Field Value
Author BTW Media
Published 2026-07-01
Primary category arin
Categories governance; rir-watchdog; arin
SEO title ARIN and the economics of fees, reserves, and incentives
SEO description A research analysis of how ARIN's member fees, reserves, legal budget, transfer rules, legacy-resource treatment and post-exhaustion incentives shape registry behaviour in North America.
Focus keyword ARIN fees reserves incentives
Primary domain Governance
Content type Research
Topic fees, reserves, and incentives
Subject ARIN
Region North America
Time horizon 12-24 months
Impact HIGH
Confidence B / 0.88

The quiet test of a rich registry

ARIN is not the regional Internet registry most associated with open institutional crisis. Its territory contains the United States, Canada, parts of the Caribbean and the North Atlantic, and a large share of the world's cloud, hosting, enterprise, university, government, payment, security and access-network infrastructure. Its public operations look orderly beside the more dramatic fights that have surrounded other parts of the RIR system. The meetings are regular. The procedures are legible. Budgets, fee schedules and annual reports are published. The Board is elected by General Members. Staff can point to years of operational continuity and to services that most network operators use without daily controversy.

That stability is why ARIN deserves close attention. An institution in visible failure teaches by breakdown; a mature institution teaches by incentives. ARIN is a cleaner test of the post-exhaustion registry economy because the obvious excuses are weaker. It is not poor. It is not newly improvised. It is not short of documents. It operates in a commercially sophisticated region where addresses have long since become business inputs, balance-sheet concerns and transaction assets. The question is therefore not whether ARIN can keep the registry running tomorrow. It is whether a registry with money, staff, rules, reserves, legal capacity and member voting can remain mainly a neutral ledger when the resource beneath the ledger has become scarce, tradable and financeable.

The North American setting sharpens the question. IPv4 addresses in ARIN's region are no longer just technical labels assigned to networks. They are inputs for access providers, data centres, cloud platforms, mail systems, content-distribution firms, universities, public-sector networks, managed-service providers, security vendors and companies that inherited large early allocations. They are also sold, leased, pledged in commercial diligence and considered in mergers and restructurings. ARIN does not describe number resources as property, and that legal framing matters. But it does not make the resources economically weightless. Nor does calling a charge a service fee prevent that charge from affecting value, liquidity and risk.

The main argument is straightforward. ARIN's behaviour is shaped not only by mission language, community procedure or engineering necessity, but also by the economics of fees, reserves, legal capacity and institutional survival. Annual invoices finance the organisation. Transfer charges put a price on market movement. Reserve policy determines how long the institution can absorb deficits, litigation, political pressure or member resistance. Legal spending capacity affects how confidently the registry can defend its interpretation of rules. Member voting supplies legitimacy, but it also reflects a narrow and self-selected subset of the affected economy. Legacy-resource treatment tests whether historical certainty can survive modern service dependence. Post-exhaustion scarcity turns administrative rules into economic signals.

ARIN's own public materials are useful not because they settle the debate, but because they are good exhibits. The 2026 fee schedule lays out annual Registration Services Plan charges, transfer processing fees, a capped fee for pre-2024 LRSA holders and a mechanism for annual increases. The 2026 budget shows a revenue base dominated by registration maintenance fees, a budgeted operating deficit, legal and professional-service lines, and more than $36m in investment reserves. The investment policy statement explains the purpose of reserve funds, including the explicit use of operating reserves for legal challenges arising from policy enforcement and other significant outside-counsel matters. The transfer guide describes a private market filtered through registry approval. The legacy-resource page shows that some core record services remain available to legacy holders outside an ARIN agreement, while RPKI and IRR services require an agreement.

Those documents should not be read as a corporate creed or as a neutral ending to the argument. They are records of incentives. A registry that recovers costs through annual fees has one set of pressures. A registry holding a reserve near one year's spending has another. A registry that can condition transfer recognition on fee standing, agreements and recipient qualification has another. A registry that must defend itself in courts and preserve staff, systems, offices, meetings and global coordination has another still. The public vocabulary of stewardship can describe some of this, but it can also blur it. The more useful vocabulary is economic: who pays, who benefits, who is delayed, who can exit, who bears uncertainty, and who has the institutional capacity to stand its ground.

From rationing to capital administration

The old economics of IPv4 administration was rationing. The resource was finite, but it could still be issued from a regional pool. A registry could ask whether an applicant needed address space, whether the applicant would use it efficiently, whether contact records were correct and whether the request complied with policy. The administrative question was who should receive a scarce but still distributable public input. In that world, registry fees looked closer to the dues of an infrastructure association. They paid for staff, systems, meetings and registration services. Address value existed, but the registry was still mainly an allocator and record keeper.

IPv4 exhaustion changed that compact. ARIN's IPv4 waiting-list page states that its free pool depleted in September 2015. Since then, organisations seeking IPv4 have relied on limited returned space, reserved pools, in-region specified-recipient transfers, inter-RIR transfers, mergers and acquisitions, leasing, network sharing, carrier-grade NAT, address-rich corporate acquisitions and IPv6 transition plans. That mix is not a distribution pool. It is a capital environment in which the registry record helps determine whether an address block can be trusted, financed, transferred and operated.

The registry remains essential, but the nature of its essentiality has changed. In the allocation era, its main power was to approve requests from the common pool. In the post-exhaustion era, its main power is to recognise, update, service or refuse to recognise movement of already-held resources. A seller can sign a contract and a buyer can wire money, but the public record changes only if the registry process is satisfied. ARIN can ask whether the source is authorised, whether the block is eligible, whether the recipient qualifies, whether fees are current, whether the right agreement is signed, whether a waiting-list restriction applies and whether an inter-RIR counterparty operates a compatible policy. Each check may be defensible. Together they turn a ledger into an administered market.

