Summary

  • NANOG members elect six directors. Those directors appoint the Executive Director, who is the seventh voting director, and the resulting member-elected-majority Board appoints specialist committees under rules that differ markedly from body to body.
  • Expert appointment can be more workable than electing every volunteer committee. Its legitimacy is easiest to assess when the public can see the instrument, scope, selection route, term, Board connection, conflicts rule, output, review and end condition for each delegation.
  • Program and Election reveal the central trade-off. Their outputs are visible and their mandates matter, yet important parts of their decision trails—and two inconsistencies in the governing documents—remain difficult to reconstruct publicly.

Two votes at the hinge of the system

On 4 February 2026, at NANOG 96, the Board made two adjacent decisions. It accepted a slate of committee appointees and then appointed Board liaisons. Both motions passed with seven votes in favour, none opposed and no abstentions, according to the public Board minute. The identical count does not make them the same act. It also says nothing about unanimity among NANOG members, conference attendees, committee volunteers or unsuccessful candidates. It records agreement by the seven voting directors on two corporate motions.

The distinction is more than procedural. Appointment determines who serves. A liaison creates a channel between a committee and the Board. Neither fact alone establishes whether the committee may decide, must recommend, simply delivers a service or needs later Board approval. Treating appointment, liaison, delegated decision and ratification as interchangeable makes the authority map look simpler than it is.

That map starts with a ballot. NANOG members elect six directors. Those elected directors appoint an Executive Director who serves at their pleasure and becomes the seventh voting director. The arithmetic—six elected directors plus one Executive Director equals seven voting Board members—supports a precise description: NANOG has a member-elected-majority Board, not a wholly elected Board. From that Board, authority branches toward bodies that choose conference content, administer elections, oversee audit work, organise mentoring, moderate community channels and develop workshops.

The current bylaws state that amendments proposed on 23 September 2025 were adopted by NANOG members on 5 November 2025. That date belongs to the current text; it does not mean that every clause was invented in 2025. Public appointment records reaching back more than a decade show an architecture that has been altered repeatedly rather than created all at once.

The bylaws place NANOG's property, affairs and business under Board management and control. They also allow the Board, by general resolution, to delegate powers permitted by the bylaws to officers and committees. A committee's corporate mandate therefore comes from a Board act and the instrument defining it, not from an inherent claim to speak for the community. Direct member authorisation reaches the six elected directors. Beyond that point, legitimacy rests on accountable delegation.

The scope of that delegation must also be kept in bounds. NANOG describes itself as a facilitator of technical information exchange and professional relationships, not an Internet operator. Its committees exercise corporate, conference and community-channel authority inside NANOG. They do not regulate North American networks, direct employers or control routing. NANOG, Inc., its members, directors, Executive Director, staff, committee members, liaisons, attendees, sponsors, vendors, employers and network operators remain distinct actors.

Why appointment is defensible

The strongest case for appointment is practical. A professional association needs people prepared to review technical proposals, work with auditors, mentor newcomers, moderate difficult conversations and design specialised workshops. Asking members to elect every volunteer body would impose a large information burden on voters and candidates. Visibility on a ballot is not the same as competence for patient, sometimes confidential work. Campaigning could deter qualified volunteers. More elections would not automatically produce better expertise, steadier service or stronger handling of conflicts.

NANOG's committee nomination process gives that defence substance. An annual public call supplies the candidate pool. Only members in good standing may serve. Since 2024, new and returning committee members have had to sign a nondisclosure agreement. Confidentiality can protect candidate, entity, incident and personnel information; an NDA is not evidence that accountability has been suppressed. At the same time, an open call is not a committee election. Members do not vote on the appointments, and the process page does not expose comparative rankings or reasons for each nonappointment.

The 2026 call covered eight bodies: Program, Audit, Community Engagement, Election, Hackathon, Moderation, Mentorship and Workshop. It ran from 5 January to 3 February, required nominees to accept and complete a questionnaire, and said service typically demanded at least one hour a week. It announced appointment by the Board on 4 February. The later minute confirms the formal votes, but does not reproduce the appointee names, applicant rankings or reasons other candidates were not selected.

This is a serious operational model: an open route into a volunteer pool, a good-standing requirement, a questionnaire, confidentiality where needed, Board selection, defined terms in many instruments, continuing Board accountability and visible public work. Appointment can be entirely legitimate. A useful reform should make that model easier to inspect, not replace expertise with indiscriminate plebiscites or require publication of private files.

