LACNIC and the economics of policy mailing-list procedure

The useful scene is not a boardroom. It is a thread.

A policy proposal appears in LACNIC's public policy system. The title is technical, but the stakes are economic: whether IPv4 resources can be sub-assigned to third parties, how such use should be recorded, whether the holder remains responsible, which recipient must hold an ASN and IPv6 resources, what size limits apply, whether a public movement log should exist, and how long newly received blocks should remain unavailable for this kind of use. The proposal is not hidden. It can be opened in Spanish, English and Portuguese. It has version tabs, dates, authors, a status marker, a diff link, downloadable files and a visible path through discussion, first consensus assessment, last call, second assessment and board ratification. Its authors say that an earlier draft was shared on the mailing list, that comments from the mailing list and the Public Policy Forum were reviewed, that some suggestions were included, and that others were left out.

That is the old internet at work: text, revision, community memory and a public record. It is also a market for attention. A few people can read the proposal early. Fewer can read it in all three languages. Fewer still can compare version one and version two, follow the Public Policy Forum, understand the Policy Development Process, judge the economics of leasing, estimate the burden on small ISPs, decide whether to object during last call, and keep doing this across proposal after proposal. The mailing list is open, but the useful unit of participation is not access. It is sustained, procedurally literate attention.

LACNIC describes its policy development as participatory, public, transparent, open to any individual and based on consensus. It says proposals are presented and discussed on the policy mailing list before being taken to the forums where consensus is sought. The public policy development page links to the policy discussion list, the policy system and the LACNIC Policy Manual. The separate Policy Development Process sets out the roles of moderators, the Public Policy Forum, the LACNIC Board, appeals, last call and LACNIC's support obligations. On paper, the design is more open than many institutions that shape infrastructure markets. In practice, it still converts unequal attention into rulemaking power.

This article is about that conversion. It is not another article about whether IPv4 is scarce, whether transfer markets should exist, or whether leasing is a suitable financing tool. Those questions matter, but they are not the main subject here. They appear here only as the policy objects that reveal a deeper institutional economy. The subject is the procedure by which LACNIC's policy community decides how such questions become policy text. A mailing list may look like a cheap institution. It has low visible cost and high archival value. Yet it decides who can see the agenda, who can understand the consequences, who can frame a problem, who can sustain a coalition, who can object without social penalty, and whose absence is later treated as a tolerable silence.

That makes procedure a form of power. Attention is the first input. Language ability is the second. Time is the third. Repeated participation is the fourth. Archive memory is the fifth. None of these are formally sold, assigned or counted, but all of them accumulate. The person who remembers the last failed proposal, knows the moderator's expectations, reads the Spanish text before the English reply has settled, understands how a last-call objection must be substantiated, and can return after implementation feedback has appeared owns a kind of policy capital. The mailing list is open to everyone; the compounding return goes to those who can remain present.

In that sense, the mailing list is agenda-setting market infrastructure. It is not a parliament, and it should not try to become one. It is not a commodity exchange, and consensus should not be bought or counted as if it were a shareholding vote. But it is part of the governance machinery that determines how scarce number-resource rules are made. If the list works well, it lowers the cost of correction, captures technical knowledge, exposes minority concerns and leaves a record that future participants can inspect. If it works badly, it launders participation inequality into the language of community consensus.

The list as an institution of attention

The basic economics are simple. Participation has fixed costs. A person must notice the proposal, read it, understand the policy section being changed, infer the operational effect, read the earlier debate, decide whether the new text is better or worse, understand the role of the Public Policy Forum, and return at the correct moment if a last call or consensus assessment changes the stakes. For a frequent participant, these costs are part of ordinary institutional life. For a small operator, they compete with outages, customers, billing, vendor problems, power issues, compliance demands and the daily labour of keeping a network running.

Fixed costs create advantage for scale. A large carrier in Brazil or Mexico can amortize policy monitoring across a large resource base and a large customer base. A global cloud or content company can assign specialists to several regional internet registry lists. A broker or address-market participant can justify tracking a policy because one sentence may affect deal flow. A consultant can build reputation by participating repeatedly. A small Caribbean provider, a rural ISP, a municipal network, a university, a family-owned operator or a regional hoster may care more about a specific rule and still be less able to speak.

The mailing list therefore functions as a sorting device. It rewards those who can translate operational pain into procedural language. It rewards those who know when a proposal needs new text rather than another objection. It rewards the participant who understands the difference between a forum discussion, a consensus assessment, a last call, board ratification and implementation. It rewards endurance. It punishes late discovery.

None of this makes the process bad. A closed staff procedure would be worse. LACNIC needs a public place where operators, engineers, lawyers, researchers, members, critics and users can read and respond. The list creates an archive. It keeps arguments from disappearing into hallway memory. It makes it possible to see whether a proposal changed after criticism. It lets non-members speak. It gives moderators a record to interpret. The problem is not openness; it is the habit of treating formal openness as if it solved participation economics.

The difference matters most when a policy alters resource economics. A small wording change about a transfer lock, a need justification, a waiting-list priority, an RPKI condition, a public log, a WHOIS responsibility or a documentation requirement can change the cost of obtaining address capacity. The mailing list does not set a price for IPv4, but it can change the risk premium on a block. It can make leasing more visible or more opaque. It can reduce fraud or raise entry costs. It can protect small networks or make them dependent on larger intermediaries. The forum may be free to join, but its output is not free to live with.

A regional procedure under many languages

LACNIC serves Latin America and the Caribbean, a region that is not one market in any practical sense. It contains large continental economies with extensive technical communities, local counsel, data-centre investment, mobile scale and public-policy capacity. It also contains island economies where a few submarine paths, a narrow labour pool and foreign-currency payments shape connectivity. It contains Spanish-speaking and Portuguese-speaking operators, English-speaking Caribbean networks, indigenous and local-language realities outside policy rooms, and companies that may have to translate not only words but also legal categories, corporate documents and network habits.