Institutional economics treats such positions carefully. A body need not own an asset to shape the market for it. It need only control a condition necessary for reliable use, public recognition, security, customer diligence or transfer completion. Land registries, securities depositories, clearing houses, port authorities, domain registries and numbering registries all affect markets without setting market prices. Their fees may look small relative to the value beneath them, but their discretion matters precisely because market participants cannot easily route around the bottleneck.

ARIN's case is subtle because many of its rules are not facially abusive. Source verification protects against theft. Fee-currentness protects the funding base. Signed agreements define service and liability. Recipient qualification reflects conservation habits. Legacy distinctions reflect history. Reserve policy protects continuity. Member voting constrains the Board. The risk is not crude extraction. The risk is institutional drift: record-integrity rules and cost-recovery tools gradually become instruments of market control, contract migration, budget stabilisation and organisational self-preservation.

That drift is clearest when fees, reserves and legal capacity are read together. Fees are not merely invoices; they are the price of remaining inside a service and recognition system. Reserves are not merely prudence; they are the institution's ability to withstand anger, deficits, legal shocks and political pressure. Legal budgets are not merely professional expenses; they are the cash expression of how far the registry is prepared to defend its interpretation of rules. In a scarce-address market, these are policy instruments even when no policy document describes them that way.

The fee table as political economy

ARIN's 2026 Registration Services Plan is simple in presentation and rich in implications. The fee schedule sets annual charges by service category, with aggregate holdings measured across IPv4, IPv6 and autonomous system numbers. A 3X-Small organisation pays $275; 2X-Small pays $550; X-Small pays $1,100; Small pays $2,205; Medium pays $4,410; Large pays $8,820; X-Large pays $17,640; 2X-Large pays $35,280; 3X-Large pays $70,560; 4X-Large pays $141,120; and 5X-Large pays $282,240. The same table says annual RSP fees may increase by no more than 5% as approved by the Board. On May 12, 2026, ARIN announced that the Board had approved a 5% increase for 2027, citing a long-term financial strategy intended to align operating costs with revenues while maintaining cost-control targets through 2030.

The table looks like cost recovery. It is also a map of distributional politics. It shows who subsidises whom, who is most exposed to tier changes, who has the strongest reason to vote and who is likely to view ARIN's charges as negligible compared with address value. A 5X-Large holder paying $282,240 may control resources whose market value is many orders of magnitude larger. A 3X-Small holder paying $275 may have a more immediate sensitivity to percentage increases, especially if it is a small wireless ISP, community network, specialist hoster or enterprise whose network budget values predictability more than institutional ceremony.

The most important feature is not the absolute level of the fee. ARIN's charges can be defended as modest compared with the market value of IPv4 space. In a narrow sense that is true. If addresses can sell or lease at high per-address prices, even a six-figure annual registry charge may look small. But that comparison is dangerous. It shifts the justification from service cost to asset value. If ARIN charges for registry services, the relevant question is whether the fee is proportionate to the cost, risk and investment required to maintain accurate, secure and available registry functions. If fees are justified by the value of the assets recorded, the registry begins to resemble a rent collector on privately used infrastructure inputs.

ARIN's budget still presents the service-cost story as the formal accounting frame. In the 2026 budget, registration maintenance fees are budgeted at $29.64m on a GAAP basis, compared with total revenue and support of $32.628m. Network transfers contribute $1.745m. Registration allocation and assignment fees are only $495,000 after accounting adjustments. In other words, ARIN is funded primarily by the standing customer base, not by new issuance. That is exactly what one would expect after exhaustion. The pool is no longer the business model. The stock of existing registrations is.

This creates a structural survival incentive. ARIN needs a broad paying base. It needs annual maintenance fees to rise enough to cover a staff-heavy, software-heavy, security-conscious and governance-heavy institution. It also needs large holders not to exit cooperation, litigate, withhold payment or mobilise against fee growth. The Board may approve increases, but every increase is a political act because customers cannot choose a competing North American registry for the same resources. The fee resembles association dues. The market position resembles a utility.

The annual 5% cap is therefore useful but incomplete. It constrains the slope of fee growth. It does not answer the base question: what cost standard should govern a monopoly-like registry service? A capped increase can still compound. A balanced-budget target can still justify greater revenue from a captive base. A fee schedule can be public and still create cross-subsidies that particular groups reject. Legacy holders experience the loss of a cap differently from post-ARIN customers. Transfer recipients experience one-time processing fees differently from annual maintenance. Small operators experience percentage increases differently from hyperscale firms. A public price table is not automatically a neutral one.

Transfer fees and the administered market

ARIN's transfer charges show how cost recovery and market approval become intertwined. The 2026 schedule lists a $500 non-refundable source transfer request fee for 8.2 merger, acquisition and reorganisation transfers; a $500 source fee for 8.3 specified-recipient transfers within the ARIN region; and a $500 source fee for 8.4 inter-RIR transfers out of the ARIN region. The schedule states that the $500 processing charge does not guarantee approval. It also says the source and recipient must pay outstanding annual fees before evaluation or completion. The recipient then pays a separate transfer processing fee after approval and before allocation, with the fee scaled by aggregate IPv4 transfer size. A /24 costs $187.50; larger than /24 through /22 costs $375; larger than /22 through /20 costs $750; and the table rises by size until larger than /6 costs $192,000.

Seen narrowly, this is a plausible attempt to charge for work. Transfers require diligence. Staff must verify authority, documentation, organisation records, agreements, fee status, source eligibility, recipient qualification and technical updates. Larger transfers may carry greater institutional risk or more complex records. A fee schedule that scales with block size can be defended as a proxy for complexity, consequence or the value of staff attention.

