Reasons Given After the Sanction

Summary

• Reasons must exist contemporaneously: The holder and the reviewer must know which facts, rules and risk assessments actually caused the sanction at the time it was imposed.
• Retrospective reconstruction harms review because it allows institutions to improve the case after seeing the appeal, operational impacts or the procedural situation.
• A decision record should separate rule violation, evidence, proportionality, rejected alternatives, customer impact and authority to act, with each reason dated and approved before the sanction.
• Later-discovered facts can support a new decision or continued measure, but should not be smuggled in to retroactively justify the original sanction.

The Timing of Reasons is the Timing of Accountability

Every sanction has a history after it is imposed. Staff can tidy the file. Lawyers can sharpen the theory. Executives can identify a broader institutional risk. Public criticism can show which argument sounds most legitimate. A later explanation can be more polished than the actual decision made. This very polish makes timing so important.

A registry sanction should be judged on the reasons that existed at the time of the decision. The holder needs those reasons to understand what was found, which rule was applied, and what could have changed the outcome. A reviewer needs them to test whether the institution acted within its powers. The community needs them to distinguish disciplined enforcement from retrospective justification.

This is not a demand for perfect prose before action. Operational institutions often decide under pressure. But even a brief decision can name the decisive facts, the rule applied, the harms feared, the alternatives rejected, and the official who approved the action. These elements create a contemporaneous record. Without them, review becomes a contest of memory and institutional storytelling.

The governance danger is subtle. Retrospective reasons can be true, in the sense that they identify real concerns. They may even be concerns that the registry could have relied on. The deficit is that the holder was sanctioned before those reasons were adopted as the basis of the decision. Review then tests a decision that was not actually made.

A Reason is Not a Slogan

Many sanction notices use institutional slogans: registry integrity, community protection, policy compliance, preserving trust. These phrases may describe values, but they are not reasons. A reason connects facts to authority and remedy. It says what happened, why the rule applies, why the evidence is reliable, and why this sanction rather than a milder measure follows.

For registry enforcement the distinction is crucial. A holder cannot respond to the value of registry integrity. But it can respond to an alleged forged document, a missing customer file, out-of-region use, an unpaid invoice, an unauthorised transfer, or a refusal to cooperate with an audit. A reviewer cannot verify an appeal to trust. It can verify whether the institution had evidence of a violation and whether the withdrawal was proportionate.

The notice should therefore avoid abstract conclusions that stand alone. It can begin with the institutional interest but must quickly move to the operational chain. The chain has five links: rule, fact, evidence, analysis, and consequence. If a link is missing, the holder remains in doubt. If analysis is missing, the sanction appears automatic.

A slogan can be useful in public communication after the decision. But it should not replace the decision record. Public readers may need a simple statement; the affected holder needs a verifiable one. The institution should not confuse these two audiences.

The Old Administrative Law Lesson Still Applies

Public administrative law systems have long recognised the problem of post-hoc rationalisation. In the United States, the Chenery principle is often summarised as the rule that an agency's action must be upheld on the grounds invoked by the agency at the time of its action. A private RIR is not a government agency in the same constitutional position, and the analogy must be used carefully. But the institutional lesson is transferable.

The lesson is not that every registry must import the doctrine of one jurisdiction. The lesson is that review loses integrity when the decision-maker can substitute reasons after a challenge. If the reviewer accepts new reasons as if they were the original reasons, the institution learns that weak initial decisions are safe. It can act quickly, wait for resistance, and then later build the best case.

Registry governance is particularly vulnerable to this problem, because the same institution often investigates, decides, maintains the file and defends the decision. If reasons are allowed to evolve without labelling, there is no clean way to know when staff moved from suspicion to finding, or when the board approved one theory over another. The sanction becomes a moving target.

The proper borrowing from administrative law is therefore modest and strong: write the reasons at the time of action, label later material as later material, and do not allow later material to retroactively cure an earlier lack of reasons. This discipline is compatible with private contracts, community governance and emergency measures.

The Decision Record Should Show the Path, Not Just the Destination

A final sanction letter often holds the destination: membership terminated, resources deregistered, transfer refused, services suspended. A reviewable decision record holds the path. It shows the evidence examined, the objections received, the alternatives evaluated, and the authority used. The path matters because two decisions with the same destination can be radically different in legitimacy.

