Introduction

In December 2012 the World Conference on International Telecommunications (WCIT-2012) convened in Dubai by the International Telecommunication Union (ITU) ended in controversy when 55 member states, including most Western democracies, walked away without signing the Final Acts, which constitute the new International Telecommunication Regulations (ITRs)(Schiller 2013; Pfanner 2012). The sticking point leading to this outcome was the apparently innocuous resolution which did no more than invite member states:

1. to elaborate on their respective positions on international Internet-related technical, development and public-policy issues within the mandate of ITU at various ITU forums including,inter alia, the World Telecommunication/ICT Policy Forum, the Broadband Commission for Digital Development and ITU study groups; [and]
2. to engage with all their stakeholders in this regard (ITU 2012).

The fact that 89 other member states signed the ITRs led to a Wall Street Journal article proclaiming this as 'America's first big digital defeat' (Crovitz 2012).

WCIT-2012 was but the most recent iteration of an apparently perpetual controversy concerning authority over central Internet resources, particularly the domain name system (DNS). In the late 1990s, an endemic lack of clarity regarding authority for managing the DNS resulted in a crisis in DNS governance (Mueller 2002; Lindsay 2007: 36-48). While the ITU, a UN-based organisation, was involved with an early attempt to establish a DNS governance structure, known as the International Ad Hoc Committee, this process was effectively sidelined when the US government established its own policy process. The US policy process eventually led to the formation of the Internet Corporation for Assigned Names and Numbers (ICANN) ' a private, not-for-profit corporation incorporated under Californian law ' which is responsible for the Internet Protocol (IP) address space and for overseeing the DNS (Lindsay 2007: 46-48; 64-5). 

Pursuant to an agreement with the US Department of Commerce, ICANN remained subject to US government supervision. The US government role in DNS governance was reinforced by a 2005 statement, entitled US Principles on the Internet's Domain Name Addressing System (NTIA 2005) which affirmed that:

The United States will continue to provide oversight so that ICANN maintains its focus and meets its core technical mission.

The statement was a reaction to a UN-based process, known as the World Summit on the Information Society (WSIS), which had been initially proposed by the ITU in 1998, but which was eventually held in two phases over the period 2003-2005 (Lindsay 2007: 92-4; Klein 2004). As part of this process, the Working Group on Internet Governance (WGIG) was established to investigate and make proposals for action on Internet governance (Lindsay 2007: 28).  In response to the US government's oversight of ICANN, the European Union (EU) issued the following statement at a 2005 WGIG meeting:

' the EU believes that a new cooperation model is needed in order to concretise the provisions in the WSIS Declaration of Principles regarding the crucial role of all actors within Internet governance (Lindsay 2007: 58).

Prior to the second phase of the WSIS, in late 2005, a compromise between the US and the EU began to emerge. The EU supported the privatisation of DNS governance and the role of ICANN, but advocated 'phasing out the oversight functions of the US Department of Commerce over ICANN' (Lindsay 2007: 58). The second phase of the WSIS resulted in the formation of a UN forum for multi-stakeholder dialogue, known as the Internet Governance Forum (IGF). The IGF is a powerless talk-fest, which convenes every year to discuss worthy issues. ICANN, on the other hand, retains responsibility for key decisions relating to the DNS, being in the midst of managing a massive (and complex) expansion of gTLDs (ICANN 2012; Lipton and Wong 2012).

Despite the reluctance of the US government to relax its supervision of ICANN, it was always envisaged that management of the DNS would be fully privatised. In 2009, largely in response to the opposition of the EU and other national governments to the continued US government role, the contract between the Department of Commerce and ICANN known as the Joint Project Agreement was not renewed, but was replaced with the Affirmation of Commitments (NTIA and ICANN 2009). The Affirmation is an unusual document, which appears to have no legal effect, but which nevertheless includes a commitment from ICANN to remaining a not-for-profit corporation headquartered in the United States (Froomkin 2011). While the Affirmation signalled an end to direct US oversight of ICANN, the US government retains a degree of potential legal influence over ICANN in the form of a contract for the performance of the IANA function, which includes assignment of IP numbers and protocol numbers (DOC and ICANN 2012). Moreover, the Department of Commerce remains in a contractual relationship with Verisign in relation to the operation of the authoritative root zone server (Froomkin 2011).

As the US government progressively relaxed its supervision of ICANN, the role of government representatives within ICANN itself was being expanded. From its inception, ICANN's complex internal structure incorporated a Governmental Advisory Committee (GAC), membership of which is open to all national governments, as well as international organisations, such as the ITU and the World Intellectual Property Organization (WIPO). As part of structural reforms introduced in 2002, changes to ICANN's Bylaws expanded the GAC's powers so that, for example, in the event of conflict between an ICANN board decision and GAC advice, the board is required to negotiate with the GAC (Lindsay 2007: 74-5; Froomkin 2011). As Froomkin has explained the evolution of ICANN's institutional relationship with governments:

' over time ICANN allied itself with non-U.S. governments as a way to extract the U.S. from its directly controlling role, and also as a way to head off non-U.S. support for alternatives to ICANN based on the ITU or the United Nations. In ICANN's latest evolution, rather than being fully privatised, the DNS is instead semi-internationalized (Froomkin 2011: 209).

The ongoing struggles over DNS governance are clearly multi-layered. In essence, however, they reflect a fundamental conflict between two models of international decision-making: the traditional top-down process, where institutional authority is derived from a treaty between sovereign nation states, and the emerging multi-stakeholder model, in which governance is based on negotiations between state and non-state actors, including representatives of the private and voluntary sectors. There is no simple definition of a multi-stakeholder organisation. According to Waz and Weiser, however, a multi-stakeholder organisation must, at the mimimum, satisfy the following two criteria:

1. representation (or, at a minimum, openness to representation) from a diversity of economic and social interests; and
2. a representational role for civil society, generally defined as relevant stakeholders other than government (Waz and Weiser 2012: 336).

