Summary

  • NANOG’s 2010–2011 transition gave a named corporation custody of the association’s trademark, domain, archives, contracts, finances and operating systems, replacing an arrangement in which Merit had carried much of that institutional responsibility.
  • The votes that ratified the transition authorized a defined internal constituency to govern the association. Neither those ballots, later membership elections nor the open mailing list established a mandate over attendees’ employers, autonomous networks, routing choices or North America as a political community.
  • Incorporation delivered observable benefits: assets could attach to a legal person, service providers could change without dissolving the institution, financial accounts could be audited, and recurring elections could be recorded. Those benefits do not erase early transparency concerns, missing instruments or the continuing difference between custody of an archive and proof that every item survived.

Five objects at the boundary

On 1 February 2011, Merit Network and NewNOG, Inc. signed an agreement that placed a small but consequential set of institutional assets on the other side of a legal boundary. Merit’s announcement named the NANOG trademark, the meeting archives and the nanog.org domain. A partner announcement also referred to the list and website contents. The agreement took effect on 7 February. After years in which NANOG had operated as an activity hosted by Merit, a separate corporation could now be identified as the principal custodian of the public name, the digital address and the accumulated record of the forum.

The transfer is easy to overstate because NANOG’s public identity is larger than the list of things that changed hands. Network operators met under its name. Technical discussions could shape professional judgment. People took knowledge back to their employers. Yet the surviving governing documents describe a transaction in institutional property and responsibility. They do not describe a sale or delegation of routers, fibre, address space, routing tables, interconnection decisions or employer authority. NewNOG acquired the means to keep the association running; it did not acquire the networks of the people who gathered there.

That difference is the heart of NANOG’s legal transition. A forum can be influential without being sovereign. An association can govern its meetings, membership, committees and assets without governing the industry whose practitioners attend. A corporation can speak for itself without speaking for a continent.

The narrowness of the transfer does not make it trivial. A trademark needs an owner. A domain needs a responsible holder. Conference venues impose contractual and financial exposure. Archives require custody. Staff and contractors need supervision. Revenue must be received, spent and reported through some accountable principal. Before 2011, Merit supplied much of that institutional shell. Afterward, NewNOG did. The important question is therefore not whether NANOG became powerful in the abstract, but which responsibilities became attachable to a named legal person, which people authorized that change, and where that authorization stopped.

Even the ceremony’s location in institutional memory carries a warning about the record. Merit’s contemporaneous announcement says the agreement was signed at NANOG 51 in Miami. NANOG’s later archived history calls the occasion NANOG 52. A May 2011 Board update identifies the next Denver meeting as NANOG 52, supporting Merit’s numbering. The conflict should remain visible rather than being quietly repaired. It is a small discrepancy, but it demonstrates why legal custody of an archive and perfect archival coherence are not the same claim.

From Regional-Techs to a hosted activity

NANOG’s official history traces its lineage to NSFNET Regional-Techs meetings beginning in 1987. It says the NANOG name followed a broader charter in February 1994. That history establishes an institutional chronology: a technical forum existed before there was an independent NANOG corporation. The sequence matters because it separates three things that are often collapsed in retrospect—a community of practice, an activity administered by a host, and a legal entity able to own property in its own name.

By 2005, an archived charter described NANOG explicitly as an activity hosted at Merit. The page is labelled a May draft for comment, while NANOG’s official history says an amended charter was adopted and the first elections were held in October. The arrangement it records nevertheless gives a detailed picture of the hybrid structure.

Merit supplied staff, meeting logistics, registration, financial management and mailing-list hosting. It consulted on meeting locations and administered the machinery that made the gathering recur. Leadership later said Merit also carried venue-contract liability, maintained servers and archives, funded activities and kept personnel in place.

This was not a simple hierarchy in which Merit decided everything and the operator community decided nothing. Under the charter, people who had recently registered for meetings could qualify to vote. An elected Steering Committee held defined responsibilities, and programme and list functions included elected participation. At the same time, a Merit appointee shared specified powers and Merit retained central operational roles. The structure distributed authority: elected committees had meaningful control over content and moderation, while the host carried many of the legal, financial and logistical obligations.

