The Fee Payer Who Cannot Vote

Summary

• Regional registry charges attach to different things: a legal membership, an LIR account, an address holding, an independent-resource record, an associate status or a downstream service. Electoral rights attach to separately defined legal classes.
• APNIC expressly offers paid Non-Member accounts that do not receive online election votes. ARIN Service Members must obtain and retain General Member status before voting. AFRINIC Associate Members pay annual fees but attend member meetings as observers rather than ordinary electors.
• RIPE NCC comes closest to fusing service and membership, yet even there separate charges and downstream payments do not create extra votes, and suspension for non-payment removes an existing member's franchise.
• Excluding some payers can be defensible. A coherent association needs eligibility, good-standing and contact rules. The legitimacy defect arises when an institution treats its revenue base, service population or regional reach as though all three were represented by the electorate.
• A credible account should publish four denominators every year: distinct paying legal entities, eligible voting members, ballots cast and materially affected non-voting customers. Without them, claims of fee-funded member accountability cannot be tested.

Start with the invoice, then ask what it buys

The simplest story about a membership association runs in a circle. Members pay. Members vote. Elected governors approve the budget. The institution spends the members' money. If members dislike the result, they replace the governors. That circle can be a powerful form of accountability because financial contribution and constitutional control are held by the same identifiable constituency.

Regional Internet Registries complicate the story. They are not only associations. Depending on the region and the instrument, they are also service providers, corporate committees, registration counterparties, custodians of records, organisers of policy forums and coordinators within a global numbers system. Money enters through annual membership fees, per-account contributions, resource-related schedules, maintenance charges, sign-up fees, sponsorship arrangements and payments passed through local or national intermediaries. The person who bears the cost is not always the legal member. The legal member is not always eligible to vote.

The eligible member does not always cast a ballot.

An invoice proves a narrower proposition than governance rhetoric often asks it to prove. It shows that a named party owes money under a specified schedule or agreement. It may keep an account in good standing, maintain a contractual relationship or fund a defined service. It does not, without another rule, establish a right to nominate a director, approve accounts, amend constitutional text or vote in an election.

This is not merely semantic. If the populations of payers and voters differ, the institution needs to say which one authorises which decision. A budget may be legally approved by eligible members even though some revenue comes from non-voting accounts. A fee schedule may be adopted by voters whose charges are partly passed to customers. A service change may be financed by every account but reviewed electorally by only one class. Each arrangement can be valid. None should be described as universal representation.

Five systems use five different constitutional maps

Comparison requires caution because the five RIRs do not use the word member in the same way. RIPE NCC is a Dutch membership association in which conclusion of the standard service relationship ordinarily establishes membership. ARIN distinguishes Service, General and Trustee membership, with election rights concentrated in General Members in good standing. APNIC offers both Member and Non-Member service relationships and weights Member votes by tier. LACNIC's bylaws distinguish voting Active members from Adhering members who may speak but not vote.

AFRINIC distinguishes Registered, Resource and Associate Members, with different meeting and voting positions.

Those differences are not drafting trivia. They determine whether payment is a condition of voting, merely one condition among several, or wholly separate from the franchise. A fee can be necessary but limited public evidence. It can be paid by a non-member. It can preserve a service without creating governance rights. It can be owed by an intermediary whose customers ultimately absorb the expense. It can also be waived or discounted without diminishing a member's constitutional status.

The disciplined question is therefore not, "Do members pay fees?" They generally do. It is, "For every class of recurring payer, what legal step turns payment into voice, vote and remedy?" The answer must be given class by class. A high-level annual report that lists members, resources and revenue cannot substitute for that map.

RIPE NCC: the strongest fusion still has seams

RIPE NCC is the case most likely to support the intuitive circle. Its Standard Service Agreement identifies the member relationship, its Articles of Association give each unsuspended member one vote, and the General Meeting approves charging schemes. An ordinary member with one or several LIR accounts therefore belongs to the association and can vote once if not suspended. The institution is not inventing a constituency after collecting the money; the legal relationship and the constitutional franchise are closely connected.

But the seams matter. The 2026 billing procedure charges an annual contribution per LIR account, not per corporate vote. A member with several accounts receives several invoices and one vote. Separate charges may also attach to independent assignments, autonomous system numbers and legacy records. The payment unit can be an account or record while the electoral unit remains the legal member. That choice may be entirely sensible, but it disproves any simple equation between euros contributed and votes held.

