Timothy M. Denton

Success Through Understanding Technology

The Multi Stakeholder Approach to Policy Development

The Multi-stakeholder Approach to Policy Development


Every few hundred years or so, something occurs that transforms how people live. Sometimes the event is a technological change, sometimes it is war or revolution, or a plague, or a discovery. The printing press was one such change. The line between the invention of the printing press, the publishing of the Bible into vernacular languages, and the Reformation is clear. The Internet appears to be of equivalent importance to printing, though whether its consequences will matter as much as the Reformation, is as yet unknown.

The immediate effect of the Internet’s distributed architecture, uniform standards and globally unique addressing system was a disparity between its globe-embracing reach and the jurisdiction of the territorial state. Rather than having its management handed over to multi-lateral organs of states, the management of the Internet’s core addressing functions was entrusted to something novel: the multi-stakeholder (MSH) institution. As the Internet is principally a method of linking computers through trusted protocols and unique addresses, it made sense to adopt styles of decision-making borrowed from the world of engineering standards. These are quite distinct from state-based forms, as will be seen.

The multi-stakeholder (hereafter MSH) process is, in the view of many, a genuine innovation and adds to the repertory of ways in which decisions can be taken, consensus formed or advice developed. Some hope its advent may cause us to change the processes by which we habitually make decisions, including who makes them.

Normally decisions are made by politicians, law-makers, bureaucrats, judges, regulators and others invested with the power of the sovereign, or of the state. Citizens are normally supplicants before the organs of the state. Special procedures and specialized personnel, such as lawyers, are normally used to address courts and regulators. While elected officials are sensitive to the needs of their electorates, there are large areas of government where the citizen addresses the official not as an equal participant, but as a petitioner, if at all.

The Sovereign, or his representatives, is in the room. He or she can be represented by any number of titles and persons, but is always there. The focus is on the Decider, whoever he or she or they may be.

What if the Sovereign got out of the room for a while? What if all the players discussed face to face until a consensus emerged? Would this lead to greater legitimacy in the decisions and outcomes? Would it be more effective than traditional institutional outcomes? Are there circumstances in which a multi-stakeholder approach might be preferred to others? What are they?

This report explores these questions.


What this report is (and is not) about

This report is the result of my reading on the subject and of interviewing selected practitioners and authorities who have been associated with Internet or state-based governmental institutions, often both. The work occurred in the late spring and summer of 2017.

The report is intended as a pragmatic guide to what an MSH process consists of, and when and where it may be appropriate to use the MSH approach, as distinguished from others. It will seek to explain what the MSH process is and distinguish it, as much as may be sensible, from other, more well-known, types of consultation and decision-making.

I follow Aristotle’s dictum that one should pursue a subject with no more precision than the subject can bear. It is my belief that MSH processes can be identified by a cluster of attributes, so that an MSH process may be distinguished from normal interest group consultation. Yet as we are dealing with an emerging phenomenon, we must avoid treating MSH processes as if they were an ideal Platonic form: immaculate, mathematical, and never to be confused with some other type of consultative or deliberative process.

The issues under consideration are whether MSH processes are sufficiently distinguishable from other forms of state-based rule making and consultation, and whether they have been or can be more effective compared to those other forms. The first question can be answered clearly in the affirmative. To the second question, the answer is: it depends on the goal of the convenor of the process.

  • Multistakeholderism is a process, not an ideology

Milton Mueller wrote that, “as an ideology that can guide change, multistakeholderism is both radically incomplete and flawed” because “it does not provide any guidance on the substantive policy issues of Internet governance”.[1] I agree with Professor Mueller in the sense in which he speaks. I agree that if one is to consider the MSH approach to be an ideology, one would be walking on air. The subject of this report is not however an ideology but a type of process for arriving at consensus among a broad range of interested parties and facilitating better decisions. Is it sufficiently definable that we may speak of it meaningfully? I consider it to be so.

The MSH approach does not provide any guidance on the substantive issues it may deal with, unless you happen to consider the process of how you reach decisions to be a question of substance. In some cases, it is.

In this regard Laura DeNardis has highly useful observations:

It is important to view multistakeholderism not as a value in itself applied universally but as a question of what form of administration is necessary in any given context. Certain areas of Internet governance should jurisdictionally be overseen by national governments or via international treaties. Other areas are effectively administre3d by the private sector and non-profit institutions. In these cases, transparency and accountability are values necessary to create the legitimacy for those groups to establish public policy for the Internet.[2]

•           The cluster of attributes of a multi-stakeholder process can be identified

The MSH approach is a process for seeking consensus and gathering evidence for decisions.

