Summary

  • ICANN's adopted FY27 budget projects $165.1 million in ordinary operating funding. About $162.0 million, or 98.1 percent, comes from gTLD transaction fees, registry fixed fees and registrar accreditation fees. The projected RIR contribution is $0.8 million, less than one-half of one percent.
  • That funding concentration can shape attention, staffing, risk language and the visibility of results. It does not prove that domain registrants precisely subsidize number services, because published IANA and corporate cost reporting does not allocate every shared expense among names, numbers and protocol parameters.
  • Number governance has a deliberately different constitutional form. The five RIR communities develop regional policy; a global number policy must pass through all regions; the ASO checks and forwards it; and ICANN's Board has a bounded ratification role. The numbering service is separately disciplined by an agreement between ICANN and the five RIRs.
  • Financial convenience must not become jurisdiction. The proper safeguards are identifier-class cost reporting, service-level evidence, separate explanations for shared spending, conflict disclosure and explicit proof of authority whenever ICANN proposes action affecting number resources.

A corporation with three identifier duties and one dominant payer base

ICANN's name creates an impression of balance. The corporation coordinates assigned names and numbers, while Public Technical Identifiers performs the IANA functions for names, numbers and protocol parameters. In legal purpose, these identifier classes sit within one mission. In political economy, they do not finance the institution on anything close to equal terms.

The imbalance is visible in the adopted ICANN FY27 Budget, which took effect on 1 July 2026. ICANN projects $165.1 million in funding for its ordinary operations. Legacy and new gTLD transaction fees account for $116.8 million. Registry fixed fees add $28.6 million. Registrar application, annual accreditation and per-registrar variable fees add $16.6 million. Together those three domain-contracting streams provide about $162.0 million, or 98.1 percent of the operating total.

The same budget records $2.1 million in country-code contributions, $0.8 million in RIR contributions and $0.2 million in meeting sponsorships and other receipts. Even without counting country-code contributions as domain-name funding, the contracted gTLD sector supplies nearly the whole operating base. The number-community contribution is less than one-half of one percent of ordinary operating funding.

This is not a hidden condition. ICANN publishes the figures and explains the contracts behind registry and registrar fees. Nor is dependence itself evidence of misconduct. A public-benefit corporation may rely on a concentrated revenue base while administering duties that benefit a wider public. Central banks, sector regulators, standards bodies and professional associations often collect money from a subset of those affected by their work.

The governance question begins one step later. Revenue changes what an institution can count, forecast and defend. It creates recurring relationships with payers. It makes the health of one market material to hiring, travel, reserves, legal capacity and community support. If the institution also has a narrower, differently constituted responsibility for number resources, financial scale can be mistaken for constitutional scale. The risk is not a crude purchase of votes. It is that the funded centre of gravity gradually supplies the language, assumptions and administrative habits used everywhere else.

The FY27 ledger measures concentration, not ownership

The FY27 figures deserve careful reading because each category means something different. Transaction fees rise or fall with annual increments of additions, renewals and transfers that survive the applicable grace period. Registry fixed fees attach to registry agreements. Registrar accreditation fees attach to the right to remain an ICANN-accredited registrar, while variable fees spread a set amount and transaction charges across eligible registrars.

These are not voluntary donations from disinterested benefactors. They arise from contractual and accreditation relationships in the generic domain-name sector. The fee base therefore links ICANN's financial position to domain volume, the number of contracted parties and negotiated fee schedules. ICANN's FY27 plan assumes 244.4 million legacy and new gTLD transactions and lists higher-than-planned domain transactions as a financial opportunity. The FY26 third-quarter report likewise attributed stronger funding to domain-name transactions, registrar accreditation and application fees.

Yet a fee is not a share certificate. Registries and registrars do not acquire a residual ownership claim over ICANN because they supply its operating income. Directors remain fiduciaries of the corporation. Contracted parties participate in the GNSO through defined structures, but they do not obtain a veto over the ASO's regional policy communities. The Bylaws do not calculate authority according to dollars remitted.

The distinction protects both sides. Domain businesses should not be told that every payment buys a right to determine unrelated identifier policy. Number communities should not be told that low direct contributions make their authority conditional on the tolerance of the names sector. The corporation should not treat its own ability to finance a project as evidence that the project lies within its remit.