Seen economically, the schedule confirms that ARIN is not merely passively recording private bargains. It charges both the source-side request and the recipient-side recognition. It ties processing to fee currentness. It states that financial terms between private parties are for those parties to negotiate, while all transfers must comply with current ARIN policy and only compliant requests may be processed. That is not a free market. It is a market under administrative recognition.

The difference matters because the transfer market is where North American scarcity is priced. IPv4.Global's public pricing-data page described continued strength in the IPv4 market in 2026, with large-block prices moving upward slowly, small and medium block prices stable, and demand and transaction volume strong across block sizes. That is a broker's market note, not a neutral economic census. But it captures a reality that ARIN's transfer activity also reflects. The 2025 annual report counted 540 8.2 transfers, 1,795 in-region specified-recipient transfers and 445 inter-RIR transfers during 2025. The registry record has become a transaction layer for scarcity.

ARIN's strongest claim in this layer is anti-fraud and record integrity. It must confirm that the source really controls the resource, that the recipient exists, that merger documents are real, that authorisations are valid, that no dispute is being smuggled through a routine update and that routing-security records will not create dangerous residue. These are ledger functions. They are essential. A registry that simply rubber-stamped asserted transfers would make hijacking, corporate impersonation and title disputes more likely.

The more debatable layer is recipient qualification and needs-based control. ARIN's transfer guide says IP addresses and ASNs may be transferred when a company acquires assets using the resources, when unused IPv4 or ASNs are released to a specified recipient who qualifies under current ARIN policy within the region, or when unused IPv4 or ASNs are released to a specified recipient outside the region who qualifies under the recipient RIR's policy. That language imports allocation-era judgments into private transactions. It keeps ARIN in the business of deciding whether the buyer qualifies, not merely whether the seller is authorised and the record can be safely updated.

There are reasons for such rules. Without recipient qualification, speculative accumulation, shell buying, circular transfers or purely financial warehousing could increase. But every such rule has costs. It can delay transactions, reduce buyer pools, make financing harder, push demand into leasing structures, reward firms with specialist policy staff and favour incumbents that already understand registry procedure. A rule designed to prevent abuse may also entrench larger, better-advised actors. The institutional question is whether each qualification rule protects the ledger or preserves the registry's historic role as allocator after allocation has largely ended.

The recipient processing fee deepens the issue. If the fee is cost recovery, it should map to work performed. If it is scaled to block size, it begins to resemble a transaction levy on scarcity. The levy may be small relative to deal value, and charging larger transfers more may be administratively efficient. But the theory should be explicit. The more a fee tracks resource value rather than operational cost, the harder it is to maintain that ARIN is only a service provider and not a toll point in the address market.

Reserves as insurance and power

Reserve funds are usually defended in the language of prudence. A critical registry should not run with a thin cash buffer. It must survive revenue shocks, vendor failures, investment volatility, technology incidents, office moves, data-centre changes, security problems, litigation and delays in fee collection. No serious operator should want the North American registry to be financially fragile. The question is not whether ARIN should hold reserves. It is how reserves change incentives.

ARIN's public numbers are large enough to matter. The 2026 budget shows investment reserves of $36.578m at the beginning of the year and an estimated $36.586m at the end, after estimated earnings and withdrawals. The same budget shows total operating expense on a GAAP basis of $36.072m. The reserve is therefore close to a year's GAAP operating expense. The 2025 budget had shown a beginning reserve balance of $36.092m and an estimated ending balance of $35.044m after larger withdrawals. In 2026, ARIN budgeted a GAAP operating deficit of $3.444m, offset partly by investment earnings and planned reserve use.

The investment policy statement explains the structure. ARIN divides funds into an Operating Reserve Fund and a Long Term Reserve Fund. The long-term reserve exists to provide financial stability, and the policy says the Board has directed that the long-term reserve should be equal to or greater than the previous year's operating and capital budgets on a cash basis. The operating reserve exists to meet operational needs in the current budget year and to support legal challenges resulting from policy enforcement and other legal transactions requiring significant outside counsel. Withdrawals above $2m require Finance Committee approval. The operating reserve is invested conservatively; the long-term reserve has diversified return targets and risk controls, including a prohibition on direct investment in publicly traded telecommunications or technology companies in major indices because of perceived conflict with member organisations.

This is financially literate governance. It is also politically consequential. A reserve roughly equal to a year's budget gives ARIN independence from immediate member anger. It lets the institution run deficits while adjusting fees. It allows management and the Board to litigate, resist pressure, fund system projects and maintain staff. It improves continuity. It also weakens short-term member discipline. A registry with no reserves may be too fragile. A registry with ample reserves may be too insulated.

The legal-purpose clause is especially revealing. Many nonprofits hold reserves for emergencies. ARIN's policy specifically contemplates legal challenges arising from policy enforcement. Policy enforcement is where a registry's authority meets member and holder expectations. If ARIN reviews resource use, denies a transfer, suspends services, revokes resources, enforces payment terms or interprets a contract, a holder may challenge it. The reserve fund is therefore partly a litigation capacity attached to registry powers that can affect address value and market movement.

That capacity can be valuable. A registry must be able to resist fraudulent claims, bad documentation, pressure campaigns and meritless lawsuits. It should not collapse because one well-funded holder sues. But the same capacity has a second side. If the downside of aggressive enforcement is funded by the fee base that must obey the enforcement, the registry may be more willing to test boundaries. Members collectively fund the legal capacity that can be used against individual members. That is sometimes unavoidable in associations. It is also a moral-hazard problem that deserves explicit oversight.

