Comments of the Internet Governance Project


Docket No. 060519136-6136-01


The Continued Transition of the Technical Coordination and Management of the Internet Domain Name and Addressing System




The Internet Governance Project (IGP) is an interdisciplinary and international collaboration of academic researchers. The Project’s partners possess scholarly and practical expertise in global governance, Internet policy, and information and communication technology. Several of our partners have extensive experience in ICANN, having served on the GNSO Council, the ICANN Nominating Committee, and various ICANN advisory committees and task forces. Other IGP partners have served in a variety of other ICT-related international organizations.

IGP commends NTIA for its opening to public comment on the future US role in managing the Internet domain name and addressing system. We are impressed with the scope and pertinence of the questions asked in the Notice of Inquiry (NOI). We have actively encouraged Internet users and stakeholders from around the world to comment in the proceeding.


The essence of our comments can be summarized as follows: The transition of DNS administration envisioned by the White Paper has gotten stuck somewhere in the middle of the process. This halfway house is an unstable and undesirable place. We urge the U.S. Commerce Department to complete the transition of DNS to a nongovernmental, multi-stakeholder regime and to clarify and limit the role of national governments, including the U.S. government. To pave the way for the transition, we ask the U.S. to ameliorate and internationalize its oversight in ways that will strengthen the accountability of ICANN to the global Internet community. We ask that two new principles related to accountability be explicitly recognized as part of ICANN’s charter. One pertains to the basic human right of freedom of expression; the other would be a general principle of accountability.


The numbered points below correspond to the numbers in the NTIA NOI.


  1. Principles
    1. Are the 1998 White Paper principles still relevant?


The White Paper articulated four principles to guide the process: 1) stability, 2) competition, 3) private, bottom-up coordination, and 4) representation. Three of the four principles (stability, competition, and representation) continue to provide ICANN and its surrounding community of interest groups with meaningful policy guidance. (We note under Heading 4 below that the principle of representation is still imperfectly realized.) The most serious problems and contradictions, however, have arisen around the principle of private, bottom-up coordination. That principle was valid when first formulated and is still important. But it has not been implemented properly and is threatened by various developments since 1998.


Private, bottom-up coordination was a crucial guiding principle in the White Paper process. The US government recognized the need to open up and internationalize participation in the technical coordination functions. It proposed to do so by relying on private, multi-stakeholder mechanisms for policy making. This path avoided traditional intergovernmental organizations and mechanisms, which were perceived as insufficiently global and too politically contentious, slow and burdensome, and in some cases as potentially threatening to the freedom and openness of the Internet.


In the original understanding at ICANN’s creation, the Commerce Department was supposed to serve as a trustee for the interests of the global Internet community while mediating the thorny and often hostile relationship between the new ICANN and VeriSign, the dominant registry operator with considerable de facto control over the root.[1] We understand the reasons why ICANN, a rather bold experiment in a new form of global governance, needed external supervision for its first few years. And we understand why the US assumed that stewardship role by default, due to its legacy role in subsidizing early TCP/IP development and establishing the IANA and Network Solutions contracts and relationships. But the special US stewardship role was supposed to end. What was supposed to be a temporary situation has now gotten stuck in place. Last year, the USG asserted for the first time a principle that it will retain a special, unilateral oversight role indefinitely. This new assertion deviates from and contradicts the White Paper principle of private, bottom up coordination.


The special authority of the US cannot be rationalized by appeals to “stability.” We know from the WGIG Report and the Tunis Agenda that many in the international community, including our allies in the European Union, strongly object to the special US role. This creates political instability around DNS. Another indicator of instability caused by U.S. control is the potential fragmentation of DNS caused by additional DNS roots in China and the Arab world. Clearly, non-US actors who feel threatened by the dominant US role can and will take steps to create alternative technical management arrangements. Further, the pre-eminent position of the U.S. government destabilizes ICANN’s policy making processes. It allows one government to interfere with ICANN decisions based on its own domestic political calculus, as occurred during the .xxx top level domain decision. Equally bad, it offers U.S.-based interest groups an opportunity to bypass or short-circuit ICANN’s organic policy making processes. It is important for ICANN’s international legitimacy that all interest groups operate on a level playing field. We also know that the lingering presence of one national government – the U.S. – in the heart of Internet administration encourages and provokes other governments to assert their own right to have the same powers. All of these consequences of the stalled transition are inimical to long term stability of the administration of Internet resources.


