Multistakeholder as Governance Groups: Observations from Case Studies

Authors: Urs Gasser, Ryan Budish, and Sarah Myers Westº
Berkman Center for Internet & Society at Harvard University

Table of Contents

I. Introduction
II. Case Study Process and Overview
A. Research Process and Methodology
B. Terminology
C. Scope
D. Overview of Case Studies
III. Observations on Structural Elements
A. Purpose and Context
1. Purposes of Governance Groups
2. Cultural and Contextual Factors
3. Types of Legitimacy
B. Formation
C. Operation
1. Decision-Making Procedures
2. Conflict Resolution
3. Interfaces and Coupling Management
4. Knowledge and Memory Management
5. Motivational Issues
6. Communication
D. Outcomes
1. Outputs, Outcomes, and Impact
2. Unintended Consequences
3. Operational Continuity
IV. Critical Success Factors
A. Inclusiveness
1. Formal Inclusiveness: Identifying and Including Stakeholders
2. Voting, Consensus, and the Use of Decision-Making Processes
3. Effect of Alternative Participatory Mechanisms
4. Resource Asymmetries
B. Transparency
C. Accountability
D. Legitimacy
E. Effectiveness
V. Summary of Observations
VI. Conclusions

I. Introduction

This paper synthesizes a set of twelve case studies of real-world governance structures.[1] Together with the underlying case studies, it is the result of a globally coordinated, independent academic research pilot project by the Network of Interdisciplinary Internet & Society Research Centers (NoC). Facilitated by the Berkman Center for Internet & Society at Harvard University, this study examines existing multistakeholder governance groups with the goal of informing the evolution of—and current debate around—the future evolution of the Internet governance ecosystem in light of the NETmundial Principles and Roadmap, discussions at the Internet Governance Forum, and the NETmundial Initiative, as well as other forums, panels, and committees.

Internet governance is an increasingly complex concept that operates at multiple levels and in different dimensions, making it necessary to have a better understanding of both how multistakeholder governance groups operate and how they best achieve their goals. With this need in mind, at a point where the future of Internet governance is being re-envisioned, this project aims to deepen our understanding of the formation, operation, and critical success factors of governance groups (and even challenge conventional thinking) by studying a geographically and topically diverse set of local, national, and international governance models, components, and mechanisms from within and outside of the sphere of Internet governance, with a focus on lessons learned.

The research effort is grounded in a diversity of global perspectives and collaborative research techniques. Adhering to objective and independent academic standards, it aspires to be useful, actionable, and timely for policymakers and stakeholders. More broadly, the Network of Centers seeks to contribute to a more generalized vision and longer-term strategy for academia regarding its roles in research, facilitation and convening, and education in and communication about the Internet age.

[1] The synthesis and each case study can be downloaded at http://ssrn.com/abstract=2549270.

II. Case Study Process and Overview

A. Research Process and Methodology

Responding to the dynamics of the current debate about the future of Internet governance, the objectives of this pilot project have evolved over time as we have gained a better understanding about the interaction between the formation and operation of governance groups and the contextual environments in which such groups operate.

Initially drawing upon the Panel on Global Internet Cooperation and Governance and Governance Mechanisms (“Panel Report”),[2] the research began with the guiding question of whether a set of best practices that built off of the Panel Report could be distilled from a series of select real-world case studies examining various governance groups. In that mode we sought case studies that would help us better understand three components of the Panel Report: (1) how governance groups best match challenges with the organizations, experts, networks, and governing bodies/entities most able to help develop legitimate, effective, and efficient solutions; (2) how to structure the flow of information and knowledge necessary for successful governance; and (3) how different governance groups approach coordination between regional and global (or national and regional) governance networks in order to avoid conflicting directives.

Building upon twelve case studies, this synthesis concludes that there is no single best-fit model for multistakeholder governance groups that can be applied in all instances. Rather, it reveals a range of approaches, mechanisms, and tools available for both the formation and operation of such groups. The analysis demonstrates that whether governance groups meet their objectives depends to a large degree on the careful selection, deployment, and management of suitable instruments from this “toolbox.” As governance groups pass through different phases of operation, conveners and facilitators must remain alert to changes in circumstances that necessitate adjustments to the approaches, mechanisms, and tools that they deploy in order to address evolving challenges from inside and from outside. The case study series provides insights into how those instruments can be deployed and adjusted over time within such groups, and highlights how important contextual factors they interact with may be successfully managed within given resource restraints.

The case studies come from the global academic institutions that constitute the Network of Centers. After a series of planning meetings and learning calls among the participating Centers, the research team met in May 2014 to collaboratively identify and select case studies based on a set of criteria emerging from the current debates about the future of Internet governance, as described above. The case studies (described in greater detail below) represent a wide variety of initiatives, including public and private sector efforts, operating at national and international levels in regions all over the world, addressing issues with varying levels of technical complexity.[3] The selected case studies also cover a wide range of activities, including drafting legislation, developing policy in the absence of formal regulatory authority, defining and exploring the scope of a problem, and building connections among stakeholders. While some of the case studies come from areas related to Internet governance, others come from outside the Internet governance ecosystem. Although they differ in many respects, the governance groups documented in the case studies share a common commitment to using multistakeholder processes to solve complex issues.

Each of the case study authors conducted desk research and interviews in order to develop the case studies. In several cases, the case studies reflect direct input from participants and leaders in the group’s operation. Once drafted, the case studies underwent several rounds of peer review and served as key inputs for a meeting of the Network of Centers on October 1-2, 2014. At the meeting, the case studies were discussed and key findings were identified and synthesized.[4] That discussion formed the basis of this synthesis paper.

[2] Panel on Global Internet Cooperation and Governance Mechanisms, Panel Report, http://internetgovernancepanel.org/panel-report.
[3] The process is documented here: http://networkofcenters.net/events.
[4] The notes from the meeting are available here: http://networkofcenters.net/event/evolution-internet-governance-ecosystem.

B. Terminology

At the outset of this research, we conducted a review of existing literature to inform our analysis of the cases. In particular, our review focused on the rich literature on multistakeholder governance, examining the body of scholarship that provides frameworks for, or definitions and critiques of, the term. In its most basic interpretation, multistakeholder governance implies the incorporation of representatives from multiple groups in discussions and decision making.[5] Multistakeholder partnerships can, in their idealized form, facilitate global governance by bringing in a range of resources and competencies to address common problems.[6]

In the context of Internet governance the multistakeholder approach has taken on greater salience since the World Summit on the Information Society (WSIS). Alongside the rise in prominence of non-state actors in global affairs, the UN resolution providing for WSIS provided an initial legal basis for multistakeholder participation in Internet governance discussions. As initially used, the term “multistakeholder” was associated with private-sector leadership, but its definition was somewhat vague. By the Tunis Agenda its meaning became more consolidated to mean incorporation of governments, business entities, civil society and intergovernmental organizations.[7] The Internet Governance Forum (IGF) further expanded the scope of stakeholder representation to include the academic and technical communities.

Though the literature reflects general agreement on core principles, it also suggests some divergence as to the processes that constitute multistakeholder governance, particularly across different disciplinary perspectives. Governance approaches come in many shapes and sizes: in addition to multistakeholder, governance processes may alternately be described as distributed, decentralized, or collaborative.[8] They also may differ depending on the governance context: governance in the economic domain may look quite different and operate at different scalar levels than governance in the environmental domain. Thus, while the most basic definition of multistakeholder governance is generally agreed upon at the core, understandings of governance practices are shaped by a variety of interpretations and perspectives at the periphery.[9]

Given this lack of consensus, we elected not to take a definitional approach to the work; we did not try to match our case studies onto predefined templates of how a governance group should look, nor did we focus only on those elements that aligned with a preselected definition. Instead, we sought to contribute to our shared understanding by taking a phenomenological approach and focusing on the functions of the groups as well as the elements critical to those functions. One benefit of this bottom-up approach, which is particularly relevant in the context of the global research design of this study, is that it accommodates the varying translations of the term “multistakeholder” into different languages, in which it frequently takes on different nuances and connotations.

[5] Minu Hemmati. Multistakeholder Processes for Governance and Sustainability: Beyond Deadlock and Conflict (Sterling, VA: Earthscan Publications, 2002).
[6] Thomas Hale and David Held. The Handbook of Transnational Governance: Institutions and Innovations. (Cambridge: Polity Books, 2011).
[7] Markus Kummer, “Multistakeholder Cooperation: Reflections on the Emergence of a New Phraseology in International Cooperation” The Internet Society, 2013, http://www.internetsociety.org/blog/2013/05/multistakeholder-cooperation-reflections-emergence-new-phraseology-international.
[8] For examples, see:
Multistakeholder governance: Laura DeNardis and Mark Raymond, “Thinking Clearly About Multistakeholder Internet Governance” (November 14, 2013), http://ssrn.com/abstract=2354377; Minu Hemmati, Multistakeholder Processes for Governance and Sustainability: Beyond Deadlock and Conflict (Sterling, VA: Earthscan Publications, 2002); Carlos Afonso, Virgilio Almeida, and Demi Getschko, “The Origin and Evolution of Multistakeholder Models,” IEEE Internet Computing 19, 1 (January 2015), 74-79.
Decentralized governance: G. Shabbir Cheema and Dennis A. Rondinelli “From Government Decentralization to Decentralized Governance” in G. Shabbir Cheema and Dennis A. Rondinelli, eds., Decentralizing Governance: Emerging Concepts and Practices (Brookings Institution Press: Washington, DC., 2007).
Distributed governance: Frederick M. Abbott, “Distributed Governance at the WTO-WIPO: An Evolving Model for Open-Architecture Integrated Governance,” Journal of International Economic Law vol. 3(1) (2000): 63-81; Jon Marshall, “Negri, Hardt, distributed governance and open source software,” PORTAL Journal of Multidisciplinary International Studies vol. 3, No. 1 (January 2006), http://epress.lib.uts.edu.au/journals/index.php/portal/article/viewFile/122/84; Paquet, Gilles, “The New Governance, Subsidiarity and the Strategic State” OECD Forum for Future Conference, Governance in the 21st Century, (Spring, 2000), http://www.gouvernance.ca/publications/00-60.pdf.
Collaborative governance: Chris Ansell, and Alison Gash, “Collaborative Governance in Theory and Practice,” Journal of Public Administration Research and Theory vol. 18(4) (2008): 543-571; John Donahue, “On Collaborative Governance,” Corporate Social Responsibility Initiative Working Papers, No. 2 (March 2004).
[9] See Roxana Radu, Jean-Marie Chenou, and Rolf H. Weber, eds., The Evolution of Global Internet Governance (Zurich: Schulthess 2013) (including several articles describing different approaches to and types of multistakeholder governance).