The region's unevenness is not a footnote to procedure. It is the terrain on which procedure operates. In a large country, a registry-policy question may be one file among many handled by an association, a legal team, a regulatory office or a network-strategy group. In a small market, the same question may reach one engineer-manager after hours, after the generator has failed, after an upstream contract has changed, and after a public agency has asked for a service extension. A proposal may be public in three languages and still be practically louder in the places where people already have time, peers and confidence to discuss it.

The policy system acknowledges part of this diversity. Proposals appear with Spanish, English and Portuguese access where translations are available. The LACNIC policy system shows language buttons, version histories, download links and status paths. That is meaningful. It is better than a single-language archive and a closed meeting culture. Yet multilingual publication does not remove multilingual cost.

Translation changes timing. A participant who reads the original language first has a head start. A participant who waits for translated material may arrive after the frame has settled. Even when a translated version exists, informal understanding often moves through side conversations, forum comments, list replies and relationships that may not be equally accessible. A Brazilian participant following Portuguese and English may not experience the same debate as a Spanish-speaking participant who followed the original thread from the first day. An English-speaking Caribbean operator may be technically fluent and still feel peripheral if most social context forms elsewhere.

The cost is not only linguistic. Legal and commercial terms travel poorly. "Assignment", "sub-assignment", "transfer", "temporary transfer", "need", "holder", "recipient", "service contract", "NIR" and "public log" may have stable meanings inside the policy text and different instincts in business practice. A small operator may understand exactly why it needs a rule but hesitate to enter a public argument if the terminology feels like someone else's terrain. A government network may have to ask whether a published comment exposes procurement practice. A university may have no counsel familiar with address-policy language. A small hoster may know the operational answer but not the consensus vocabulary.

Time zones and travel add another layer. The Public Policy Forum is important because mailing lists can flatten nuance and because live debate helps participants test whether objections are practical or rhetorical. But a forum linked to a regional meeting is not costless. Travel, visas, hotels, conference time, employer permission and social familiarity matter. Remote participation helps. It does not fully replicate the confidence of being in the room, catching a moderator's tone, speaking to the authors, and hearing why a particular objection is or is not gaining traction.

LACNIC cannot abolish these differences. No regional institution can make an English-speaking island operator, a Brazilian access network, a Mexican enterprise provider, a Chilean university, an Argentine cloud firm and a Central American wireless ISP experience policy at identical cost. But it can stop treating the list as if subscription were the relevant measure. The relevant measure is whether affected parties can understand and influence the issue before silence becomes procedural closure.

What the PDP actually filters

LACNIC's Policy Development Process is not casual. It assigns important duties to moderators. The PDP text says moderators must publish the Public Policy Forum agenda on the Public List at least one week before the forum, prepare a report after the forum, introduce community proposals for discussion, remind participants that the minimum debate period is eight weeks, and note that no proposal may be called to consensus unless it has been presented and discussed in at least one Public Policy Forum. After the debate period, moderators have up to two weeks to say whether consensus has been reached. If consensus is reached, a four-week last call follows. After last call, moderators have up to one week to confirm whether consensus remains. Board ratification then decides whether to accept the proposal and work with staff on implementation, or reject it and ask for deeper analysis.

Those steps are safeguards. They prevent a proposal from becoming policy merely because a few people moved quickly in a thread. They force public discussion, a forum presentation, a consensus assessment, a last-call period and board attention. They also create filters. A person who does not know the eight-week clock may arrive late. A person who does not understand the Public Policy Forum threshold may misread list debate as merely advisory. A person who believes last call is a real opening may discover that only certain objections are treated as timely and substantive. A person who does not know how to write a grounded objection may be read as offering an opinion rather than evidence.

The filter is most powerful because consensus is not a vote. This is a strength in technical governance. A simple vote can reward factional mobilization, low-information flooding, or corporate turnout. Rough consensus can privilege reasoned objection and technical merit. But the lack of a vote also means the population being interpreted is the active participation record, not the affected economy. Moderators interpret a conversation among those who showed up. They cannot interpret the silence of those who did not know, could not afford attention, or feared public exposure.

This is why the phrase "the community" should be used with care. In a PDP record, the community is not a census of every LACNIC member, every downstream user, every small island provider, every state network, every university and every enterprise that may later depend on the rule. It is the visible policy community assembled by the list, the forum, the proposal history and the moderators' reading of the record. That narrower community can make good technical decisions. It should not be confused with the whole incidence map of the rule.

LACNIC's last-call text is revealing. Its purpose is to give the community a brief final chance to comment, especially for those who did not previously do so. It allows editorial changes that preserve the substance of the text that reached consensus and says exceptional objections may arise if a previously undiscussed aspect is discovered, but objections must be substantiated rather than unsupported technical opinion. This is reasonable. It protects the process from endless reopening. It also raises the bar for a late participant who may be precisely the kind of affected operator the earlier stages failed to reach.

This is the central tension. Procedure needs closure. Markets need predictability. A proposal cannot remain open forever because every possible small network might one day notice it. Yet closure is legitimate only if the record has reasonably tested the issue against affected interests. The more a proposal touches resource economics, the less persuasive it is to say that nobody objected. The better question is who had a practical chance to object before the closure mechanism began to work.

The compressed clock

The timing rules look generous until one sees them through the calendar of a small operator. Eight weeks of discussion is a long period for a list. It is not necessarily long for a company that must first recognize that a proposal matters, assign someone to read it, compare the current text with the manual, ask whether the issue affects present contracts or future plans, decide whether a public comment is safe, translate the comment if necessary, and return after the proposal changes. Four weeks of last call is long enough for a participant who has followed the debate. It is short for a participant who discovers the consequence only when the last-call notice appears.