The five links that determine accountability

A roster answers who serves today. It does not, by itself, show the five links that matter most.

The Board first appoints committee members. That identifies the selector but not necessarily the comparative criteria, duration or renewal test. It may then assign a liaison, creating an information path without automatically giving that person a committee vote or daily command. A committee may receive authority to decide within a bounded field, as Program does when choosing content. It may instead recommend while the Board retains final approval, as Workshop does on certain commercial and intellectual-property questions. Finally, it may produce or report something the public can observe: a conference programme, election result, audited statement, mentoring activity or moderation statistic.

These links distribute authority differently. Board ratification can protect accountability because elected governors retain responsibility for consequential decisions. It can also concentrate several controls in one place if the Board appoints members, revises the charter, chooses a liaison and makes the final decision. That concentration does not prove day-to-day interference. It does make the boundary between delegated and retained power worth publishing.

The current structure is not governed by one reusable committee rulebook. The public committee index lists Audit, Community Engagement, Compensation, Election, Hackathon, Mentorship, Moderation and Workshop. Program is the only bylaw-named standing committee. Compensation is a Board subcommittee. Audit, Community Engagement, Election, Hackathon, Moderation and Workshop have public ad-hoc charters. The Mentorship charter available to the public is not labelled ad hoc. Calling all nine standing, or all nine ad hoc, would erase distinctions in their actual instruments.

Conflict and tenure rules are just as body-specific. Bylaw 15.2 expressly covers Directors and officers: it requires disclosure of certain organisational affiliations and nonparticipation in the corresponding transaction or arrangement. It is not a universal volunteer-conflict clause. Audit refers potential conflicts to the Board; Election contains a candidate-specific recusal rule; Compensation excludes defined people from membership. Other charters rely on their own confidentiality, conduct or escalation language.

Standing Committee members normally serve staggered two-year terms, with roughly half the seats filled each year. They may serve no more than two full consecutive terms, ordinarily no more than five continuous years, and may return after at least a year away. The Board fills vacancies and can remove a Standing Committee member early with at least five votes. Those rules matter for Program. They do not become an ad-hoc removal standard merely because they appear in the same bylaws. Audit and Community Engagement have their own two-term ceilings. Hackathon, Moderation and Workshop permit renewable terms without stating that cap.

Mentorship's public charter does not state a committee-member term at all.

Even a general Board explainer shows why the controlling instrument matters. The Board responsibilities page correctly says that the Board appoints committee members, but describes two-year Board terms. Current bylaw 4.3 says three years. That mismatch is public-document drift, not a choice between equally current rules. The bylaw controls the term claim.

Program: visible selections, less-visible reasoning

Program offers the clearest example of delegated decision-making. Under the bylaws, the Board determines the number and location of programmes while the Program Committee selects their content. The committee's public page shows the body and current roster, but the bylaw supplies the division of authority. Programme selection is consequential: it shapes which operational lessons, technical arguments and speakers reach the NANOG stage. It is still conference authority, not regulatory power over the networks discussed there.

The standing committee must have at least sixteen members. Members must be in good standing, and eligibility includes attendance at a NANOG conference during the preceding year. The Board appoints the members after the Board election; the committee organises itself and elects its officers. Its chair participates on the Board ex officio without a vote. That chair is not the Executive Director, who is a voting director, and neither role is synonymous with a liaison.

The governing text also contains a conspicuous tension. Article 7 says at least one member of every standing or ad-hoc committee must be a Board member. Article 7.1 says no elected director may serve concurrently on a standing committee and permits the Board to appoint a liaison to observe and facilitate communication. The public record does not explicitly reconcile the general Board-member requirement with the standing-specific exclusion and liaison option. A current roster cannot settle the meaning.

The correct public conclusion is that the two provisions sit in tension, not that a liaison secretly votes or that the committee has no Board connection.

Program's output is unusually measurable. NANOG's 2024 annual report says that the committee reviewed 206 submissions and accepted 105. The ratio, 105 divided by 206, is about 51.0 per cent. It is an acceptance rate, not a quality score. Published programmes and aggregate totals show that the work occurred and at what scale; they do not expose individual scores, committee votes, recusals, conflicts or any appeal decision.

