Summary

  • The June 1998 White Paper was a United States government statement of policy. It expressly said that it was not a substantive rule, contained no mandatory provisions and did not itself carry the force of law. Agreements, corporate instruments, contracts and operational acceptance did the later implementing work.
  • Its central bargain was narrow but important: coordinate IP-number policy, the authoritative root, new top-level-domain policy and necessary protocol parameters through a private not-for-profit body, because unique identifiers require common administration and a national government monopoly was increasingly indefensible.
  • The document rejected a monolithic structure for Internet governance and said the new corporation should not expand beyond functions then exercised by IANA. That language rules out reading it as a charter for general global regulation, even though identifier decisions inevitably had economic and political effects.
  • “Global participation” did not mean a vote for every Internet user. The White Paper proposed functional and geographic representation, councils, open associations and other mechanisms; it left membership design unresolved and gave the eventual board final authority over council recommendations.
  • The initial corporation therefore borrowed legitimacy from several incomplete sources: Commerce recognition and agreements, California non-profit law, inherited technical practice, stakeholder consultation, transparent procedure and continued operational acceptance. None alone amounted to a world electorate.
  • Private coordination did not make ICANN owner of the root, domain names, IP addresses or protocol parameters. The White Paper contemplated policy and direction within recognised administrative systems but did not declare property title, create a general asset-confiscation power or displace rights available under national law.
  • Contemporary materials expose the bargain's weaknesses. Commerce's October 1998 letter flagged unresolved membership, financial accountability, transparency, conflicts, geographic representation and country-code concerns. ICANN's November changes promised minutes, reasons, membership work and independent review, while admitting that important mechanisms remained to be designed.
  • The durable legitimacy test is instrument-specific: identify the coordinated function, the rule or agreement authorising action, the affected party's procedural and national-law rights, the evidence of bottom-up support, the anti-capture safeguards, the remedy and the connection to stability or interoperability. The White Paper is a boundary, not a blank cheque.

One sentence still controls the argument

The most revealing sentence in the 1998 Statement of Policy is not its invitation to create a new corporation. It is the sentence that says the policy did not propose a monolithic structure for Internet governance. The Department of Commerce doubted that the Internet should be governed by one plan, one body or even a series of plans and bodies. What it sought instead was a stable means of addressing narrow questions around the management and administration of Internet names and numbers.

That distinction has been blurred in both directions. Admirers sometimes retell 1998 as the birth of a globally legitimate institution empowered to make whatever rules a unified Internet needs. Critics sometimes reply that ICANN was merely a private California corporation with no authority beyond ordinary contract. Each description omits an essential fact.

The White Paper contemplated a genuine concentration of coordinating authority. It said a new corporation ultimately should oversee the authoritative root, determine policy for adding top-level domains, coordinate technical parameters and set policy for and direct allocations of IP number blocks to Regional Internet Registries. Those functions affect market entry, reachability and institutional status. “Technical” did not mean trivial.

Yet the same text limited the project. It tied central coordination to unique identifiers and universal connectivity. It said the new corporation's responsibilities were not to expand beyond functions then exercised by IANA. It preserved applicable national law and rejected a substantive regulatory regime in the instrument itself.

The bargain was therefore narrower than government and broader than clerical service. Understanding that middle category is the only sound way to judge what Commerce offered, what ICANN's founders built and what later institutional legitimacy can honestly claim from 1998.

Read the legal label before the political language

The White Paper's formal status matters. In its administrative-law section, Commerce said that public comments had led it to remove aspects of the earlier proposal and issue a general statement of policy rather than impose a substantive regulatory regime. It then stated that the policy was not a substantive rule, contained no mandatory provisions and did not itself have the force and effect of law.

This is not a footnote to be brushed aside by the document's historical influence. It tells us how the bargain was expected to become operative. Commerce would recognise an appropriate private body by entering agreements. Existing government agreements with Network Solutions and other actors would be altered. Databases, software and expertise would be made available. The new corporation would adopt articles, bylaws and policies. Registries, registrars and users would encounter obligations through contracts and recognised administrative arrangements.

The White Paper could shape those instruments without itself being one. It stated the government's selection principles and transition intentions. It invited private stakeholders to form a body. It described functions, structure, representation and safeguards that Commerce would consider before agreement. Political weight and legal force were therefore different.

That distinction disciplines claims on both sides. It is too strong to say that one policy notice directly vested a private corporation with sovereign jurisdiction over the world's Internet identifiers. It is too weak to say that the notice did nothing because it was not binding law. A government that controlled or contracted for important administrative dependencies announced how it would use agreements, recognition and transition arrangements. Private actors organised in response. The result could change practical authority without creating a statute.

A later decision should consequently identify its nearer legal and institutional basis. Was it authorised by ICANN's articles or bylaws, a registry or registrar agreement, an IANA arrangement, an adopted policy, a national law or a specific government direction? Invoking “the White Paper” alone skips the instrument that made the power enforceable.

The inherited system was neither private innocence nor settled sovereignty

In 1998 the identifier system was already distributed across public contracts, research institutions, private companies and technical communities. The White Paper described IANA under United States government arrangements, Network Solutions operating key generic-domain registration and root-zone functions under a National Science Foundation cooperative agreement, and Regional Internet Registries administering address blocks downstream from IANA.