C. Scope

This project takes a networked approach to academic research. By activating a global network of interdisciplinary research centers, we were able to conduct in parallel research on twelve different case studies, and then collaboratively consider lessons from those case studies. Although the work was conducted in a highly compressed timeframe, we endeavored to demonstrate that such an interdisciplinary and global approach can offer nuanced, helpful, and timely insights to the stakeholders involved in Internet governance debates or concerned with the implementation of governance groups more broadly.

Of course, such an approach is not without methodological limits. First, our conclusions are bounded by the samples chosen. We chose only cases that were multistakeholder in whole or in part, and did not study groups with entirely different structures. Nor did we seek examples of failed governance groups; although not all of the groups analyzed in our research achieved all of their intended goals, they all were viable in their operations and sustainable throughout their existence. Accordingly, this research does not endeavor to identify factors that may trigger failure and instead focuses on factors that seem to enable success. Second, as with most case study-centered research, this project does not assert to be a statistically representative sample of models, regions, or experiences. Despite these limitations, we believe this synthesis and the accompanying case studies provide timely insights to the current debate over Internet governance, while also laying a foundation for future research that will be able to more fully explore the concepts, considerations, and questions raised in this work.

D. Overview of Case Studies

This synthesis paper and its foundational case studies are not standalone documents, but are intended to be read as interacting elements. For the convenience of readers, below we provide short abstracts for each case study; we recommend reading the relevant case studies in their entirety where further information is necessary.

Aviation – The Worldwide Slot Guidelines – Giovanni Sartor and Hanna Schebesta, European University Institute (“Aviation Slotting”):[10] This case study explores the Worldwide Slot Guidelines (WSG), which represent a set of agreements made under the auspices of the International Air Transport Association (IATA) in order to allocate airport capacity. The right to use airport capacity for the purpose of takeoff and landing operations at airports is commonly called slots, with each airport having a finite amount of airport capacity. Thus, the slot allocation process is one of resource management: the WSG were developed as a global industry standard to address a key constraint in the aviation domain. The actual process of slot allocation is complex and involves organizations and stakeholders at multiple levels, all falling under the umbrella of IATA, a trade association representing 240 airlines. It designates particularly congested airports, creates roles of airport coordinators, and establishes management principles for slot allocation. Because they manage a valuable economic resource, the transparency and independence of coordinators is of particular concern: this has resulted in the institutionalization of a negotiating process to ensure accountability among key stakeholders. Through an examination of the history and key points of change in the process, the WSG case presents an example of private sector coordinated resource management.

[10] The Aviation Slotting case study is available for comment at https://publixphere.net/i/noc/page/IG_Case_Study_Aviation_Slotting

The Evolution of Governance Structure in Cryptocurrencies and the Emergence of Code-Based Arbitration in Bitcoin – Markus Oermann and Nils Tollner, Hans-Bredow-Institute (“Bitcoin”):[11] This case study describes the process by which Bitcoin revised its core code to accommodate a new feature called “multi-signature transactions.” Bitcoin is a cryptocurrency, which was introduced in 2009 and has spread rapidly since then. It is based on open-source peer-to-peer software that that leverages a network of user accounts (wallets) set up on peripheral sites in which the units of account (Bitcoins) may be stored after they are produced and transmitted. It is a goal of cryptocurrencies in general to operate without a central agent, which makes it complicated to resolve disputes. To address this, Bitcoin added multi-signature transactions. This case study examines how that change was made. In particular the case study explores how an open source community is able to maintain a stable codebase that can serve as a basis for an entire form of currency, while still making necessary changes. The case shows that the decision-making processes regarding transformations of Bitcoin’s governance structure are not as transparent as one might expect given the cryptocurrency’s commitment to open source. However, major changes to the code are publicly discussed in the Bitcoin developers’ community and the authors could not identify a single case of a decision in which the core development team deviated from the consensus of the community.

[11] The Bitcoin case study is available for comment at https://publixphere.net/i/noc/page/IG_Case_Study_Bitcoin_and_Autonomous_Systems.

Creative Commons (CC) – Federico Morando, Nexa Center for Internet & Society at Politecnico di Torino (“Creative Commons”):[12] This case study examines the formal structure of Creative Commons (CC) as an organization stewarding the development of the most well-known and widespread set of standard public copyright licenses. In particular, the case study explores the interaction of CC with its own community and with a broader set of international stakeholders during the drafting of the fourth version of the Creative Commons Public Licenses, which took place between mid-2011 (with public discussion starting in November 2011) and November 2013. This case study highlights how a transnational organization can operate outside of the scope of formal regulatory authority, but with the purpose of creating an output (CC licenses) that have legal weight. In such an environment credibility and legitimacy are paramount, necessitating opportunities for public consultation, transparency, and even admitting mistakes.

[12] The Creative Commons case study is available for comment at https://publixphere.net/i/noc/page/IG_Case_Study_Creative_Commons.

Fighting Spam the Multistakeholder Way: A Case Study on the Port 25/TCP Management in the Brazilian Internet – Ronaldo Lemos, Carlos Affonso de Souza, Fabro Steibel, Juliana Nolasco, Institute for Technology and Society at Rio de Janeiro State University (“CT-Spam”):[13] This case study explores how CGI.br, the multistakeholder Brazilian Internet Steering Committee, addressed through a collaborative decision making and educational process the difficult issue of spam propagation. Although the technical solution to the spam problem in Brazil was relatively clear, convincing stakeholders to adopt the solution was challenging. Telecommunications companies and ISPs initially resisted this recommendation out of concern around the costs of switching and the challenges of communicating the change to end users. This case study examines how a multistakeholder process involving telecommunications companies, ISPs, consumer rights associations, and government ministries and agencies was able to overcome these obstacles while simultaneously respecting consumer rights, freedom of speech, and commercial competition. CGI.br’s Anti-Spam Working Commission (CT-Spam Commission) highlights how collaborative governance can be applied in an iterative and educational fashion. Addressing the spam issue would require buy-in and cooperation from a variety of parties, and by engaging all stakeholders, the CT-Spam Commission was able to identify the concerns of stakeholders and then develop a variety of educational materials, technical reports, and policy changes in order to address those concerns. By developing the policy in such a fashion, the CT-Spam Commission was able to gain the support of the telecommunications companies and ISPs without regulatory oversight. Ultimately, with the buy-in of key stakeholders, implementation of the Port 25/TCP recommendation in 2013 led to a dramatic decrease in spam in Brazil.

[13] The CT-Spam case study is available for comment at https://publixphere.net/i/noc/page/IG_Case_Study_Fighting_Spam_the_Multistakeholder_Way.

Enquete-Kommission Internet und digitale Gesellschaft (Enquete Commission on Internet and Digital Society) – Kirsten Gollatz, Sarah Herweg, Jeanette Hofmann, Humboldt Institute for Internet and Society (“EIDG”):[14] This case study explores the Enquete Commission on Internet and Digital Society (EIDG Commission), which is a parliamentary inquiry body of the German Bundestag that conducted its work from May 2010 until April 2013. Enquete Commissions are special bodies of the German Bundestag, which form an interface between policy-makers and academic and professional experts to examine through discussion forums and working groups broad and complex matters of society that cannot be sufficiently dealt with within the regular parliamentary framework. The Commission was introduced in response to rising awareness regarding the challenges for politics and society that digitization brings along, which became apparent in a controversial draft bill before the parliament that called for the mandatory blocking of websites with child pornography-related content. The Commission consisted of 34 individuals, 17 democratically-elected members (members of parliament) and 17 experts and practitioners representing industry, trade unions, civil society and academia. Furthermore, the public was invited to participate in the Commission’s consultations “to a special degree,” which included efforts at transparency and an online engagement platform. It is too early to assess the policy impact of the Commission’s work, as it is still unclear to what extent the new parliament will include the Commission’s policy recommendations in its legislative initiatives. However, as a mechanism for bringing together heterogeneous sources of knowledge and expertise for the purpose of enlightening decision-making processes, the general model of an Enquete Commission can be regarded as an effective and legitimate mechanism for parliaments and possibly other deliberative bodies. Going forward, finding the appropriate implementation mechanisms that will help inform evidence-based policy making and provide justification for spending further resources should be taken into account in the process design.

[14] The EIDG case study is available for comment at https://publixphere.net/i/noc/page/IG_Case_Study_EnqueteKommission_Internet_und_digitale_Gesellschaft.

Multistakeholder Approaches to Water Resource Management in the White Volta River Basin – Rebekah Heacock Jones, Ryan Budish, et al, Berkman Center for Internet & Society (“White Volta”):[15] This case study examines the deployment of the decentralized integrated water resource management (IWRM) model for the management of water resources at both the local and transboundary levels of the White Volta River Basin. The Volta River Basin spans across six countries—Benin, Burkina Faso, Cote d’Ivoire, Ghana, Mali, and Togo. One of the Basin’s four main sub-basins is the White Volta in Northern Ghana. The case study investigates and compares governance at the sub-basin level through White Volta Basin Board (WVBB) and at the Basin level through the Project for Improving Water Governance in the Volta River Basin (PAGEV). It also discusses the development of IWRM in water governance more broadly. The case study concludes that PAGEV was more successful in implementing IWRM principles than the WVBB, but questions remain as to their decision-making processes and the full extent of stakeholder participation.