There are also people behind the direct payer. An LIR can recover registry-related costs through prices charged to customers or End Users. Those customers may pay a specific annual maintenance amount under a sponsorship agreement or absorb the cost through a broader connectivity bill. Their money supports the chain, but they do not become RIPE NCC members merely because the LIR passes on an expense. The LIR's vote remains its own.

Finally, payment works as an exclusion device as well as an entry condition. Under the Articles, failure to pay the annual contribution automatically suspends the member; during suspension it cannot attend the General Meeting or exercise voting rights. This is a conventional good-standing rule. Yet it shows that the franchise is not a reward for historical contribution. It is a current legal status activated by compliance with constitutional and financial conditions.

One account, several fees, one corporate voice

The distinction between fee units and voting units exposes a common analytical error. Suppose one company maintains four LIR accounts. The charging scheme may assess four annual contributions. The company still casts one vote. A second company with one account also casts one vote. This is not a hidden defect; RIPE NCC publishes both principles. It reflects a choice to protect equality among legal members rather than convert financial contribution into proportional political power.

That choice has merits. If every paid account produced a ballot, organisations could acquire more influence by opening additional accounts. Wealthier members would have a straightforward route to buying power. Keeping one vote per member blocks that route within one legal entity and prevents the budget from becoming a shareholder register.

The same choice also creates an obligation of candour. The institution cannot use total invoices, LIR accounts or resources as evidence that its electorate represents each financed relationship. Those are different denominators. In 2024, RIPE NCC reported 19,993 active members and 20,991 active LIRs, expressly noting that one member can hold more than one LIR. By June 2026, it reported 20,056 members and 20,782 LIR accounts. The aggregate gaps do not reveal who held the extra accounts or how much each paid. They do establish that counting accounts as members would be inaccurate.

The May 2026 General Meeting report adds another denominator: 3,421 members registered to vote and 3,049 cast ballots. That is a substantial electoral event, but it is not the entire paying service population. The proper account should show the progression from legal members to unsuspended members, registered voters and actual ballots, while separately showing LIR accounts and non-member beneficiaries. Each count answers a different legitimacy question.

ARIN: paying for service does not preserve the ballot

ARIN makes the separation more explicit. Its public membership guidance says that membership is not required to obtain direct Internet number resources and does not confer an advantage in obtaining them. It distinguishes Service Members from General Members and reserves election voting to General Members in good standing whose organisations have designated a valid Voting Contact by the relevant deadline.

This means an organisation can have a recognised service relationship without automatically holding the election franchise. A Service Member must request General Membership if eligible. It must then maintain the participation conditions that ARIN introduced with the membership changes that took effect around the 2022 transition. Following the 2023 election and later elections, a General Member that casts no ballot in any of the previous three elections reverts to Service Member status, though it may apply again.

There is a defensible institutional logic here. A long list of nominal members who never update contacts and never participate can weaken election integrity. Requiring an affirmative request and periodic participation makes the electorate more current. It also means that paying, receiving service and voting are different states. A Service Member can remain financially and operationally connected to ARIN while lacking the ballot available to an active General Member.

ARIN's framework therefore rebuts two opposite simplifications. It is wrong to say that everyone receiving service controls the association. It is also wrong to say that payment alone buys a vote. The franchise depends on classification, application, good standing, contact designation and participation history. If public communications describe ARIN as accountable to "the members," they should specify which class is acting and how many service organisations are outside that electorate.

Participation conditions improve one measure and narrow another

Active-electorate rules often produce a favourable statistic: turnout among eligible voters rises because dormant organisations are removed from the denominator. That can be real improvement. An election in which a larger share of qualified voters participates is easier to administer and less vulnerable to forgotten credentials.

Yet the same reform can narrow the population described as politically represented. If a service organisation pays charges but does not complete a General Membership request, misses the contact deadline or loses status after three non-voting elections, its absence is no longer counted as abstention. It disappears from the voting denominator. Reported turnout can rise even if the share of all service-connected organisations casting a ballot remains unchanged.

Neither denominator is inherently correct for every purpose. Corporate validity should use the legally eligible electorate. Claims about engagement among service users should use the service population. Claims about regional legitimacy require an even broader account of affected operators and customers. Problems begin when the highest percentage is selected without naming its denominator.