That division worked well enough to support a recognised forum for years. It also created a dependency. If the host selected venues, signed obligations, employed staff, ran financial administration and held the institutional assets, continuity depended on the host relationship as well as on the willingness of volunteers and attendees to keep participating. The technical community could exercise influence over programme content while lacking an independent legal hand with which to sign the venue contract. Its public identity could feel communal even though important property and exposure sat inside another institution.

The distinction is not an accusation against Merit. Hosting is a common way for an activity to begin and mature. Merit itself was a nonprofit host, and an existing organisation could pool staff, systems and financial capacity for a forum that did not yet need—or did not yet want—its own corporation. The later transition leaders praised what Merit had provided. The serious question was whether an activity of NANOG’s size and longevity still wanted its institutional continuity to depend on the same arrangement.

What a single meeting put at risk

The operating exposure became visible in the figures for NANOG 49, discussed during the 2010 transition period. The meeting drew 607 attendees: 505 paid, 102 had fees waived, 199 were newcomers, and participants came from 26 countries. Those figures describe one event, not the entire region and not a long-run attendance trend. They do, however, show the scale of the undertaking that had to be booked, staffed, financed and administered.

The same meeting recorded revenue of $409,061 and expenses of $423,340. The arithmetic is direct: $409,061 minus $423,340 equals a deficit of $14,279. That result does not prove why NANOG incorporated, nor does it represent annual corporate performance. It shows something more concrete and more modest: a large technical meeting created real financial exposure, and some legal person had to carry it. Under the hosted arrangement, that person was Merit.

Participant notes from NANOG 50 also record questions about who would have authority to sign as Executive Director, how financial controls would work and what paid membership would mean. The notes identify themselves as unofficial and acknowledge possible errors, so they cannot substitute for executed legal instruments. Their value is different. They show that formalisation did not magically remove governance concerns. It made certain concerns more explicit by forcing participants to ask who could bind the organisation, who would oversee money and what constituency would elect its leaders.

The cost and attendance figures also discipline claims about representation. Six hundred and seven attendees at a meeting can demonstrate logistical scale. They cannot establish that the meeting represented every network in North America. Twenty-six countries can demonstrate geographic reach. They cannot turn an operators’ gathering into an interstate body. Waived registrations, paid registrations and first-time attendance are event categories; none is identical to membership, electoral eligibility or a delegation from an employer.

This is why the legal container should be judged first against the responsibilities it was built to hold. Could it sign? Could it receive revenue? Could it bear a deficit? Could it maintain staff and archives? Could it survive the replacement of one service provider with another? These are demanding tests even though they are narrower than authority over networks. NANOG’s transition becomes more intelligible when viewed as an answer to them.

Separation as continuity, not conquest

The best independent account of the separation debate does not present it as a campaign for regional power. Ashwin Jacob Mathew reports that practical dissatisfaction helped drive the discussion: venue scheduling, the replacement of familiar staff and concerns about transparency. He describes an initial period in which the Steering Committee and an advisory group considered separation, followed by questions on the nanog-futures list and a formal consultation at NANOG 49.

That account complicates any ceremonial story in which incorporation was simply the inevitable next stage of maturity. If early deliberation generated transparency criticism, then member control was not born fully legible. People could support independence and still question how the decision was being developed. Moving authority out of a host did not itself guarantee that every internal choice would be open, well explained or widely understood.

At the same time, Mathew’s account supports the practical case for independence. A forum whose recurring events depended on staffing, venue choices and host decisions had reasons to seek a principal that existed specifically for the forum. The transition could be motivated by a desire for continuity and control over operating conditions rather than by a desire to claim broader public authority. Indeed, the latter claim would sit uneasily with the narrow assets and powers recorded in the legal documents.