The better reserve policy distinction is functional. Continuity reserves should be protected and boring: payroll, core systems, data availability, security, disaster recovery and essential public services. Investment reserves should be transparent enough that members understand risk, liquidity and governance. Legal reserves should be accompanied by reporting of categories, triggers and authority, without disclosing privileged case detail. Members do not need every invoice. They do need to know whether reserves are preserving the ledger or funding contested exercises of discretion.

The legal budget beneath a quiet institution

ARIN's ordinary legal line is modest. In 2026 the GAAP budget lists legal at $284,000 and consulting plus other professional services at $1.008m. In 2025 legal was budgeted at $359,000 and consulting plus other professional services at $932,000. Those numbers are small beside total operating expense, salaries and benefits, reserves, or the market value of resources whose recognition ARIN affects. The modest line can be read as reassuring. ARIN does not budget like a litigation machine.

But annual legal cost is not the same as contingent legal power. The budget shows normal expectations; the reserve policy shows stress capacity. A registry can budget a moderate legal amount in ordinary years while keeping the ability to tap reserves for significant outside counsel if a policy-enforcement decision is challenged. That is not a criticism by itself. It is a description of institutional architecture.

The economic issue is liability asymmetry. A registry decision can have consequences much larger than the annual fee or the staff time involved. A transfer delay can disrupt a sale. A service stop can damage operations and reputation. Removal from public lookup services after non-payment can affect counterparties and customers. Revocation can destroy expected asset value. A refusal to update records can complicate acquisition financing. A contract interpretation can alter a legacy holder's bargaining position. Yet no critical registry can realistically become a full insurer of every downstream economic loss, and ARIN is not a general court for corporate disputes.

This creates a familiar problem: high-consequence authority with bounded institutional exposure. The registry must have discretion because the ledger must not become a channel for fraud. But if remedies are narrow, process is slow and practical review is difficult, the affected holder bears much of the downside while the institution bears mainly legal and reputational cost. Reserves help the institution bear legal cost. They do not necessarily help the injured counterparty if the registry made a wrong or excessive decision. That is why accountability should focus on whether the strongest powers are narrow, rule-bound, timely and reviewable.

The North American legal environment makes this more important, not less. Bankruptcy courts, mergers, asset sales, security interests, corporate restructuring and sophisticated commercial litigation are normal. IPv4 blocks can appear in diligence schedules, purchase agreements and restructuring negotiations. ARIN can describe its role as registration-right recognition rather than property adjudication, and that distinction matters. Yet market actors still need certainty. Legal uncertainty around registry recognition is priced like any other transaction risk.

ARIN's transfer guide points users to a legal perspective on bankruptcy-related transfers. That is a reminder that the registry sits at the edge of corporate law without being a general commercial court. Courts may approve sales or restructurings. Buyers may pay money. Lenders may rely on schedules. The registry still has to process the record. As IPv4 value rises, that interface will matter more often. A narrow, predictable and well-reported approach lowers market friction; a broad and opaque one raises the premium paid to lawyers and insiders.

Billing, revocation and the price of standing

ARIN's billing rules show how a modest invoice can carry large consequences. The billing page says annual fees are due for registry services for Internet number resources covered by an RSA or LRSA, usually on the last day of the anniversary month. Invoices are emailed 60 days before the due date. Reminder messages are sent 30 days before and on the due date. At 30 days past due, collection notices continue every 30 days, with additional attempts by email, standard mail, certified mail and phone. At 100 days past due, ARIN sends a Pending Deregistration Notice specifying when revocation will occur.

The revocation page adds the hard edge. At 120 days past due, if full payment has not been received, ARIN stops providing services and removes the resources and associated records from public lookup services such as Whois and RDAP. At 180 days past due, ARIN terminates the RSA, revokes the included resources and returns them to available inventory for reissue, subject to reinstatement details and whether resources have already been issued elsewhere.

This is not merely accounts receivable. It is the payment discipline of a monopoly-like registry service. ARIN cannot operate if fees are optional. A credible revocation path is necessary to prevent free-riding. But the consequence is larger than the unpaid amount. A missed, disputed or misdirected invoice can eventually affect public records, recognised control, services and address value. The escalation schedule turns billing into governance.

For most compliant customers the risk is remote. Invoices are predictable, reminders are ample and serious holders have every reason to pay. Edge cases still matter. Companies reorganise. Billing contacts change. Organisations dissolve, merge or enter bankruptcy. Legacy histories are messy. Small operators miss mail. Internal approvals delay payment. Fraudsters may try to manipulate account control. A fee dispute can become a service dispute. In such cases the registry's process determines whether a small administrative failure becomes a large market loss.

The economics are asymmetric. The annual fee may be small relative to address value, but the enforcement consequence can be severe. Due process is therefore not a luxury. Public-lookup removal at 120 days and revocation at 180 days may be reasonable if notices are reliable, contacts are current and dispute channels work. As IPv4 value rises, the market will reasonably ask for more evidence that accidental loss, hostile contact manipulation, corporate confusion and contested authority cannot easily produce irreversible harm.

Standing also becomes an asset attribute. A block with clean billing, current contacts, reliable organisation records, agreement clarity and no unresolved ARIN issue is more valuable than a block with administrative uncertainty. ARIN's billing system therefore shapes market quality. It encourages professional address management, specialised brokers, lawyers and services that keep records current. That is efficient in one sense. It also raises the administrative cost of holding address space, especially for smaller or older holders that did not originally treat IP resources as financial assets.

Legacy-resource certainty and the fee-cap question

Legacy resources are the constitutional memory of the North American registry system. ARIN inherited many resources issued before ARIN existed. Its legacy page states that when ARIN was formed in December 1997, the Board decided to provide registration services for those resources without requiring original holders to enter a Registration Services Agreement or pay service fees. It also says legacy holders not under an ARIN agreement can still maintain unique registration in Whois and RDAP, update public data, manage reverse DNS, maintain records through ARIN Online and access DNSSEC. They cannot access ARIN's RPKI or IRR services unless the resources are under an ARIN agreement.