A serious annual report would publish the full bridge: organisations with current service relationships; Service Members; organisations eligible to request General Membership; approved General Members; General Members in good standing; valid Voting Contacts; ballots cast. It should state where an organisation can appear more than once and where several services are consolidated into one legal entity. That table would make ARIN's active-membership design assessable without assuming that every exclusion is illegitimate.

APNIC's Non-Member is not a paradox

APNIC offers the clearest named example of a payer without a vote. Its Non-Member fee page describes organisations that want access to resources without becoming full APNIC Members. Some governments and not-for-profit bodies may face rules that prevent them from joining other organisations. They can hold a direct service relationship, satisfy the same resource-management criteria and pay a Non-Member annual account fee.

The word Non-Member does not mean informal, free or invisible. APNIC publishes a distinct Non-Member Fee Schedule. From 2025 through 2027 it set staged changes to base fees, with a separate Non-Member base higher than the Member base. Least Developed Country discounts can reduce the amount for qualifying organisations. The relationship is deliberately documented and priced.

APNIC's online voting terms are equally clear: Non-Member account holders are not eligible to vote online. The Member is the constitutional voter, acting through authorised contacts and subject to the election rules. Payment and registration provide contract rights and services, not corporate membership.

This arrangement may serve organisations that cannot or do not wish to join. It also demonstrates why "fee payer" cannot be used as shorthand for "member." A Non-Member may contribute more under the published base schedule than a similarly situated Member and still have no election vote. The difference is a product of legal classification, not the size of the invoice.

A non-voter can still hold a remedy

Electoral exclusion does not place an APNIC Non-Member outside law. The Non-Member Resource Services Agreement defines the parties, service obligations, notices and internal response or appeal steps. A non-voter can possess contractual rights that an ordinary entity in an open meeting lacks. This matters because governance debates often collapse vote, voice and remedy into one scale of inclusion.

They are separate. Voice is the ability to submit a policy proposal, comment, speak or entity. Vote is the formal power to select officeholders or decide a corporate matter. Remedy is the ability to require reconsideration, enforce a contract or obtain relief from a court or other competent body. Payment may support one or more of these without supporting all three.

APNIC's open policy process permits interested people to participate and distinguishes consensus from an election vote. That openness can give a Non-Member meaningful influence over resource policy. It does not let the Non-Member cast Member election votes. Conversely, a Member can hold many tier-weighted votes but never contribute substantively to policy discussion.

The institutional design question is not whether every right should be identical. It is whether the package is intelligible and proportionate. A payer excluded from elections should know which decisions remain open to it, what evidence decision-makers must consider, and where an adverse service act can be reviewed. Electoral exclusion is more defensible when voice and remedy are real rather than decorative.

APNIC also shows that more payment can mean more votes

APNIC's Member structure complicates the anti-plutocratic ideal in another direction. Its active tier document assigns votes according to chargeable address holdings: one for Associate, two for Very Small, then four, eight, 16, 32 and 64 through Extra Large. Annual fees are also calculated by holdings. Within the Member population, financial and resource scale therefore correlate with formal voting weight.

The correlation is not exact. Fee formulas, discounts, non-chargeable resources and tier boundaries matter. A vote is allocated by published membership rules, not by attaching a ballot to each dollar. Still, APNIC deliberately treats scale as constitutionally relevant. It is the opposite of RIPE NCC's one-member-one-vote rule, where multiple paid accounts do not increase a single member's ballot.

Both systems can give reasons. APNIC can argue that members responsible for larger holdings and higher charges carry greater registry exposure. RIPE NCC can argue that corporate equality prevents large incumbents from dominating small operators. The comparison reveals that there is no natural law connecting payment and voice. Institutions choose a proxy for political standing and embed it in their rules.

That choice should be tested against the actual interests at stake. Address holdings may be a poor proxy for customer count, critical-service dependence, route scale or the cost of an erroneous record. A small address holder can operate essential infrastructure. A large holding company can outsource operations. Fee-weighted voice is not self-justifying merely because a schedule is mathematical.

LACNIC: speaking status is not electoral status

LACNIC's bylaws draw the line through membership categories and resource pathways. Active A members and founding members have voting rights under the published constitutional structure. Adhering members may take part in the General Assembly with voice but no vote. Organisations receiving address space through qualifying direct or national-registry paths can enter Active A status; an organisation receiving only an ASN does not automatically become a member under the current guidance.

The result is a layered political geography. Two organisations may both rely on accurate regional records and both pay costs connected to number services, yet one holds a vote and the other only a speaking right or contractual position. Among Active A members, address scale can produce between one and eleven votes. Once again, payment, resource status and franchise are related without being identical.