Transition leaders made a similar case in April 2010, although their announcement was advocacy by interested proponents and should be read as such. They said Merit had entered contractual obligations on NANOG’s behalf, carried venue liability, supplied staff and servers, maintained funds and archives, and supported operations. Their proposed answer was a nonprofit that would take responsibility gradually. Those statements are partly corroborated by the older charter, but their praise of the transition’s expected effects is not independent proof that every benefit followed.

Jesse H. Sowell offers a useful counterpoint at a wider analytical level. His work argues that privately motivated, bottom-up arrangements can create public goods. Openness and transparency remain separate questions; neither automatically follows from private initiative. Applied carefully here, that proposition explains how a membership corporation could support a broadly valuable technical forum without becoming a public authority. NANOG could make conferences, archives and an open list more durable. The benefit could extend beyond those who cast ballots, while the corporation’s mandate remained internal.

This is the strongest explanation of the transition, not a concession to be mentioned and discarded. Legal form can be useful precisely because it narrows responsibility. A named corporation can be sued, audited, contracted with and held to its own rules. It can retain an archive and employ or supervise the people who administer a meeting. The fact that it does not govern routers is a boundary that makes its real duties clearer, not evidence that it has no public value.

What Delaware created—and what it did not

The inspected certificate of incorporation names NewNOG, Inc. as a Delaware non-stock nonprofit corporation. It states charitable and educational purposes and provides for no capital stock. This instrument created a legal person designed to conduct exempt-purpose activity. It did not, by itself, establish federal tax exemption, and the copy examined does not show a visible certified filing stamp.

NANOG’s official history places incorporation on 11 May 2010. The same history compresses the legal sequence by describing NewNOG as incorporated “as a 501(c)(3).” The certificate and later records require more care. State formation and federal tax recognition were separate events. A May 2011 Board update said the Internal Revenue Service had recognised NewNOG, while IRS-derived data place federal exemption in April 2011. The determination letter itself was not examined. The defensible chronology is therefore a Delaware corporation formed in 2010, followed by reported federal recognition in 2011.

That separation is more than a lawyer’s technicality. NewNOG could exist as a corporation before the federal tax decision. If exemption had been denied, the corporation would not simply have vanished; its tax position and its legal personality were distinct. Conversely, charitable status did not expand its mandate over third-party networks. It described the federal treatment of the corporation and its purposes, not a grant of regulatory jurisdiction.

The certificate helped solve a specific ownership problem. An activity can have a familiar name and stable customs without being able to hold assets apart from its host. A corporation can own property, enter obligations and govern its internal affairs through officers and directors. The change made it possible to attach the NANOG assets and liabilities to an organisation whose stated purpose was NANOG’s educational work.

It did not turn every person who used the name NANOG into a corporate member. It did not make list subscribers shareholders; there was no capital stock. It did not assign attendee networks to the corporation. It did not bind employers to Board decisions. Nothing in the examined governing record delegates control of routing, interconnection or address resources. The legal powers were meaningful because they concerned NewNOG’s own property, affairs and business.

Ratification and the missing denominator

The transition was not completed by filing alone. In October 2010, voters considered both a charter measure authorising the transition and NewNOG bylaws. The charter measure passed by 210 votes to 16. The margin was 194 votes, calculated as 210 minus 16. Among the 226 ballots cast on that measure, the affirmative share was 92.9 percent, calculated as 210 divided by 226.

The bylaws measure passed by 169 votes to 26. Among 195 ballots cast, the affirmative share was 86.7 percent, calculated as 169 divided by 195. Those are strong majorities among the people who voted. They document ratification by the defined electorate. They do not disclose turnout because the eligible-voter denominator was not published in the inspected result.

That missing denominator sets a precise limit. It would be wrong to describe the vote as universal consent by operators, but it would also be wrong to dismiss the result simply because the total eligible population is unavailable. The record establishes what the ballots established: large majorities among votes cast supported the transition charter and bylaws. It does not establish how many eligible people abstained, why they abstained or whether the electorate resembled every network active in North America.