That arrangement is a compromise. It recognises historical expectations while reserving some modern services for contracted customers. It is not pure neutrality. It uses service access as an incentive to sign. As routing security becomes more important, the value of RPKI and IRR access increases. A legacy holder may sign not because the historical question has disappeared, but because operational expectations have changed. Service design can therefore convert practical need into contractual migration.

The legacy fee-cap change makes the incentive more visible. ARIN's 2026 fee schedule says organisations with active Legacy Registration Services Agreements entered before January 1, 2024 receive the same RSP for legacy resources, but with a total annual fee cap. For 2026 that cap is $250, regardless of the number of IPv4 legacy resources held under such an LRSA, and the cap will increase by $25 per year. Any legacy resources brought under agreement after January 1, 2024 fall under regular RSP fees with no cap. ARIN's legacy-resource page says the legacy fee cap expired on December 31, 2023 for new additions, while pre-2024 LRSA-covered resources retain limits.

This is a quiet but important change in property-like certainty. The pre-2024 cap protects a class of holders who accepted a contract under an earlier fee promise. New entrants face the regular fee table. That may be fair. ARIN cannot freeze every future service price forever. Yet it also shows how contract timing changes asset economics. Two holders with similar historical resources can face different annual costs because one signed before a date and another signs after it. The difference affects expected cash flows, willingness to sign, transfer pricing and bargaining power.

The central issue is not whether legacy holders should pay nothing. They benefit from the registry. They may use services. The registry incurs costs. The issue is whether ARIN can use operational dependence to pull historical resources into a contemporary fee and contract regime without eroding the certainty that made those resources valuable. If the answer is yes without strong limits, legacy certainty becomes conditional on changing service needs. If the answer is no categorically, ARIN may be forced to provide modern services without adequate funding or legal clarity. Neither extreme is attractive.

A credible middle path would treat basic ledger continuity as non-negotiable, advanced services as separately defined, fee changes as predictable and historical distinctions as durable. The risk is that bundling becomes too attractive. As RPKI, IRR, automation and security functions become more central, ARIN can reasonably say that they require agreements. Holders can reasonably say that modern operational safety should not be used to rewrite historical expectations. The fee schedule is where this argument becomes arithmetic.

NRS, LARUS and Lu Heng's public commentary has pressed this point in more political language, arguing that registries should protect the public record rather than expand gatekeeping over economically valuable number resources. Those are interested voices; their commercial and advocacy positions should be weighed accordingly. But the underlying institutional question is legitimate. Historical certainty is not an anti-registry slogan. It is a market requirement. Buyers, sellers, users and lenders need to know whether old resources remain old in a legally and operationally meaningful sense, or whether modern service dependence gradually dissolves the distinction.

Member power, member limits

ARIN's governance gives members real tools, and that matters. The membership page says there are Service Members, General Members and Trustee Members. Membership is not required to obtain direct resources from ARIN and does not confer advantage in obtaining them. It is also not required for participation in the policy process, suggestions or consultations. General Members in Good Standing can vote in ARIN elections through a Voting Contact. ARIN also requires General Members to keep participating: after the 2023 annual election and each election thereafter, General Members that did not cast a ballot in any of the previous three ARIN elections revert to Service Member status, although a blank ballot counts as participation.

The 2025 annual report gives useful scale. As of December 31, 2025, ARIN reported 25,085 Service Members and 1,472 General Members. It also said ARIN served roughly 40,000 organisations and managed about 8m registration records, including public Whois and internal database records. The gap between the service base and the voting base is not a scandal. General Membership must be requested, maintained and used. But the gap is a reminder that member voting is real without being universal.

This distinction is crucial when ARIN decisions affect more than members. A transfer rule can affect buyers, sellers, brokers, customers and counterparties. A legacy-service rule can affect downstream networks. A revocation rule can affect users who have no vote. A fee increase can affect an organisation's budget even if it never became a General Member. A routing-security service rule can affect networks that filter or validate routes. The ARIN community is larger than the electorate, and the affected economy is larger than the ARIN community.

Member democracy therefore cannot be the whole legitimacy story. It disciplines the Board. It lets engaged customers remove or support trustees. It gives institutional decisions more credibility than pure staff rule. But it does not turn a membership corporation into a public regulator. It does not mean silence equals consent. It does not mean the most affected parties are the most represented. It does not prove that cost allocation is fair because a Board approved it.

The political economy of participation is uneven. Large holders have more at stake, more counsel, more compliance capacity and more reason to monitor fee schedules. Small operators may have less time and more immediate business pressures. Brokers and leasing firms may understand transfer incentives better than ordinary enterprises. Legacy holders may engage only when a service change touches them. Staff and long-term community participants understand procedure better than occasional voters. None of this delegitimises ARIN. It shows why transparency must do more work than voting alone.

The most useful member power is not just the ballot. It is the ability to demand budget clarity, reserve discipline, legal-spending categories, transfer-performance metrics, fee-impact analysis and plain explanations of policy choices. In a post-exhaustion registry, accountability must become more economic. Members should not merely ask whether services are stable. They should ask who pays, who benefits, who bears legal risk, who is delayed, who faces contract leverage, who receives residual returned space and who is pushed into the transfer market.

