Summary
- The Number Resource Society needs stable core revenue for the authoritative number ledger, authentication, security, dispute handling, interoperable registration data and continuity. Automatic sunsetting should apply to separately identified non-core charges, not to the institution's minimum ability to keep the registry safe and available.
- A review date is limited public evidence. A true sunset makes the charge legally and operationally expire unless operators affirmatively renew it after receiving a comparable cost statement, performance record, incidence analysis, alternatives assessment and wind-down plan.
- The boundary must be decided before money is raised. Work becomes non-core when the Society could stop it without impairing unique registration, authorized record changes, essential RPKI service, minimum public data, dispute review or tested succession. Novelty alone does not make an activity peripheral, and familiarity does not make it essential.
- Renewal authority should belong to the operators who bear the charge, under a franchise that cannot be enlarged or narrowed for the renewal itself. Board sponsorship, consultation and budget approval are useful, but they are not substitutes for payer approval where the programme is contestable.
- Comparative systems contribute mechanisms rather than ready-made law: RIPE NCC members adopt an annual charging scheme, APNIC publishes a resource-linked fee schedule under board authority, and the Australian Government's cost-recovery framework requires activity-specific cost documentation and ongoing explanation of revenue and expense variance.
- Expiry must be credible. Staff commitments, contracts, reserves, grants and software should be structured so the programme can stop without threatening the core ledger. A charge that can never end because the institution has made termination dangerous is not genuinely temporary.
- The rule should prohibit evasion through relabelling, bundling or migration into the core fee. A materially similar activity remains subject to the original sunset until a valid constitutional decision changes its status, and unspent balances return to payers or reduce the next core assessment.
The fee is an institutional mandate expressed in money
Fees are often presented as arithmetic. The board estimates expenditure, allocates a prudent reserve, selects a charging unit and sends invoices. Yet every line also distributes authority. A permanent charge allows an activity to recruit staff, sign contracts, build constituencies and define success. Once those commitments exist, ending the activity appears disruptive even when the original justification has weakened.
This is particularly important for a number-resource institution. Operators cannot replace the recognized registry relationship as easily as they can change an ordinary supplier. A network may disagree with a programme while still needing accurate records, reverse delegation, authenticated account access, transfer completion and route-security credentials. The institution's positional dependence gives a fee decision more force than a normal club subscription.
That force does not make every fee illegitimate. Shared registry services have large fixed costs and collective benefits. Security engineering protects operators who never suffer a visible incident. Continuity reserves matter precisely when ordinary revenues or systems fail. A functioning appeal body may hear few cases yet discipline every first-instance decision. Charging every function transaction by transaction would underfund readiness and create bad incentives.
The constitutional issue arises when essential dependence finances optional expansion. If the same unavoidable invoice funds a broad advocacy programme, general technical development, sponsorship portfolio or speculative service, operators must either subsidize the activity or endanger access to the core relationship. Formal membership approval may soften that concern, but only if payers can see the contested amount, vote on it separately and make non-renewal operationally possible.
An automatic sunset gives that choice a structure. It says that the institution may experiment and respond to new needs, but a non-core claim on all operators does not become permanent merely because it survived one budget year.
Core funding and non-core funding require different defaults
The Society's minimum mission should not depend on a recurring vote that could fail for tactical reasons. Core funding should cover the coherent current record of IPv4, IPv6 and autonomous system number holdings; secure authentication; authorized changes and transfers; public registration data at the required level; essential route-origin authorization functions where offered; correction and review; incident response; audit trails; and tested continuity to a successor.
These functions can still be audited, budgeted and challenged. Core does not mean unconstrained. It means that complete expiry would create a direct risk to uniqueness, evidence, security or service continuity. Their charging basis should therefore persist under the constitution until replaced by another valid core scheme, with enough reserves to prevent an election or budget dispute from interrupting operations.
Non-core work is different. It includes an activity the Society could discontinue without losing the minimum service. Depending on design, that may include broad public-policy advocacy, grants unrelated to registry reliability, media production, sponsorship, general digital-development programmes, commercial research products, conference expansion, investment promotion or a service available from competitive suppliers without compromising the registry record.
The boundary is functional, not rhetorical. Training on secure account use may be core if poor practice creates immediate registry risk. A long-running conference may be non-core even if it has become culturally important. RPKI engineering can be core because credential continuity and accurate authorization state directly support the recognized resource record. A general campaign about Internet values may remain peripheral even when its goals are admirable.