The NTIA process must come to grips with the inherent contradiction between the principle of “private, bottom-up coordination” and the pre-eminent position of one national government. It must also recognize that in the post-WSIS period, the changing role of ICANN’s Governmental Advisory Committee (GAC) threatens to undermine that important principle. (We comment more on that issue under heading #4.)


There are basically three ways to improve upon the current system of unilateral U.S. oversight:


  • Amelioration. This path leaves US control in place for the time being but takes concrete steps to make it more internationally acceptable. The U.S. could, for example, issue a new policy that carefully clarifies how far its assertion of unilateral authority goes, and make many of the processes surrounding the exercise of this power more transparent. For example, the U.S. could issue a formal commitment not to make any changes in the root zone file that contradict or bypass the outcomes of legitimate ICANN processes; it could institute a neutral, competitive bidding process for the award of the IANA contract; it could open up and make more transparent discussions surrounding the signing of the root zone in the implementation of DNSSEC.


  • De-nationalization. This is the original vision of ICANN. It would terminate the MoU, end unilateral U.S. oversight and complete the transition to a private sector actor under California Public Benefit Corporation law. Most advocates of this position recognize that some additional ICANN reforms are needed to ensure that the private entity remains accountable to the affected Internet stakeholders. (See our discussion under new principles, 1b, below.)


  • Internationalization. This option would involve more governments in ICANN’s supervision via a formal intergovernmental agreement(s). This could be something as narrow as a host country agreement for ICANN, or as broad as a framework convention on Internet governance that establishes international agreement on the basic public policy parameters within which ICANN operates.


IGP is on record as favoring the second option (de-nationalization) as an end state.[2] Completing the transition to an Internet identifier system coordinated by a responsible, accountable, global, nonprofit, private entity should be the goal. However, both of the other two pathways – amelioration of the U.S. role and some forms of internationalization – could and should be used to supplement the transition to a de-nationalized ICANN.


We recognize that in some quarters, options pertaining to internationalization are met with charges that administration of the Internet will be “taken over” by an inefficient and corrupt UN bureaucracy, or that Internet policies will somehow be dictated by authoritarian governments. These fears are unfounded and only impede constructive debate of the real alternatives before us.


As strong advocates of individual rights and the openness and liberty of the Internet, IGP recognizes – perhaps more clearly than many others, because we do not exempt the US government from these concerns – the dangers that governmental actions can pose to the internet. But we also understand the simple fact that cooperation and agreement among all the world’s states in some form is unavoidable if the Internet is to remain a globally interconnected and orderly communication medium. This fact is already recognized in the current ICANN regime. The GAC is open to all the world’s governments, including the authoritarian ones. Country code TLDs are derived from an international standard set by an organization with its secretariat in Geneva, Switzerland. During the World Summit on the Information Society both the U.S. and ICANN actively sought to achieve agreement and support for the ICANN regime among other states. Likewise, if ICANN or its successor is to be held accountable, multiple governments will need to agree on applicable laws and accountability mechanisms, due to the inherently global scope of Internet coordination.[3]


Thus, the issue is not whether the ICANN regime should pursue international agreement among states regarding the nature and parameters of the regime, but which substantive policies and issues it should seek agreement on.


IGP believes that the U.S. should seek international support for the original four principles, including especially the principle of private, bottom up coordination. We believe that those principles need to be supplemented with two additional principles, which are outlined in the next section. We also believe that the roles of the At Large and Governmental Advisory Committees need to be clarified and/or reformed.


    1. Should there be new principles?


Yes. We ask that two new principles be explicitly recognized as part of ICANN’s charter. One pertains to basic human rights related to freedom of expression; the other would be a general principle of accountability.


With respect to human rights, it has become evident that ICANN’s authority over Internet identifiers and domain names can be and sometimes is abused to restrict or control freedom of expression. While these cases are exceptional and mostly peripheral so far, ICANN’s technical leverage over identity on the Internet has and will probably continue to create intense pressure on ICANN to exploit that leverage to regulate expression in one way or another. We therefore believe that ICANN needs a technical coordination equivalent of a First Amendment, clearly prohibiting it from making any rule or developing any policies that have the intent or effect of suppressing specific kinds of content or virtual freedom of association. ICANN’s activities should be content neutral. Such a principle would not, of course, prevent national governments from regulating speech within their own sovereign territories, nor would it prevent the negotiation of international treaties on content-related matters. It would simply prevent the global apparatus of domain name and IP address administration from being seized and leveraged to enforce, expand or supplement any form of content regulation.