Budget politics after exhaustion

The 2026 budget makes ARIN's post-exhaustion position visible. Total revenue and support on a GAAP basis is budgeted at $32.628m, while total operating expense is $36.072m. Salaries, benefits and fringe account for $23.77m on a GAAP basis, tied to an approved employee count of 106. Engineering operations and infrastructure are $5.203m after accounting adjustments. Travel and meetings are $2.007m. Professional services total $1.292m, including $284,000 for legal. Industry support is $608,000, including $255,000 for ICANN and $190,000 for the NRO. General and administrative expense is $2.584m. The budget expects a GAAP operating deficit of $3.444m and estimated reserve withdrawals of $2.010m to balance the cash plan.

The budget is not extravagant on its face. A critical registry with 40,000 organisations, millions of records, security services, transfer processing, software systems, help desks, meetings, international coordination and governance obligations will not be run on a tiny staff. Salaries dominate because the institution is knowledge-intensive. Engineering costs are material because registry systems must be secure, available and maintained. Outreach and meetings are part of the multistakeholder culture, whether one admires or doubts that culture.

The issue is trajectory. A registry after IPv4 exhaustion does not have a growing free-pool issuance business. It has a mature maintenance base, transfer processing, routing-security services, software modernisation and governance obligations. If costs rise faster than maintenance revenue, the Board must choose among fee increases, cost cuts, reserve withdrawals and service changes. ARIN's approval of the 2027 5% increase should be understood in that context. It is not evidence of crisis by itself. It is a predictable response to a cost structure.

Yet fee increases are not politically neutral because the customer base is sticky. A network cannot decide that its ARIN-issued or ARIN-administered resources should now be served by another North American registry. The exit option is limited. In ordinary markets, excessive price increases invite substitution. In registry markets, they invite political opposition, litigation, non-payment, calls for reform, transfer structuring or grudging compliance. The absence of easy exit is why the annual budget is a governance document, not merely an accounting document.

Budget politics also affect institutional survival incentives. A stable staff and reserve cushion make ARIN resilient. They also create a natural interest in preserving the size and scope of the organisation. This is not sinister. Organisations defend their functions, budgets and relevance. After exhaustion, a registry may compensate for the loss of allocation centrality by expanding service scope, routing-security programmes, education, participation work, international coordination or compliance activity. Some expansion may be valuable. Some may be self-preservation described as public service. The difference should be tested by evidence, not assumed from institutional language.

The budget's network-transfers revenue line is also revealing. At $1.745m, it is material but much smaller than maintenance revenue. ARIN is not mainly funded by transfer tolls. That reduces one possible conflict: it is not financially dependent on maximising transfer volume. But it creates another priority. A registry funded by standing holders has its strongest survival incentive in annual maintenance revenue and fee adequacy, not necessarily in transfer-market liquidity. Unless members demand efficient market metrics, transfer friction can remain a secondary issue.

Legal certainty and the North American transfer premium

The North American transfer market depends on more than price. It depends on legal and administrative certainty. Buyers want confidence that the source can transfer, that ARIN will recognise the transfer, that the recipient will qualify, that records are clean, that RPKI and reverse DNS changes can be managed, that no hidden fee dispute will surface and that a court-approved transaction will not be trapped in registry ambiguity. Sellers want confidence that ARIN will not reopen unrelated historical questions or delay closing beyond commercial tolerance. Brokers want predictable queues and documentation standards. Lawyers want representations that can be underwritten. Lenders want confidence that address-related value is not illusory.

ARIN's transfer guide tries to create certainty by specifying categories. Section 8.2 transfers follow mergers, acquisitions and reorganisations. Section 8.3 transfers are in-region specified-recipient transfers. Section 8.4 transfers are inter-RIR transfers where reciprocal, compatible, needs-based policies exist. The guide says that once ARIN has a signed RSA and all applicable fees, resources are transferred within two business days. That final two-business-day statement is useful. It tells the market that once prerequisites are satisfied, record updates should be quick.

The uncertain part is not usually the final two days. It is everything before them. What documentation is sufficient? How does ARIN treat older corporate records? How flexible is it when names, jurisdictions, assets and network use do not fit a clean template? How fast does staff respond? How predictable are recipient qualification decisions? How often are requests abandoned or denied? How often does a dispute or billing issue emerge late? How much should parties involve ARIN before signing? These practical questions determine transaction cost.

ARIN publishes transfer logs, statistics and guidance, but market accountability could go further. Aggregate data on transfer request duration, approval rates, withdrawal reasons, denial reasons, documentation deficiencies, appeal use, fee-related holds, post-approval timing and inter-RIR mismatches would be valuable. It would not require revealing private sale prices or confidential documents. It would allow the market to distinguish ordinary diligence from administrative drag.

This matters because gatekeeper risk can be invisible when each file is confidential. One delayed transfer may be explained by bad documents. A pattern of delay may indicate under-staffing, overly broad review, unclear guidance or policy mismatch. Without aggregate metrics, participants rely on brokers, lawyers and anecdotes. That favours insiders. Better reporting would reduce the premium paid to procedural familiarity and make the transfer market more competitive.

The ledger-versus-gatekeeper distinction is practical here. A ledger-focused transfer process would ask whether the source is authorised, whether the recipient exists and can be recorded, whether fees and agreements are in place, whether the resource is eligible, whether the public record can be updated safely and whether routing-security consequences are clear. A gatekeeper-focused process asks broader questions about the recipient's business plan, future need, regional economics and policy philosophy. The former lowers transaction cost. The latter may express conservation values, but it raises capital friction. ARIN's burden is to justify each gate with evidence of harm avoided.

Post-exhaustion incentives: why scarcity keeps the registry relevant

IPv4 exhaustion might have made registries less important. In theory, once the free pool was empty, the allocation role should have shrunk, leaving record maintenance, transfer processing, routing-security services and statistics. In practice, exhaustion kept registries central because scarcity made the record more valuable. A registry entry now supports not only uniqueness, but also asset diligence, routing-security controls, transaction eligibility, service access and fee standing. Scarcity did not reduce registry power. It changed its basis.

