Summary
- NANOG has two distinct affiliation controls. Candidates disclose relevant relationships before an election; Directors and officers later disclose an affiliation before a related Board discussion and refrain from a decision when Article 15.2’s reasonable-person threshold is met.
- The current unauthenticated election archive preserves employer or organisation labels unevenly. Recent pages direct members to authenticated biographies, which supports a finding about external retention—not an allegation that candidates withheld information or that voters lacked access.
- A limited public record could preserve relationship categories, dates, decision subjects and participation outcomes while leaving full biographies and private details with the audiences that need them.
Two moments, two different tests
NANOG’s affiliation rules are more considered than a simple employer field on an election form. They operate at two moments, for two audiences and for two different purposes. Before an election, a disclosure helps members evaluate a candidate. Once a person is a Director or officer, a fresh disclosure helps the Board deal with a particular discussion or decision. Treating those moments as one control obscures what each is designed to do.
The first moment is set out on NANOG’s public page describing Board responsibilities and candidate eligibility. Candidates are expected to disclose all affiliations relevant to NANOG, including their main employer and other major relationships. NANOG’s own example shows why “employer” is too narrow a shorthand: a person employed by a nonprofit and sponsored by a vendor should disclose both organisations. The requested information is therefore a map of potentially relevant relationships, not merely the name on a payslip.
That requirement has a long institutional history. A 10 August 2011 election announcement requested candidate biographies for the election web page and required disclosure of a principal employer and other major relationships relevant to NANOG. The rule predates the annual public pages compared below. It reflects a durable premise: members should receive declared information about material organisational ties rather than having to infer them from reputation, conference encounters or outside searches.
The second moment begins only after institutional authority has been acquired. The current NANOG bylaws place the corporation’s property, affairs and business under the Board’s management and control. That authority is substantial, but bounded. The Board governs NANOG; it does not govern the employers named on election pages, third-party networks or North American network operations as a whole. The current Board has seven voting members—six elected Directors and the Executive Director—and a candidate for an elected seat must be a NANOG Member in Good Standing. Election confers an office within NANOG. It does not make the officeholder a delegate of an employer, a commercial sector or the continent.
Article 15.2 addresses the point at which an organisational relationship intersects with Board business. Before a Board discussion that implicates or relates to an organisation or association with which a Director or officer is involved or affiliated, that relationship must be disclosed to the entire Board. The rule then applies a further test to participation in a related decision: the Director or officer must refrain if a reasonable person would conclude that the affiliation could interfere with the person’s independent judgment.
The wording matters. Article 15.2 does not create an automatic employer disqualification. It requires a relationship to be disclosed in relation to a discussion, and it makes nonparticipation depend on a reasonable-person conclusion about independent judgment in a decision. A name that appeared in an election biography may alert the institution to a question, but it does not answer that question for every later agenda item. Equally, the passage of time since an election does not make a relevant relationship disappear. The Board-stage disclosure refreshes the record when context has become concrete.
Article 15.1 sits nearby but covers a narrower subject. It governs certain pecuniary-benefit transactions involving a direct or indirect financial interest of a Director or officer, subject to statutory exceptions. It should not be enlarged into a general ban on professional relationships. Employment may establish an affiliation while the available record says nothing about a transaction within Article 15.1. A financial interest may be relevant without being, by itself, a determined conflict. The nature of the transaction, the interest and any applicable exception still matter.
The two-stage structure is sensible because an election disclosure cannot anticipate the Board’s entire future agenda. Members can use a candidate’s declared relationships to ask about experience, incentives, obligations and approach. Later, the Board can examine an actual organisation, a defined subject and the officeholder’s current circumstances. The ballot-stage map informs electoral judgment; the Board-stage rule governs conduct in context. Neither can do the other’s work.
What the current public record does not consistently supply is the connection between them. An unauthenticated reader can find the candidate rule, portions of the annual election history and the bylaw governing later conduct. But the public trail does not reliably show whether an election-time relationship remained current, became relevant to a defined Board matter, prompted disclosure to the Board or resulted in participation or nonparticipation. Nor does it show whether the officeholder had authority to represent the affiliated organisation.
