Summary

  • AFRINIC’s five-seat Policy Development Appeal Committee was left with Luc Missidimbazi, Adam Nelson and Paulos Nyirenda after Wafa Dahmani and Jean-Robert Hountomey resigned. Two transfer-policy appeals were already outstanding, and a third appeal arrived on 21 April 2021.
  • On 14 April, the Board unanimously adopted Resolution 202104.606. It exceptionally treated the three remaining members as a sufficient quorum pending review of the Terms of Reference and the filling of the two vacancies. This was a temporary dispensation, not a permanent amendment of Version 2.
  • The apparent repair did not produce continuation. On 6 May, the three remaining members said the earlier suspension and the Board’s language of permission had created uncertainty. They declined to proceed until the committee was fully constituted with five members.
  • The durable lesson is not that continuity should have been abandoned. It is that a private appeal body needs a pre-authorised bridge specifying emergency quorum, docket preservation, permitted acts, reasons, expiry, review and handback—without inflating an internal correction mechanism into sovereign or judicial power.

L3 — Three survivors and three live appeals

The arithmetic could hardly have been clearer. The Policy Development Appeal Committee had five seats. Two occupants had resigned. Three members remained. And, by the time the remaining committee explained its position on 6 May 2021, three appeals were waiting: two outstanding matters concerning transfer policy and a further appeal filed on 21 April about confirmation of consensus in the selection of Policy Development Working Group co-chairs. An internal route intended to examine disagreements about the actions of PDWG chairs had business before it, but its ordinary composition had fractured.

The institutional answer looked, at first glance, as simple as the arithmetic. On 14 April, AFRINIC’s Board decided that the surviving three could proceed. Resolution 202104.606 exceptionally allowed the committee to operate with three members pending a review of its Terms of Reference, and deemed those three a sufficient quorum until the vacancies were filled. Yet permission and performance diverged. The Board authorised continuation with three; the remaining committee later refused to continue until it had five. The bridge existed in the Board’s resolution but could not carry the body it was meant to support.

That reversal matters because an appeal mechanism is tested not by the elegance of its description in settled times but by whether it remains credible when the organisation around it is under strain. A correction route must be available precisely when a disputed decision has already produced disagreement. If a resignation can remove that route, the people seeking review inherit the institutional failure even though they neither caused the vacancy nor designed the rules. If an improvised fix can itself become the subject of a legitimacy dispute, restoring a number on paper does not restore the function.

Version 2 of the committee’s Terms of Reference explains both the strength and the weakness of the arrangement. It established a Board-appointed committee of five. Two places were assigned to people who had previously served as PDWG chairs or co-chairs, while three were for people selected from the AFRINIC community. Employees, directors and current PDWG chairs were excluded. Those choices did more than set a headcount: they attempted to preserve experience, distance from current decision-makers and a mixture of relevant perspectives.

The Terms also contemplated turnover. After a resignation, a replacement was to be found through the ordinary selection process. During a live appeal, however, ordinary replacement was to be delayed unless appeals had remained in progress for six months. That safeguard is intelligible. Changing a reviewer during a contested matter can unsettle the forum, require repetition and invite suspicion that the composition is being altered around a result. But the same protection can become a trap when resignations remove enough members to cast doubt on whether the body can act at all.

Version 2 stated the committee’s five-seat composition and provided a path to replacement, yet it did not state an express numerical quorum.

Silence is not self-executing. It does not establish that three was necessarily sufficient, nor that three was necessarily insufficient, under every relevant rule. It tells us something narrower and more useful: the published Terms did not supply the numerical answer that the institution urgently needed. Once the vacancies existed, any answer would have to be derived from some other reading of authority or added after the event. That is the opening through which a practical continuity measure became a contest about who could define continuity.

The surviving names are part of the record. Wafa Dahmani and Jean-Robert Hountomey were recorded as having resigned. Luc Missidimbazi, Adam Nelson and Paulos Nyirenda remained. The reasons for the resignations, the dates of the resignation letters and the full contents of any surrounding correspondence are not established by the available record. Resignation should therefore be treated here as the trigger, not converted into a story about personal motive. The institutional consequence is known even where the private explanation is not: a five-person body had become a three-person body.

The next move intensified the difficulty. According to the committee’s later report, the Board chair wrote to the remaining members asking that committee activity be suspended. The full letter is not available in this record, so its complete reasoning and wording should not be reconstructed. What can be said is that suspension placed the forum’s ability to work behind an instruction from the same corporate structure that appointed it. Even if motivated by caution, that sequence made the eventual repair harder. The issue was no longer only how many members were enough.