This history complicates any clean privatisation story. Commerce was not taking a conventional government agency and selling it. Nor was it leaving an entirely voluntary club untouched. The government had contracts, funding history and leverage over functions on which global users depended. Technical legitimacy came from long practice and community acceptance; practical control also passed through United States institutional arrangements.

The arrangement had become harder to defend. Commercial use had grown. Network Solutions' position attracted complaints about competition. Trademark disputes were increasing. More users and operators lived outside the United States. Decisions about adding top-level domains could no longer plausibly rest on ad hoc personal authority without formal accountability. Research-agency sponsorship looked increasingly mismatched to a commercial global medium.

Pre-existing norms were themselves bounded. RFC 1591, published in 1994, described domain delegation as a public trust, expected a designated manager to serve the community and placed significant responsibility on the IANA administrator. It was an operational policy document, not a democratic constitution. Its legitimacy relied on expertise, stewardship, service and acceptance.

Commerce therefore faced a problem with no pure baseline. Keeping direct United States stewardship would preserve identifiable public accountability in one state while deepening a national monopoly over a global system. Handing functions to a private body could internationalise participation and reduce day-to-day government control while raising questions about electorate, corporate accountability and private coercion.

The White Paper's narrow bargain was an attempt to trade one concentration of power for a constrained coordinator, not to abolish power.

Four functions defined the proposed centre

The document's scope is clearest when broken into the functions it said required coordination. The first concerned policy and direction for allocation of IP number blocks to Regional Internet Registries. The second concerned oversight of the authoritative root server system. The third concerned policy for determining when new top-level domains enter the root. The fourth concerned assignment of other technical parameters needed for universal connectivity.

Each function involves uniqueness. Two incompatible allocations of the same address block create conflict. Divergent authoritative roots can send the same name to different destinations. A protocol parameter used with inconsistent meanings can break communication. Coordination is not merely a preference for tidy administration; it helps preserve a common namespace and interoperable network.

The functions are nevertheless different. Address allocation was already hierarchical and regional. Root-server operation was dispersed even though authoritative change required coordination. New top-level-domain policy involved market structure, intellectual property and competition. Protocol-parameter assignments sat within standards relationships, and commenters warned that the Green Paper had overstated IANA's role relative to the IETF.

The final White Paper acknowledged the boundary. It said coordination of protocol parameters should preserve stability and interconnectivity, but expressly disclaimed expansion beyond IANA's then-current functions. That correction is important evidence of public comment doing substantive work. The proposed centre was not entitled to absorb adjacent standards powers simply because they involved Internet technology.

The same reasoning should apply within each category. Authority to maintain uniqueness does not automatically include authority to regulate content carried over an address, decide national speech law, direct business use of a domain, claim beneficial ownership of a number resource or govern unrelated cybersecurity conduct. A restriction needs a closer connection to the coordinated function and a proper source of authority.

The four functions created a centre. They also supplied the first boundary around it.

Coordination solved a real collective-action problem

It is fashionable to describe every central Internet function as an avoidable hierarchy. The White Paper was more realistic. Some common decisions had to be made if users were to experience one interoperable system. The relevant question was who would make them, under what procedures and with what limits.

A single authoritative root provides coordination around the set of top-level delegations. Allocation hierarchies prevent simultaneous official registration of the same number resource to different parties. Parameter registries let independent implementations attach the same meaning to protocol values. None requires one institution to run every server or network. They require authoritative agreement at defined joints.

Alternative roots and private registries existed and contributed to debate. Their existence did not eliminate the economic value of common recognition. If most networks, applications and users consult one root or allocation hierarchy, inclusion in it has practical force. The coordinator can therefore influence market access even if no police power compels technical adoption.

The White Paper's answer was to place limited central functions in a private not-for-profit body rather than continue direct national-government control. That choice avoided converting every change into an act of United States administration. It also sought flexibility and speed. A private body could consult technical and commercial communities, form specialised councils and adapt without a full legislative act.

But privatisation moves accountability rather than making it unnecessary. A government official is constrained by public law, elections and oversight within one polity. A private coordinator serving global users cannot rely on those exact mechanisms. It needs transparent procedure, representative structure, contract limits, review, competition where possible and continuing legal remedies. The White Paper listed these substitutes without proving they would be sufficient.

Coordination was justified by the technical joint. Institutional legitimacy depended on preventing that joint from becoming a general gate over the rest of the network.

The Green Paper debate narrowed the instrument

The January 1998 Green Paper was more prescriptive. Commerce proposed a rule and invited comment on detailed provisions, including corporate membership, a transition-period number of new generic top-level domains and trademark procedures. The final document recorded that those elements had been removed in response to comments and changing technology.

Commerce reported receiving more than 650 comments amounting to roughly 2,000 pages on the Green Paper. Its June testimony characterised support for a private not-for-profit coordinator and the four functions as overwhelming or near consensus. That is evidence of substantial engagement. It is not a census of Internet users, countries, networks or affected rights holders. There was no defined world electorate, eligibility roll, turnout figure or equal ballot.

The White Paper itself was more careful than the testimony's political shorthand. Its comment summaries used terms such as “many,” “most” and “a limited number” as rough descriptions, not complete quantitative analysis. It recorded objections to United States location, calls for international governmental management, concern over root authority, criticism that proposed protocol control exceeded IANA's role, and requests for human-rights and public-trust principles.