That is a public-legibility problem, not proof of a private failure. The record does not establish that a required disclosure was missed, that Article 15.2 was ignored or that members voted without adequate information. It establishes a narrower point: an outside reader has difficulty following the operation of the two controls across time.
An affiliation line opens inquiry; it does not close it
Governance analysis goes wrong quickly when “affiliation”, “interest”, “conflict”, “recusal” and “representation” become synonyms. They describe different facts or decisions. Keeping them separate protects both the institution’s accountability and the individual’s agency.
An affiliation is the broadest of these categories. Under NANOG’s election rule it can include the main employer, a sponsoring vendor or another major organisational relationship relevant to NANOG. Its value is that it identifies where a question might arise. It does not reveal the terms of the relationship or prove that the organisation endorsed the candidacy, instructed the candidate, financed a position on NANOG business or expects a particular vote.
A financial interest is more specific, but it is not self-executing. Article 15.1 uses financial-interest language in the setting of certain pecuniary-benefit transactions. One cannot move from “this person works for that organisation” to “this transaction is prohibited” without the facts that the clause makes relevant. Employment can affect incentives, but a general observation about incentives is not a finding about a transaction.
A conflict determination is narrower again. Under Article 15.2, the institutional question is tied to a discussion or decision involving an affiliated organisation and to whether a reasonable person would conclude that the relationship could interfere with independent judgment. This is neither an innocence presumption that makes affiliations immaterial nor a guilt presumption that converts a salary into control. It is a contextual test.
Decision-specific disclosure makes that test possible. A biography prepared for an election may be months or years old when the Board confronts a relevant matter. The relationship may have ended, changed form or acquired new significance. Another affiliation may have arisen since the election. Requiring disclosure to the entire Board before the related discussion puts current information before the people who must conduct the institution’s business. It is not a duplicate of campaign disclosure.
Nonparticipation is the consequence when the reasonable-person threshold is met. The bylaw’s language is precise: the person is to refrain from participating in the decision. Public commentary should not silently expand that into absence from an entire meeting, withdrawal from every discussion or a moral judgment about the person. At the same time, a vote tally that records an abstention cannot establish Article 15.2 nonparticipation unless the reason is given. People abstain because of conflicts, absence, uncertainty, strategy and other considerations. Similar-looking entries can represent different governance events.
Authorised representation is different from all of these. An employee is not necessarily an employer’s delegate. A candidate biography does not prove authority to bind the named organisation, speak on its behalf or cast votes under its instruction. If such authority exists in a particular matter, it should be stated as a separate fact. If it is not stated, neither readers nor the institution should imply it from a logo or job title.
The opposite inference is no safer. Saying that a Director serves in an individual capacity does not demonstrate the absence of professional incentives, dependencies or loyalties. People exercise judgment; they also inhabit institutions. An affiliation record is useful because it leaves both realities open to examination. It becomes misleading only when a relationship is treated as the answer rather than the beginning of a decision-specific inquiry.
This is why employer maps attract more certainty than they can bear. A list of firms can prompt good questions about where knowledge and exposure come from. It cannot prove a voting bloc, corporate instruction or capture. It cannot show that the candidates collectively represent the membership. It cannot show that a varied set of labels constitutes balanced governance either.
Representativeness requires a denominator, and the public election pages provide none. They do not disclose the complete employer, sector or ASN composition of NANOG members, voters, candidates, Directors or North American operators. They therefore cannot sustain a claim that a public candidate list mirrors, underrepresents or overrepresents the relevant population. No causal line can be drawn from an affiliation label to an election result, sponsorship arrangement, Board vote, agenda choice, routing policy or operating outcome on the evidence available here.
Trying to repair the institutional record with personal web searches would make the evidentiary problem worse. Professional profiles can be stale; names can be confused; public biographies can omit quiet but material relationships; current employment can be mistaken for employment at election time. More fundamentally, such reconstruction transfers responsibility from the institution to the observer and turns an audit of governance records into a dossier about individuals. The appropriate question is not whether a determined researcher can assemble a plausible employment history.
It is whether NANOG can preserve the minimum institutional facts needed to explain how its own controls operated.