The persistent problems with DNS governance arise, in large measure, from questions over the legitimacy of the ICANN multi-stakeholder model. This article examines this issue in the context of an explanation and analysis of the disputes associated with ICANN's decision to approve the .XXX gTLD which, to date, has been the most controversial top level domain.

Legitimacy

Although it is often used in a loose, colloquial sense, legitimacy is a complex concept, which has changed over time. For example, in pre-modern times the legitimacy of authority structures was based on shared metaphysical conceptions, such as the doctrine of divine right (Modak-Truran 2013). The 1648 Treaty of Westphalia, which concluded the Thirty Years War, is customarily taken as the beginning of modern international law, with the legitimacy of international agreements henceforth deriving from territorially-based sovereign nation states (Berman 2003; Wilson 2009).

With the evolution of democratic governments, modern concepts of legitimacy are based on two principles: the rule of law and popular sovereignty (or democratic governance) (Habermas 1996). The answer to the question of whether a power structure is legitimate therefore has two parts.

• First, there is the positive or descriptive answer that a power structure is legitimate to the extent that it derives its authority from a recognised legal source, such as a treaty or a constitution. In this sense, the rule of law can be regarded as establishing limits on unlimited or arbitrary power, especially to protect the rights of individuals.
• Second, there is the normative answer that a power structure is legitimate so long as it reflects principles of justice. As it is difficult or impossible for there to be agreement on what these principles might be, it is common for this normative conception of legitimacy to be subsumed under a procedural view of justice, whereby legitimacy is conferred by a democratic decision-making process.

There are well-known problems with dualistic modern concepts of legitimacy. The positive answer to the question of legitimacy ' that a power structure is legitimate because it is lawful ' is notoriously circular. Once a lawful source of authority is identified, then it becomes important to ask why that source is legitimate. For example, the legitimacy of an international organisation, such as WIPO, is based on an international treaty. Why, then, is a treaty regarded as a legitimate source of authority' The answer is that it derives from sovereign nation states. But why are agreements between nation states regarded as a legitimate source of power' The answer to this is either the simple positivist answer ' because the nation state is recognised as a source of lawful authority ' or some recourse to a normative source of power ' such as nation states derive their authority from popular sovereignty (the 'will of the people'). Once normative conceptions of legitimacy enter into the picture, difficulties arise from reconciling incommensurable perspectives. For example, in liberal democracies, the rule of law incorporates the protection of individual rights, which necessarily requires limits on popular sovereignty. What precisely are those limits, and who decides what they should be' Moreover, as far as international decision-making is concerned, it is simply unfeasible for decisions to be based on anything like a democratic mandate, even in the loose sense of electorates voting for representatives on an international body. As globalisation increases the importance of developing solutions to global problems, the international processes required become tragically more distant from the people which, according to the democratic principle, remain the source of legitimate decision-making.

Within a broadly Kantian tradition, there have been sophisticated attempts to reconcile the contradictions in modern concepts of legitimacy, notably by Rawls (2005) and Habermas (1996). Against these attempts to develop coherent accounts of legitimacy, however, postmodern theorists have argued that universal theories of legitimacy obscure differences between interest groups, such as differences in class, ethnicity and gender (Rosenfeld 2011). This has led to calls for a re-conceptualisation of concepts of legitimacy. Rosenfeld, for example, has argued that:

' as the need for greater convergence on a transnational scale joins the need for greater room for divergence on  a national or infra-national scale, appropriate standards of legitimacy and normative validity adapted to these new circumstances are called for (Rosenfeld 2011: 4).

Similarly, Modak-Truran has argued that in radically pluralistic societies, there is a need for the emergence of new, post-secular conceptions of legitimacy (Modak-Truran 2008).

Conflicts over DNS governance may therefore be seen against a background of contestation over conceptions of legitimacy. While a UN-based organisation, such as the ITU, has the imprimateur of its sovereign nation state members, it is remote from other stake-holders, and potentially cumbersome and unresponsive in its decision-making. It is, however, clearly based on a recognised lawful source of authority. On the other hand, it is impossible to point to a persuasive legal source for ICANN's authority over the international resource that is the DNS. Any legitimacy that ICANN has is not based on a positive legal source, but arises solely from the extent to which stakeholders, including national governments, recognise ICANN as legitimate. This means that the ICANN multi-stakeholder model is inherently unstable, as it must have continual recourse to contestable normative accounts of legitimacy.

In the absence of a convincing external positive legal source of authority, ICANN has incorporated a version of the rule of law within its constituent legal documents which, in general, require it to operate in a fair and transparent way. The clearest example of this is probably Article 4 of ICANN's Articles of Incorporation, which provides that ICANN:

shall operate for the benefit of the Internet community as a whole, carrying out its activities in conformity with relevant principles of international law and applicable international conventions and local law and, to the extent appropriate and consistent with these Articles and its Bylaws, through open and transparent processes that enable competition and open entry in Internet-related markets. To this effect, the Corporation shall cooperate as appropriate with relevant international organizations.

Given the impossibility of ICANN directly implementing a form of popular sovereignty, the multi-stakeholder model establishes an analogue of this in the form of a pluralistic internal structure, involving three supporting organisations and four advisory committees. To an extent, the privileged position of the GAC within ICANN's internal structure can be seen as an attempt to internalise an external source of legitimate authority. As Froomkin has suggested, however, the privileged role of the GAC, which confers a degree of legitimacy, has the potential to undermine the pluralistic internal structure, by providing an avenue for undue influence:

On the one hand, the existence of the GAC provides a source of external supervision over ICANN's activities; on the other hand, GAC also provides a route by which governments ' might be able to harness the [DNS] root to some extraneous end (Froomkin 2011: 218).  

