Summary

  • On 2 March 2022, AFRINIC’s Board adopted an amended election process through Resolution 202202.678 by four votes in favour, none against and three abstentions, with publication ordered for the following day.
  • The process contained serious operational safeguards, but ballot controls could not establish which seats were lawfully open, the proper term for each winner or whether the resulting succession pattern would preserve quorum.
  • The Board had just acknowledged a conflict involving casual vacancies and the sequence in Articles 13.6 and 13.7, while its minutes recorded a warning that too many simultaneous elections could leave the next Board without quorum.
  • The strongest repair would have been a signed seat-sequence certificate joining the operative bylaws, every seat’s history, the open-seat decision, winner terms and a forward quorum forecast before voting began.
  • This was private corporate governance for a technical bookkeeper and coordinator. It mattered for continuity and accountability, but it created no sovereign, regulatory, punitive or public-law authority over operators, resource holders or Africa.

The vote and the unresolved input

On the 2 March 2022 continuation of a Board meeting that had opened on 26 January and continued on 21 February, AFRINIC’s directors adopted a new Election Process and Guidelines as amended. The official act was Resolution 202202.678. Four directors voted in favour, none voted against and three abstained. The roll call names the affirmative votes as BE, AO, SM and HY and the abstentions as OO, EK and AA. The minutes identify the seconder only as “A”; the public record does not justify expanding that initial.

The Board also directed the chief executive to publish the process by Thursday, 3 March, and to launch a two-week expression-of-interest call for the 2022 Nominations Committee.

Those details make the decision look brisk and administrative: review a rulebook, amend it, vote, publish, begin staffing the election. Yet the sequence of the agenda makes the constitutional problem impossible to treat as a remote afterthought. Immediately before adopting the process, the Board had dealt with the extension of the incumbent in Seat 6. One director objected that members lacked recourse against the extension, while another warned that too many seats elected at the same time could leave the next Board facing a quorum problem.

The next resolution, Resolution 202202.677, recorded that two casually filled seats becoming vacant at the 2022 annual meeting for three-year terms would breach the election-sequence principle in Articles 13.6 and 13.7. The Board asked its Governance Committee to develop workable solutions by the end of April.

The election rulebook thus arrived at a moment when the Board knew that the input to the election—the set of seats and the succession pattern produced by their terms—was unsettled. Abstention from Resolution 202202.678 does not by itself reveal why any director withheld support, and the record should not be made to say more than it does. But the institutional juxtaposition is clear. A detailed process was adopted in the same sitting in which a sequence conflict was acknowledged and a future-quorum warning was heard. The central issue was not whether election officials could count ballots.

It was whether the office structure on the far side of the count had been independently reconciled with the bylaws.

That difference matters because an election transforms votes into offices. A ballot can be authentic, a member can be eligible, a proxy can be verified and a tally can be accurate, yet the institutional result can remain unsafe if the contest opened a seat at the wrong time or attached the wrong term to its winner. Procedure protects the journey from voter to tally. It does not necessarily protect the separate journey from seat history to lawful vacancy, or from declared winner to a Board capable of acting through the next cycle.

The strongest case for the rulebook

The benign case should be stated without qualification: AFRINIC needed an election process that people could use. Its annual meeting might be physical or virtual. Members needed to know who could nominate, who could stand, when comments could be made, how eligibility would be checked, how proxies would operate, when electronic voting would open, who would serve as trustees, how votes would be tallied and when results would be announced. Without those details, every contested step could become an occasion for improvisation.

The published process addressed those needs in substance. It cited the Board’s authority under Article 13.2 to establish election arrangements and said it applied to Board elections, as well as to other elections following the process with necessary adaptations. It identified Seats 1, 2 and 7 as the open seats for 2022. It defined member good standing, distinguished the act of nomination from the status of a candidate and assigned the Nominations Committee responsibility for eligibility assessment and the final slate. The timetable included Board approval and later provided a point at which Board input was non-binding advice.

That distinction matters because the same institution whose seats are being filled should not silently blur advice, approval and final candidate selection.

The process also recognized that participation had more than one channel. It dealt with face-to-face and virtual meetings, proxy voting and electronic voting. For Board and Governance Committee elections, the named trustees were the chief executive, the Nominations Committee chair and AFRINIC legal counsel. Electronic voting was to open at least fourteen days before the annual meeting and end on election day. Responsibilities, dates and custody points were placed where members could inspect them. Those measures could reduce ambiguity, deter casual manipulation and create records against which complaints could be tested.

Limiting the published contest to Seats 1, 2 and 7 was also a meaningful restraint. The Governance Committee’s later communiqué recommended a maximum of three seats at an annual meeting and listed the same three for 2022. That later recommendation was not a court ruling, and it should not be treated as one. Still, the correspondence shows that the published open set was not recklessly broad. It reduced the immediate danger that six elections would be synchronized in one meeting. Operational prudence was present.

