1. Bottom-up oversight of multistakeholder processes.

    (This is a version of a NetMundial 2014 contribution

    One of the barriers to accepting free standing multistakeholder governance organizations is oversight.  Accountability requires that there be oversight.  But how can oversight be done in a bottom-up multistakeholder organization without inflicting external control, for example the so-called new space, upon the bottom up multistakeholder organizations.

    Often, as soon as the word ‘oversight’ is mentioned governments jump in and assert themselves as the ultimate overseers of all things related to Internet public policy.  From the point of view of the multistakeholder models of international  participatory democracy, it is impossible to accept government oversight of multistakeholder organizations.  If one stakeholder group is solely responsible for oversight, the group becomes a unistakeholder entity, no matter what appearance those running around at the bottom of the organization might give.  Government oversight of multistakeholder organizations would destroy their multistakeholder character.

    There are, however, methods by which organizations do create bottom-up oversight. Some like the IETF have multiple levels of appeal on every decision; where the decisions of working groups are can be appealed to the Internet Engineering Steering Group (IESG), a group chosen by a multistakeholder Nominating Committee (Nomcom) and approved by the Internet Architecture Board (IAB), itself having been chosen by the multistakeholder Nomcom and approved by the Internet Society (ISOC) Board of Trustees.  The dissatisfied appellant can further appeal a decision by the IESG to the IAB.  Since the IESG and the IAB have very different mandates from the community, it would be a mistake to think that the IAB rubber stamps the decisions of the IESG; anyone who has observed the relations between these groups only has to recall IETF history to know that relations have not always been warm or even cordial.  

    In case the appeals process is somehow perverted, the IAB decision can be further appealed to the Board of Trustees of ISOC on process issues, a group that is chosen by a combination of nominating committees and votes in the various stakeholder groups that make up the Internet Society.  Each of these appeals and results is published transparently for the entire Internet community to read, in a manner that allows comment on the IETF list.  Each of these appeals can be discussed in the open plenaries held three times a year, if so desired by the community.  Anyone who has ever followed the IETF list or an IETF plenary knows how vox populi can be quite persuasive in this environment. The history of the IETF shows that, on occasion, it has persuaded either the IESG or the IAB to reconsider its decisions.   It should be noted that there are also methods within the IETF for recall of the members of the IESG and the IAB. A chain of transparent appeals such as this, with a strong public voice is one form of bottom-oversight that has been shown to work.  A series of public and transparent appeals are, in fact, one of the tools that any organization can use to provide oversight.

    The Affirmation of Commitments between the US Department of Commerce and the Internet Corporation for Assigned Names and Numbers (AOC) establishes a new method for bottom-up oversight of a multistakeholder organization.  This process is entering its 4th year.  I had the privilege of being part of the second Accountability and Transparency Review Team (ATRT2) that is stipulated in the process and write the following from my personal experience as a member of that group.

    As outlined in the AOC, each of the Support Organizations and Advisory Committee nominates several members to the various review teams: transparency and accountability, the security and stability of the names and number systems, competition and consumer trust, and registrant directory operations compliance with international laws and norms (aka WHOIS).  From the candidates presented by the various stakeholder groups, the Chair of the Board or the CEO, depending on the team, and the Chair of the Government Advisory Committee (GAC) pick a review team.  To this group they add a few experts to balance out the mix of skills.  While not everyone is happy about the selection process and it does seem to favor governmental actors, at last in terms of the ATRT, it is by and large a process accepted by the community.  The groups work together for approximately a year to produce their reports.  They are served by a group of professional ICANN staff and are constrained by an ICANN Board determined budget.  The ATRT2 report discusses how this is occasionally an impediment to the group being able to do as complete a job as might be hoped.

    These reviews run in a 3 year cyclical manner with the ATRT running every third year, for a year. Currently ICANN is in the second cycle of AOC reviews.  Each of these reviews provides the opportunity for multiple public consultations and several public reviews of the work in progress.  The final reports are published in multiple languages and opened for a final public review before the ICANN Board receives them.  The Board then reviews the reports, and the comments and, with the advice of Staff, decides how it will respond to the reports.  By the time the Board acts, the review teams have been disbanded. The results of Board and Staff actions are reviewed by the next review teams.

    In addition to reviewing the state of transparency and accountability in ICANN, ATRT is responsible for reviewing the way in which ICANN had responded to the previous AOC reviews and for reviewing any changes that had been made in response to the previous ATRT review. It should be noted that review team reports only recommend actions to the Board, which the Board is able to accept or not to accept.  Any refusal by the Board or ICANN staff to implement the changes requested is subject to review by the next review team as the cycle repeats.  That is, ATRT3 will review the changes made since ATRT2, just as ATRT2 reviewed the changes since ATRT1.