When the substantive issues are not clear, where discussion is needed even to define the order of priorities, let alone conclude on specific points, the MSH process shows itself to have several advantages over more state-directed or lawyer-dominated forms of consultation and decision making. This was the observation of a considerable proportion of my interviewees.

The MSH process has had certain successes when the structures of how to reach consensus were not already in place, or where the challenge was reaching consensus in a novel situation, such as relative statelessness of Internet rule-making. It is also highly appropriate where, as in the case of much of the Internet, no one government may have full jurisdiction, does not know enough of the implications, or where it shares effective power with other private sector actors who alone can implement a decision. Where an apparatus for making decisions already exists, such as a regulatory agency, the MSH process can offer advantages where the fundamental issue is whether consensus can be achieved, provided that the divisions among the stakeholders are not insuperable. Even if the MSH process does not succeed in generating consensus, it can identify issues with considerable precision, as well as who takes what positions on them..

  • The MSH process may be used inside territorial jurisdictions

The question has since arisen, whether the MSH process may have advantages in some cases, even where the Sovereign may act through its normal processes to arrive at a decision.

“If multistakeholderism means only that people who are strongly impacted by policies should be actively heard from, then it is nothing but normal pluralist politics”[3]

To which I would respond, firstly, that normal pluralist politics may benefit from improved processes. The question is whether there is the likelihood of better outcomes when an MSH process is essayed rather than other consultative rule-making processes. With significant qualifications, my answer would be yes.

The MSH process has proved to have particular advantages when, as in the case of much pertaining to the Internet, the very structures of decision-making and consensus formation had to be created from scratch. It also has advantages when the issue is not restricted to making the particular set of rules and decisions, but where consensus about the need for rules and decisions has not yet agreed upon.

The MSH process has been particularly useful where the multi-national institutions that would give it cover did not exist, or operated in such a way as to exclude all stakeholders but states from the deliberations. This does not imply that an MSH process would be less effective within a given state’s jurisdiction. Inside a given jurisdiction, players in an MSH process would usually have a common culture, legal framework and a set of expectations that would assist, rather than hinder, the kinds of consensus that an MSH process seeks to arrive at. Thus MSH processes could be employed inside a national or state jurisdiction, possibly more easily than under an international umbrella.

  • The MSH process is neither a chimera nor a myth – there are conditions whereby it can be a success, and has been a success

I have also read some academics describe the MSH process as a chimera and a myth. Some have pointed out that tri-partite processes of labour, management and state have existed since the 19th century. [Nothing to see here, move on]. I doubt that anything written here will soften their views.

For myself and for most of the interviewees in this project, the MSH process has come to prominence because the Internet immediately linked the world in a framework of global communications, detached from the nation state, so that the institutions that were appropriate to governing it had to be improvised on the fly.

As it turned out, the natural rulers of the Internet’s core functions were engineers and computer scientists, and they had their own ways of sorting issues that had nothing to do with sovereignty or powers of compulsion.

It may be true that something like MSH processes pre-existed the Internet. It is also evident that the MSH process can be vitiated in any number of ways, which we will discuss. Nor do we ever escape the not-so-hidden role of Leviathan at the beginning and end, and sometimes the middle of the MSH process. One school will assert that MSH processes are chimerical because they can be vitiated in any number of ways. Others will say that, because you never get away from the state’s deciding role, MSH processes amount to a group delusion. I do not share those attitudes, and nor did most of my interviewees, however cynical and experienced they are in politics and MSH processes.

Most of the people interviewed belong to the cautious, realistic and qualified school of multi-stakeholderism. Even the deep skeptics acknowledged its existence, its features and its relative novelty, albeit with less belief that it would ultimately be all that significant.

My contentions are that

  1. the MSH approach has a definable set of characteristics and
  2. these characteristics are sufficiently distinguished from normal processes of rule-making that the MSH process deserves to be considered a new thing in the world.
  3. A process need not satisfy every criterion to be considered “multistakeholder”, but it should satisfy many of them
  4. The effectiveness of the MSH process depends on the type of problem the Convenor of the process is trying to solve.