Thirteen years of uneven public retention
The current public election archive is best understood as a retention map. It is not a reconstruction of the information presented to voters at the time, a compliance score for candidates or a count of conflicts. Fields can disappear during migration, designs can change and biographies can move behind authentication while an annual page remains online.
The official past-elections index provides annual navigation through 2025, a valuable basis for historical access. The existence of a year page, however, cannot establish that every original biography or employer field survived. With that limitation fixed, the present unauthenticated surfaces show the following complete 2013–2025 pattern:
| Current public year page | Public candidate names | Names with adjacent employer or organisation labels | Biography access statement on the public page |
|---|---|---|---|
| 2013 | 9 | 9 | Not assessed |
| 2014 | 5 | 0 | Not assessed |
| 2015 | 6 | 0 | Not assessed |
| 2016 | 6 | 0 | Not assessed |
| 2017 | 9 | 9 | Not assessed |
| 2018 | 4 | 4 | Not assessed |
| 2019 | 5 | 0 | Not assessed |
| 2020 | 6 | 6 | Member login for separate biographies |
| 2021 | 5 | 0 | Member credentials for biographies |
| 2022 | 6 | 0 | Separate member-only biography page |
| 2023 | 5 continuing + 1 withdrawn | 0 | Member login to view biography |
| 2024 | 6 | 0 | Separate member-only biography page |
| 2025 | 6 | 0 | Separate member-only biography page |
The 2013–2020 sequence is notably non-monotonic. The present 2013 page pairs nine names with nine employer or organisation labels. The 2014 page retains five names but no adjacent employer labels; the 2015 and 2016 pages similarly show six names apiece and no adjacent labels. All nine names have labels again in 2017, as do all four names in 2018. The five-name 2019 summary has none. The six labels return beside six candidates in 2020. This is neither a steady expansion of public disclosure nor a steady retreat from it. It is uneven retention on the current public surface.
The 2020 page is especially useful because it shows that two access layers can coexist. It publicly pairs six candidates with Juniper Networks, Q7, Oracle, Itaunas Telecom Consulting, Amazon Web Services and StackPath, while also directing logged-in members to separate complete biographies. The public employer lines did not require publication of the full member biography; the authenticated biography did not require removal of the public line.
From 2021 through 2025, every page reviewed preserves candidate names and repeats the affiliation-disclosure rule while describing biographies as available through member authentication. The 2021 page lists five candidates and refers to member credentials. The 2022 page lists six and points to a separate member-only biography page. The 2023 page identifies five continuing candidates and one withdrawn candidacy, with login required to view a biography. Withdrawal is an election-status fact only; it says nothing about affiliation or disclosure. The 2024 and 2025 pages each list six names and describe a separate member-only biography page.
Those last two pages also ask candidates to acknowledge the public Board and staff travel policy and an expectation that they can participate under it. That requirement concerns the practical conditions of service. It is not evidence that an organisational affiliation exists, that a conflict was determined or that a candidate is independent.
The annual table measures only what a person can see today without authentication. It says nothing conclusive about what voters saw in 2013 or any later election, what candidates submitted, what the Election Committee received or whether every required relationship was disclosed. It does not count employer delegates, determined conflicts or Board diversity. It is not a representative sample of operators. Its zeros mean “no adjacent label on this current public page”, not “no employer”, “no biography” or “no disclosure”.
The causes of the pattern remain unknown. Privacy policy, web migration, editorial format, authentication design or some other reason might explain why one year retains labels and another does not. The pages do not permit a choice among those explanations. The contents of the member-only biographies also remain outside the public evidence: their completeness, detail and treatment of material relationships cannot be inferred from the access statements. A careful audit leaves those questions open.
That restraint cuts in both directions. It would be wrong to say that names-only pages prove concealment or candidate noncompliance. It would also be wrong to say that authenticated access for voters answers every question about durable public accountability. Information can be sufficient for an electorate during an election yet limited public evidence for an outside reader trying years later to understand how a relationship intersected with Board authority. The archive reveals an internal-access/external-retention distinction, not a verdict about motives.