The scope of the question also mattered. Voters were deciding how NANOG would govern itself and whether to move its activity into NewNOG. They were not voting on routing policies for each employer. An engineer’s ballot could authorise a charter without binding the engineer’s autonomous system. Even a unanimous association vote would remain a vote over the association unless a separate delegation extended it.

This is where the word “community” can blur categories. Conference attendees, recent registrants, mailing-list readers, individual members, eligible voters and actual voters may overlap, but they are not interchangeable. Nor are individuals necessarily voting as mandated representatives of their employers. The 2010 result is significant because it gave the transfer an internal electoral basis. Its legitimacy should be judged within that scope rather than inflated into a regional plebiscite or reduced to a meaningless private tally.

The handoff was an event and a period

The February agreement supplied a clear legal milestone. It named parties, a signature date, an effective date and several important assets. Yet independence did not arrive as a single instant in which every operational function changed hands. Transition leaders had described a gradual handoff. The May 2011 Board update said registration, membership, mailing-list and finance systems would be migrated. The corporation had principal responsibility, but the work of moving systems continued.

NewNOG contracted Association Management Solutions to provide association and meeting support. The executed service contract was not examined, so its full terms, safeguards and allocation of duties remain unknown. What the Board’s report does show is institutional portability: the corporation could choose a service provider to execute important tasks while retaining its own identity and responsibility. Independence did not require every registration screen, ledger entry or meeting task to be performed directly by directors.

This division between principal and provider is one of the transition’s most concrete benefits. Under the older arrangement, the host and the activity were tightly joined in operations. Under the corporate arrangement, NANOG could remain NANOG while contracting for support. A provider could in principle change without requiring the public identity, archives and membership corporation to move into a new host’s legal personality.

The May update also reported that all NANOG intellectual property originating with Merit had transferred. That is a strong contemporaneous Board claim, consistent with the named trademark, domain and archive assets in other announcements. It is not the same as an itemised inventory. The full agreement was not examined, so its warranties, exclusions, liabilities and preservation duties are unknown. No checksum or item-by-item archive manifest establishes that every web page, post or recording was present and intact at handoff.

Mathew reports another practical bridge: an initial $250,000 loan from ARIN, repaid by NANOG 56. The amount and repayment should remain attributed to his independent account because the executed loan instrument and repayment documentation were not examined. Participant notes from 2010 record questions about financing, but they do not provide a complete contract. The reported loan illustrates that corporate independence still depended on relationships and transitional support. A separate legal person did not mean financial self-sufficiency on day one.

The counterfactual is useful. Merit might have continued hosting NANOG under a revised agreement, preserving meetings and archives. Another established nonprofit might have received the assets. Either path could have changed custody without creating a membership corporation. Incorporation was therefore not the only imaginable route to continuity. Its distinctive contribution was to combine independent custody with an internal electoral structure and a principal dedicated to NANOG’s own affairs.

Attendance ceased to be the ballot box

The move to NewNOG altered the constituency mechanism. Under the 2005 charter, voting eligibility was linked to recent meeting registration. After the transition, individual membership carried the rights to vote, run for office and serve, while membership was not required to attend a conference. An ARIN transition announcement made that separation explicit.

The difference changed incentives. A person could attend for technical exchange without joining the corporation. A person who wanted formal electoral rights had to become a member and remain in good standing. Attendance no longer automatically served as the gateway to governance. This created a clearer association boundary, but it also meant that the population benefiting from NANOG events could be larger and different from the population choosing directors.

The first member-election record in October 2011 lists four candidate vote totals: 84, 89, 74 and 86. It also records two bylaw results, one passing 79–0 and another 78–8. These counts show an operating member electorate and an association willing to publish results. As with the 2010 votes, the eligible-voter denominator is unavailable, so turnout cannot be calculated. The numbers cannot tell us how many members declined to vote or how representative the members were of the wider operator profession.

Nor should conference access be confused with list access. The current mailing-list policy describes a public, archived, technical forum open to all. The historical bylaws also distinguished the broad forum from corporate membership. The list’s present audience claim and public availability demonstrate continuing reach, but current policy cannot prove that every historical subscriber or every archived item survived the 2011 migration.