The MSH process can be situated on a sliding scale from complete lawyerization, judicialization, compulsory power and regard for precedent at one end, to complete lack of compulsion, lack of sovereignty, and forward orientation at the other.

These contentions will not satisfy those who seek ideal forms or who insist on comprehensive conceptual clarity. Many of my interlocutors insisted that there was something fuzzy at the core of the MSH idea, and I agree with them. But fuzziness is not an insuperable bar to something being useful or being real. Others pointed out that there was no agreed definition of multi-stakeholder governance. Yet from the inception of international forums discussing the Internet[4], the term “multistakeholder” has been used, and the legitimacy of business, technical, and civil society membership in these forums has been accepted. So is everyone just blowing smoke, or is there substance to the MSH concept? The interviewees in this consultation were persuaded they were speaking of real, observed and describable processes.

The following section tries to elucidate what those criteria are, and to see whether and how an MSH process is properly or usefully distinguished from other forms of state based rule making, standards setting, and law making.

Traditional Forms of Decision Making

Traditional forms of decision-making most often involve the state, though not in all cases. They include

  1. Law making
  2. Subordinate regulations passed in virtue of laws
  3. Judicial processes
  4. Commissions of inquiry
  5. Economic regulation, such as regulation constraining actors thought to have market power, or those with economic privileges
  6. International treaty-based forums
  7. Standards setting

Of the seven forms listed above, only in standards setting is it possible for the state to accept the results of private decisions without deciding the issues for other players.

The nature and extent of the role of the state in the other six forms of decision-making distinguishes standards-setting and an MSH process from the others.

It will be my contention that the role of the state in MSH processes is, or at least should be, one of limited and very careful intervention and – most important - non-intervention.

The various forums for rule-making can be characterized by the following:

  • Who is qualified to speak? What are the restrictions on who speaks?
  • To whom is the speech directed? Who must be persuaded?
  • Is there an audience? Is the audience a participant, or is it passive?

Traditional decision-making and rule-making forums are characterized by

  • Formal legal procedures, with elaborate rules governing the presentation of evidence, the timing of responses to the acts of other parties, and well-understood legal criteria by which arguments will be judged.
  • Lawyerization – specialist presenters are relied upon to present argument or guide the presentation of argument. Frequently, only licenced professionals, usually lawyers, may be employed to make arguments. At other times, bureaucrats may substitute for lawyers.
  • In state-based treaty organizations, only states and their representatives are qualified to speak. Private sector, commercial, or non-state actors speak by permission only. Frequently private actors have to petition the state to accredit them as representatives of states in order to be allowed to speak.
  • Precedent: what has gone before, whether by way of process or by way of previous substantial decisions, has authority over the proceedings conducted in the present.
  • Speech and argument are frequently directed exclusively to the representatives of the sovereign in the room: the judges, the regulators, or the officials. Argument is from many parties to a limited set of people who participate in a decision-making capacity, but not as equals or near equals.
  • There is an audience, but the audience is not a participant. It is distinctly subordinate, and may be told to clear the room or to be quiet. The audience is there to see justice done, or to report back to principals and superiors, but it has no powers of intervention (save for largely involuntary reactions like gasps or laughter).


Law-making is so different in parliamentary systems compared to presidential systems that it is difficult to summarize their features. In many systems, legislators actually create the texts of laws, such as the US Congress does, with the help of staff. In others, legislation is drafted by civil servants after extensive private consultations with stake-holders. But whether legislation is drafted by elected politicians and their staff, or ministers and their staff, the will of the legislature is supreme, however it may be expressed, and whoever prepares the texts of the statute. Leviathan is in the room, deliberating with parts of itself, and outsiders are petitioners first, and subjects always. The petitioners, or lobbyists if you prefer, are addressing the state, not each other, and no one is seeking consensus, when a majority of legislators will suffice. The decisions of legislatures carry all the authority of the sovereign state, and may compel obedience.


Engineering Standards processes

The first thing to be said about multi-stakeholder processes is that they have emerged, in the Internet context at least, from the kind of discussion forums dominated by engineers, rather than by lawyers. From this difference much else follows.

The Internet is the creation of engineering standards for the exchange of information between Autonomous Systems – computers under common control. This will surprise many people who imagine the Internet is something physical, or the information carried by networks according to standards, but not that the Internet really is a collection of standards. The need for globally unique identifiers meant that there would be a centrally coordinated administration of address assignments. Domain names added another central point of management. Domain names and addresses at first were managed by a single individual working on a set of rules developed as needed, but soon came to be managed by not for profit corporations, often membership-driven. The institutions concerned in the management of the Internet’s central resources and standards -names, numbers and protocol parameters – were from the start composed of commercial interests of various kinds.

