Summary

  • On 8 July 2016, AFRINIC’s Board unanimously adopted Resolution 201607.290, reconstituting its Finance, Remuneration and Audit standing committees after Sunday Folayan presented a membership table that had been discussed informally during a Board retreat.
  • The roster placed chief executive Alan Barrett on Finance and Remuneration but not Audit. It also connected Finance and Audit asymmetrically: the Audit chair was to observe Finance, while the Finance chair became a full Audit member.
  • That structure could reduce information silos, preserve continuity and help the Board connect financial planning with assurance. It could also blur the line between preparing financial choices and independently challenging them if chairs, conflicts, recusals, delegated limits and reporting duties were not explicit.
  • The public record proves the names, committee assignments, proposer, seconder and unanimous vote. It does not establish the selection rationale, identify the chairs at the vote, reproduce the retreat discussion or exact committee charters, or show how well the committees later performed.
  • AFRINIC remained a Mauritius private company carrying out a useful but narrow bookkeeping and coordination function for Internet number records. Its committee structure mattered to members and operators, but neither that practical importance nor a Board vote created sovereign, regulatory, police, punitive, confiscatory or adjudicative power.

The oddity at the centre of the table

Imagine the Board’s work spread across three smaller rooms. In one, directors and the chief executive review financial planning, management and reporting. In another, directors and the chief executive consider executive pay and performance. In the third, directors examine financial-reporting integrity, internal controls, risk, internal audit and information systems. The rooms are not sealed. One chair may listen in on Finance without becoming a member there; the other chair enters Audit as a full member. Information can cross the corridor in both directions, but formal standing does not cross symmetrically.

That is the concrete institutional design created by Resolution 201607.290. At its meeting on Friday 8 July 2016, AFRINIC’s Board reconstituted the Finance, Remuneration and Audit standing committees. The decision mattered because these committees divided access to some of a company’s most consequential information: money, controls, risk, executive performance and pay. It also mattered because their recommendations could shape what the full Board saw before it acted. Yet the proper scale of the event must be kept in view.

AFRINIC is a Mauritius private company performing a useful, narrow bookkeeping and coordination function for Internet number records. The resolution arranged its internal oversight. It did not legislate for Africa or confer power to regulate, police, punish, confiscate or adjudicate.

The unusual cross-link between Finance and Audit is the best way into the resolution because it turns a list of initials into a set of control questions. Was the observer role designed to give Audit sight of financial planning without drawing it into Finance’s decisions? Was full membership for the Finance chair on Audit designed to keep auditors informed about the assumptions behind budgets and reports? Did the difference reflect deliberate separation, convenience, individual expertise or something else? The published resolution does not answer. It establishes the topology, not its rationale.

That distinction is the foundation of serious governance analysis. A roster can show who was placed where and which formal connections were created. It cannot, by itself, show how members behaved, whether challenge was robust, whether conflicts were disclosed, whether anyone recused, whether recommendations were accepted, or whether the arrangement succeeded. Reading more into the table would turn structure into accusation. Reading less would miss the fact that structure determines who receives information and where dissent has a chance to form.

Nine people, ten places, three kinds of work

The resolution named four people to Finance: Krishna Seeburn, Haitham El Nakhal, Christian Bope and Alan Barrett. It added the chair of the Audit Committee as an observer, without naming that chair in the operative wording available in the public record. Remuneration consisted of Lucky Masilela, Sunday Folayan and Alan Barrett. Audit consisted of Andrew Alston, Seun Ojedeji and Abibu Ntahigiye, with the chair of the Finance Committee joining as a full member. Across the three rosters, nine named people occupied ten listed committee places because Barrett sat on two committees.

The official minutes do more than expand initials. They record that Sunday Folayan presented a table of committee members which had been discussed informally during the Board retreat. They also record the mechanics of adoption: Haitham El Nakhal proposed the resolution, Andrew Alston seconded it, and the Board passed it unanimously. Those facts give the decision a traceable corporate path. A proposal was presented, moved, seconded and agreed.