It was also whether the committee, once properly constituted, depended on a fresh grant of operational permission from the Board.

The Board did not reach for the least subtle option. Its 14 April minutes record that legal-team recommendations included suspending Version 2, dissolving the committee and creating an ad hoc committee to deal with the outstanding appeals. Those steps would have replaced the existing forum more completely. The Board instead chose a narrower-looking course: preserve the same three people, treat them as a temporary quorum, keep the filling of the vacancies in view and link the dispensation to a review of the Terms.

Seun Ojedeji’s recorded intervention identifies the practical concern. He asked that his comments be recorded, said quorum was the issue, proposed stating it as three and warned against losing the trust of the remaining members. Vika Mpisane proposed the resolution, Ojedeji seconded it, and the minutes record unanimous approval. The result was a specific corporate act directed at a specific committee in a specific period of reduced membership. It was not a general alteration of what Version 2 would permanently require.

Nor did it decide any of the appeals, change their merits or establish that the surviving members had in fact continued working.

The committee’s response on 6 May made the failure of the repair unambiguous. The three did not treat Resolution 202104.606 as a sufficient basis on which to resume the appeals. They acknowledged that the Board had a role in constituting the committee, but disputed the proposition that, once properly constituted, the Board could grant or deny permission for it to perform its stated activities. They said the suspension and subsequent resolution had imposed undue uncertainty and burden. Their conclusion was operational, not merely rhetorical: they would not proceed until the body had again been fully constituted with five members.

That position must not be exaggerated into a judgment that the Board acted unlawfully. The committee’s report states an institutional interpretation; it is not a court ruling. Equally, unanimity in the Boardroom does not prove that the Board’s interpretation was accepted by the body expected to use it. The episode contains two competing ideas of legitimate continuity. One begins with corporate sponsorship: the Board appoints and organises the committee, so it can keep that machinery available during an unforeseen vacancy.

The other begins with decisional autonomy: after constitution, an appeal body cannot appear to operate only when its corporate sponsor switches it on. Both respond to a real risk. Neither is sealed here as the final legal answer.

The three dockets sharpened the stakes without resolving that dispute. Two were described as outstanding transfer-policy appeals. The third, submitted on 21 April, concerned confirmation of consensus in selecting PDWG co-chairs. Their merits, the parties’ arguments and their eventual outcomes lie outside this account. No quantified delay, financial loss, resource impact or outage can be attributed to the episode from the available evidence. What the dockets demonstrate is pressure on the mechanism: this was not an abstract seminar about committee design.

Requests for internal review existed while the forum’s capacity to review them was contested.

An appeal route of this kind has a modest but important purpose. It can test whether a private policy process followed its own rules and whether a chair’s judgment of consensus should stand within that process. It can preserve reasons, allow an objection to be heard and create a bounded correction without forcing every procedural disagreement into a wider institutional confrontation. Version 2 used terms such as “adjudicate” and described outcomes as final and binding, but those words belong inside AFRINIC’s private PDP.

They do not turn the committee into a public court, any more than a unanimous Board resolution turns the Board into a legislature.

This boundary is fundamental. AFRINIC is a private technical bookkeeper and coordinator. Accurate records, interoperable administration and the reduction of avoidable operational conflict are worthwhile functions. The Board may organise internal machinery within the company’s governing arrangements, and the Appeal Committee may perform the internal review assigned to it. Neither body thereby acquires sovereign, legislative, regulatory, police, prosecutorial, punitive, confiscatory or public-law adjudicative authority. The importance of a correction path comes from disciplined administration, not from any claim to rule a region.

The April–May sequence can therefore be stated without melodrama. A rulebook created a five-seat internal review body and a vacancy mechanism, but no express numerical quorum. Two members resigned. The chair instructed the survivors to suspend activity. The Board considered dissolving the body and appointing an ad hoc alternative, then unanimously chose a temporary three-member dispensation instead. The surviving three rejected the permission framing and left three live appeals for a restored five-member committee. The attempt at continuity was real, narrow-looking and unsuccessful in use.

This is why the event deserves attention on its own terms. The interesting question is not who won a constitutional contest in miniature. It is why an arrangement designed to correct disputed procedure had no agreed way to correct its own loss of capacity. The body could describe who belonged, who could not belong and how a vacancy would eventually be filled. It could not point to a previously accepted bridge from five to three while live appeals remained protected. At the moment it needed resilience, its governing text offered principles that pulled in different directions but no common operational answer.