The Internet is inherently not bound to the territorial borders of states, nor was it the creation of lawyers or treaties. It leaped ahead and outside of the state system from the moment of its creation. It was the creation of computer scientists. For the engineers, the model of governance at hand was the way they have habitually settled technical issues.

The most perfect example of the MSH form of rule-making is the Internet Engineering Task Force. It is governed by expertise and status accorded to engineering-scientific talent: what the political scientists call an epistemic [knowledge] community.

  • Little or no restriction applies formally to who may attend or participate, save the cost of getting to meetings or participating electronically.
  • It is expected that commercial entities will be present. Commercial interests are held to be legitimate, and normal.
  • State-based entities have no extra status, nor are they given more weight than their technical arguments may merit.
  • Precedent may have some persuasive force in how procedural disputes are to be handled, but the focus is on designing the future, not conforming to the past. Justice is not being sought, good design is.
  • There is no hard and fast distinction between the audience and the participants, indeed, everyone present is expected to contribute.
  • There is no specialist professional class of arguers and no specialist class of deciders.
  • Consensus need not be perfect but, in the absence of authoritative rule-makers and opinion deciders, consensus is as good as it gets. Hence “rough consensus and running code” expresses a deep truth about how the IETF works.
  • Absence of agreement and other problems can be handled by the establishment of other working groups, by sidelining the troublesome in irrelevant groups, and by abandoning the idea. There is no force other than self-interest that compels agreement.
  • Standards have no legal force; they are accepted or they are not, ignored or adopted, as the interests of the parties may direct.

I cite Dick Beaird on this subject: The IETF participants “are assisted by the fact they are operating within the laws of physics”[5]. We now turn to those processes where the laws of physics do not necessarily set the outer boundaries.

Multi-Lateral processes

Multi-lateral organizations are those which have been set up by states, which are conceived as the sole authorized representatives of power in exclusive territorial units. Private interests are permitted to join delegations and speak only with the permission of their respective governments. Thus while their processes need not be dominated by lawyers, and argument is not directed to a single sovereign with deciding power, multi-lateral organizations are the organizations we have evolved to let Sovereigns speak in the same room.

Multi-lateral processes, I will venture to say, resemble multi-stakeholder processes in one essential regard: all states are equal, but some are more important than others. They are in this sense slightly akin multi-stakeholder processes, though with only one kind of stakeholder allowed, sovereigns.

Multi-stakeholder (MSH) processes

Where do we locate the MSH process in a spectrum from court-room trials at one end and places like the IETF at the other?

It is evident that an MSH process partakes of many of the attributes of an engineering standards forum, and is furthest removed from a trial in a court room. It sits closest to the engineering standards process in the linear spectrum. The properties that make it so include:


  • freedom to speak, write and otherwise contribute
  • speech is not organized around specialized interlocutors, such as lawyers
  • speech is not exclusively addressed to agents of the sovereign in the room
  • records of contributions, emails, are open and accessible
  • meetings are announced, secret conclaves are generally shunned


  • stakeholders self-identify as such
  • there may be a pre-existing set of stakeholder constituencies that self-organize, elect their chairmen and officers, and caucus within the constituency first before seeking consensus with other constituencies
  • in the absence of such pre-existing constituency organizations, it may be necessary to allow time for them to self-organize

Role of government

The MSH process is not like an engineering standards process in several ways:

  • It has been convened for a purpose by an outside, usually governmental, authority
  • While the agenda of the MSH process may be refined by the stakeholders, and otherwise adjusted, the convenor may want a decision or a consultation, or the development of consensus, to occur
  • The convenor will probably want to establish in advance of the process boundary conditions for the range of acceptable outcomes
  • The convenor will want to strike a balance between clarity in range of acceptable outcomes (his purposes), and enticing the major powers among the stakeholders to attend (their incentives to attend and take the process seriously)
  • The convenor may also play a role in recognizing some groups and not others, or subsidizing some groups to attend and not others, or favouring the election of some people and not others as agenda setters or chairmen of constituency groups. [The convenor does so at considerable risk].
  • The convenor, or some trusted body, will pay for the maintaining email listserves, conference calling facilities, and other forms of connection and record-keeping.