The constitution should state the core classes narrowly and require a higher-threshold operator decision to add a permanent class. Everything else begins as non-core. That order prevents the board from avoiding sunset simply by calling a desired programme essential.
Review is not expiry
Many institutions promise to review fees annually. A draft budget is published, comments are invited and the board approves a new total. That is useful transparency, but it does not reverse the presumption of continuation. If silence leaves the charge in force, the programme owns the status quo.
A true sunset changes legal consequence. The resolution authorizing a non-core fee states an end date. After that date, the Society lacks authority to invoice, collect or commit new expenditure from that charge unless a renewal resolution has passed. A late report does not extend it. A board statement that the activity remains valuable does not extend it. Inclusion in a draft budget does not extend it.
The distinction changes behaviour before the vote. Management has a reason to preserve cost records and define measurable outcomes because continuation will need them. Contracts must contain termination or scaling clauses. Reserves cannot be treated as an indefinite endowment. Operators know when evidence and participation will matter. Opponents do not need to assemble a majority to repeal an entrenched charge; proponents must earn affirmative renewal.
Automatic expiry is not a judgment that the activity failed. It is a rule about who bears uncertainty. For the minimum ledger, uncertainty should normally favor continuity. For a peripheral programme funded through a dependent relationship, uncertainty should favor release of the payer unless current evidence supports another term.
A scheduled discussion may accompany sunset, but it cannot replace it. The authorizing instrument should identify the exact moment authority ends and the limited actions permitted afterwards: settling accrued liabilities, preserving records, completing an orderly wind-down and returning the balance. Those powers should not include launching new work under the expired mandate.
Existing RIR models show different allocations of fee authority
Current regional registries do not use one constitutional model. The variation is instructive because it separates three questions that are often collapsed: who proposes the fee, who approves it, and who controls the expenditure it supports.
The RIPE NCC publishes a specific charging-scheme approval procedure. Management prepares a draft, the Executive Board considers it, and the General Meeting adopts the scheme. Members receive the proposed text before the meeting. The adopted scheme applies to its financial year and continues until a replacement is adopted. The RIPE NCC Articles of Association separately place adoption of the coming year's charging scheme on the General Meeting agenda while the Board manages the association and adopts the budget and activity plan.
This design gives members a direct decision over the charging instrument, but it is not an automatic sunset. Continuation until replacement protects revenue continuity. It can also make the old scheme the fallback if members reject a new one. NRS can borrow annual member decision while changing the default only for separately identified peripheral charges.
APNIC allocates authority differently. Its current member fee schedule links annual and transaction charges to resource holdings and service categories. APNIC's explanation of its Executive Council states that the Council develops the activity plan and budget, receives member feedback and has responsibility for final budget adoption. This model illustrates board-led financial administration under published rules, not direct annual payer approval of every programme.
The comparison shows that transparency and voting can exist at different points. NRS should make the contested non-core charge itself the unit of approval rather than asking one aggregate budget vote to carry every institutional choice.
Annual charging votes are useful but can still bundle missions
An annual vote creates rhythm. It forces the institution to publish a scheme, explain expected income and expose changes before invoices arrive. It is better than a standing power to set any fee at any time. Yet a single annual scheme can still combine indispensable registry revenue with contested expenditure.
Suppose operators are offered one resolution funding record maintenance, cyber defence, a continuity reserve, a policy fellowship, broad advocacy and an international event. Rejection could leave the Society without enough income to protect credentials. Approval can then be described as endorsement of every component. Neither inference is sound. Operators may approve because the ledger cannot be put at risk, not because each peripheral programme has a mandate.
Automatic sunset addresses this by requiring separability. The core scheme can remain the stable base. Each material non-core programme receives a named charge or a clearly quantified component, its own end date and a renewal resolution that can pass or fail without disabling the ledger. The invoice may collect amounts together for administrative efficiency, but the legal authority and accounting should remain distinct.
Materiality should prevent fragmentation into trivial votes. The constitution can set a low percentage of annual core expenditure or a fixed inflation-adjusted amount below which incidental work is absorbed and disclosed. Several related activities that form one programme should be assessed together. The board should not divide a large expansion into small pieces to stay below the threshold.
Conversely, a board should not combine unrelated temporary charges into a single renewal package. Bundling allows a popular security pilot to shelter an unrelated communications campaign. The voting paper should let operators accept the former and reject the latter.
Cost recovery disciplines the entity of a charge
Public cost-recovery systems cannot be transplanted directly into a private transnational association. They operate under legislation, ministerial responsibility and public finance controls. Their documentary logic is nevertheless useful.
