NANOG’s Members Can Govern NANOG

Summary

NANOG members possess substantive rights over NANOG, Inc. Members in good standing elect six directors, vote on bylaws, can open amendment and recall processes through defined petitions, may run for office and can serve on committees. Partial participation does not erase those rights or nullify a properly constituted corporate vote.
The same governing record fixes the limit. NANOG is a facilitator, not a network operator; membership is optional and individual; and members, eligible voters, meeting registrants, attendees, mailing-list readers and North American operators are different populations. A NANOG ballot does not, by itself, bind third-party networks or turn its electorate into a proxy for the region.

An overwhelming answer to a narrow question

Ninety-six percent sounds like near unanimity. One hundred and twenty-five ballots sounds small. Five hundred and eighty-three eligible voters supplies the frame that both figures need. In NANOG’s separate 2020 special election on bylaws, the organization reported all three numbers: 583 eligible voters, 125 votes cast and 96 percent approval. Dividing ballots by eligible voters produces roughly 21.4 percent participation, an author calculation rather than a percentage published by NANOG. The 96 percent figure describes approval among the relevant votes.

It is not turnout, and it is not evidence that 96 percent of every eligible member—still less 96 percent of North American network operators—supported the amendment.

That contrast is tempting material for a legitimacy takedown. It should instead discipline the claim. A voluntary corporation does not ordinarily require every eligible member to cast a ballot before a valid electorate can act. Unless its rules say otherwise, abstention is not a veto. A member who does not vote remains uncounted, not automatically opposed. The amendment result can therefore be decisive inside the institution even when most eligible voters did not take part. What the participation denominator changes is the language available outside that setting. “Participating members approved the amendment” is supported.

“The networking community approved it” is not.

The official 2020 election record also reports a Board election with a different electorate: 553 eligible voters and 176 votes cast, approximately 31.8 percent by the same simple calculation. The two elections must remain separate. The record does not reconcile why one electorate contained 553 eligible voters and the other 583. Membership status can change, and eligibility rules or timing can differ. Whatever the explanation, it is not available in the published figures. Candidate percentages should not be reverse-engineered into turnout either; their tabulation denominator is a different question. One year offers a useful lesson in denominator discipline, not a longitudinal story about rising or declining participation.

This distinction—valid decision, bounded description—is the central institutional fact. It protects NANOG’s ability to govern itself through rules and ballots rather than ambient applause. It also protects everyone outside the electorate from being silently enrolled in a mandate they did not confer. The result becomes more credible, not less, when its scope is stated accurately.

The strongest case for a bounded electorate

A durable technical forum needs more than a conference hall, a mailing list and a shared professional identity. It must hold property, contract for services, manage staff, preserve records, organize meetings and make decisions when the crowd is not assembled. An identifiable legal body can do that work. An identifiable electorate can hold that body to written rules. Without those boundaries, “the community” can become a flattering name for whoever has the microphone, controls the budget or happens to be present when a decision is made.

NANOG’s current bylaws create such a structure. Members in good standing vote in elections, can run for the Board, may serve if elected and can serve on committees. Annual elections can concern Board candidates and proposed bylaw amendments; interim elections can address amendments or director removal. Members can initiate an amendment ballot through a petition signed by at least 30 members or 1 percent of the membership, whichever is greater. Recall uses the same petition threshold, followed by a two-thirds supermajority of participating voters. These are not ceremonial privileges. They identify who can start a formal process, what threshold opens it, who may decide and how demanding the final vote must be.

There is an important asymmetry in those rules. Access to the ballot is not passage. Thirty members—or 1 percent when that is larger—may put an amendment question before the electorate, but the proposal still needs the required majority member vote. A recall petition can start a removal question, yet the two-thirds requirement applies to participating voters, not to every person on the membership roll and certainly not to all operators affected by NANOG’s work. The architecture permits minority initiative without handing the petitioners the final word.

Open membership strengthens the internal case. NANOG describes membership as available to any interested individual in Internet operations, engineering or research who applies and pays dues. The membership policy sets out the practical programme and its privileges. This is preferable to a hidden or hereditary electorate. A person need not own an incumbent organization or occupy a reserved institutional seat merely to apply. Formal membership is a visible status rather than an inference drawn from reputation.