It would therefore be wrong to dismiss the 2022 process as empty form. The process was a real governance control for a private member corporation. It made election work legible and repeatable. The right criticism is narrower and stronger: the document could govern only the values supplied to it. It could say how to elect Seats 1, 2 and 7, but it could not prove, merely by repeating that list, why each seat was lawfully open, what term its winner should receive or how the resulting expiries interacted with every other seat.

Two kinds of integrity

Operational integrity asks whether the announced rules were followed. Were nominators and candidates eligible? Was good standing measured at an identified point? Were comments handled within the announced period? Were proxy instructions authenticated? Did electronic voting open and close as stated? Were trustees properly assigned? Does the tally reconcile with the votes received? Can the result be reconstructed from retained records? These are the familiar questions of election administration.

Constitutional-input integrity starts earlier. Which bylaw version governed on the date of the decision? What did it require for the normal term and staggered sequence of every elected seat? Was an incumbent elected by members, appointed to a casual vacancy or continuing under a separate Board act? When did that person’s legal tenure begin and end? Did filling a casual vacancy alter the seat’s cycle, or only provide an occupant until the next specified point? If a seat returned to election, was the next term a full term or a remainder? What would the combined expiry pattern do to quorum one, two and three years later?

Articles 13.5, 13.6, 13.7 and 13.14 were relevant to terms, sequence and casual-vacancy treatment. The Board’s neighbouring decision demonstrates that at least one interface among those provisions was contested or considered defective in its effect. The later Governance Committee paper identified four risks: misreading the casual-vacancy provision, inability to raise quorum when many seats were open, elected terms shorter than three years and a broken election sequence. Its recommendation was an institutional response, not an external adjudication. It nevertheless clarifies why the inputs deserved a control of their own.

Imagine a perfectly administered vote over an incorrectly composed open-seat list. No ballot need be false. No proxy need be defective. The trustees may perform faultlessly. Yet if one winner is placed into a term that should not have begun, or another receives a shortened term inconsistent with the governing arrangement, private corporate power has been redistributed on an unstable basis. Conversely, extending an incumbent to protect sequence changes who holds office before a member ballot. Either direction requires a reasoned link from bylaw text to seat history to term result. Election mechanics cannot supply that link after the fact.

The public material does not establish that Resolution 202202.678 was unlawful, that the 2022 ballot was invalid or that the act caused AFRINIC’s later loss of quorum. It does not contain the amended draft as reviewed before the vote, a version hash, tracked changes, the complete legal opinions or a distinct seat-sequence certificate. Those items may have existed privately; absence from the public packet is not proof that they did not.

The defensible conclusion is more exact: the public act demonstrates a known input risk and a detailed voting process, but it does not show the independent reconciliation that would allow members to audit how the first was kept from contaminating the second.

Why private governance still matters

AFRINIC’s role must be kept within its proper boundary. It is a private technical bookkeeper and coordinator. Its defensible work concerns accurate registry records, uniqueness, bounded coordination, continuity, auditability and correction. An internal election selects office-holders for a member corporation. It does not create a legislature for Africa, confer authority over network operators or turn directors into public officials. AFRINIC has no sovereign, regulatory, police, prosecutorial, punitive, confiscatory or public-law adjudicative power.

That limit does not make Board succession trivial. A technical coordinator can still depend on a Board to supervise staff, approve budgets and authorize the organisation’s limited registry service. If the line of corporate authority becomes uncertain, members and counterparties must spend time establishing who may decide, sign or oversee. Operators may have to conduct additional diligence around continuity even when no network interruption has occurred. The private character of the institution is the reason for precise governance, not an excuse to imitate public power or to disregard corporate form.

NRS’s account of AFRINIC’s private character helps place member scrutiny on the right plane: the issue is accountability inside a corporation, not obedience to a territorial sovereign. Heng Lu’s thin-bookkeeper analysis adds the restraint that uniqueness coordination cannot become punishment or political rule. LARUS identifies the practical transmission path by which opaque registry governance can impose continuity questions and diligence costs on operators, without proving that this resolution caused any incident.

BTW’s treatment of the later quorum crisis supplies a boundary for this inquiry: the later breakdown has its own history and should not be reverse-engineered into proof against the 2022 rulebook.

The event owned here is smaller but consequential. The Board adopted an operationally mature election process while constitutional succession inputs were under active repair. The useful question is not whether better ballot technology could have solved the issue. It could not. The missing control belonged between the bylaws and the ballot: a visible, signed account of why these seats, for these terms, in this sequence, would produce a lawful and quorate Board.