The Australian Government's Cost Recovery Policy treats cost recovery as charging non-government recipients for some or all of the efficient cost of a specific activity. It requires a cost recovery implementation statement before charging begins, ongoing reporting, stakeholder engagement and explanation of material variance between revenue and expense. The framework distinguishes a charge for a specific service from a levy funding activity for a group.
The most useful element for NRS is specificity. A charge should have an identifiable activity, payer group, authority, cost model and reporting record. It should not be a free-floating amount justified by the overall usefulness of the institution. The continuing statement becomes a common entity for operators, management and auditors: this is the service being funded, these are its efficient costs, this is the demand forecast, and this is what changed.
Private registry economics require adaptation. NRS may legitimately pool costs across operators, hold reserves and subsidize participation where collective registry quality improves. It need not mimic government accounting or claim that every payer receives an identical benefit. But it should be able to show why a non-core activity belongs to the charged group and why the chosen amount corresponds to an efficient, bounded programme rather than institutional appetite.
At renewal, the original cost statement should be updated rather than replaced by a new narrative. Variance is evidence. Persistent underspend, repeated overcollection or cost growth without commensurate outcomes should change the renewal decision.
Every temporary charge needs a fee identity card
Before the first invoice, NRS should publish a compact constitutional record for the charge. It should identify the activity, legal authority, payer class, calculation unit, expected gross revenue, expected cost, start date, sunset date, performance measures, accountable executive, independent reviewer, renewal electorate and expiry disposition.
The record should also state why the activity is non-core. That may seem counterintuitive: institutions usually explain importance, not dispensability. Yet the point of temporary authority is to preserve a clear answer about what happens if the programme ends. The statement should identify which core services remain unaffected and which benefits will cease.
The payer class must be precise. All members, active resource holders, transfer applicants, certification users and event entities are not interchangeable constituencies. A fee imposed only on transfer recipients should not be renewed by members who never pay it without a second test of payer consent. A general levy may use the broad operator electorate if the benefit is genuinely collective and incidence is published.
The calculation unit should avoid hidden claims about the asset. A charge based on IPv4 holdings can distribute cost by size, but it can also look like rent on scarcity. The identity card should say whether the metric approximates service burden, ability to pay, risk contribution or distributive policy. It should not imply that NRS owns the economic value of the number resource.
Finally, the record should contain a unique public name and version history so relabelling is visible. If the same purpose, payer class and benefit survive under a new title, the sunset follows the substance. Institutional memory begins with the ability to compare what was promised with what was delivered.
Cost evidence must show more than expenditure
An institution can always prove that it spent money. Renewal asks whether the amount was efficiently required for the approved purpose. The dossier should therefore separate direct cost, shared overhead, capital expenditure, reserve contribution, procurement commitments and wind-down liability.
Direct cost includes staff time, contracted services and assets used for the programme. Shared overhead should use a disclosed allocation rule. Assigning a large fraction of executive, office and security cost to a small programme can make it appear indispensable; assigning none can make it appear artificially cheap. The rule should be consistent across programmes and tested by the auditor.
The dossier should compare budget, actual cost and forecast for the next term. It should explain demand errors, currency effects, procurement delay, vacancies and scope change. If revenue exceeded cost, operators should see the balance and proposed disposition. If cost exceeded revenue, they should see whether the deficit was caused by temporary implementation or a structurally weak model.
Cost per output is helpful but not sufficient. A fellowship can report cost per entity; an advocacy programme can report cost per submission; a measurement service can report cost per observation. The renewal case must also show outcome and connection to the payer group. More output may represent busy expansion rather than value.
An independent auditor need not decide policy. Its role is to attest that the figures reconcile to accounts, allocation methods were applied consistently, liabilities are complete and management did not move costs between core and non-core lines to influence the vote. Operators then decide whether the evidenced activity deserves another compulsory term.
Incidence evidence reveals who really carries the fee
The named payer is not always the economic bearer. A large network may absorb a flat fee easily, while a small operator passes it to customers or leaves the market. A transfer fee may be paid by the recipient but reflected in the negotiated price. A resource-based charge may alter decisions about account consolidation, returns or corporate structure.
Renewal should therefore include an incidence report. At minimum it should show payments by operator size, jurisdiction group, resource class and corporate group; fee as a proportion of the operator's total NRS invoice; exemptions; arrears; waivers; closures plausibly associated with the charge; and concentration among the largest payers. Privacy-protective aggregation can prevent disclosure of individual commercial data.
