Summary

  • AFRINIC’s Version 1 Terms made a genuine institutional advance: a standing five-seat Appeal Committee, separate from current PDWG chairs, with specialist eligibility pools, exclusions, randomised candidate ordering, recusal, publication duties and power to annul a chair decision when the PDP had not been followed.
  • The independence claim stopped short of the control surface. The Board appointed every ordinary member, could fill shortfalls from the AFRINIC community, could replace any or all members, selected a temporary replacement after recusal and received a committee request for extra time.
  • The sound response is neither to abolish expert review nor to elevate it into a private court. It is to place protected terms, reasoned removal, independent conflict substitution, autonomous support, fixed timing rules and a complete public record around a remedy limited to correcting the PDP.

L3 — Five seats, fourteen days and a standing appeal body

The starting point was a promise already present in AFRINIC’s Policy Development Process, not an appeal power invented from nothing by the Board in September 2017. A person dissatisfied with an action taken by the chairs of the Policy Development Working Group was first expected to discuss the disagreement with the chairs or the working group. If that effort did not resolve the matter, the person could appeal to a Board-appointed Appeal Committee. The filing needed support from three other participants in the relevant discussions and had to arrive within two weeks of the challenged decision. That sequence matters.

It placed informal correction first, but it also recognised that asking the original decision-makers to reconsider themselves could not exhaust the available remedy.

The promise left consequential design questions unanswered. It did not itself settle how large the committee should be, who should sit on it or whether reviewers should be assembled only after a dispute arose. Version 1 of the Terms of Reference, dated 18 July 2017, said that the PDP was silent on size and composition and treated those matters as choices for the Board. The selected Board record then fixes 13 September 2017 as the date on which the Board approved the committee’s terms and composition. These are distinct institutional moments. The first is the date carried by the charter text; the second is the recorded act of approval.

Neither date, on the evidence available, establishes the day on which a functioning committee first became able to receive and decide a case.

That last distinction prevents a neat chronology from becoming a false one. AFRINIC’s official archive dates Version 1 to 18 July 2017, dates Version 2 to 26 November 2017 and lists the earliest previous-member terms as running from December 2017 to June 2018. The archive therefore supports a sequence of instrument, approval record and later listed terms. It does not prove that the body was operational on 18 July or 13 September. Nor does it establish that the people whose terms are listed from December were necessarily the precise composition approved in September.

The surviving record is strong enough to analyse the design, but not to manufacture a commencement date or a membership history.

Within that design, the choice of a standing body was sensible. A fourteen-day filing period is too short to combine comfortably with a fresh search for reviewers. An ad hoc panel created after each complaint would force selection to take place in the shadow of the very dispute it was meant to review. Potential candidates, appointing officials and participants would all know the identity of the appellant, the policy dispute and the interests affected. Delay would collide with the filing clock; tactical argument over the panel could overwhelm the procedural question. Version 1 instead reasoned that the committee had to exist in advance.

Standing capacity was not bureaucratic ornament. It was the practical answer to a compressed remedy.

The five seats were designed around experience of policy process rather than generic prestige. Two were for past AFRINIC PDWG chairs or co-chairs. Two were for current or past chairs, co-chairs or equivalent figures from the policy groups of other Regional Internet Registries, with a rule that both could not be associated with the same RIR. The fifth was for a current or past chair or vice-chair of the NRO Number Council, which also serves as the Address Supporting Organization Address Council, and that person had to come from outside the AFRINIC service region.

The architecture sought people capable of reading a record of mailing-list exchanges, meetings, objections and chair reasoning without needing a tutorial in the mechanics of rough-consensus assessment.

That expertise had real value. Rough consensus is neither a referendum nor a simple majority. Chairs must identify substantive objections, decide whether objections have been answered, distinguish repetition from unresolved technical concern and determine whether the group can move forward despite a lack of unanimity. A reviewer unfamiliar with those tasks could mistake volume for weight or numerical support for procedural validity. Former chairs and peer-process leaders were more likely to understand what an adequate record should contain, which procedural steps mattered and how a chair’s reasons related to the PDP.

The design was therefore capable of producing informed scrutiny rather than a ceremonial rehearing.

