Summary
- Several RIR processes sensibly require a complainant first to raise disagreement with the chairs or working group. This can clarify misunderstandings and correct obvious errors without a formal case.
- RIPE's current appeals route then moves to the Working Group Chairs Collective, excludes the relevant chairs, proposer, appellant and RIPE Chair team from that decision, and permits referral to the RIPE Chair if the collective cannot resolve the dispute. AFRINIC uses a standing Board-appointed appeal committee; LACNIC's published process has historically routed formal appeal to its Board.
- The first step becomes illegitimate when it has no deadline, demands persuasion of the original decision-maker, narrows issues before independent review, exposes complainants to retaliation or allows chairs to cure the record after seeing the appeal.
- A sound system labels the stages “reconsideration” and “appeal,” preserves the original record, stops relevant deadlines, permits direct escalation for bias or urgency, uses independent selection and recusal, and grants remedies capable of changing the process rather than merely publishing criticism.
A familiar door with the wrong name
Someone believes working-group chairs ignored an objection, misstated the discussion or declared consensus that the record cannot support. The process tells them to contact those chairs first. This instruction can look absurd: ask the people who made the disputed judgment to judge it again. Yet the first return is not inherently defective. Decision-makers can correct a factual error, explain a misunderstood step or reopen discussion faster than a formal panel.
The defect begins when this encounter is called an appeal. Reconsideration asks the original decision-maker to revisit their work. Appeal asks a different authority to review it. Both can be valuable, but they answer different legitimacy needs. A system that merges them allows institutional language to promise independence while the entity remains before the same people.
RIR policy development relies heavily on volunteer chairs. They moderate lists, guide authors, summarise objections, manage meeting time and determine whether rough consensus exists. Their judgment cannot be replaced by a simple count. Precisely because that judgment is broad, a entity needs a credible route when they allege it was exercised unfairly or outside the agreed process.
Returning first to the chairs should therefore be a short, documented opportunity for correction—not a loyalty test, a mandatory negotiation without time limit or a chance to rewrite the record before anyone independent sees it. The appeal system becomes legitimate at the moment authority genuinely moves.
What the published procedures do
The RIPE Policy Development Process says concerns should first be raised with the relevant working-group chairs. If unresolved, anyone may submit an appeal to the Working Group Chairs Collective within four weeks of the appealable action. The appeal is public. The collective decides by consensus, and the relevant chairs, proposer, appellant, RIPE Chair and Vice Chair recuse. At least five chairs participate. If the matter remains unresolved, it is referred to the RIPE Chair for a final decision.
RIPE's review of the 2019-04 appeal is particularly valuable because it examines a process after real use. In 2020, the proposer appealed the Anti-Abuse Working Group chairs' determination that the proposal lacked consensus. The Working Group Chairs Collective upheld the decision. The later review identified practical lessons and exposed how a procedure that looks complete on paper behaves under pressure.
The AFRINIC Policy Development Appeal Committee is a standing, Board-appointed body. Its governing text requires a person first to discuss disagreement with the chairs or working group. If unresolved, an appeal supported by three working-group entities can be filed within two weeks. The committee may annul a chairs' decision when the process was not followed.
The LACNIC PDP likewise requires initial consideration by chairs and has historically permitted a formal appeal through the Board, with a four-week decision period. Across these systems, the initial return is common, while the second-stage institution and independence safeguards differ.
Reconsideration can be the fastest remedy
Many disputes are not constitutional crises. A chair may have missed an email sent near a deadline, linked the wrong proposal version or described a entity as satisfied when they were not. The complainant may have misunderstood which phase was open. A direct exchange can correct the record in days.
Reconsideration also respects chair responsibility. Chairs should not be stripped of authority whenever a entity uses the word appeal. They need a chance to explain the consensus analysis and correct an error without forcing a panel to reconstruct a long discussion. Formal review consumes volunteer time and can polarise a community.
The first step can improve an eventual appeal. The entity identifies the exact action, evidence and requested remedy. Chairs provide reasons. Issues that remain disputed become clearer. An independent body receives a focused record rather than a general accusation that the process felt unfair.
These benefits depend on structure. The reconsideration request must be acknowledged, answered on a short deadline and preserved. It must pause or protect the formal filing period. Chairs must not demand that the complainant withdraw criticism as a condition of correction. If the dispute alleges bias, retaliation or misconduct by the chairs, direct escalation should be available.