[15] The White Volta case study is available for comment at https://publixphere.net/i/noc/page/IG_Case_Study_Water_Management_in_Northern_Ghana.

Multistakeholder Governance and Nodal Authority – Understanding Internet Exchange Points – Ben Wagner, European University Institute and Patricia Mindus, Philosophy Department, Uppsala University (“IXP”):[16] This case study considers Internet exchange points (IXPs) as an example of governance processes in action. Internet exchange points are the points of connection between different Internet networks, which enable different networks to exchange traffic at a shared facility without cost to either party through a process known as “peering”. Three different IXP governance models representing large and influential IXPs are compared: the DE-CIX in Frankfurt, CAIX in Cairo, and KIXP in Nairobi. DE-CIX, the largest IXP in the world, is a subsidiary of the German Internet trade association eco, and is thus “owned” by the Internet industry in Germany. Though well functioning, this has meant that key stakeholder groups such as civil society, and the academic and technical communities are excluded from participating in discussions over policy decisions. In contrast, the Cairo Internet Exchange Point (CAIX) is run by a public authority, the Egyptian Ministry of Communications and Information Technology. Though it is governed by a broad set of stakeholders including private sector, government, and civil society representatives, its decision-making processes are somewhat opaque. Lastly, KIXP was founded by a Kenyan network engineer and is governed by a local trade association. While set up with multistakeholder coordination under the leadership of the private sector, its day-to-day operations and governance fall under private sector control. By tracing out the plurality of models used for IXP governance and comparing the processes of developing peering relationships, this case provides unique lessons for the governance process, particularly surrounding trade-offs between inclusiveness and effectiveness.

[16] The IXP case study is available for comment at https://publixphere.net/i/noc/page/IG_Case_Study_Multistakeholder_Governance_and_Nodal_Authority_Understanding_Internet_Exchange_Points.

Towards a Cyber Security Policy Model – Israel National Cyber Bureau Case Study – Daniel Benoliel, Haifa Center of Law and Technology (HCLT), University of Haifa Faculty of Law (“INCB”):[17] This case study examines the elements of the creation of a cyber security policy model at a national level. It uses Israel’s recently established National Cyber Bureau (INCB) cyber command funneled by its national cyber policy as a case in point. In so doing the brief offers a cross-section comparison between leading cyber security national policies of the United States, United Kingdom, Canada, Japan and the Netherlands. It further introduces comparable policies including the balancing of cyber security with civil liberties, cyber crime policy, adherence to international law and international humanitarian law, forms of regulation (technological standards, legislation, courts, markets or norms) and prevalent forms of cooperation (intra-governmental, regional, public-private platform (PPP) and inter-governmental cooperation).

[17] The INCB case study is available for comment at https://publixphere.net/i/noc/page/IG_Case_Study_Towards_a_Cyber_Security_Policy_Model_Israel_National_Cyber_Bureau_INCB.

A Bill of Rights for the Brazilian Internet (“Marco Civil”) – A Multistakeholder Policymaking Process – Ronaldo Lemos, Carlos Affonso de Souza, Fabro Steibel, Juliana Nolasco. Institute for Technology and Society at Rio de Janeiro State University (“Marco Civil”):[18] This case study explores the events that led to the passing of the “Marco Civil da Internet” legislation, which became law in Brazil in April 2014. The first of its kind anywhere in the world, Marco Civil is a landmark “Bill of Rights” for the Brazilian Internet users. The legislation is unique both due to its substance and the process by which it was created. From a substantive standpoint, Marco Civil applies to the Internet key principles such as freedom of expression, net neutrality, and due process. But perhaps more important than the substance is the way in which the bill was drafted and enacted. This case study explores the unique features of that drafting process. Marco Civil was conceptualized in 2007 and, over the course of many years, was drafted using an open multistakeholder process through which members of the public, government, global and local Internet companies, civil society, and others engaged in negotiations over the legislation’s text, largely mediated through online platforms. Although the law was ultimately adopted, the process was not without its challenges. For example, attracting early contributors and participants was difficult, as was debating contentious topics openly among the stakeholders, who frequently disagreed. After many delays, Marco Civil was eventually passed. This case study will explore the factors that enabled a complex piece of legislation to be drafted with input and consultation from a distributed and diverse collection of stakeholders.

[18] The Marco Civil case study is available for comment at https://publixphere.net/i/noc/page/IG_Case_Study_A_Bill_of_Rights_for_the_Brazilian_Internet.

The Global Multistakeholder Meeting on the Future of Internet Governance (NETmundial) – Marília Maciel, Nicolo Zingales, and Daniel Fink, Center for Technology & Society at FGV School of Law, Rio de Janeiro (“NETmundial”):[19] This case study examines NETmundial, the Global Multistakeholder Meeting on the Future of Internet Governance, which was held in Sao Paulo, Brazil on April 23-24, 2014. The meeting was convened by 1net, a coalition of stakeholder groups involved in Internet governance discussions, in partnership with the Brazilian Internet Steering Committee (CGI.br), in response to revelations of mass surveillance of communications by the United States. It sought to develop a set of universally-acceptable Internet governance principles as well as a way forward for the evolution of the Internet governance system, which together could serve as a framework for the governance and use of the Internet. It convened 930 participants from 110 different countries, representing civil society, the private sector, academia, the technical community, governments and intergovernmental organizations, as well as over 1000 remote participants from 23 countries around the globe. It also employed a content contribution platform that sought to crowdsource inputs from stakeholders for the production of the outcome document. The meeting served as a demonstration of the multistakeholder process in action: in the production of the outcome document, stakeholders with a diverse range of backgrounds and interests collectively negotiated the inclusion or exclusion of highly-sensitive and complex issues. While the process of achieving rough consensus involved sometimes messy debates and there were procedural imperfections, the case is informative for its structured production of bottom-up multistakeholder outcomes.

[19] The NETmundial case study is available for comment at https://publixphere.net/i/noc/page/IG_Case_Study_NETMundial.

Swiss ComCom FTTH Roundtable Case Study – Urs Gasser, Ryan Budish, David R. O’Brien, et al., Berkman Center for Internet & Society (“Swiss ComCom”):[20] This case study explores the dimensions of how the Swiss government—in particular, the Swiss Federal Communications Commission (ComCom) and the Swiss Federal Office of Communications (OFCOM)—used a multistakeholder process to organize private sector firms to begin deploying in a coordinated fashion a fiber optic network connected to every home in Switzerland. Although fiber optic connections were available in some areas of Switzerland, telecommunications and cable companies in many regions of the country had not yet upgraded older network infrastructure to fiber or connected it to homes. The large capital investments required, competing interests among stakeholders, and the lack of coordinated efforts to serve common interests were all contributing factors. The Swiss government was eager to see the national infrastructure upgraded, but lacked the formal authority to regulate the fiber optic networks under its communications laws. In lieu of regulation, the Swiss government convened a series of voluntary roundtable meetings between 2008-2012 to achieve a consensus-based solution. The Roundtables, and related technical working groups, attempted to bring together the critical stakeholders and surface the key issues that needed to be resolved. Using Swiss public records, reports, and personal accounts of those directly involved, this case study examines the roles and approaches used in the Roundtables and evaluates the process used as a model of ad hoc, problem solving and distributed governance.

[20] The Swiss ComCom case study is available for comment at https://publixphere.net/i/noc/page/IG_Case_Study_Swiss_ComCom_FTTH_Roundtable.

Turkish Internet Improvement Board – Leyla Keser, Mehmet Bedii Kaya, IT Law Institute, Istanbul Bilgi University (“TIIB”):[21] This case study focuses on a multistakeholder working group that the Turkish Internet Improvement Board created in order to generate innovative, bottom-up approaches for fixing the Internet law in Turkey. In 2007, Turkey adopted its first comprehensive Internet content regulation legislation: Law 5651, the “Regulation of Publications on the Internet and Combating Crimes Committed by Means of Such Publications.” Since its adoption, Law 5651 has come under scrutiny for enabling overly broad restrictions on web content, its establishment of criminal liability for hosting providers who fail to adequately restrict content, the threats it posed to Internet-related innovation in the country, the cybersecurity vulnerabilities that resulted from the widespread use of circumvention tools by Turkish Internet users in response to content restrictions under the law, and related data protection concerns. The law also called for the creation of the Internet Improvement Board. In 2011, the Board created the “5651 Working Group,” comprised of seven representatives from government, civil society, academia, and the private sector, to explore possible solutions to these problems. This case study examines how the working group collaborated with a broader group of stakeholders to develop a resolution, delivered to the Ministry of Transportation, Maritime Affairs, and Communication, proposing amendments to Law 5651.

[21] The TIIB case study is available for comment at https://publixphere.net/i/noc/page/IG_Case_Study_Turkish_Internet_Improvement_Board.

III. Observations on Structural Elements

In this section, we explore the key structural elements of these case studies, trace through some of the nuanced differences that distinguish them, and identify areas they have in common. We conduct this evaluation across several dimensions, which emerged "bottom-up" from the cross-sectional analysis and discussion of the cases (as further described in the Methodology section above). First, we examine some of the many reasons multistakeholder groups are formed. Second, we describe the process of formation, and the ways in which the convening process and initial establishment of the groups can have downstream effects on their subsequent operation and outcomes. Third, we analyze some of the key aspects of their operation, including how they approach decision-making, conflict resolution, and internal and external communications. Lastly, we discuss their outcomes (both intended and unintended), how they sought to measure their success, and their continuing impacts after their formal operations concluded.

A. Purpose and Context

1. Purposes of Governance Groups

The case studies reveal diverse motivations drove the formation of governance groups, ranging from supplementing existing institutions and drafting legislation to more open-ended approaches that seek to define or inform the scope of an issue. In several cases, governance groups were created to help advise or inform existing political institutions in order to help supplement a perceived deficiency in the process. For example, in the Marco Civil, TIIB, and EIDG case studies, multistakeholder bodies were created as a response to controversial legislation. The groups in each of those cases were intended to help bring specialized knowledge to bear on the legislative process with the ultimate aim of informing legislative bodies and drafting legislation.