They do not disclose the substance of the retreat discussion. The table presented to the Board is not reproduced in the public material examined here. Nor does the published act provide a competency matrix, appointment criteria or individual reasons for allocating each person to a particular committee. “Discussed informally” should not be converted into “improperly decided” or “secretly arranged”. Informal preparatory discussion can be entirely ordinary. The precise point is narrower: a reader cannot reconstruct from the published material what alternatives were considered or why this configuration was preferred.

The committee names also matter because they describe distinct control functions rather than decorative titles. AFRINIC’s 2016 annual report said the Audit Committee assisted the Board with the integrity of financial reporting, internal controls, risk management, internal audit, information systems and information-technology governance. It described Finance as monitoring significant financial planning, management and reporting, then making recommendations to the Board. Remuneration was described as recommending policy on executive and senior-management pay, including performance-related elements.

Those mandates touch one another without being identical. Finance helps the Board decide how resources should be planned, managed and reported. Audit helps the Board test whether reporting and controls can be relied upon and whether risks are being identified. Remuneration links performance expectations to executive reward. In a well-designed company, none operates as a miniature sovereign. Each is an internal lens through which the Board receives information and advice while retaining ultimate corporate responsibility.

June shows that the committees were already working surfaces

The July resolution did not invent these subjects from nothing. Minutes from 6 June 2016 show predecessor committees handling live assignments before the reconstitution. Finance had been asked to complete work on the fee structure by the end of July and report to the Board. Remuneration was handling chief-executive evaluation and key performance indicators and had been asked to circulate a master document concerning chief-executive recruitment. Audit was expected to produce a report for the Board that would inform a report to the community.

These June assignments give the July roster practical weight. A Finance seat offered access to work on fees and financial planning. A Remuneration seat came with proximity to executive evaluation, performance measures and recruitment material. An Audit seat carried a role in preparing assurance information for the Board. They were not merely honorary panels waiting to be given a subject.

But chronology sets a firm evidentiary limit. The June minutes concern work before the July reconstitution. They do not prove that the July committees completed those tasks, adopted the same methods, reached any particular conclusion or carried every predecessor assignment forward unchanged. A committee name may remain constant while its people and working dynamics change. Reconstitution is precisely the moment when continuity must be demonstrated rather than assumed.

The annual report supplies the institutional reason for such moments. It says Board committees assist the Board under Article 15.3 of AFRINIC’s bylaws, are re-formed when a new Board is elected, have defined lifetimes and have terms of reference describing their role, membership and any delegated power or authority. Board turnover in 2016 therefore gives a straightforward explanation for the July action: the company refreshed committee membership after a change in the Board. Routine, however, is not the same as trivial.

A recurring control decision deserves a reproducible record because it repeatedly determines who will filter information for the Board.

There is another useful distinction here. The annual report’s statement that committees, as an institutional class, had defined lifetimes and terms of reference is evidence of a governance framework. It is not the same as publication of the exact charter, term, adoption date or delegated limits for each committee as reconstituted on 8 July. The report tells readers what the system was meant to contain. The resolution tells them who was assigned. The public chain between those two points remains incomplete.

Finance and Audit: information crossing in unequal forms

The most revealing feature of the roster is not ordinary dual membership. It is the explicit design of two chair-to-committee connections. The Audit chair was an observer of Finance. The Finance chair was a full member of Audit. These are different legal and practical positions inside a committee even when the individuals involved are equally conscientious.

An observer can receive papers, hear discussion and understand the assumptions behind financial proposals, depending on the applicable rules. Observation can give Audit early warning of issues that will later affect reporting or controls. It can also allow an audit perspective to enter Finance’s deliberations without giving the Audit chair a vote or making that chair responsible for Finance’s recommendation. In principle, this could support both information flow and a measure of distance.

A full member of Audit normally has a more participatory status than an observer: inclusion in the committee’s membership, and potentially in its quorum, deliberation and recommendations, subject to the charter. The published resolution makes the Finance chair a full Audit member but does not reproduce the rules that defined the consequences of that status. This link could give Audit valuable knowledge of budgeting choices, financial assumptions and reporting constraints. It could also make independent challenge harder if a person involved in shaping a financial position later took part in assessing controls or reporting connected to it.

