The Essential-Facility Remedy That Preserves Competition

Summary

• Essential-facility liability is exceptional and jurisdiction-specific. European Union law applies a strict indispensability, elimination-of-competition and objective-justification inquiry to an outright refusal under Bronner, while United States law is deeply cautious about compulsory dealing and judicial administration. Dependence alone does not establish an infringement.
• Internet number governance contains two distinct layers. Uniqueness coordination must prevent duplicate issuance and preserve authoritative history. Registration services around that shared state can still support competition in service quality, verification, support, portability, security and price.
• The first remedy should be non-discriminatory interoperability: the incumbent must publish usable technical and contractual terms, offer competitors the same timeliness and information it gives its own operations, and permit qualified providers to read and submit the records necessary to serve holders safely.
• Portability is the decisive competitive right. A holder should be able to change its qualified registration provider while retaining its number resources and continuity, subject to authentication, dispute protection and synchronized confirmation that prevents two providers from acting as current authority at once.
• Functional separation is appropriate when an integrated incumbent can still discriminate through hidden priorities, degraded access, bundled fees or confidential information. Separate staff, accounts, decision rights and oversight can protect rivals without transferring the entire institution to the state.
• Structural separation or public takeover should be a last resort. A remedy order needs measurable outcomes, security exceptions, price principles, independent monitoring, rapid dispute resolution, transition testing and a sunset review so intervention opens competition without becoming permanent detailed management by a court.

The remedy question begins after the metaphor

“Essential facility” is a powerful phrase because it compresses an economic argument. One undertaking controls an input that rivals need, reasonable duplication is unavailable, access is denied or degraded, and competition in another activity suffers. The phrase is also dangerous because it can make the conclusion sound self-proving. A facility can be important without being legally indispensable. A disappointed competitor can prefer access without competition as a whole being threatened. A court can identify exclusion yet lack the information to set a safe price or design daily operations.

The remedy inquiry must therefore begin with a precise infringement. What service is controlled? Which market depends on it? What alternatives exist? Is the complaint about outright refusal, discriminatory terms, delay, degradation, tying, denial of portability or exclusion from a decision forum? Did the incumbent create the facility through risky investment, inherit a coordination role, receive exclusive recognition or combine several functions over time? What objective security, capacity or integrity reasons justify limits?

These questions matter acutely for Internet numbers. RFC 7020 describes the Internet Numbers Registry System and its role in allocation and registration of globally unique number resources. The non-duplication requirement is not an ordinary commercial preference. Competing providers cannot each issue contradictory current claims and ask networks to choose casually. A remedy must preserve a common, verifiable result.

That constraint does not prove that one integrated institution must supply every surrounding service. It defines the technical invariant around which competition should be designed. The remedy opens access to coordination under equal rules while preventing duplicate authority. The goal is competitive registration, not competitive inconsistency.

Liability is exceptional, and the legal tests are not interchangeable

European Union competition law prohibits abuse of a dominant position under Article 102 TFEU. The official EU treaty text includes unfair conditions, limitation of markets or technical development, and dissimilar conditions that place trading parties at a competitive disadvantage. Dominance itself is not unlawful; abuse must be established.

For an outright refusal to grant access, Bronner set a demanding standard. Later EU judgments summarize the conditions as indispensability, a refusal likely to eliminate all competition by the requesting undertaking and absence of objective justification. Indispensability asks whether realistic alternatives exist, not whether access would merely be more convenient or profitable.

The strictness protects long-term competition. If every successful facility must immediately be shared, firms may invest less in building difficult infrastructure. Courts also risk converting competition law into continuous sector regulation. A registry remedy must take those concerns seriously even where the historical role was coordinated, nonprofit or community recognized rather than built for conventional investor return.

United States law is more skeptical still. The Federal Trade Commission's guidance on refusals to deal explains the general freedom to choose business partners and the limited circumstances in which monopoly maintenance can create liability. Verizon v Trinko emphasized the difficulty of compulsory dealing and noted that the Supreme Court had not recognized an essential-facilities doctrine. A regulatory mechanism already able to compel sharing also weakens the case for an additional antitrust duty.

Australia supplies a different model. The National Access Regime described by the ACCC creates routes to reasonable access for certain nationally significant infrastructure services, including declaration, undertakings and arbitration. It shows that access can be governed through explicit legislation rather than stretched from a general judicial doctrine. None of these regimes automatically governs a global number registry; together they reveal why the legal route must be identified before the remedy is designed.