This is especially true when the proposal sits inside ordinary operational noise. A small ISP may be dealing with storms, power interruptions, vendor renewal, peering disputes, customer complaints, tax filings, a bank query and a router migration at the same time. Policy debate is not experienced as a civic duty on a clean calendar. It is another demand on a scarce engineering and managerial budget. A large firm can assign policy attention as work. A smaller operator must steal it from something else.

The clock also interacts with language. If the first meaningful discussion happens in one language, translation and local explanation may trail the frame. By the time an English-speaking Caribbean network or a Portuguese-speaking operator outside the main circles reads a clean version, the strongest positions may already have hardened. A later comment may then be treated as reopening an issue rather than as evidence that the early stage did not reach far enough.

Proposal revision makes the timing problem harder. Revision is good: it shows learning. But each new version resets the cost of understanding. LAC-2025-5 changed size limits, regional-use language, responsibility provisions and restrictions. Those changes matter. They may make the proposal more acceptable, but they also require fresh reading by every affected participant. A repeat participant can compare versions quickly. A newcomer may have to reconstruct the whole debate. If the list then treats the second version as familiar because the active participants have been discussing the issue for months, procedure has imported insider time into public time.

The calendar should therefore be analyzed as an economic instrument. A debate period allocates the burden of attention. A last call allocates the burden of late discovery. A consensus assessment allocates the risk of silence. Board ratification allocates the final institutional responsibility. None of these allocations are wrong by themselves. They become questionable when their costs are invisible.

One practical remedy would be a "changed-incidence" note with every new version of a high-impact proposal. Instead of merely showing a diff, the authors and moderators could state which groups face a different burden under the new text. If a maximum block size changes from /21 to /22, who is affected? If an IPv6 requirement is added, which recipients may be excluded? If regional-use language is added, what does that mean for cross-border Caribbean service or multinational networks? If responsibility remains with the holder, what kind of contract or operational capability is implied? This would not slow every proposal. It would lower the cost of deciding whether a new version matters.

Another remedy is a stronger late-discovery path. Last call should remain final in normal circumstances, but moderators should explicitly ask whether any late objection is tactical or evidentiary. A sophisticated participant who waited until the final stage to relitigate an old preference should not be rewarded. A small operator that first understood a fixed-cost burden at the final stage should be treated as a signal that notice was incomplete. The distinction requires judgement, but consensus already requires judgement.

The compressed clock is not proof of bad faith. It is the natural consequence of trying to make policy through a public but specialized forum. The danger is pretending that a published date range means the same thing to every participant. It does not. Time is cheaper for some institutions than for others. In a mailing-list economy, cheap time becomes influence.

The proposal thread as economic evidence

LAC-2025-5 is a useful example because it is not abstract. The proposal concerns sub-assignment of IPv4 resources to third parties. Its public text says the existing LACNIC Policy Manual prohibited sub-assignments outside the original recipient's infrastructure. It says IPv4 exhaustion had produced informal leasing practices without a record of the transaction. It identifies risks such as outdated records and resources leaking to other RIRs. It says small ISPs are among the most affected. It proposes a mechanism for transparent and traceable sub-assignment, including WHOIS identification, a public movement log, responsibility remaining with the LACNIC member making the sub-assignment, a /24 minimum, a /22 maximum in the second version, recipient ASN and IPv6 requirements, utilization justification, regional-use limits, critical-infrastructure exclusions and restrictions on recently received blocks.

The substance of the proposal belongs to leasing and scarcity economics. The procedure tells a second story. Version one did not reach consensus. Version two says the authors reviewed comments received through the mailing list and the Public Policy Forum, tried to incorporate suggestions while preserving the proposal's initial spirit, changed the maximum block size, clarified the use of the existing manual, specified the log, added regional-use language and improved restrictions. The public system displays a discussion period from December 2025 into January 2026, a first consensus assessment, a last-call period and ratification in March 2026.

That is not a rubber stamp. It shows revision. It shows that comments mattered. It shows that the list and forum can alter text. It also shows how much procedural stamina is required. A participant worried about the first version had to remain present long enough to evaluate the second. A small ISP that would benefit from cheaper access had to decide whether the restrictions were workable. A resource holder had to judge whether remaining responsible for the recipient was tolerable. A network that leases informally had to decide whether visibility was safer than opacity. A Caribbean operator had to infer whether regional-use language helped or limited it. A broker had to understand how a public log changed settlement risk. A security participant had to consider whether WHOIS and RPKI authority would follow operational reality.

All of that had to be compressed into the rhythms of a mailing list, a forum and a last call. The public record may be sufficient to justify policy adoption. But it should also make observers ask what kind of organizations had the capacity to shape the revisions. Were comments dominated by repeat participants? Did small operators speak directly or were they invoked by others? Did English-speaking Caribbean concerns appear in the same density as larger Spanish- and Portuguese-speaking markets? Were dissenting comments treated as evidence, bargaining, delay or discomfort with market reality? The answers matter because the proposal itself is about making an informal market legible.

Another proposal, LAC-2025-4, illustrates the opposite path. It proposed priority on the IPv4 waiting list for applicants that had already deployed IPv6. It did not reach consensus. LACNIC staff comments on the public page described high operational impact, very high legal impact and medium system impact, including the burden of verifying more than 1,700 organizations on the waiting list and concern about equality in processing requests. That public staff comment is valuable because it brings implementation costs into the record. It also shows how staff analysis can influence whether a proposal survives. The economics of procedure include not only list participants but also the registry's ability to make implementation risk visible.