The danger of forced persuasion
An informal rule can become a substantive barrier when the complainant must convince the original chairs before gaining access to appeal. The chairs may characterise the issue as content disagreement rather than process error, declare it resolved and argue that no appeal remains. A entity unfamiliar with procedure may accept that conclusion even though the independent body could review it.
The threshold should be exhaustion, not satisfaction. A complainant has exhausted reconsideration when they sent a sufficiently specific concern and received a response or the response deadline passed. They need not persuade the chairs that their own decision was wrong. The appeal body, not the challenged decision-maker, determines jurisdiction.
Forms and guidance should avoid language such as “try to resolve the matter to the chairs' satisfaction.” They should state that reconsideration is intended to clarify or correct, that the right to appeal remains, and that the original chairs cannot block filing. A central secretariat can receive both notices and track deadlines without deciding merits.
This distinction is especially important in rough-consensus systems. A chair can honestly believe every objection was addressed while an appellant honestly believes the summary omitted the strongest one. The existence of that disagreement is the reason for review, not evidence that review is premature.
Deadlines can expire while people are being polite
Volunteer communities reward informal conversation. A complainant may spend days exchanging careful messages, waiting across time zones and avoiding escalation. If the formal appeal clock continues running from the original action, the effort to resolve the matter can destroy the right to review.
The rule should toll the filing deadline when a written reconsideration request is received. Either the remaining time resumes after the chair response, or a fresh, defined period begins. The process should not depend on whether a mailing-list thread is considered formal enough. A simple subject label or web form can create certainty.
Chairs should answer within a short period, perhaps seven days for ordinary decisions and faster where implementation or last call will proceed. If more time is needed, an independent administrator—not the challenged chairs alone—should confirm the extension and protect the appeal right.
The underlying proposal may also need a temporary hold. Not every complaint should freeze policy; that would invite strategic delay. A stay should depend on potential irreversibility, seriousness, likelihood that review would become meaningless and harm from pause. The appeal body or neutral officer should decide quickly.
Preserve the decision before it can be improved
Once chairs receive a challenge, they may naturally clarify their summary, add references or explain reasoning more fully. Correction is desirable, but it can obscure what was originally decided. An appeal body must know whether the contemporaneous record supported the action, not only whether a better rationale was assembled afterward.
The original consensus declaration, proposal version, relevant messages, meeting transcript and poll should be preserved at filing. Later corrections should be versioned and dated. Chairs can concede an error, withdraw the decision or issue a replacement. They should not silently edit the page so the appellant appears to have challenged words that no longer exist.
This is not adversarial evidence preservation for its own sake. It allows institutional learning. If chairs routinely need to add reasons after complaints, consensus announcements need a stronger template. If appeals arise from broken links or ambiguous dates, administrative design can improve.
The preservation package should be proportionate. It need not copy every mailing-list post when stable archives exist. It should identify the exact record and checksum or timestamp essential items. Private moderation material should be handled under clear confidentiality rules and disclosed to the appellant as far as fairness allows.
Independence is about relationships, not labels
A body called an appeals committee may still be close to the original chairs. Working-group chairs across an RIR collaborate, select one another, share employers or have taken positions on the proposal. A Board may appoint reviewers and later receive their policy recommendation. Formal separation is only the beginning.
RIPE's current recusal list recognises several direct conflicts and excludes the relevant working-group co-chairs, proposer, appellant and RIPE Chair team from the collective's decision. It also permits an external facilitator exceptionally. These are meaningful controls. Yet other chairs may have participated in the underlying discussion or publicly supported a side. They should disclose and recuse where their involvement would reasonably impair review.
AFRINIC's standing committee model provides continuity and a defined mandate, but Board appointment and composition create their own accountability questions. Selection criteria, terms, conflicts and replacement rules should be public. LACNIC's Board route offers a clear institutional destination but can mix corporate oversight with policy-process review.
No design is automatically independent. A credible body has members selected before the dispute where practical, diverse enough to resist one network of relationships, bound by disclosure, and removable only under transparent rules. Its decision does not require approval from the people it reviews.
Small communities cannot promise perfect distance
The specialist pool is limited. People qualified to understand a years-long address-policy discussion may have commented on it. Excluding everyone with prior knowledge can leave reviewers who understand procedure but not the dispute. Independence must be balanced with competence.