ICANN's weak claims to external legitimacy, as well as the dangers of possible capture by stakeholders, including governments, mean that considerable attention is necessarily focused on the transparency and fairness of ICANN's processes. This emphasis on process is entrenched in ICANN's constituent documents, including the core values set out in Article I.2 of ICANN's Bylaws, which include:

• Seeking and supporting broad, informed participation reflecting the functional, geographic and cultural diversity of the Internet at all levels of policy development and decision-making.
• Employing open and transparent policy development mechanisms that (i) promote well-informed decisions based on expert advice, and (ii) ensure that those entities most affected can assist in the policy development process.
• Making decisions by applying documented policies neutrally and objectively, with integrity and fairness.

In sum, ICANN's weak formal claims to legitimacy mean that questions will always be asked about its authority to make decisions about the DNS. The multi-stakeholder model is only sustainable, then, if ICANN continues to be recognised by the majority of influential stakeholders, including (but not confined to) national governments. In the absence of convincing external sources of legitimacy, ICANN's continued authority therefore rests, to a considerable extent, on how well it manages its stakeholders. Possibly more importantly, ICANN's tenuous claims to authority mean that it must be meticulous in complying with fair and transparent policy processes. Much more than traditional treaty-based international organisations, ICANN's legitimacy depends upon the extent to which its decisions are regarded as legitimate which, in turn, depends upon the legitimacy of its decision-making processes. The disputes involving the application for the .XXX gTLD must be seen in this context.

The .XXX Application Process

From a policy perspective, ICANN's most significant function is to decide on the introduction of new gTLDs into the DNS. This is set out in Article 3 of ICANN's Articles of Incorporation, which provides that one of the organisation's objects is:

performing and overseeing functions related to the coordination of the Internet domain name system ('DNS'), including the development of policies for determining the circumstances under which new top-level domains are added to the DNS root system.

In 2000, ICANN called for applications for new gTLDs (Lindsay 2007: 12-13). ICM Registry Inc., a company incorporated in Toronto, Canada, submitted an application for .XXX, which was intended to be used for the purpose of 'adult' entertainment, as an unsponsored TLD (uTLD). In the event, in November 2000, the ICANN board approved only seven of the 44 applications it received for new gTLDs, rejecting the .XXX application. Subsequently, in December 2003 ICANN issued a Request for Proposals (RFP) for additional new gTLDs, this time confined to sponsored TLDs (sTLDs). An sTLD is a specialised TLD that has a sponsor representing the community that is most affected by the TLD. The RFP included selection criteria in relation to technical standards, financial standards and sponsorship requirements.  In 2004, ICANN received ten sTLD applications, including one from ICM Registry for a .XXX sTLD, which nominated the International Foundation for Online Responsibility (IFFOR) as the sponsoring organisation. By this stage, Stuart Lawley, a British entrepreneur based in Florida had taken effective control of ICM, and Joan Irvine, formerly the executive director of the Association of Sites Advocating Child Protection (ASACP), was executive director of IFFOR (Richards and Calvert 2011). In support of the application, Lawley said that it was a: 

significant step towards the goal of protecting children from adult content, and [to] facilitate the efforts of anyone who wishes to identify, filter or avoid adult content. Thus, the presence of a '.XXX' in a web address would serve a dual role: both indicating to users that the website contained adult content, thereby allowing users to choose to avoid it, and also indicating to potential adult-entertainment consumers that the websites could be trusted to avoid questionable business practices (ICM Registry v ICANN 2010:[14]).

ICANN established an independent panel of experts, known as the Evaluation Panel, headed by Liz Williams from Australia, to review the 2004 applications against the selection criteria. The Evaluation Panel found that, although it met the other criteria, ICM's application failed to satisfy all of the sponsorship criteria, mainly in that it was not clear that ICM represented 'a clearly defined community.' (ICM Registry v ICANN 2010:[15]). Given the deficiencies in some other applications for the new sTLDs, however, the ICANN board resolved to give all applicants a further opportunity to resolve perceived deficiencies.  Consequently, ICM submitted a revised application on 7 December 2004.

On 1 June 2005, the ICANN board, by majority, unconditionally resolved to authorise ICANN's President and General Counsel to enter negotiations with ICM relating to a proposed registry agreement, to be submitted to the board for approval.  At that stage, despite numerous opportunities and invitations to do so, no members of the GAC expressed any reservations, or indeed made any other comments, in relation to the .XXX application. Following the ICANN board's resolution to proceed, however, at a GAC meeting, national representatives expressed serious concerns about an 'apparent shift' in policy towards approving new sTLDs, clearly referring to the .XXX application.  While negotiations between ICANN's General Counsel and ICM on the terms of a proposed agreement proceeded, the Assistant Secretary for Communications of the U.S. Department of Commerce and the Chairman of the GAC in his 'personal capacity' sent two important letters to ICANN requesting a delay in the decision on the .XXX proposal.  Subsequently, other governments, including the Australian government, expressed reservations about ICM's application.  The board decision was delayed and, in May 2006, the ICANN President wrote to the GAC indicating that a final decision had not been made.