I observe that the MSH process borrows heavily from the style of proceedings in engineering forums like the IETF. With important qualifications, participation is more open than in any forum more dominated by legal process.

Participation is not reserved for special professions and presenters of argument; there is no audience - there are in theory only participants; commercial interests are accepted as normal, diverse and competing points of view are argued without privileges for some points of view. Truly stupid points of view are shunned.

The stakeholders

In Internet circles, the composition of the stakeholders has included representatives of the technical communities, businesses, civil society groups, and governments. It must be emphasized that the parties with direct commercial interests in the outcome cannot be excluded. It is they who must agree to whatever constraints that might be put upon their profits or property rights[6].

Several interviewees emphasized that in an MSH process, all were included, though not all were equal. Inequality is in part determined by the degree to which a party may be impacted by the consensus or may have to be relied upon to implement the agreed upon thing

We shall address the question of representation and representivity further in this paper. Nevertheless, any issues that might exist in an international forum, including a global forum, shrink to relative insignificance if an MSH process is confined to the territory of a nation-state.

In Internet circles, the groups that have had influence and which must be persuaded to participate have included:

  1. A technical community that develops the software standards and applications for a digital environment. They include the IETF, the W3C, and the Internet Society, and should include any of the many developers who inhabit the technical forums of the Internet.
  2. Advocacy groups focused on digital rights, such as for privacy, or affordable access and connectivity.
  3. Internet-based businesses, such as content providers (Netflix, Google), Internet service providers of all sizes. Generally, these groups favour global access, interoperability and open markets.
  4. Various institutional actors may also be included if they show or have shown interest in the particular Internet-related issues being addressed by the MSH process: the Regional Internet Registries, ICANN, the OECD.

These are rough groupings. No definitive list can or should be devised, as the intention of an MSH process is to include a wide enough spectrum of interests that the resulting consultation, or decision, benefits from the legitimacy imparted to the process from the inclusion and participation of the relevant actors.

In short, the invitations may go out, but who shows up will be out of the control of the Convenor, and ought to be.

Representation and Representivity

A word about representivity, which concerns the adequacy of representation. The legitimacy of deliberations and proceedings is frequently challenged when the groups partaking are held not to be sufficiently representative. Failure to represent can take several forms. A group can be held not to represent the full scope of affected stakeholders or held not to actually represent or to be acknowledged to represent the stakeholders that they claim to represent. It will be a delicate practical problem for the Convenor to decide how many to invite to the MSH process, or whether anyone can be excluded. Something artful can be arranged to ensure some groups show up without necessarily excluding others. Arithmetically it is impossible that a smaller group of any description whatever (the representatives) can be as diverse as the larger groups (the represented). Representation substitutes for participation by all conceivable groups, and it cannot be otherwise.

Second, an engineering standards process is less beholden than some others to geographical, sexual, cultural, regional, or religious representation. Concerns for representivity are bounded because the organization as a whole is concerned with the laws of physics and whether and how problems can be solved by better or worse technical ideas. It might split on Apple versus Microsoft lines, but is less likely to split on the other considerations mentioned, than bodies not bound to the laws of physics

An MSH process has to be concerned with more than technical competence. It has to devise politically and technically acceptable solutions or compromizes. The ruled have to give their qualified blessings to the rules. The legitimacy of the process generates public acceptance of the outcomes. But representation, particularly of consumers and the public, will not be perfect. Hence the importance of getting the right people in the room but not excluding others.

Records must be kept

Another important attribute of openness or transparency is the degree to which all meetings are open and ideally receptive to remote participation; that all discussions are recorded, that a trail of email or other documents is available and searchable; that email lists may be freely joined; that a secretariat exists and acts neutrally among parties. Keeping track of the discussions and recording who said what, when is an essential element of the MSH process. Multistakeholder processes without adequate records to demonstrate the openness would fail the critical test: do participants have confidence in the process?

Open records and invitations to meetings also allow progress to occur. If a group of thirty or so people has had an open meeting, kept recorded discussion or discussion thread, and come to a conclusion, it is then possible for the group or group leaders to tell a dissident that the discussion is closed. The issue may be raised again in the general assembly at any time by any participant. But the group must have power to declare that a particular point has been settled, if indeed this is the case.