The final choice did not simply count submissions. Commerce evaluated them against its policy objectives: stability, competition, private bottom-up coordination and representation. It retained a United States-incorporated body for continuity, rejected both continued national operation and a treaty organisation, and left substantive detail to later private processes.

This was consultation as institutional design, not constituent instruction. It could improve legitimacy by revealing arguments and changing the proposal. It could not produce the democratic authorisation of people who never knew of the proceeding, lacked effective access or did not regard Commerce as their representative.

That boundary matters whenever “community consensus” is traced back to 1998. The record supports a contested, informed policy choice. It does not establish unanimous or globally measured consent.

Private did not mean owner

The White Paper preferred responsible private-sector action to government control for functions requiring coordination. That preference is often mistaken for a transfer of property. Administration, policy authority, contractual control and ownership are different legal relationships.

The proposed corporation would oversee or coordinate functions. Commerce would enter agreements and seek international support. Network Solutions would recognise the new body's policy role under amended arrangements. IANA staff and resources would support continuity. None of those steps required a declaration that ICANN owned the root, every domain, all address space or the protocol registries.

Domain registrations have economic value and may support contractual or legal claims. Address holdings can carry substantial commercial value. A coordinator can affect that value through delegation, accreditation, transfer, revocation or registration status. Consequence does not itself establish beneficial title in the coordinator.

The White Paper was explicit in one relevant area: nothing in domain registration agreements or operation of the new corporation should limit rights that registrants or trademark owners could assert under national laws. It contemplated alternative dispute resolution through registration agreements, not an abolition of courts. Incorporation in the United States was not intended to displace other countries' applicable law.

For IP numbers, the document proposed policy for and direction of allocation to Regional Internet Registries. It did not define number resources as property, create a power of compulsory acquisition, specify compensation for deprivation or settle contractual rights between an RIR and a holder. Those questions require other instruments and applicable law.

Calling the coordinator a private body therefore did two things: it identified legal form and rejected direct government administration. It did not transform the coordinated subject matter into the corporation's estate. A steward can exercise consequential administrative authority without owning the thing administered. That is precisely why procedural limits and review are necessary.

Recognition was an agreement, not an election

Commerce did not wait for a worldwide vote to identify the new corporation. The White Paper said the United States was prepared to recognise an appropriate body by entering an agreement and seeking international support. Private stakeholders were invited to organise it; Commerce would assess commitment to the stated principles.

ICANN filed articles in California on September 30, 1998. Its initial directors were selected during formation, not elected by all Internet users. Commerce reviewed ICANN's submission, competing proposals and public comments. In October it called ICANN a significant step but identified serious concerns. In November ICANN amended its governing documents and sought the transition agreement.

The November 25 Memorandum of Understanding then constituted an agreement between Commerce and ICANN. It did not recite a popular mandate. It said Commerce had determined that the project could be done most effectively with ICANN because ICANN had the relevant purpose and best demonstrated an ability to accommodate broad and diverse interests. The parties would jointly design, develop and test transition mechanisms.

Recognition supplied practical and governmental legitimacy. Commerce had authority over its contracts and transition choices. ICANN had corporate capacity under its articles and bylaws. The agreement connected the two. International input and later operational acceptance could broaden legitimacy, but they did not retroactively turn recognition into election.

This selection method was defensible only as an interim and conditional arrangement. Stability made a sudden global constitutional convention implausible. Someone had to choose a body with technical and organisational capacity. The White Paper therefore demanded evolving governance, geographic and functional diversity, open processes and protections against capture.

The problem appears when temporary recognition is narrated as permanent constituent power. Commerce could choose a counterparty for functions under its influence. It could not cast ballots for the rest of the world or confer powers that it did not possess. ICANN's durable authority would have to be renewed through performance, bounded rules, participation, contracts and acceptance.

Global participation was not a global electorate

The White Paper wanted international participation and a board broadly representative of the global Internet community. It proposed representation of number registries, domain registries, registrars, the technical community, service providers, commercial and non-commercial users and individuals. It expected geographic diversity and restricted voting government officials to advisory roles.

Those are serious commitments. They acknowledge that a corporation established under one country's law needed input beyond that country's firms and officials. They also reject a board captured by a single technical, commercial or national faction.

But representation was functional as well as territorial. The paper spoke of membership organisations in names, numbers and protocols, direct user interests, councils and open associations. Directors might be elected from membership or through other mechanisms ensuring broad representation and participation. It did not specify universal suffrage, population-weighted national delegations or one-network-one-vote.

That ambiguity was deliberate and costly. A world electorate is difficult to define. Is the voter a person, domain registrant, address holder, company, network, government or technical contributor? Individuals can hold many domains; companies can control millions of users; most end users never contract with ICANN. Identity and turnout at global scale would be expensive and vulnerable to capture. Functional representation can bring expertise but gives organised incumbents structural advantages.

The White Paper left these trade-offs to the interim institution. Commerce's October letter then warned that ICANN's proposed board was merely encouraged, not required, to establish open membership. Many commenters believed bottom-up coordination and representation would be unlikely without some membership-based structure. ICANN's November response promised an advisory committee to develop membership criteria, rights and responsibilities - an admission that the electorate was not yet designed.