The facility must be defined as a service, not an institution's entire identity

A court that defines “the RIR” as the facility has already made the remedy too broad. An institution may perform allocation, registration, public directory, reverse DNS, routing-security, membership, policy, training, measurement, grants and outreach functions. A competition concern about access to current registration does not establish that every function must be opened, separated or controlled.

The relevant facility could be narrower: the authoritative confirmation that a number resource is uniquely associated with a recognized holder; the ability to submit and validate a transfer; the interface needed to maintain current registration; or access to a security entitlement tied to recognized status. Each supports different downstream competition and carries different risks.

The claimant must also identify the dependent market. A competing registration provider may need coordination access to serve holders. A transfer facilitator may need timely status confirmation but not authority to write current records. A security service may need validated public data but not confidential account material. An operator may need portability to escape poor service rather than wholesale access to serve rivals.

Narrow definition prevents opportunistic bundling. The incumbent cannot argue that because one sensitive function should remain restricted, every service must remain exclusive. The rival cannot demand confidential or security-critical access unrelated to competition. The court can connect each obligation to demonstrated foreclosure.

It also improves administrability. An order can specify which requests, fields, acknowledgements and time limits must be available rather than requiring the incumbent to “share the registry” in the abstract. Security exceptions can attach to particular operations. Prices can reflect the actual incremental and common costs of the opened service. Monitoring can test measurable parity.

Uniqueness is the invariant; institutional monopoly is not

The Internet needs a coherent answer about whether a number resource has already been issued and which current registration is recognized within the coordinating service. That does not require every interaction to occur through one customer-facing organization. Other industries separate shared coordination from competitive service without allowing duplicate current rights.

Payment systems can support multiple banks while using common clearing and settlement arrangements. Domain registration allows multiple registrars to serve registrants under common registry rules, although the analogy has limits. Electricity markets can separate network access from generation and retail. Securities systems distinguish intermediaries from final settlement. These examples do not transfer mechanically, but they demonstrate a design pattern: competition around a shared authoritative core.

For numbers, the shared core should perform atomic confirmation. A change becomes current only when the common coordination function verifies that the prior state, authorization and proposed new state are consistent. Two providers cannot simultaneously complete conflicting changes. Every accepted act receives a verifiable sequence and history. Providers compete in obtaining and validating instructions, customer support, security, evidence handling and complementary services.

The incumbent may claim that integration is necessary for accuracy. It should prove which link is technically necessary. If the same organization must operate the final uniqueness check, competitors can still obtain equal interface access. If independent operation is feasible under common standards, the remedy can assign the neutral core to a separated entity. The analysis should follow evidence rather than institutional tradition.

The rival must accept reciprocal discipline. It cannot issue provisional claims as though they were final, conceal failed updates or continue acting after a holder switches. Qualification, logging, financial responsibility and security testing are legitimate conditions when applied equally. Competition does not mean unverified access to a critical coordination function.

Foreclosure often appears as degradation rather than a closed door

An incumbent rarely needs to say “no competitor may enter.” It can provide nominal access that is slow, incomplete, expensive or unpredictable. A rival may receive an interface without timely documentation, a transfer channel without status visibility or public data without the private confirmations needed to resolve errors. The result is constructive exclusion.

The United States Department of Justice's current Merger Guideline 5 discussion lists ways a firm can limit rivals' access: denial, degradation, worse terms, restricted interoperability, lower reliability and delayed information. Although merger review is a different legal setting, the list is useful for diagnosing how an integrated registry could favor its own service without an explicit refusal.

EU case law also distinguishes outright refusal from unfair access terms. In Slovak Telekom, the Court examined conduct around local-loop access in a regulated telecommunications setting and explained why not every constructive refusal is assessed identically to a pure refusal under Bronner. A number-registry case would require its own facts, but the lesson is that form should not hide disadvantage.

Registry degradation can include unexplained queue priority, repeated requests for documents not demanded from the incumbent's own service arm, delayed confirmation, selective maintenance windows, incompatible authentication, hidden changes, bundled fees, incomplete history, discretionary qualification or denial of routing-security functions needed to offer a credible service.