For example, a registry may issue withdrawal after finding forged documents, missing credible customer dependency, and no available milder measure. It may also issue withdrawal out of staff frustration with a difficult holder, incomplete evidence, and a vague belief that the holder does not align with policy goals. The destination may be identical. The reasons are not. Review must see the difference.

The record should capture who made which judgment. Technical staff may identify a data anomaly. Compliance staff may decide that the anomaly breaches an agreement. Legal counsel may advise that termination is contractually possible. Senior management may decide that customer risk is acceptable. A board may approve the action. If all of this is collapsed into an institutional voice, accountability is hidden.

A path record also supports learning. If a court or reviewer later overturns the sanction, the institution can see whether the error was factual, legal, procedural or proportional. Without a path, an overturn becomes politics or embarrassment. With a path, it becomes maintenance of the governance system.

Reasons Must Address the Holder’s Response

A common sign of post-hoc justification is a decision that ignores the holder’s strongest response. The registry sends a notice, receives evidence or arguments, imposes a sanction, and later explains the decision with a theory that was not put to the holder. The holder’s actual response is treated as irrelevant because the institution has changed the question.

A contemporaneously reasoned decision should address material responses. It need not address every sentence. It should name the arguments that, if accepted, would have changed the outcome, and explain why they were rejected. If the holder submitted corporate documents, the decision should say whether they were accepted, limited public evidence, or contradicted. If the holder proposed a customer handover, the decision should say why it was limited public evidence.

This is not only fair to the holder. It protects the institution from confirmation bias. A staff team that must answer the strongest counterarguments before a sanction is more likely to spot errors. A board that sees unresolved objections before voting is less likely to approve a predetermined outcome. A reviewer can see that the dispute was actually joined.

If the institution later discovers a stronger reason, it should note that the new reason is new. It can open a supplementary decision process. What it should not do is pretend that the old response did not address a reason that was never communicated to the holder.

Proportionality Reasons Are Separate from Violation Reasons

A finding of a violation does not automatically justify the most severe sanction. A registry decision should separately reason why the chosen consequence is proportionate. This is particularly important in withdrawal, where the consequence can affect customers, routing trust, market value and regional confidence.

The violation reason answers whether the holder violated a rule. The proportionality reason answers what should happen because of that violation. The same violation can support a warning, a corrective order, a transfer block, a service restriction, staged termination or immediate withdrawal, depending on risk and history. If the decision record only holds the violation, later defenders can add proportionality after the fact.

Proportionality reasons should consider curability, prior notifications, holder conduct, customer exposure, urgency, evidentiary strength and milder measures. They should also explain why delay would not protect the same interests. A sentence that the violation is serious is not enough. Seriousness is an input, not the full analysis.

Separating the two reasons improves remedies. The holder can admit the violation but challenge the remedy. A reviewer can uphold the finding but modify the sanction. The institution can learn that its evidence rules are solid but its remedy ladder too steep. Without separation, every challenge becomes all-or-nothing, encouraging exaggeration on both sides.

Board Approval Should Not Paper Over Gaps in Staff Reasoning

In many registry systems, severe sanctions require top-level or board approval. This approval is valuable but does not cure a missing staff record unless the approving body receives and adopts adequate reasons. A board minute that says the action was approved after discussion may show authority; it may not show reasoning.

The approving body should see the proposed reasons before the action. It should know the holder’s principal response, the evidence still uncertain, the customer risk assessment, and the alternatives examined. If the body relies on legal advice, the decision record can note the nature of the advice without disclosing privileged details. If the body changes the reason, the final notice should reflect the adopted reason.

Otherwise, board involvement can create a shield. Staff can say the board decided. Directors can say they relied on staff. The holder sees only the sanction. The reviewer sees institutional opacity. Accountability dissolves into collective assent.

A better practice is an approval memorandum with explicit findings. The board or senior officer can adopt, reject or modify staff findings. The final decision then states the adopted reasons. This does not overlegalise governance. It makes clear which reasons belong to the decision-maker authorised to impose the consequence.

Emergency Reasons Can Be Short but Not Absent

Emergencies test the reasons rule. A registry may need to act before a full record can be compiled. That does not mean it acts without reasons. It means the initial reasons are shorter, focus on the immediate risk, and are followed by a prompt fuller explanation.