The right test distinguishes exposure from commitment. Reading the list or attending the meeting is not a conflict. Authoring text, advising a party, declaring a position, sharing a material employer interest or participating in the challenged decision may be. Reviewers disclose relationships, and a neutral facilitator decides contested recusals.
External reviewers can help with process questions, but they should not import another region's policy preferences. A panel might combine experienced chairs from unrelated groups, former community leaders and an independent procedural member. Training and a written standard reduce dependence on personal familiarity.
Quorum rules matter. If recusals leave too few reviewers, the process should appoint substitutes rather than pressure conflicted members to remain. RIPE's minimum participation requirement is one way to avoid a tiny remainder deciding for the whole collective.
Scope must be neither a second debate nor a shield
Appeals commonly review whether the process was followed and whether a consensus determination was reasonable; they do not decide the technical merits of the proposal anew. This boundary protects bottom-up discussion from being replaced by a small panel.
But process and substance cannot be separated mechanically. If an appellant says chairs ignored a material objection, reviewers must understand the objection well enough to determine whether it was addressed. They need not decide whether the proposal is good policy. They must assess whether the chair's treatment had a supportable basis in the record.
An appeal standard might ask whether the decision identified the correct proposal version, considered all material channels, accurately summarised support and objection, applied the published definition of consensus, handled conflicts, gave required notice and offered reasons capable of review. It should also ask whether any error could have affected the result.
The body should reject attempts to use appeal as extra speaking time on a losing argument. It should not reject a procedural claim merely because evaluating it requires reading technical material. A claim's label does not determine scope; the requested institutional judgment does.
Support thresholds can protect and exclude
AFRINIC requires support from three working-group entities for an appeal. Such thresholds can deter frivolous or purely personal cases and show that a process concern has community significance. They can also make review inaccessible to a lone entity whose exclusion is itself the problem.
Threshold design should reflect the remedy and risk. A challenge to a general consensus call may reasonably require limited support. A complaint about denial of speaking opportunity, discrimination, undisclosed conflict or retaliation should not depend on colleagues publicly aligning with the complainant.
Support should attest that review is warranted, not that supporters agree with the proposal or desired result. People may believe chairs reached the correct policy outcome but used a defective process. The form should make that distinction clear.
Private expressions of support may be necessary where retaliation is credible, with identities verified by a neutral officer. The final decision can report the count without exposing names. Abuse controls should focus on evidence and conduct rather than erecting a threshold so high that only organised factions can appeal.
Public filing can chill legitimate complaints
Publishing appeals promotes transparency and allows the community to understand a challenge. RIPE requires public submission to relevant lists and publication on its site. This is appropriate for disputes about public consensus records, but full publicity can expose appellants to hostility, especially when chairs are respected community figures.
The system should separate the public issue from protected details. The appeal, response and evidence about public discussion can normally be public. Personal data, moderation reports, harassment evidence and confidential employment conflicts may require restricted handling. Redactions should be explained and available to reviewers and parties under fair conditions.
A entity should be able to seek procedural advice confidentially before filing. The adviser must not decide merits or become an advocate. This lowers the knowledge barrier and helps frame specific claims.
Anti-retaliation expectations should cover meeting access, mailing-list moderation and future chair treatment. Enforcement can use the code-of-conduct mechanism where appropriate. The fact that an appeal failed should never be treated as evidence of misconduct unless the filing itself was abusive under a defined standard.
Remedies determine whether review is real
An appeal body that can only issue observations may improve future practice but cannot protect the current entity. AFRINIC's published authority to annul a chair decision if the process was not followed is concrete. RIPE's process can uphold or reject an appeal and ultimately reach a final decision. Other systems should be equally clear about consequences.
Available remedies should include correcting the record, repeating a consensus call, extending discussion, restoring a proposal to an earlier phase, changing facilitators, requiring recusal, reopening a missed participation channel and annulling a decision. The remedy should target the defect rather than award policy victory to the appellant.
If a consensus call is annulled, the proposal does not automatically become accepted. It returns to a lawful stage. If chairs failed to consider an objection, they or replacement facilitators evaluate it after proper discussion. Review protects process authority rather than transferring policy power to the panel.
The decision should explain implementation status. If the policy already took effect, the body may need to recommend temporary handling while the community reconsiders. A remedy that arrives after irreversible action requires a candid account of what cannot be undone and how future cases will be treated.
