding the interplay of network, target and contextual variables in explaining differences in goal attainment of two STANs. However, I take issue with the degree of centrality accorded to investors in McAteer and Pulver’s analysis. While shareholder pressure, through the filing of shareholder resolutions in partnership with socially responsible investors was a key advocacy strategy, it was not necessarily the only strategy employed by Amazon Watch and Amnesty International respectively. Here, shareholders were approached by INGOs to join a pre-existing advocacy campaign, and their relationships with affected communities were brokered by INGOs. Hence, although empirically and conceptually very rich, McAteer and Pulver’s analysis does not offer very much insight into power differentials and diverging sources of authority within the STAN.  129 It is also unclear how these relationships might shift when shareholders invite NGOs to join pre-existing investor-driven campaigns, rather than when shareholders are invited to join pre-existing NGO campaigns.  Morton Winston’s (2002) survey of NGO strategies for promoting corporate social responsibility also includes shareholder activism. He acknowledges that since large pension funds often hold large number of shares in many companies, they can often exert considerable economic as well as moral influence at annual shareholder meetings (Winston, 2002, p. 80-81). However, while helpful in providing a broad-view of corporate-focused NGO activism, Morton (2002) like Rodman (2002), treats NGOs and investors interchangeably, offering few in-roads into understanding power relations between the two. He writes for instance: Shareholder activism makes use of the accountability of corporations to their investors in order to widen the sphere of corporate accountability to other stakeholders, and is regarded by a number of NGOs as an effective way in which to influence corporate policy and practice (Winston, 2002, p. 81).   Ultimately though, the focus for Morton is on NGOs and not investors. He concludes by noting that NGOs involved in corporate social responsibility campaigning should “combine stigmatization with other pressure tactics, such as shareholder resolutions and boycotts, with parallel attempts to change corporate thinking about human rights by means of dialogue, rational persuasion and the sharing of best practices” (Winston, 2002, pp. 85-86), but that they are unlikely to be successful in the long-run unless they mobilize consumers and governments. Hence, in Morton’s estimation, shareholder pressure may be a viable campaigning tactic, but investors (as shareholders) are not necessarily an important mobilization target or ally for NGOs.  Morton’s analysis, like much of the literature, emphasizes the relationship between firms and NGOs, but leaves the relationship between investors and NGOs largely ignored.   130 Emma Sjöström’s (2007) work on this subject also adopts INGOs rather than investors as the entry-point. She explores how civil society organizations use financial markets to pressure corporations regarding their social responsibilities, particularly in the area of human rights compliance. Her paper includes two case studies: the first shows how Amnesty Business Group (ABG) bought shares in 12 Swedish firms in order to use shareholder rights as a mechanism for pressuring them to adopt a policy on human rights, whereas the second case evaluates how Friends of the Earth (FOE) put pressure on the construction company Balfour Beatty regarding a controversial dam project in Turkey. FOE first sought to influence shareholders and then became a shareholder itself. Like Morton (2002), Sjöström distinguishes the NGOs as confrontational (FOE) or non-confrontational (ABG) in their approach, noting that less powerful actors (here, NGOs) can achieve sought-after change by translating the problem at hand so that it fits the ideology of actors (in this case firms) who have the power to resolve it.  Sjöström’s (2007) treatment of power relations does not just focus on material imbalances between NGOs and firms. She observes that through shareholder activism, civil society organisations can educate investors on risks with a company’s approach to human rights, and the investors in turn can be a powerful force for pressuring corporations to change. In other words, by helping the financial sector to become more knowledgeable about human rights, and how these rights relate to corporate risks and opportunities, the financial sector can be an ally to civil society organisations (Sjöström, 2007, p. 165). For Sjöström, this education mechanism taps into the ideology of financial actors. Ideologies guide action and can be operationalized through the pursuit of ideals. These are principles or values that organisations can actively pursue as goals (Sjöström, 2007, p. 166). She notes that the profit maximising imperative of firms can be an impediment to a self-view that transcends economic rationality. This is contrasted with the  131 ideology of civil society organizations such as FOE which may be grounded in interconnectedness of ecosystems and the intrinsic value of non-human life, or ABG which may be guided by the view that every person enjoys a certain set of universal rights. Despite competing ideologies, civil society organisations capable of translating their own guiding logic to that of firms, can be successful. She writes: “ABG and FOE are not talking about human rights primarily as a moral issue, but are directing attention to risk and shareholder value” (Sjöström, 2007, p. 167).  