The strongest defence of member-only biographies
The case for authenticated biographies deserves to be stated without caricature. NANOG members elect the Directors. If a candidate biography is primarily an electoral document, giving the full version to the electorate can be a proportionate access choice. The general public does not automatically need every professional detail that a member may reasonably consider before casting a vote.
Biographies can contain more than an employer line. They may describe previous roles, volunteer service, sponsorship, institutional relationships and other context that becomes sensitive when detached from the election. Accuracy does not eliminate privacy risk. An old relationship can be presented as current; a sponsor can be mistaken for a principal; an employer can be portrayed as directing an individual. Permanent searchability can intensify those errors long after the relationship or candidacy has ended.
Restricting the fuller narrative can also improve candour. A candidate may be more willing to explain a complicated relationship to a defined electorate than to compress it for unrestricted distribution. A short public label can itself distort. A person might work for one organisation, serve personally in another and receive relevant support from a third. Publishing one brand while suppressing the context can invite the very employer determinism that disclosure is meant to resist.
The internal accountability case is substantive, too. Members are not merely passive recipients of a biography. They can ask candidates questions, compare answers and vote. It is reasonable for an association to place the richest election disclosure with the group that has formal electoral authority. Nothing in the public pages reviewed establishes that members were denied those materials; the recent pages say the opposite about access, although they do not reveal the contents.
Privacy also protects participation. Network governance benefits from practitioners whose knowledge was acquired inside firms, nonprofits and other institutions. If every disclosed connection is treated publicly as an accusation of corporate agency, capable candidates may avoid service or offer unhelpfully thin disclosures. Employment does not erase personal judgment, and an organisation should not build an accountability system that encourages readers to treat every company as the real candidate.
The defence becomes strongest when it is paired with differentiated access, not when it is asked to justify public silence indefinitely. Members may need the full election narrative. The Board may need confidential facts to apply Article 15.2. The public need not receive either set in full. But a continuing governance rule creates a legitimate interest in a smaller institutional record showing what happened when a relevant relationship met a decision.
That smaller record would concern the institution’s conduct, not the person’s life story. It could state that a current organisational affiliation was disclosed before discussion of a bounded agenda subject and record whether the officeholder participated in the decision. It need not publish private contact information, compensation, professional history or the content of Board deliberations. The difference is not cosmetic: one is a dossier; the other is an account of how authority was exercised.
Privacy and public accountability therefore need not be traded as indivisible blocks. Data minimisation asks what the public purpose actually requires. For historical audit, the answer can be much less than a complete biography: a relationship category, the period or update date, a decision subject described at a safe level of generality and the participation outcome. Such a record would leave the fuller disclosure with members and the Board while preserving continuity in the public archive.
Where the public bridge becomes necessary
NANOG already publishes other records in the name of durable accountability. Its financial-reports page describes audited financial statements and tax returns as part of ethical transparency and links annual audits for 2016 through 2024. Those publications do not prove anything about candidate disclosure or conflict enforcement. They do show that NANOG recognises a public audience whose interest in institutional stewardship extends beyond the entities in a particular election or transaction.
The organisation also maintains a public Board-minutes index with extensive recent PDF coverage, including years from 2020 through 2026. An index is not a transcript, and the availability of many PDFs does not mean every Board discussion is public. Minutes are selective by design. Some subjects may be treated in executive session or recorded at a higher level of generality for legitimate reasons.
Even so, minutes are the obvious place an outside reader might look for the second half of NANOG’s affiliation architecture. A limited text screen of eight public sets of minutes illustrates the problem. The meetings were dated 19 January 2024, 20 October 2024, 17 January 2025, 18 April 2025, 18 July 2025, 16 January 2026, 1 February 2026 and 17 April 2026. Across those eight documents, the text did not contain an explicit conflict, recusal, affiliation or disclosure label.
That observation has a deliberately small perimeter. Eight documents are not the full minutes archive. A keyword screen cannot detect an unlabelled process, a disclosure handled elsewhere, an action described in different language or a matter omitted appropriately from public minutes. It cannot establish that no conflict determination or nonparticipation occurred. It cannot say which Article 15.2 situations arose outside the sample, what executive sessions concerned or whether any affiliated organisation was involved.