The founding bargain promised a route towards representation, not proof that representation had already been achieved.

“Consensus” was an anti-capture device, not unanimity

The White Paper said supermajority or even consensus requirements might protect against capture by a self-interested faction. It also proposed separate robust councils for names and numbers, open petitions, transparent decisions and a final board authority to approve or reject recommended policy.

Consensus in this setting was procedural. It meant more than a simple majority of whoever appeared at a meeting, but it did not mean every affected Internet user agreed. The paper did not define a fixed threshold, eligible population or veto for each constituency. Its concern was that a narrow faction could exploit the coordinator's unique position.

This creates a tension. High thresholds can block capture, but they can also entrench incumbents. A registry coalition might prevent competition. A vocal minority might describe delay as lack of consensus. Staff might convert an ambiguous record into a preferred interpretation. A board might cite broad support while disclosing neither dissent nor the weight assigned to each group.

The White Paper's answer was not consensus alone. It coupled consensus with open and transparent decisions, recorded reasons, petition rights, diverse councils, conflict protections and board responsibility. The board retained final authority, which meant it could not outsource accountability to a vague community mood.

A legitimate consensus claim therefore needs an issue definition, eligible affected groups, accessible notice, evidence of views, treatment of objections, conflict disclosures, a reasoned account of convergence and a review route. It should state whether support is among contracted parties, a supporting organisation, meeting entities, commenters or a wider measured population.

The founding text warned against capture but did not solve the measurement problem. It would be a misuse of 1998 to treat the word “consensus” as a source of unlimited policy authority. The stronger reading is narrower: where unique coordination gives one body unusual leverage, decisions need broader and more inspectable support than ordinary private management.

The board was final, and therefore accountable

The White Paper imagined policy councils developing and recommending matters within their competence. It then said the elected board should have final authority to approve or reject council recommendations. This was not a pure federation in which constituent groups retained sovereign vetoes.

Final board authority can protect the institution from a supporting organisation that serves its own commercial interest. It can reconcile conflicts between stability, competition and representation. It can ensure that one specialised group does not impose costs on the rest of the system. It also creates the risk that a small board will invoke the Internet community while departing from the people closest to the issue.

The paper therefore demanded sound, transparent and publicly recorded decisions. It expected the board to evolve as the Internet constituency changed. It discouraged government voting control while permitting advisory participation. It treated conflict and geographic balance as design requirements.

ICANN's November 1998 amendments made some of those ideas concrete. The corporation promised open public meetings associated with regular board meetings, roll-call voting on important matters, timely minutes showing votes and reasons, stronger conflict controls, geographic diversity and work towards independent review. The formal board meetings themselves were not planned to be open, reflecting a claimed need for candid deliberation; publication and public sessions were the compensating controls.

These controls mattered because the board did not answer to a world parliament. Reasons let outsiders compare action with articles, bylaws and stated principles. Conflicts reveal whose interest shaped a decision. Independent review provides a path from disagreement to institutional remedy. Evolution prevents the initial founders from becoming a hereditary class.

The legitimacy of final authority thus depended on observable restraint. The more consequential the decision, the less sufficient it was for the board to say that a council, staff report or general consensus required it.

California incorporation supplied law, not universality

The White Paper expected the new body to be headquartered and incorporated in the United States as a not-for-profit corporation. It justified that choice through continuity: the relevant functions and expertise were then concentrated there. It also said United States incorporation was not intended to displace applicable law in other countries.

Corporate form supplied essential things that “the Internet community” could not. ICANN could own property, employ staff, sign contracts, sue and be sued, adopt bylaws, appoint directors and operate under a recognised legal order. Fiduciary and procedural rules attached to the entity. Courts had a jurisdictional basis for at least some disputes.

It did not make California law a constitution enacted by global users. Foreign states retained law over conduct and rights within their jurisdiction. Contract parties could invoke agreed forums and remedies. Country-code managers and national governments had relationships not reducible to a California corporate vote.

The White Paper also rejected a special antitrust immunity so long as the corporation's policies were reasonably based on and no broader than necessary for legitimate coordinating objectives. That sentence is a powerful scope constraint. A private coordinator could not justify every restraint by calling itself a technical authority. Competition remained one of the founding principles, and ordinary law remained relevant.

The November 21 amendment to ICANN's articles said the corporation would operate for the benefit of the Internet community as a whole, conform to relevant international principles and applicable conventions and local law, and use open and transparent processes enabling competition and entry. This was a corporate commitment, not a grant of sovereignty. Its value lay in creating a standard against which corporate action might be reviewed.

Legal personality made coordination executable. It did not make the corporation the legal representative of humanity.

Commerce did not disappear at the moment of recognition

The transition is often described as if the United States published a principle and stepped away. The 1998 documents show a more conditional arrangement.

The White Paper expected agreements between government and the new corporation, a ramp-down of Network Solutions' cooperative agreement, transfer or access to data and software, consultation with other governments, and a security review of the root system. It described actions that only government or existing contractors could take.

The November MOU was explicitly a joint project. Commerce required assurance that the private sector had the capability and resources to assume the responsibilities. The parties would design, develop and test mechanisms, methods and procedures, then document the results. Commerce would provide expertise, coordinate with Network Solutions, participate in root-security work, consult internationally and maintain oversight of functions performed directly or under United States agreements until further arrangements were made.