Governance groups can play an important executive role in carrying out policies. For example, in the INCB case study, multiple actors joined together in the execution of cybersecurity policy. Stakeholder involvement at the implementation stage provides an opportunity for the actors that may be affected by a policy to shape its execution.

In contrast, in other case studies the governance groups were intended not merely to supplement or support, but to replace existing institutions or fill in gaps between institutional authorities. For example, the Bitcoin, NETmundial, and Creative Commons case studies all describe governance groups that operate outside of the domain of existing institutions. In fact, as the Bitcoin case study describes, the primary goal of the Bitcoin developers was to create an alternative currency that was specifically not dependent on existing institutions.

Within the case studies there are other, less explicit but nonetheless important purposes driving the formation of governance groups. For instance, we see examples of more open-ended purposes such as building networks and relationships between partners or advancing knowledge about an issue. Although not the primary purpose of the group, the Creative Commons case study describes the important goal of helping to build a global network among knowledgeable, interconnected individuals. Similarly, governance groups can allow members to develop common shared definitions and understandings, overcoming language barriers that frequently develop among members from different domains. This may help facilitate greater understanding between partners, reduce the potential for conflict, and smooth the way forward in tackling complex issues.

2. Cultural and Contextual Factors

A key thread throughout this paper is the importance of cultural and contextual factors in shaping the functioning and the outcome of governance groups. Of particular importance seem to be: preexisting relationships between stakeholders, the relationship between the governance group and governmental institution, the allocation of resources, and geopolitical factors.

How the group is situated within existing institutions or normative structures can have downstream effects on the relationships between stakeholders and the results inside the governance group. For example, in the EIDG case study, the Commission existed within a parliamentary framework and was made up of both members of parliament and outside experts. The members of parliament knew each other as repeat players and partners within the political system, and most expert representatives were selected on the basis of their political party affiliation. The fact that the participants’ relationships were largely defined by politics ultimately made the Commission itself more political.

The case studies demonstrate a range of different relationships between government and governance groups. This relationship may differ depending on the historical and political context of the issues involved. Because cybersecurity directly implicates national security interests, the INCB in Israel illustrated much stronger governmental control of both the structure of the process and eventual outcomes. Governance groups acting in areas in which government has not traditionally held a strong position may include less government involvement (NETmundial, for example).

The allocation of resources (including time) inhibited full participation of some stakeholders in several case studies. For example, at NETmundial, a lack of clarity regarding the availability of funds until late in the process meant that some potential participants were deterred from attending the meeting in person. Similarly, a lack of both time and funding at NETmundial limited the ability of the organizers to offer multilingual support during the online consultation period, limiting participation to only those who were proficient in English.

Geopolitical factors and crises influenced the governance groups in a number of the case studies. For example, the Snowden disclosures prompted the formation of NETmundial and added new urgency to the passage of the Marco Civil. And, as noted previously, several of the governance groups were prompted by controversial legislation.

3. Types of Legitimacy

Legitimacy is an important aspect of governance groups because it helps explain why a group’s decisions might be adopted and followed even in the absence of formal enforcement mechanisms. Some governance groups utilize traditional mechanisms of legitimation, including legal frameworks, checks and balances, and voting processes. However, other groups choose to (or must) seek their sources of legitimacy elsewhere.

Interestingly, most governance groups in the case studies appear to rely upon multiple and overlapping sources of legitimacy, and move between them depending on the circumstances and context. For example, the EIDG not only used traditional sources of legitimacy derived from the German constitutional framework providing for Enquete commissions, but also generated expertise-based and procedural legitimacy through its knowledge creation process, which included open meetings and opportunities for public feedback. It is important to note, however, that governmental-based legitimacy for a governance group is derived from the legitimacy of the government itself; in other words, a government that is not viewed as legitimate in its exercise of power, cannot confer legitimacy upon a governance group merely by association. Moreover, this sense of legitimacy is entirely perspective-dependent, as different groups may have different assessments of a government’s legitimacy and thus different assessments of an affiliated governance group’s legitimacy.

Legitimacy can manifest through input legitimacy or output legitimacy. Input legitimacy refers to perceptions of a fair process: for example, the involvement of a wide range of stakeholders on equal footing, and with a shared understanding of the procedural rules, aided the input legitimacy of the CT-Spam group and the NETmundial conference through the inclusion of a wider range of viewpoints. By contrast, other cases illustrated the role of output legitimacy, in which the quality of the output of a process ensures it is seen as legitimate among stakeholders. The coordinated and efficient deployment of FTTH throughout Switzerland has meant that the ComCom roundtables were largely seen as a success, despite the fact that some stakeholders claimed the process disenfranchised them.

Lastly, context proved important in determining the centrality of legitimacy. We found that legitimacy becomes most important when conflicts arise during the multistakeholder process. In these instances, the groups had to assert stronger forms of legitimacy. Where there was broad approval of the process and its objectives, “lighter” forms of legitimacy tended to be more accepted.

B. Formation

The architecture of the groups examined in the case studies varied notably, despite the fact that they all shared a multistakeholder approach. Each group interpreted and implemented this approach in highly contextualized ways. These choices have consequences for the decision-making process and eventual outcomes.

The way in which multistakeholder groups are convened can have important effects on the ultimate outcome of the process. For example, in the TIIB case, the Turkish Internet Improvement Board was authorized by law to create a multistakeholder group in order to research issues relating to the widely criticized law and propose amendments to it. Though in many ways the group was intended to form a critique of the law, its formation necessarily predicated close involvement of the government, which may have challenged its legitimacy. The CT-Spam group was formed in the wake of a failure to convince key stakeholders of the importance of an issue, which allowed the issue to become a crisis. Because of this, a multistakeholder approach proved to be the only way to develop a sustainable solution to a critical infrastructure problem Brazil was facing.

The composition of the group is a key element of the formation process. The case studies illustrate a wide variety of approaches: some sought to obtain as broad-based participation as possible through voluntary processes, while others were more selective in choosing representatives of stakeholder groups to participate in more intimate discussions. On the far end of the spectrum, the NETmundial process incorporated over 930 attendees and over 1,000 participants who joined the meeting through remote hubs. On the other end, ComCom selected a small group of members to participate in its round tables, seeking a balance between enough participants so that decisions could be broadly accepted but not so many participants that the process would take too long. In each case, the breadth of participation was calibrated to fit the issue at hand.

Inclusion criteria influence who may be considered a potential participant or stakeholder. ComCom limited participation in the roundtables to only those groups that had already begun the process of making investments in FTTH, and furthermore only allowed CEOs to serve as representatives of these groups. Similarly, participation in the airport slotting guidelines development process is inherently limited to the small group that can own major airports or airlines, and only the delegates from these stakeholders with authority to make decisions on their behalf are allowed to participate in the slotting process at the bi-annual Slot Conferences. These restrictive inclusion criteria ensured participants in these highly technical matters had the necessary expertise and ability to make informed, binding decisions. However, they limited the diversity of voices in the process, which may have inhibited the creativity of outcomes. Additionally, it is worth noting that the transparency of the inclusion criteria and the identity of the individual or organization setting such criteria may also affect the perceived legitimacy of the group as a whole.

While all the case studies examine groups that can be called multistakeholder, the groups demonstrate a diverse range of structures and internal architectures. Some groups created topical divisions, segmenting off more technical issues into working groups that informed broad-based plenary processes. For example, the EIDG created twelve working groups to produce reports on a variety of topics. These reports were later discussed and voted on in plenary sessions. Others created participatory divisions. In addition to general participation at NETmundial, organizers created several committees that helped to structure and operate the event, including the Executive Multistakeholder Committee, which shaped the ultimate outcome document. At the meeting itself, participants were categorized by stakeholder group, and through rotation of microphones each group was given equal voice regardless of the total number of participants it had. Finally, some groups had no hierarchies whatsoever. During the drafting stage for the Marco Civil legislation, for instance, there was no clear established hierarchical structure for participation; rather, an open platform channeled public conversation in a productive manner, enabling the legislation to be drafted in a bottom-up fashion.

C. Operation

The case studies describe not just the structure of the multistakeholder organizations, but also their operation in addressing a variety of complex issues. Although all of the case studies reflect collaborative and distributed attempts to address a specific problem or set of problems, the mode in which each group operated and functioned in order to address that problem differed. This section describes the operational systems and tools that each group used in order to reach the agreement necessary to create its outputs and address the issue at hand.

The operational process itself has several components. Because every group at least implicitly seeks to identify and implement solutions to problems, decision-making processes are a critical part of group operation. Decision making, however, is only one of several operational components. Additionally, some groups have systems for addressing internal group conflict, pathways for interacting with other groups and organizations, processes for building and maintaining institutional memory, approaches for motivating participants, and systems for enabling both internal and external communication.

1. Decision-Making Procedures

The case studies exemplify a spectrum of decision-making procedures, ranging from consensus-oriented deliberative approaches to a single point of unilateral control. This spectrum exists not just across case studies but within them, indicating that most groups utilize a range of procedures that are matched to the particular stage of the process and the context.

The strongest example of consensus-oriented decision making among the case studies is NETmundial. Prior to the meeting in Brazil, participants submitted public comments. An Executive Multistakeholder Committee (EMC) was responsible for synthesizing these comments into the draft outcome document. During the drafting process, the EMC required consensus on every single paragraph. Following public deliberation and input at the plenary sessions, the EMC further revised the document into the final outcome document. Although consensus was strictly adhered to, the decision-making process was limited to the relatively small EMC.

Other case studies describe groups that begin with consensus procedures but switch to other approaches as necessary. The EIDG and the TIIB are examples of this mixed use of procedure, using consensus on the less controversial matters and voting on the rest. In the case of the EIDG, this meant that individual working groups often reached consensus on the factual assessment portions of their reports. For the recommendations, however, there was a greater chance of disagreement, particularly on current political issues. These conflicts would be escalated out of the working group to the EIDG plenary, which would use majority voting to decide what text would be included in the working group report. At times, voting would proceed sentence-by-sentence through the report. Similarly, the Creative Commons Corporation tries to achieve rough consensus within the community. Should that not work, however, the Corporation reserves the right to make decisions on its own.