Caucuses, Constituencies and Sub-Groups

It may be necessary to encourage groups to find ways of convening in caucuses or interest groups. This will have the effect of causing discussion and resolution of issues to take place within kindred interest groups. It is more likely that interests can be aligned within caucuses or subsections of the whole that profess similar ideals and goals

The formation of caucuses or sub-groups within the general assembly is a natural event. The IANA transition, for example, benefited from the pre-existence of constituencies and working groups within the ICANN structure which had had long experience in electing their leaderships, keeping minutes and discussion threads, maintaining websites, and working with other constituencies within the ICANN structure. Leaders, positions and personalities were known to each other.

One of the significant difficulties I can envisage is that, if an MSH process is convened and there has not been experience among the stakeholder groups in forming caucuses aligned on mutual interests, it may take considerable time for some of the groups to self-organize. The IANA transition process was delayed in part because one of ICANN’s constituencies needed a long time to self-organize, and this was so because it had not previously been in the habit of working together.

The other advantage of constituencies within an MSH process is that it leads to the creation of a set of chairmen or spokesmen who can work collectively to manage issues and drive progress. Internal decision-making structures within a constituency allow for the discussion and resolution of issues in manageable bits. The leaders of such groups, having been selected by their constituencies for that role, are in the position to make the necessary trade-offs and compromizes that constitute politics.

The role of the state

Where the MSH process differs from engineering standards-setting is that the state has most often played a part in setting up the process, in assisting some groups to be present and not others. Above all, some agent of the state has probably selected the boundary conditions of the MSH process. These boundary conditions include such matters as:

  • The reasons why an MSH process was selected.
  • The goals that are expected to be achieved by the process; the criteria of success or failure.
  • The outcomes that are considered unacceptable to the Convenor

The question that arises throughout an MSH process is the degree of responsibility left to the participants. The autonomy granted by government to the MSH process is the most delicate and important matter, in my view. Larry Strickling, former head of the NTIA and a strong proponent of MSH processes,[1]told me that they key question for participants is: Are we involved in a consultation? Or making a recommendation? Or in making an actual decision?

Where has decision-making been vested? If decision making has been vested with the group or groups in the process, then everyone is forced to the table. It is vital, he said, that there be no separate channel for a collateral attack at the end of the process.

The art of holding an MSH process is to give enough incentive to the participants to commit at the front end.

Yet, numerous interviewees emphasized that one never gets completely away from the state. The entire process has been set up for a reason: either a consultation or a decision. Yet, as we saw from the IANA transition process, the state in the person of Larry Strickling of the NTIA established boundary conditions: conditions which, if met, would cause him to accept the results, and if not met, would have caused him to maintain US hegemony over the IANA functions contract.

It follows that an MSH process does not exclude the state or states from the process. Here we must be careful about the intentions of the state, which remains in most cases the decider. From my discussions with participants in various Internet forums, an MSH process is as much about the intention of a government which might engage in it as any other factor. Government capture of the process might wreck it. But lack of government participation, as sponsor, as listener to the results, would obviate the incentive of the stakeholders to participate.

We need to be careful about how we speak of the participation of states in an MSH process. Two cases present themselves: inside a nation-state, and internationally.

If an MSH process is being conducted inside a national jurisdiction, it is much easier for representatives of the government holding the process to leave the room, and let the other players sort out the issue. Or the actual deciders can leave the room while possibly other portions or sectors of government might participate, if that were found agreeable to the government holding the MSH process to act that way. In either case the stakeholders are not required to deal with states inside the MSH process, where all players have equal rights but some players are more important than others.

If the context of an MSH process is international, so that multiple states are involved, then we have a quite different situation. The most important recent example was the MSH process on the IANA transition. There the United States held a unique role as the lone state with authority over the IANA. Other governments assembled in the Government Advisory Committee (the GAC), which acted within the bounds of the IANA transition process. Now that the IANA has transitioned away from US jurisdiction, it is difficult to imagine the circumstances in which one state will ever again hold formal decision-making powers over other states in an Internet body.


You do not finally escape the state

The idea that the state might be the Convenor of an MSH process, and the ultimate decider of its outcome, will be found unsatisfactory by some. A libertarian might challenge the idea that an agency of the state might set terms for the MSH process. Is not the whole purpose of the MSH process to replace state-based decision making?