Sjöström acknowledges that a risk posed by this translation of moral issues into financial risk may signal corporations are obligated to act only when it is in their financial interest to do so, and that an emphasis on such issues may lead to the neglect of other more pressing human rights issues. However, her study does not interrogate how power leveraged by civil society groups may shift when they take on the persona of investors through share ownership. Importantly, the perspective of investors remains opaque in this account.  Anastasia O’Rourke (2003) explores whether shareholder activism can deliver the change desired by activists and which strategy seems to work best for this purpose (p. 228). In providing context for her study, O’Rourke (2003) provides a helpful breakdown of the composition of shareholder activist groups, noting they may be composed of individual share owners, NGOs, socially responsible investment (SRI) funds, institutional investors such as public pension funds, private pension funds and insurance companies and umbrella investor groups such as the Council of Institutional Investors (CII). The author conducts a case study of a shareholder proposals filed at BP Amoco between 2000-2002, aimed at getting the firm to halt the development of the Northslope oil field in Alaska and redistribute the investment to the firm’s solar energy division. O’Rourke (2003) examines this shareholder activity as part of a larger campaign launched by a consortium of activists including Greenpeace, the WWF, SRI funds, public interest associations  132 and individual investors. This group organized a campaign website, held meetings with other shareholders ahead of the vote and were in contact with senior management at BP to encourage a positive response to the proposal. Given the study’s emphasis on campaign outcomes, intra-network politics (particularly on the change in campaign leadership from Greenpeace to WWF in 2002) and relationships between investors and INGOs are not explored in any detail. Like Sjöström (2007), O’Rourke (2003) also tends to treat shareholder activists as a cohesive group, noting “the shareholder activist community currently seems to need a dedicated research effort to establish firm connections between CSR and core business issues in order to bring more shareholders on board” (p. 238) but notes that “As a society, the broader questions still need to be addressed as to what point we are willing to allow a model of ‘shareholder democracy’ to take over from other democratic processes of controlling corporations’ behavior and guiding them towards developing the motive for sustainability” (O’Rourke, 2003, p. 238).  In general, the trend in the literature examining NGOs in the context of shareholder activism is to treat them as a homogenous group with other actors, including institutional investors (see also Guay et al., 2004; Ivanova 2015; 2016; Waygood and Wehrmeyer, 2003; Waygood, 2013; Voorhes, 2005). NGOs are treated as central nodes within activist networks that adopt the language and tools of finance in order to influence firms. Investors are seen as secondary actors, who join or adopt an NGO campaign, but rarely initiate or lead these efforts. As the proliferation of IGNs on sustainability issues suggests in the next section, these assumptions may not always hold. Moreover, it is clear that the existing literature has left the governance mechanisms of IGNs largely unexplored, making the present study a unique contribution to advancing knowledge on IGNs as pathways for transnational political action.    133 4.2 Analytic framework To systematize my analysis, I evaluate INGO-IGN relations along two dimensions a) the degree to which civil society actors, in this case INGOs, are recognized as a contributing force towards the IGN’s mandate and whether a formal mechanism exists to include INGOs in the initiative's governance or membership framework (recognition) and b) the actual level of INGO participation in activities toward the implementation of the IGN’s mandate. Working definitions and the operationalization of these concepts within my analysis are elaborated next.    Defining recognition The concept of ‘recognition’ occupies an important role in the discipline of political science, and political theory in particular. It denotes the idea that the idea of ‘self’ at the individual or group level, is informed in part by others’ recognition (or absence of recognition) of this self. From interpersonal relationships, to multiculturalism, identity politics, new social movements, economic inequality, human development and diverse modalities of power, theories of the ethics and politics of recognition have been widespread (Hayden and Schick, 2016). In the area of global governance, Greta Snyder (2016), for example, focuses on the role the politics of recognition plays in constructing global struggles against injustice. She argues that recognition politics serves at least two important functions: integrative and performative.  