2. Conflict Resolution

As described above, at the terminal point of making decisions about outcomes, groups can rely on the decision-making procedures described above to resolve conflicts. However, sometimes conflicts between stakeholders emerge at points in the process where there is no final decision to be made or where voting or other outcome-oriented responses would be inappropriate. One approach for helping to resolve these kinds of non-terminal conflicts is highlighted in the TIIB case, where the Board used a “cooling off” period to defuse tensions in conflicts where no solution was readily apparent. Such an approach does not in and of itself resolve the underlying conflict, but it provides both space and time so that a resolution might be easier to identify or more palatable to the parties.

3. Interfaces and Coupling Management

The groups we studied operate within a network of existing institutions. These groups often have formal and informal mechanisms for interacting with those existing institutions, particularly government bodies. The EIDG and TIIB both offer examples of formal interfaces between the groups and government. In both cases, the groups were designed to help improve legislative outcomes, so it was important to have established, formal pathways between the groups and the government. The primary interface mechanism in these cases was having governmental stakeholders play active roles within the groups, participating on equal footing with other stakeholders. Interestingly, neither group had a formal mechanism for turning proposals into legislation, but the active engagement of governmental actors at least created a pathway to ease the creation and introduction of such legislation.

In other cases the interfaces between groups and government are more informal. In the Swiss ComCom Roundtables, for example, active governmental involvement was limited to convening the Roundtables and the working groups. The primary governmental interface was the threat of future regulation in the event that stakeholders failed to reach a suitable agreement. In this case, the parties arguably participated within the Roundtable process out of a desire to avoid a more formal interface with government; the lack of a formal interface represented a form of interaction in and of itself, albeit a more passive one. Government is not the only institution with which groups interact in the case studies. For example, the planning and operation of NETmundial required interfaces with the so-called I * organizations (e.g., ICANN, IETF, ISOC, IAB). Representatives of the I * organizations played a significant role within NETmundial, helping to shape the meeting and serving on several of the key committees. In several instances, platforms such as GitHub in the Bitcoin case or Adhocracy in the EIDG case have played crucial roles in sustaining interfaces. These platforms in turn enable interactions with existing communities of stakeholders or other interested parties.

4. Knowledge and Memory Management

The case studies evidence a spectrum of approaches to maintaining knowledge and building institutional memory. These approaches differ primarily in how easily each approach can be leveraged to affect the groups’ future decisions. At one end of the spectrum, institutional memory is largely stored within the individual memories of staff. Creative Commons embodies this approach, where such human-centric institutional memory is supplemented by publicly-archived e-mail lists and documentation of all important decisions. Such an approach makes leveraging this institutional knowledge more challenging because it relies upon the availability and memory of certain key individuals.

Relying instead on a more comprehensive, public, online database of material makes leveraging institutional knowledge easier. NETmundial is a good example of this approach. In that case all of the input documents were not only available online, but were accessible in formats that allowed third parties to analyze and assess contributions. The use of such a database lowers the cost of acting on institutional knowledge and reduces points of failure.

Still one step further is the use of technology to automatically log not just every discussion but the state associated with every decision, so that the group can easily undo bad decisions. This approach was taken by Bitcoin, which used GitHub for tracking both the debate related to decisions about the core Bitcoin code and the decisions themselves. Because GitHub is designed for code-based projects, it was a particularly well-suited tool for Bitcoin.

5. Motivational Issues

Successful operation of the groups described in the case studies required committed and motivated participation from the stakeholders and participants. In most of the cases the primary motivation was a shared desire to solve the problem that the group convened to address. However, the full scope of motivating factors is likely as diverse as the number of participants (if not more so, given that the individuals and organizations profiled in the case studies often held multiple explicit and implicit motivations). Often, these motivations may be layered and co-existing as the actors involved in a case may have primary, secondary, or tertiary motivations shaping their actions.

As an example, one layer of motivations revealed in the cases clustered around two dimensions: monetary and non-monetary. Some of the parties in the Swiss ComCom Roundtables, for example, recognized that it would be better if deployment of fiber optic cable to the home could be done faster, with less disruption, and avoid regulation. A faster, less disruptive, and less regulated deployment would make it more likely that the utilities and telecom providers would lower their initial costs and begin profiting from their investments sooner. Achieving these goals, however, would require cooperation, and the Roundtables provided that mechanism. In contrast, other motivations are non-monetary. For example, ComCom itself may have been motivated more by the opportunity to exert influence over an area outside its regulatory scope than any monetary incentives. Other forms of non-monetary motivations can be seen among Creative Commons participants, who were in part motivated by academic goals (i.e., extending the licensing standard to include science research formats) and personal goals (i.e., the opportunity to participate in a strong interconnected network of individuals and experts who share common values and interests).

6. Communication

Within the case studies we observe two forms of communication: internal communication among participants and external communication with the broader public. The case studies demonstrate that both are important and that both are challenging to do effectively.

Internal communication helps participants contribute more effectively to the outcome of a group. Unclear communications can lead to wasted or duplicated effort. NETmundial offers one example, where vague and inadequate communications regarding the roles of the various committees left some participants duplicating work or changing tasks at the last moment. Untimely communications can leave participants feeling underprepared, as was reportedly the case in some of the Swiss ComCom Roundtables.

External communication helps outside observers understand the process and outcome, ultimately helping build support for the work of the group and enhancing the perceived legitimacy of the outcome. When successful, communication (often enabled by Internet technologies) can even extend the breadth and depth of participation in the process itself, enabling far-flung interactions and creating a multidirectional communications stream. In the case of NETmundial, technology allowed the organizers to collect content submissions and comments on the draft outcome document. Platforms such as Adobe Connect enabled remote participation for both the planning meetings and the event itself. Another example is the Marco Civil’s use of e-Democracia and Cultura Digital for enabling communication and online participation.

In some of the case studies, external communication and interactivity was challenging. In EIDG, for example, the Commission’s online communications and participation tool was significantly delayed until the process was already well underway. Once it was made public, it was not fully integrated into Commission’s working procedures. Although it generated 30,000 visitors and 3,250 active participants who performed 80,000 actions on the platform, the overall extent of participation did not meet the original expectations for the platform. Moreover, not all Commission members were equally willing to incorporate public feedback into their outputs, largely on account of the perceived increased effort this would take.

D. Outcomes

The tangible products (e.g., reports, proposals, agreements, guidelines, legislation, etc.) of a multistakeholder initiative are important, but our case studies show that participatory processes often create more ephemeral outcomes, too. Some outcomes are shared across all stakeholders, and some accrue more to individuals. Some outcomes are intended, and some are not. Some outcomes culminate the actions of the group, and some develop into separate ongoing actions. Ultimately, whether these outcomes represent “success” for the group depends on a variety of factors, tied to the purposes for which the group was originally formed and the legitimacy with which it operated.

1. Outputs, Outcomes, and Impact

Multistakeholder groups can yield many different kinds of outputs and outcomes, sometimes simultaneously. The case studies generally reflect coordinated efforts to address particular problems. Thus, a common output is a document that embodies the decision or identified solution. We see this in the Creative Commons case, where the outcome was version 4.0 of the licensing standard, and in the Aviation Slotting case, where the outcome was a set of guidelines for managing international slotting and busy airports.

Not all outputs are tangible, however. For instance, NETmundial, although it had a tangible output document, it also had an intangible output in the form of impact on other Internet governance fora, such as the UN’s Internet Governance Forum (IGF). The IGF has largely eschewed concrete outcomes and instead focused on promoting dialogue between participants, largely out of the belief that achieving consensus on specific issues would be too difficult. However, in the wake of the success of NETmundial, several participants at IGF 2014 suggested that IGF could and should rethink its objectives vis-à-vis outcomes. This sense of empowerment that NETmundial helped to generate at other venues is itself an important outcome.

Although NETmundial had global impact, an outcome need not be so widespread to be successful. In fact, some outcomes are limited to a small group of people or may even be personal in nature. For example, the Marco Civil case study illustrates how the process of discussion and collaborative drafting assisted in the building of ongoing alliances. Those alliances became important for obtaining the eventual passage of the bill and monitoring its enforcement. Similarly, the Creative Commons case study shows how one outcome helped participants build a network of knowledgeable peers. In both cases, an outcome was the construction of relationships and alliances.

2. Unintended Consequences

In some cases the intended outcomes are not the only ones that occur. For example, in the Swiss ComCom Roundtables, the final four-fiber agreement ended up having significant unintended effects on the utility companies and the utilization of their fiber. Because of the utility companies’ lack of experience as network managers and a competitive disadvantage against Swisscom, who could negotiate license agreements at a national instead of local level, the utility companies found themselves with under-utilized fiber. Similarly, the Aviation Slotting case study shows that the slot allocation standard has sometimes created unintended local problems related to antitrust and other regulatory concerns. In both cases, the unintended consequences seem to emerge from interactions between the group’s decision and external factors (and a failure to properly anticipate such interactions). In the case of the Swiss ComCom Roundtables, the issues emerged when the four-fiber plan interacted with content licensing agreements. In the case of the Aviation Slotting guidelines, the issues emerged when the guidelines interacted with existing local institutional structures.

3. Operational Continuity

Once a group has identified and implemented a solution, the group does not necessarily need to continue operating. With the exception of EIDG, our case studies do not include any examples of organizations that completely disbanded after identifying solutions to the problems they were formed to address. This may be a result of the small sample size, but it may also reflect the fact that few solutions are self-executing. For example, in both the TIIB and the Marco Civil case studies, continued operation of the alliances and networks that formed throughout the processes was necessary in order to keep pushing for legislative progress and to monitor enforcement. In other cases, such as the Swiss ComCom Roundtables, active pressure and work was no longer needed, but the group was never formally disbanded, leaving open the possibility for it to be reconstituted if needed in the future., In other cases there is never a complete solution, and the outcome is ongoing management functions, such as with the INCB. And even with EIDG, although the commission was formally disbanded, its work lives on through pending legislative proposals and the formation of a permanent legislative committee on digital society within the parliament.