    The ATRT2 report gives a very detailed description in its review of both the  state of the organization and the degree to which ICANN complied with the first cycle’s reviews.  The determination was mixed, with many changes only being made as the ATRT2 review was ongoing.  Some progress had been made, but not as much as the first ATRT review had recommended.  As ATRT2 second report is still in review until 15 March 2014, the Board has not had much opportunity to act on it yet, though to its credit it has started working on some of the items in anticipation.

    For a bottom-up review mechanism like the AOC to work, those responsible for the organization must take the report and the obligation to accept and implement the recommendations seriously.  Whether this is the case for ICANN, remains to be seen, and will be part of the proof of whether this is a workable method.  According to a blog piece by the President of ICANN, Staff has every intention of completing the necessary changes.

    There are other requirements for such a bottom-up review process to work.  One of the handicaps ATRT2 experienced was the need to start its review process from a blank sheet.  We had to request reports from ICANN on every issue, including those that had been documented in the first ATRT report.  We were also unable to build upon any of the research and data collection that had been done by the first review and in the intervening reviews as that data had not been retained and processes had not been continued. During the ATRT2 review, several research efforts, for example, research on the policy development process and on measurement practices were undertaken.  It is to be hoped that that the data from these will be available to ATRT3 so that the team will not need to again start from a blank page and will be able to look at longitudinal trends in ICANN’s transparency and accountability.  Continuity of bottom-up oversight enables a better view of an organization and thus better accountability.


    To do a complete and thorough review, a review team needs full access to organization data.  This is sometimes difficult as the team needs to know what data is available in order to ask for it.  In order to do an adequate review it is better to have an organization anticipate and offer the necessary information as opposed to having a gate in the process that only gives the team information when it is correctly requested and the request has been approved up the corporate and legal chain of command.  ICANN has tended toward the gating methodology as opposed to an anticipatory and open method of offering data, though for the most part after a bit of pushing and pulling, the necessary information was obtained.

    ICANN is settling into the process of ongoing AOC reviews. The process needs improvement, and the ATRT2 report contains recommendations for how this can be done.  For a more in depth review, readers can consult the ATRT2 Transparency and Accountability review.  In my view the method shows promise yet to be fully delivered on.

    The strongest bottom-up oversight would be achieved by combining the methods of the IETF and ICANN.  IETF offers a strong appeals mechanism and the ability to remove officers from their roles if they do not fulfill the responsibilities.  ICANN has a very weak appeals mechanism, in fact an review and repair of that mechanism was recommended by both ATRT1 and ATRT2.  ICANN processes allow for review of Board decisions, but only for process infractions, there is no means of request a review within ICANN of a decision on a substantive basis as can be done in the IETF.  And while there is an Independent Review Process for Board decisions and Staff actions, it is reputed to require a deposit in the order of a million dollars and can only deliver a decision in favor of the appellant if malfeasance of some other bad faith effort can be proven.  ICANN has no mechanism for removing Board members for the manner in which they fulfill their responsibilities other than not to renew their 3 year terms for a second or third time (ICANN Board members can serve a maximum of 9 consecutive years).  

    ICANN on the other hand offers an innovative set of bottom up community reviews, whereas the IETF only initiates reviews of its processes when there is an indication that the process is somehow broken using the ‘if it isn’t broken don’t fix it’ philosophy.  It has, over the years initiated several working groups to review its processes when the community felt there was a problem. I had the opportunity of chairing one of its Nominating committee review working groups.  The recommendations of this WG were accepted and implemented.

    Bottom-up oversight of multistakeholder organizations is possible.  No method is perfect and use of several methods, the so-called belt and suspenders method, gives an organization its best chance of being accountable to its community without needing to resort to a governmental and hierarchical form of oversight.  An additonal element that might be considered as part of any bottom-up oversight mechanism is affordable access to an independent mediation board for issues that can’t be resolved in any other manner.

    Moving forward, no new spaces are required to oversee organizations such as IETF and ICANN, though both can improve their bottom-up methodologies by borrowing elements from the other.  An activity that would be helpful would be for workshops to be held within the IGF context on the various methodologies of bottom-up multistakeholder oversight and review so that organizations would be able to learn from each others experiences. 

    The practice of bottom-up multistakeholder review is young, but has a promising future.