LAC-2023-7, a temporary-transfer proposal, shows persistence and attrition. It passed through multiple versions, was presented at LACNIC 42 and LACNIC 43 or 44 stages depending on version, did not reach consensus and was ultimately withdrawn. Its text discussed leasing, temporary transfers, regional benefit, RPKI, geolocation, IRR information, MANRS practices and the need of smaller entities. The fact that a proposal can accumulate five versions and still leave the field is healthy in one sense: procedure can say no. It is also expensive. Only a small subset of the region can follow a multi-version debate closely enough to know whether the final withdrawal was a victory for caution, a failure to solve a real problem, or a sign that the active list could not digest the market it was discussing.

Staff comments as price signals

Registry staff comments are sometimes treated as implementation notes. In a scarcity economy they are also price signals. When LACNIC staff say a proposal has high operational impact, the market should hear more than "this is hard for the office." It should hear that the rule may impose verification costs, ticket delays, evidence burdens, legal exposure, system changes and support complexity. Those costs do not remain inside the registry. They reappear as waiting time, legal work, uncertainty, failed applications and conservative behavior by members.

LAC-2025-4 makes the point. A proposal to reward IPv6 deployment by moving applicants up the IPv4 waiting list sounds attractive if framed as encouragement for transition. Staff comments shifted attention to verification and equality. How would LACNIC determine effective IPv6 deployment? What evidence would be enough? Would network testing be required? How would more than 1,700 waiting-list organizations be reviewed? Would a queue designed for predictability become a queue exposed to challenge? Would applicants that lack the staff to document IPv6 deployment be pushed down even if they have real IPv4 need? These are not narrow office questions. They are economic-incidence questions.

Staff comments are valuable because staff see failure patterns that the list may not. They know where applicants misread policy, where documents are incomplete, where legal or contractual ambiguities appear, where system changes are expensive and where support queues will grow. The active community may debate principles; staff can describe friction. In a registry, friction is cost.

But staff comments must also be bounded. A staff warning can become a soft veto if the community treats operational discomfort as decisive without examining alternatives. A rule may be costly and still worth adopting. A legal risk may be real but reducible by narrowing text. A system change may be heavy but justified if it removes a larger market distortion. Staff expertise should discipline the record, not replace consensus. The proper role is to make cost visible in categories that participants can evaluate.

For high-incidence proposals, staff impact notes should therefore be written with market readers in mind. They should say what information will be requested, who must provide it, what systems must change, which parties face new obligations, what delays are likely, which legal risks are specific, which are speculative, how NIR-related cases might differ, what implementation data will be published, and how the policy could be reviewed. This would make staff comments less like institutional caution and more like settlement information.

Staff comments should also disclose uncertainty. If LACNIC does not know how many small operators would use a sub-assignment mechanism, it should say so. If it cannot predict whether a public log will reduce opaque leasing, it should say what evidence would be monitored. If it expects a surge in support tickets, it should estimate the type. If privacy or commercial sensitivity is a concern, it should identify the data field rather than invoke sensitivity in general terms. Precision protects both staff and participants.

The mailing list needs staff comments because volunteers do not see the whole registry service surface. Staff comments need the mailing list because staff may underweight market consequences that members experience outside the ticketing system. The strongest procedure lets the two correct each other. A weak procedure lets either staff or active participants dominate the field without enough evidence from those who bear the cost.

In LACNIC's regional context, staff comments can also reduce large-country bias. If the registry can publish anonymized evidence that a proposed document requirement would fall unevenly on certain countries, languages or small-account categories, the list can debate the actual incidence rather than relying on anecdotes. If staff can show that a proposed log can be maintained at low cost, opponents cannot exaggerate burden. If staff can show that a waiting-list change would create legal risk, supporters must answer the risk rather than moralize IPv6. Public staff analysis is therefore not bureaucracy. It is part of the price system of policy.

Repeat players and the cost of memory

Every open procedure creates repeat players. They are not necessarily villains. In a technical community, they are often the people who keep history alive. They remember why an old rule was written, which compromise failed, which wording caused trouble, which implementation detail is harder than it looks, and which proposal is merely a recycled idea with a new title. Without repeat players, the list would be a place of amnesia.

But memory is power. A repeat participant knows when a problem should be framed as a manual change, when it belongs in an operational request, when it might be better as a working group, and when a moderator is likely to ask for clearer text. A newcomer may have a better view of the burden but a weaker ability to express it. A small ISP may say that a requirement is costly; a repeat participant may answer that the issue was settled years ago. Both may be right. The first is describing present incidence; the second is describing procedural history. If the process privileges history too strongly, it becomes conservative by default. If it ignores history, it wastes time and creates unsafe rules.

Policy entrepreneurs sit at the intersection. They draft proposals, circulate ideas, respond to comments and keep issues alive. They provide a necessary public good. A proposal about sub-assignment, waiting-list priority or temporary transfers will not exist unless someone turns frustration into text. Yet the person who writes the problem statement also defines the economic frame. A proposal described as transparency and traceability asks opponents to explain why opacity is acceptable. A proposal described as preventing speculation asks opponents to argue against caution. A proposal described as IPv6 reward asks opponents to explain why early deployment should not matter. A proposal described as small-ISP access asks opponents to confront development claims.

Framing can be legitimate and still powerful. It decides which objections look selfish, which look technical, which look moral and which look procedural. A holder worried about continuing responsibility may appear to be avoiding accountability. A small operator worried about justification burdens may appear to be asking for weak controls. A registry staff comment about legal exposure may appear to stop a community preference. A broker worried about process complexity may appear self-interested even when it reveals real transaction friction.