That arrangement is narrower than an immediate transfer and more substantial than passive observation. ICANN received recognition as the selected collaborator and a route towards greater management responsibility. Commerce retained leverage, oversight and control over its own agreements during the test period.

The conditional structure answers one legitimacy problem and creates another. Testing protects stability and lets a public authority require non-arbitrary procedure before relinquishing influence. Continued United States oversight means the supposedly private global coordinator is still partly dependent on one national executive.

The White Paper treated this as transition, with the government role expected to end after a stable handover. Whether and when later arrangements fulfilled that intention is a separate historical question. For understanding 1998, the important point is that “private coordination” was an objective pursued through public-private agreements, not a state of nature declared in June.

The MOU made procedure part of the deliverable

The November agreement did not limit the project to moving servers and files. It required the parties to develop mechanisms for transparent, non-arbitrary and reasonable private technical management. It said standards and practices should not be applied inequitably or single out a party without substantial and reasonable cause. It required sufficient appeal procedures for adversely affected members of the Internet community.

ICANN agreed to work on external neutral review where decisions conflicted with its bylaws, on competition in registration services, on participation by affected parties, and on methods for soliciting, evaluating and responding to comments. Commerce retained responsibility for advice on open public proceedings and general oversight during the transition.

These clauses reveal what the founding bargain considered necessary to legitimate private power. A unique coordinator cannot rely only on consent to contract because affected parties may have no practical alternative. It owes consistency, reasons, participation and appeal.

The clauses also have limits. “Sufficient” appeal was not self-defining. Review focused on conflicts with bylaws and articles, not a general appellate court for every disputed outcome. Public input did not give each commenter a vote. The agreement was between Commerce and ICANN; a third party's direct enforcement rights depended on law and contract.

Still, it is wrong to treat accountability as a later decorative addition. The 1998 implementation agreement made procedural assurance one of the things to be designed and tested. The institution had not completed that work at birth, but its need was part of recognition.

A later policy that has no accessible reasons, treats equivalent parties differently or lacks meaningful review cannot defend itself simply by pointing to the White Paper's preference for private action. The preference came paired with procedural obligations.

Competition was part of legitimacy, not an optional market policy

One immediate target of the transition was Network Solutions' exclusive position in generic-domain registration. The White Paper expected the cooperative agreement to be ramped down and competition introduced. It contemplated equal access, shared registration, registrar accreditation and terms under which registries and registrars would operate.

This matters because the new coordinator itself would possess a unique role. Privatisation could remove one monopoly while creating a gatekeeper over competitors. The founding principles tried to distinguish necessary central coordination from services that could support market entry and consumer choice.

The coordinator could define technical and contractual qualifications for registrars and registries, but those conditions had to serve stability and legitimate coordination rather than protect incumbents. The White Paper's rejection of special antitrust immunity reinforced the point. A policy broader than necessary for coordination could face ordinary competition scrutiny.

Competition was not absolute. A common root cannot be duplicated within one authoritative system merely to create two providers of the same decision. Security and interoperability may justify uniform requirements. New top-level domains can impose external costs. The challenge is to locate the narrow central choice and open the surrounding services.

That design has a procedural consequence. An excluded applicant needs criteria known in advance, a record of evaluation, conflict disclosure, reasons, comparable treatment and review. A market cannot discipline the sole coordinator that decides who may enter the market. Accountability must do the work competition cannot.

The White Paper therefore offered more than a preference for private business. It offered a constitutional economy of sorts: centralise only the joint that must be common, expose the rest to entry, and constrain the coordinator through law and procedure. It was an institutional design principle, not a transfer of public control to any incumbent company.

Domain policy had consequences without becoming legislation

The White Paper devoted substantial attention to trademarks, registrant contact data and dispute resolution. It anticipated registration agreements requiring certain information, court jurisdiction for infringement disputes and alternative procedures for cyberpiracy or cybersquatting. It also preserved rights under national law.

This is a good example of the coordinator's intermediate character. ICANN would not enact criminal or property law. Yet its accreditation and registry agreements could standardise obligations across a large market. A registrant might encounter a mandatory dispute clause because registrars needed accreditation and registries needed root recognition. Private ordering could have regulation-like reach.

The proper question is not whether that effect is “government.” It is whether the rule falls within the coordinated function, rests on a valid instrument, was adopted through the promised procedure, treats parties fairly and leaves appropriate legal rights intact. The White Paper itself did not adjudicate future disputes.

The document's trademark proposals also show why narrow technical coordination inevitably touches social interests. Adding a top-level domain changes competition and naming opportunity. Registration data affects privacy and enforcement. A dispute procedure can determine who controls a valuable name. Engineers alone cannot resolve the normative choices.

That does not authorise a general social-policy mandate. It requires the coordinator to identify the operational connection and limit collateral effects. Where a national court, legislature or international agreement is the proper decision-maker, private coordination should not displace it merely because compliance can be enforced through the root or a contract.

The founding bargain accepted policy around the identifier system. It did not convert technical leverage into general lawmaking competence.

Number coordination did not settle number title

IP addresses appeared in the White Paper alongside DNS functions. The document described IANA allocating blocks to Regional Internet Registries and proposed that the new corporation set policy for and direct allocation of IP number blocks to those registries. That was not an incidental reference.