The report should examine behavioural effects. Did a transaction charge reduce frivolous requests or deter legitimate transfers? Did a fixed programme levy weigh disproportionately on very small accounts? Did a discount improve participation from underserved economies? Did a resource-based component encourage efficient returns, or merely price a scarcity value unrelated to service cost?
Incidence matters constitutionally because votes can conceal distribution. One organization, one vote gives equal formal voice while contributions differ. Resource-weighted votes can align payment and power but allow large holders to dominate. Neither rule is automatically correct. The renewal dossier should expose the mismatch rather than pretend that the voting formula resolves it.
Where a small class supplies most revenue, a dual threshold may be justified: majority approval by eligible operators plus a majority of represented fee value, with a cap preventing one corporate group from controlling the second count. The purpose is not to sell votes. It is to prevent either a numerous low-paying class or a concentrated high-paying class from unilaterally imposing the burden on the other.
Benefit must be traced without pretending it is individually consumed
Collective services often benefit operators indirectly. A secure registry reduces systemic risk even for a member that submits no request. Policy participation can improve rules that a quiet operator later relies on. Training can reduce errors by counterparties. The renewal test should not demand a retail receipt for every payer.
It should demand a benefit route. The programme must identify the condition it seeks to change, the activity it performs, the observable result and the reason that result belongs to the charged group. For example, a targeted training programme might reduce authentication failures among account administrators, lowering fraud and support cost. That chain is closer to the core mission than general technology education with no measured registry effect.
Benefits can be distributed unequally if the rationale is explicit. Subsidizing participation from smaller economies may improve regional evidence and policy legitimacy for everyone. A route-security support programme may initially assist less capable operators while increasing the coverage and reliability of the shared system. The institution should state the redistribution rather than disguise it as identical service.
The weakest claim is reputational uplift for the Society itself. Visibility can support trust and recruitment, but almost any expansion can be justified as enhancing reputation. Renewal should require evidence connecting communications or sponsorship to a defined registry outcome. General prominence is not a sufficient compulsory benefit.
Operators should also see alternatives: voluntary subscription, restricted donation, competitive procurement, partnership, beneficiary payment or termination. A programme may be valuable yet unsuitable for universal fee funding. Sunset creates the moment to distinguish those conclusions.
Operator approval must be an actual power to refuse
Consultation is not approval. A board can receive comments, publish responses and proceed unchanged. For a compulsory non-core charge, the operators who bear the relationship should have a decision whose negative result ends authority.
The electorate should be fixed before the renewal campaign begins. Eligibility can depend on recognized operator status, good standing and a representative designated by a record date. It should not depend on support for the programme, attendance at a meeting or a discretionary membership conversion controlled by the board. Corporate groups should not multiply votes by creating nominal accounts, and customers hidden behind sponsoring operators need a stated representation route.
Approval should be charge-specific and informed. The full renewal dossier should be available far enough in advance for independent analysis. Management may advocate continuation, but it should publish a neutral summary and provide equal access to material facts. The board members sponsoring the charge should disclose relevant programme relationships and procurement interests.
A simple majority of votes cast may be too weak when turnout is very low. A denominator safeguard can require participation by a minimum share of the eligible electorate. A supermajority of all eligible operators may be too rigid and make change impossible. A balanced rule could require a majority of votes cast, a modest participation floor and a majority across more than one operator-size band.
Whatever rule is chosen must be knowable at authorization and unchanged during the term. The board cannot alter voter eligibility to make renewal easier. The separate problem of who controls the electorate is itself constitutional and should be protected by a higher amendment threshold and delayed effect.
Renewal must be separate from the core budget
The most important ballot design is an unbundled choice. Operators should first receive the core budget and fee required to keep the minimum service safe. Each expiring charge should then appear as a separate resolution with amount, term and scope.
This prevents continuity pressure from deciding the policy question. A member should be able to vote against an advocacy levy without voting against account security. The board should not announce that rejection of one peripheral fee will force indiscriminate cuts to the registry. If the programme has been honestly separated, its termination plan should show otherwise.
Shared overhead complicates separation but does not defeat it. The non-core programme can pay a proportionate contribution to finance, security and facilities while active. On expiry, the core budget may retain genuinely fixed overhead that cannot be removed immediately. The renewal dossier should identify that stranded amount and the period over which it will disappear.