The repeat-player problem is therefore not motive. It is cost. People who can return repeatedly will define not only outcomes but also the acceptable language for outcomes. A small island network may not have the chance to say, in the right vocabulary, that a public log is good but the utilization proof is too high, or that an IPv6 requirement is sensible for larger recipients but difficult for a temporary operational need, or that regional-use restrictions may collide with how Caribbean networks buy capacity and serve customers across borders.

Good moderation can mitigate this. Moderators can ask whether claims about small ISPs are supported by direct testimony. They can encourage authors to state what comments were rejected and why. They can keep a confident repeat participant from occupying the room. They can ask staff to explain implementation risk without letting staff preference become the whole record. They can distinguish a late tactical objection from a late discovery by an affected party. They cannot make attention equal, but they can keep inequality visible.

NIRs and indirect participation

LACNIC's region also has national registry pathways, most visibly in Brazil and Mexico. That structure can lower local service costs. A national registry can provide language familiarity, domestic institutional knowledge, local relationships and a more accessible interface for many operators. It can explain regional rules through local practice and support members that might otherwise struggle to deal directly with a regional office.

But local mediation also changes participation evidence. An operator may experience a LACNIC policy through a national registry's forms, instructions, support staff and domestic community rather than through the regional mailing list. Its concerns may be raised locally, summarized by someone else, or not raised at all. The regional archive can then understate the real volume or character of concern. Silence on the LACNIC list may not mean local agreement. It may mean that the relevant debate happened closer to the operator and did not travel upward in a visible way.

This matters for proposals that affect transferability, sub-assignment, RPKI, WHOIS, documentation or waiting-list treatment. A policy can be formally uniform while its administrative cost differs by path. If a resource holder is in a national registry environment, the practical burden may depend on local forms, local support, the timing of local notices, the relationship between regional and national records, and how public logs or RPKI authority are represented to counterparties. A buyer or lessee may price the path differently if it expects additional explanation or uncertainty.

The point is not that NIRs are barriers. They can be bridges. The point is that bridges have their own structure. A regional consensus assessment should ask whether national communities have received usable notice for high-impact proposals and whether any local observations should be summarized in the regional record. Such a summary need not give a national registry a veto. It would simply prevent the regional list from treating its own archive as the whole participation field.

This is particularly important for small operators inside larger countries. A Brazil-based or Mexico-based small ISP may be closer to a national registry but still far from regional policy discussion. Large firms in those same countries may have both local and regional access. If the policy record simply shows comments from a country, it may hide size concentration within that country. National participation is not the same as small-operator participation.

Indirect participation also matters in the Caribbean. Many Caribbean networks may interact with LACNIC directly rather than through a national registry, but their participation may still be mediated by regional operator groups, upstream providers, consultants, brokers or informal technical relationships. A small provider may not post on the policy list but may rely on a consultant who does. That can help. It can also distort the signal if the consultant's incentive differs from the operator's. Procedure should welcome intermediated expertise while trying to record whose burden is being described.

For high-incidence proposals, LACNIC could publish a short "reach map": which language versions were available when, whether national registry or operator-group channels were alerted, whether Caribbean-facing notice was issued, whether staff received off-list questions that reveal common confusion, and whether any affected category appears absent. This is not complicated public administration. It is a way to measure whether a mailing-list debate is actually regional.

Absence is not consent

Consensus is not a vote and should not be treated as a commodity. It is a judgment that visible objections have been addressed well enough for the process to move. In a technical policy setting, that may be the only practical way to make decisions. But consensus becomes dangerous when absence is treated as consent.

Silence has many meanings. It may mean agreement. It may mean that the issue is too technical for the affected party to assess quickly. It may mean the operator did not see the message. It may mean translation lag. It may mean the relevant employee was not allowed to post publicly. It may mean the company fears revealing its resource position or leasing arrangement. It may mean the participant objected in a hallway, a national forum or a private support exchange but not on the public list. It may mean exhaustion. It may mean resignation.

In a region with unequal attention costs, silence is especially weak evidence. A large operator can choose silence strategically. A small operator may have silence imposed by capacity. A global company may remain quiet because it can adapt to almost any rule. A small network may remain quiet because it cannot analyze the rule until it hurts. A policy system that reads all these silences alike will bias its legitimacy claims toward those with organized voice.

The risk is not theoretical. IPv4 scarcity policies are full of areas where absence can hide exposure. Transfer conditions affect future buyers and sellers, many of whom will not follow the list until they transact. Waitlist rules affect applicants that may be waiting for years and may not have policy specialists. Leasing visibility affects lessees who may not be LACNIC members or may not think of themselves as policy participants. RPKI requirements affect downstream route acceptance and leased-resource users. Abuse-contact rules affect small teams that may already struggle with complaint volume. Legacy regularisation affects holders whose old documents may be fragile and whose public posture may be cautious.

The procedural answer is not to wait for perfect participation. It is to record absence honestly. A consensus assessment for a high-incidence proposal should say not only who supported and opposed, but which groups appear underrepresented. Were there comments from small ISPs? From the Caribbean? From Brazil and Mexico outside the largest players? From holders, recipients and operational users? From security operators? From those likely to lease? From those on the waiting list? From public-sector or university networks? From the NIR-related communities? The absence of a group does not veto a proposal. It changes how confidently the process should speak.

This is not a demand for identity quotas or a turn toward parliamentary representation. LACNIC is a registry, not a legislature. The point is evidentiary discipline. A consensus call should not say, even implicitly, that the region agreed when the record shows only that the active policy community converged. That narrower claim may still be enough. It is also more honest.

Moderation: discipline and its shadow

Mailing lists need discipline. Without moderation, a policy list can become unreadable. Spam, personal attacks, repetitive claims, off-topic politics, unsupported allegations and circular quarrels raise the cost of participation for everyone. Small operators are not helped by a chaotic forum. They are often the first to leave it. The legitimacy of a mailing list depends partly on the ability of moderators to keep discussion focused and civil.