The distinction between voice and vote deserves respect. Allowing an Adhering member to speak can enrich deliberation and place evidence on the record. But it cannot be described as equivalent to deciding the outcome. An organisation that can argue but cannot vote lacks the formal power to select directors or determine a resolution when the rules require ballots.

LACNIC should therefore report category counts alongside financial data. The 2024 Assembly minutes show weighted votes on institutional reports and a fee-adjustment proposal. They do not disclose the category and control distribution behind every vote. Without those denominators, an observer cannot tell how much of the financial constituency had electoral power or how concentrated that power was across address tiers.

AFRINIC: an annual fee can buy observer status

AFRINIC's Associate Member provides another direct test. The constitution defines an Associate Member as a person or organisation with a substantial interest in number-resource management and AFRINIC's mission that signs the relevant agreement. AFRINIC's fee guidance publishes annual Associate Membership amounts for individuals and for small, medium and large organisations. These members do not use assigned number resources under the registration service relationship.

The Associate Member page says that the class receives notice of member meetings and may attend as an observer. It can use training, consultancy and technical-expertise services where qualified. It does not receive the ordinary electoral position of Registered and Resource Members. A corporate Associate Member can therefore pay thousands of dollars a year and remain outside the ordinary vote.

That arrangement is not incoherent. Associate status is voluntary, and the member receives a defined package. AFRINIC may reasonably reserve control over resource-service governance to Resource Members and directors holding Registered Member status. The important point is that the word member and the fact of payment do not settle the electoral question even inside one corporation.

The distinction also warns against crude cross-regional tables. A column labelled "members" will combine an AFRINIC Associate observer, an unsuspended RIPE NCC voter, an ARIN Service Member and an APNIC Member with between one and 64 votes. A comparative report must translate each class into common rights: pay, receive service, speak, nominate, vote, inspect, appeal and exit.

The downstream payer is the least visible class

Direct non-voting accounts are at least legible. Their agreements and fee schedules can be read. The harder constituency is the customer behind an LIR, NIR, host, access provider or enterprise group. That customer may finance the regional relationship through a contract price without ever receiving an RIR invoice. It may use addresses drawn from the provider's allocation and depend on the provider to maintain records, reverse delegation and security credentials.

If a registry fee rises, the provider may pass the increase through. If an account closes, the customer may face renumbering or loss of administrative support. If policy changes, the provider may alter contract terms or deployment. The customer's exposure can be substantial even though its legal relationship is entirely downstream.

Calling the provider a member does not make every customer represented. The provider's vote is a corporate right exercised in its own interest. It may consider customer welfare, but there is no general presumption that it has consulted customers or received authority to vote for them. A provider may prefer lower compliance costs while a customer prefers stronger continuity protection. Their interests can diverge.

This does not require a ballot for every subscriber. It requires accurate language. The provider is an intermediary, not an automatic democratic proxy. When a proposal imposes material downstream costs, the institution should seek customer-impact evidence directly and require members to disclose the basis of any representative claim.

Why the company law answer is not enough

A corporate lawyer can answer the narrow question quickly: voting rights belong to the class identified by the constitution, subject to governing law. A customer or non-member payer does not acquire a ballot merely because money ultimately reaches the corporation. That answer is important. It protects institutions from unbounded claims by anyone who can trace a payment.

But legal validity does not exhaust institutional legitimacy. RIR decisions concern services with unusual coordination value. There is generally one recognised regional institution for a given relationship, and changing providers is not equivalent to changing an ordinary software subscription. A payer excluded from the electorate may have limited ability to exit without changing addresses, sponsorship or business structure.

That dependence raises a second question: what safeguards compensate for narrow voting standing? Administrative law offers a useful design vocabulary even when the registry is private. Consequential decisions can require notice, evidence disclosure, reasons, conflict controls, proportionate transition periods and independent review. Contract law can protect defined service expectations. Competition and consumer law may matter in particular jurisdictions. Open policy processes can hear non-members.

The point is not to convert an association into a state. It is to match safeguards to consequences. The less credible the payer's exit, the less confidently the institution can rely on corporate voting alone.

Equal votes and equal fees solve different problems

Debates about charging schemes often become debates about fairness. One-member-one-vote says that each legal member has equal corporate dignity. A flat per-account fee says that each account consumes a broadly comparable base service. A resource-based fee says that scale should contribute more. A weighted vote says that scale should confer more political power. These propositions can be combined in several ways, but they are not interchangeable.