Not necessarily. If the MSH process offers superior chances for the creation of trusted consensus solutions, relative to a normal regulatory process, then despite all its drawbacks: duration, complication, fuzziness of outcomes, its openness to gaming of various kinds; then an MSH process may offer to governments a better way to formulate policy.

As Dick Beaird told me “The multi-stakeholder process is a form of public, transparent consultancy. You don’t escape government… I don’t know how you escape the authority question”[7]

We must be careful about too much categorical thinking in relation to the role of the state in MSH processes. Leviathan may leave the room for a while or engage in much less hierarchical behavior within the room, in order for the process to work out. In the words of Robert McDowell, a former FCC commissioner, “Governments should not own the table”.[8] That expression captures well the sense of government pulling back from framing all questions, and determining all matters, and sharing with other players, in some smaller or greater measure, responsibility for the outcomes.


Drawbacks, disadvantages

My interviewees pointed to the following issues, some of which are inescapable. Quotation marks indicate direct statements by an interviewee. As a rule I have kept them anonymous for reasons that will appear obvious.

  • The MSH process cannot cause people to come to agreement when their interests are sufficiently at variance.
  • There will be cultural attitudes that prevent agreement on certain issues. Differing approaches to privacy in the United States and Europe were cited as an example of an unbridgeable gap.
  • The MSH process is consensus driven, where consensus is hard to achieve.
  • “The MSH process is not inherently more just or moral than others which are more directly governmental in nature.”
  • “The Geneva types can turn it (the MSH process) into a representivity game”.
  • “Its frailty is the tyranny of the minority”.
  • Nothing says that stakeholders are of equal importance. Power is asymmetrical.
  • Those with money and expertise will always be advantaged.
  • You must have clearly identified stakeholder constituencies/subgroups in order to narrow the range of outcomes and expedite decision-making.
  • “Don’t discount the value of something someone says because of the interests he represents”
  • “It is a challenge when one of the idiots is a major player or a state.”
  • “The pace of MSH processes can be glacial. They tend to self-complicate, at a fractal pace. Each working group uncovers new problems”.
  • “Not every decimal point should be chased. A good-enough solution can be right in front of you”.
  • There will be great reluctance to admit failure. In some cases the inability to declare failure will drag the process on.
  • MSH process enthusiasts believe it is inclusive but it is not; those with money and time can show up and build coalitions; those who cannot afford it are out of the game.
  • “If you can’t declare the conditions for failure or success at the outset, you can’t succeed.”
  • “The MSH process can work only with a modicum of trust and respect.”




In the MSH process, the means whereby we arrive at a rule is the process of legitimizing its acceptance. In certain novel situations that the Internet has generated, where there was neither a pre-existing forum nor an agreed agenda, the use of the MSH forum has bestowed legitimacy on decisions that could not otherwise have been taken.

The promise of the MSH process is that when states will be faced with similar problems: incomplete jurisdiction, inadequate understanding of the nature and priority of the problems, a need to engage the public in the possible solutions, and a degree of technical complexity that challenges the knowledge base of most participants, the MSH process will prove suitable.

The MSH process is not swifter than a process of inquiry by an efficient investigator, it is not as knowledge-driven as an engineering standards forum, it is not as concerned with the protection of the weak by the operation of law as is a trial, and its standards of evidence may be rougher- hewn. Nevertheless, the successful MSH process has the effect of legitimizing its outputs, and narrowing differences, if the Convenor can have the wit and patience to frame the conditions for success and then let the parties engage in it.

The Convenor has a delicate task, to attract the kinds of parties to the process that can adequately deal with the topic, and to leave enough freedom to the assembly to devise the outcomes in the confidence that the stakeholders will be heeded. Thus the Convenor must find terms on which he can bind himself to accept a range outcomes, and leave no incentive to any party to go around the process. If these conditions can be met, the MSH process offers a useful method for addressing complex and novel problems.



Table for the Convenor’s Consideration

  Who defines the issue

Who is affected by the issue

Who needs to be involved in the issue Who decides the issue Who implements the decision
Private Sector          
Technical experts          
Civil Society          

Appendix A



[1] See the list of interviewees given in Appendix A for details of the person and the interview date



1.         Bill Graham, Victoria BC, former writer for the Internet Governance project, June 16, 2017 and 27 August, 2017, This email address is being protected from spambots. You need JavaScript enabled to view it., 778 433 3539 (My thanks to Bill for his careful comments,from which the written report has greatly benefitted).