LACNIC's PDP gives moderators important responsibilities: introduce proposals for discussion, manage timing, judge consensus, publish last calls, communicate outcomes and support the public forum. These responsibilities are not clerical. They shape the market for attention. A moderator who asks for evidence can improve the record. A moderator who lets one style of speech dominate can narrow it. A moderator who frames late comments as obstruction can discourage affected parties. A moderator who never closes debate can make policy impossible.

The distinction between process discipline and agenda control is delicate. Anti-spam moderation is legitimate. Removing abuse is legitimate. Asking participants to substantiate claims is legitimate. Keeping a thread on a specific proposal is legitimate. But process discipline becomes agenda control if certain kinds of economic argument are treated as improper because they sound commercial, if dissent is socially framed as anti-community, if criticism of registry discretion is treated as hostility, or if experienced participants can apply tone pressure that newcomers cannot resist.

Social pressure matters because mailing lists are public reputational spaces. A network engineer may hesitate to question a proposal if senior figures have already praised it. A small provider may fear that disagreement with a registry-adjacent consensus will affect how peers see the company. A lessee may not want to reveal dependence on a lease. A resource holder may not want to discuss unused space. A broker may speak and be discounted because of obvious interest, even when the broker sees real transaction data that others do not. A public list can chill speech without any formal ban.

The cure is not neutrality as silence. Moderators should intervene more, not less, when the record is skewing toward social dominance. They can ask participants to separate technical objections, economic incidence, implementation burden and personal preference. They can invite authors to answer the strongest version of an objection rather than the easiest version. They can ask whether a claim about small operators has direct evidence. They can remind the list that disagreement is not disloyalty. They can ask staff to state legal or operational concerns in categories rather than broad caution.

Moderation should be judged by the cost it imposes on weak participants. A list that is polite but socially closed is not healthy. A list that is noisy but open to newcomers is also not enough. The test is whether an affected operator with limited time can understand the issue, make a credible point and see that point answered in the record.

Archives, version histories and the price of auditability

The great virtue of the mailing-list institution is memory. LACNIC's policy system does more than display current text. It shows proposal lists, statuses, dates, versions, language access, downloadable files, diff analysis and links to the policy discussion list. The PDP says LACNIC supports the Public List and the Public Policy Forum by maintaining the discussion list and its archive, updating web pages with process information, proposals under discussion and archives of previous proposals and their statuses, maintaining the Policy Manual and its update history, and helping with moderator elections.

These are not decorative features. They are anti-capture infrastructure. A public archive lets later participants see whether a claim was made early or invented late. A version history shows whether authors actually incorporated comments. A diff makes small but material wording changes visible. A public staff comment exposes implementation risk. A status record shows whether a proposal failed, was withdrawn, reached consensus or was ratified. A policy manual history shows how today's constraint arrived.

But auditability has its own cost. Archives are only useful to those who can search, read and interpret them. A long archive can become a moat for insiders. The more historical material exists, the more powerful procedural memory becomes. A newcomer may be told to read years of debate before objecting. That may be fair if the issue was fully covered. It may also be a way of converting memory into exclusion.

The answer is better summarization, not weaker archives. For each high-incidence proposal, the public record should contain a plain-language summary of the issue, the economic incidence, the main objections, what changed between versions, what was rejected, what staff warned about, and what assumptions should be checked after implementation. The summary should be available in Spanish, Portuguese and English. It should not replace the archive. It should make the archive usable.

Minority reports can help. They need not be legal briefs. A minority report can be a short statement by a dissenting participant, a group of operators or the moderators summarizing unresolved concerns. It should distinguish technical risk, economic burden, legal exposure, process concern and implementation uncertainty. It should be preserved with the proposal. That would prevent consensus from erasing dissent and prevent dissenters from later claiming that nobody heard them.

Impact statements are equally important. LACNIC staff comments on LAC-2025-4 are a good example of the value of public implementation assessment. A proposal that would reorder a waiting list of more than 1,700 organizations based on IPv6 deployment would require verification, raise legal concerns about equality in processing, and affect a queue designed to provide transparency and predictability after the 2020 IPv4 exhaustion event. Those are not merely internal implementation details. They are the economics of the proposal. Public comments of that kind should be standard when resource economics are at stake.

Implementation feedback closes the loop. After a policy is ratified, the list should not be allowed to forget it. Did the new sub-assignment mechanism reduce opaque leasing? Did small ISPs use it? Did holders find responsibility too risky? Did public logs improve trust or create privacy concerns? Did IPv6 requirements exclude legitimate recipients? Did support tickets rise? Did processing times differ by country, language or size? Without feedback, the mailing list becomes a place of ex ante confidence and ex post silence.

Scarcity policy through a procedural filter

Mailing-list procedure matters because IPv4 scarcity has pushed registry policy into the economics of capital, even when the language remains technical. LACNIC's IPv4 waitlist began after the last available IPv4 block was assigned on August 19, 2020. The current waitlist environment is a long queue for small amounts of recovered space, not a meaningful supply channel for most near-term operational demand. That makes transfers, sub-assignments, leases, corporate reorganizations, routing-security authority and documentation practices more important.

Transfer policy is shaped by procedure. LACNIC's current transfer rules allow IPv4 transfers between LIRs and end users, inside the region and with other RIRs, subject to conditions. The minimum transferable block is a /24. A regional recipient must justify the IPv4 resources under the applicable policies. LACNIC or the corresponding registry verifies the holder and checks for disputes. Intra-regional transfers require a signed legal document. Inter-RIR transfers require coordination with the other registry. LACNIC maintains a public transfer log. Transferors face a one-year ineligibility period for new IPv4 allocations or assignments. Previously transferred addresses cannot be transferred again for one year. LACNIC-issued resources are locked for three years after allocation or assignment. Legacy resources transferred into the region lose legacy status.