The caveat belongs beside the defence: openness is an eligibility rule, not an audited social outcome. It does not establish universal enrolment, equal affordability, equal practical access or a representative distribution by employer, geography, autonomous system, career stage or network type. It does not show that an applicant’s employer delegated a vote. Still, the absence of those broader proofs does not empty membership of meaning. It means the membership is what the rules say it is: a constituency of enrolled individuals with corporate rights, not a miniature parliament of the industry.

Good standing makes that electorate administrable. Under the bylaws, timely dues and the absence of suspension define a member in good standing. NANOG’s member-voting guidance directs users to confirm status and expiry through their profile and membership-management interface. That may sound mundane, but procedural clarity matters when ballots determine directors and governing documents. An electorate cannot be accountable if no one can establish who was entitled to act.

Good standing must not be made to carry a meaning it does not have. It is not a certification of technical expertise, meeting attendance, active participation, independence, demographic representativeness or merit. It is an institutional status. NANOG has one formal membership class, but one class on paper does not prove equal influence in practice. Some members may have more time, money, visibility, organizational support or familiarity with the process than others. The available record does not quantify those differences. Formal equality provides a common rule; it does not by itself establish equal voice.

The membership rules also provide exits and sanctions. Membership can end through resignation, expiry, nonpayment after notice or expulsion. Suspension and expulsion involve notice, an opportunity to be heard and a final Board decision. These provisions make the boundary enforceable rather than rhetorical. A 2018 amendment package shows NANOG publicly describing proposed changes across 34 sections, 19 of which it characterized as legally significant, including provisions on good standing, termination, suspension, expulsion, committees and elections. But a proposal summary is not the current constitution. Present powers must be read from the adopted bylaws. Nor can the existence of a procedure tell us how frequently discipline has been used or whether any particular case was fair; the published material supplies no basis for either inference.

This design deserves a defence on its own terms. An enrolled constituency can replace informal status with inspectable rights. Annual elections, petition routes, candidate requirements, published results and archives give members mechanisms with which to challenge or change the institution. A Board with a defined corporate mandate can preserve conferences, communications, finances and staff more reliably than an unincorporated crowd. None of this requires pretending that the electorate is universal.

Indeed, voluntary membership works precisely because a person can accept the rights and obligations of membership while another entity can remain outside them.

Members govern through a layered corporation

The internal mandate is real, but it is not direct democracy over every operating choice. NANOG’s Board has seven voting members: six elected directors and the Executive Director. The distinction matters. It would be inaccurate to say that every voting Board seat is directly elected by members. Members select six directors; the governing structure also makes the Executive Director a voting Board member. In turn, the Board appoints the Executive Director, selects officers and appoints committee members under the bylaws.

The Board manages and controls NANOG’s property, affairs and business and may delegate powers within that corporate framework.

This layered allocation answers a common objection from the opposite direction. Member voting is not cosmetic simply because directors and officers make many choices after an election. Representative corporate governance necessarily mediates authority. The electorate does not vote on every contract, programme slot, staffing decision, archive rule or expenditure. It chooses most of the voting directors and retains defined powers over bylaws and removal. The Board then exercises management authority over the organization’s affairs. Committees perform assigned functions. Officers and staff carry out operations.

A mandate can be substantive without being plebiscitary.

The Board’s authority to publish procedures implementing the membership programme illustrates the same division. Members receive rights from the governing structure, while the Board administers the conditions under which the programme operates. Neither side is unlimited. Administrative procedure cannot silently erase a bylaw right. A member vote cannot silently assume every power assigned to directors, officers or staff. Accountability depends on knowing not only who votes but which decision belongs at which level.

Candidate rules make the allocation more concrete. The 2020 Board-candidate materials required candidates to be members in good standing, to attend at least two of three annual meetings while serving and to disclose relevant affiliations. Such requirements can help voters assess capacity, participation and conflicts. They do not transform an affiliation into a corporate ballot cast by an employer. Disclosure says who a candidate is connected to; it does not prove that the employer directed the candidacy, authorized the individual to represent it or captured the resulting Board. It does not prove independence either. Those are separate empirical questions.

The Election Committee sits inside this structure too. NANOG’s current committee description says it assists the Board with director elections, amendments to governing documents and other election matters under Board direction. That role can be practically important—elections require administration—without becoming a separate constitutional source of authority. “Assists” and “under direction” do not describe an independent public electoral commission, and the committee cannot enlarge the constituency merely by running the process.