IV. Critical Success Factors

Having described the common techniques, approaches, tools, and patterns observed within the case studies, we now turn toward critical success factors.

The governance bodies that have been objects of the case studies are almost all subject to specific scrutiny because they operate outside or on the fringes of common governance structures. To be deemed successful they are not only measured by the degree to which they have achieved their objectives and thus proved to have been “worth the effort.” But they also must show that they have integrated safeguards into their structures and procedures that are at least functional equivalents of those present in the common governance environment in which they operate. These safeguards include ensuring inclusiveness, transparency, accountability and legitimacy, all of which go beyond mere effectiveness. In both the case of meeting objectives and including safeguards, the definition of “success” is inherently context-dependent, as what is successful for one group may not align with the objectives, inclusiveness, transparency, accountability, or legitimacy of another.

Although context dependent, the case studies help illuminate the interactions between a group’s structural elements and the context in which it operates. While our case studies describe groups that were largely successful (as defined within their own context), the groups and activities represented within those case studies were not without fault. In almost all case studies, some techniques and tools worked better than others; and in some cases experiments and approaches simply failed.

A. Inclusiveness

One metric by which to evaluate governance groups is the degree to which they are inclusive of a diverse array of participants and stakeholders. That is not to say that more inclusiveness is necessarily better, but to instead consider the extent to which the structure and tools of the governance group enabled the greatest degree and diversity of participation within the constraints and needs of the group. There are many avenues for participation, each offering different kinds and depths of involvement. Perhaps the most obvious avenue is through formal inclusiveness—having a seat at the table and being able to contribute to the discussion. Offering input, however, is different from decision making; another avenue of participation is through being able to vote on or otherwise affect the outcome or decisions of the group. Participation is sometimes mediated through other means, such as technological tools or through sub-entities like working groups. Regardless of the available avenues of participation, inclusiveness requires consideration of how well the stakeholders can actually make use of these options.

1. Formal Inclusiveness: Identifying and Including Stakeholders

One avenue through which a governance group can increase inclusivity is by formally inviting a larger array of individuals and organizations to participate in the discussion as stakeholders. As noted in Part II, all of the case studies describe groups that are multistakeholder. But the word “multistakeholder” does not convey the nuance of arrangements and the complexity of how these groups are crafted. For example, “multistakeholder” does not mean that every possible stakeholder is included. Although reasons for this varied, in all of the case studies, the conveners exercised some degree of control and discretion over the number and types of participants.

There are many reasons why conveners would want to exercise control over the selection of participating stakeholders. Perhaps the most common rationale for participant curation is to ensure that participants have a significant connection to the topic at hand. To some extent, this is obvious: a “stakeholder” is literally someone with a stake in the issue or outcome. However, we see from the case studies that determining the level of connection sufficient to generate a stake, and excluding the rest, often results in greater specificity in selection. For example, governance groups have enforced a nexus between topic and potential stakeholder by looking at the level of economic investment (Swiss ComCom Roundtables) or at technical expertise or involvement (IXPs). And in the Israel National Cyber Bureau case study, the nexus requirement was directly related to the sensitive nature of the national security topics.

Topical relevance is only one reason for curation that we observe in the case studies. Other rationales tend to be more practical in nature. For example, NETmundial faced space and funding limitations that required selecting from a larger pool of potential participants. In the case of the TIIB and the EIDG, the significant government involvement in those governance groups meant that participants were selected for their compatibility with the policymakers involved.

Even the most open of the governance groups still involve a basic level of participant curation. For example, the Bitcoin case study describes the process of revising Bitcoin’s core codebase as a “meritocracy,” where anyone can submit code revisions. However, only a handful of participants are core developers, and only a core developer can invite a participant to become elevated to the status of core developer. Thus, while basic participation was open to all, the overall organization was focused on a more select group.

Whatever the reasons for the curation of participants, these decisions are rarely static. Instead, the inclusiveness of governance groups often changes over time. For example, the Swiss ComCom Roundtables were formally limited to the CEOs of companies that had already invested in fiber optic cable deployment. This decision excluded some organizations, such as homeowners associations, that had a strong interest in the outcome. During the course of the Roundtables, ComCom reached out to these organizations and even invited some to make presentations to the Roundtables.

The dynamic nature of participant inclusion highlights an important observation from the case studies: the conveners of the observed governance groups often approach decisions about inclusiveness instrumentally. In other words, the conveners use different configurations of participants in order to achieve strategic aims. Interestingly, our case studies do not evidence many, if any, detrimental effects from this instrumental approach. This could be because a successful outcome helps erase any concerns about inclusiveness. However, the more likely reason is that achieving a successful outcome often requires the full buy-in and participation of critical stakeholders, so any instrumental approach to inclusiveness is bounded by the need to maintain the support of those stakeholders necessary for developing and implementing a solution. We see this in the CT-Spam case, where CGI.br required the full support of the ISPs, technical, and consumer communities in order to implement its desired solution.

2. Voting, Consensus, and the Use of Decision-Making Processes

Formal participation in a multistakeholder process is important for stakeholders because it offers the opportunity to shape the final outcome. Thus, equally important to stakeholders are the mechanics of decision making that determine that final outcome. The array of decision-making processes is described above in Part II. What we observe from the case studies and their deployment of processes is that consensus is generally preferred, but that governance groups will revert to voting where consensus cannot be reached and a solution can still be implemented over objections.

That consensus is generally preferred is not a surprise. For example, in both the Swiss ComCom Roundtables and the CT-Spam cases, the lack of regulatory authority to force implementation of the solution meant that a successful outcome required the voluntary cooperation of all stakeholders. Because of this, consensus was required to ensure that the stakeholders would buy into, and take ownership of, the solution.

Governance groups switch to voting only when a solution can be implemented over the objections of stakeholders. Voting as a decision-making process suggests that there is disagreement regarding the outcome. It is only a reliable option when full acceptance of an outcome is not necessary or when dissenters can be compelled to accept a resolution. This often occurs when there exists a recognized regulatory or legal authority to impose a solution over objections. For example, the EIDG could switch to voting for two reasons: (1) the Commission’s outputs were reports and recommendations for parliament as opposed to items that the commission members could directly act on; and (2) to the extent that the German parliament wanted to act on recommendations of the EIDG, the parliament has the authority to enact laws without the full cooperation of all stakeholders. In that regard, the choice of decision-making process is closely related to the ability of the governance group to implement any given outcome. Switching to voting may come with the additional cost of changed participant behavior; the move from consensus may signal the acceptability of more adversarial processes and positions. In other words, a switch to voting when there is disagreement regarding the outcome may itself trigger greater disagreement. In the case of EIDG, the switch to voting pressured the participants to align with their political allegiances.

In some cases, governance groups use different decision-making processes at the same time in different parts of the same organization. For example, the Bitcoin case study describes several components: the Bitcoin code development on GitHub and the more formal Bitcoin Foundation. The development of Bitcoin code relies on GitHub forums to facilitate decision making. Through this medium, consensus is the simplest method for decision making, as voting would be challenging to implement in that environment. However, the Bitcoin Foundation, an organization that supports the development of Bitcoin, is a formal entity incorporated under the law of the District of Columbia, with formalized roles and legal obligations, and therefore a more formal role for voting and process. Although the level of interaction between the development of Bitcoin code and the Foundation is not entirely clear, we once again observe an increased significance of voting where it is supported by legal authority, even in an organization that generally uses consensus.

3. Effect of Alternative Participatory Mechanisms

Enabling in-person stakeholder participation is only one method of inclusion. The case studies demonstrate an array of alternative participatory mechanisms, some more effective than others. In several case studies, the use of mediating technologies was effective at expanding the array of participants to a wider community of stakeholders. For example, the Marco Civil case study describes how tools like eDemocracia allowed for collaborative drafting among members of the public.

That said, participatory technologies will not themselves automatically generate greater inclusiveness. Like any tool, participatory technologies must be used with intent and purpose. The remote hubs at NETmundial are an example where the organizers worked to ensure that individuals at the remote hubs would be able to participate actively and meaningfully in the discussion as though they were present in the room. Where online tools are not fully integrated into the workflow or embraced by the organizers, they were less effective. For example, in the EIDG case study, the delays in deploying the Adhocracy online platform meant that when it was ready, organizers had already developed their workflows. Participatory tools are not a panacea, but are instead a useful tool that can be deployed when their use is carefully considered and designed into the governance groups as a whole.

The case studies also contain examples of alternative participatory mechanisms that are not technology-based. The primary example from the case study is the use of working groups. Generally these working groups are used as a way to reduce friction in the decision-making process. For example, in the Swiss ComCom Roundtables, a working group was used to simplify decision making regarding less controversial technical decisions. Similarly, TIIB used a working group to bypass the more constraining membership requirements of the Board in order to include a broader array of viewpoints and perspectives. In contrast, the EIDG used working groups not as an alternative means of participation, but as the primary operational structure for the governance group.

4. Resource Asymmetries

Formal participatory roles or tools can increase inclusiveness only if participants and stakeholders have the ability and resources to take advantage of those opportunities for participation. Financial resource limitations constrained who was able to afford to attend the NETmundial meeting. Although funding was eventually secured to help support the attendance of those who could not afford to attend, by then it was too late for many to change plans. Similarly, both time and finance constraints affected who could participate in the EIDG working groups; participants who had funding to support their attendance and provide staff support could contribute more actively, while those who were simply volunteering on top of their normal jobs were more limited. And in the ComCom Roundtables, knowledge asymmetries about the telecommunications industry placed the utility companies at a disadvantage in negotiations. In some cases the constraints are less explicit. For example, in the Bitcoin and IXP case studies, technical expertise was an implicit limitation on participation.