Recognition politics is integrative in that it fosters solidarity and enables individuals to recognize themselves as part of a cohesive political grouping (such as the ‘global left’ in Snyder’s study). By performing a democratic will, the group can engender a kind of global democracy even in the absence of global democratic institutions. In doing so, Snyder illuminates  134 the importance of real-world political institutions like the World Social Forum and campaigns like the anti-War protests of 2003. Similarly, Matthew Weinert (2015) uses recognition theory to challenge and extend the notion of world society, which captures a web of relations between diverse actors operating outside the formal rubric of the state. Within this study of IGNs, I draw on Snyder’s (2016) distinction between the integrative and performative aspects of to gauge whether INGOs are recognized as a vital social partner for aligning capital markets with long-term sustainability. I argue that this analysis matters because it speaks to the capacity of INGOs to serve as democratizing forces within a key institutional form that is driving global governance arenas of sustainable finance and human rights.    Defining participation  Waygood’s ideas about the potential benefits of INGO-investor advocacy coalitions resonate with the broader scholarship on the role of INGOs in global politics. Some scholars suggest INGOs deserve to be embraced by transnational governance bodies because they have better access to more ideas and issue expertise that can improve the results of deliberations (Esty, 1998). INGOs are seen as providing organized channels of political influence, enhancing civic participation, and raising public awareness about particular issues (Ayres, 2003). Moreover, INGOs, as representatives of civil society, contribute to the “democratic governance of the global economy” through increasing the transparency of governing authorities, enhancing public accountability through citizen monitoring, and enhancing representation of the marginalized communities (Scholte, 2004). To fulfil this democratizing function in global governance, certain institutional choices and provisions may affect how INGOs participate, or indeed, if they participate at all.  135 As Avant, Finnemore and Sell (2010) explain, participants in global governance engage in four types of activities: 1) setting agendas and creating issues; 2) making rules; 3) implementing mandates; and 4) evaluating, monitoring, and adjudicating outcomes (14-16). Through this rubric, the authors (2010) offer a helpful framework for measuring the institutional determinants of participation by INGOs in IGNs. It is important to note that each of these four practice areas correspond to a distinct phase in the larger process by which new norms, such as those related to corporate social and environmental responsibility (CSR), are diffused transnationally. This three-phased norm life-cycle process (Finnemore and Sikkink, 1998) comprises issue emergence (agenda setting), norm-diffusion/cascade (rule-making, implementation and enforcement) and norm internalization (evaluating outcomes).   Rather than assuming that the issues requiring global governance are obvious, international relations (IR) scholars have shown the variety of ways in which “agenda-setting” activities involve a struggle between actors to persuade others that a problem or issue exists (Avant, Finnemore and Sell, 2010, 14, also Price, 2003 and Carpenter, 2007). According to Carpenter (2007), in order for an issue to emerge onto a transnational advocacy agenda, two steps are required. First is the construction or definition of an issue as problematic and the second is the adoption of that issue as a subject of advocacy. Hence, the “‘governance of a problem cannot begin until someone defines a problem as an issue and succeeds in placing it on a consequential agenda” (Finnemore, Avant and Sell, 2010, 15).61 Carpenter locates issue emergence and its constituent effects as being temporally and logically prior to norm-diffusion                                                 61 In addition to issue-based characteristics such as salience or urgency, norm entrepreneurs (Nadelmann, 1990, Lessig, 1995, Finnemore and Sikkink, 1998) and issue gatekeepers (Carpenter, 2010) also influence outcomes in transnational agenda-setting.  136 through building campaigns, negotiating agreements, and holding targets (corporations in this case) accountable to new norms.  However, taking a different view, I propose that Carpenter’s valuable insights about issue-emergence can travel to the norm-diffusion phase as well. Even after an issue emerges onto the transnational advocacy agenda (such as climate change or corporate compliance with human rights), we do not know why norm promoters/entrepreneurs, working through platforms such as IGNs, may choose to diffuse particular aspects of that agenda over others (Carpenter, 2007, 101). I suggest that IGNs occupy a unique space in the norm-life cycle on issues related to CSR because they can serve simultaneously as norm-entrepreneurs and norm-diffusers. As norm-diffusers, they tend to have a common principled commitment to the obligations of corporations to fulfil certain environmental and social responsibilities, and see their role as providing a particular degree of influence of firms to adopt these responsibilities. In this role, they are not raising CSR, responsible investment, human rights or climate change as a “new” subjects of advocacy per-se. However, as norm-entrepreneurs, they are capable of broadening the scope and content of what is considered CSR or responsible investment by using IGNs as platform to new issues, tools and approaches.  