The correct inference from the ledger is therefore limited but consequential. ICANN is financially exposed to the domain-name market. That exposure creates incentives requiring disclosure and control. It does not establish corruption, establish an exact subsidy, or transfer authority. Those further claims need separate evidence.

Transaction revenue can shape attention without issuing an instruction

Institutional incentives seldom arrive as a written demand from a payer. They operate through planning assumptions. A revenue line that supplies almost all ordinary funding receives close monitoring. Its volume forecasts affect whether vacancies are filled, which projects can be launched and whether expenses must be reduced. Staff develop expertise in the contracts, disputes and commercial conditions surrounding that line. Directors receive repeated briefings about its risks.

This focus is rational. ICANN would fail every identifier community if it ignored the market that pays its bills. Financial resilience supports the technical staff, legal capacity, cyber security, facilities and corporate continuity needed to perform IANA work. The concern is not that domain revenue receives attention. The concern is that attention becomes a proxy for public significance.

Number-resource governance presents different signals. IPv4 scarcity, IPv6 deployment, autonomous system number administration, registry accuracy and inter-RIR coordination matter enormously to network operators, but they do not generate a comparable ICANN transaction stream. Many of the economic relationships sit at the RIR level. Policy debate occurs in regional meetings and mailing lists. Operational performance is measured against the IANA Numbering Services agreement. The success indicators are continuity, accurate registry changes and faithful implementation of globally agreed policy, not growth in an ICANN fee line.

An institution trained to ask how an initiative affects contracted-party funding can therefore ask the wrong first question about numbers. The relevant number-policy questions are who authorized the action, whether all five regional communities agreed where global policy is required, whether the ASO completed its verification role, whether the Board stayed within its defined review power and whether PTI met the service agreement. Affordability matters after authority, not before it.

Attention can also shape public narrative. A large meeting, outreach effort or technical study financed from ordinary operations may be described as evidence of ICANN's leadership across unique identifiers. That description can be accurate when the work is within the mission and requested by the affected community. It becomes hazardous when expenditure is used to imply that ICANN originated the mandate. Paying for a convening function does not make the convener the policy principal.

Number governance entered ICANN through a pre-existing regional system

The number community was not invented as a department of the domain-name corporation. The RIR system predates ICANN. APNIC, ARIN and the RIPE NCC were already operating when the first ASO agreement was signed in 1999. LACNIC and AFRINIC later completed the present five-region structure. The ASO's official history describes the formation of the ASO as bringing ICANN into an existing and functioning system of global address-policy coordination.

That origin is reflected in today's institutional design. The current ICANN Bylaws say the ASO advises the Board on policy issues concerning the operation, assignment and management of Internet addresses. They define the ASO as the entity established by the 2004 memorandum between ICANN and the NRO. The NRO, in turn, is the coordinating body of the five RIRs.

The ASO Address Council is the NRO Number Council. Each RIR region supplies three members, two selected by the regional community and one appointed by the RIR board. The council oversees the global policy development route, advises the ICANN Board on number policy and nominates people for ICANN Board Seats 9 and 10. This is a federation of regional authority with a link into ICANN, not a policy bureau receiving delegated discretion from ICANN's chief executive.

Regional policies remain regional. Anyone may propose policy in an RIR's open forum, and each community decides under its own documented rules. A policy becomes global only when it requires action by IANA or another ICANN-related body and obtains agreement in every RIR region through their respective procedures. The ASO Address Council then verifies the required steps and forwards the result.

This architecture explains why budget share cannot be the measure of policy weight. The legitimacy of a global number policy comes from convergence across affected regional communities. If the smallest financial contributor has completed the same bottom-up public deliberation as the largest, its regional consent is not discounted. Equal regional passage is a safeguard against converting wealth, membership size or proximity to ICANN into unilateral global authority.

The ASO asymmetry is constitutional, not an accounting error

Within ICANN, the ASO looks unusually light. It does not maintain a policy bureaucracy comparable to the structures serving generic names. The NRO supplies the secretariat functions described in the ASO memorandum. RIRs finance regional meetings, policy staff, registry operations and much of the human effort required to develop number policy. ICANN's budget therefore captures only a fraction of the resources sustaining the number-governance system.