Member authority is best understood as a chain with identifiable links. An individual enters under an application-and-dues rule. Good standing establishes formal status. Eligible members can vote or use petition rights. The result authorizes a specified corporate decision. Directors, officers, committees and staff then act within powers allocated by the governing document. The chain is neither trivial nor infinite. Its legitimacy comes from remaining attached to the entity of the vote.

A constituency built in public, but not all at once

NANOG did not always possess this exact institutional form. Its official historical timeline says that NewNOG was incorporated on May 11, 2010 as a community-membership educational nonprofit, that a membership policy was adopted on January 4, 2011 and that the first membership elections occurred in October 2011. The chronology shows the construction of a corporate constituency: incorporation supplied a legal body, policy supplied membership terms and elections supplied a route for members to choose directors and act on governance.

Chronology, however, is not ratification evidence. An August 2011 announcement archive described a majority vote of eligible members and the route by which the Board or a petition meeting the 30-member-or-1-percent threshold could place a bylaw amendment before voters. It also announced that the first-year Board-created membership structure would be offered to the 2011 membership electorate for acceptance or rejection, with only NANOG members eligible to vote. That announcement was prospective. It proves that NANOG said a vote would occur and identified the intended electorate and choice. It does not, on its own, prove that the question appeared on a ballot, that voting was completed or that the structure was ratified.

That restraint matters because institutions often narrate origins too cleanly. An announced vote is not a completed vote; a first membership election is not automatically evidence of every contemporaneous ballot outcome. The membership’s authority today does not need an embellished origin story. Current powers rest on the current adopted bylaws. Historical notices explain how the constituency and amendment mechanism were described at the time, but they cannot certify an unobserved result.

A September 2012 election notice offers a later snapshot of the then-current arrangement: members could petition for bylaw amendments, candidates had to be in good standing and three of six elected seats were open under the two-year structure then in place. It is contemporaneous evidence that petition and candidacy mechanisms were presented to members as operating rights. It cannot be used to backdate today’s precise term or election rules. Constitutions change; historical notices belong to their own time.

The 2018 package further shows governance being revisited rather than treated as sacred inheritance. NANOG said 34 sections were involved and 19 contained legally significant changes. The breadth of that package does not prove that every proposal was wise or that each sentence was adopted unchanged. What it demonstrates safely is institutional attention to the machinery of standing, termination, discipline, committees and elections. Current text, not the explanatory summary of a proposal, answers present-tense questions.

The past-elections index places annual records from 2005 through 2024 on one public surface and says members shape NANOG through voting. The archive is a meaningful accountability mechanism: decisions leave records that can be inspected rather than surviving only as recollection. But an index does not provide consistent eligible-voter and ballot denominators for every year. It cannot support a trend line merely because the years are listed. The public member-minutes index likewise exposes minutes or decks for meetings from 2013 through 2025. Its existence supports a claim about an archive surface, not the content of every linked document.

Together these records show an institution making its internal authority legible through named members, status rules, elections, petitions, published results, meeting archives and amendable bylaws. That is a stronger basis for governance than an undefined appeal to “the community.” The very specificity that strengthens NANOG internally also gives an observer the means to resist inflating its decisions externally.

The denominator ladder

Technical communities accumulate audiences that overlap without becoming identical. A person can read the mailing list without attending a meeting. A registrant can fail to appear. An attendee can decline membership. A member can fall out of good standing. An eligible voter can abstain. A participating voter can approve one ballot and skip another. A candidate may disclose an employer without possessing any mandate to cast an organizational vote. Each step answers a different question.

A notice before NANOG 71 provides a vivid example. The September 2017 attendee archive reported 1,159 registered individuals, 316 registrants who said this would be their first NANOG meeting and 296 registrants identified as members. These were pre-meeting registration data. They are not an audited attendance count. The 296 figure is the overlap between registrants and people identified as members in that record; it is not NANOG’s full membership and not an electorate. It cannot be divided into the 2020 ballot counts, drawn from a different event and year, to manufacture a participation rate.

The notice’s description of membership as optional is institutionally important. It tells a prospective entity that appearing in the conference ecosystem and joining the corporation are different choices. That distinction allows the forum to be broad while keeping formal decision rights administrable. It also prevents the organization from claiming that everyone who registered thereby consented to member decisions. Registration records an intention before a meeting; it is neither proof of attendance nor a constitutional delegation.