A February 2019 selection minute provides a different denominator. The Board reviewed 31 Program submissions, expanded the committee to a minimum of 20 and maximum of 25, then made three one-year and thirteen two-year appointments. Three plus thirteen equals sixteen appointments. It would be unwarranted to call the other fifteen merit rejections: eligibility, withdrawal, alternate status and other conditions are not fully reported.

Program is therefore not a black box. Its authority is named, the Board appointment route is stated, standing-committee terms are bounded, officers are internally chosen and programmes are public. Yet the public can more easily see what the committee produced than how it reached an individual selection or how the Board evaluated a renewal. Output transparency is comparatively strong; decision traceability is thinner.

Election: an appointed body inside the electoral loop

Election presents a more sensitive loop. Members elect the six directors who make up the Board's elected majority. The Board then appoints the body that administers the next election. Appointment is not evidence that the committee is controlled in its daily work. It does mean that independence must be demonstrated through bounded authority, conflicts discipline, ballot security, retained decisions and reviewable outputs rather than assumed from the committee's name.

Bylaw 10.3 requires an Election Committee of at least three people, appointed by a majority of elected directors whose terms are not expiring. It says that the committee serves until the election concludes. The Election charter also requires at least three Board-appointed members in good standing, including a Board member not then up for re-election. It assigns the committee responsibility for election forms, rules, candidate forums, quorum and ballot security. A committee member who accepts a Board nomination must recuse and leave the committee.

Those are meaningful controls. The restriction on participating directors narrows the appointment decision, and the recusal rule addresses a direct candidate conflict. Public results reveal the electoral output. But the Board retains at least some rule approval. In May 2025, on the committee's recommendation, the Board formally adopted ranked-choice voting for future Board elections, as the decision record shows. The committee recommended; the Board decided.

The term is described in two unreconciled ways. The charter states two-year terms capped at two consecutive terms. Bylaw 10.3 says service continues until the election concludes. One could imagine those provisions applying to different aspects of service, but the public documents do not say so. The discrepancy is a documentary inconsistency, not evidence that an election was invalid or the committee unlawful.

The 2024 annual report supplies another useful calculation: 165 votes cast from 698 eligible voters, or about 23.6 per cent. It also names two elected directors. Turnout describes participation in that election. It is not a committee approval rating, a measure of satisfaction or evidence that NANOG's electorate represents every attendee, employer or network operator.

Election again separates outcome from trace. Winners can be announced, forms and forums administered and ballot security assigned even where an outside reader cannot reconstruct challenge handling, individual recusals or the operative relationship between the two term provisions. Those public gaps do not show that internal controls are absent. They show where confidence still depends more on institutional assurance than on a record others can retrace.

Seven bodies, seven different allocations of control

The remaining committees confirm that one body's safeguards cannot be imported into another.

Audit is Board-appointed under an ad-hoc charter. It has at least three members in good standing, including at least one Board member, and staggered two-year terms capped at two consecutive terms. Potential conflicts go to the Board. The committee must meet at least twice a year, appoint the independent auditor, review the audit, and send the final audited statements and its auditor-contract recommendation to the Board. Selecting an auditor, executing the contract and accepting the statements are distinct steps. The 2024 report says Audit received the auditor's report and forwarded it after review. A 2025 community presentation later reported selection of a new auditor, completion of review of the 2024 audit and no issues or concerns. Those outputs do not reveal every comparison, conflict referral or deliberation.

Community Engagement is Board-appointed under a charter requiring at least one Board member and two other members in good standing. Its two-year terms are capped at two. It works on engagement, diversity, acceptance, policy and strategic recommendations, while direction and charter changes require Board approval. Quarterly reporting to both Board and community creates a regular review surface. Selection rankings, individual votes and a committee-specific appeal route are not public in the instrument. Nor do overlapping concerns establish that Community Engagement is a legal rename of the former DEI committee.

Compensation is categorically different. Its 2025 charter makes it a Board subcommittee, reconstituted annually at the first Board meeting. At least three elected directors are voting members; external experts may participate without a vote. The Executive Director's direct reports, and people whose compensation the Executive Director determines, cannot serve. The committee's final report and compensation recommendation go to the full voting Board for approval or rejection in executive session. Confidential personnel reasoning remains private, but the authority path—restricted membership, annual cycle, recommendation and full-Board decision—is visible.