The 2004 ASO Memorandum of Understanding assigns the NRO secretariat responsibility and defines global policy as policy agreed by all RIRs that requires action by IANA or another ICANN-related body. It also establishes a narrow Board interaction. When the ASO forwards a proposed global policy, the Board may accept it, ask for changes, reject it with stated concerns or fail to act within the specified period. A reaffirmed proposal faces a higher rejection threshold, and a second rejection leads toward mediation.

That design intentionally leaves much of the cost and authority outside ICANN. If one looked only at ICANN's accounts, number governance might seem underfunded. If one looked only at the ASO's place in the Bylaws, it might seem administratively thin. In reality, five regional institutions perform the substantial work.

The asymmetry is therefore not simply a gap to be filled by expanding ICANN. It is a separation of functions. ICANN provides top-level coordination, IANA registration services, Board ratification at a defined point and a home for ASO participation in wider corporate governance. The RIR system supplies regional policy formation, membership accountability, registry service, operational knowledge and the resources supporting those roles.

This separation can create friction. A distributed system may be slower to explain and harder to represent in a single budget chart. It may produce inconsistent regional practices. One RIR's institutional failure can affect confidence across the system. None of those problems gives the best-funded corporation a default right to absorb the missing function. Reform must follow the authority assigned by community agreement.

The $0.8 million contribution is not the price of number authority

ICANN's budgets have carried an RIR contribution near $0.8 million for many years. Historical plans described the contribution as a mutually agreed or traditional financial contribution and also noted that RIRs funded ASO meetings, staff support and travel. The FY27 budget keeps the line at $0.8 million.

It would be a mistake to read that number as the full cost of the number community's relationship with ICANN. It excludes the cost of operating five registries, maintaining regional registration records, running policy fora, supporting the NRO, selecting council members, participating in ICANN meetings and overseeing IANA numbering performance. It also does not isolate the cost to PTI of performing number registry changes within a team and technical environment shared with other IANA duties.

It would be equally mistaken to treat the contribution as a licence fee purchasing policy autonomy. The legal and institutional bases for RIR authority are the Bylaws, the ASO memorandum, regional constitutions and policies, and the numbering-service agreement. A larger contribution would not enlarge those powers. A smaller contribution would not erase them.

This principle matters in both directions. If domain fees cover corporate services used by number coordination, the names sector may reasonably ask for transparent cost allocation. It may not demand number-policy control as repayment. If the RIRs contribute less than the apparent cost of services they receive, the number community may reasonably explain the value and costs it supplies elsewhere. It may not dismiss every question about shared spending as hostile to autonomy.

Legitimacy requires a non-proprietary account of money. Payers deserve evidence of stewardship, not jurisdiction over other communities. Affected communities deserve control under agreed rules, not immunity from financial scrutiny. ICANN owes both an explanation detailed enough to prevent budget contribution from becoming a political myth.

Cross-subsidy is a question to measure, not a slogan to deploy

The phrase "cross-subsidy" can describe several different things. One is a deliberate transfer: revenue collected from one activity is assigned to another. A second is shared overhead: several activities use the same legal, human-resources, security and financial capacity, while one revenue stream happens to finance the corporation. A third is an accounting result: costs cannot be separated economically, so a common pool pays for an integrated service. A fourth is political rhetoric, used to suggest that the paying sector owns the supported function.

ICANN's published figures establish funding concentration. They do not by themselves identify which of those meanings applies to each dollar spent on number coordination. The FY27 IANA budget is $11.9 million, including $11.5 million for PTI and $0.4 million for specified support activities outside PTI. PTI performs naming, numbering and protocol-parameter functions. The public total does not present a complete expense allocation showing the marginal and fully loaded cost of each identifier class.

Some expenses are plainly shared. A security control can protect systems serving multiple registries. A finance team can account for PTI as a whole. A software engineer may improve a platform used across request types. Splitting those costs requires an allocation rule, and different defensible rules will produce different results.

Other expenses are separable. Staff time devoted to a numbering request, number-service measurement, ASO support or a dispute under the numbering agreement can be tagged. Travel support can be attributed to a constituency. A dedicated contract, audit or study can be identified. Publishing those figures would improve debate without pretending that every shared cost has a natural owner.

Until that allocation exists, strong claims should be bounded. It is fair to say that domain-name fees finance the corporate pool from which IANA and number-related support are paid. It is not yet fair to state a precise net subsidy from domain registrants to the RIR system. The RIR contribution, RIR-borne governance costs, shared overhead and benefits flowing back to the domain sector all have to be counted.