Nor should a fee be bundled with a popular discount. A resolution that continues a peripheral programme while reducing another charge can buy support without establishing the programme's value. Redistribution decisions should be voted independently or clearly shown as inseparable parts of one coherent charging model.
The minutes should record results against the full electorate, not only percentages among valid ballots. Operators need to know whether renewal reflects broad assent or a narrow active minority. A valid low-turnout vote may still authorize the term under the adopted rule, but institutional claims should describe its mandate accurately.
The term should match the uncertainty and commitment
One sunset period will not fit every activity. A six-month pilot may be enough to test a narrowly scoped service. A training or measurement programme may need two years to show behavioural change. A capital-intensive shared facility may require a longer horizon to avoid waste, but the fee should not outlive the useful comparison period merely because management prefers certainty.
Three factors should determine duration. First is evidence latency: how long until the claimed outcome can be measured? Second is reversibility: how much harm or stranded cost follows termination? Third is institutional risk: how much authority, money and constituency can accumulate during the term?
High-risk programmes should have shorter decision intervals even when the underlying project is long. Operators can authorize a multi-year objective while approving annual fee tranches against milestones. Low-cost work with slow outcomes may receive a longer term if commitments remain cancellable and reporting is strong.
The clause should include an outer limit for repeated renewal without constitutional reconsideration. After, for example, two renewals, the institution must either end the activity, convert it to voluntary finance or seek a higher-threshold decision adding it to the permanent mission. Endless temporary renewal can be as entrenched as a standing fee.
Conversion to core should be exceptional. Longevity, staff expertise and repeated approval do not prove indispensability. The proposal must show that ending the activity would now directly impair unique registration, evidence, security, review or continuity, and it should explain whether the institution itself created that dependence.
Expiry must have a defined financial effect
A sunset is meaningless if authority to spend ends but authority to collect continues. The instrument should close both. No invoice for a period after expiry may include the charge, and no new programme commitment may be entered after the cutoff.
Accrued liabilities still need payment. Staff notice, vendor termination, final audit and data preservation can continue under a limited wind-down appropriation established at the start. That amount should be held within the programme's own balance, not taken unexpectedly from core continuity reserves.
Unspent funds should follow a precommitted rule. Small balances can reduce the next core invoice proportionately. Material balances can be returned or credited to the original payer class. Donation to another programme should require fresh consent because it changes the purpose for which the money was collected.
Assets purchased by the programme should be inventoried. Equipment or software genuinely useful to core operations can transfer at independently assessed value, preventing the programme from inflating its apparent cost or gifting a hidden subsidy to the core. Other assets can be sold, licensed or retired under published rules.
Records should survive. Cost statements, contracts, performance data, votes, dissent and closure findings form the evidence for later proposals. Expiry ends authority, not accountability. The final report should compare original forecasts with actual results and state which obligations remain.
A clean financial effect also protects future experimentation. Operators may be more willing to authorize a pilot if they know that rejection later will produce a real end rather than a semantic rebudgeting.
Contracts and staffing must not make non-renewal impossible
Institutional entrenchment often occurs through commitments rather than rules. A board authorizes a two-year fee but signs a five-year lease, hires permanent staff without redeployment planning and promises grants beyond the sunset. At renewal, management argues that cancellation would waste money or mistreat employees. The formal sunset remains, but the economic decision has already been made.
The authorization should cap commitments at the term plus the approved wind-down. Contracts need termination rights aligned to expiry. Capital purchases should have residual value or a transfer plan. Grant agreements should make future installments contingent on renewal. Staff should be hired on terms consistent with applicable law and honest workforce treatment, with advance plans for reassignment where skills support the core.
This does not require precarious employment or inefficient month-to-month procurement. It requires the board to price commitment honestly. If a programme cannot operate responsibly without a five-year contract, operators should authorize a five-year exposure or reject it. A nominal annual sunset should not conceal longer liability.
The independent cost statement should list commitments by termination date and maximum exit cost. Renewal materials should show which obligations resulted from decisions made after the previous vote. Unexpected lock-in is a governance failure even when procurement was technically lawful.
Boards also need a duty not to sabotage expiry. They should not delay wind-down preparation in the expectation that operators will renew rather than face disorder. Readiness to end is part of programme management, just as continuity is part of core registry management.
Reserves must protect the ledger, not perpetuate peripheral work
Core continuity reserves are justified by the need to preserve essential services through revenue shock, litigation, disaster, security incident or operator transition. A peripheral programme should not acquire an open-ended claim on that reserve merely because its fee expires.