Each of these details can be defended on administrative grounds. Together they define the economics of mobility. A one-year lock changes liquidity. A three-year lock changes planning. Recipient justification changes which buyers are easiest to approve. A public log changes transparency and bargaining. Inter-RIR coordination changes settlement risk. Legacy status loss changes valuation. The mailing list and forum are where such rules are framed, challenged and revised. If participation is unequal, the market effect is unequal too.

Waitlist rules are filtered the same way. A queue looks neutral because order is visible. Yet policy can change what counts as fair order. LAC-2025-4 asked whether IPv6 deployment should move an organization upward. The proposal framed the idea as a reward, not a punishment: a carrot rather than a stick. Staff comments reframed the problem around verification cost, legal risk and equality of processing. The proposal did not reach consensus. That outcome may be correct. The important lesson is that mailing-list procedure had to decide whether scarce recovered IPv4 should be distributed by chronological queue, demonstrated IPv6 effort or another principle. That is resource allocation by procedural debate.

Leasing visibility is even more dependent on procedure. A registry can either ignore leasing, prohibit forms of it, regularize parts of it, or make responsibility visible without regulating price. LAC-2025-5 moved toward visibility for third-party sub-assignment. LAC-2023-7 explored temporary transfers as an equivalent or near-equivalent route. Both illustrate how the policy list becomes the forum for deciding whether an informal market is brought into the record, and at what cost. Too little visibility leaves stale records, abuse confusion and RPKI uncertainty. Too much control can turn the registry into a gatekeeper over private use arrangements.

RPKI and security obligations add another control surface. Routing-security rules are not merely technical hygiene. A resource holder's ability to create or maintain route origin authorizations affects whether upstreams, customers and counterparties treat a route as trustworthy. A sub-assignment rule that requires a recipient to have an ASN and IPv6 resources, or that links WHOIS entries to ASN information, changes who can present operational control. A temporary-transfer proposal that mentions RPKI, geolocation, IRR information and MANRS practices changes the burden of legitimate temporary use. Mailing-list participants who think primarily in security terms may see necessary discipline. Small operators may see fixed costs. Both views must be in the record.

Legacy regularisation, abuse contacts and documentation burdens complete the picture. A rule that improves record quality can still fall regressively if it requires translated documents, old corporate proof, public exposure of a small team, or repeated interaction with staff. The fact that a requirement is good for the registry does not prove it is proportionate for every resource holder. Mailing-list procedure should force that distinction into the open.

Small islands, large-country gravity

LACNIC's procedure must be judged in its region, not in an imagined average. Large-country gravity is real. Brazil and Mexico have larger technical communities, larger operators, more lawyers, more address demand, more event presence and stronger capacity to participate repeatedly. Argentina, Chile, Colombia, Peru and other larger markets also contain sophisticated networks and policy participants. These communities are essential to LACNIC. They provide expertise, scale, funding and operational knowledge. They can also dominate the attention economy.

Small-island dependency is different. A Caribbean provider may combine access, hosting, public-sector connectivity, enterprise support and resilience planning with a staff that would be small by continental standards. A /24 may not sound large to a global participant, but it can be a meaningful operating input for a local service, a tourism-related network, a payment platform, a public agency, a disaster-resilience path or an enterprise customer. A policy that adds ten hours of reading, two document cycles and uncertainty about public logs can matter.

Travel culture amplifies the gap. LACNIC events and LACNOG-adjacent spaces can be extremely valuable. They build trust, share practice and make regional coordination possible. But if policy influence depends heavily on attending meetings, the process rewards those with travel budgets and employer permission. Fellowships and remote participation can reduce the gap. They do not erase the advantage of people who have been in the room for years.

Large firms also have another advantage: they can survive complexity. If a sub-assignment rule is imperfect, they can hire counsel, adjust contracts, build tooling, pay fees and wait. If a small provider faces the same complexity, it may simply not use the official path. It may lease through informal arrangements, rely on upstream space, accept weaker documentation, or delay expansion. A policy intended to bring practice into the light can fail if the compliance cost is too high for the intended users.

This is why small-operator analysis should not be a sentimental add-on. It is the stress test. Does the proposed rule work for an operator without a policy department? Does it work for an English-speaking Caribbean network? Does it work for a Portuguese-speaking small ISP outside the main policy circles? Does it work where corporate documents are older, paper-heavy or not easily translated? Does it work where the relevant person is the same person who handles routing incidents and customer escalations?

Large-country expertise should not be discounted. Many policy problems require experienced participants. But the record should distinguish expertise from authorization. A few knowledgeable people can identify a good rule. They cannot, by speaking often, become proof that the weaker parts of the region have absorbed the burden.

What transparency can and cannot do

Transparency is a repeated theme in LACNIC policy debates. It is often the right answer. A public transfer log makes address movement visible. WHOIS registration of sub-assignment can make responsibility clearer. Version histories make proposal changes auditable. Mailing-list archives make debate inspectable. Public staff comments reveal implementation concerns. These are real safeguards against capture and arbitrary power.

But transparency is not always enough. It can shift cost to the reader. It can expose small participants to reputational risk. It can make private commercial arrangements legible without making them fair. It can produce a perfect archive that only insiders can use. It can create confidence among those who already know how to navigate the system while leaving weaker participants outside.

The policy question should therefore be "useful transparency for whom?" A public log that records start and end events for sub-assignments may help counterparties and reduce stale records. If it also creates fear that small networks will reveal business dependencies, adoption may fall. WHOIS identification linked to an ASN may improve responsibility. If updating that information is slow or difficult, informal use may continue. Version histories show what changed. If changes are not summarized in plain language, only dedicated participants will see why they matter.