The burden should fall on whoever makes the numerical claim. A statement that "names pay for numbers" needs a cost method and treatment of RIR-supplied resources. A statement that "numbers pay their full way" needs the same. Financial transparency should reduce political leverage, not supply a new form of it.

The IANA budget combines services that have different principals

The IANA brand can obscure an institutional fact: the three main service families do not share one policy principal. Naming has the post-transition structures built around the IANA Naming Function Contract, the Customer Standing Committee and naming-function reviews. Numbering has the five RIRs, the NRO, the ASO and the IANA Numbering Services agreement. Protocol parameters are governed through the relationship with the IETF and the standards it produces.

PTI performs operational work across these families, which creates efficiency and a coherent technical culture. But a common operator does not merge the sources of authority. The person carrying out a registry change must follow the policy and service arrangement applicable to that registry.

The PTI FY26 plan usefully describes numbering services as administration of Internet number-resource registries in accordance with global policies and mutually acceptable guidelines. The work includes allocations of IP address and autonomous system number blocks to RIRs, handling returned resources, maintaining general number registries and administering the relevant reverse-DNS zones. That is an operational description, not a general grant of policy power.

The same distinction should appear in financial decisions. A shared IANA sustainability project may improve staffing, systems and resilience for all customers. Its governance case should still state which community requested each element, which agreement authorizes it and how costs are allocated. "IANA needs it" is not enough when the proposed work changes the rights or obligations of one service community.

The integrated budget also makes interruption risk a common concern. Number communities benefit from ICANN's reserve, cyber security and organizational continuity even though those resources are funded mainly by names. Domain communities benefit from the credibility of a stable operator serving all unique identifier systems. This mutual benefit is a sound reason for shared infrastructure. It is not a sound reason to collapse policy boundaries.

A service agreement gives the number community a different lever

The IANA Numbering Services agreement has been in effect since 1 October 2016. It was created because the former United States government contract had supplied performance and reporting requirements that would disappear at transition. The five RIRs and ICANN chose a direct agreement rather than relying solely on corporate assurances.

The agreement sets performance expectations, escalation and dispute-resolution routes. It renews automatically in five-year terms unless either side gives the required notice. In the event of non-renewal or termination, the RIRs select a successor operator for numbering services. PTI performs ICANN's service obligations under a subcontract, but ICANN remains the counterparty to the RIRs.

This arrangement matters for budget incentives because it gives the number community a service-based source of discipline that is not proportional to its contribution line. The RIRs do not need to outspend registries and registrars to insist that numbering requests be acknowledged, implemented accurately and reported. Their standing comes from being the contracting service community.

The NRO's IANA Numbering Services Review Committee adds public evidence. It examines performance data, invites comment across the five regions and reports to the NRO Executive Council. The 2025 review found full performance on several request and allocation measures while recording lower performance on a reverse-DNS measure. This type of record is more useful than a general claim that the service is valuable. It identifies what the operator promised and what occurred.

The successor-operator provision is also constitutionally significant. It prevents ICANN's control of the corporate budget from becoming permanent control of number operations. Transition would be complex and costly, and the existence of a clause does not make exit easy. Still, the legal possibility changes the relationship. Number service is not merely a discretionary program sustained at the pleasure of the names-funded corporation.

Mission language is the first boundary on spending

ICANN's Bylaws Mission describes number authority with care. ICANN coordinates allocation and assignment at the top-most level of IP numbers and autonomous system numbers. It provides registration services and open access for global number registries as requested by the IETF and RIRs. It facilitates global number-registry policy development by the affected community and performs related tasks as agreed with the RIRs.

Three limiting ideas are built into that language. The coordination occurs at the top-most level. The policy is developed by the affected community. Related tasks depend on agreement with the RIRs. The next sentence in the Bylaws is even clearer: ICANN shall not act outside its Mission.

Budget authority cannot override these limits. A project does not become lawful because the Board approved an appropriation. Financial capacity does not convert "as requested" into "as directed by ICANN." A strategic objective written broadly enough to cover unique identifiers cannot erase the more specific allocation of roles.

This sequencing should govern every proposal touching numbers. First identify the exact mission clause, ASO provision, global policy or agreement authorizing the act. Then identify the affected community decision. Then explain the requested operational result and cost. If the authority cannot be stated without relying on the size of ICANN's budget or the general importance of Internet stability, the proposal is not ready.