Hackathon is Board-chartered as ad hoc. Its charter requires at least four Board-appointed members and representation from both Board and Program. Terms normally last two years; successive reappointment is allowed without the two-term ceiling used by Audit and Community Engagement. The committee develops formats, topics, projects, tutorials, technical environments and sponsor coordination. The Board retains approval over charter and direction changes. Events and results display delivery, but the charter gives no fixed report cadence, universal removal rule or automatic sunset. Renewability is not tenure or evidence of impropriety; it is a different public control design.

Mentorship has another structure. Its charter is not labelled ad hoc. The Board appoints committee members, and the committee in turn appoints and trains community members as mentors. It must use programme metrics and surveys, an NDA and the Code of Conduct, while programme recommendations return to the Board for consideration and final approval. The public charter does not state committee-member terms, removal conditions, a formal approval date or a fixed Board-report cadence. Those are unknowns, not proof of perpetual service or absent review. The 2024 report described semi-monthly meetings, matching, meeting presentations, an Ambassador launch and a help desk; the 2025 presentation listed Ambassador, mentoring and newcomer orientation. Activity is visible, while entity-level success, complaints, reassignment and appeal handling are not.

Moderation is Board-chartered as ad hoc. Its charter provides for at least four Board-appointed members, including a Board member, with normally renewable two-year terms and no stated universal cap. The remit covers NANOG Discord and mailing lists, not official social-media accounts. Serious matters escalate to staff, while platform statistics and summarised incidents go to the Board liaison. Staff escalation does not make the liaison the moderator, and reporting does not show that the Board decided each case. Privacy may justify withholding incident details; a summary path is nevertheless not the same as a public appeal process.

Workshop is likewise a Board-approved ad-hoc body. Its charter calls for at least four Board-appointed members with Board and Program representation. Its normally two-year terms are renewable without a stated cap. The committee selects and develops workshops, shepherds content and coordinates delivery. Recommendations involving fees, commercial sponsorship or intellectual property return to the Board. A recommendation is not final authority. Public workshops and a formalised submission-and-evaluation format show work, but not individual scoring or appeals. Similar subject matter does not prove that Workshop legally inherited all of the dissolved Education Committee's powers.

Across all nine bodies, variation can be sensible: audit, compensation, moderation and programme selection face different risks. Yet the public instruments offer no common view of term length, renewal ceiling, removal, conflicts, report cadence, appeal and sunset. Different design is defensible. Leaving the differences difficult to compare is less so.

Where the final decision actually sits

The comparison becomes clearer when the nine bodies are grouped by the kind of authority they exercise rather than by the word “committee.” Program holds a defined selection power: within the number and location of programmes determined by the Board, it chooses content. Election administers a corporate process but returns at least some rule changes to the Board. Audit selects and works with an auditor, then sends statements and a contract recommendation onward. Compensation formulates a recommendation whose legal and financial consequence depends on the full voting Board. Workshop proposes on fees, sponsorship and intellectual property.

Mentorship makes programme recommendations for Board consideration. Moderation handles routine platform work while escalating serious matters to staff. Hackathon and Community Engagement operate within direction that the Board can formally change.

Those are not small drafting differences. They tell a member where to direct a question. If a conference submission is declined, the relevant delegated decision appears to sit with Program. If the issue is how many programmes NANOG will run or where, it sits with the Board. If an election rule changes, a committee recommendation may explain the proposal, but a Board record may contain the decisive act. If an auditor is chosen, the committee's selection should not be confused with execution of the contract or acceptance of the audited statements.

If a workshop raises sponsorship terms, the committee can supply expertise without acquiring the Board's final corporate authority.

This separation also prevents two opposite errors. The first is to assume that a Board-appointed body is merely advisory and therefore harmless. Program's content choice is a real delegated decision, and routine moderation decisions affect access to NANOG's own channels. The second is to assume that appointment means the Board personally directs every operational choice. Charters assign work to committees precisely because daily judgement has been delegated. A reviewable system must show both the reality of committee discretion and the line beyond which the Board or staff takes over.

The Board connection changes with the function as well. Compensation is composed as a Board subcommittee and returns its recommendation to the full Board. Audit and Community Engagement expressly include Board members. Hackathon and Workshop require both Board and Program representation. Moderation includes a Board member and reports summaries to a liaison. Election requires a Board member not up for re-election, while its appointment vote excludes elected directors whose own terms are expiring. Program presents the unresolved standing-committee language and gives its chair nonvoting Board access.

“Board involvement” is therefore too blunt a category to establish either independence or control. Membership, liaison, ex officio access, reporting and final approval must be named separately.