The word “could” is essential. Cross-membership does not prove compromised independence. Many organisations deliberately overlap committees to prevent specialised groups becoming informational islands. In a small Board, the available pool of skills may be limited. A director with financial expertise may be the sensible person for both planning and assurance work. The question is not whether overlap is inherently improper. It is what safeguards turn overlap from a source of blurred accountability into a controlled bridge.

Those safeguards are concrete. Which matters could the observer attend? Did the observer receive every paper or only relevant ones? Could the observer speak but not vote? Was observer attendance counted for quorum? When Finance’s prior work was under review, could its chair participate in Audit’s deliberation, or was recusal required? Who chaired each body, for what term, and who substituted when a conflict arose? Which recommendations identified dissent? Did the full Board receive enough information to understand how cross-members had managed competing duties?

None of these questions accuses any named person. They are design tests prompted by the structure that the Board itself adopted. A governance system should not depend on outsiders trusting that sensible people will intuit the right boundary. It should set the boundary in advance, record its use and permit the Board and members to verify that it operated.

The chief executive’s two seats—and the absent third

Alan Barrett, AFRINIC’s chief executive, was a member of Finance and Remuneration and was not named as a member of Audit. That placement forms a second control line through the table. Management participation can be necessary because directors need accurate, timely information from the people running the company. Finance without the chief executive could lack operating context. Remuneration may need information about organisational performance, senior-management responsibilities and implementation constraints.

Audit, meanwhile, is often expected to preserve a clearer line of independent challenge over the records and controls produced by management.

The roster is consistent with such a logic, but it does not prove that this was the Board’s rationale. Nor does exclusion from Audit establish that all independence questions were solved. Committee effectiveness depends on more than whether the chief executive’s name appears on the membership list. It depends on who set agendas, which sessions management attended, whether non-management members met separately, how papers were prepared, who controlled access to underlying records and how recommendations reached the Board.

Remuneration creates a particularly obvious need for procedural clarity. A chief executive can provide indispensable operational information while also being the subject of evaluation, performance measures or pay policy. The June minutes show that the Remuneration Committee was handling chief-executive evaluation and key performance indicators, as well as a recruitment master document. Barrett’s July membership should therefore direct attention to recusals, agenda segmentation and the difference between supplying information and participating in a recommendation concerning oneself.

It would be wrong to infer from the roster that Barrett took part in any decision from which he should have been excluded. No such conduct is established. The accountable question is whether the rules stated when he could contribute, when he should leave, how a recusal would be recorded and how the remaining members would form a recommendation. A public summary could answer those questions without exposing personal compensation figures or confidential employment discussion.

Finance presents a different but related balance. The chief executive is normally central to financial planning because management prepares budgets, implements spending choices and understands operating needs. Directors nevertheless require the ability to challenge assumptions and ask for alternative scenarios. A Finance Committee with management at the table can be better informed; it can also become too dependent on management’s framing if members lack independent papers, sufficient time or direct access to financial expertise. Again, the membership list identifies the pressure point. It does not tell us how the pressure was managed.

What the resolution distributed

Read at the right scale, Resolution 201607.290 distributed four things: attention, information, challenge and recommendation. It gave smaller groups responsibility for looking closely at subjects that a full Board could not examine with the same frequency or depth. It positioned individuals to receive documents and hear explanations. It created places in which assumptions might be questioned before a matter reached the Board. And it shaped who would frame a recommendation for collective decision.

Attention is scarce in any Board. Committee formation is a claim about which subjects need concentrated time and which people should supply it. Finance, Audit and Remuneration were sensible categories because a registry company depends on financial continuity, reliable internal controls and capable executive management. The annual report’s descriptions show an intended division of labour: Finance looked forward and across planning and management; Audit examined reporting integrity and control; Remuneration focused on pay and performance policy.