In this unique position, what is not clear is whether the composition of IGNs affects the types of issues they advance, or indeed how they advance them. Therefore, opening up the black box of IGNs and studying how INGOs participate in key functional areas may not only shed light on norm-diffusion dynamics, but also serve as a broader illustration of how investors are wielding power in the context of global governance.  The authority to select stakeholders and to define stakeholder categories is critical for transnational decision-making, as it grants the power to determine if and how stakeholders can participate in a partnership or coalition (Dingwerth, 2007). Understanding when and how investors choose to  137 include social partners in their transnational activism may also shed light on how these processes of participation themselves are contentious and possibly historically contingent (Bruhl 2010; Dingwerth 2007, p. 125).    Data, research strategy and measures My analysis of how INGOs are recognized and included in IGNs proceeded via a qualitative review of thirty-two IGN websites. Through purposive sampling (Palys, 2008), I sought to build a foundation for analysis that covered a broader universe of cases of IGNs than examined by MacLeod and Park (2011) (which was restricted to four cases focused on environmental politics). While the resulting list of IGNs cannot claim to be exhaustive, the universe of cases was generated based on the unique characteristics of the target population and objectives of the study (Palys, 2008). Specifically, the objective was to achieve maximum variation (heterogeneous purposive sample) in order to capture the full range of IGNs in the first instance. Starting with one of the earliest examples of a formalized global IGNs (the International Coalition for Corporate Responsibility), I followed a chain-referral method to build up the universe of IGNs in the sample. This method included identifying additional IGNs listed within “members” or “partners” sections of websites of IGNs. This approach included a number of limits – it focuses on initiatives that have an on-line presence and the search was restricted to English websites (though a few non-Anglophone European IGNs were captured using this method). IGNs that are regional and global in scope were identified through the sampling method described (see Table 4.1 for a full list of initiatives). 62 Fusch and Ness (2015) note that                                                 62 Note: As the focus of this study is on transnational advocacy and global governance, I did not include IGNs that operate primarily within a specific domestic context. National IGNs such as the United States’ and United  138 there is no one-size-fits-all method to reach data saturation because study designs are not universal. To enhance the validity of one’s findings, they recommend that scholars aim for “thick and rich” data. To achieve this standard for the present study, case illustrations are used to support key claims.  Table 4.1 List of IGNs studied 1. 2º Investor Initiative 2. Global Unions Committee on Workers' Capital 3. Asia Investor Group on Climate Change 4. Inter-faith Centre on Corporate Responsibility (ICCR) 5. Association for Sustainable & Responsible Investment in Asia 6. Investor Group on Climate Change (Australia & New Zealand) 7. African Sustainable Investment Forum 8. International Investor Group on Climate Change 9. Carbon Disclosure Project (CDP) 10. Intentional Endowments Network 11. Coalition for Environmentally Responsible Economies (CERES) 12. International Corporate Governance Network 13. Climate Bonds Initiative 14. Investor Alliance for Human Rights 15. Confluence Philanthropy 16. Investor Environmental Health Network 17. Ecumenical Council for Corporate Responsibility (ECCR) 18. Natural Capital Finance Alliance                                                 Kingdoms’ Sustainable Investment Forums or the Canadian Responsible Investing Association were excluded for this reason, even though they appear on the membership lists of some global IGNs.   139 19. Emerging Markets Investor Alliance 20. Responsible Investment Association Australasia 21. European Responsible Investment Network 22. The Asian Corporate Governance Association 23. Eurosif 24. The Green Infrastructure Investment Coalition 25. Farm Animal Investment Risk and Return (FAIRR) 26. The International Integrated Reporting Council Investor Network 27. Global Impact Investing Network 28. United Nations-backed Principles for Responsible Investment 29. Global Investor Coalition on Climate Change  30. UNEP Finance Initiative 31. Global Sustainable Investment Alliance 32. Workforce Disclosure Initiative   The method used in this study was qualitative content analysis. Holsti (1969) provides a broad definition of content analysis as the application of scientific methods to documentary evidence, whereas Krippendorff (1980) defines this approach as “a research technique for making replicable, valid inferences from data to their context” (in Kim and Kuljis, 2010, 370). Qualitative content analysis is primarily interpretive in nature.  