At the top of a hierarchical allocation system, coordination is necessary to avoid duplicate official allocation. Policy can decide which regional institution receives blocks and on what aggregate basis. The role has important operational consequences and can shape downstream scarcity.

But the White Paper did not define what legal interest a downstream holder possesses. It did not say that addresses were property of ICANN, the United States, an RIR or a registrant. It did not establish a worldwide transfer code, a compensation rule, a security interest, a bankruptcy priority or a power to seize holdings for unrelated conduct.

The distinction between administrative authority and asset title is essential. A land registry records title under law; it does not own every parcel. A securities registrar can change an entry under valid instruction; it does not gain beneficial ownership of the security. Number resources are not identical to either category, but the analogy exposes the missing step. Power to coordinate unique registration does not prove ownership.

Later number-resource action must rest on RIR policies, membership or service agreements, corporate succession, applicable law and the relevant IANA arrangements. If an institution cancels or transfers a registration, it should identify the rule, factual basis, authority, notice, review and remedy. The White Paper's general allocation language cannot substitute for that chain.

Nor should critics erase the coordination power that was contemplated. Direction of top-level allocation is real. The accurate conclusion is bounded: 1998 supported administration of a unique hierarchy; it did not settle every proprietary consequence of participation in that hierarchy.

Country-code domains exposed the national boundary

Country-code top-level domains made the “global but private” model especially awkward. The White Paper recognised that many ccTLDs were administered by governments or private entities with the relevant government's acquiescence. It expected international participation and said incorporation in the United States would not displace applicable foreign law.

Commerce's October 20 letter to ICANN identified ccTLD policy as an unresolved concern. It stated an assumption that national governments would continue to have authority to manage or establish policy for their own ccTLDs, except where policy adversely affected universal connectivity. ICANN's submission had been silent, and Commerce requested clarification.

This is a direct boundary around global coordination. The common root requires a record of the delegated operator and stable technical service. That need does not automatically give the coordinator authority over all substantive national naming policy. A government or local community may determine eligibility, dispute rules or public-sector use, subject to law and the technical requirements of common connectivity.

The boundary is not effortless. A redelegation dispute can involve local legitimacy, service failure, government demands and global stability at once. The root coordinator must decide which operator to list. Refusing to decide is also a decision. Narrow coordination therefore still requires criteria, evidence, notice and review.

What it cannot do is treat the necessity of one root entry as proof of general jurisdiction over the country's digital policy. The White Paper's bargain was to coordinate the common technical joint while respecting law and territorially grounded authority beyond it.

The ccTLD issue demonstrates the article's larger thesis. A single coordinator can be necessary without becoming a world government. Its legitimacy depends on knowing where technical unity ends and another institution's lawful authority begins.

Rights remained plural and enforceable elsewhere

The White Paper did not offer an Internet bill of rights. It described participation, transparent governance, registration information, contractual dispute mechanisms and access to national courts. It expected the corporate charter and bylaws to provide internal constraints.

Affected parties therefore had different rights from different sources. A registry or registrar might enforce a contract. A registrant or trademark owner might invoke national law. An applicant might seek reconsideration or independent review under ICANN's governing rules. A corporation could challenge unequal treatment under an agreement. A government could act within its legal jurisdiction. Commerce could enforce the transition arrangement.

Plural remedies are less elegant than one global court, but they fit the narrow bargain. ICANN was not given a general judiciary. Its review function would compare action with articles, bylaws and adopted procedure. Courts retained authority over legal claims. Contracts translated policy into obligations and supplied agreed dispute routes.

The arrangement risks gaps. A user may be practically affected without contractual standing. Internal review may find a procedural defect but lack power to award full relief. Litigation may be expensive or jurisdictionally uncertain. A policy may be formally valid yet oppressive to a weak constituency. Representation and public comment were meant to reduce those gaps, not eliminate them.

Legitimacy analysis should identify the remedy before praising the rule. Who can challenge? On what ground? Before which body? Can the reviewer see the full record? Can it stay a harmful act, order reconsideration, declare inconsistency or award relief? What national-law claim remains?

The White Paper's refusal to displace national rights was an important limit. It also meant that equal global treatment could not be guaranteed by one institution. The bargain accepted legal pluralism in exchange for avoiding a global sovereign.

The October letter is the missing founding document

Commerce's October 20 response to ICANN's proposal is indispensable because it records doubts before recognition. The Department called the submission a significant step and said comments generally supported moving forward. It then listed serious concerns: representational and financial accountability, transparency, conflicts of interest, geographic representation and the role concerning country-code domains.

Membership was central. ICANN's proposal encouraged but did not require an open structure. Commerce said many commenters doubted that bottom-up coordination and representation could be achieved without membership-based accountability. It asked how the board would answer to the Internet community.

Finance mattered because a private coordinator could impose costs through its unique position. Commenters worried about budget transparency and controls. Conflict rules mattered because supporting organisations might place representatives on the board and influence policies benefiting their sector. Minutes and reasons mattered even for decisions below the threshold of major public effect.

Geographic diversity was not treated as ceremonial. Commerce asked how developing regions and other communities would gain equitable representation through a transparent selection method. On ccTLDs, it demanded clarity about national authority.