Information follows membership. Papers, briefings and discussion make committees informational junctions. The Finance–Audit cross-links could make those junctions more connected. The chief executive’s Finance and Remuneration seats could bring management knowledge into the room. Sunday Folayan’s position on Remuneration while also presenting the overall committee table could support Board-level continuity. Yet the public materials do not identify the access rules that would show how information actually moved.

Challenge is the difficult element because it cannot be inferred from the number of names. Three people can challenge one another more rigorously than ten, or simply agree more quickly. Independence is not a personality trait that can be read from a roster. It is partly produced by procedure: clear remit, access to evidence, the right to commission or request further work, protected dissent, conflict rules and a route for concerns to reach the full Board.

Recommendation completes the chain. The annual report describes Finance and Remuneration in recommendation terms and Audit as assisting the Board over specific assurance domains. Committees do not replace the Board merely because they specialise. A good record therefore connects each recommendation to a Board response: accepted, rejected, returned for more work or noted. Without that connection, readers see committee existence but not corporate accountability.

The resolution’s consequence lies in this control chain, not in any imagined command over the Internet. AFRINIC’s registry function can make accurate administration and continuity important to network operators. Dependence on a registry service can give internal decisions practical leverage. But leverage arising from reliance on bookkeeping infrastructure is not sovereignty. A committee’s authority is whatever bounded corporate remit the company lawfully delegates internally; it is not authority over African states, the African public, operators as a political class, Internet users or the number resources recorded in the registry.

The retreat, the motion and the limits of unanimity

The adoption sequence looks orderly. Sunday Folayan presented the table. Haitham El Nakhal proposed the resolution. Andrew Alston seconded it. The Board passed it unanimously. Each fact matters because it makes the formal decision traceable and identifies participants in the corporate act.

Unanimity can reasonably support a modest conclusion: the directors who voted agreed to the roster. It may indicate that the arrangement was considered workable and uncontroversial by those present. It may also reflect preparatory discussion at the retreat that resolved questions before the meeting. The public record does not provide enough detail to choose among those possibilities.

What unanimity cannot do is enlarge the nature of the institution. A unanimous private-company Board vote is still a private-company Board vote. It is not an election by an African public, consent from every member, endorsement by all network operators, authorisation from governments, or a mandate from end users. Calling a broad set of participants “the community” cannot transform consultation, attendance or expertise into sovereign delegation. Community participation may improve information and expose the Board to practical experience. It does not create a continental principal capable of granting public power to a registry.

That boundary is not a semantic quarrel. It determines the standard by which the resolution should be judged. If the Board were a legislature, analysis might ask whether it had constitutional power to constitute regulatory committees. It was not. The relevant questions are corporate: Did the Board allocate oversight coherently? Were conflicts managed? Did committees remain accountable to the Board? Could members see enough of the chain to evaluate stewardship? Were roles thin, reviewable and reversible?

The retreat reference deserves the same discipline. Informal discussion before a formal vote is neither surprising nor inherently suspect. Boards need spaces to compare skills, workloads and preferences. But when that discussion produces the architecture of oversight, a concise public rationale is valuable. It can say, without exposing private deliberation, which competencies the Board sought, why overlap was necessary and which safeguards accompanied it. The minutes record the existence of prior discussion but not that explanatory bridge.

Routine reconstitution is the strongest contrary case

The fairest reading in the Board’s favour begins with turnover. AFRINIC’s annual report says committees were re-formed when a new Board was elected, and the 2016 Board change took effect at the beginning of July. Reconstituting the committees on 8 July could therefore have been routine housekeeping performed promptly after directors changed. A Board does not need a crisis or controversy to refresh its working groups.

The composition can also be defended on operational grounds. A small organisation may need the same financially literate director in more than one committee. The Finance chair’s full membership of Audit could provide context that prevents audit work from becoming detached from planning realities. The Audit chair’s observer status on Finance could expose assurance concerns early without merging the two bodies completely. Barrett’s Finance seat could ensure the committee received management information directly. His Remuneration seat could help with workforce and organisational context, subject to proper exclusion from personal matters.