The remedy should therefore require parity of outcome and input, not merely a formal connection. Rivals need the same material information, service levels, change notice, security support and error correction that the incumbent supplies internally, subject to genuinely different risk. Performance data should be measured by request type and provider so discrimination becomes observable.

The objective is competition, not comfort for a particular competitor

Competition law protects competition, consumer welfare under the applicable legal standard, innovation and market opportunity; it does not guarantee every rival's business model. A remedy should be granted only where opening access is likely to improve competition rather than transfer value to one claimant.

The authority should ask whether multiple qualified providers can enter, whether holders can switch, whether service quality or price can improve and whether the remedy reduces the incumbent's ability to exclude. If only one applicant could use the access and no broader qualification route exists, the order may create a bilateral privilege rather than a market.

Demand matters. Operators may value a new provider for multilingual support, regional expertise, stronger security, faster transfer handling, better evidence services or lower fees. The remedy should permit differentiation in those dimensions. It should not standardize every customer experience until rivals become identical contractors.

Entry also needs viable scale. A provider should be able to aggregate customers across appropriate regions or service categories without obtaining separate bespoke permission for each request. Qualification requirements should be published and proportionate. Insurance, capital or technical conditions should correspond to measurable risk rather than incumbent size.

The incumbent's legitimate interests remain. It should receive cost-based compensation for neutral coordination services, protect confidential information, enforce security standards and prevent duplicate or unauthorized acts. The order must not require it to subsidize rivals' optional products or surrender unrelated intellectual property. Competitive access is reciprocal responsibility, not confiscation.

The first remedial rung is transparency with a reference offer

Before ordering separation, an authority should require a complete public reference offer for essential coordination access. The document should define services, eligibility, technical interfaces, authentication, security requirements, prices, service levels, maintenance, change control, confidentiality, liability, suspension, termination and dispute resolution.

Transparency reduces two forms of exclusion. Rivals can assess entry without months of private negotiation, and the incumbent cannot change terms selectively after seeing a competitor's plan. A published offer also gives the court and monitor a baseline against which to test parity.

The reference offer should separate mandatory coordination from optional products. A rival should not have to buy training, analytics, membership benefits or unrelated support to obtain essential access. Fees should be itemized. Discounts available to the incumbent's own service arm or favored partners should be visible and objectively available on equivalent terms.

Change control is critical. The incumbent should publish proposed interface and rule changes with adequate testing time, except for narrow security emergencies. Rivals should receive the same documentation and test environment as internal teams. Emergency changes need prompt explanation and retrospective review.

Transparency alone will not cure strategic delay. The order should make the offer enforceable, allow complaints and require version history. Material ambiguity should be resolved by an independent body rather than by the incumbent's unilateral interpretation. If the reference offer conflicts with the substantive non-discrimination duty, the duty prevails.

This first rung is relatively light because it preserves integration. It should be tried where opacity and bargaining asymmetry are the main concerns. If data show persistent unequal treatment, the remedy escalates.

Non-discrimination must compare the incumbent with its own service arm

A general promise to treat competitors fairly is difficult to enforce. The comparison should be explicit: the incumbent must provide a qualified rival with access equivalent in timeliness, quality, completeness, reliability and support to what it provides its own downstream or customer-facing operations.

Equivalence of input is strongest where the incumbent's service arm and rivals perform the same task. They should use the same interface, validation rules, maintenance notices and queue. If internal use requires a different route for legitimate technical reasons, the monitor should be able to compare outcomes and inspect the justification.

Equivalent treatment does not require identical handling of unequal risk. A provider with repeated credential failures may face enhanced controls. The incumbent must identify the risk rule in advance where possible, apply it consistently and offer cure and review. “Security” cannot remain an unreviewable category that appears only when a rival gains customers.

Non-discrimination also covers information. The incumbent should not use confidential data obtained through the neutral function to target a rival's customers, anticipate its products or disadvantage its negotiations. Access controls, staff separation and purpose limitations can protect commercially sensitive information.

Service-level metrics should include acknowledgment, validation, completion, rejection, error correction, outage, support response and change notice. Results should be reported separately for the incumbent's own arm and every provider, with confidential customer data protected. Persistent unexplained gaps trigger presumptions, credits or escalation.

The remedy should prohibit retaliation. A holder that chooses a rival must retain equal access to public registration, reverse-DNS administration, routing-security services and review, subject to the same rules. The incumbent cannot make switching economically impossible by withdrawing complementary functions that are part of the essential service.