This is the survival incentive at the heart of the ARIN test. A registry formed to allocate and record resources must justify itself after allocation declines. It can do so by providing excellent registry services, improving security, publishing useful data, maintaining public accountability and processing transfers efficiently. It can also do so by expanding the meaning of stewardship, preserving qualification rules, tying services to agreements, increasing outreach and defending its role in global Internet governance. Some of those activities are useful. Some can become self-reinforcing.

The 2025 annual report illustrates the modern service stack. ARIN reported work on TLS support, retirement of FTP access, data-centre migration, improvements to RPKI, IRR Auto-Manager, ASPA support in the testing environment, a ROA change log, RDAP changes, ARIN Academy and community outreach. These are not allocation-era activities. They are service, security, education and legitimacy activities. They may improve the registry. They also create a larger operational surface to fund and defend.

Routing security is especially important. ARIN reported 8,140 organisations signed up for RPKI services as of December 31, 2025, with 98% using Hosted RPKI. That is a useful public service. It also increases dependence on ARIN-operated systems and contract status. If hosted RPKI, IRR automation and related services become more central to routing hygiene, ARIN's service boundaries matter more. A registry that withholds a security service unless a holder signs an agreement is not merely selling convenience. It is shaping operational risk.

This does not mean ARIN should provide every service without contract or fee. Security services have liability, authentication and operational implications. But the economics are clear. The more the registry adds valuable services around the ledger, the more it can influence behaviour through bundling. The organisation's survival incentive aligns with service expansion. The market's need for certainty aligns with narrow, reliable record functions. The challenge is to expand services without turning optional value into coercive dependence.

Post-exhaustion also changes the politics of conservation. Conservation was straightforward when resources were handed out from a dwindling pool. It is more complicated when resources move between private parties. A needs test can prevent speculative warehousing. It can also suppress legitimate forward planning and make address supply less liquid. A 60-month waiting-list transfer lock-up can prevent arbitrage on residual returned space. It can also reduce flexibility for small networks that later merge or change strategy. A fee-currentness rule protects the institution. It can also turn a billing issue into a transaction barrier. Scarcity makes every rule dual-use.

The ledger versus gatekeeper risk

The clearest way to assess ARIN is to separate ledger functions from gatekeeper functions. A ledger function protects uniqueness, record accuracy, verifiable control, contactability, reverse DNS, routing-security coherence, transfer history and public trust. A gatekeeper function decides who may receive, trade, access, qualify, exit, re-enter or use services under conditions that go beyond record truth. Every registry needs some gatekeeping. Fraud prevention is gatekeeping. Minimum documentation is gatekeeping. The danger is excess gatekeeping: using stewardship language to exercise economic control that is not necessary for the ledger.

ARIN is not uniquely exposed to this risk. All regional registries face it. ARIN is the clean test because its legal, financial and market setting is mature. If a stable, well-documented registry in a rich region cannot keep the boundary clear, the problem is structural rather than accidental.

The boundary can be tested with simple questions. Does a rule prevent false records, duplicate claims, hijacked resources, broken contactability, unsafe routing-security state or unpaid service consumption? If yes, it is likely a ledger rule. Does a rule judge business need, restrain private transfer for scarcity-management reasons, induce contract migration, preserve regional supply, protect institutional revenue or advance a broad conception of community stewardship? If yes, it is a gatekeeper rule and should face stronger economic justification.

Transfer source verification is ledger protection. Recipient needs assessment in a private sale is gatekeeping. Billing notices are ledger-service finance. Public-lookup removal and revocation are high-consequence enforcement and need procedural safeguards. Legacy Whois and RDAP maintenance is ledger continuity. RPKI and IRR access tied to agreement status is service bundling with gatekeeper potential. Reserves for continuity are prudent. Reserves for policy-enforcement litigation are institutional power and require oversight. Member voting is accountability. Treating member participation as full public consent is overreach. Fee schedules are cost recovery. Using asset value to justify fee increases is rent logic unless made explicit and debated as such.

This distinction does not produce easy answers, but it improves the debate. ARIN defenders can argue that each gate is necessary. Critics can ask for evidence. Members can demand metrics. Market actors can price risk. Courts can understand the difference between recordkeeping and economic control. The public can see where a registry is indispensable and where it has acquired power by institutional convenience.

The weakest answer would be to collapse everything into stewardship. Stewardship is too elastic. It can justify accuracy, conservation, service expansion, fee increases, transfer denial, legacy contract pressure, legal fights and political positioning. A serious registry should not ask the market to accept so broad a word without accounts. Stewardship must be decomposed into functions, costs, risks and limits.

What accountability would look like

ARIN already discloses more than many comparable institutions. It publishes fee schedules, budgets, annual reports, corporate documents, service statistics, transfer information, policy materials, election rules, investment policy and member participation rules. That disclosure is a strength. But the next stage of accountability should match post-exhaustion economics, not allocation-era self-description.

First, fee-setting should become more economically explicit. A public fee consultation should separate cost recovery, reserve targets, inflation, staffing, security investment, service expansion, transfer-market work and cross-subsidy choices. Members should be able to see whether a fee increase is driven by salaries, capital projects, reserve restoration, legal risk, outreach, global coordination or new services. A 5% cap is helpful, but a cap is not analysis.

Second, reserves should be disclosed in a way that distinguishes continuity insurance from institutional war chest. The investment policy is already public. The missing layer is regular reporting of reserve usage by purpose, especially legal-purpose withdrawals and significant policy-enforcement expenses. Confidentiality can protect case details, but not the basic category of spending. Members should know whether reserves are preserving services or funding contested exercises of discretion.