The concerns of national governments, and especially the US government, which had previously supported the .XXX application, were influenced by the mobilisation of anti-pornography campaigners, including the religious right, which organised mass email campaigns directed at the Department of Commerce. The campaigns occurred against the backdrop of preparations for the first phase of the WSIS which, as explained above, served as a focus for criticism of the influence of the US over DNS governance. In September 2005, for example, a representative of the European Commission wrote to the ICANN Chair claiming that the .XXX decision had been made without sufficient consultation with the GAC (ICM Registry v ICANN 2010: [30]). Public opposition to the application was not, however, confined to moral conservative lobby groups, as some free speech advocates feared the proposed TLD would increase the potential for censorship, as countries could easily block access to the entire domain (Richards and Calvert 2011).

Following a considerable amount of prevarication, including the exchange of communications between the GAC and ICANN's President, the application came before the ICANN board on 30 March 2007. Meanwhile, negotiations towards a registry agreement with ICM had been proceeding with, for example, ICM providing commitments to suppress child pornography. By a majority of nine to five, the ICANN board resolved to reject ICM's application on the basis that the proposed registry agreement failed to meet the sponsorship criteria of the RFP, but also that ICM's application failed to adequately address public policy issues raised by the GAC, including law enforcement compliance issues, and the possibility that ICANN might be required to oversee Internet content. In relation to this latter point, then ICANN Chair, Vint Cerf, stated that he was persuaded, 'that there were very credible scenarios in which the operation of IFFOR and ICM might still lead to ICANN being propelled into responding to complaints that some content on some of the registered .xxx sites didn't somehow meet the expectations of the general public'' (ICM Registry v ICANN 2010:[48]). Dissenting from the decision that the application should be rejected, however, board member Susan Crawford stated that:

I am troubled by the path the board has followed on this issue ' ICANN only creates problems for itself when it acts in an ad hoc fashion in response to political pressures. ICANN ' should resist efforts by governments to veto what it does ' The most fundamental value of the global Internet community is that people who propose to use the Internet protocols and infrastructures for otherwise lawful purposes, without threatening the operational stability or security of the Internet, should be presumed to be entitled to do so (ICM Registry v ICANN 2010:[52]). 

ICM Registry v ICANN

As part of the attempt to institutionalise accountability, Article IV.3 of ICANN's Articles requires ICANN to establish a process for independent third-party review of board actions that are alleged to be inconsistent with the Articles of Incorporation or the Bylaws (Lindsay 2007: 81). The Bylaws further require ICANN to establish an Independent Review Panel (IRP) to be operated by an international arbitration provider appointed by ICANN, to be known as the IRP Provider. Any person who is 'materially affected' by a board decision or action that is alleged to be inconsistent with the Articles or Bylaws may submit a request for independent review. In May 2004, ICANN entered an agreement with the International Center for Dispute Resolution (ICDR), which is the international division of the American Arbitration Association, to provide the functions of the IRP (ICANN 2004).

Following a 2006 resolution of the ICANN board to effectively defer a decision on the .XXX application, ICM filed a request for reconsideration of the decision, but was persuaded to withdraw the request when ICANN's general counsel indicated that the board would consider a revised draft registry agreement. After the board definitively resolved to reject the application, on 6 June 2008 ICM filed a request, pursuant to Article IV(2) of ICANN's Bylaws, for an independent review of the board decision. The request came before a three member ICDR panel, consisting of Judge Stephen Schwebel (Presiding), Jan Paulsson and Judge Tevrizian. By majority, with Judge Tevrizian dissenting, the panel decided that the ICANN board resolution to reject the .XXX application was inconsistent with ICANN's Articles and Bylaws. Before reaching this conclusion, however, the panel was required to address three important preliminary matters: whether the review process was a binding arbitration; whether ICANN board decisions are entitled to deference; and the applicable law of the proceeding.

On the first issue, the panel unanimously concluded that a decision of the IRP is not a binding arbitration.  While the ICANN Bylaws and review procedures are ambiguous, the panel found more indications that an IRP declaration is purely advisory.  The highly qualified nature of the terms used to describe the IRP process led the panel to conclude that ultimate decision-making authority in relation to requests for review rests not with the IRO, but with ICANN. For example, while ICANN's articles refer to 'arbitration', which implies that a decision is binding, the authority of the IRP under the Bylaws is to 'declare' whether the action of the board is inconsistent with the Articles or Bylaws, and not to 'decide' or 'determine'. Moreover, Section 3(15) of the Bylaws simply provides that, 'Where feasible, the Board shall consider the IRP declaration at the Board's next meeting'.

On the second issue, a majority of the panel concluded that ICANN board decisions are not entitled to deference and that the IRP must engage in de novo review.  While Californian law, to which ICANN is subject, embodies the 'business judgment rule', whereby directors who act bona fide in the interest of a corporation are shielded from liability, the majority decided that this default rule did not apply because ICANN was not an ordinary non-profit Californian corporation. In this respect, the majority emphasised that:

The Government of the United States vested regulatory authority of vast dimension and pervasive global reach in ICANN (ICM Registry v ICANN 2010:[136]).

The dissenting panelist, however, applied the domestic Californian 'business judgment rule' to conclude that ICANN board decisions should be given substantial deference.                 

The third, and potentially most important, preliminary issue centred on a dispute between the parties concerning the meaning of the phrase 'relevant principles of international law' in Article 4 of ICANN's Articles of Incorporation, which is extracted in full above. On the one hand, ICM claimed that Article 4 acts as a choice-of-law provision and, referring to Article 38 of the Statute of the International Court of Justice, submitted that international law includes 'the general principles of law recognised by civilised nations', including the principle of good faith.¹ According to this argument, by rejecting the .XXX application ICANN had not acted in good faith, and was therefore acting contrary to principles of international law (Simma and Alston 1992). ICANN, on the other hand, contended that Article 4 was not a choice-of-law provision and that, as a private party, ICANN is not subject to international legal principles that apply only to sovereign states.