An emergency decision should note the triggering facts, the immediate harm, the authority for interim measures, why prior notice was not practicable, and when post-action review will occur. If the institution cannot name these elements, it may have urgency as a feeling rather than a record. The holder should receive the emergency reasons as soon as the protective measure is not frustrated.

The post-action fuller explanation should not rewrite the emergency. It should review it. Did the immediate harm exist? Was the measure narrower than withdrawal? Did new facts support continuation? Should the measure be lifted, narrowed, or converted into an ordinary sanction process? Each question looks forward from the emergency record, not backwards from the desired outcome.

Emergency decisions are legitimate when disciplined. The danger is not speed itself. The danger is a sanction-first, reasons-later culture, where emergency vocabulary becomes a shortcut on evidence. Short contemporaneous reasons are the antidote.

Later Facts Can Support a New Decision

A strict timing rule does not blind the registry to new information. If later investigations uncover forged documents, concealed transfers, customer harm or false statements, the institution may rely on those facts prospectively. It can open a supplementary notice, continue an interim measure, or impose a new sanction. The key is labelling.

The new decision should say what is new, when it was discovered, how it changes the analysis, and what response the holder can make. It should not be silently attached to the original reasoning. If the original sanction was deficient, the institution may need to cure the deficiency through a re-decision, not by pretending the original decision had the later record.

This distinction protects both sides. The holder cannot escape accountability merely because evidence arrived after the first letter. The registry cannot escape review of the first decision by improving it after a challenge. A reviewer can decide whether the original measure was justified at the time and whether a continued or renewed measure is justified today.

In practice, this may require two remedies: one for the past decision and one for future continuity. A reviewer may criticise the first record but permit a new process based on new facts. That outcome is more precise than declaring one side perfectly right. Timing discipline makes precision possible.

The Public Statement Should Not Precede the Decision Record

Registries under pressure often face public demands for explanations. Members, operators and the media want to know why severe action was taken. Public communication is legitimate, but it should not become a parallel reasoning channel that differs from the notice to the holder. If the press release contains the real reasons and the decision letter does not, the process has failed.

The public statement should align with the decision record, handling confidential details carefully. It may describe categories of violations, procedural steps and continuity protections. It should avoid new allegations that were not in the notice, particularly if the holder had no opportunity to respond. Public legitimacy does not require trial by press release.

This restraint also protects the registry. Public exaggeration can create defamation risk, harden positions, and make later corrections look like retreat. A measured statement that refers to the decision process is less dramatic but more durable. It tells the community that the institution acted on a record, not on narrative momentum.

If public communication reveals a reason that is not in the decision record, the institution should treat that as a deficiency to be corrected. Either the reason was not decisive and should not be public, or it was decisive and should have been in the notice. There is no healthy third category where decisive reasons first appear in public communication after the sanction.

Review Bodies Should Reject Reconstructed Clarity

The reasons rule ultimately depends on reviewers. If an appeals body, an independent panel or a court accepts every post-sanction improvement, institutions will adapt. They will learn that the initial notice can be thin because the review process gives time to perfect the reasoning. The cost of a poor initial reasoning then falls on the holder.

A reviewer should ask three timing questions. What reasons were communicated before or at the sanction? What reasons appear for the first time in the review? Which later reasons are truly new facts and which are improved arguments on old facts? The remedy should follow the answers.

If the original reasons were sufficient, the sanction can be measured against them. If later facts justify a prospective measure, the reviewer can allow a new decision path. If the institution has only reconstructed a better theory from pre-existing material, the reviewer should not treat that theory as the original basis. It can remit, suspend, narrow or require a new decision.

This discipline does not make enforcement impossible. It makes enforcement readable. A registry that records its reasons will pass the test. One that acts first and writes later may still have serious concerns, but must bring those concerns through a proper decision. Review is not a writing workshop for already-imposed sanctions.

The Record Should Outlast the Dispute

The ultimate audience for contemporaneous reasons is not only the holder or the reviewer. It is the future institution. Years later, staff will need to know why a resource was withdrawn, why a record was changed, why a customer handover was allowed, or why a court criticised the process. A reconstructed narrative written after litigation may win a moment, but it degrades institutional memory.

A durable record helps compare cases. Similar violations should receive similar analysis, unless facts differ. Different treatment should be explainable. Without contemporaneous reasons, every case becomes exceptional because no one can see the actual basis for past action. This invites allegations of bias and makes policy learning difficult.