Unanimous adoption may show that directors considered the arrangement balanced and practical. Detailed terms of reference, conflict procedures, chair appointments, calendars and working papers may have existed inside the company even though they are not reproduced in the public material available for this analysis. Committee members may have challenged one another effectively. The Board may have received careful reports and acted on them. Nothing in the roster disproves these possibilities.

This contrary case is strong because it explains the architecture without alleging grand motives. It also clarifies why the missing evidence matters. If cross-membership was chosen to prevent silos, a charter and concise rationale would demonstrate the design. If management participation was carefully bounded, recusal rules and a decision register would show it. If the committees delivered useful recommendations, Board responses and a year-end effectiveness review could establish the outcome. Transparency would not merely satisfy critics; it would substantiate the sensible case.

The absence of those documents from the public sources reviewed here does not prove that no private documents existed. It proves only that the public control chain cannot be reproduced from the material cited. This is not a demand to disclose every salary discussion, audit vulnerability or commercial detail. It is a demand for enough structural evidence to distinguish an intentional control bridge from an unexplained overlap.

The charter problem: a framework without the instrument

The 2016 annual report gives readers a broad assurance about committee governance. Committees had defined lifetimes and terms of reference that described role, membership and delegated power or authority. This suggests that Resolution 201607.290 sat within an established formal framework, not a wholly improvised arrangement.

Yet the report’s institutional description cannot substitute for the exact instruments applicable to the July committees. The resolution itself does not attach their terms. The public materials cited do not establish when each term began or ended, who held each chair at the moment of the vote, what quorum applied, whether observer rights were defined, how conflicts were treated or what matters were reserved to the full Board. The annual report reproduces the July-to-December rosters consistently with the resolution, but a consistent roster is still only one layer of the control system.

This gap has particular importance because the resolution uses two different membership concepts—observer and full member—without publicly setting out their operational consequences. If the observer could attend all Finance sessions, receive confidential papers and speak, that is one design. If the observer attended only selected items, that is another. If the Finance chair counted toward Audit’s quorum and voted on recommendations, that differs from a role limited by automatic recusal. Without the charter or a concise public summary, outsiders must supply assumptions that the Board could have resolved with a few lines.

Delegation also requires precision. The annual report says terms of reference described any delegated power or authority, but the committee descriptions emphasise assistance, monitoring and recommendations. Those words are compatible with a Board retaining final decisions. The exact line matters: could a committee approve anything, or only recommend? Which spending, personnel or assurance matters were reserved? When did an issue have to return to the Board? Publication of delegated limits protects both the committee and the organisation because it prevents responsibility from dissolving between them.

The adjacent Board actions on 8 July underline the need not to mix instruments. Resolution 201607.291 separately reconstituted four ad-hoc committees. Resolution 201607.292 approved a delegation-of-authority document dated that day and kept that document internal. Those decisions provide context for a Board organising its machinery, but their contents cannot be imported into Resolution 201607.290. The unpublished delegation document cannot be used to fill gaps in the standing committees’ remit. Different resolutions must remain different evidentiary objects.

Why the missing chain matters without creating public power

It is tempting to choose between two exaggerated positions. One treats an Internet registry’s committees as obscure corporate administration that nobody outside the Board need understand. The other inflates registry governance into a quasi-governmental system ruling a region’s Internet. Both miss the real institutional problem.

AFRINIC’s work is narrow but useful. It keeps and coordinates records that help operators manage Internet number resources. Reliability, continuity and accurate administration matter because networks depend on stable operational facts and predictable registry service. A poor internal decision about finance, controls or executive incentives can therefore concern members and operators even when the registry has no public authority. Practical consequence does not require sovereignty.

This is why committee records deserve scrutiny in proportion to the registry’s actual function. Finance oversight bears on whether the company can sustain its service and explain fees. Audit oversight bears on whether financial information, controls, risks and systems receive credible challenge. Remuneration oversight bears on whether executive incentives align with organisational duties rather than personal discretion. These are familiar private-company concerns intensified by operational dependence, not evidence that the company owns or governs the Internet.