The association thus has several concentric but non-equivalent constituencies. The open list can include nonmembers. Conferences can admit nonmembers. Members can have corporate rights without carrying formal mandates from employers. Voters are the subset of eligible members who cast ballots. Directors receive authority over corporate affairs from that internal electorate. None of these facts makes the outer circles unimportant. They simply prevent influence, access and legal authority from being treated as the same thing.

Imagine that every attendee had automatically become a member. The electorate would have been broader, but it still would not necessarily have represented every network in North America. Some networks would send several people, some one, some none. Individuals might vote from personal judgment rather than employer instruction. The representational limit is not an accident of paid membership alone; it follows from the difference between an association of people and a legislature of networks.

A public name, a DBA and a corporation

The name sequence reveals how institutional identity can sit on several layers. The 2010 certificate created NewNOG, Inc. The 2013 bylaws’ revision history says “NewNOG” was replaced by “NANOG” in November 2011 under a doing-business-as designation, while the legal name remained NewNOG, Inc. The association could therefore present the familiar NANOG name publicly without pretending the original corporate name had already changed.

In 2019, NANOG’s annual report said the organisation officially changed its name from NewNOG, Inc. to NANOG, Inc. Independently audited financial statements for that year use NANOG, Inc. and describe it as a Delaware nonprofit corporation formed in 2010. Together, those records corroborate the resulting corporate identity. They do not reproduce the Delaware amendment instrument or establish its exact filing date. That remains unknown.

This progression—NewNOG corporation, NANOG DBA, NANOG corporation—shows why a brand should not be mistaken for a legal person. “NANOG” existed as a forum name long before NewNOG was incorporated. For a time, it was the public name under which NewNOG did business. Later it became the formal corporate name. The audience could experience continuity while the legal layer evolved underneath.

The continuity was useful. It meant that a change in corporate mechanics did not require the professional community to abandon a recognised identity. But the familiar name could also encourage overreading. “North American Network Operators’ Group” sounds geographically comprehensive. The governing documents do not convert that breadth of aspiration into authority over all North American operators. The corporation’s name identifies its field and forum; it does not define a political constituency.

The 2013 bylaws say this directly in institutional terms: NANOG is not itself a network operator. They define an educational forum mission and vest control of property, affairs and business in the Board. That is substantial authority. The Board can govern what belongs to the corporation. The same clause also supplies a boundary. Corporate control over NANOG’s business is not operational control over a member’s or attendee’s network.

Formality has to be maintained

Creating a corporation answered the custody question at one point in time; it did not eliminate the need for later governance work. The 2018 bylaw consultation is evidence of that maintenance. NANOG published a legally reviewed package of proposed changes addressing member discipline, Board removal, committees, staff and election procedures. The proposal page demonstrates legal review and member consultation. It does not prove that every proposed provision was adopted or implemented.

That distinction matters because institutions are not made accountable merely by possessing bylaws. Rules have to be current, interpreted and applied. Consultation can expose a question without resolving it. Legal review can improve clarity without proving fair outcomes. The 2018 record is best understood as evidence that the corporation’s governing framework continued to receive attention, not as a final verdict on all later practice.

The later financial record offers a more measurable view of corporate continuity. For 2019, independently audited statements recorded revenue of $3,854,462 and expenses of $3,839,110. Revenue exceeded expenses by $15,352, calculated as $3,854,462 minus $3,839,110. Total assets were $5,754,246 and net assets were $4,615,113. The statements also describe three major meetings, one-day events, the website, list, archives and training.

Those numbers show that by 2019 NANOG was not a paper corporation waiting passively behind a familiar brand. It was operating at a scale far above the finances of a single 2010 meeting, maintaining programmes and subjecting annual accounts to independent audit. The audit was commissioned for NANOG and provides assurance about the financial statements for that period. It does not establish representative authority, historical archive completeness or the quality of every organisational decision.