The 19 January 2024 minutes contain an executive-session heading. The 20 October 2024 minutes are another document in the limited screen. The presence of an executive session says nothing by itself about the subject, a relationship or misconduct. It is not circumstantial proof of secrecy in the pejorative sense. Nor would the absence of an executive session prove that every relevant relationship was publicly discussed.
The 17 January 2025 minutes use motion tallies that distinguish votes in favour, votes opposed and abstentions. This procedural detail still cannot explain an abstention’s reason. A zero abstention count does not prove that no relevant affiliation existed. A nonzero count would not prove a recusal: the person might have been absent, uncertain, strategically opposed to voting or affected by another circumstance. Article 15.2 nonparticipation should be identified as such rather than reverse-engineered from arithmetic.
The 18 April 2025 minutes record an executive session, and the 18 July 2025 minutes record entry into executive session before the official meeting resumed. Again, those entries establish procedure, not subject matter. They provide no basis for claiming that a conflict, affiliated organisation or concealed decision was involved.
The 2026 documents carry a smaller lesson in archival precision. The meeting record officially dated 16 January 2026 has a canonical filename containing “2026-01-26”. The date of the meeting is 16 January, despite the filename. The 1 February 2026 minutes state that there was no executive session, and the 17 April 2026 minutes complete the eight-document set. Neither the presence nor absence of an executive session answers whether a relevant disclosure was made.
It would be counterproductive to demand minutes that reproduce sensitive deliberations. Detailed accounts can expose confidential commercial terms, chill candid discussion or turn a prudent participation decision into reputational punishment. The public interest can be served by a much narrower marker: that a relevant relationship was disclosed in connection with a defined subject and what participation outcome followed. The Board’s private reasoning and the individual’s private financial details can remain private.
Without such a marker, silence is radically ambiguous. It could mean no relevant affiliation existed. It could mean an affiliation existed but the decision did not involve the organisation. It could mean the affiliation related to the matter but did not meet the reasonable-person threshold for nonparticipation. It could mean the threshold was met and the person refrained. It could mean the relevant information was handled in a record outside the public minutes. All are compatible with an unlabelled public entry.
That ambiguity is why silence cannot support an allegation. It is also why silence cannot demonstrate the operation of the rule. The entity of better reporting is not to remove every uncertainty; institutional life does not permit that. It is to preserve the result of a defined procedure so that readers do not have to choose between suspicion and faith.
What nonprofit guidance clarifies—and what it cannot prove
External guidance is useful here only as a vocabulary for separating steps. The IRS Form 1023 instructions say that adoption of a conflict-of-interest policy is not required for tax-exempt status and that the included sample is not prescriptive. The sample also explains that a financial interest is not necessarily a conflict. Disclosure of an interest, determination by disinterested people and nonparticipation are treated as distinct parts of a procedure.
The IRS Form 990 instructions similarly distinguish having a written policy, facilitating disclosure, monitoring or enforcement and imposing participation restrictions. These are reporting concepts, not a finding about NANOG and not a governance form that NANOG is shown to be legally required to adopt. The guidance cannot establish what happened in any NANOG meeting, whether Article 15.2 was triggered or whether the organisation enforced a rule correctly.
As an analytical comparator, however, the separation is valuable. “NANOG has a policy” is one claim. “A relationship was disclosed” is another. “The relationship was found capable of interfering with independent judgment in this decision” is another still. “The person did not participate” describes a consequence. Conflating them creates false positives—every job looks like a conflict—and false negatives—the existence of a bylaw looks like proof that every case was resolved and documented.
NANOG’s own design already recognises much of the sequence. The election rule provides relationship information before members vote. Article 15.2 requires a fresh disclosure to the entire Board before a discussion involving an affiliated organisation. Its reasonable-person standard makes the participation decision contextual. Its instruction to refrain from a decision supplies a conduct rule when that threshold is reached. The public-record question is how to preserve a modest trace of this existing sequence, not how to invent a conflict regime that NANOG lacks.