As an unobtrusive and highly contextual method, qualitative content analysis enables researchers to investigate phenomena without influencing the research procedure or outcomes, and it therefore may be less biased compared to other techniques such as questionnaire surveys and interviews. If data is readily  140 available and accessible, as with web-based content, it is also relatively simple and economically efficient technique (Ozaki and Rivas, 2002 in Kim and Kuljis, 2010). This research approach has been applied by international relations and business ethics scholars alike. For example, Charli Carpenter (2007) conducted a content analysis of 36 advocacy websites in the children and armed conflict issue area to advance claims related to agenda-setting in transnational advocacy networks. Similarly, Jose and Lee (2007) conducted a content analysis of the environmental reports of Fortune’s Global 200 companies in order to identify which environmental issues become the subject of environmental disclosures as well as how environmental concerns are institutionalized and structured.   Within my study, recognition was evaluated by examining whether civil society organizations, non-governmental organizations or other social partners such as trade unions were specifically mentioned a particular IGN’s statement of purpose through website content that describes the organization’s vision, mission and mandate. This variable was measured dichotomously (yes/no). Following Avant, Finnemore and Sell’s (2010) rubric, I measured participation along four dimensions: agenda-setting, rule-making, activity implementation and evaluation.  To determine if INGOs are involved in agenda-setting, I took a historical approach. The formation of an IGN can serve as an indication that a group of actors has been successful in identifying the issue being addressed by the IGN as a subject worthy of advocacy. From this view, an IGNs existence itself is an indicator of ‘successful’ agenda-setting. Therefore, examining the role of INGOs in an IGNs founding can shed light on whether they were part of this agenda-setting process. Here I reviewed available website information on an organization’s creation story. In general, this included content from IGN website sections entitled “about us”, “background” or “history”.  As well, INGO participation was evaluated dichotomously as well  141 (yes/no) – either INGOs played a role in establishing the IGN or they did not.  With regard to rule-making, I explored the degree to which a formal mechanism exists to include INGOs as members of the IGN and within its governance mechanisms. For this element, I also paid attention to whether any restrictions were placed on membership (such as through second-tier membership categories with limited influence over governance decisions). The coding structure for membership denoted whether INGOs were permitted as members (yes), whether there were any restrictions placed on membership (yes, with restrictions) or whether they were excluded altogether (no). The coding structure for governance mechanisms denoted whether INGOs were included (yes) or excluded (no). Data for this component of the analysis was drawn from website sections that related to “membership” or “signatories” and “governance” or “about us”.  INGO participation within an IGNs activities was evaluated using information disclosed on website sections pertaining to activities, programs, projects, tools, issues, themes, policy, events and conferences. Here I adopted a very low threshold and simply mapped whether NGOs were directly involved in any mandate-implementation activities or not (yes/no). Finally. monitoring and evaluation activities were deemed relevant only for initiatives where IGN members were obligated to submit an annual report or disclosure as a condition for continued membership, or where an advocacy target’s compliance with an IGN’s demands and requests was required. In both instances, I reviewed related IGN website content to determine if INGOs were involved in such activities. This form of participation was measured dichotomously (yes/no).  One limitation of the coding scheme deployed in my study is that the degree or intensity  142 of participation across the four functional areas was not covered. For instance, within agenda-setting and mandate implementation, I did not specify whether one or many INGOs were involved and whether this involvement was sustained over a long period of time. In-depth case studies are especially well-suited for process-tracing how participation may vary over time, and would be a good follow-up research project that builds on this initial exploratory effort.    Description of sample  Across the sampled IGNs, a few features are noteworthy: the relatively recent addition of IGNs to the global governance landscape, the geographic concentration of IGNs and a fair degree of isomorphism in the preferred institutional identity for IGNs (DiMaggio and Powell, 1983).63  First, IGNs are a relatively new phenomenon, as the mapping exercise revealed that an overwhelming number of IGNs (75%) were founded after the year 2000 (see Figure 4.1) and fifteen of the 32 IGNs in the study sample were formed after the global financial crisis of 2008. There is also a very distinct thematic trend, where over 40% of IGNs are dedicated to environmental issues and climate change.  The first exclusively human-rights focused IGN, the Investor Alliance for Human Rights launched in 2017.                                                   