The majority of the panel refrained from deciding whether or not Article 4 is a choice-of-law provision, but went on to determine that it required ICANN to act consistently with principles of international law, including general principles recognised as a source of international law, such as the principle of good faith. In reaching this decision, the majority recognised ICANN's highly unusual status, interpreting Article 4 as indicating 'an intention that the Panel understands to have been to subject ICANN to relevant international legal principles because of its governance of an intrinsically international resource of immense importance to global communications and economies' (ICM Registry v ICANN 2010:[140]). The dissenting panelist, on the other hand, agreed with ICANN's submissions that Article 4 is not a choice-of-law provision, and that as a private corporation ICANN could not be subject to the body of law governing sovereign nation states. That said, this difference of opinion on the interpretation of Article 4 was not determinative, as the majority of the panel considered that the principle of good faith was not only found in international law and in general principles that are the source of international law, but also in domestic Californian corporate law.   

After dealing with these preliminary issues, the panel turned to the central issue of whether or not the decision to reject the .XXX application was inconsistent with ICANN's Articles and Bylaws. Although the majority had decided that ICANN is required to act in good faith this was not the basis of its decision. Instead, the majority emphasised the importance of ICANN complying with the core value relating to procedural fairness which, as set out in Article I.2.8 of the Bylaws, requires ICANN to make decisions 'by applying documented policies neutrally and objectively, with integrity and fairness'. On the facts, the majority concluded the sTLD application process established by ICANN was a two-step process and that, by adopting the resolutions of June 2005 to proceed with negotiating a registry agreement, the ICANN board had determined that ICM's application met the sponsorship criteria. Consequently, the majority decided that, once it was determined that ICM satisfied the criteria, it should have proceeded to negotiate a registry agreement. In reaching this conclusion, the majority formed the view that the majority of the ICANN board had mistakenly concluded that, even after deciding that negotiations should proceed, the question of whether or not the sponsorship criteria had been met was open to reconsideration. Nevertheless, despite the documentary record, the majority of the panel did not consider that the board succumbed to pressure from individual governments or the GAC, stating that:

It is not possible to estimate just how influential expressions of governmental positions were. They were undoubtedly very influential but it is not clear that they were decisive. If the Board simply had yielded to governmental pressure, it would have disposed of the ICM application much earlier. The Panel does not conclude that the Board, absent the expression of those governmental positions, would necessarily have arrived at a conclusion favorable to ICM. (ICM Registry v ICANN 2010:[150]). 

The dissenting panelist, emphasising the importance of giving due deference to board decisions pursuant to the Californian 'business judgment rule', contrastingly decided that the June 2005 resolutions had not conclusively determined the sponsorship issue, thereby allowing the board to subsequently revisit the issue.

ICANN's Decision to Approve .XXX

As explained above, the IRP unanimously decided that its decision was not a binding arbitration. Moreover, pursuant to Article IV.3.15 of ICANN's Bylaws, the board was required only to 'consider' the IRP declaration at the next board meeting 'where feasible'. Nevertheless, following the IRP declaration the issue was considered at the March 2010 board meeting, held in Nairobi, where the board directed ICANN's President and general counsel to finalise a report on options for further consideration and to make the report available for public comment. The resulting options paper made it abundantly clear that, given the advisory nature of an IRP declaration, ICANN considered itself free to determine how to proceed and to revisit the key findings of the IRP (ICANN 2010). In particular, the options paper rejected the possibility of proceeding directly to execution of the negotiated registry agreement, reasoning that ICANN was required to conduct further due diligence, and take due account of advice from the GAC. More than 13,700 public comments were received on the options paper.

ICANN was in an unenviable position in responding to the IRP declaration, and not merely because of the need to evaluate the large number of public comments. On the one hand, an independent review process had formed the view that ICANN had already decided that the .XXX application satisfied all relevant criteria, including sponsorship criteria, and that re-opening this issue was an abuse of process. On the other hand, the IRP declaration was non-binding, and there continued to be strong opposition to the introduction of a .XXX TLD, including from governments. Moreover, in responding to the IRP declaration, according to its Bylaws ICANN was required to consult with the GAC, and in the event of a conflict, negotiate 'in good faith and in a timely and efficient manner'. 

In the event, it is unsurprising that on 25 June 2010 the ICANN board accepted the declaration of the majority of the IRP and directed ICANN staff to proceed with the .XXX registry agreement. It is simply inconceivable, although formally free to act contrary to an IRP declaration, that the board would decide to do so. If nothing else, this would leave any new decision subject to IRP merits-based review. Accordingly, a proposed registry agreement was posted for public comment in August 2010 and a consultation process was initiated with the GAC. The continued opposition of GAC members to the .XXX application was made abundantly clear in a letter from the GAC Chair to the ICANN Chair in March 2011, which stated that:

• There is no active support of the GAC for the introduction of a .xxx TLD.
• While there are members, which neither endorse nor oppose the introduction of a .xxx TLD, others are emphatically opposed from a public policy perspective to the introduction of a.xxx TLD (Dryden 2011).

The GAC was also concerned that ICANN might be required to supervise .XXX content. The registry agreement came before the ICANN board at the March 2011 ICANN 40 meeting in San Francisco, which was accompanied by public protests from the US Free Speech Association and the Californian adult entertainment industry (Richards and Calvert 2011: 537-8). Despite continued opposition from sections of the public and the GAC, a 9-3 majority of the board (with 4 abstentions) approved the registry agreement. In its summary of the outcomes of the San Francisco meeting, ICANN stated that:

The Board approved the dot-xxx domain after years of agonized debate. The decision was not easy and several Board members spoke out against it, not to mention members of the adult industry based in California. Governments had also made it clearly (sic) they did not support the creation of dot-xxx. Ultimately, ICANN approved the application because to have done otherwise would have undermined its own independent review and raise significant questions about the organization's accountability (ICANN 2011).