Term design reveals another choice about how expertise is renewed. Capped service can rotate access and limit entrenchment, but it can also force out scarce expertise. Renewable service can preserve knowledge, but it places more weight on the quality and visibility of renewal review. Annual reconstitution suits Compensation because it is a Board function tied to a recurring personnel cycle. Staggering suits Program and Audit because continuity matters while membership changes. Interim appointments, such as Moderation's initial terms ending in February 2025, can bridge the creation of a new body to the regular selection cycle.

None of these arrangements is inherently illegitimate. Each calls for a different public answer to the same question: who reconsidered the appointment, using what stated criteria, and when?

Removal and end conditions are where the comparison is thinnest. The five-vote early-removal rule supplies a clear path for Standing Committee members. Audit, Community Engagement and Election publish term caps, but a term cap is not itself an early-removal or appeal process. Renewable terms for Hackathon, Moderation and Workshop do not identify a sunset or continuation test. Mentorship's committee terms are not stated in its public charter. The ad-hoc label describes how several bodies were constituted; it does not disclose the event that ends each mandate.

A body can remain useful for years, but periodic usefulness should be demonstrated by a dated continuation decision rather than inferred from its continued appearance on a website.

Conflict controls likewise need to follow the decision being made. Election's rule is concrete: accepting a Board nomination requires recusal and departure from the committee. Audit routes potential conflicts to the Board. Compensation excludes people whose relationship to the Executive Director would compromise the body's composition. Bylaw 15.2 controls Directors and officers. These provisions respond to different risks and different actors.

A single generic statement that “conflicts are managed” would conceal the variation; extending the Director-and-officer rule to every volunteer would invent coverage the published text does not provide.

A comparable public record need not force identical governance on all nine bodies. It should instead expose the reason for each difference. Why does one role have a two-term ceiling while another is renewable? Why does one committee report quarterly while another has no stated cadence? When is a liaison an observer, when is a Board member part of the committee, and when is the committee itself made of directors? Which decisions are suitable for an appeal, and which are reviewed through Board acceptance or periodic reappointment?

Answering those questions would make specialised design intelligible without pretending that a programme-selection appeal, a confidential compensation assessment and a moderation incident should follow one universal procedure.

A system made by revision, not a fixed hidden constitution

The historical record shows repeated adaptation rather than one permanent structure.

In 2011, the Board selected eight Program Committee members for two-year terms, after which the committee would select its own chair and vice-chair, according to the contemporaneous announcement. That establishes appointment plus internal officer selection, not the committee's origin date. In 2012, a selection call and appointment notice covered Program, Communications and Development. Most terms were two years, with one one-year Development appointment alongside them. The material does not establish when Communications or Development later ended.

In 2017, the Board expressly created a NANOG On The Road ad-hoc committee and appointed four people to two-year terms, as its public notice records. No public dissolution date appears in the evidence here, so the body belongs to history rather than the current list. A 2018 bylaw-amendment proposal contemplated Board participation on committees, an Election Committee and transfer of Communications work to staff. A proposal establishes what was under consideration, not the vote result or the exact date when current language took effect.

The 2019 Program record then exposed its 31-submission funnel, size change and sixteen appointments without disclosing scoring, recusals or individual reasons. In 2021, February minutes say the Board and Program chair considered incumbent performance, expiring terms, size, composition and new submissions before the Board unanimously approved slates for Program, Mentorship, Education and Scholarship. Names and broad factors are public; candidate-level weights and votes are not. May minutes show Election and Audit charters entering Board review. A pending review proves a route through the Board, not the final adoption date.

In 2023, a February record shows Board approval of a revised Audit charter and liaison, together with an omnibus slate for Program, DEI, Education, Mentorship, Scholarship and Elections. A November minute records unanimous adoption of a Hackathon charter and appointment slate. Yet the 2023 annual report still described Hackathon as a Program subcommittee. The November charter created a separately chartered ad-hoc body, and the 2024 operational reboot was another stage. “Created in 2024” would wrongly collapse those phases.

In 2024, the Board adopted an updated Mentorship charter in April, shown in the April minute. In October it established Moderation, approved its charter, appointed three people to interim terms ending in February 2025 and separately approved the Board-election ballot, according to the October record. The annual report records committee and liaison appointments, several charter actions, creation of Moderation and dissolution of Education and Scholarship. Dissolving Scholarship's committee does not show that scholarships ended. Dissolving Education does not establish that Workshop inherited its duties.