63 DiMaggio and Powell (1984) define isomorphism as “a constraining process that forces one unit in a population to resemble other units that face similar environmental conditions” (149). Their hypothesized mechanisms for institutional isomorphic change (coercive, mimetic and normative) could offer a useful framework for future research seeking to explain why such isomorphism is observed in the responsible investment domain (DiMaggio and Powell, 1984, 150-154).  143  Figure 4.1. Growth of IGNs, 1970-2017 Second, perhaps due to pragmatic reasons in acting as the collective voice of investors, IGNs are mainly headquartered in global financial capitals, with 70% being located in the United States (New York and Boston area), United Kingdom (London area) and Switzerland (Geneva) (see Figure 4.2). Although members of these IGNs may be more geographically dispersed, participation from organisations outside of Europe, North America and Australia is quite limited, with the exception of dedicated regional networks. It should be noted however, that even in groups like the Asian Corporate Governance Association, the presence of regional subsidiaries of European and North American investors, banks and asset managers somewhat skews the actual number of truly domestic participants.64 In the context of human rights, this level of geographic concentration in IGNs also reveals a considerable distance between groups and affected                                                 64 A study of membership lists using network analysis could be a fruitful avenue for future research that seeks to document intra-network and cross-network politics, gatekeepers and influencers within IGNs. 051015202530351970s 1980s 1990s 2000s 2010s TotalNumber GlobalRegionalTotal 144 communities that are often the subject of an investor’s concerns and the investors themselves. Examples include advocacy by European and North American Investors on land grabs in the Brazilian Amazon Rainforest and adverse impacts for local communities in Brazil (CERES, Press release, 11 November 2016), or the investor engagements with multinational food companies about slave labor in their Ghanaian and Ivory Coast cocoa supply chains (UNPRI, 2016, p. 7).    Figure 4.2. Where are IGNs headquartered? Third, although there is some variation in the how IGNs choose to structure their collective action platform, there appears to be a preference for non-profit status, with 84% of IGNs opting for this model either as a non-profit trade body, a non-profit association or as a sub-project of a pre-existing non-profit IGN or charitable foundation (See Figure 4.3). For-profit companies, informal networks and public-private partnerships are less frequently used to structure investor activism. It should be noted, that MacLeod and Park’s (2011) reference to both informal and formal pathways for INGOs to engage with IGNs is important given prevalence of 10% 3%7%7%3%7%35%28%AustraliaCanadaFranceHong	Kong	South	AfricaSwitzerlandUnited	KingdomUnited	States 145 the “trade-association” model. This model usually restricts participation for non-financial actors, making informal participation pathways for INGOs potentially more salient.  Figure 4.3. IGNs studied by organisation type  4.3 Findings Having identified the universe of cases within the study sample, I now move onto mapping the actual degree to which IGNs recognize and allow for the formal and informal participation of INGOs in transnational shareholder activism in each of these cases.   Recognition  Within the study sample, at the aggregate level there no significant difference between IGNs that recognize civil society organizations (as a proxy for INGOs) as necessary partners in advancing sustainability within global capital markets and those that do not. Fifteen IGNs do 0 1 2 3 4 5 6 7 8 9 10Company (for profit)Informal network (non-profit)Industry association (non-profit)Mult i-stakeholder  organisation (non-profit)Distinctly-goverend project from a non-profit IGN orfoundationPublic-Pr ivate PartnershipGlobal Regional 146 make this recognition compared to seventeen that do not. For example, the Investor Environmental Health Network allocates differentiated roles between actors in the IGN, but nonetheless is inclusive in its approach vis-à-vis civil society. The group describes itself as a “collaborative partnership of investment managers, advised by nongovernmental organizations, concerned about the financial and public health risks associated with corporate toxic chemicals policies” (IEHN, 2018, ¶1). A similarly inclusive self-description is offered by The International Integrated Reporting Council (IIRC), which presents itself as “a global coalition of regulators, investors, companies, standard setters, the accounting profession and NGOs” (IIRC, 2018, ¶1).  Other IGNs do not appear to reference civil society at all, and define themselves in exclusive terms.  For instance, although the Responsible Investment Association of Australasia describes its work as shifting “more capital into sustainable assets and enterprises and shape[ing] responsible financial markets to underpin strong investment returns and a healthier economy, society and environment,” (RIAA, 2015, p. 2) the organization makes no commitment to working with social and environmental partners in delivering this goal. Rather, it views itself as “the peak industry body representing responsible and ethical investors across Australia and New Zealand” (RIAA, 2018, ¶1).  