Additionally, ICANN and the surrounding community of governments and other stakeholders need to articulate a new principle regarding accountability. Such a principle would recognize that there needs to be effective external checks on ICANN’s actions and budgets once it is no longer subject to the MoU, and that certain limits needs to be imposed on ICANN’s ability to shield itself from accountability via its contracts. No single accountability mechanism is sufficient; a number of external and internal checks need to operate in tandem. In the original design of ICANN, individual membership and Board elections by those members was a key internal accountability mechanism. Those checks and balances were eliminated in the ICANN “reforms” and now the Board plays a major role in selecting its own members. The nature and form of any new external checks – or reinstitution of the old model of elections – needs to be further discussed and debated. Because of its monopoly on the DNS root ICANN might, as a condition of entry into the domain name market, use contracts to absolve itself from any liability for arbitrary or unfair actions in the selection of registries or the making of policies related to registration. ICANN should be prohibited from making itself unaccountable via contract. We find the existing accountability mechanisms of ICANN inadequate, as per the discussion under Heading 2 below.


Until these two principles are implemented, ICANN is not ready for release from the MoU. We believe that a new MoU should be prepared with a fairly short time frame (one year) and that the new MoU include instructions to ICANN to develop and implement these principles.


  1. Has ICANN accomplished the tasks set out for it in the MoU?


ICANN has accomplished most of the tasks set out in Amendment 6 of the Department of Commerce MoU, with a few notable failures.


ICANN has failed to develop a predictable, uniform and objective process for the addition of new generic top level domains. Implementation of a new process was supposed to commence by December 31, 2004. ICANN still has not defined a policy, much less a scalable procedure that could be implemented soon. Some progress toward policy has been made recently in the GNSO. But the failure to reach the milestone in the MoU is evident.


ICANN has also failed to adequately resolve the problem of creating accountability mechanisms. The recent creation of an Ombudsman’s office, though welcome in certain respects, has not provided the kind of external check on arbitrary decisions that is needed. There are two fundamental flaws in the Ombudsman’s office. One is that the Ombudsman is an employee hand-picked by ICANN’s management, and thus lacks the requisite independence. In effect, the Ombudsman is accountable to ICANN and not to those who are affected by ICANN’s decisions. Second, the Ombudsman has defined the scope of his decisions and mediation processes too narrowly. This means that some of the most important accountability issues do not fall within his scope. The independent review process is as yet untested, but it too seems inadequate. It is too much an internal part of the ICANN organization, in part because the ombudsman has to agree to initiate the independent review process. There is at least one case where he refused to do so. There are also structural problems related to limiting ICANN’s budget.



  1. Are the MoU tasks and milestones still relevant? Should new or revised tasks/methods be considered in order for the transition to occur?  On what time frame?


In the NOI this heading mentions both the June 2005 US principles and the DNS “transition.” As noted above, there is a direct contradiction between the US principles, which contemplate indefinite U.S. governmental supervision of DNS management, and the concept of a transition to a privatized DNS administration. Either we are making a transition to a nongovernmental system or we are stuck in a US-controlled regime. We cannot have it both ways.


With that caveat in mind, we believe that the two new principles we proposed in heading 2 should become new milestones, and that the new TLD process should be retained as a milestone. (In fact, the accountability principle may or may not be considered a new milestone, as it maps well onto section II.C.4 of Amendment 6 of the existing MoU).


We believe that the time frame for accomplishment of these milestones should be short – as a starting suggestion, one year – but it is more important to do it right than to do it quickly.



  1. Are stakeholder groups involved effectively? Are there additional stakeholder groups?


There are important issues about two stakeholder groups: individual Internet users and governments.


    1. Individual Users

Individual Internet users are not involved effectively. The At Large Advisory Committee (ALAC) is supposed to represent the individual Internet user. Its proposed structure under the “reformed” ICANN is not workable. Although the main rationale for creating an ALAC was to represent individuals, individual users are excluded from the new ALAC. ALAC’s focus on noncommercial organizations overlaps in an unproductive way with the remit of the GNSO’s Noncommercial Users Constituency. The interim ALAC board's terms are indefinite, and several of them are appointed by the board itself. The process and budget of ALAC are not transparent.