Interoperability turns access from permission into capability

Legal access is useless if systems cannot exchange authenticated instructions and confirmations reliably. Interoperability is therefore the central conduct remedy. It defines the minimum common functions that let qualified providers serve holders while a shared core preserves uniqueness.

The standard should cover identity and authority assertions, resource identifiers, current-provider status, proposed changes, consent, dispute holds, completion, rejection reasons, historical confirmation, revocation status, security events and error correction. It should specify secure transport, cryptographic authentication, replay protection, time synchronization and version negotiation without prescribing every provider's internal technology.

Standards should be openly documented and implementable without discriminatory licensing. A neutral test suite should verify conformance. Providers need a test environment that behaves like the live service without exposing real holder data. Conformance results and material exceptions should be published.

Governance of the standard must not remain solely with the incumbent service arm. Qualified providers, operators, security experts and affected users should participate under conflict rules. The final uniqueness requirement is non-negotiable, but implementation choices should be contestable where they can favor one provider.

Interoperability includes semantic consistency. Every provider must mean the same thing by current authority, pending change, dispute hold and completed transfer. Ambiguous status can create competing claims even if messages travel correctly. Definitions, state transitions and error meanings require shared agreement.

The remedy should be technology-neutral over time. It can specify outcomes, compatibility and change discipline while allowing improved protocols. A sunset review can retire obsolete requirements only when replacement preserves equal access and portability.

Portability is the remedy that creates a market

Access allows a rival to connect. Portability gives the holder power to choose. Without it, competition remains theoretical because an operator cannot leave the incumbent without surrendering the resource, disrupting security services or reopening settled entitlement.

A portability right should allow a holder to move its registration service relationship to any qualified provider while retaining its current number resources, historical continuity and associated essential functions. The switch changes the service provider, not the unique allocation. The shared core records one current provider at a time.

The transfer needs strong authentication. The current provider receives notice and can raise defined objections such as fraud, a binding court order, an unresolved authority dispute or unpaid charges directly tied to permitted switching conditions. It should not block portability because the holder buys fewer optional services or criticizes the incumbent.

Time limits make the right real. Routine uncontested switches should complete within a published period. Silence cannot become indefinite veto. Objections receive rapid independent review. Emergency security concerns can pause the act narrowly, with expiry and reasons.

Continuity must cover linked services. Existing public registration, reverse DNS and routing-security authorizations should remain valid through a coordinated transition, unless the holder requests change or a separate lawful basis requires it. Credentials can be reissued through dual control without creating overlapping power to alter the same record.

Exit data should be portable in an authenticated, human-readable and machine-readable form. It should include current details, authority evidence, history needed for service, pending disputes and effective restrictions. The outgoing provider retains records required by law and accountability but cannot withhold the operational material needed for continuity.

Portability must never permit duplicate allocation

The strongest objection to provider competition is the risk that two firms recognize incompatible states. The remedy must answer that objection directly rather than dismissing it as incumbent rhetoric.

Every provider-switch or holder-transfer act should use an atomic commit at the shared uniqueness core. The act checks the current state, verifies authorization, records the new state and invalidates the prior provider's authority to submit future changes. If any check fails, no partial current state appears.

Pending operations need serialization. A holder cannot switch providers halfway through an undisclosed transfer and ask both to complete it. The common service should expose pending status, permit authorized cancellation and define precedence for court holds, fraud review and ordinary changes.

History should be append-only in the institutional sense that prior states remain auditable even when corrected. That does not mean every personal contact remains public. Authorized reviewers need to reconstruct who could act, which provider submitted the change, what checks passed and when current authority moved.

Technical resilience requires independent validation, tested recovery and geographically diverse continuity arrangements. A remedy that opens customer service while leaving one untested point of failure has not preserved competition or stability. At the same time, multiple replicas must converge on one accepted state rather than becoming independent issuers.

Liability rules should allocate error. A provider that submits an unauthorized change bears defined responsibility; the neutral core bears responsibility for accepting an invalid state contrary to its checks; the holder bears responsibility for compromised credentials under fair terms. Caps should reflect control and insurability rather than eliminate accountability.