Version 1 also imposed significant eligibility exclusions. AFRINIC employees and directors could not serve as ordinary committee members. Current AFRINIC PDWG chairs or co-chairs were excluded, as were authors or co-authors of the proposal under appeal. Each exclusion addressed an obvious route by which the original decision chain or a direct substantive interest might enter the reviewing body.

Removing current chairs was especially important: an appeal against chair action cannot be a meaningful second look if current occupants of the appealed office sit in judgment over their own institutional peers or over a decision in which they participated.

Candidate ordering added another modest safeguard. Within the eligible pools, Version 1 used random permutation to establish the order in which candidates would be approached. If candidates declined, the search widened. Random ordering did not make the appointments independent of the Board, but it did reduce the opportunity to choose a favoured person first from a pool of otherwise eligible experts. That is worth crediting precisely. Institutional criticism becomes less useful when it treats every internal safeguard as a sham.

A randomised approach sequence can restrain hand-picking even when it does not resolve who defines the pool, makes the appointment or controls the resulting term.

The committee’s process contained other visible checks. A member who had a conflict was required to recuse. After an appeal arrived, the committee had to publish a timeline for its work. It was required to report to the PDWG through the Resource Policy Discussion mailing list, and appeals and findings had to be published. The filing itself had to identify the challenged decision, name the supporters, describe the unsuccessful attempt to resolve the dispute, state detailed grounds and include supporting material. It went to the committee with a separate copy to the discussion list.

These requirements forced an appellant to make a procedural case on a record, while publication made it harder for the dispute simply to disappear into private correspondence.

The remedy, too, was more than advice. If the committee found that the PDP had not been followed, it could annul the chair decision. Version 1 described that ruling as final and binding. Within AFRINIC’s private policy arrangements, that gave the review practical consequence: the committee was not confined to expressing disappointment while the original act remained untouched. Annulment could restore the process to the point at which proper procedure had to be followed.

A participant therefore had a route that was separate from the current chairs, informed by people familiar with policy development and capable of reversing a procedurally defective chair action.

But the nature of that remedy must be kept exact. The committee reviewed whether the PDP had been followed. It was not granted a general licence to substitute its preferred policy outcome for the chair’s rough-consensus judgment, still less to legislate for a continent. “Final and binding” described the endpoint of an internal appeal within a private, membership-based technical coordination system. It did not create a sovereign judgment, displace public courts or confer regulatory, policing, prosecutorial, punitive or confiscatory power. AFRINIC keeps useful records and supports coordination over Internet number resources.

Neither a charter nor the solemnity of an appeal vocabulary transforms that role into public authority.

This boundary strengthens rather than weakens the case for good internal procedure. Private decisions can still matter greatly. A policy outcome may affect how operators plan for address resources, transfers, compliance work and capital deployment. An uncertain process can consume volunteer time, prolong mailing-list conflict and lead parties to seek more costly forums. The fact that a committee is not a court does not make its fairness unimportant.

It means that its legitimacy must come from bounded authorisation, clear rules, visible independence and a remedy fitted to the private process it administers—not from borrowing the language or imagery of a judiciary.

The best reading of the 2017 act therefore begins with genuine progress. The PDP’s earlier promise of conflict resolution acquired a specified body. The body stood apart from the current chairs. Its seats drew on relevant experience, its ordinary membership excluded obvious institutional and proposal conflicts, its candidate order incorporated chance, its conflicted members had to step aside, and its proceedings culminated in a published, consequential remedy. None of that proves independence in operation, because the record does not show an appeal being heard in 2017 or how any member behaved.

It does show that the Board addressed a real procedural gap with more care than a purely ad hoc or chair-controlled review would have offered.

The question left open was not whether the committee existed on paper as a different body. It was whether that difference survived the practical levers by which a live reviewer could be constituted, changed, supported and timed. Structural separation from PDWG chairs answered only the first half of the control problem. Version 1’s safeguards reduced direct conflicts at the reviewer level, yet the charter located appointment and turnover at the Board. To judge the design fairly, one must follow those powers through the life of a case rather than stop at the committee’s name or the qualifications of its members.