RIPE NCC's model can charge several accounts while assigning one member vote. APNIC's model links holdings to both fees and vote tiers. LACNIC uses resource categories to determine electoral weight. ARIN narrows voting through active General Membership rather than address scale. AFRINIC separates non-resource Associate contributors from ordinary electoral members.

No formula removes the need for judgment. Flat fees may burden small operators. Progressive fees can create difficult valuation questions. Equal member votes can understate customer exposure behind large networks. Weighted votes can entrench incumbents. Participation requirements can improve active turnout while shrinking the electorate.

An institution should state which problem each rule is intended to solve. A charging scheme should fund predictable services without disguising political allocation. A franchise rule should identify a defensible constituency without treating revenue as consent. When the same variable controls both, the institution should explain why.

The argument for excluding some payers

There are strong reasons not to award a vote to every source of funds. First, downstream payments are often impossible to trace cleanly. A customer pays for a bundle of connectivity, support and infrastructure; only a fraction relates to registry costs. Counting it as a direct contribution would invite double representation through both customer and provider.

Second, corporate governance needs stable legal identities. Election administrators must know who is eligible, who may act for an organisation and how disputes are resolved. A shifting population of indirect users would make the register fragile.

Third, some non-member arrangements exist precisely because the organisation cannot or does not wish to join. Imposing membership could conflict with its legal constraints. Fourth, automatic fee-based votes could enable the purchase of influence through multiple accounts or trivial paid services. Fifth, ordinary commercial contracts do not generally convert every customer into a shareholder or association member.

These objections defeat a simple "one invoice, one vote" reform. They do not justify ignoring non-voting payers. The right response is to separate electoral control from other safeguards: consultation, reasons, customer-impact review, contractual standing, independent appeals and credible exit. Representation has more than one instrument.

The argument for treating payment as constitutionally relevant

The opposite case is also serious. A recurring payer is not a casual observer. It sustains the institution and may be locked into a recognised service relationship. If it lacks both a vote and a practical alternative, the institution receives compulsory-looking revenue without the discipline normally supplied by voice or exit.

Payment can also reveal a concrete affected interest. Open forums sometimes attract entities with expertise but no exposure to the costs of implementation. A payer has at least some stake in service quality, price and continuity. Excluding that stake from budget and board decisions can produce a constituency of speakers without power.

The correct conclusion is not that the payer's interest overrides every other one. It is that the interest should appear in the constitutional account. If a class pays but cannot vote, the institution should say why, identify its alternative rights and disclose how much revenue and operational dependency the class represents.

That disclosure would improve debate. It could show that a non-voting class is small and well protected by contract. It could instead show that a large share of revenue is collected from organisations with no electoral path. Both are empirical possibilities. Neither should be decided by rhetoric.

Four denominators, published together

Every RIR should be able to publish four comparable annual denominators. The first is distinct paying legal entities. This count should separate members, direct non-members, sponsored End Users and other billed relationships while avoiding duplicate counting across multiple accounts.

The second is eligible voting members on the register cutoff date. It should identify exclusions for class, suspension, missing contact, participation requirement or other rule. The third is ballots cast, with invalid or unused entitlements explained. In weighted systems, both organisations voting and voting units used should be shown.

The fourth is materially affected non-voting customers. This is the hardest measure and need not identify individuals. Providers could report bounded ranges of downstream organisations, independent-resource users and customer allocations, subject to audit and privacy controls. The aim is not surveillance. It is to estimate the constituency whose costs are carried through intermediaries.

Published together, the denominators would prevent category substitution. Revenue discussions could use payer counts. Election reports could use eligible voters and ballots. Impact assessments could use affected-customer estimates. No single number would be asked to prove everything.

Rights should be mapped beside every fee class

Counts alone are limited public evidence. A public rights matrix should place each fee class against the decisions it can influence and the remedies it can invoke. Columns should include election vote, constitutional amendment, budget approval, policy proposal, meeting speech, access to records, internal appeal, independent review and termination or sponsor change.

Such a matrix would reveal important compensations. An APNIC Non-Member lacks an online election vote but holds a direct service agreement and specified appeal path. A downstream RIPE NCC customer may participate in open policy discussion but lack contractual standing against the association. An ARIN Service Member can seek General Member status if eligible. An AFRINIC Associate can attend as observer and use specified services. These are not equivalent packages.