The record also teaches the community what enforcement means. When reasons are clear, members can adjust behaviour. When reasons appear only after the sanction, members learn uncertainty. They may overcomply, litigate early, or distrust routine notices. The cost of poor reasons is then borne by the entire system.

The registry’s authority is strongest when it can say: This is what we knew, this is what we decided, this is why we acted, and this is what later review changed. That sentence requires reasons at the time of decision. Anything else asks the community to trust not only the institution’s power but also its power to rewrite the explanation after using it.

Versioned Reasons Prevent Silent Substitution

A practical way to control post-hoc justification is versioning. Every notice, decision memorandum, board approval, emergency order and appeal response should have a date, an author or approving body, and a status. If reasons change, the change should be visible. The record should show whether a paragraph was part of the original decision or added after a challenge.

Versioning is an ordinary document discipline but has governance power. It prevents a later reader from confusing a procedural explanation with a decision reason. It lets the holder see what it could respond to. It lets a reviewer separate contemporaneous reasons from supplementary reasons. It also protects staff from allegations that they invented reasons, when the version history shows an honest evolution.

The institution can still correct errors. If the first notice cited the wrong policy section but facts and analysis were clear, a corrected version can note that. If a new investigation adds a reason, the supplementary version can open a response period on that reason. What versioning prevents is silent substitution: old reason out, new reason in, with no acknowledgement that the review situation has changed.

For high-stakes sanctions, version control should be part of the case file from the start. The cost is low. The benefit is a record that can survive appeal, judicial review and later institutional memory, without anyone having to trust a reconstructed chronology.

The Appeal Record Should Not be Larger than the Decision Record Without Explanation

An appeal will naturally add additional material. The holder submits arguments, the registry responds, experts explain, and reviewers ask questions. The problem arises when the registry’s appeal record contains decisive factual and analytical material that was available before the sanction but missing from the decision record. This discrepancy requires explanation.

A reviewer should ask why the material was not in the original record. Was it actually discovered later? Was it privileged advice now summarised? Was it an internal assumption not written down? Was the sanction imposed before staff completed analysis? Each answer has different consequences. Some may justify supplementation. Others may show that the original decision was premature.

This rule encourages institutions to assemble the record before acting. It does not require that every conceivable argument be written up at length. It requires that the decisive reasons appear. If the appeal record becomes the first full decision record, then appeal has been converted into first-instance decision-making. That deprives the holder of a genuine internal decision to challenge.

The size comparison is not mechanical. A complex appeal may be longer because the holder raises many points. The warning signal is not length itself. It is the new decisiveness: the point that now carries the institution’s reasoning was missing when the holder was sanctioned.

Secret Decisive Memoranda are Incompatible with Meaningful Review

Institutions sometimes rely on internal memoranda that are not disclosed because they contain legal advice, security details, or frank staff discussion. Confidentiality can be legitimate. But if a memorandum contains the decisive factual or policy reason for the sanction, the holder and the reviewer need a usable version of it. A secret decisive memorandum makes the public decision hollow.

The registry can protect sensitive material through redaction, summary, privilege logs, or reviewer-only access. It can separate legal advice from facts. It can disclose the conclusion and the non-sensitive basis without revealing strategy. What it should not do is issue a thin letter while the real reason remains hidden, and then ask a reviewer to confirm the sanction on faith.

Secret decisive memoranda also distort internal governance. A board may approve a sanction based on an oral summary of a memorandum that staff later characterise differently. Legal counsel may argue that the memorandum supports one theory while operational staff understood another. Disclosure discipline forces the institution to decide what its reason actually is.

Meaningful review requires that the reason be available in a reviewable form. If the reason cannot be disclosed at all, the institution should consider whether the sanction can be fairly proceded, whether a narrower interim measure is possible, or whether an independent reviewer can test the confidential basis before harm becomes irreversible.

The Remedy for Inadequate Reasons Should Match the Defect

A finding of missing reasons does not always require a permanent defeat for the institution. Remedies should match the defect. If the problem is unclear wording but the contemporaneous record is sufficient, a clarification may suffice. If the decision omitted the proportionality analysis, the reviewer can remit for a new decision on remedy while preserving the violation finding. If the original record contains no decisive reasons, the sanction may need to be overturned or re-decided.