The 2024 record supplies a later comparison and a rare published electoral denominator. NANOG reported 698 eligible voters and 165 votes cast. The participation ratio was 23.6 percent, calculated as 165 divided by 698 and rounded to one decimal place. This result describes one election. It cannot, without further evidence, prove apathy, capture, exclusion or illegitimacy. A turnout figure identifies the breadth of a particular electoral mandate; it does not explain the reasons behind participation.

IRS-derived data for 2024 report revenue of $3,072,544 and expenses of $3,425,606. The deficit was $353,062, calculated as $3,072,544 minus $3,425,606. The same record lists assets of $3,597,867, liabilities of $1,058,265 and net assets of $2,539,602. These figures describe a later financial condition and a continuing public accountability trail. They do not establish that the 2010–2011 transition caused the result, and they should not be stretched into an assessment of reserve adequacy.

Taken together, recurring elections, an open archived list, published operational records and audited financial statements are counterevidence to the claim that legal form merely enlarged bureaucracy. They show durable activity attached to the corporation. They still do not show that incorporation changed routing performance, security outcomes or interconnection. Continuity of an institution and outcomes on independently operated networks are different evidentiary categories.

The archive is an asset, not an alibi

Archives occupy a special place in this story because they are both institutional property and the evidence through which institutional memory is judged. The transfer announcements named meeting archives, list material and website contents. The Board later said the Merit-originated intellectual property had moved. Current policies continue to describe an archived technical list. These facts support continuity of custody.

They do not prove item-level completeness. The full transfer agreement might contain warranties or preservation duties, but it was not examined. No inventory establishes that every historical file crossed the boundary. The NANOG 51/NANOG 52 discrepancy demonstrates that an archive can preserve conflicting descriptions. A transferred collection can be valuable, actively maintained and still incomplete or internally inconsistent.

That is not a reason to discount the transfer. A named custodian is a precondition for durable care. When responsibility is legible, people know which organisation can maintain the domain, preserve meeting materials and answer for the collection. The legal person creates a target for accountability even if the record cannot prove that every item survived.

But custody should never become an alibi for claims the archive cannot support. A video collection does not by itself establish attendance. A list archive does not show that all affected operators consented to a policy. A transfer announcement is not the complete agreement. A later institutional history is not automatically more reliable than a contemporaneous release. The archive helps define what can be known; it also reveals the boundary of that knowledge.

This disciplined use of the record strengthens rather than weakens the case for legal continuity. The honest institutional claim is that NewNOG became the named custodian of identified assets and that NANOG continued to maintain public information systems. The extravagant claim—that every item was complete, every stakeholder represented and every later benefit caused by incorporation—is unnecessary and unsupported.

Useful authority, bounded authority

NANOG’s transition worked because the corporation did not need sovereignty to be useful. It needed title, signature authority, accounts, officers, committees and an electorate for its own affairs. It needed to be able to select service support, carry conference exposure and retain the institutional assets when personnel or providers changed. These are forms of authority with observable objects.

The distinction between legal authority and professional influence remains essential. A respected NANOG discussion can influence engineers. A presentation may change how an attendee understands a risk. Peers may treat committee members or directors as trusted conveners. None of that informal influence is created by the Delaware certificate, and none is the same as legal power to compel a network’s decision.

No examined governing record establishes a separate delegation allowing NANOG to represent particular employers or autonomous networks. That is an absence in the available record, not proof that no separate instrument could ever exist. The proper conclusion is limited: the documents establishing and governing the corporation concern its own mission, members, property, affairs and business. They do not supply a wider mandate.

The same restraint applies to public policy. A nonprofit can contribute expertise, host debate and develop collective understanding without being a regulator. Its members can participate in policy discussions without converting their association into a government. A regional name can describe where a professional forum concentrates its attention; it does not create borders, jurisdiction or popular sovereignty.

This boundary protects both the association and the networks. NANOG can be evaluated against duties it can actually perform: maintaining the forum, controlling its assets, keeping accounts, holding elections and supervising operations. Networks remain responsible for their own engineering and commercial choices. If a network adopts an idea encountered at NANOG, causation must be shown through evidence of that decision, not inferred from conference attendance or corporate form.