Qualification must protect integrity without recreating exclusion

Not every applicant should gain write access to essential coordination. Qualification can require organizational identity, technical capability, security controls, trained staff, incident response, financial responsibility, insurance and acceptance of audit and dispute rules. The challenge is to prevent standards from becoming an incumbent-designed barrier.

Each requirement should map to a risk. A minimum capital rule needs evidence that it supports recoverable liability or continuity. A geographic presence rule needs a legal or operational rationale. Experience can be demonstrated by personnel and testing rather than years of prior registry operation that no entrant could possess.

Applications should follow published criteria, fixed time limits and reasoned decisions. Rejection and suspension need independent review. The incumbent's customer-facing arm should satisfy the same requirements and undergo the same tests; grandfathering should expire after a transition period.

Qualification can be tiered. A provider handling only public contact maintenance presents different risk from one authorized to submit holder transfers or routing-security changes. Entrants can begin with limited functions, establish performance and expand. Tiers should not trap rivals below commercially viable service.

Reciprocal recognition across appropriate jurisdictions can reduce duplicative cost. A neutral body can certify common controls while local law addresses mandatory differences. The authority should watch for standards captured by established providers through committees, fees or proprietary test tools.

Sanctions should be proportionate. A specific interface can be suspended after a credential compromise without expelling the provider from every service. Serious or repeated misconduct can justify broader restriction, but the reasons and evidence must be reviewable. Holder continuity should be protected during provider failure.

Pricing is where an access remedy can quietly fail

An incumbent can comply with access in form and defeat it through price. A fee may be excessive, discriminatory, opaque, bundled or structured so a rival cannot compete with the incumbent's retail offer. Conversely, a price set too low can force the neutral function to subsidize competitors and discourage investment in security and resilience.

The remedy should first define the service being priced. Essential coordination charges should be separated from optional support, policy participation, analytics and community programs. Common costs can be allocated through a transparent methodology tied to use, capacity or holder scale.

Comparable users should face comparable terms. Volume discounts need objective cost justification and availability to rivals that can meet the same conditions. The incumbent's own service arm should account for the same internal transfer price. Free internal use combined with high external fees is a classic route to hidden discrimination.

Long-term price control by a court is undesirable. The order can set principles, require audited accounts and assign disputes to a specialized independent arbitrator or authority. A price cap may be necessary during transition, followed by periodic review informed by actual cost and market entry.

Quality belongs with price. A cheap service delivered too slowly to be useful is not access. Service credits and remedies should address missed levels, but they should not become the rival's main revenue. Repeated failure triggers operational correction or escalation.

The authority should also prevent margin squeeze where applicable under the governing law. It should examine whether a reasonably efficient rival can offer competitive service after paying the access charge, while respecting the distinct legal tests that different jurisdictions apply. The objective is sustainable entry, not a guaranteed profit.

Security exceptions need evidence, review and expiry

Critical infrastructure remedies must permit immediate defense. A provider may need to block a compromised credential, isolate malicious traffic or suspend an interface that threatens integrity. A rigid equal-access command should not require continuation during an active attack.

Security clauses can also become the easiest route around the order. The exception should require a specific risk, proportionate response, contemporaneous record, notice as soon as safe and review by personnel independent of the commercial service arm. Restrictions should expire unless renewed on current evidence.

The incumbent should apply equivalent controls internally. If a vulnerability affects all providers, its own arm should face the same limitation. If only one provider is affected, the reasons should identify the relevant difference without exposing exploit detail.

The monitor needs access to restricted evidence and security expertise. Public reports can state frequency, duration and outcome without revealing methods. A pattern of security holds concentrated on a growing rival deserves investigation, but statistical disparity alone does not prove abuse.

Recovery obligations should be defined. The incumbent supplies a cure path, testing and prompt restoration. A provider cannot be left suspended indefinitely because the original notice omitted the controls needed for reinstatement.

Security investments should remain fundable. Access charges can include efficient common security cost, and standards can rise as threats change. The incumbent should not use unnecessary proprietary controls to exclude, while rivals should not demand a frozen low-cost environment that shifts risk to holders and networks.

Dispute resolution must move at network speed without sacrificing fairness

An access right can be lost through delay. A rival that waits months for interface restoration or a holder that cannot switch during a commercial crisis may leave the market before final judgment. The remedy needs a specialized, rapid forum.