Consider an agenda item that concerns an organisation named in a Director’s old election biography. The item might involve a contract, sponsorship, meeting service or only an incidental reference. The affiliation might be current or might have ended. The Director might possess confidential information, have no financial stake, or hold some other role not visible in the biography. The old employer line cannot resolve any of those facts. The current disclosure and the reasonable-person test exist because relevance is created by the combination of relationship and matter.
The resulting public entry should be equally contextual. “Participated” would not certify independence as a general character trait. It might mean the Board concluded the threshold was not met, that the relationship was no longer current or that the matter did not meaningfully involve the organisation. “Did not participate” would not establish wrongdoing. It would show that the control produced a precautionary outcome for that decision. Good records describe institutional acts without converting them into verdicts about people.
The distinction also disciplines language about representation. Directors inevitably bring professional experience to a Board. That is often why members elect them. Experience acquired at an employer may inform a view without being an instruction from that employer. Unless authority to act for an organisation is expressly established, “affiliated with” is the supportable description. Where an officeholder really is authorised to represent an organisation in a particular capacity, the record should identify that fact separately rather than allowing readers to infer it from employment.
This approach refuses two easy stories. One story sees a company name and assumes capture. The other sees an individual office and assumes perfect insulation from organisational incentives. Neither follows from an affiliation label. A procedural record is valuable because it shows what the institution did when the relationship became relevant, leaving claims about motive or control to evidence that can actually support them.
The minimum public record
A useful public bridge need not be a large register of people and inferred ties. It can be a compact institutional template with fields that prevent unlike concepts from collapsing into one another. The following design is a recommendation based on the public evidence, not a description of NANOG’s current practice.
The election-side entry should identify the current employer and other material organisational relationships covered by the candidate rule. Each should carry an effective date or last-confirmed date. Time is essential: an election label should not become a permanent identity after a person changes jobs, ends a sponsorship or leaves an outside role. Updates should supplement the historical record rather than silently rewriting what was known at the election.
“Other material relationships” should be a category, not an invitation to publish a full résumé. It can record that a relevant organisational role, sponsorship or comparable connection exists without disclosing compensation, contact information or immaterial personal history. This respects NANOG’s own example, which recognises that a nonprofit employer and a sponsoring vendor can both matter, while avoiding the fiction that the principal employer exhausts all relevant ties.
The entry should show when it was last updated. That gives the officeholder a clear way to correct stale information and allows readers to distinguish an election-time declaration from a current status. Where a correction is necessary, the archive can preserve the fact and date of the correction without retaining private detail that no longer serves an accountability purpose.
The decision-side entry should identify the subject at a useful but bounded level. “Service-provider contract”, “event sponsorship” or another general category may suffice where negotiations or commercial terms are confidential. The purpose is not to recreate deliberations. It is to explain why an organisational relationship entered the Article 15.2 process.
It should then record the participation outcome in the bylaw’s own conceptual terms. Did the Director or officer participate in the decision or refrain? If the item involved discussion but no decision, the record should say that rather than inventing an abstention. If an abstention occurred for an unrelated reason, it should not be relabelled as conflict management. A consistent outcome field would do more for auditability than a raw vote tally that leaves the reason unknowable.
Finally, the record should distinguish affiliation from authority to represent. A field might say that no representational authority was stated, or describe a narrow authority if one actually existed. The default must not be that employment turns a NANOG officeholder into an employer’s agent. Making the field explicit would protect organisations and individuals from an inference that the underlying disclosure never established.
These elements—current employer, other material relationships, update date, decision-specific disclosure, participation outcome and authority represented if any—could be divided across an election archive and a decision register linked by dates. They do not have to appear on a permanent person profile. Members could retain access to richer candidate biographies; the Board could receive the confidential detail needed for its judgment; the public layer could preserve only the structured subset needed to understand institutional continuity.
The template should include its own cautions. A relationship is not a conflict determination. Nonparticipation is not an admission of misconduct. Participation does not certify independence beyond the decision recorded. “No authority stated” does not prove that incentives are absent; it means only that representation has not been established. Those distinctions should be part of the record’s explanation so that added transparency does not produce more confident misreading.