In 2025, the Board approved the Workshop charter and dissolved Education in the January minute. The February record describes a selection workshop, one omnibus motion accepting appointees and a separate motion assigning liaisons for eight committees; the shared whiteboard slate is not reproduced in the minute. A March minute records one two-year Hackathon appointment. One person's term is not a system-wide rule. In June, the Board referred a hands-on vendor-experience idea to Workshop for feasibility work, required a proposal to return for Board review and changed the Moderation, Election and Workshop liaisons. That June record shows both retained Board authority and the adjustable nature of liaison assignments.

In 2026, the public call led to the separate 7-0-0 appointment and liaison motions with which this article began. The sequence makes sourcing and formal Board action visible. Comparative criteria, the full slate inside the minute, rankings and reasons for unsuccessful candidacies remain outside the public record.

Across these dates, Communications, Development, On The Road, Education, Scholarship and DEI appear and recede. Hackathon changes form; Mentorship is re-chartered; Moderation and Workshop are established; liaisons are reassigned. The record shows adaptation, not a timeless committee constitution or automatic legal succession. It does not prove that Community Engagement is DEI renamed, that Workshop inherited Education, or who now governs scholarship selection after the Scholarship Committee's dissolution.

Output transparency and decision traceability are different

NANOG publishes enough to defeat the claim that its committee system is a complete black box. Programmes and acceptance totals are public. Election results and turnout are reported. Audit work reaches audited statements and Board acceptance. Mentorship reports matching, Ambassador, help-desk and orientation activity. The 2025 community presentation named programme types and responsible bodies, Audit status, Community Engagement activities, Mentorship programmes, Moderation's scope and Workshop's formalised submission and evaluation format. These records show that delegated work occurred.

They do not answer every accountability question. A programme reveals what was selected, not how individual proposals were scored or whether someone recused. An election result reveals winners, not necessarily how challenges were handled. An audit outcome shows assurance work, not confidential auditor comparisons. Mentoring activities show delivery, not entity-level success, complaints, reassignment or appeals. Moderation statistics may show scale while appropriately withholding identities. Output transparency asks whether the product of delegated work is visible.

Decision traceability asks whether the authority, criteria, conflicts, intermediate actions, retained approvals and review route can be reconstructed. An institution can do well on one and unevenly on the other.

That distinction also leaves room for privacy. Publication of candidate questionnaires, unsuccessful applicants' identities, personnel assessments or incident details could harm individuals and reduce candour. Authority does not need to be secret merely because evidence considered in a decision is confidential. Aggregate candidate counts, dated appointment motions, recusal standards, term rules, the identity of the reviewing body and anonymised outcome measures can often remain public.

Affiliation is another poor shortcut. Nothing in these records shows that a committee volunteer acted on an employer's instructions simply because an employer was known. Professional background may supply expertise; it does not prove representation, capture or self-dealing. The better unit of analysis is the decision: who had authority, under which instrument, subject to what conflict rule, with which final approval and what reviewable result.

The unknowns must remain bounded. The public material does not disclose complete candidate questionnaires, demographics, rankings, renewal reasons, individual committee votes, recusal records or appeal outcomes. It does not establish a universal ad-hoc removal, conflict, appeal, reporting or sunset rule. Mentorship terms and fixed Board-report cadence remain unstated. Communications, Development and On The Road lack precise discontinuation dates. Compensation's individual assessments and voting reasons remain confidential. None of these absences proves concealment, institutional failure or that no internal process exists.

It means an outside reader cannot reconstruct that control from the public record.

What corporate law and reporting practice can—and cannot—show

The wider corporate framework explains why Board appointment is not inherently suspect. Delaware's nonstock-corporation provisions adapt references to shareholders and boards to member and governing-body structures where appropriate. Its board and committee provisions generally place corporate affairs under Board management or direction and allow committees created by Board resolution to exercise delegated authority within statutory limits. These provisions are structural comparators, not a legal opinion on NANOG's compliance.

The 2025 Form 990 instructions distinguish member governance rights from governing-body action and ask whether meetings and written actions of a governing body, and of committees authorised to act for it, were contemporaneously documented. That reporting guidance does not require NANOG to publish committee minutes, use a particular appointment method or disclose confidential deliberations. It does underline the difference between a body that advises and one authorised to act.