The 2011 transition therefore did not answer every question of openness, transparency or representation. Early critics were right to ask how separation was discussed and how executive and financial powers would be controlled. Later electoral and financial publication makes aspects of responsibility more visible, but formal visibility does not prove that every decision was correct. The corporation created a structure within which accountability could be demanded. It did not guarantee the answer.

What the alternatives reveal

Several plausible alternatives clarify what incorporation actually contributed. Merit might have remained host under new terms. If so, conferences and archives might have continued, but principal custody would still have rested within the host arrangement unless title changed separately. Another existing nonprofit might have received the trademark, domain and archives. That could have improved portability without creating member elections.

Assets could also have moved to an independent custodian while governance remained appointed rather than electoral. That would separate continuity from member accountability. Conversely, NewNOG could have incorporated and failed to receive federal exemption; legal personality would have existed even though tax treatment differed. Each alternative removes one link from the historical combination and shows that custody, elections, hosting and tax status are distinct achievements.

The actual transition combined them over time. A Delaware corporation was formed in 2010. Defined voters ratified the charter and bylaws. Merit and NewNOG signed the asset agreement in February 2011. Federal recognition was reported later that spring. Operating systems continued to migrate. A first member election followed in October. The NANOG public identity was used under a DBA before the corporation eventually adopted NANOG, Inc. as its legal name.

This was not a clean leap from informality to completeness. It was a sequence of legal, electoral and operational acts, each with its own evidence and limits. The certificate did not transfer the assets. The transfer did not prove archive completeness. IRS recognition did not create the corporation. Membership did not equal attendance. Election results did not establish regional representation. A service contract did not make the provider the institutional owner.

The strength of the arrangement lies partly in keeping those acts separate. A legal person can change providers without changing its identity. Members can elect directors without controlling nonmembers’ networks. The open list can serve people who never join the corporation. The association can preserve records of a professional field without claiming to own the field.

The corporation behind the community

By 2011, NANOG had acquired something both narrower and more durable than authority over the Internet: an institutional home it could call its own. The trademark, domain and named archives no longer depended on being held inside Merit. Contracts and financial exposure could attach to NewNOG. Registration, membership, list and finance systems could be migrated under a principal whose existence was dedicated to the forum. Elections could fill offices defined by corporate rules.

The transition’s legitimacy is strongest when described at that scale. The 2010 votes show substantial approval among ballots cast, while the missing denominator prevents a turnout claim. The 2011 election shows member governance in operation, while the absent eligible total prevents another. The later 23.6 percent participation figure precisely describes one election and nothing more. These are not weaknesses to disguise; they are boundaries that keep the institutional claim credible.

The practical benefits are equally specific. By 2019, NANOG had independently audited accounts, multimillion-dollar operations and recurring programmes. By 2024, it continued to publish electoral and financial measures. Its mailing list remained public, archived and separate from membership. These are observable forms of continuity. None needs to be converted into a claim that incorporation improved routing or made the Board representative of every operator.

Important unknowns remain. The exact Delaware filing date and instrument for the 2019 name change were not established. The full 2011 transfer agreement was not examined. Item-level archive completeness cannot be verified. Eligible-voter totals for the 2010 and 2011 ballots are missing. The executed ARIN loan terms are unavailable in the record considered here. No governing document examined provides a delegation from particular networks or employers.

Those gaps do not undo the central result. They define it. The transition made ownership and responsibility more legible by giving NANOG’s institutional assets a named corporate custodian. It provided a vehicle for contracts, finances, staff oversight, committees, elections and succession. It also left the autonomous networks where they had always been: outside the corporation, controlled by their own operators and employers.

That is the constitutional modesty of an operators’ association. It can create a shared public benefit without becoming public government. It can hold the forum together without holding the networks. NANOG’s legal container mattered not because it captured a region, but because it made clear who was responsible for the association—and, just as importantly, what remained beyond its reach.

Sources