This letter prevents a triumphalist reading in which Commerce simply confirmed that ICANN embodied world consensus. Recognition was conditional on institutional repair. The concerns also reveal that contemporaries understood the difference between consultation and accountability. Listening to comments was not enough if the board, budget and conflicts remained insulated.

The document does not prove that every criticism was correct or every proposed remedy feasible. It proves that the founding bargain was contested on exactly the questions that a private global coordinator had to answer.

November's changes were promises under construction

ICANN responded by amending its articles and bylaws and explaining the changes in a November 23 letter. It promised work on membership, more public sessions, timely minutes with individual votes and reasons, mandatory development of independent review, stronger conflict rules and geographic diversity.

The candour of the letter is useful. On membership, ICANN said an advisory committee would propose alternatives for criteria, rights and responsibilities. On independent review, it recognised the need for recourse when ICANN or staff broke their own rules but said the mechanism had not yet been figured out. The bylaws made creation mandatory; implementation would follow public proposals.

These were meaningful commitments, not completed institutions. The distinction matters because later histories sometimes project mature accountability mechanisms back onto the founding moment. In November 1998 the corporation was still designing how the public would participate and how an affected party could obtain outside review.

The board also kept formal voting meetings closed while promising associated public meetings and complete minutes. That compromise reflected a tension between candid deliberation and direct observation. The value of the arrangement depended on whether minutes actually conveyed reasons and whether public input changed decisions.

Conflict protections included a rule intended to prevent directors associated with a supporting organisation from supplying the decisive margin for that organisation's proposal. Funding mechanisms would face notice and comment and were not to shape policy by their nature. These controls recognised that functional representation can reproduce sector capture inside the board.

The MOU two days later should therefore be read as a test arrangement with an institution under construction. Commerce selected ICANN to develop the model, not because every legitimacy question had been conclusively answered.

Founders could start the corporation, not constitute the world

Every institution needs a first meeting before its ordinary selection rules exist. ICANN's incorporators and initial board solved that bootstrapping problem. Their authority to create a California corporation was ordinary and legally intelligible. Their claim to coordinate global dependencies required more.

Founders brought technical relationships, professional reputation, government access and organisational capacity. They could draft articles, appoint initial directors, hire counsel and negotiate. They could not make themselves elected representatives of people who had never selected them.

The White Paper addressed the gap with interim status and evolution. It expected an initial board representing functional and geographic diversity, followed by mechanisms for electing a permanent board. It even anticipated that interim directors might be barred for a period from serving on the permanent board. The objective was to prevent the bootstrap from hardening into incumbency.

Practical continuity pulled the other way. Experienced people knew the systems and counterparties. Excluding all founders could endanger stability. Developing global membership could take longer than expected. Each delay could be explained as necessary, while each delay also extended unelected authority.

The proper legitimacy account is therefore temporal. Initial authority can rest on necessity, competence, recognition and limited duration. Durable authority needs compliance with founding limits, representative renewal, review, performance and continued acceptance. The standard should become stricter, not looser, as emergency conditions recede.

This framework avoids personalising the issue. The question is not whether founders were benevolent. It is whether the institution converted borrowed authority into accountable authority. The White Paper supplied a direction for that conversion but no automatic certificate of completion.

Technical acceptance is powerful but not democratic consent

An identifier coordinator gains authority when networks, registries, software and users rely on its outputs. The root is useful because resolvers and operators converge on it. Number registries are effective because the routing and addressing community recognises the allocation hierarchy. Protocol registries work because implementers use common assignments.

This operational acceptance is a real source of legitimacy. A coordinator that consistently preserves stability, treats parties fairly and produces interoperable records earns trust. A competing body with a grander constitutional claim but no adoption cannot coordinate the system.

Acceptance is not equivalent to free consent. Switching from the widely recognised root or allocation hierarchy can be technically and economically prohibitive. An end user may rely on ICANN-coordinated DNS without knowing its policies. A registrant may accept a contract because every practical registrar includes it. Network effects give the coordinator leverage.

The White Paper's procedural safeguards responded to that leverage. Competition should exist around services even where the central identifier decision remains unique. Representation should include users who cannot individually exit. Transparency should expose reasons. Review should provide a remedy. National law should remain available.

The institution should therefore avoid a circular argument: everyone follows the coordinator because it is legitimate, and it is legitimate because everyone follows it. Operational acceptance proves effectiveness and reliance. It does not by itself prove fair distribution of authority or lawful treatment in a particular case.

The founding bargain joined effectiveness to constraint. Stable common identifiers justified a coordinator; dependency increased the coordinator's accountability burden.

The narrow mandate still allowed consequential judgment

Calling the White Paper narrow should not sanitise it. Deciding whether to add a top-level domain, which registry operates it, what accreditation conditions apply, how an address hierarchy is directed and which procedure resolves a naming conflict can transfer large economic opportunities.

The phrase “technical coordination” can hide distribution. A stability requirement may favour an incumbent with existing scale. A trademark rule may favour rights holders over registrants. A fee can deter small entrants. A geographic-diversity rule can change who exercises board power. An address policy can affect scarcity and market access.

The White Paper understood some of this. It made competition, representation, transparency and anti-capture safeguards coequal with stability. It expected legal expertise in competition and intellectual property as well as technical expertise. It refused special immunity for restraints broader than legitimate coordination.