The lesson is modest. Members need not vote on every operational choice for delegation to be legitimate. A Board can appoint specialists and still remain accountable. The essential record identifies the authorising instrument, the limits of the grant and the matters retained by the Board. Public legitimacy adds a separate question: how readily can members and affected entities who were not in the room retrace that chain?

A delegation ledger would make the model easier to defend

NANOG already publishes most of the raw ingredients. The bylaws identify central powers. Charters explain purposes and composition. Nomination pages show how volunteers enter the pool. Board minutes record appointments, liaison changes, charter decisions and retained approvals. Annual and community reports show outputs. The weakness is fragmentation: a reader must find documents from several years, decide which version controls and avoid transferring one committee's rule to another.

A compact public delegation ledger could solve that problem without changing the appointment model. For each current body, it should identify the current instrument and approval date; delegator; delegated and retained decision rights; eligibility and selection route; term, staggering and renewal ceiling; Board-member or liaison role; applicable conflict and recusal rule; reporting cadence; complaint or appeal route where one is specified; removal authority; renewal test or sunset; latest appointment minute; latest public output; and unresolved conflicts among documents.

The ledger should be concise, versioned and linked. It should not publish private questionnaires, identities of unsuccessful candidates, personnel assessments, sensitive incidents or confidential deliberations. Where no public appeal route is specified, it should say exactly that rather than claim no internal route exists. Where two authoritative documents disagree, it should display both formulations until NANOG resolves them. Renewal, re-chartering and dissolution should be preserved as dated transitions rather than overwritten.

Such a ledger would allow different controls to remain different. Audit could display capped terms, conflict referral, its twice-yearly meeting floor and Board-facing outputs. Compensation could show annual reconstitution, voting restrictions and full-Board ratification while protecting personnel reasoning. Program could show its standing rules and content-selection power. Election could place the two term formulations side by side. Hackathon, Moderation and Workshop could state how renewable appointments are reviewed. Mentorship could add its term and review arrangement when those become public.

Community Engagement could connect quarterly reports to any formal Board response.

Most importantly, the record would stop a liaison being mistaken for an appointer, or an appointment being mistaken for a delegated decision. It would reveal when a committee makes the final call, when it recommends, when staff receives an escalation and when the voting Board must approve.

The finding, then, is neither scandal nor acquittal. Appointing specialists is not proof of capture, and NANOG's committees produce meaningful public work. The system's strongest defence is its practical combination of open volunteer calls, good-standing eligibility, Board appointment, specialised expertise, bounded terms or charters, Board accountability and visible output without forcing members to elect every working body.

The weakness lies in the variable public trace after direct member authorisation stops at the Board. Purpose, roster and output are often clear. Selection rationale, committee-specific conflicts, renewal reasons, removal paths, appeals and end conditions are much less consistent. The two February 2026 motions capture the difference in miniature: the Board chose committee members, then chose liaisons. A mature delegation record would let a reader follow both acts onward—from the member ballot, through a Board with an elected majority, into each specialist body, through a bounded decision and back to a reviewable result.

SEO and social

SEO title: After the Ballot: How NANOG Delegates Power to Appointed Committees

SEO description: How NANOG's member-elected-majority Board appoints nine specialist bodies—and why public outputs do not always reveal the full path from delegation to decision and review.

Social title: Who Decides After NANOG Members Vote?

Social description: Appointment can be the right way to staff expert committees. NANOG's harder governance test is whether each body's authority, conflicts, retained powers and end conditions can be traced.

Image metadata

Alt text: Synthetic editorial illustration of a membership ballot leading to a seven-seat Board docket and branching into nine distinct committee folders marked charter, term, report and review.

Caption: One ballot authorises a Board with an elected majority; separate appointment and liaison decisions connect nine specialist bodies with different rules.

Accessibility description: A clearly synthetic overhead editorial scene shows a paper membership ballot on the left linked to seven empty Board seats at the centre. Six seats carry election markers and one carries an appointment marker. Separate paths lead from the Board docket to nine colour-distinct folders, each carrying small markers for its charter, term, public report and review status; unresolved fields are shown as open markers rather than filled with invented information. There are no real people, organisation logos, employer symbols or glowing network graphics.

Synthetic provenance: This AI-generated conceptual editorial image represents an abstract governance process. It is not a historical photograph and does not depict a real meeting, committee member, author, employer relationship or NANOG logo.