The forum should be able to issue interim preservation, order information, inspect confidential material, test technical evidence and require temporary access. It should publish standard filing requirements and target times by urgency. Routine service-level disputes can receive abbreviated review; suspension, discrimination and portability blocks need faster attention.

Independence matters. The incumbent cannot be final judge of whether it discriminated. A monitor can resolve operational questions, while a tribunal or court retains authority over contested legal findings and sanctions. The division of roles should be explicit.

Confidentiality must protect holder data, security and rival plans without hiding precedent. Decisions can publish the legal and technical rule with sensitive facts redacted. Repeated issues should lead to reference-offer clarification rather than endless bilateral litigation.

Costs should discourage abuse without deterring entry. The forum can shift costs for frivolous claims, require focused evidence and offer fee relief for smaller providers raising credible systemic concerns. Mediation may settle price or implementation issues but should not replace an enforceable determination where exclusion continues.

Appeal should not automatically stay every operational order. The reviewing body can preserve access under safeguards when interruption would irreversibly harm competition, or pause a dangerous order where security evidence is strong. Reasons should address both market and continuity risk.

Independent monitoring converts an order into observable behavior

Courts are not network operations centers. They should set duties and measurable outcomes, then use an independent monitor with technical, economic and governance expertise to inspect compliance. The monitor should not become the permanent manager of the institution.

The mandate should cover reference-offer terms, interface parity, service levels, internal transfer pricing, information separation, portability, qualification and security exceptions. The monitor needs access to records and staff, authority to test systems and a protected budget not controlled case by case by the incumbent.

Reports should distinguish public findings from confidential annexes. Providers and holders need enough information to trust the remedy, while attackers and competitors should not receive sensitive detail. Material noncompliance should be reported promptly rather than buried in annual statistics.

The order can use escalating consequences. A first service-level failure may require cure and credit. Repeated discrimination can shift burdens of proof, impose penalties or trigger functional separation. Deliberate concealment can support contempt or reopening of structural relief.

The monitor should also identify overreach. If a requirement creates security risk, prevents efficient improvement or imposes cost unrelated to competition, the monitor can recommend modification. Remedy governance must learn without allowing the incumbent to relitigate the infringement through every implementation choice.

Sunset provisions discipline monitoring. The authority should review whether entry occurred, portability works, concentration changed and the incumbent can still foreclose. Duties necessary for interoperability may continue as sector rules; intrusive supervision should end when competitive conditions make it unnecessary.

Functional separation is the proper backstop for persistent discrimination

If transparency, parity and monitoring fail because the incumbent remains structurally able and motivated to favor itself, functional separation becomes appropriate. The neutral coordination function and the competitive registration service remain within a broader group but operate with distinct staff, management, accounts, information access and incentives.

The neutral function should have an independent board or committee, a mandate limited to uniqueness and common services, and a duty to treat all qualified providers equally. Its budget should be transparent and funded through non-discriminatory charges. Customer-specific information should not flow to the incumbent service arm except on terms available to every provider.

Employees should not alternate casually between neutral access decisions and commercial sales. Senior appointments, bonuses and performance measures should reflect the separated mandate. Shared security and infrastructure may remain where efficient, but access and cost allocation need audit.

Functional separation is less disruptive than divestiture. It preserves technical expertise and economies of scope while reducing the channels of discrimination. It is also reversible if evidence later supports reintegration under safeguards.

The remedy needs more than organizational boxes. Decision rights, service interfaces, procurement, incident handling and dispute response must be mapped. A nominal subsidiary whose parent can override access decisions immediately is not independent. The monitor should test real behavior.

Separation should not isolate the neutral function from users. Providers, holders and operators need structured input on common standards, while adjudication of individual access disputes remains impartial. Participation should not become a route for incumbent rivals to coordinate prices or exclude new entrants.

Structural separation is justified only after a feasibility test

Divestiture or transfer of the neutral core may be necessary when integrated control makes non-discrimination impossible, prior orders have failed and a viable independent entity can operate safely. Structural relief can remove incentive and ability to favor an affiliate more cleanly than endless conduct supervision.

It also carries severe risks. Separating systems can interrupt service, divide expertise, create unclear liability and weaken incident response. A newly independent core may lack funding or become captured by its largest customers. If legal recognition or contracts do not travel, the separated entity may possess technology without authority.