Reasoned decisions create precedent without binding doctrine
Appeal outcomes should identify the claim, standard, relevant facts, reasoning, remedy and any dissent or uncertainty. This helps chairs understand expectations and entities assess future cases. It also constrains inconsistency.
RIR communities may resist legalistic precedent, reasonably fearing that volunteer processes become courts. A reasoned archive need not create rigid binding doctrine. It can state that future panels should consider prior decisions and explain material departures. That is enough to make like cases more coherent.
The archive should be searchable by issue: notice, recusal, consensus summary, version mismatch, participation, deadline, scope and remedy. Personal disputes should not dominate titles. A periodic review can identify recurring design flaws and recommend procedural amendments through the community's normal governance route.
The RIPE review after the first use of its appeal procedure demonstrates the value of learning from a live case. Procedures reveal ambiguities only when tested. Every region should conduct a post-case review that includes appellant, chairs, reviewers and observers without relitigating the outcome.
The final reviewer can recreate the same problem
A multi-stage process may begin with the original chairs, move to a collective and end with a single community chair or Board. The final stage can concentrate power again. RIPE excludes the RIPE Chair and Vice Chair from the collective, then permits referral to the RIPE Chair if the collective cannot resolve the dispute. This provides closure, but it demands strong reasons and conflict safeguards.
A final reviewer should not have advised the original chairs, advocated on the proposal or selected the appeal outcome informally. If conflicted, a predetermined substitute should act. The standard of review should be clear: reconsider the matter, review for procedural reasonableness or resolve only the collective's deadlock.
Finality is necessary. Endless appeal would paralyse policy. It is legitimate when the route was independent, timely and capable of remedy. Final does not mean beyond public criticism; decisions remain part of the governance record and the community can amend the procedure prospectively.
Where a Board is final reviewer, it should distinguish its corporate fiduciary role from policy merits. It reviews whether the community process worked under the adopted rules, not whether directors prefer the outcome.
Election legitimacy reaches into policy appeals
Chairs and Board members who hear appeals are selected through governance processes. Their legitimacy depends on election rules, terms, disclosure and accountability. An appeal system cannot be more independent than the institutions that appoint its reviewers.
Chair candidates should disclose relevant policy authorship, employer interests and prior appeal roles. Election materials should explain the adjudicative responsibilities attached to the role. Members often vote on visible facilitation skill without realising that chairs may later make quasi-adjudicative consensus judgments.
Boards appointing committees should publish criteria, applications where appropriate, conflicts, terms and reasons. Appointment should not occur after the Board knows which dispute the committee will hear unless unavoidable. Staggered standing membership reduces case-specific selection.
Removal protections matter. A reviewer should not fear losing a role for annulling a popular chair's decision. Conversely, members need a process to address persistent bias or incapacity. Transparent term and recall rules preserve both independence and accountability.
A clean two-stage design
Stage one is chair reconsideration. A entity files a concise notice identifying the action, alleged error, evidence and requested correction. A neutral secretariat timestamps it, preserves the record and tolls the appeal deadline. Chairs respond within seven days, correct, withdraw or affirm the action with reasons. Direct escalation is available for alleged chair bias, retaliation or urgent irreversible harm.
Stage two is independent appeal. Filing requires no permission from the chairs. A standing panel or pre-defined selection mechanism confirms jurisdiction, disclosures and recusals. It decides any stay, publishes a timetable and invites focused submissions. Review concerns process and defensibility of consensus judgment, not a fresh policy vote.
The panel can affirm, annul, remit with instructions, require recusal or correct the record. It decides within a public target, explains extensions and publishes reasons with necessary privacy protection. The underlying policy timeline resumes according to the remedy.
Afterward, a separate procedural review asks what the case teaches. It does not change the outcome. Recommended rule amendments return to the community. This separation prevents a panel from expanding its own power through case decisions.
Evidence that the system works
An institution should publish aggregate appeal data: reconsideration requests, corrections, formal appeals, time to decision, stays, outcomes, recusals and remedies. A low appeal count is not automatically success. It may show trusted chairs, or inaccessible review.
Surveys and interviews can test awareness, fear of retaliation and clarity. The institution should examine whether first-time entities, remote attendees or speakers using other languages can use the route. It should track abandoned complaints caused by deadline confusion.