2.         Professor Jeanette Hofmann, Professor, Internet politics, Freie Universität Berlin, tel. 49 30 25491 288, June 16, 2017

3.         Mike Godwin, R Street, Washington DC, 1050 17St NW Washington DC, Suite 1150, This email address is being protected from spambots. You need JavaScript enabled to view it., 202 763 3307 20 June, 2017

4.         Larry Strickling, former head of NTIA, This email address is being protected from spambots. You need JavaScript enabled to view it. , 312 343 1500, 21 June 2017

5.         Daniel Sepulveda, former US Ambassador to ITU and internet institutions for Obama regime, 21 June 2017, This email address is being protected from spambots. You need JavaScript enabled to view it.

6.         Sally Wentworth, Internet Society, This email address is being protected from spambots. You need JavaScript enabled to view it. tel 703 439 2146, 22 June, 2017

7.         Ambassador David Gross, This email address is being protected from spambots. You need JavaScript enabled to view it., 202 719 7414, 5 July 2017

8.         Byron Holland, President of CIRA, This email address is being protected from spambots. You need JavaScript enabled to view it., 5 July 2017

9.         Hiroshi Esaki, WIDE project consortium, Japan, http://hiroshi1.hongo.wide.ad.jp/hiroshi/ 6 July, 2017

10.       Tim Polk, director for Internet Security Standards, National Institute of Standards and Technology (NIST), This email address is being protected from spambots. You need JavaScript enabled to view it., 301-906-2813, 6 July 2017

11.       Robert McDowell, former commissioner, FCC, partner in the Communications regulatory, Digital media & entertainment, Telecommunications, etc. practices at Cooley LLP, This email address is being protected from spambots. You need JavaScript enabled to view it., 201 251 4640, 7 July 2017

12.       Richard Beaird, former Senior Deputy United States Coordinator for International Communications and Information Policy, and Office Director for Multilateral Affairs within the U.S. Department of State’s Department of International Communications and Information Policy (CIP), associate at Wiley Rein, Washington, This email address is being protected from spambots. You need JavaScript enabled to view it., tel. 540 461 8639, 10 July, 2017

13.       Akinori Maemura, General Manager of Internet Development Department at Japan Network Information Center (JPNIC), This email address is being protected from spambots. You need JavaScript enabled to view it. 81 90 7014 5225, 11 July 2017

14.       Gordon Goldstein, Managing Director - Head of External Affairs at Silver Lake Investments, and Senior fellow, Council on Foreign Relations, This email address is being protected from spambots. You need JavaScript enabled to view it. 1 877 668 4493 x 807 935 015, 12 July 2017

15.       James Bladel, Vice President of Policy, Go Daddy Inc., This email address is being protected from spambots. You need JavaScript enabled to view it. , 21 July, 2017

16.       Kathy Brown, President, Internet Society, This email address is being protected from spambots. You need JavaScript enabled to view it., 703 447 4677, 9 August, 2017

17.       Richard Shockey, Chairman, SIP Forum, This email address is being protected from spambots. You need JavaScript enabled to view it., 703 593 2683, 25 August 2017

18.       Professor Fen Hampson, Centre for International Governance Innovation, co-director, Global Commission on Internet Governance, This email address is being protected from spambots. You need JavaScript enabled to view it., 613 769 1004, 29 August 2017

19.       Professor Milton Mueller. Georgia Institute of Technology, This email address is being protected from spambots. You need JavaScript enabled to view it.. 315 254 3242, 12 September 2017

20.       Allan MacGillivray, Director of Policy, Canadian Internet Registration Authority, 14 September, 2017, This email address is being protected from spambots. You need JavaScript enabled to view it., 613 760 3300


  1. Mueller, Milton Networks and States: The Global Politics of Internet Governance, The MIT Press, 2010, page 264
  2. Laura DeNardis, The Global War for Internet Governance, Yale University press, 2014, page 227


  1. Muller, ibidem, page 266
  2. See for instance the Tunis declaration of 2005, paras 73 and following at http://www.itu.int/net/wsis/docs2/tunis/off/6rev1.html
  3. Interview 10 July 2017
  4. Ambassador David Gross, 5 July, 2017, James Bladel, VP Go Daddy, 21 July 2017
  5. Richard Beaird, in discussion, 10 July 2017
  6. Interview July 7, 2017
You are here: Home Publications Reports Uncategorised The Multi Stakeholder Approach to Policy Development