The feasibility test should identify assets, staff, intellectual property, contracts, data rights, security functions, funding, governance and transition milestones. It should model failure and recovery. Holders should not bear a forced change of resource status merely because institutional ownership changes.

The buyer or successor must be neutral and qualified. Sale to the largest downstream provider simply relocates the conflict. A nonprofit, mutual or regulated utility form may fit, but legal form alone does not guarantee independence. Appointment, voting, funding and review rights matter.

Transition should use parallel testing, reconciliation and rollback. The old and new operators can compare outputs without both exercising final authority. Cutover occurs only after independent assurance that no duplicate state can emerge and emergency recovery is credible.

Structural relief should remain connected to the proven harm. There is no basis to transfer unrelated training, measurement or community programs unless their combination caused foreclosure or separation is necessary for viability. Competition remedies are not opportunities to redesign an institution without evidentiary limits.

Public takeover is not the default cure for private concentration

When a private institution controls essential coordination, nationalization can appear decisive. The state has public-law duties, taxation power and administrative machinery. But a global number service crossing jurisdictions does not fit easily within one government's control.

Public takeover can create new exclusion. Foreign operators may distrust a national authority, sanctions policy may shape access, domestic courts may assert broader reach and political priorities may change. Replacing a private monopoly with a state monopoly does not create provider competition or portability.

Government can still have a role. Legislatures can establish access rights, courts can enforce competition law, authorities can supervise neutral terms and public agencies can support continuity. States may coordinate where critical public services depend on the registry. Those roles need not include operating every registration function.

A multinational treaty body is also not automatically superior. Negotiation can be slow, representation uneven and amendment difficult. The existing technical system could remain dependent on one political settlement. Institutional legitimacy requires accountable design, not merely public status.

The least restrictive remedy is therefore preferable: open common access under enforceable non-discrimination, make holders portable, separate the neutral core functionally if needed and reserve public or structural control for proven failure. This sequence preserves plural service and limits political concentration.

Any public intervention should state its boundary. It should protect uniqueness, continuity and competitive access rather than claim ownership of number resources or authority over autonomous routing. Transparent judicial review and time limits reduce mission expansion.

A model remedial order can be precise without managing every decision

A court order should begin with findings: the defined essential service, relevant market, incumbent control, exclusionary conduct, harm to competition, absence or inadequacy of alternatives and rejection of objective justifications. Remedies should map directly to those findings.

The operative provisions can require a public reference offer within a fixed time; non-discriminatory access for all qualified providers; open technical specifications; equal service levels and information; separation of essential and optional charges; a portability right; continuity of linked services; an atomic uniqueness check; protected treatment of confidential data; and reasoned security exceptions.

The order can appoint an independent monitor and specialized dispute resolver, require audited accounts, define reporting, prohibit retaliation and set escalating consequences. It can require the incumbent service arm to use equivalent interfaces and qualification standards after a transition period.

Functional separation can be conditional. If specified parity measures fail for repeated periods, or if deliberate discrimination is found, the incumbent must implement a predesigned separation plan. Structural relief receives a separate hearing based on evidence from the conduct-remedy period.

Modification and sunset clauses matter. The authority can adjust technical details on the monitor's recommendation while preserving the competitive outcomes. Formal review after defined intervals asks whether entry, switching and service parity exist. Intrusive duties end when no longer necessary; core interoperability can migrate into durable industry rules.

The order should reject claims it does not decide. It does not authorize duplicate allocation, establish universal property rights, command networks to route, exempt providers from security requirements or make the court the daily policy body. Precision protects the remedy from both evasion and expansion.

Success should be measured through entry, switching and parity

The remedy is successful if competition becomes possible and durable, not if the incumbent produces documents. Measures should include qualified provider entry, time to qualify, customer switching, completion and failure rates, price dispersion, service innovation, concentration and exit.

Parity metrics compare acknowledgment, validation, completion, support, outages and security restoration between the incumbent's own service arm and rivals. Portability metrics show routine completion time, objections, upheld blocks, abandoned requests and continuity incidents. Interoperability metrics cover conformance, version adoption and errors.

Quality must accompany quantity. A flood of undercapitalized providers causing security incidents is not successful competition. Measures should include unauthorized changes, data exposure, recovery, complaint resolution and provider failure. Qualification can tighten if evidence shows a risk, subject to equal application.