The mailing list expands the outer circle further. NANOG’s usage guidelines describe an open list with a public archive and an audience above 10,000. They also say that contributions express individual opinions and that NANOG disclaims responsibility for those opinions. A large technical audience can create influence: operators learn, compare incidents, test explanations and circulate norms. Yet reach is not enrolment. The audience estimate is not an audited subscriber total, a member roll, an eligible-voter list or endorsement of any message. The disclaimer reinforces the point that a communication platform hosts voices without adopting each one as institutional speech.

“Community” often compresses all these rings into a single warm noun. The compression is useful in casual conversation and dangerous in a mandate claim. Pre-meeting registrants are not established attendees. Attendees are not necessarily members. Members are not necessarily in good standing. Members in good standing are not necessarily eligible in every ballot, and eligible voters are not necessarily participating voters. Mailing-list readers may sit outside every other category. North American operators may never enter a NANOG venue at all.

The 2020 numbers demonstrate why the distinctions are operational, not pedantic. There were 553 eligible voters and 176 ballots in the Board election. There were 583 eligible voters and 125 ballots in the separate bylaw election. The unresolved difference warns against moving casually among membership totals, eligibility and participation. The proper response is not suspicion masquerading as fact. It is to preserve the reported categories and admit the missing reconciliation.

The same caution applies to abstention. The record does not say why eligible people did not vote. They may have been satisfied, disengaged, busy, conflicted, unaware, indifferent between choices or prevented by circumstances not captured in the published material. Assigning a motive to the silent majority would turn absence into invented testimony. Abstention narrows what can be said about expressed preference; it does not automatically express opposition or consent.

What 96 percent can—and cannot—authorize

A bylaw is an internal governing instrument. When the required NANOG electorate approves an amendment under NANOG’s rules, the result can validly alter that instrument. The 96 percent figure communicates the balance among the counted votes on the proposition. The 125-ballot count communicates the reported number of participating voters. The 583 figure communicates the reported eligible electorate for that special election. All three are necessary for a candid account; none cancels the others.

The internal result should not be dismissed merely because approximately one eligible member in five cast a ballot. An organization that conditions validity on a defined voting process can receive a binding answer from those who participate. To argue otherwise without pointing to a violated election rule would impose a new condition after the fact. Limited participation may motivate outreach, easier voting, clearer information or member inquiry. It does not independently establish invalidity.

At the same time, the vote does not furnish the broader claim that “NANOG’s community” approved the amendment unless that phrase is explicitly limited to participating members. More than 10,000 mailing-list readers were not the electorate. The pre-meeting registrants in the earlier record were not the electorate. Conference attendees as such were not the electorate. Employees of network operators were not collectively the electorate. Eligible members who abstained did not cast approving ballots. Precision is not an attempt to diminish the outcome. It is the difference between reporting a vote and manufacturing a constituency.

The same logic governs the Board result. The 176 participating voters could choose directors under the election’s rules. Their ballots did not transfer ownership of their employers’ routers, addresses, contracts or customers to the resulting Board. An elected director receives an office in NANOG, not a public commission over Internet operations. The election establishes accountability for NANOG’s conferences, funds, policies and institutional actions. It cannot establish consent by networks that never delegated such power.

The phrase “members set the direction of NANOG” is defensible when “NANOG” means the corporation and its programmes under the bylaws. It becomes misleading when it slides toward “members set the direction of North American networking.” NANOG’s regional relevance may be substantial; its forum may shape ideas and professional practice. Influence, however, is not command. A technically persuasive norm spreads because operators adopt it, not because a corporate electorate legislated for their production systems.

Mandate inflation often occurs through an unmarked change of entity. The sentence begins with a legitimate subject—members—and a legitimate verb—elect, amend, guide—then swaps NANOG’s affairs for “the community,” “the ecosystem” or “the region.” Three questions expose the swap. Who authorized the action? Through which rule or instrument? Over what entity? If the answer is eligible NANOG members, acting through a bylaw election, over NANOG’s governing document, the mandate is clear. If the entity suddenly becomes third-party routing policy, customer service or continental representation, another source of authority is required.

The constitutional edge is written in NANOG’s own description

The most important boundary does not come from an external critic. NANOG’s bylaws say the organization is not itself a network operator but a facilitator of discussion, learning and technical communication among networking professionals. They describe a purpose of providing North American forums for education and knowledge sharing in the Internet-operations community. Those statements locate NANOG’s value: it creates places and channels in which practitioners can exchange knowledge. They also locate its limit: it does not operate the networks whose professionals participate.