It is certainly possible to overcome these limitations and asymmetries, but it generally requires both time and money. For example, additional funding eventually provided for expanded participation at NETmundial. More challenging is addressing knowledge asymmetries and engaging in capacity development, as this can take years, depending upon the starting circumstances. The WVBB case study mentions how the process of building capacity in water resources management involved developing a core of teachers who could then work over the course of years train the necessary team of skilled professionals who could implement and manage the system. Addressing these asymmetries in governance groups is important for increasing inclusiveness, but the difficulty of doing so should not be underestimated.

B. Transparency

In the same way that the tools of inclusiveness are often deployed in instrumental ways to achieve strategic objectives, transparency is also an instrumental tool within the case studies. The case studies show that often the level of transparency that a governance group uses can be moderated to the specific needs and circumstances that the group faces over its lifecycle.

Within the case studies, transparency is the effort at communicating with both the stakeholders and the public more broadly with respect to the decision-making process of the group. The approaches to transparency within the case studies spanned a wide spectrum, from very limited openness to fairly radical depths of transparency. The Swiss ComCom Roundtables represented a more limited approach to transparency: meetings were closed, and the parties agreed that public communications about the Roundtables would come from ComCom and not from individual stakeholders. Thus, the public record of the Roundtables is dominated by a single source of information mediated through a single entity. NETmundial and Marco Civil, by contrast, represent more transparent approaches. For example, in NETmundial all of the input statements were made publicly available online where they were searchable and open for research and analysis. Similarly, in Marco Civil, organizers required stakeholders to publish their comments online in a way that clearly identified the stakeholder and their position, allowing an opportunity for feedback and criticism.

Although the case studies used different approaches to, and levels of, transparency, in many cases the driving motivation was the same: what approach would best support the identification and implementation of a solution? In Marco Civil, for instance, the organizers believed that given the complexity of the negotiations, transparency would be necessary to help lower the barriers to reaching an agreement. In the ComCom Roundtables, ComCom believed that open meetings would make it more difficult for the stakeholders to compromise in the discussions. This fear may not have been unfounded, as the EIDG case study notes that the use of public meetings often led to political posturing on the part of the stakeholders, at the expense of engaging more deeply in the substance of the debate.

Accordingly, it seems that in some circumstances transparency can aid the deliberative process, while in other circumstances it may be a hindrance. The case studies suggest that one explanation of this divergence may be the background and previous experiences of the stakeholders and participants. In the context of the Internet space more broadly, transparency is a more established norm, and stakeholders who are more familiar with that environment may be more inclined to believe that transparency assists in the deliberative process. For example, the Bitcoin case study featured a community of highly technologically adept stakeholders predisposed to be more comfortable with having debates about the code in public forums. In contrast, EIDG and ComCom involved more politicians and established telecom incumbents, respectively, who may not be as steeped in the norms of transparency. This suggests that the level of transparency may be a reflection of the norms of the community instead of something that in the abstract aids or hinders deliberation; it may be self-fulfilling that a community that is primed to be suspect of transparency will find it to be hindrance to a successful deliberation.

Different contextual environments may also influence how decisions about transparency are perceived more broadly. In general, regardless of the level of transparency that a governance group used, there did not appear to be a significant negative impact on the outcome or on the perception of the group as a whole. This is not to suggest that all approaches were without fault. As noted previously, the transparency of the EIDG workshops was criticized for producing an overly political environment. In the ComCom Roundtables, some participants criticized the lack of transparency as a way for ComCom to control the messaging and create an appearance of greater authority. And even within the context of a fairly transparent approach, participants at NETmundial criticized the transparency of the drafting process and the selection process for committee representation. However, in no instances were transparency decisions (or lack thereof) viewed as fatal to the work of the group. This suggests that at the very least the level of transparency within the case studies was generally well-matched to the expectations of the community affected by the work of the governance groups.

C. Accountability

Although accountability has not emerged as a primary issue in these case studies, we do see various “traces” of it. While the analysis of the case studies does not point to a singular notion of accountability, we can identify a number of tools through which accountability is imposed or procedural limitations restraining the unbridled operation of a governance group. These limitations can take many forms, including voting for (or removing) key decision makers, leaving the governance group for alternatives, subjecting group decisions to oversight panels, offering opportunities for public deliberation and protest, and requiring various forms of transparency, among others.

One of the most common forms of accountability that we see in the case studies is the threat of defection to alternate forums. For example, the Bitcoin case study describes how the developers of the Bitcoin core code were inherently conservative in their actions due to the paramount concern of maintaining a unified Bitcoin architecture. The key developers explicitly described how the relatively low transaction costs of forking the code and creating a competing system limited their actions. Similarly (although less explicitly), the Creative Commons and NETmundial case studies describe situations where the outcomes (CC licenses and the NETmundial principles, respectively) were not self-executing and required continuous reaffirmation, meaning that people must continually opt in to their use. For instance, publishers of scientific data must continually opt in to the use of the Creative Commons 4.0 license for each new data set, meaning that abandoning Creative Commons licenses is always an option.

In addition to the threat of defection, we also see several other forms of accountability in the case studies. For instance, the use of public deliberation and criticism is described above and below. The case studies are mixed with respect to the effectiveness of the tools used to enable enhanced public deliberation. As a lever of accountability, public deliberation can be no more effective than the tools that enable that participation in the first place. In other words, if public participation is not fully realized or implemented within the governance group, it is unlikely to serve as a significant source of accountability. Transparency is yet another form of accountability that we observe in the case studies, and which is described more fully above.

To a lesser extent we also observe the use of both voting and oversight authority as levers of accountability within the case studies. For example, in EIDG, half of the Commission members were elected members of the German parliament. Although the public support of those elected members was likely tied to many more issues than just their activities on the Commission, it remained possible for those members to be removed from office. Additionally, both the EIDG and Marco Civil case studies offer examples of oversight authority. Both cases resulted in legislative proposals that required the action of a separate legislative body in order to become law. This separation between drafting and enactment provided accountability to the governance groups by limiting their ability to unilaterally implement their proposals.

In looking more broadly at the use of levers of accountability within the case studies, we observe a relationship between the methods of accountability used and the other defining elements that we explore elsewhere in this paper. For example, where formal regulatory or legal authority underlies the operation of the governance group, the use of voting or oversight committees appears to fit quite naturally as an accountability tool (e.g., EIDG and Marco Civil). However, in the absence of regulatory authority, it is often easier for stakeholders to leave the governance group and create alternatives, meaning that defection and transparency might be more effective levers of accountability (e.g., Swiss ComCom, Bitcoin, and Creative Commons). Accordingly, the use of accountability tools should not be viewed as an independent matter, but should instead be viewed in the context of the overall operation and structure of the governance group as a whole.

D. Legitimacy

Perceptions of the legitimacy of governance groups also proved an important contextual factor. As highlighted above, many of the governance groups we examined relied on multiple sources of legitimacy, which frequently included a combination of traditional sources, such as government, as well as non-traditional sources such as expertise or broad participation. In Section III above, we reviewed the range of sources of legitimacy in the case studies; here we reexamine legitimacy’s role as a factor in the success of the governance groups.

The case studies highlight the variety of ways in which legitimacy can shape a governance group’s success. In some cases, input legitimacy was a critical source of legitimacy. For example, in the wake of the Snowden revelations, most stakeholders in the Internet governance community agreed upon the need to stage a public discussion about Internet governance. However, the appropriate process and platform for holding that debate was highly contested, making the legitimacy of the NETmundial process a critical component.

In other cases, output legitimacy is more central to the success of the group. For example, in the Swiss ComCom Roundtables, the outcome was an agreement that improved the deployment of fiber optic cable across Switzerland. Despite concerns over inclusion of stakeholders within the process, the Roundtables have been perceived largely as legitimate due to the success of the outcome.

Interestingly, in some case cases legitimacy was itself a key deliverable for the group. At NETmundial, it was important to demonstrate that multistakeholder processes could be used to create principles for governance, making the legitimacy of the output and the process a goal in and of itself. In contrast, legitimacy was not an end in and of itself for the CT-Spam governance group. This was largely because once implemented, the Port 25/TCP change did not require continued buy-in and reaffirmation from stakeholders, and had relatively little capacity to cause public harm and great potential to enhance the public good. Although legitimacy in the process was necessary to obtain buy-in from the key stakeholders, it was not necessary as an outcome.

In addition to highlighting the ways in which legitimacy may be foregrounded at different stages of the governance process, the case studies also indicate that the centrality of legitimacy is linked to the alignment between the interests and preferences of stakeholders. In domains that are more politicized or where the positions of stakeholders are more divergent, legitimacy is more frequently called into question. In these instances, “stronger” forms of legitimacy become necessary for the perceived success of the process.

E. Effectiveness

Judging the overall effectiveness of governance groups is challenging because effectiveness must be assessed with respect to the groups’ unique context, objectives, and needs. The case studies covered a wide-range of outcomes, spanning convening to legislation. But whether those outcomes were effective depends on how well the governance group met initial expectations and achieved its goals. There is no one definition of effectiveness that can be applied across all the case studies. Rather, success is highly context-dependent, and should be measured at two levels: internally and externally. Internal definitions of success are generally more process-oriented, focused on whether the multistakeholder group operates as effectively as possible given its context. External definitions of success are generally objective-oriented, focused on whether or not a stated goal was achieved.

The very process of clearly defining a mission and outcomes can positively contribute to the overall effectiveness of a governance group. Vagueness around a group’s objectives can obfuscate the process of achieving consensus, and increases the vulnerability of a group to be influenced by political pressure. For example, a number of EIDG members expressed regret regarding what they perceived to be the overly broad focus of the Commission. They felt that if the group had a more clear and narrowly defined objective, the Commission might have achieved more significant results. Similarly, the mission can be more clearly defined by separating more technical decisions from policy ones. This division can improve effectiveness by freeing the stakeholders to reach consensus on less controversial decisions without resolving all of the political concerns that may have motivated the formation of the governance group in the first place.