The distinction also changes the direction of accountability. A regulator may justify coercive power through law, jurisdiction, procedure and appeal. A private registry should keep its power thin: limited to service, contractual and corporate functions; reviewable through records and corporate governance; and reversible when a decision proves mistaken. Internal committees are useful precisely because they can discipline a Board’s own choices. They become dangerous conceptually when their recommendations are presented as commands carrying public legitimacy they do not possess.

“Community” does not solve this problem. Meetings, consultations and participation can provide evidence, expertise and criticism. They can help directors understand operational consequences. But a shifting community is not a sovereign electorate, and participation is not a transfer of governmental authority. The Board’s unanimous vote proves corporate agreement. Its legitimacy must therefore rest on bounded purpose, competent stewardship, transparent process and accountability to the relationships the company actually has—not on a claim to speak for a continent.

The economic and operational stakes in each room

The Finance Committee’s work sits close to the company’s ability to continue operating. Financial planning allocates scarce resources. Reporting enables the Board to compare plans with results. The June fee-structure assignment suggests that committee work could also affect what members were asked to pay, though the record does not establish what the July committee ultimately did with that assignment. A robust Finance process would expose assumptions, compare scenarios, identify sensitivities and explain recommendations to the Board.

Audit operates on a different time horizon and epistemic footing. It asks whether reported figures and underlying controls deserve confidence; whether risks are recognised; whether internal audit has access and independence; and whether information systems and IT governance support reliable operations. The annual report’s description is broad enough to show that Audit’s field extended beyond year-end accounts. For an organisation providing registry services, information-system governance can be directly relevant to continuity and record integrity.

Remuneration connects institutional aims to human incentives. Policy on executive and senior-management pay, especially performance-related elements, determines what behaviour is rewarded and how performance is measured. The June references to evaluation and key performance indicators show the committee dealing with the machinery that translates organisational priorities into executive assessment. Poorly specified measures can encourage the wrong optimisation; well-specified measures can align management with continuity, accuracy, financial discipline and service.

The committees’ subjects therefore intersect. A performance target may have budget consequences. A budget choice may alter a control environment. An audit finding may reveal that an incentive or financial plan created excessive risk. Cross-membership can help the Board see those connections. The danger is not connection itself but unmarked role-switching. A person who advocates a financial choice in one room may need to disclose that role, refrain from leading assurance over it in another, or ensure that a fresh challenger is present.

For members and operators, the practical interest lies in whether these processes keep the registry accurate, solvent and dependable. No source cited here attributes an outage, financial loss, control failure or infrastructure harm to Resolution 201607.290. It would be irresponsible to invent one. The stakes are prospective and structural: committee architecture affects the probability that risks are noticed and that the Board receives candid advice before a problem grows.

This prospective character is why records should be designed before failure. After an incident, minutes and charters become forensic evidence. Before one, they are control tools. A clear recusal rule can prevent a conflict from contaminating a recommendation. A decision register can reveal that the Board repeatedly deferred a control issue. A year-end review can identify that cross-membership produced excessive workload or insufficient challenge. Governance should not wait for damage to make its own architecture legible.

What can be proved—and what remains inference

The strongest claims about the resolution are simple. The Board reconstituted three standing committees on 8 July 2016. Finance comprised Krishna Seeburn, Haitham El Nakhal, Christian Bope and Alan Barrett, with the Audit chair as observer. Remuneration comprised Lucky Masilela, Sunday Folayan and Alan Barrett. Audit comprised Andrew Alston, Seun Ojedeji and Abibu Ntahigiye, with the Finance chair as a full member. Folayan presented a table previously discussed informally at the retreat. El Nakhal moved the resolution, Alston seconded it, and it passed unanimously.

The annual report supports the roster and explains the committees’ institutional purposes. The June minutes show predecessor work in fee structure, executive evaluation and recruitment, and an audit report. Together these records establish that the July act allocated oversight over active subjects and linked two committees explicitly.