Each non-core charge should include only the reserve needed for credible wind-down and bounded operational volatility. The target, use conditions and maximum should be published. Once the balance reaches its cap, collection should decrease automatically unless operators approve another use.
Large surpluses are a warning. They may reflect prudent forecasting, delayed work or lower demand. They can also allow a programme to continue after payer authority ends. The sunset instrument should prohibit using accumulated funds to run a new term without renewal. Money collected under the old term may settle the old term and close it; it cannot become an endowment for the same activity.
Deficits also require discipline. Management should not routinely cover them from core fees and then present the programme as self-financing. If the board approves an emergency bridge to avoid unlawful default, the amount and reason should be disclosed and charged against any later renewal revenue. Repeated deficit is evidence that the fee model or activity is unsound.
This separation supports trust in the core reserve. Operators should know that money held for registry succession will not be consumed to preserve a contested event, campaign or grant portfolio. Financial resilience depends as much on protected purpose as on balance size.
IPv4 scarcity must not become a convenient tax base
IPv4 holdings are visible, differentiated and economically significant. That makes them an attractive fee metric. A larger holding may correlate with account complexity, transfer activity, security exposure or ability to pay. It may also have little relationship to the cost of a peripheral programme.
NRS should distinguish charging for service burden from charging for scarcity value. If a fee rises with IPv4 holdings, the renewal statement should identify the cost or distributive reason. A general claim that larger holders benefit more from the Internet is too broad. It could finance almost any institutional ambition.
Scarcity-linked charges can affect the transfer market. A recurring holding fee may encourage returns or consolidation, change bid prices and disadvantage holders whose resources support low-margin networks. A transfer percentage can resemble a toll on asset value rather than payment for verification. Those effects may be intended policy, but they should not emerge accidentally from a revenue decision.
The operator vote should therefore receive counterfactual models: flat account charge, transaction charge, resource-size tier, income-independent levy and voluntary beneficiary funding. The models should show revenue concentration, small-operator burden and likely behavioural response. No basis is neutral; the point is to choose it transparently.
Automatic sunset is especially valuable where scarcity provides a growing base. Rising market value can make a modest percentage produce revenue far beyond efficient cost. Expiry forces the institution to return to the service and ask whether the amount still bears a defensible relationship to the approved activity.
Optional programmes should prefer optional finance
Some non-core activities may be valuable enough to continue but not appropriate for universal compulsory funding. Sunset should open alternative finance rather than create a binary choice between levy and disappearance.
A specialized analytics service can use subscription revenue. A conference can charge attendance and accept disclosed sponsorship under conflict controls. A training programme can combine beneficiary fees with grants targeted to operators unable to pay. Research can be commissioned by a consortium. Public-interest work can receive restricted donations so long as donor influence and exit risk are disclosed.
Optional finance has costs. It can fragment access, favor wealthy entities and expose the programme to sponsor agendas. Those risks belong in the renewal comparison. Universal funding may still be the best choice when broad participation creates collective value that cannot be captured individually.
The key is that usefulness does not answer who should pay. An activity can advance Internet development while remaining outside the minimum number-registry bargain. Requiring a separate financing decision keeps the Society from treating its invoice as the default answer to every worthy goal.
Voluntary migration should be planned before sunset, not improvised after rejection. The dossier can identify prospective subscribers, grant restrictions, minimum viable scale and safeguards against pay-to-play access. Operators can then compare a universal renewal with a credible alternative rather than with abrupt closure alone.
Emergency charges need a narrow exception, not a blank cheque
A severe security incident or continuity event may require expenditure before an operator vote can be organized. The constitution should permit a temporary emergency assessment, but the exception must be harder to abuse than ordinary non-core authority.
The trigger should be objective: an event creating imminent risk to unique records, authentication, essential certification state or service availability that cannot be addressed from the approved core contingency. The board should record the evidence, maximum amount, payer basis and actions funded. General financial pressure or a delayed programme is not an emergency.
The assessment should expire quickly, perhaps within ninety days, unless operators ratify a longer term. Unspent money returns or credits automatically. The board cannot use emergency revenue to launch adjacent improvements once the immediate risk is contained.
Independent review should begin while the charge is active, not months later. Security-sensitive detail may be protected, but operators should receive enough information to understand the event class, amount and necessity. A reviewer with access to withheld evidence can attest whether the trigger was met.
Core contingency remains the first line because repeated emergency invoicing would undermine predictable finance. The exception exists for hazards beyond a prudent reserve, not as an alternative to maintaining one.