Third, legal spending should be presented with more useful categories. The annual legal line is too compressed. Categories such as routine corporate counsel, contracts, transfer and bankruptcy matters, policy enforcement, resource review, governance, employment, vendor disputes and external institutional matters would help. The point is not to second-guess every invoice. It is to let members see whether legal capacity is mainly defensive, administrative or control-oriented.

Fourth, transfer-market metrics should be improved. ARIN can publish aggregate data on request processing times, approval rates, withdrawal reasons, denial reasons, documentation deficiencies, inter-RIR mismatches, fee-related holds and appeal outcomes. This would reduce insider advantage and clarify whether ARIN's administered market is efficient. It would also let ARIN defend itself against vague criticism with evidence.

Fifth, legacy-service boundaries should be stated in economic language as well as contractual language. If RPKI and IRR access require an agreement because of liability, authentication, security or cost, say so plainly. If fee caps are changing because old pricing no longer covers service cost, show the cost logic. If historical certainty is preserved for pre-ARIN resources, explain exactly what remains durable and what changes when modern services are requested. Ambiguity is expensive.

Sixth, member power should be widened through information even if the voting base remains narrower than the affected economy. The 1,472 General Members recorded at the end of 2025 cannot carry the full legitimacy burden for decisions affecting 40,000 organisations and many more downstream users. Better disclosure, plain-language economic analysis and accessible budget and transfer reporting would make non-voting participants less dependent on insiders.

None of this requires turning ARIN into a public utility regulator. It requires recognising that ARIN has utility-like features. A monopoly-like registry funded by annual fees, holding large reserves, operating critical services and controlling recognition in a scarce-asset market should expect utility-like scrutiny, even if its legal form is nonprofit and its culture is multistakeholder.

The next 12-24 months

The next two years are likely to test fee legitimacy more than technical continuity. ARIN's services are unlikely to collapse. The more probable pressure points are quieter: the 2027 fee increase, continued reserve withdrawals or restoration, member reaction to cost growth, the durability of the legacy fee distinction, transfer-market efficiency, routing-security service dependence and the political meaning of General Member participation.

The budget path is the first watchpoint. If operating expenses continue to exceed revenue, ARIN will need repeated fee increases, stronger cost control, reduced services, more reserve use or some combination. Repeated reliance on reserves would be defensible during a transition but risky as a habit. Repeated fee increases would be defensible if tied to documented service needs but vulnerable if members see mission creep. The institution should not assume that service stability alone will quiet economic questions.

The legacy boundary is the second watchpoint. Pre-2024 LRSA holders with capped fees have one set of incentives. Holders bringing legacy resources under agreement after January 1, 2024 face regular RSP fees. As routing security becomes more important, pressure to sign agreements may rise. ARIN should avoid any appearance that advanced services are being used to convert historical certainty into contemporary revenue. The safer strategy is to make service cost, liability and historical treatment explicit.

The transfer market is the third watchpoint. IPv4 demand remains strong enough that friction will be noticed. ARIN does not have to maximise liquidity at any cost. It does have to show that its gates prevent concrete harms rather than preserve allocation-era habits. The more public metrics it provides, the easier that defence becomes.

The reserve and legal lines are the fourth watchpoint. A reserve close to one year's spending is prudent, but it is also power. If legal disputes rise, members will want to know whether their fees defend the ledger or defend discretion. If legal spending remains modest, ARIN can still improve category reporting before a shock occurs. The time to clarify legal-spending accountability is before a major dispute consumes attention.

The final watchpoint is rhetoric. ARIN should resist the temptation to treat official, community or stewardship language as a conclusion. Those words can describe process, but they do not answer the economic question. The registry must show that fees match legitimate costs, reserves match continuity needs, legal powers are bounded, transfer gates are proportionate, legacy certainty is respected and member participation is meaningful without being overstated.

A stronger registry by narrower claims

The argument against gatekeeper drift is not an argument for a weak registry. It is an argument for a stronger registry by narrower claims. ARIN is most defensible when it is precise: it keeps records accurate, protects uniqueness, verifies transfer authority, maintains services, supports routing security, publishes data, runs fair procedures, collects reasonable fees and preserves continuity. It is less defensible when scarcity tempts it to adjudicate economic worthiness, use service dependence as contract leverage, treat member participation as full social consent or justify fee growth by reference to the market value of resources it did not create.

North America does not need an ARIN that is starved, timid or unable to defend the ledger. Fraud, hijacking, false transfers, broken contacts, routing-security mistakes and non-payment are real problems. A registry that cannot respond would make the market more dangerous. But North America also does not need an ARIN that gradually becomes an address-capital gatekeeper while describing every decision as stewardship. The value beneath the ledger is too high, and the exit options are too limited, for that to be accepted on trust.

The best future for ARIN is a leaner legitimacy: publish the costs, state the incentives, narrow the gates, defend the record, report the legal risks and separate continuity reserves from institutional self-protection. If ARIN can do that, it will be a useful model for the post-exhaustion registry economy. If it cannot, its stability will become evidence that even a mature registry drifts toward gatekeeping when fees, reserves, legal capacity and scarce resources sit in the same institutional hand.

The North American case is therefore not a sideshow. It is the cleanest test of the registry model after abundance. ARIN's fee table shows who pays. Its reserves show how long the institution can stand its ground. Its legal budget shows how discretion can be defended. Its transfer rules show how private markets remain administered. Its legacy policy shows whether history still constrains contemporary power. Its member system shows how accountability works and where it stops. The ledger remains indispensable. The gatekeeper risk is that indispensability becomes a reason to ask fewer questions just when the asset beneath the record requires more.