Mission limits also protect ICANN from payer pressure. A registry or registrar coalition cannot demand an unrelated number intervention merely because its fees make the intervention affordable. The Board can point to the Bylaws and the RIR agreements. A disciplined remit is a defence against both institutional ambition and funder ambition.

The global policy route keeps money outside the merits

The global number-policy route is designed to test consensus, not purchasing power. A proposal can begin in an RIR forum or with the ASO Address Council. It must move through the policy procedures of all five RIR communities. The Address Council reviews whether the required steps and consensus conditions have been met before sending the proposal to ICANN's Board.

The Board's role is consequential but bounded. The published Global Policy Development Process permits the Board to accept, request changes, reject with a statement of concerns, or allow a proposal to take effect through inaction. If the Board rejects, it must explain which significant viewpoints were not adequately considered. The ASO and RIRs then assess those concerns. If all RIRs reaffirm and resubmit, the Board needs a two-thirds vote to reject again, after which mediation follows.

This is not the generic-names policy route with number terminology substituted. It reflects the RIR federation. ICANN's Board checks whether the globally coordinated result meets the agreed standard and may raise defensible concerns, but it does not originate regional allocation rules by funding a study or staff team.

The structure also answers a common spending argument. ICANN may bear costs for Board analysis, legal review, IANA readiness or public communication when a global policy arrives. Those expenses do not make the resulting policy an ICANN product. They are the cost of performing ICANN's assigned step in a larger constitutional sequence.

Conversely, the RIRs' financing of regional deliberation does not allow them to bypass the global route when coordinated IANA action is required. Financial autonomy and policy autonomy are not absolutes. Each institution must complete the step assigned to it.

Board seats and budget votes reject a pay-to-play theory

The ASO nominates two voting directors to ICANN's Board, Seats 9 and 10. It also participates in the Empowered Community alongside the ccNSO, GNSO, ALAC and GAC. Those rights do not shrink when the RIR contribution falls as a share of ICANN revenue.

This formal equality is a deliberate rejection of corporate pay-to-play. The GNSO contains the sectors supporting almost all ordinary funding, but it receives one Decisional Entity position. The ASO, whose direct contribution is small, also receives one. Their constituencies, selection methods and policy duties differ, yet both can participate in the constitutional powers assigned to the Empowered Community.

Budget rejection makes the point concrete. The Empowered Community may reject ICANN and IANA budgets under defined petition, support, forum and voting rules. An individual can ask the ASO or another Decisional Entity to carry a petition under that body's procedures. The ASO thus has a route to challenge the financial plan that depends on institutional support, not on matching the dollars collected from domain names.

The power is difficult to use and should not be romanticized. A petition must clear time limits and coalition thresholds. Rejecting a budget is a blunt remedy. The caretaker provisions preserve continuity, and a Decisional Entity must weigh disruption against correction. Still, the ASO's place in the mechanism is evidence that contribution level is not meant to determine constitutional voice.

The same principle should guide informal influence. Board members nominated through the ASO owe duties to ICANN rather than instructions to the RIRs, but they can bring number-community knowledge into deliberation. Other directors should not treat that knowledge as a niche interest because the associated revenue line is small. Expertise and affected-community authorization are relevant measures; fee volume is not.

Spending categories can conceal the identifier split

ICANN's activity-based reporting improves on a ledger that lists only salaries, travel and professional services. The FY27 budget identifies $63 million for community and engagement, $27 million for policy implementation and $26 million for the technical mission, with the rest supporting corporate operations. About half of the technical-mission category is dedicated to IANA functions.

These categories help readers understand purpose, but they still do not reveal every identifier allocation. Community support can serve the GNSO, ASO, ccNSO, advisory committees and cross-community work. Policy implementation is heavily concerned with contracted parties in generic names. Technical work can support IANA, the L-root server, DNS research and other initiatives. Shared legal and financial costs sit elsewhere.

The absence of an identifier view has two effects. First, advocates can overstate subsidy because no public table supplies a precise counterfigure. Second, management can overstate balance because a broad "unique identifiers" label makes name-heavy spending look evenly distributed. Neither inference is reliable.

A better presentation would add an identifier lens beside the traditional and activity views. Expenses should be tagged as names, numbers, protocol parameters, shared IANA, shared corporate or other mission work. Shared amounts should identify the allocation method. The table should show direct spending, allocated overhead and non-ICANN resources supplied by partner communities where those resources are material to interpreting the result.