The matrix would also expose empty rights. A theoretical speaking right is weak if agendas close before affected parties receive notice. An appeal is weak if it cannot preserve service while facts are reviewed. An exit is weak if it requires surrender of continuity. Institutional design should evaluate rights by practical effect, not by the number of boxes checked.

Good standing must be reviewable

Because payment status can switch voting rights on or off, billing administration becomes election administration. A misapplied invoice, disputed category or delayed bank transfer can affect the register. Institutions should therefore publish cutoff dates, notice periods, cure opportunities and an expedited review route for good-standing disputes.

The reviewer should be able to inspect the invoice, payment evidence and governing rule without taking direction from a candidate or incumbent board member. Decisions should be reasoned and published in anonymised aggregate form. If a dispute cannot be resolved before voting closes, a provisional ballot or equivalent preservation mechanism may be appropriate where governing law permits.

This is not an invitation to let chronic debtors vote indefinitely. It is recognition that disenfranchisement can result from administrative error as well as genuine default. The more consequential the election, the stronger the case for a documented chain from invoice to exclusion.

The same principle applies to ARIN's contact and participation conditions and to category decisions elsewhere. Eligibility rules are legitimate only if affected organisations can know, correct and challenge their status before the moment of decision.

No vote should be inferred from silence

Institutions sometimes treat payment without complaint as acceptance. The inference is weak where the payer cannot change provider easily and has no vote. Paying an invoice can mean only that interruption is more costly than compliance. It does not prove support for the fee, the board or the policy funded by it.

Silence by an indirect customer is even less informative. The customer may not know which regional decision affected its bill. Its provider may not disclose the connection. The relevant consultation may occur in a language, timezone or forum the customer does not follow. Absence from that forum cannot be converted into consent.

This is why reasons matter. When governors adopt a charge or service change, they should state which classes supported it, which classes lacked votes, what objections were received and how downstream effects were assessed. A valid member majority remains the legal decision-maker. The reasons record prevents that majority from being inflated into unanimous service-user approval.

What the evidence cannot establish

Public documents establish the architecture: classes, fees, votes and formal remedies. They do not provide a complete current five-RIR census of who pays without voting. Some fee schedules change annually. Membership status changes. A single organisation can appear in several accounts, regions or corporate entities. Downstream costs can be bundled and invisible.

The evidence also cannot establish that non-voters are systematically mistreated. Many providers consult customers. Non-member contracts can protect service interests. Open policy forums can be more influential than low-turnout elections. Some organisations deliberately choose non-voting status. Formal exclusion is a fact; practical power requires further study.

Nor does cross-regional comparison support a single legal conclusion. The institutions operate under different corporate laws and constitutional texts. A right enforceable for a Dutch association member may not map directly to an Australian corporate committee, a Virginia nonstock corporation, a Uruguayan association or a Mauritian company.

The defensible claim is narrower and important: the fee relation and the governance qualification diverge in documented ways. Any statement that fee-funded RIRs are controlled by the people who pay them must identify which payers it means.

From fiscal constituency to legitimate institution

A legitimate registry institution does not need to grant an equal vote to every person touched by a charge. It does need to reveal the boundaries of its electorate and protect those on the other side. The minimum package is straightforward: harmonised payer and voter counts; a class-by-class rights matrix; reviewable good-standing decisions; customer-impact analysis for material proposals; reasons that distinguish legal approval from broader acceptance; and remedies capable of preserving continuity while a dispute is heard.

These measures would improve member power rather than dilute it. Members could see who finances the institution, where costs are passed downstream and whether the electorate is carrying the consequences of its decisions. Boards would have evidence beyond the loudest meeting interventions. Non-voters would know where to speak and what relief is available.

The longer-term question is whether registry service can be made portable enough that voice and exit reinforce one another. Number Resource Society offers a restrained future direction by arguing that operators should have clearer control, accurate records and meaningful alternatives to concentrated authority. That proposition should be judged by interoperable design, independent governance and tested continuity, not by advocacy alone.

Until such alternatives exist, the constitutional burden remains with the incumbent institutions. They should not call every invoice a membership mandate. The fee payer who cannot vote is not an anomaly at the edge of the system. It is a recurring legal position created by the system's own classifications. Accountability begins by counting that person accurately, naming the rights withheld and supplied, and refusing to treat payment as consent.