This remedy flexibility is important because registry disputes affect continuity. A reviewer should avoid both extremes: rubber-stamping reconstructed reasons and automatically striking down every enforcement action for a wording defect. The goal is to restore accountable decision-making. Sometimes that means a new notice, sometimes a narrower measure, sometimes immediate reversal.

The reviewer should also consider interim protection during the re-decision. If the registry’s concerns are serious but the reasons were inadequate, a temporary transfer block or a holding order can protect the system while proper reasons are issued. If the sanction itself is causing disproportionate harm and urgency was not shown, the measure should be suspended.

Matching the remedy to the defect encourages better practice. Registries learn that inadequate reasons have consequences. Holders learn that a procedural victory does not eliminate proven risk. The review system becomes corrective rather than theatrical.

Internal Disagreements Should Be Recorded, Not Erased

High-stakes decisions often contain internal disagreements. Technical staff may assess a risk lower than legal counsel. Customer teams may warn about transition damage. Board members may ask for a narrower measure. Compliance staff may disagree on curability. Erasing these disagreements from the record can make the final decision look cleaner, but less truthful.

The decision record does not have to publish every internal debate. It should record material dissents or unresolved uncertainties that bear on the sanction. If a team warned of high customer harm, the final reasons should explain why the institution proceeded. If staff disagreed on evidentiary strength, the final decision should name the basis for choosing one view. If directors imposed conditions, those conditions should appear.

Recording dissents protects decision-makers. It shows that alternatives were considered. It prevents later claims that no one raised an obvious risk. It also helps reviewers understand the judgment under challenge. A reasoned decision can acknowledge uncertainty and still act. What undermines confidence is false certainty.

In technical governance, dissent is often a sign of expertise, not disloyalty. A registry that preserves material disagreements in the decision path shows maturity. It tells the community that sanctions are not manufactured consensus but accountable judgments made in the face of known trade-offs.

Contemporaneous Reasons are the Cheapest Accountability Tool

Independent review, litigation, and public campaigns are expensive. Contemporaneous reasons are comparatively cheap. They require the institution to write down the basis for its action before or at the time of action. This modest discipline can prevent months of dispute over what the decision was.

The cost of writing reasons is also a test of readiness. If the institution cannot explain the sanction before it is imposed, the case may not be ready, except for a narrow emergency measure. If it can explain the sanction, the explanation will help the holder, the reviewer, and the community understand the decision, even if they disagree.

Reasons do not have to be literary. They must be specific, dated, and tied to evidence. They should show violation, authority, remedy, and rejected alternatives. They should label later material as later. They should allow a reviewer to say whether the institution acted on the reason it now defends.

That is why post-sanction reasons are not harmless housekeeping. They can change the subject of review. They can transform a weak decision into a stronger story without giving the holder the process it was due before the consequence. A registry that values lasting authority should prefer the less glamorous path: make the decision record before using the decision power.

Reasons Disciplining Institutional Emotion

Sanctions often arise after frustration. Staff may have chased a holder for months. Communication may have become confrontational. Public criticism may have mounted. The holder may have behaved badly. Institutional emotion is understandable, but it is not a reason. A written decision forces the institution to translate frustration into rule, fact, evidence, and remedy.

This translation is valuable even when the holder deserves a sanction. It prevents the record from becoming a catalogue of irritation. It asks which facts actually matter and which merely explain why trust broke down. It also helps senior decision-makers distinguish a difficult relationship from a legally or contractually sufficient reason.

A holder’s tone, delay, or public statements can be relevant if they affect cooperation, risk, or credibility. They should be identified as such. They should not become a substitute for proof of the underlying violation. Reasons create this boundary. They let the institution say with discipline that the sanction follows from this failure, not from the accumulated inconvenience of the dispute.

In a community system, emotional discipline is a public good. Members may disagree with the outcome but can see that the institution acted on reasons, not retaliation. That perception is often as important as the immediate case.

The Quality of Reasons Should Be Reviewed Before Publication

Before a severe sanction is sent, someone not involved in the investigation should review the reasons for quality. The reviewer could be a senior staff member, a legal advisor outside the enforcement team, or a governance officer. The task is not to rewrite the outcome. It is to ask whether the decision can be understood, responded to, and reviewed.