They do not establish why each person was selected. They do not identify the chairs in the resolution’s operative text. They do not show which skills, conflicts, workloads or independence considerations were assessed. They do not reproduce the retreat table, the exact July terms of reference, meeting schedules, quorum provisions, recusal rules, access controls, minutes, deliverables or decision logs. They do not show whether recommendations were accepted or how effective the committees were.

This separation between proof and inference should survive in both criticism and defence. A critic cannot treat overlap as proof of capture, collusion or failed independence. A defender cannot treat a formal roster or unanimous vote as proof of rigorous challenge. Both claims leap over the missing operating record.

Nor should absence from the cited public record be recast as proof that an internal document never existed. Private companies can retain confidential charters, papers and minutes. The proper conclusion is about public reproducibility: members and observers cannot verify the full chain from appointment through committee work to Board response using the published material identified here.

That is a meaningful finding even without an allegation. Governance quality is not only the quality of decisions reached in private. It is also the organisation’s ability to show, at an appropriate level, how responsibility was allocated and exercised. A public summary can protect confidential details while revealing chairs, terms, delegated boundaries, attendance, recusals, recommendations and Board disposition.

A record designed around decisions, not publicity

What would a stronger record look like? It would begin with the approved roster but not end there. Each committee would have a published chair, term and appointment rationale expressed through competencies rather than personal praise. A brief conflict statement would identify relevant categories of interest and explain how they would be handled. The purpose is not to publish private financial information; it is to show that the Board considered role compatibility before assigning access and influence.

Next would come the committee charter or a faithful public summary. It would define remit, matters reserved to the Board, any delegated approvals, access to staff and advisers, the status of observers, quorum, voting, recusal and reporting. For Finance and Audit, it would state how the two chair links operate when one committee reviews work connected to the other. For Remuneration, it would explain how management supplies information while being excluded from decisions concerning its own pay or evaluation.

A calendar would connect standing purpose to actual work. The June assignments offer examples of deliverables that could be tracked: fee-structure work, executive-performance material and an audit report. The July committees could publish whether an inherited task was adopted, revised, completed or returned to the Board. Dates need not expose confidential content. They would show whether a “standing” committee actually stood behind a programme of work.

Concise minutes or a decision register would then capture attendance, recusals, principal evidence considered, recommendations and any dissent. A recommendation need not disclose sensitive numbers to identify its subject and direction. The Board’s response should be recorded alongside it. That closes the accountability loop: a committee advised; the Board accepted, rejected, amended or deferred; responsibility remained visible.

Finally, a year-end effectiveness report would test the architecture itself. Did cross-membership improve information flow? Did it cause repeated recusals or concentrate workload? Did committees receive papers in time? Were there gaps between Finance’s planning assumptions and Audit’s risk assessment? Did Remuneration’s performance measures align with service continuity and control quality? This review could recommend changes before the next annual reconstitution.

Such a record is not bureaucratic excess. It is proportional to the control surface. Much of it could fit in short tables. Its value lies in separating legitimate confidentiality from avoidable opacity. Personal compensation amounts, security-sensitive audit details and commercial information can remain protected while the existence of procedures and the disposition of decisions become visible.

A counterfactual July decision

Consider how Resolution 201607.290 might have read under this stronger approach. The Board would still name the same Finance, Remuneration and Audit members. It would also identify each chair and the term ending at the next specified Board transition. A short annex would list sought competencies: finance, risk, governance, executive oversight and operational knowledge. Relevant conflicts would be declared or recorded as none, with continuing disclosure obligations.

The Finance–Audit cross-link would receive a paragraph. The Audit chair’s observer role would permit access and voice but not voting or quorum status. The Finance chair’s Audit membership would be paired with a rule requiring recusal when Audit reviewed a matter in which that chair had led Finance’s recommendation. An independent member would lead such an item. Whether those exact rules would have been appropriate is for the Board to decide; the point is that the design choice would be explicit and testable.