Quality matters more than volume. Decisions should meet timeliness and reason-giving standards. Remedies should actually occur. If a panel orders a new consensus call, the archive should link it. If implementation continued despite a stay, the breach needs explanation.
Periodic simulation can test readiness before a high-conflict case. Chairs, secretariat and potential reviewers walk through filing, preservation, recusal and communication using fictional facts. This is administrative preparation, not rehearsal of preferred outcomes.
The appellant needs the reasons they are challenging
An appeal right is hollow if the original decision contains only a conclusion. “Consensus reached” or “no consensus” does not reveal which objections chairs considered, what evidence they relied upon or why contrary views did not prevent closure. The appellant is forced to guess at the reasoning and may be criticised for challenging the wrong thing.
Consensus declarations should therefore be reasoned before any complaint. They should identify the exact text, participation channels, material objections, responses and chair judgment. Reasons do not need to reproduce the entire discussion. They must be sufficient for a entity and reviewer to trace the path from record to outcome.
If reasons were inadequate, an appeal body should decide whether to annul, remit for a fresh decision or accept supplemental explanation. The choice depends on prejudice. A missing citation may be cured; a rationale first invented after the appeal may show that the contemporaneous judgment was not properly made.
Reason-giving also disciplines the initial return to chairs. Reconsideration becomes a comparison between a specific challenge and a published explanation, not a private argument about tone or trust. The chairs can concede that a point was missed without treating the complaint as an attack on their general competence.
Language and time-zone inequality follow the case
Policy appeals often impose tighter deadlines and more formal writing than ordinary participation. This can magnify existing language and regional inequality. A fluent insider can produce a procedural brief quickly; a remote entity may need translation, transcript access and advice before identifying the appealable action.
The institution should provide plain-language guidance, examples of valid claims and a filing form that accepts concise facts. Translation support should be available for core submissions where feasible, with the authoritative version and any ambiguity identified. Deadlines should account for when a usable transcript or written decision became available, not merely when words were spoken at a meeting.
Hearings should not require travel. Asynchronous answers and remote sessions should carry equal status. Time zones can rotate where a panel spans regions. Accessibility accommodations, captions and readable evidence packages are part of due process, not optional meeting services.
The standard should remain substantive. Assistance helps a person state a claim; it does not lower the evidence required. Equality means that procedural sophistication is not the hidden qualification for review.
Multiple appellants should not become a popularity contest
A controversial consensus call may produce several complaints. Treating them as separate appeals wastes effort and risks inconsistent outcomes. Merging them without consent can also erase distinct claims or allow one well-resourced appellant to dominate.
A neutral case manager should group challenges to the same action while preserving each appellant's allegations and requested remedy. Common evidence is filed once. Appellants may coordinate or submit separately. The panel decides shared process questions together and addresses unique issues explicitly.
The number of appellants should not determine the result. Ten repeated filings do not prove a process defect, just as a lone filing does not prove insignificance. Support and opposition can inform context, but review applies the adopted standard to the record.
Public communication should avoid framing the case as chairs versus a faction. The issue is institutional: whether a specified action complied with the process and a defensible consensus judgment. This protects entities from political escalation and keeps the remedy focused.
Implementation during appeal needs a default rule
If policy implementation begins while appeal is pending, a successful challenge may arrive too late. If every filing stops implementation, opponents can create delay through weak claims. The process needs a default and a fast exception test.
One approach is a short automatic standstill after a consensus declaration, long enough for reconsideration notice. After that, implementation proceeds unless a neutral reviewer grants a stay. The stay test considers likelihood of a reviewable defect, irreversibility, harm to users, public interest and the time required for decision.
The reviewer can tailor relief. They may pause one contested provision, preserve applications without deciding them, or allow technical preparation while preventing final action. Reasons and duration should be public. A stay is not a prediction that the appellant will win; it protects the value of review.
Where no stay is granted, the institution should warn affected users that the policy is under appeal and preserve the ability to remedy cases. Completed allocations or registration changes may be difficult to reverse, making early and narrow protection preferable to broad retrospective disruption.
Chairs should not control the appeal's public narrative
Because chairs manage the mailing list and meeting, they may also control the channels through which an appeal is announced. Even neutral moderation can appear self-protective when the moderator is a party. Communication authority should move with adjudicative authority.