In other words, the flaws in ICANN's internal processes identified by the IRP left the board with no alternative but to approve .XXX.  This is confirmed by statements made by board member Rita Rodin Johnstone, who explained her decision to vote in favour of the application as follows:

The role of a Director is not to be popular or lauded. Nor is it to inject personal beliefs or biases into debates. It is to respect established processes and apply them equally to all. If the Board were to vote against .XXX notwithstanding the findings of the IRP, it would turn its back on a process the ICANN community established to be a check and balance on its decisions. This would be a violation of its fiduciary duties, and no governmental pressure or popular belief could assuage that breach (Johnstone 2011: 523-4).

The .XXX TLD was delegated to the root zone in June 2011 and was opened to public registrations in December 2011, with there now being more than 220,000 total registrations in the domain.

The Manwin Disputes

ICANN's decision to approve ICM's application for .XXX was by no means the end of the legal disputes involving the TLD. Manwin Licensing International S.A.R.L. is a company incorporated in Luxembourg, founded by Fabian Thylmann, which specialises in search engine optimisation for adult entertainment sites and which has agreements, among others, with Playboy Enterprises, Inc. It also operates several adult websites, including YouPorn.com.

After the failure of negotiations for a profit-sharing agreement with ICM proposed by Thylmann, in November 2011, Manwin filed a suit under US antitrust law in a Californian federal district court against both ICM and ICANN. Amongst other matters, the complaint seeks an injunction against the .XXX gTLD on the basis of: the alleged anti-competitive costs imposed on trade mark owners by the need for defensive registrations of domain names in the .XXX TLD; and an alleged lack of competition in the process awarding the sole .XXX registry agreement to ICM. In early 2012, ICM and ICANN filed motions to dismiss Manwin's complaint and in August 2012, the Californian District Court accepted two of the seven motions to dismiss the complaint, but held that ICANN is subject to US anti-trust law (Manwin Licensing International v ICM Registry 2012). The court further indicated that the case would proceed largely on the basis of the allegations relating to defensive registrations. In a separate suit, in October 2012 ICM filed a complaint against Manwin, alleging that Manwin's decision to boycott the .XXX TLD was anti-competitive. In this dispute, the Californian court granted Manwin's motion to dismiss, on the basis that ICM's complaint had not sufficiently alleged anti-competitive harm (Murphy 2013).

Meanwhile, in addition to the antitrust suit, in November 2011 Manwin lodged a request with the ICDR for a review of the decisions to approve the .XXX application and to grant ICM a registry agreement, pursuant to ICANN's independent review process. This is only the second application for review of an ICANN decision, following ICM's successful use of the review process. At the time of writing this article, the dispute centred on Manwin's standing to bring a request for independent review, with ICANN arguing that Manwin is not 'materially affected' by the ICANN decisions (ICANN 2013). ICANN is also arguing that any review would re-open issues that were definitively resolved by ICDR's determination of the ICM application for review.

Conclusion

ICANN is a not-for-profit corporation, that is incorporated under California's Nonprofit Public Benefit Corporation Law, but which is responsible for managing and making public policy decisions about a global resource of immense economic and social significance. This is already unusual, but ICANN's internal governance structures further serve to emphasise the extraordinary nature of this organisation. In accordance with the multi-stakeholder model, ICANN must coordinate input from the private sector, from national governments and from the non-profit sector. Although it is a private cooperation, and has been freed from US government supervision, it is subject to input from governments via the GAC, which has a privileged position within ICANN's governance structure. ICANN is committed to remaining based in the US and, consequently, being subject to US domestic law, including US anti-trust law. Yet, in accordance with its Articles of Incorporation, ICANN is subject to public international law, and to principles of international law, that conventionally apply only to sovereign nation states.

As a private corporation with a multi-stakeholder internal structure, ICANN has weak claims to formal legitimacy. In response to the legal positivist question, 'what is the lawful source of ICANN's authority'' it is simply impossible to give a clear answer. As a result, as long as it exists in its present form, without a firm foundation in traditional public international law, such as would be provided by an international treaty, ICANN will be subject to questions about its legitimacy. The controversy surrounding the role of the ITU in WCIT-2012 will not be the last time that the ITU or nation states seek a greater role in Internet governance, including DNS governance.

As a result of its weak claims to legitimacy under positive law, ICANN must rely on normative claims to a much greater extent than more traditional public decision-makers. ICANN's constitutional documents ' its Articles of Incorporation and Bylaws ' attempt to entrench a version of the rule of law by imposing procedural requirements in the form of rules requiring accountability and transparency. An important component of these procedural safeguards is the ability for affected parties to have ICANN decisions reviewed by an independent review process. ICANN's internal structure also attempts to implement a degree of accountability through the openness of its 'bottom-up' policy development processes. A further important element in ICANN's claims to legitimacy is its accountability to sovereign national governments, through the privileged position given to the GAC.

As the case study of the process for approving the .XXX application analysed in this article indicates, however, there are flaws in ICANN's claims to normative legitimacy. The flaws are both structural and behavioural. First, if ICANN is to be truly accountable, the decisions of the IRP should be made binding, not merely advisory. There is, after all, little point in having an independent review process if the review decisions can be effectively ignored. Second, there is clearly scope for improvement in more precisely specifying the criteria for review of ICANN decisions. In particular, the wording of Article 4 of ICANN's Articles of Incorporation, which subjects ICANN to all relevant principles of international law, international conventions and local law, appears to be impossibly vague and indeterminate. This seems to be why the majority of the IRP in ICM's application preferred to base its decision on the more specific requirement in ICANN's Bylaws for it to apply its documented policies neutrally and fairly. To be clear, while there is no reason to object to an organisation such as ICANN being subject to public international law, some thought must be given to what this actually means. Moreover, if ICANN is regarded as a truly international regulatory body, whose decisions are subject to multiple internal checks and balances, it seems unusual that it should be subject to the domestic laws of one particular nation state. After the lengthy and expensive process, complete with a thorough independent review, leading to the approval of .XXX, it is odd that a Californian court may have the final say. Although it seems highly unlikely that Manwin's anti-trust complaint against ICANN and ICM will succeed, managing such actions is, if nothing else, a drain on resources that could be better spent elsewhere. 