The quality review should ask whether the rule is cited, facts are dated, evidence is identified, holder responses are addressed, the remedy is justified, customer impacts are considered, and the appeals route is clear. It should also ask whether a public statement says more than the decision record. If the answer is yes, the record should be corrected before the measure.

This internal check is not an independent appeal. It is pre-decision hygiene. It catches the avoidable errors that later make the institution look arbitrary. It also protects reviewers from receiving a deficient record after harm has occurred.

The check should be briefly documented. A note that reason quality was reviewed, with any changes made, helps later accountability. It shows that the institution treated the explanation as part of the sanction, not as post-sanction public relations.

A Decision That Cannot Be Explained May Need a Narrower Remedy

Sometimes an institution senses a real risk but cannot yet explain enough to impose the strongest sanction. The answer should not be to impose the sanction first and write later. The answer can be a narrower interim remedy: secure records, freeze transfers, require contact updates, maintain customer continuity, or seek independent review of confidential evidence.

Narrower remedies create time without pretending the final case is ready. They can protect the registry from asset flight or harm while preserving the holder’s ability to respond. They also create a record of restraint. If the case later supports withdrawal, the institution can show it did not rush from suspicion to maximum consequence.

This is particularly important when evidence is complex or partly confidential. The inability to explain everything publicly may be real. But the more explanation is withheld, the more cautious the remedy should be, unless immediate harm is shown. Secrecy and severity are a dangerous combination.

The rule is simple: if the reasons record cannot carry the remedy, choose a remedy the record can carry. Then build the record. The registry’s power is not weakened by waiting until reasons and remedy align. It is strengthened because the final decision is harder to attack as reconstructed.

The Discipline Should Be Routine Before It Is Tested

The hardest time to build a reasoning practice is during a high-profile dispute. Every word is strategic. Every omission is suspected. Every correction looks like retreat. The discipline should be routine before it is tested. Routine rejections, minor service restrictions, and ordinary compliance findings should all use the same basic structure: rule, fact, evidence, response, analysis, remedy, and appeals path.

Routine use builds muscle memory. Staff learn to state facts without exaggeration. Managers learn to ask about proportionality before approving escalation. Holders learn what a serious decision looks like. Reviewers, when they later become involved, see familiar records instead of ad hoc inventions. The institution does not have to design accountability under fire because it already practises it in smaller cases.

This routine discipline also reveals policy gaps early. If staff cannot explain why one defect gets a warning and another gets a suspension, the remedy ladder may need revision. If decisions repeatedly rely on confidential evidence, the disclosure protocol may need overhaul. If holder responses are rarely addressed, the notice process may need revision. The act of writing reasons exposes weaknesses that slogans hide.

For severe sanctions, the same structure can be expanded. The decision may require more evidence, top-level approval, customer risk analysis, and legal review, but should still be recognisable as the same institutional form. That consistency matters. A holder should not face a completely different process only when the stakes become existential.

Reasons at the time of decision are therefore a habit, not a heroic act. A registry that waits until litigation to become reasoned will sound reconstructed even when sincere. A registry that reasons every day can show its work when the hard case comes. The community does not have to agree with every decision to trust the pattern. It can see that power flows through a record before it flows through a sanction.

A Reasoned Loss Is More Durable Than a Silent Win

A registry can win a dispute by acting quickly and giving little explanation, if the holder lacks the means to challenge. That is a fragile win. Other holders will see the silence. Members will wonder if the same treatment could befall them. Staff will learn that thin records are acceptable. The institution has preserved an outcome while weakening the pattern that makes future outcomes acceptable.

A reasoned loss is often more durable. If a reviewer says the registry had a real concern but chose the wrong remedy or failed to answer a material objection, the institution can repair the process and act again if needed. The community sees correction not impunity. Staff learn the boundary. The holder receives a decision it can understand, even if it remains dissatisfied.

That is why contemporaneous reasons are not just defensive. They are how an institution turns power into precedent. Every severe sanction teaches the community what future enforcement will look like. A silent win teaches fear. A reasoned decision, even if partly overturned, teaches standards.

The Final Record Should Indicate Future Use

A final decision should also say how it may be used later. If it is a precedent for future sanctions, the institution should say which principle carries forward. If it is case-specific, it should say that. This prevents later staff from converting a narrow sanction into a broad rule without public deliberation.