The secretariat or appeal facilitator should publish the filing, timetable, evidence links and decisions. Original chairs retain the right to respond under the same limits as the appellant. They should not choose the subject line, summarise the appellant's case for the community or decide which related comments are admissible.
Ordinary conduct rules still apply. A separate moderator can enforce them, with review of restrictions that affect the case. This prevents an allegation of unfair moderation from being answered through more moderation by the same people.
After the case, chairs resume their normal role unless the remedy says otherwise. Temporary transfer of communication power is not a finding of wrongdoing. It is a structural protection that keeps the proceeding visibly fair.
Losing should not close the underlying policy question
An appeal may fail because chairs followed the process and reasonably assessed consensus. That outcome does not prove the proposal's merits are wrong or forbid a revised proposal. Review concerns the challenged action, not ownership of the policy agenda.
The decision should say what remains open. An appellant may return with new text, evidence or participation through the normal process. Chairs can manage repetitive submissions under ordinary rules, but should not cite a failed appeal as a permanent bar.
Similarly, a successful appeal does not vindicate the policy. It restores lawful deliberation. Both distinctions reduce the winner-takes-all pressure that makes appeals personal. Entities can accept procedural defeat while continuing substantive advocacy.
Archives should link subsequent proposals without implying they are relitigation misconduct. Governance improves when review clarifies procedure and ordinary participation remains available.
Informal mediation needs consent
Before formal review, a neutral mediator may help chairs and complainant identify misunderstanding or agree on correction. Mediation can preserve relationships in a small community, but it should be voluntary. A entity should not have to bargain away a public process concern in a private room.
The mediator must not later sit on the appeal panel or report confidential concessions as admissions. Any resolution affecting the public record should be published: correction, renewed discussion, recusal or withdrawal. Private details can remain protected, but the community needs to know why the institutional action changed.
Mediation pauses deadlines only by written agreement and for a defined period. Either party may end it and proceed to appeal without adverse inference. These limits keep a conciliatory tool from becoming another layer controlled by people more comfortable with insider negotiation.
Recusal decisions need their own neutral owner
A reviewer may believe their prior participation was minor; an appellant may see a close alliance with the chairs. Letting the challenged reviewer decide alone undermines confidence, while allowing any allegation to force recusal invites panel manipulation.
Disclosures should be made before substantive submissions. A designated ethics officer, panel members not affected by the challenge or an external facilitator decides contested recusals under a published standard. The decision and basic reasons are public, with employment or personal details minimised.
Substitutes should be named through a pre-existing order. Ad hoc replacement after the merits are known can be as consequential as the original conflict. If no qualified panel remains, the process should permit external appointment with clear limits.
Recusal is not a finding of misconduct. It protects the decision and the volunteer. A consistent mechanism makes stepping aside ordinary rather than reputationally explosive, which in turn encourages more honest disclosure.
Costs should not determine who can appeal
Most RIR policy appeals rely on volunteer submissions and do not charge a filing fee, an important feature of open participation. Costs still arise through legal advice, translation, time away from work and the effort of reconstructing an archive. A technically free process can remain practically inaccessible.
The evidence package, plain-language guidance and neutral procedural assistance should reduce those costs. Panels should resist unnecessary formality and repetitive filings. Where an oral session is useful, remote access should be the default and no party should gain extra time merely because it can retain counsel.
Cost awards or threats of liability would sharply chill review and should be reserved, if available at all, for clearly abusive conduct under a published standard. Losing a good-faith appeal is part of accountable governance, not a debt to the institution.
Conclusion: let reconsideration be honest about itself
Sending a complaint first to the same chairs can be sensible. They know the record, can correct simple errors and remain responsible for the process they lead. It becomes illegitimate only when the institution pretends that this return is independent review or allows it to consume the right to appeal.
The labels should tell the truth. The first stage is reconsideration. The second is appeal. Between them, authority changes hands, conflicts are disclosed, the original record is preserved and deadlines remain usable. The appellant does not need the chairs' agreement to leave their jurisdiction.
RIR policy depends on chair judgment because rough consensus cannot be reduced to arithmetic. That makes independent review a support for the model, not an insult to volunteers. Chairs who give reasons and accept correction gain legitimacy. Entities who lose an appeal can still see that someone other than the challenged decision-maker heard the claim.
An appeal that always returns to the same chairs is a conversation. It may be helpful, civil and even corrective. It is not enough. A trustworthy process knows when the conversation must end and review must begin.