The efficiency of the governance group has an impact on its effectiveness, but whether a governance group operates with sufficient speed appears to be a relative concept. In some cases, multistakeholder processes may have provided efficiencies for policymakers by filling in where regulations were not able to keep up with technological development. For example, the Swiss ComCom Roundtables benefitted from the relative efficiency of governance group as opposed to the more lengthy process of amending the telecommunications law to address fiber optic cable. In other cases, multistakeholder processes can be slow and arduous due to the need to derive input and achieve consensus from stakeholders who may or may not easily reach agreement. Though this can create frustrations among stakeholders, the pace of progress should be carefully calibrated with other objectives. For example, the authors of the CT-Spam case note that the technological solution to the spam problem was implemented in many countries around the same time, but those countries with greater centralized regulatory authority were able to implement the technical change far faster than Brazil. However, the multistakeholder group recognized that in Brazil a top-down process, while speedier, would have proved ineffective in the long run because the participants would not have fully supported or implemented the solution. In fact, the high-speed process at NETmundial may have negatively influenced the full inclusion of views by leaving insufficient time for stakeholders to achieve internal consensus. Thus, some level of efficiency may need to be sacrificed for the sake of long-term effectiveness.

Operating at multiple levels of scale can also contribute to the effectiveness of groups. For example, Swiss ComCom and the EIDG used smaller working groups to develop concrete solutions with more technical knowledge before passing these solutions up to broader higher-level plenary groups. Expertise was closely tied to questions of scale: in some instances, such as the EIDG, groups found it beneficial to develop solutions at a technical level before issues could become politicized, while in the CT-Spam group actors were hesitant to implement technical solutions without having input from a larger array legal and policy advisors. In still other cases, such as the TIIB and the INCB groups, efficiencies were achieved at the expense of full participation by limiting the number of stakeholders who could participate.

V. Summary of Observations

The analysis of the case studies from a variety of governance contexts reveals a complex and nuanced picture. We distill those findings below in three categories: (1) their purpose and context; (2) their formation and operation; and (3) their critical factors for success. With respect to the purpose and context for and in which multistakeholder groups are formed and operated, we identify the following key findings:

• Multistakeholder groups can serve a broad range of purposes, ranging from open-ended missions (e.g., gaining a deeper understanding of complex issues by incorporating views of multiple actors) to issue-specific tasks (e.g., execution of a particular policy). The primary mission is often supplemented by secondary purposes such as network building or formation of communities of knowledge and practice.

• Contextual factors, including culture, are key variables that have the potential to shape almost all dimensions of multistakeholder groups, from formation to dissolution. In particular, pre-existing relationships, cultures of problem-solving, and the availability of resources need be taken into account.

• Finding alternative sources of legitimacy is critical for groups that cannot build upon traditional sources.

We observe significant variations in the case studies when it comes to the formation and operation of governance groups. In many cases these differences are shaped by the unique purposes and contexts of the group. The key takeaways from the analysis include the following:

• There is no standard way to form multistakeholder groups. While in all cases the composition of the group is a critical element of both the formation and the ultimate outcome of the process, inclusion criteria as well as internal structures and architectures vary greatly from one case study to the other.

• The multistakeholder groups analyzed in this synthesis share a common collaborative mode and bring together distributed resources. The operational process itself, including systems and tools, is managed very differently from group-to-group, however. For instance, decision making and conflict resolution procedures range from consensus-driven to hierarchical, with all permutations in between. Similarly, interfaces with other institutions, communication strategies, and mechanisms of knowledge management range from formal to informal.

• With respect to outputs and outcomes, the case studies indicate the range of what multistakeholder groups can contribute to the world, including tangible outputs such as documents, specifications, standards or code and intangible outputs such as endorsement or empowerment. Not all of these outputs are intended; multistakeholder groups might also lead to unintended consequences, including complicated interactions with existing laws and regulations (e.g., antitrust issues).

In addition to (and interacting with) these structural elements, the analysis of the case studies has led to a series of observations regarding the critical factors for success of multistakeholder groups. These takeaways are:

• Inclusiveness (including mechanisms for participation) and transparency are critical factors to be managed and adjusted throughout the lifecycle of governance groups. The case studies revealed different formal and informal mechanisms of inclusion, as well as mechanisms of participation—such as consensus or voting. Similarly, transparency as an effort to communicate both within the group and with those outside is also an instrumental tool. Importantly, transparency and inclusiveness are dynamic characteristics and need to managed throughout the entire process, from formation and operation to the afterlife of a multistakeholder group.

• Accountability and legitimacy of governance groups are highly reflective of and dependent on contextual factors. The analysis of the case studies does not lead to a coherent concept of accountability, but highlights various elements of accountability, including direct approaches such as voting, and indirect mechanisms such as “defection” and opportunities for public criticism. These elements of accountability interact with other factors for success, such as transparency, and they need to be considered in the context of the overall operation and structure of the governance group. The case studies also highlight different kinds of legitimacy, ranging from input to output legitimacy, and demonstrate that governance groups might rely on overlapping sources of legitimacy at any given moment.

• The case studies suggest that effectiveness is defined with the context of the individual governance groups. In process terms, effectiveness describes the ability to operate within particular time constraints, but also manage pace over longer periods of time, and the ability to deal with multiple levels of scale, all of which might change over the life of a multistakeholder group.

VI. Conclusions

As noted in the introduction, this document provides a high-level synthesis of a case study series that seeks to deepen our understanding of the formation, operation, and effectiveness of governance groups. The goal of this work, taken as a whole, is to inform the evolution of and current debate around the Internet governance ecosystem in the light of the NETmundial Principles and Roadmap, as well as the discussions at the Internet Governance Forum (IGF), the NETmundial Initiative, as well other forums, panels, and committees.

Using the final report of the Panel on Global Internet Cooperation and Governance and Governance Mechanisms as a conceptual starting point, the case studies were selected to ensure diversity of cases along several axes emphasized in the Panel’s report. In particular, the participants in this collaborative research effort wanted to ensure that the case studies collectively inform us about: (1) how governance groups best match challenges with the organizations, experts, networks, and governing bodies/entities best able to help develop legitimate, effective, and efficient solutions; (2) how to structure the “flow of information and knowledge” necessary for successful governance; and (3) how different governance groups deal with coordination between regional and global (or national and regional) governance networks in order to avoid conflicting directives.

As to be expected, the real-world case studies examined in the context of this research effort only partially mapped onto the conceptual model outlined in the Panel’s report. Specifically, the analysis and discussion of the case studies has led to a more diverse and nuanced set of observations regarding the structure and operation of such groups, as well as the factors that are important for their successful operation. Indeed, the analysis of a diverse set of real-world case studies suggests a far richer set of options than we originally anticipated in the formation and operation of governance groups. From the diverse cases we have studied, no clear dependencies or hierarchies have emerged between the different approaches available to define and manage the purpose, structure, process, or performance of governance groups.

At the most fundamental level, the review of the case studies suggests that governance groups work best when they remain sensitive throughout their entire lifecycle to the surrounding (and typically dynamic) contextual and cultural conditions, the availability of support systems, and the opportunities for and trade-offs related to inclusion, transparency, and accountability. The groups that are most sensitive to these factors and how they change over time are best positioned to be flexible and adaptive to new circumstances and needs. At the same time, the detailed analysis of various structural elements and mechanisms makes clear that there is no single model or way to run governance groups that would be of strong prescriptive value across contexts. Rather, our study sheds light on the various “tools” available in the toolbox when it comes to formation, operation, and outcomes of multistakeholder governance groups, and this synthesis strives to highlight the interplay between those elements. Similarly, the analysis of a series of critical factors—including inclusiveness, transparency, accountability, legitimacy, and effectiveness—does not suggest a single approach that promises success across or even within governance contexts. While we cannot offer clear-cut “templates” or “model groups” based on our analysis, we instead offer observations and considerations with respect to a broad range of approaches, mechanisms, and tools and discuss these findings in the context of real-world situations.

What then is the secret to a governance group’s success if it is not the application of a specific model or tool? Our case studies indicate that the most robust, effective, and legitimate groups are those that demonstrate sensitivity to the cultural environment in which they operate. In other words, it is attentiveness to both the “multi” and the “stakes” of “multistakeholder”—creatively and dynamically harnessing the multiplicity of forces and motivations into a collaborative body. This places a premium on leadership that can not only guide the group toward the its purpose, but can simultaneously and continuously screen the environment for enhanced sources of legitimacy, stabilizing relationships, and enabling partnerships.

Such conclusions are of course limited by our research intentions and the samples chosen. Despite those constraints, the case studies help us understand that despite challenges to their creation and operation as well as often-difficult regulatory environments, there remains a vibrant, open, and complex space for multistakeholder groups and their successful operations. Indeed, it is sometimes in the most challenging environments where formal regulation cannot or does not exist that the benefits of multistakeholder governance groups can be most impactful. We hope that the lessons from the case studies can help the conveners and leaders of future governance groups better understand and dynamically deploy the full range of tools and mechanisms available to them.

---

ºACKNOWLEDGEMENTS

This synthesis and case study series would not have been possible without the dedication and hard work of many contributors from the Network of Centers (NoC) and the Berkman Center. In particular, we would like to thank our case study authors, who dedicated countless hours to this project and whose efforts were instrumental in its completion. Many thanks also to our friends and collaborators Herbert Burkert, Jens Drolshammer, Raimondo Iemma, Juan Carlos De Martin, Sarah Myers West, Mayte Peters Schomburg, Wolfgang Schulz, Stefaan Verhulst, and Jonathan Zittrain for guidance, collaboration, and support.

We wish to extend our gratitude to Rebekah Heacock Jones, David O’Brien, and Annie Pruitt for conceptual guidance and editorial work, and Gretchen Weber for communications support. Special thanks are due to Ryan Budish for coordinating this research effort and providing extensive research and editorial support.

The Berkman Center is grateful for initial seed funding for this project provided by ICANN, and acknowledges the use of portions of project and general support grants, respectively, by the MacArthur Foundation and Ford Foundation.

Urs Gasser