Summary

  • NANOG's archive documents conflict disclosure and self-recusal in 2004; the public record reviewed here does not establish the current rule or its operation.
  • Aggregate, proposal-level reporting could make recusal more observable without naming reviewers or exposing confidential submissions.

NANOG's own archive shows that the organisation recognised this problem early. A 2005 account of the selection process says a conflict-of-interest policy had been adopted in May 2004. Reviewers were expected to disclose conflicts during discussions and recuse when they considered that appropriate. The record is valuable. It is not a current operating report.

The 2004 process was more than an informal conversation

The archived description lays out a real workflow. Committee members rated proposals from one to five, added comments, and could not see one another's ratings until a later selection call. The committee could accept, reject or request changes, with a further review after revisions. Conflict disclosure and self-recusal sat inside that process.

That architecture separated individual scoring from group deliberation and created a route for improving a submission. It also left important choices with reviewers: what counts as a conflict, when recusal is sufficient, and who replaces the missing judgment. The archive does not show that those choices were abused. It shows where accountability depends on rules that the public cannot fully inspect.

Current pages describe review, not conflict handling

NANOG's current call for content tells submitters to use the Program Committee tool, provide an abstract and draft slides, and work with an assigned shepherd. The committee performs an initial review, continues developing the material and decides what reaches the agenda. This is a clear public description of the submission path.

It does not state whether the 2004 conflict policy remains unchanged, whether affiliations are collected in a standard form, whether a recused reviewer loses access to the proposal, or whether replacement review is mandatory. Absence from the page is not proof that controls are absent. It means the public page cannot verify how the control operates today.

Expertise and independence are not opposites

A strict rule excluding anyone who shares an employer category, market or technical specialty with a submitter would hollow out peer review. A routing operator can recognise operational evidence that a generalist misses. A vendor engineer can distinguish a deployment account from a disguised product pitch. A researcher can detect when a measurement claim outruns its method.

The useful boundary is therefore not distance from the field. It is disclosed proximity plus a decision rule. Direct involvement in the proposed work, a reporting relationship, a close commercial stake or a live dispute may justify recusal. General knowledge of the same technology usually should not. The rule needs enough specificity that two reviewers reach broadly comparable decisions.

Aggregate disclosure can protect confidential proposals

NANOG need not publish rejected submissions or name every reviewer who stepped aside. It could report, for each meeting, the number of proposals reviewed, the number that triggered at least one declared conflict, how many ratings were excluded, and whether every affected proposal received a minimum number of unconflicted reviews. It could also publish the current policy version and the role responsible for resolving uncertainty.

Those figures would not prove that every decision was correct, that the process was fair or that the mechanism was effective. They would make its operation more observable and auditable. They would also let the committee explain exceptional cases—such as a tiny specialist pool—without revealing a submitter's confidential draft.

The mandate ends at the programme

The Program Committee's authority is substantial but narrow: it curates NANOG's meeting programme. Selection can confer visibility and shape what attendees discuss. It does not certify a technology, bind an employer, or establish the position of North American network operators.

Other official records clarify who carries that responsibility. A 2008 vacancy notice required members to review every submission unless excused. A 2010 election notice described a sixteen-person committee, while a 2019 announcement recorded appointments and the committee's choice of chair and vice-chair. A 2020 governance page says the Board selects committees and appoints a voting director as ex-officio liaison. These records locate institutional responsibility; they do not document today's recusal mechanics.

That boundary matters when evaluating conflicts. A weakly documented review process can damage confidence in the agenda without proving control of the wider industry. The remedy is a better process record, not an accusation. NANOG already published the essential idea in 2004: disclose and recuse. The unfinished task is to make the current rule, its substitution path and its aggregate operation visible enough to be tested.

Sources and limits