Holder experience matters. Operators should be able to identify their provider, understand charges, obtain records, switch and appeal without specialized institutional knowledge. Surveys can supplement hard measures but should not replace them.

Innovation should remain visible. Rivals may offer better multilingual service, stronger identity verification, insurance, transaction support or routing-security integration. The neutral core should not absorb those products merely because they succeed. Its mandate remains common coordination.

The authority should publish a baseline before relief and regular comparisons after it. If no entry occurs despite compliance, the original market theory or remedy may be wrong. If entry occurs but portability remains unused because holders fear retaliation, enforcement needs deeper examination. Evidence, not attachment to the original order, should guide modification.

The Number Resource Society offers an institutional destination

A Number Resource Society can host the neutral rules needed for competitive registration without becoming a state owner of every function. It can qualify providers, maintain interoperability requirements, oversee the uniqueness core, operate independent review and publish performance.

Its constitution should separate common coordination from competitive services. The Society should not offer a favored retail registration arm unless that arm is functionally separated and subject to identical terms. Provider voting must be balanced so incumbents cannot exclude entrants and entrants cannot weaken security standards for short-term gain.

Holders need direct rights: transparent terms, service portability, access to their records, continuity during switching, reasons for denial and independent review. Those rights make provider competition meaningful rather than a negotiation among institutions.

The Society should preserve one current authoritative state and complete history while permitting multiple qualified service relationships. Its technical mandate is narrow: prevent conflicting current authority, authenticate changes, support common essential functions and recover from failure. It does not need to dictate operators' routing choices or commercial use of lawfully registered resources.

Funding should come from transparent common charges apportioned by justified measures. Optional provider services remain separately priced. Audit and public reports test neutrality, security and concentration. Courts and competition authorities retain external jurisdiction under applicable law.

This destination is consistent with the remedy ladder. A conduct order can create open access and portability now. Functional separation can prepare a neutral core. If structural independence becomes necessary, the Society provides a governed home rather than an improvised receivership.

Conclusion: open the service, preserve the invariant

The central remedial choice is not simply access or break-up. It is how to create contestable registration service while preserving the one property that cannot be compromised: globally coherent uniqueness. A remedy that duplicates authority is technically unsound. A remedy that leaves holders unable to switch is competitively empty.

The legal threshold remains demanding. Essential-facility concepts differ across jurisdictions, and an authority must prove dominance or control, indispensability where required, exclusion, competitive harm and lack of objective justification. Importance alone does not create liability. The remedy follows the infringement rather than the institutional label.

When liability is established, the first response should be a public reference offer, equivalent access, non-discrimination, open interoperability and provider portability. Qualification protects integrity if standards are objective and reviewable. Atomic confirmation prevents duplicate allocation. Transparent pricing, security exceptions, rapid disputes and independent monitoring make the rights usable.

Functional separation is the correct escalation when integration sustains discrimination. Structural separation becomes appropriate only after conduct relief fails and feasibility is proved. Public takeover is not a shortcut: it can replace commercial concentration with political concentration while leaving operators without choice.

The disciplined remedy is therefore pro-competitive and institutionally modest. It opens the common service on equal terms, lets providers compete around it, lets holders leave without surrendering continuity and keeps the court out of daily technical management. It preserves one coherent answer about unique number resources while ending the assumption that only one organization may provide every service connected to that answer.

Sources and scope

The number-registry function and uniqueness requirement are grounded in RFC 7020. The European legal discussion uses Article 102 TFEU, Bronner, Case C-7/97, Slovak Telekom, Case C-165/19 P and the Commission's Article 102 enforcement materials. Those sources require attention to the exact conduct and do not make every important service legally indispensable.

The United States comparison uses the Federal Trade Commission's refusal-to-deal guidance, the Department of Justice's archived essential-facility discussion and the current Merger Guideline 5 access analysis. The Australian comparison uses the ACCC's official description of the National Access Regime. These legal systems supply comparisons, not a universal cause of action or choice of law for number registries.

The reference offer, equivalence standard, interoperability specification, portability right, atomic confirmation, qualification tiers, price principles, security exception, monitoring, conditional functional separation, structural feasibility test and Number Resource Society design are governance recommendations. Their adoption would require authority under applicable competition law, contract, corporate instruments or legislation. No claim is made that registration status alone establishes universal ownership, that a court should direct autonomous routing decisions, or that provider competition permits duplicate number allocation.