“Not a network operator” is not a declaration of insignificance. Facilitation can change an industry. A forum can accelerate incident learning, help engineers discover common problems, introduce peers, build trust and make better operational practices easier to adopt. Its conferences, lists and committees can become important infrastructure for cooperation. But facilitation works through communication and voluntary uptake. It is categorically different from possessing the authority or technical control to configure third-party networks.

The regional mission has the same double meaning. “North American forums” describes whom and what the institution seeks to serve and where its convening role is situated. It does not create political representation of North America. Geography in a mission statement is not a franchise granted by every network, user, company, government or state in that geography. An organization can be central to a region’s professional life without becoming the authorized voice of all affected parties.

The Board’s power is expressly about NANOG’s property, affairs and business. A corporate Board needs that authority to operate. Yet the nouns mark the perimeter. Nothing in these governing records extends the Board’s control to a nonmember’s routes, infrastructure, engineering decisions, customers or public-policy positions. A member’s individual decision to join supplies no such extension. A person may bring experience from an employer or autonomous system, but experience is not authorization and affiliation is not title.

This distinction is especially important in Internet operations, where coordination and control are easily confused. Operators may converge on a practice after discussion. They may accept advice from respected peers. They may coordinate incident response through relationships nurtured at meetings. None of those outcomes requires NANOG to own networks or compel entities. Voluntary cooperation is often effective precisely because authority remains distributed. Calling that cooperation “governance” without specifying the governed entity risks converting influence into sovereignty after the fact.

A broader institutional statement might still credibly reflect operator views, but the burden would change. One would need to know whom the statement sought to represent, how those people or organizations were selected, what authorization they gave, how dissent was handled and whether the sample plausibly covered the claimed constituency. A NANOG membership vote alone answers none of those questions for all North American operators. It answers a different, valid question about NANOG’s internal decision.

Voluntary membership is a feature, not a continental franchise

The appeal of open membership is easy to understand. If any interested individual may apply and pay dues, people who want formal rights have a visible route into the institution. That arrangement carries real weight in assessing internal legitimacy. It weakens the danger that a closed circle can define “the community” while offering outsiders no route to participate. It makes the electorate inspectable and the obligations reciprocal: a member enters under known terms, acquires voting and petition rights, and remains subject to standing rules.

Voluntariness also preserves the forum’s wider openness. A conference entity need not buy a corporate vote merely to exchange technical knowledge. A mailing-list reader need not accept an institutional platform to read or contribute. People can value NANOG’s convening role while choosing not to join NANOG, Inc. That separation is not evidence of institutional failure. It permits a broad professional forum and a manageable corporate constituency to coexist.

But voluntary enrolment is affirmative. The person joins NANOG as an individual under NANOG’s terms. It is not a census of all relevant professionals. It is not an election of delegates by employers. It is not a mapping from autonomous systems to votes or a geographic apportionment. The available sources disclose no comprehensive annual reconciliation of membership totals, member demographics, employer concentration, ASN distribution or geographic composition. They do not show whether any member was authorized by an employer or network to vote on its behalf.

These unknowns should produce modesty, not insinuation. There is no evidentiary basis here for declaring the membership captured by a company, demographic group or technical faction. Candidate affiliation disclosures are useful precisely because relationships deserve visibility, but visibility is not proof of control. Nor does the lack of representativeness data prove a benign distribution. Both celebratory and accusatory claims would exceed what is known.

One formal membership class supports a narrow equality claim: the bylaws do not describe multiple classes with different formal franchises. It cannot support the larger claim that every member exerts equal practical influence. Informal networks, reputation, time, committee access and resources may matter. The available material does not measure them. Institutional analysis is strongest when it distinguishes formal design from social effect instead of pretending one determines the other.

The option not to join has analytical significance too. Accessible nonmember spaces cannot be used as evidence that their users silently consented to representation by members. Participation in a conference discussion is not acceptance of another person’s authority over unrelated assets or positions. Reading a mailing list is not the equivalent of signing a corporate ballot. Voluntary membership can offer a fair internal route without becoming a delegation from everyone who remains outside it.

The 2025 cycle confirms rights, not a universal constituency

Recent election communications show continuity in the membership boundary without supplying a complete numerical picture. An August 2025 notice required NANOG membership to submit nominations and said the cycle continued ranked-choice voting. A September notice required logged-in member credentials for statements of support. Those controls connect formal participation in candidate selection to the enrolled constituency.