Nor should the form become compliance theatre. More boxes do not guarantee better judgment. The categories must be updated, applied consistently and interpreted conservatively. A decision subject should be specific enough to make relevance intelligible but not so detailed that it discloses confidential negotiations. A relationship field should draw on information provided through the institution’s process, not speculative searches. Participation outcomes should record what happened without assigning blame.
The uneven older archive should not be “completed” retrospectively from social media, conference pages or employer announcements. For 2013–2025, NANOG can accurately preserve what each present public page contains and identify what is unknown. A prospective template can begin with verified current information and retain it consistently. Honest gaps are preferable to a seamless history assembled from evidence that was never part of the institutional record.
There will also be hard cases. Disclosure of the organisation’s identity might itself reveal confidential negotiations. A general subject category may then be the appropriate public entry, with detail held by the Board. A relationship characterisation may be disputed or change during a term; the officeholder should have a correction process, and the record should date the update. A matter may be discussed without a formal decision. None of these cases makes a public record impossible. They show why the record must describe procedure rather than offer a moral score.
Continuity is the accountability test
NANOG’s rules already reject the two most damaging simplifications. By requiring relevant affiliations at candidacy and again before related Board discussion, they reject the idea that organisational relationships never matter. By tying nonparticipation to a particular decision and a reasonable person’s conclusion about interference with independent judgment, they also reject the idea that employment automatically disqualifies someone.
The public archive should preserve the same balance. Missing employer labels on a current page do not prove that a candidate concealed a relationship or that a member voter lacked a biography. An employer name does not prove instruction, authorised representation or capture. An abstention does not prove a recusal. An executive session does not prove secrecy, conflict or misconduct. Eight minutes documents without an explicit label do not prove that Article 15.2 was never used.
Important facts therefore remain unknown: the full contents and completeness of member-only biographies; the reasons fields differ across migrated pages; the motives for authentication; the application of Article 15.2 outside the eight-document screen; the subjects of executive sessions; and whether any employer instructed, funded, constrained or authorised an individual in a particular NANOG matter. The public evidence also cannot determine constituency composition or connect affiliation causally to elections, sponsorship, agendas, routing or operating outcomes.
Those unknowns are not a reason to abandon audit. They define its proper limits. NANOG can protect private biographies and confidential deliberation while publishing enough institutional information to show that its decision process occurred. Relationship categories, dates, a bounded decision subject, a participation outcome and any actual representational authority would answer the procedural question without pretending to answer questions of motive.
The central finding is therefore both restrained and demanding. NANOG already has a candidate-disclosure rule and a Director/officer disclosure-and-nonparticipation rule. What the present unauthenticated history lacks is a consistent public link between the relationship first disclosed to voters and the later moment when that relationship may have met Board authority.
Closing that gap would not decide whom a Director serves by reading an employer name. It would do something more useful: show how the institution applied judgment when a relationship became relevant. That is the point at which affiliation transparency stops being a list of organisations and becomes an accountable record of governance.
SEO and social metadata
- SEO title: From NANOG’s ballot to its Boardroom
- SEO description: NANOG has candidate-affiliation and Board nonparticipation rules. The public audit question is how a disclosed relationship connects to a later decision.
- Social title: What an employer name can—and cannot—tell NANOG voters
- Social description: Affiliation disclosure can make relevant relationships visible without treating employment as proof of corporate instruction, representation or capture.
Image copy
- Alt text: A synthetic editorial diagram with a generic candidate card, two unbranded organisation cards, a Board agenda card and separate participated and did-not-participate outcomes.
- Caption: An affiliation identifies a relationship worth examining; accountability depends on showing what happened when that relationship reached a specific decision.
- Accessibility description: A synthetic editorial illustration uses layered translucent cards to follow a generic governance sequence. A candidate card connects to two disclosed-organisation cards without names or logos, then to a Board agenda card. The final layer divides into clearly labelled participation and nonparticipation paths. No real person, company or historical event is depicted.
- Synthetic provenance: This synthetic editorial illustration represents an abstract disclosure and decision process. It is not a historical photograph, does not portray a real candidate, uses no corporate logos and does not imply that any employer controls an individual.