Such reporting would not need to assign every minute of staff time. Materiality thresholds and sampled time studies could produce a credible range. The aim is not false precision. It is to prevent a $165 million names-funded corporation and a distributed RIR system from arguing past each other with incomparable totals.

Reserves weaken immediate dependence but preserve historical provenance

ICANN's Reserve Fund is another reason to avoid a simplistic annual view. The FY27 budget begins with a projected reserve of $203 million and aims to keep the balance above twelve months of operating expenses. Reserves can sustain continuity when transaction revenue falls or an unexpected event creates large costs.

That cushion reduces short-term payer leverage. A temporary fall in registrations does not immediately force ICANN to abandon IANA operations or accept a registry demand. The Board has time to reduce expenses, use contingency and protect critical services. Financial resilience is therefore an accountability safeguard as well as a treasury objective.

But reserve money has provenance. It accumulated from prior operating surpluses and investment returns in a system financed mainly by domain-name fees. Moving current expenditure into the reserve does not make the underlying dependence disappear. It changes the time at which the link is visible.

Program funds require further separation. The new gTLD program is designed as a cost-recovery program financed by application fees. Auction proceeds are earmarked for the grant program. The FY27 budget projects $384 million in new gTLD program collections and $55 million in program expenses. Those large figures sit beside ordinary operations but are not a general purse for number governance.

Ring-fencing matters because otherwise the apparent size of ICANN could support exaggerated claims in both directions. A critic might count application money as proof that the number function is awash in domain wealth. A manager might cite total funds under management as capacity for a number initiative. The governing restrictions on each fund are more relevant than the consolidated balance.

Financial incentives appear in four recurring governance choices

The first incentive concerns agenda. Problems that affect funding forecasts or contracted parties receive repeated executive and Board attention. Number issues without a comparable revenue effect may appear episodic unless the ASO, RIRs or service measures keep them visible. A balanced agenda should therefore include service and policy obligations independent of financial materiality.

The second concerns staffing. Expertise grows where work is frequent and budgets are easy to justify. ICANN naturally employs substantial generic-names contractual, compliance and policy capacity. Number expertise may remain concentrated in PTI and a smaller set of liaison roles. That can be efficient, but proposals crossing into RIR governance should not be assessed solely by teams whose daily experience is contract supervision in names.

The third concerns evidence. Domain-name activity yields transaction counts, fee receipts, compliance cases and contract milestones. Number-policy legitimacy is often evidenced through regional consensus records and service-level reports. If senior decision-makers privilege familiar corporate measures, they may undervalue the records showing authorization in a federated community.

The fourth concerns remedy. When a domain initiative underperforms, ICANN can often renegotiate, enforce a contract, adjust a fee or reprioritize staff. When a number-governance concern arises, authority may lie with an RIR community, the NRO Executive Council, the ASO Address Council, the numbering review committee or a court in an RIR's jurisdiction. ICANN cannot turn every problem into a corporate corrective action.

These incentives are manageable if named. They become dangerous when financial scale is presented as neutral competence. The institution that can hire counsel, commission a report and host a meeting may appear to be the natural decision-maker. Governance asks a prior question: who has the right to decide?

Spending claims need a three-part test

Any claim that ICANN should act on a number-resource matter because it funds relevant work should pass three tests.

The authority test asks for the legal or agreed basis. Is the action top-level number coordination, registration service requested by the RIRs, facilitation of an affected-community policy, a related task agreed with the RIRs, an ASO responsibility or another expressly assigned duty? A general reference to Internet stability is limited public evidence where the Bylaws provide a more specific rule.

The accounting test asks what money is actually being discussed. Is it ordinary operating income, a restricted new gTLD fund, auction proceeds, the IANA budget, a dedicated RIR contribution or a shared cost? Does the claim concern direct expense, allocated overhead, reserve use or partner-supplied resources? Without that answer, "ICANN pays" can hide several unlike transactions.

The consequence test asks whether the proposed spending changes authority. Funding translation, meeting access, measurement or technical implementation may strengthen an existing community decision without displacing it. Funding a body to set regional allocation policy, recognize representation without the agreed route or supervise an RIR beyond the governing documents would be different. The more a grant changes who decides, the stronger the authorization required.