They do not turn a support statement into an employer endorsement, and the credential condition alone does not answer every question about the supporter’s membership status. The notices also caution against casual comparison with 2020 candidate percentages. Ranked-choice counting and older plurality-style figures do not share an interchangeable tabulation method. Without compatible denominators and transfer data, a trend assembled from headline percentages would be fiction dressed as arithmetic.

NANOG’s November 2025 result announcement named Leslie Daigle, John van Oppen and Michael Costello as the winners. That outcome can be reported. The notice does not provide the eligible electorate, turnout, ballot count or ranked-choice transfer rounds. None should be invented. Transparency is not all-or-nothing: the public knows who won and can inspect the announcement, while the numerical participation picture remains incomplete.

That incomplete picture does not invalidate the result. It limits secondary analysis. An observer can say that the member-gated nomination and support process, ranked-choice method and announced winners belong to the 2025 corporate election. The observer cannot responsibly calculate participation, compare the winners’ support with the separate 2020 figures or make claims about the wider operator population. Restraint preserves the value of what is known.

Accountability without mythology

The healthiest defence of NANOG membership does not require portraying it as the entire community. Its virtue is nearly the opposite: it turns an otherwise diffuse audience into a bounded group capable of making specified decisions. It gives members rules for standing, voting, candidacy, amendments and recall. It gives the corporation a Board capable of managing property and affairs. It publishes election outcomes and maintains archive surfaces. This is institutional accountability in practical form.

The structure also creates questions worth asking from within its own terms. Do members know when ballots open? Are voting tools usable? Do petition thresholds remain reachable as membership changes? Are candidate affiliations informative? Do member archives make decisions understandable? Do the Board’s implementing procedures respect bylaw rights? How many eligible members participate, and why do others abstain? These are governance questions, not attacks on the existence of governance.

The published record cannot answer several of them. It does not disclose a complete annual membership and good-standing reconciliation. It does not explain the different 2020 electorates. It does not give reasons for abstention. It does not say how frequently amendment petitions, recall, suspension, expulsion or other termination routes have been used. It does not show whether nonmembers influence choices through consultation. It does not establish causal effects of elections on routing, reliability or public policy. These absences are limits on analysis, not evidence that events never happened.

Nor does the record establish the demographic, employer, operator, ASN or geographic composition of members; the precise practical barriers to participation beyond the formal rule; or whether any institutional statement was separately authorized by a representative sample of operators. The complete 2025 electorate, ballot count and ranked-choice transfers are also unavailable in the result notice. A mandate claim should not convert any of these blank spaces into a convenient fact.

Accountability also depends on separating institutional voice from hosted speech. The mailing list’s individual-opinion rule means that a widely read post is not automatically NANOG’s position. A conference talk is not automatically a member resolution. A committee’s work is not automatically a mandate from all attendees. A Board statement is an institutional act under whatever corporate authority applies, but it is not automatically a statistically or politically representative view of every operator. Each form of speech has a speaker and a basis. Trust grows when the labels remain attached.

A practical test for any claim of mandate

Four questions can prevent both cynicism and inflation when NANOG acts.

First, what is the decision? Electing a director, changing bylaws, recalling a director, appointing a committee, setting a conference programme and recommending an operational practice are not interchangeable acts. The bylaws allocate them to different actors. Naming the decision stops the word “governance” from concealing a change in authority.

Second, who is the constituency for that decision? Members in good standing may be the relevant class, but a particular ballot may have an eligibility date and a participating subset. Board members may appoint an officer. A committee may make a delegated programme choice. Mailing-list entities may offer individual technical opinions. “NANOG community” is not a substitute for this accounting.

Third, what instrument carries the authority? A bylaw, election rule, Board delegation, contract or voluntary technical agreement can each support a different claim. Reputation and trust may cause advice to be followed, but they do not silently create legal control. If the instrument is a NANOG membership ballot, the result attaches to the corporate subject assigned to that ballot.

Fourth, where does the effect stop? NANOG can govern its membership status, governing documents, offices, committees, property and programmes under its rules. It can advocate, educate and facilitate. It can influence entities through evidence and persuasion. But without another instrument it cannot bind a third party’s routers, routes, assets, customers, staffing, engineering choices or policy positions. The fact that an external operator may be affected by an industry discussion does not enrol that operator in the electorate.