Third, there is considerable scope for concern about the role played by national governments, and the GAC, in the .XXX application process. There is little point in institutionalising procedural safeguards if an important group in the policy-making process does not fully comply with, or is unaware of, the proper processes. To begin with, it seems that national governments paid insufficient attention to the process of approving the .XXX application, to the extent that they were surprised by the June 2005 board decision that the application complied with all relevant criteria. Once they became aware, national governments and the GAC attempted to have the issue re-opened, largely on grounds that had little to do with the express criteria established by ICANN for approving new sTLDs. Therefore, as Froomkin has suggested, the GAC's role is a two-edged sword: it confers some much-needed legitimacy on ICANN, but it also opens the gate to undue influence over policy decisions. In this, it is important to bear in mind the considerable power imbalance between ICANN and national governments; if a sufficient number of influential national governments were sufficiently annoyed by ICANN decisions, they could simply replace ICANN, possibly with a traditional treaty-based organisation over which they might have more influence. ICANN is therefore subject to a degree of pressure, potentially not always transparent, from national governments and their representatives on the GAC.

Fourth, given ICANN's weak claims to formal legitimacy, it must be absolutely scrupulous in fairly applying its own processes. A pre-requisite to this is, needless to say, familiarity with those processes. In relation to the .XXX application, it seems that a majority of the ICANN board were unaware of the nature of the June 2005 decision for negotiations with ICM to proceed, incorrectly forming the view that the entire application was open to reconsideration at any time before the registry agreement was concluded. Over and above procedural fairness, ICANN's weak institutional legitimacy means that it must continually prove itself as a competent and effective manager of the DNS. If ICANN's performance in managing major initiatives - such as the current fraught process for introducing new gTLDs - is flawed, then ICANN's legitimacy is inevitably called into question. As one commentator has already suggested:

While private sector coordination of the DNS was always a noble goal, ICANN's handling of gTLDs - perhaps one of the most important aspects of managing the DNS - has demonstrated the organization's inability to operate without some form of governmental oversight. As the gTLD program moves forward, should problems arise that could have been addressed through the ignored advice of governments via the GAC, ICANN may face a serious threat to its role in the DNS. The failure of such a large and public program may be all that is needed for a transfer of ICANN's powers to a wholly government-run DNS management system (Duchesne 2012: 182).             

The problems with ICANN's internal structures and processes, however, do not necessarily mean that an international treaty-based or UN organisation is better than the ICANN multi-stakeholder model. As explained above, concepts of legitimacy, including international legitimacy, are by no means fixed for all time. In particular, the concept that international legitimacy can only derive from sovereign, territorially-based nation states is historically contingent. Moreover, international processes based on the need for agreement between nation states have proven to be cumbersome and incapable of dealing with endemic global problems including, but not confined to, climate change. In order to deal appropriately with serious global problems, as well as more diffuse sources of global power, in which nation states are only one among a range of influential stakeholders, it is necessary for new concepts of legitimacy to emerge.

If, as the sociologist Zygmunt Bauman suggests, what he calls 'liquid modernity' includes persistent anxiety about social structures and identities that were formerly considered to be 'solid' and certain, it follows that traditional 'solid' concepts of legitimacy are also up for grabs (Bauman 2000). In a global community where everything is in motion, and little is stable, ICANN can be seen as a paradigm of an organisation that must continually re-negotiate its legitimacy. If more traditional international structures continue to fail, and ruptures between groups of nation states ' such as that which emerged at WCIT-2012 ' become more persistent, the legitimacy of established international institutions will be increasingly called into question. While the ICANN experiment depends upon it constantly proving its 'legitimacy' through establishing and complying with stringent procedural safeguards, and effectively and efficiently performing its global role, there is nothing to say that a more traditional international organisation, such as a treaty-based body, would be fairer or more effective. After all, the ITU is necessarily involved in an exercise of continually negotiating its relevance with stakeholders, just as much as ICANN. In short, ICANN's flawed process for approving the .XXX application, and the procedural checks and balances implemented as part of that process, can be seen as an illustration of the difficulties encountered in the transition from established concepts of legitimacy to new forms of legitimacy, which are not yet fully developed. ICANN can therefore be regarded as a proving ground for key problems facing global decision-making, as in the new international order all claims to legitimacy are increasingly transient and questionable.

References 

Bauman, Zygmunt. 2000. Liquid Modernity. Cambridge: Polity Press.

Berman, HJ. 2003. Law and Revolution II: The Impact of the Protestant Reformations on the Western Legal Tradition. Cambridge, MA: Belknap Press.

Crovitz, Gordon L. 2012. America's First Big Digital Defeat. The Wall Street Journal, 16 December 2012. Accessed: 29 March 2013. Available from:http://online.wsj.com/article/SB10001424127887323981504578181533577508260.html.

Department of Commerce and ICANN. 2012. Department of Commerce/ICANN Contract for the IANA Function, Contract No. SA1301-12-CN-0035, 1 October 2012.

Dryden, Heather. 2011. GAC Advice on .XXX. 16 March 2011.