Audit must test the boundary as well as the accounts
A conventional financial audit can confirm that revenue and expenditure are fairly presented while missing the central governance problem. NRS also needs a boundary audit.
The reviewer should test whether costs assigned to the temporary programme belong there, whether core resources subsidized it, whether programme staff performed unrelated work, whether contracts exceed the term, whether the payer list matches the authorized class and whether management collected after expiry. It should compare the public fee identity card with actual operation.
The audit should inspect anti-evasion risk. New budget labels, reorganized teams and capital transfers may conceal continuation. The test is functional: substantially the same purpose, beneficiaries, activities and payer burden remain subject to sunset even if the organizational chart changes.
The reviewer also verifies the electorate and vote record. Were all eligible operators notified? Was eligibility frozen on the stated date? Were corporate groups aggregated as required? Did any payer receive an undisclosed waiver near the vote? Financial and electoral integrity meet at renewal.
Independence requires appointment and removal terms that do not leave the reviewer dependent on the programme sponsor. Operators could approve the auditor from a shortlist, or an audit committee with no programme role could appoint one under a published rotation policy. The board should respond to findings, but it should not edit the opinion.
Disputes should preserve service while testing authority
A payer may challenge classification, amount, eligibility or collection after sunset. The remedy should separate the disputed non-core charge from essential registry service wherever solvency and security allow.
NRS should not suspend authorized record maintenance or route-security access merely because an operator contests a peripheral levy in good faith. The disputed amount can be held in escrow, secured or paid under reservation while an independent reviewer decides. Deliberate refusal to pay an undisputed core fee remains a different matter.
Review should be fast enough to affect the term. A decision issued after expiry may clarify accounts but cannot protect the operator from an unlawful burden during the programme. Interim relief can stop collection of the contested component without disrupting the ledger.
Reasons should identify the authorizing resolution, payer status, calculation and evidence considered. If the dispute concerns whether work is truly core, the decision should apply the constitutional function test rather than defer automatically to the board's label.
Courts may ultimately interpret contracts, corporate authority and mandatory law. A clear sunset record helps them avoid choosing technical policy. They can determine whether the institution possessed authority, followed the adopted decision and preserved agreed service boundaries. Continuity becomes an explicit protected interest rather than a rhetorical reason to deny every remedy.
Anti-evasion rules must follow substance
Without an anti-evasion clause, the board can wait for expiry and place the same staff and activity inside a broad core department. It can replace a named levy with a slightly higher annual fee, call the programme a capability rather than a service, or divide it among several cost centres.
The rule should compare purpose, principal activities, beneficiary class, personnel, contracts and claimed outcomes. If a reasonable operator would recognize material continuity, the charge remains temporary. Reorganization does not restart the term.
Some migration will be legitimate. A security pilot may reveal a vulnerability that makes a function indispensable to authentication. Conversion then requires the higher-threshold core amendment, independent evidence and a statement of why the need is structural rather than created by avoidable dependence. Operators should vote on that proposition directly.
The clause should also cover indirect finance. Sponsorship income does not justify using core staff without accounting for their time. A subsidiary cannot collect the expired charge on NRS's behalf. Procurement cannot require operators to buy the same peripheral service from a designated vendor as a condition of registry access.
Public cost maps make evasion harder. Each core and non-core activity should appear in a stable catalogue with prior names and funding sources. Transparency is most useful when it preserves comparability across institutional redesign.
The strongest objection is instability
Critics can reasonably argue that repeated sunsets shorten planning horizons, invite factional campaigns and make skilled staff leave. Operators may vote against a useful fee during a wider dispute with the board. Low turnout can let an organized minority terminate a regional benefit. Vendors may charge more for cancellable contracts.
Those costs are real. The response is not to abandon expiry but to design it around the core. Essential services retain stable finance. Non-core terms can be long enough for evidence and responsible employment. Participation floors and advance calendars reduce surprise. Wind-down reserves price termination instead of denying it.
There is also a cost to permanence. Programmes accumulate without a constituency for repeal, aggregate invoices hide mission choices, and staff learn that continuation depends more on budget placement than demonstrated value. Operators disengage because consultation cannot change the outcome. Stability then protects the institution rather than the service.
The proper balance is asymmetric. The closer an activity is to unique records, security and continuity, the stronger the presumption of durable funding. The farther it travels, the stronger the case for time-limited authority and payer renewal. This is more precise than treating all spending as equally essential or equally suspect.