The proposed “Regional At Large Organization” (RALO) structure, which is supposed to be a long term solution to the representation problem for users, imposes impossibly heavy organizational overhead on individual representation. The RALO concept is completely at odds with basic, well established facts about representation and public activity. Large numbers of stakeholders with small stakes in the outcome of public policies face what is known in social science as a collective action problem. No user or domain name holder individually has enough at stake to justify intervening systematically in ICANN policy making processes. Only coordinated action can have an effect. Yet coordinating the action of a large number of people imposes costs on the organizers that may exceed any possible benefits; it also faces the problem of “free riders” – people who can benefit from collective actions while contributing nothing. These collection action problems are compounded, not avoided, by ICANN’s proposed ALAC structure, which asks individual users to invest enormous resources building regional organizations and coordinating them before they can achieve effective input.


As a purely practical matter, a return to direct election of Board members should be considered as a way of solving the problem of broad user representation. Elections provide a low-cost way for individuals to exercise input. One of the main obstacles to election implementation the first time around was cost, but ICANN’s budget now contains enough capacity to make that less of an issue.


    1. Governments

The Governmental Advisory Committee (GAC) was intended to be a purely advisory body that allowed decision making within ICANN to be well-informed about the interests and needs of governments and national and international policies. Since WSIS, a forum in which many national governments asserted their interest in formulating “public policy” for the Internet, GAC has loomed much larger in the ICANN policy process. At times, it has attempted to participate actively in policy formulation. At other times it has appeared to seek some kind of veto or last word on policies. The latter tendency has been reinforced recently by the willingness of the US and certain other governments to use GAC to promote their own policy preferences, particularly in the .xxx and Whois issues. But in reality, governments do not and usually cannot speak with a unified voice. The GAC often contains a collection of 40 to 50 different public policies. While it is useful and important for GAC to make policy developers within ICANN aware of these diverse perspectives, it is not legitimate for a few governments within it to arrogate to themselves the right to speak for all governments or to formulate a pseudo-global public policy. We fear that GAC is evolving into a parallel policy development process within ICANN, introducing uncertainty, delay, contradictions, and political gamesmanship into the process.


Our advice is to scale back GAC to the strictly limited, advisory role originally contemplated. If governments want to formulate and enforce “public policy” at the global level, they must do so through legitimate, accountable means: by openly negotiating formal international treaties and gaining ratification of those treaties in their legislatures, or work through established intergovernmental organizations. Only these kinds of interventions by governments should be able to supersede the actions of ICANN. We must be careful that government actions at the international level are constrained by democratic process. GAC should not evolve into a parallel and competing policy making process within ICANN.


There is an important role for governments in a fully privatized regime. Governments must ensure that whatever a de-nationalized ICANN does is consistent with applicable national and international laws. Governments should not attempt to interfere in the day to day workings of the privatized entity or to actively shape its decisions, but rather should act as an external, ex post check on ICANN. GAC lacks the process, the legitimacy, the quickness and often the expertise to formulate global policies for the coordination of internet identifiers.



  1. Are ICANN SOs doing their job and functioning properly?


No comments.


  1. Automation of request processing for root zone changes.


IGP supports the idea of permitting responsible parties who have been delegated ccTLDs or gTLDs to automatically update their entries in the root zone file, without DoC intervention. We note that DNSSEC implementation requires some kind of rapid and automated process for rollover of keys to sign zones. Obviously, implementation of any such system would require careful attention to security issues but does not seem to pose an extraordinary problem.


  1. How can information exchange, collaboration and enhanced cooperation among these organizations be achieved as called for by the WSIS?


Information exchange, collaboration and enhanced cooperation can be achieved in two ways. First, the US can make it clear that it intends to complete the transition to a privatized DNS and pursue the kinds of amelioration and internationalization of its unilateral authority suggested under Heading 1. Second, it can participate in good faith in the new UN Internet Governance Forum. The U.S. should embrace the Forum as a venue where cross-cutting issues are considered and international and intergovernmental organizations, including ICANN, are subjected to independent assessment on the basis of their contribution to development and human rights. Internet governance involves a heterogeneous array of public and private sector mechanisms that address different pieces of the puzzle. The Forum can view these holistically and assess all the important issues horizontally in terms of their compliance with WSIS and other public interest principles. This would allow us to identify best practices and encourage (not command) the institutions that fall short to move closer toward reasonably coherent standards of good governance in the communication-information policy domain.




[1] See Testimony of J. Beckwith Burr, Wilmer Cutler Pickering Hale and Dorr LLP House Small Business Committee Wednesday, June 7, 2006.


[2] Internet Governance Project, "Political Oversight of ICANN: A Briefing for the WSIS Summit" (November 1, 2005). Internet Governance Project. Paper IGP05-009. Available at

[3] As an example, the applicability of antitrust laws and principles to ICANN activities would be a fruitful area for intergovernmental negotiations and agreement.