Duchesne, James N. 2012. Running Amuck: Using the Debacle of Recent Top Level Domain Expansions to Argue for Greater Governmental Participation in DNS Management. CommLaw Conspectus 21: 151-184.

Froomkin, A Michael. 2011. Almost Free: An Analysis of ICANN's 'Affirmation of Commitments'. Journal on Telecommunications & High Technology Law 9: 187-233.

Habermas, J'rgen. 1996. Between Facts and Norms. Cambridge: Polity Press.

ICANN. 2004. Ninth Status Report to the Department of Commerce. 7 April 2004.

ICANN. 2010. ICANN Options Following the IRP Declaration on ICM's .XXX Application, 26 March 2010. Accessed: 28 March 2013. Available from: http://www.icann.org/en/irp/icm-v-icann/draft-options-post-irp-declaration-26mar10-en.pdf.

ICANN. 2011. Conference Summary: ICANN San Francisco. Accessed: 28 March 2013. Available from: 

ICANN. 2012. gTLD Applicant Guidebook. Version 2012-06-04.

ICANN. 2013. ICANN's Response to YouPorn's Supplemental IRP Brief and Remedies and Other Proceedings, 25 January 2013.

ICM Registry v ICANN (International Center for Dispute Resolution). [February 19, 2010]. 49 ILM 960 (2010).

International Telecommunication Union (ITU). 2012. Final Acts of the World Conference on International Telecommunications. Dubai: 2012.

Johnstone, Rita Rodin. 2011. The .XXX Factor. Cardozo Arts and Entertainment Law Journal 29: 523-525.

Klein, H. 2004. Understanding WSIS: An Institutional Analysis of the UN World Summit on the Information Society. Information Technologies and International Development 1: 3. http://dx.doi.org/10.1162/1544752043557341 

Lindsay, David. 2007. International Domain Name Law: ICANN and the UDRP. Oxford: Hart Publishing.

Lipton, Jacqueline and Wong, Mary. 2012. Trademarks and Freedom of Expression in ICANN's New gTLD Process. Monash University Law Review 38(1): 188-227.

Manwin Licensing International v ICM Registry, United States District Court, Central District of California, Case No CV 11-9514 PSG (JCGx), August 14, 2012, Order Granting in Part and Denying in Part the Motions to Dismiss (Philip S. Gutierrez, United States District Judge).

Modak-Truran, Mark C. 2008. Beyond Theocracy and Secularism (Part I): Toward a New Paradigm for Law and Religion. Mississippi College Law Review 27(1): 159-233.

Modak-Truran, Mark C. 2013. Legitimation. In Encyclopedia of Political Thought, edited by Gibbons, Michael T; Coole, Diana; and Ferguson, Kennan. London: Wiley-Blackwell.

Mueller, Milton. 2002. Ruling the Root: internet governance and the taming of cyberspace. Cambridge, Mass: The MIT Press.

Murphy, Kevin. 2012. Court rules YouPorn can sue ICANN for alleged .xxx antitrust violations. DomainIncite, 14 August 2012. Accessed: 29 March 2013. Available from:http://domainincite.com/10149-court-rules-youporn-can-sue-icann-for-alleged-xxx-antitrust-violations.

National Telecommunication and Information Administration (NTIA). 2005. U.S. Principles on the Internet's Domain Name Addressing System. 30 June 2005. Accessed: 29 March 2013. Available from: http://www.ntia.doc.gov/other-publication/2005/us-principles-internets-domain-name-and-addressing-system.

National Telecommunication and Information Administration and ICANN. 2009. Affirmation of Commitments by the United States Department of Commerce and the Internet Corporation for Assigned Names and Numbers. 30 September 2009.

Pfanner, Eric. 2012. Message, if Murky, From U.S. to the World. The New York Times, 14 December 2012. Accessed: 28 March 2013. Available from:http://www.nytimes.com/2012/12/15/technology/in-a-huff-a-telling-us-walkout.html.

Rawls, John. 2005. Political Liberalism (expanded edition). New York: Columbia University Press.

Richards, Robert D and Calvert, Clay. 2011. Adult Websites and the Top-level Domain Debate: ICANN's Adoption of .XXX Draws Industry Ire. Cardozo Arts and Entertainment Law Journal29: 527-563.

Rosenfeld, M. 2011. Law, Justice, Democracy and the Clash of Cultures: A Pluralist Account. Cambridge: Cambridge University Press. Rosenfeld, M. 2011. Law, Justice, Democracy and the Clash of Cultures: A Pluralist Account. Cambridge: Cambridge University Press. http://dx.doi.org/10.1017/CBO9781139050982

Schiller, Dan. 2013. Masters of the Internet. Le Monde diplomatique (English Edition), February 2013.Accessed: 28 March 2013. Available from: http://mondediplo.com/2013/02/15internet.

Simma, Bruno and Alston, Philip. 1992. The Sources of Human Rights Law: Custom, Jus Cogens, and General Principles. Australian Yearbook of International Law 12: 82-104.

Waz, Joe and Weiser, Phil. 2012. Internet Governance: The Role of Multistakeholder Organizations. Journal on Telecommunications & High Technology Law 10: 331-349.

Wilson, Peter H. 2009. Europe's Tragedy: A New History of the Thirty Years War. London: Penguin Books.

Endnotes 

1. Statute of the International Court of Justice, art. 38, June 26, 1945, 59 Stat. 1055, 3 Bevans 1153.

Cite this article as: Lindsay, David. 2013. 'What do the .XXX disputes tell us about Internet governance' ICANN's legitimacy deficit in context'. Telecommunications Journal of Australia 63 (3): 33.1-33.15. Available from: http://tja.org.au. http://dx.doi.org/10.7790/tja.v63i3.432