The chief executive’s participation would be similarly bounded. Barrett could provide operational and financial information to Finance and organisational information to Remuneration. He would leave for discussion and recommendation on his own evaluation, remuneration or any matter producing a direct conflict. The minutes would record departure and return without publishing private personnel substance.

The resolution or attached charter would set quorum, access rights, reporting intervals and delegated limits. It would identify which June assignments transferred to the new committees. At the next Board meetings, committee reports would appear as decisions or recommendations with responses. At year end, the annual report would provide an effectiveness summary rather than only a purpose statement and roster.

This counterfactual does not presume that AFRINIC lacked all such private practices. It illustrates what a reproducible public control chain would contain. It would allow members to evaluate cross-membership without guessing, help future directors understand inherited responsibilities and protect committee members from allegations based solely on topology.

Keeping the bookkeeper’s power thin

There is a wider lesson in this small corporate act. Institutions that maintain important records can be tempted—or encouraged by others—to describe operational dependence as public authority. The language of “community”, “region” and “resource governance” can make an internal decision sound like an exercise of continental mandate. Resolution 201607.290 supports no such interpretation.

The committees assisted a private Board. Their remits concerned the company’s finance, controls, risk and executive arrangements. Even if their work affected the quality or continuity of registry service, that effect did not create a right to coerce operators or speak for end users. Internet number records should remain a service to operational reality. The ledger describes and coordinates; it does not confer ownership of the networks whose facts it records.

Keeping power thin does not mean keeping governance weak. It means the opposite. A narrow institution should have precise, reviewable controls so that practical leverage cannot drift into unaccountable discretion. Committee roles should be explicit. Delegation should be limited. Conflicts should be managed. Decisions should be reversible when evidence changes. Records should enable those affected by the service to distinguish fact from institutional assertion.

The July architecture contains ingredients that could support thin power. Audit was separate from management membership on the face of the roster. The Audit chair observed Finance rather than joining it as a full member. Three committees divided attention rather than placing every sensitive subject in one group. The full Board formally adopted the arrangement.

It also contains ingredients that demanded clarification. The Finance chair entered Audit as a full member. The chief executive sat on the committee concerned with remuneration. The exact chairs and operating rules were not stated in the published resolution. The retreat rationale and subsequent decision chain were not reproduced. These are not verdicts. They are the points at which an internal control map either becomes auditable or remains a diagram whose arrows outsiders must imagine.

The measure of the three smaller rooms

Resolution 201607.290 should neither be inflated into a constitutional moment nor dismissed as a clerical update. It was a consequential, bounded act of private-company governance. It placed nine people across ten committee seats, assigned them access to distinct kinds of information and connected Finance to Audit through two unequal roles. It put the chief executive inside Finance and Remuneration and outside Audit. It was adopted through a recorded, unanimous Board vote after prior informal discussion at a retreat.

The sensible case for the design is readily available. Annual Board turnover required reconstitution. Cross-membership could prevent silos. Management could supply indispensable facts. Unanimity could reflect a workable allocation. Private charters and conflict procedures may have supplied safeguards that the short public record does not show.

The accountability case is equally straightforward. When the same person crosses from making or framing a financial recommendation into the committee responsible for assurance, roles need definition. When a chief executive sits on Remuneration while executive evaluation is part of the workload, recusals need definition. When an Audit chair observes Finance, access and speaking rights need definition. When committees assist the Board, recommendations and Board responses need a visible chain.

The missing public evidence does not establish misconduct, failed independence or bad outcomes. It establishes uncertainty about mechanism and performance. The appropriate response is not a dramatic claim of regulatory failure. It is a practical disclosure design: chairs, terms, criteria, conflicts, charters, quorum and recusal rules, calendar, deliverables, concise decision records, Board responses and a year-end review.

That would let members and operators judge whether the smaller rooms produced sharper attention and honest challenge or merely redistributed familiar voices. It would protect legitimate confidentiality while making stewardship reproducible. Most importantly, it would keep the institutional scale honest. AFRINIC’s committees could strengthen the company that maintains a useful registry. They could not turn that registry into a government, its Board into a legislature or unanimous agreement into a mandate from Africa.