Likewise, The Global Investor Coalition on Climate Change “advances global investor collaboration to improve investor practices, corporate actions and international policy responses to climate change” while leaving the social partners relevant to this goal invisible (GICCC, 2018, ¶3). It would appear that institutional roots may matter in explaining such variations, though they generate noticeable patterns (see Figure 4.4) It is clear IGNs with a strong industry orientation (namely companies, public-private partnerships and industry associations) are far more less likely to recognize civil society as being relevant to their vision, mission or mandate,  147 whereas multi-stakeholder organizations, informal networks and independent projects that arise from non-profit IGNs and philanthropic foundations are more likely to accord this form of recognition to civil society. While these results may not be surprising, they raise a few important implications. First, the results shows that there are two competing visions of how best to advance the global financial sustainability agenda. On the one hand, the systemic importance of institutional investors appears to warrant a focus on this group of actors, to the exclusion of others. On the other hand, because financial sustainability is shaped by a host of environmental and social factors, it seems unlikely segregated approaches will generate the desired goal of aligning capital to the needs of society and avoiding harm to the environment.   Figure 4.4. Recognition of civil society by IGN type Furthermore, the results suggest that the institutional underpinnings of an IGN matter in terms of who is recognized as a vital player in the systemic transformation of capital markets 0% 20% 40% 60% 80% 100%Company (for profit)Informal network (non-profit)Industry association (non-profit)Mult i-stakeholder  organsiation (non-profit)Distinctly governed project from a non-profit IGN or foundationPublic-Pr ivate PartnershipYes No 148 towards sustainability. If the industry-dominant IGN model prevails over multi-stakeholder approaches, the responsible investment movement risks losing out more democratic modes of deliberation that diversity would bring to bear on their strategies, principles and commitments. However, it would be erroneous to assume that the triumph of multi-stakeholder models are ethically superior or neutral, or that they are always capable of representing local communities adequately. Civil society groups who guard their independence fiercely or who may adopt a more combative stance towards the private sector may actively resist their inclusion in IGNs, thereby precluding cooperation and partnership between IGNs and INGOs.   Participation  As explained previously, Avant, Finnemore and Sell (2010) identify four functional areas intrinsic to global governance: agenda-setting, making rules, implementation and enforcement and monitoring and evaluating outcomes.  Examining these four areas within an IGN can offer valuable insights on who participates in these activities and how.   Agenda-setting  For the purpose of this chapter, I examined whether civil society groups have a role to play in setting the responsible investment agenda from a historical perspective. As per the research framework explained previously, I suggest that an IGN’s formation can itself be a signal that an issue has been adopted as subject of transnational advocacy by a set of principled actors. Based on this assumption, the presence of civil society groups in an IGN’s founding story signals their inclusion in agenda-setting within the broader responsible investment space. The data in my sample of IGNs mimics patterns with regard to recognition such that industry-associations,  149 companies and public-private partnerships tend not to involve INGOs in their inception (see Figure 4.5).  The ten multi-stakeholder non-profit organisations are evenly split in terms of INGO participation in agenda-setting. Of those that did include civil society organizations in their formative phase, two have a significant membership base of religious investors (ICCR and ECCR) and three are focused on climate action (CERES, Natural Capital Finance Alliance and the 2 degree Investing Initiative). The multi-stakeholder groups that do not reference civil society groups in an agenda-setting context are more industry-oriented in their mandate. These include Confluence Philanthropy, the IIRC, Climate Bonds Initiative and AfricaSif. Interestingly, while there is some thematic clustering in multi-stakeholder groups that include CSOs in agenda-setting (here in climate action) a similar trend is not evident across those that exclude CSOs.  Figure 4.5. CSO participation in IGN agenda-setting Rule-making Even within IGNs that recognize the potential role of civil society in advancing responsible 0 1 2 3 4 5 6 7 8 9 10CompanyInformal networkMult i-stakeholder  organisation (non-profit)Non-profit Industry associationProject of non-profit or foundationPublic-Pr ivate PartnershipAgenda Setting Yes Agenda Setting No 150 investment and sustainable finance more broadly, this recognition does not necessarily lead to a seat at the table for INGOs. This claim is advanced from the data that evaluates whether governance processes include civil society or whether they are permitted membership.