Applied to 2020, the test yields a clear result. The decision was a bylaw amendment. The reported electorate consisted of 583 eligible voters, of whom 125 cast ballots. The instrument was NANOG’s internal election process. The approval—96 percent among the relevant votes—authorized the amendment inside NANOG. The effect stopped at the corporation’s governing document. This conclusion respects both voters and nonvoters, both the institution and outsiders.

Applied to a claim about “North American operators,” the test exposes missing links. The sources do not map operators to members, members to employer authorization or ballots to a representative sample of networks. They do not identify a continental delegation, public-law franchise or contract binding absent parties. The regional forum mission establishes service and convening ambition, not political representation. An institutional position may be important and technically persuasive; importance is not a substitute for authorization.

The edge is the source of legitimacy

There is a habit in institutional debate of treating boundaries as an embarrassment. A body that admits it speaks only for its members can appear smaller than one that invokes a whole community. In practice, the bounded claim is more credible. It tells readers how authority was obtained, what decision was made and who remains free to disagree. The alternative—stretching members into attendees, readers and operators—may sound inclusive while erasing the people it claims to include.

NANOG’s structure contains a productive duality. The forum can be open and influential; the corporation can be bounded and governable. Nonmembers can learn, speak and cooperate without becoming voters. Members can elect directors and amend bylaws without acquiring control over nonmember systems. The Board can manage NANOG’s affairs without operating the Internet. These are not defects to be reconciled. They are distinctions that allow voluntary coordination to work without pretending to sovereignty.

The 2020 bylaw result captures that duality in three figures. The 96 percent approval is strong evidence of agreement among the votes counted. The 125 ballots identify the participating body. The 583 eligible voters reveal the wider internal electorate and keep the approval statistic from floating free of its base. None of the numbers voids the vote. None authorizes an extra claim about people who were not part of it.

The Board election in the same year reinforces, rather than broadens, the lesson. Its 553 eligible voters and 176 ballots belong to a separate decision. Its participation rate cannot reconcile the bylaw electorate, explain abstention or establish a trend. It does show that even within one institution and year, the phrase “the members voted” can conceal multiple populations and subjects. Careful governance reporting restores the denominators.

An institution earns legitimacy not by claiming the largest possible constituency but by honoring the one it actually has. NANOG members have a meaningful mandate: they elect six voting directors, vote on bylaws, can pursue amendments and recalls, run for office and serve on committees. Their choices help sustain a legal organization that provides valued forums for technical exchange. Those powers deserve to be taken seriously.

They also deserve to be kept in their lane. Membership is optional and individual. Good standing is a corporate status. Meeting registration, attendance, list readership, employment, affiliation and operation of a network do not automatically create or transfer a vote. NANOG’s own constitution identifies it as facilitator rather than operator. The mandate therefore ends where NANOG’s property, affairs, governing documents and delegated functions end—unless some other clearly identified party grants some other clearly identified authority.

That is not a diminished conclusion. It is the basis on which a voluntary association can speak precisely, govern confidently and cooperate widely. The 2020 electorate did not have to be North America to authorize NANOG’s bylaw. It only had to be the electorate that NANOG’s rules made responsible for that decision. The honest sentence is powerful enough: participating members approved their institution’s amendment. Everything beyond it requires a new mandate.

SEO and social

SEO title: NANOG Members’ Mandate Ends at NANOG’s Corporate Edge

SEO description: What NANOG members can authorize inside NANOG, Inc., why partial participation does not void a corporate vote, and why membership cannot stand in for every North American operator.

Social title: What Did 125 NANOG Ballots Actually Authorize?

Social description: NANOG’s 2020 bylaw vote was valid and bounded. Its denominators show how a voluntary association can have a real internal mandate without claiming authority over outside networks.

Image metadata

Alt text: Editorial illustration of a bounded NANOG member ballot box inside a meeting forum, with separate outer rings representing registrants, attendees, mailing-list readers and independent network operators.

Caption: NANOG’s electorate can authorize decisions within the corporation; broader technical participation does not confer control over third-party networks.

Accessibility description: A conceptual editorial scene places a ballot box and voting members at the center of a conference-style room. Distinct labelled rings separate eligible members from registrants, attendees, online readers and external network operators, emphasizing that the groups can overlap without being identical. The composition contains no readable ballot choices, presents no logo as an endorsement and does not suggest that outside networks are controlled by the central vote.

Synthetic image provenance: This accompanying editorial image is AI-generated as a conceptual illustration. It does not depict a real NANOG election, a specific meeting, an actual ballot, identifiable voters or documentary evidence of turnout.