Narrowness therefore concerns subject matter and institutional competence, not the absence of politics. The coordinator can make policy within the identifier joint. It must show why the policy is needed, how burdens relate to the objective, who participated, which interests were excluded, what alternatives existed and how an affected party can challenge.

This is also why neither government nor private management is automatically legitimate. A national agency can overreach; a private board can use network effects coercively. The White Paper chose private management conditionally, based on expected flexibility, bottom-up input and protection against factional capture.

The bargain should be defended without myth. It was a decision to place consequential judgment in a bounded private institution and build substitutes for electoral accountability around it.

A bounded test for later claims of authority

The White Paper remains useful when treated as a scope-and-legitimacy test rather than a sacred founding text. Any claim derived from it should answer a sequence of questions.

First, which of the coordinated functions is implicated: number allocation, authoritative root oversight, top-level-domain policy or protocol-parameter assignment? If the connection is indirect, what nearer instrument supplies authority? Second, is the act no broader than necessary for uniqueness, stability, interoperability or the legitimate market structure around the function?

Third, which constituency bears the cost, and how was it represented? A comment period is evidence of participation but not a global vote. The decision should identify support and dissent without claiming an unknown denominator. Fourth, what anti-capture protections applied? Conflicts, funding dependence, sector concentration and board composition matter.

Fifth, which enforceable rights remain? The affected party needs notice, criteria, a record, reasons, consistent treatment, review and access to national law or contract where applicable. Sixth, what remedy can correct an error? A declaration of procedural concern without restoration, reconsideration or other practical effect may be inadequate.

Seventh, what evidence supports the stability claim? Institutions often invoke systemic risk without showing the causal chain or considering a narrower action. Eighth, does the decision treat administrative authority as property ownership? If so, where is title defined? The White Paper cannot supply a missing asset-disposition power.

Finally, what preserves the ability to evolve? A founding arrangement designed as a transition cannot justify permanent insulation. Representation, procedure and review should adapt as users, markets and technical dependencies change.

This test neither freezes ICANN in 1998 nor grants it unlimited implied powers. It lets function evolve while keeping authority connected to the bargain that made private coordination preferable to government monopoly.

The countercase deserves its full weight

A defender of broader ICANN authority can make a serious argument. Unique identifiers are not static clerical entries. Threats, market structures and technologies evolve. A corporation limited to mechanically repeating 1998 procedures could fail to preserve stability. Bottom-up policy development necessarily interprets the mission. The White Paper invited an ongoing process, not a fixed list of commands.

The document also gave the board final authority and contemplated policy for adding and managing top-level domains, accrediting market entities and directing number allocations. It recognised that a global and commercially important system needed robust professional management. One cannot demand effective coordination while denying every discretionary choice.

This countercase defeats an artificially tiny reading. ICANN was not meant to be a passive typist. It was meant to coordinate policy within defined functions and respond to changing Internet needs.

It does not establish world government. Adaptation must remain connected to identifier coordination, founding principles, governing instruments and legal rights. A new problem near the DNS is not automatically a DNS coordination problem. The ability to enforce a rule through contracts or root dependence does not prove institutional competence to make the rule.

Nor does bottom-up development cure every expansion. Organised insiders may dominate participation, while dispersed users bear costs. Consensus can be manufactured through categories and meeting rules. Review can be slow or narrow. A board can describe its preferred policy as technical necessity.

The correct answer is not stasis but a burden of explanation. The farther an act moves from direct uniqueness and interoperability, and the more it affects assets, expression, competition or national authority, the stronger the authorising instrument, participation, evidence and remedy must be.

The bargain avoided a national monopoly without founding a nation

The 1998 policy confronted an institutional fact: global users depended on identifier functions still entangled with one government's agencies and contracts. Leaving that arrangement untouched would have preserved a national monopoly whose legitimacy weakened as the Internet internationalised and commercialised.

The White Paper chose a private not-for-profit coordinator. It required stability, competition, bottom-up participation and representation. It rejected a monolithic Internet-government structure, limited the functional project to names, numbers and necessary parameters, preserved applicable law and expected transparent, anti-capture procedure.

That was a coherent bargain. It allowed one authoritative system without claiming one political sovereign. It moved decisions away from direct United States administration without pretending that a world electorate had assembled. It gave a corporation real coordinating power while withholding property title and general regulatory jurisdiction.

Its weaknesses were visible immediately. The electorate was undefined. The initial board was selected, not globally elected. Finance and conflicts required controls. Country-code authority was unsettled. Independent review existed first as a promise. Commerce retained transitional oversight. Representation had to be built from unequal functional groups.

Those weaknesses do not nullify the bargain. They identify the conditions under which it remains legitimate. Coordination must stay close to the common technical joint. Policy must be no broader than necessary. Community claims must name their constituency. Decisions must expose reasons and conflicts. Affected parties must have enforceable review and national-law rights. Administrative control must not be inflated into ownership.

The White Paper authorised a path towards private coordination through later agreements; it did not legislate for the planet. It helped create an institution; it did not create a people. It recognised the need for a common ledger of identifiers; it did not transfer the underlying world to the ledger's keeper.

That narrow bargain is less majestic than a global constitution and more defensible.

Sources