Passing all three tests does not guarantee wise spending. It does prevent money from doing constitutional work silently. Board rationales should publish the answers whenever a significant number-related initiative is approved.

Better accounts would protect both the names and numbers communities

The strongest reform is an annual identifier-class financial statement. It should reconcile to the audited accounts and show funding, direct expense and allocated shared cost for generic names, country-code support, IANA naming, IANA numbering, protocol parameters, root-server work and genuinely shared functions. The method should remain stable enough for year-to-year comparison.

Second, ICANN and the NRO should publish a joint resource map. It should record not only the $0.8 million contribution but also which party pays for the ASO secretariat, regional policy development, council support, service review, travel and technical coordination. The map would not merge their accounts. It would make the institutional bargain visible.

Third, the IANA plan should identify service-family capacity and material investments. Shared technology can remain shared, but expected benefits, requesting communities and allocation assumptions should be stated. A reader should be able to see whether a major increase addresses naming demand, number-service resilience, protocol work or common security.

Fourth, Board papers on number matters should include a mandate statement before the financial-impact statement. The mandate statement should cite the Bylaw clause, global policy step, ASO recommendation or RIR agreement relied upon. If multiple authorities apply, each should be named.

Fifth, the ASO and RIRs should answer financial questions with evidence rather than treating them as encroachment. Autonomy is strengthened when communities can show the resources they supply, the policies they authorize and the service results they monitor. A distributed model cannot rely on outsiders intuiting its costs.

Finally, ICANN should publish the limits of its identifier-class figures. Allocation assumptions, ranges and unavailable data should be explicit. Honest uncertainty is better than a precise total built on undisclosed choices.

Independence requires more than a balanced spreadsheet

Even excellent cost reporting cannot remove every incentive. A corporation will still devote attention to its largest recurring income source. Contracted parties will still organize around decisions affecting their costs. Number communities will still depend on an operator housed within the names-funded corporation. Governance must therefore supplement accounting with institutional safeguards.

Conflict disclosures should identify financial and professional relationships relevant to a number matter, not merely direct personal gain. Decision records should show whether directors with current or recent roles in registries, registrars or RIRs participated and how the conflict policy was applied. Recusal should be specific rather than assumed from constituency origin.

Affected-community consent should be documented. For operational changes under the numbering agreement, the record should identify RIR consultation and the review route. For global policy, it should link the five regional outcomes and the ASO verification. For wider corporate actions, it should explain any ASO participation through the Empowered Community.

Independent evidence should remain available when institutional interests diverge. Service measures, audits and published legal analysis can test claims made by ICANN or the RIRs. Independence does not require an outsider to replace every decision-maker. It requires that reasons can be checked without accepting the decision-maker's financial interest as proof.

Exit and continuity planning are the final safeguard. The numbering agreement's successor-operator possibility should be treated as a serious continuity question, not a threat. Periodic portability exercises can identify records, credentials, dependencies and transition time without suggesting an imminent move. A service that can be transferred under agreed conditions is less vulnerable to the operator's funding dominance.

Number authority should never follow funding convenience

ICANN's domain-name revenue base is not a scandal concealed in its accounts. It is the visible result of a corporate model in which generic registries and registrars pay contractual fees while number governance remains largely distributed through the RIR system. The model can work, but only if its asymmetry is treated as a design constraint.

The FY27 budget shows concentration with unusual clarity. Nearly all ordinary operating funding comes from the generic domain-name sector. The RIR contribution is small. The integrated IANA budget does not yet show a complete identifier-by-identifier cost allocation. Those facts justify scrutiny of attention, cost attribution and expansion incentives.

They do not justify a constitutional shortcut. The affected number communities develop policy. The ASO verifies and carries global results. The Board performs the review assigned by the agreed route. PTI supplies registration services under obligations that the five RIRs can measure and enforce. None of those roles expands because domain transactions rise, a reserve grows or ICANN can afford additional staff.

The names sector deserves proof that its fees are spent responsibly. The number community deserves proof that financial dependence is not becoming policy dependence. ICANN deserves a funding model strong enough to sustain all of its assigned duties without being accused, on the basis of gross totals alone, of owning every function it supports.

The answer is not to pretend the money is balanced. It is to make the imbalance auditable and keep authority anchored in the governing bargain. Finance can enable number coordination. It cannot confer it.

Sources