Adoption should begin with classification and clean accounts
NRS should not impose sunsets retrospectively without understanding current commitments. The first step is a public activity catalogue separating minimum services, closely supporting functions and peripheral programmes. Cost allocation, staff time, contracts and reserves should be mapped to each class.
Operators then approve the initial core list under a high threshold. Activities outside it receive transitional terms long enough to gather evidence and align contracts. The transition should not presume renewal; it should make expiry feasible.
The Society should publish fee identity cards and baseline cost statements before the first sunset calendar begins. An independent reviewer should test opening balances and ensure that peripheral deficits or assets have not been hidden in the core.
Renewal dates should be staggered so operators can examine each dossier rather than face a single crowded meeting. Material programmes might come up in quarterly groups, with no more than a manageable number at one ballot. The annual core budget remains separate.
After the first cycle, the board should report not only which fees survived but what the discipline changed: narrower scopes, lower cost, alternative finance, closed programmes, improved evidence and corrected cost allocation. The goal is not a high repeal rate. It is an institution that can distinguish necessity from momentum.
Success is measured by credible choice, not constant rejection
A sunset system can become ceremonial if every charge is renewed by overwhelming margins on thin evidence. It can also become destructive if political opposition causes repeated closures unrelated to programme merit. Metrics should test the quality of choice.
Useful measures include the share of non-core revenue with a current identity card; cost variance explained; proportion of contracts aligned to term; operator participation by size and geography; time between dossier publication and vote; number of material audit findings; unspent balance disposition; and time required to complete wind-down after rejection.
The Society should track classification disputes and cross-subsidy. A rising share of staff time moved from temporary programmes into unpriced core work indicates erosion. So does repeated emergency finance for predictable activities.
Renewal results should be interpreted cautiously. A high approval rate may show strong programmes. It may also reflect dependence, bundling or an electorate dominated by beneficiaries. A low rate may show healthy correction or poor programme design. The surrounding evidence matters more than the percentage.
The best sign is credible non-renewal without registry harm. If operators can end a charge, staff can close the activity, balances can be returned and the ledger continues normally, the constitutional promise is real. That possibility disciplines every renewal even when most programmes continue.
A model automatic-sunset clause
The constitutional provision can be concise while supporting detailed schedules. It should say that no compulsory charge outside the enumerated core mission may continue beyond its specified term. Initial authority must identify purpose, payer class, amount or calculation, evidence measures, maximum commitments, reviewer, renewal electorate, end date and wind-down disposition.
Renewal requires publication of reconciled cost and revenue, performance against the original baseline, incidence, alternatives, commitments, conflicts, audit findings and a tested closure plan. The resolution must be separate from the core budget and approved under the pre-existing operator franchise.
Silence means expiry. Late review, board resolution, budget inclusion, relabelling, transfer to another entity or expenditure of accumulated balances cannot extend authority. A materially similar activity remains subject to the original outer limit.
Emergency assessment is allowed only for an imminent threat to the minimum service, capped in amount and duration, independently reviewed and submitted promptly for operator ratification. Core continuity reserves cannot finance ordinary continuation of an expired programme.
On expiry, collection and new commitments stop. Accrued liabilities and approved wind-down may be paid; records remain available; unused balances are returned or credited; and the final report compares promise with result.
The clause does not decide which ambitions are worthwhile. It decides that technical dependence is not a permanent blank cheque.
Mission discipline is strongest when money can say no
Institutions rarely announce mission creep. It arrives as a sequence of reasonable proposals: a pilot, a response to a new concern, a temporary team, a useful partnership and a small addition to the annual invoice. Each may be defensible. Together they can turn a narrow registry into a body that claims revenue for every problem adjacent to the Internet.
An automatic sunset does not require austerity or deny collective benefit. It requires the Society to preserve a durable minimum service and justify the rest at intervals chosen before institutional momentum takes over. Current cost evidence replaces inherited estimates. Incidence replaces vague claims that everyone benefits. Operator approval replaces consultation without consequence. A wind-down plan replaces the claim that continuation is the only safe option.
The discipline also improves programmes that survive. Managers know what evidence matters. Contracts remain reversible. Payers can distinguish support for the ledger from support for a peripheral ambition. The board can point to an affirmative, bounded mandate rather than infer consent from continued payment.
The Number Resource Society should be financially resilient where the Internet depends on it and constitutionally modest where its ambitions extend beyond that dependence. Stable core funding and expiring non-core authority are not opposites. They are the paired design that protects continuity without converting necessity into mission without end.