Transparency should be paired with proportionality. Record enough to protect third parties, routing security, abuse handling, accountability and settlement confidence. Do not demand every commercial detail merely because a registry can ask. Publish enough impact data to discipline debate, but do not expose sensitive operational information. Make logs useful, not punitive.

Transparency should also be paired with metrics. If a rule is adopted to reduce opaque leasing, LACNIC should publish aggregate indicators of whether it did so. If a rule is adopted to support small ISPs, usage by size and geography should be assessed. If a public log creates little adoption, the list should revisit whether the burden was too high. If staff workload rises sharply, the next proposal should know that before repeating the same assumption. The archive tells what was said; metrics tell what happened.

Reducing capture without building a parliament

The answer to unequal participation is not to turn LACNIC into a regional legislature. That would be a category error. A registry needs narrow authority, technical competence and predictable records. It should not hold votes on every distributional grievance or pretend to represent every citizen of Latin America and the Caribbean. The number-resource system works because it is specialized.

The answer is to make specialization accountable to incidence. High-impact proposals should include a participation and cost note. The note should identify who is likely to benefit, who is likely to bear fixed costs, which parties are least likely to appear on the list, whether translation is available early, whether small operators need targeted explanation, whether NIR-related paths are affected, whether Caribbean operators face different constraints, whether the proposal alters transfer liquidity, leasing incentives, waiting-list order, RPKI authority, WHOIS burden, documentation or fees.

Moderators should have an absence checklist. Before calling consensus on resource-economic proposals, they should ask whether the record contains direct input from the affected categories or only claims about them. If the latter, they can still call consensus, but the record should state the limitation. That single habit would reduce the risk of turning active-list convergence into regional consent.

Staff impact statements should be broader. They should cover operational, system and legal impact, as LACNIC already sometimes does. They should also identify external cost categories: applicant document burden, expected processing complexity, potential small-operator impact, privacy exposure, routing-security transition, public-log maintenance, likely support tickets, inter-RIR coordination and risks of activity moving outside the official record. Staff should not decide the policy by cost alone. But cost that remains unstated becomes invisible.

Minority reporting should be normalized. A proposal can reach consensus while preserving the strongest objections. This would help future review and reduce social pressure. Dissenters would know their concerns are not erased. Supporters would have a clearer statement of what trade-off they accepted. Moderators would not need to pretend that consensus means harmony.

Version histories should be accompanied by change narratives. LAC-2025-5 already provides some of this, explaining that comments were reviewed and listing major changes. That practice should become strict for high-incidence proposals: what changed, who asked for it, why it matters, and what was not included. It would make revision less dependent on personal memory.

Finally, implementation feedback should return to the same public list. If a rule created a public movement log, after a reasonable period the community should know aggregate use, support burden, common failure points and whether the expected benefits appeared. If a rule failed to attract small-ISP adoption, that is not merely operational trivia. It is evidence that the policy economics were wrong.

The limited warning from AFRINIC

AFRINIC should appear in this analysis only as a warning about legitimacy when institutional process breaks. Its governance and litigation history is not LACNIC's story. The regions differ, the institutions differ and the immediate problems differ. Still, the warning is useful. A regional internet registry can lose legitimacy not only through a catastrophic technical failure but also through the perception that rules, authority, elections, courts and service continuity no longer form a stable system.

The lesson for LACNIC is preventive. Keep core registry services insulated from political and procedural stress. Keep policy records clear. Keep moderator duties visible. Keep appeals practical. Keep board ratification from feeling like a mystery step. Keep mailing-list archives usable. Keep staff impact comments specific. Keep minority concerns attached to the record. Keep routine registry functions narrow, predictable and reviewable.

The most dangerous failures are often ordinary before they become dramatic. A few proposals where dissent is flattened. A few last calls where late affected parties feel unwelcome. A few staff comments that are too vague to evaluate. A few policies adopted without implementation feedback. A few small markets that conclude the list is not worth their time. A few rules where public transparency creates cost but no usable remedy. Over time, these accumulate into a legitimacy discount.

The AFRINIC warning is not that every registry faces the same crisis. It is that legitimacy is a stock, not a slogan. It is built in small procedural acts and depleted in the same way. Mailing-list procedure is one of those acts.

The registry as listener, not sovereign

LACNIC's development mission is real. A registry serving Latin America and the Caribbean cannot be indifferent to regional capacity, routing security, IPv6 deployment, training, applied research and inclusion. But the more expansive the mission language becomes, the more important procedural restraint becomes. A mailing list should not become a place where the active few can convert regional rhetoric into unexamined control over scarce resources.

The right posture is a disciplined listener. LACNIC should keep the list open, maintain the archives, support the Public Policy Forum, publish versions, preserve multilingual access, provide staff impact analysis, implement ratified policies, and report what happens after implementation. Moderators should protect the conversation from abuse and chaos while also protecting weaker participants from being crowded out by procedural insiders. The Board should ratify or reject with reasons that respect the community record without pretending that the active list is the entire region.

Consensus should remain consensus, not a commodity and not a vote. But consensus should be understood as a conclusion from a record, not a moral claim about community will. The record is stronger when it shows who participated, who was absent, what changed, what burden was considered, what objections remained, and how implementation will be reviewed. It is weaker when it treats a quiet thread as proof that everyone who mattered agreed.

A mailing list converts attention into policy power. That is unavoidable. The question is whether the conversion is audited. In the IPv4 scarcity era, where transfer conditions, waitlist rules, leasing visibility, fee incidence, RPKI obligations, documentation burdens and legacy regularisation can all affect working capital, that audit is not optional institutional housekeeping. It is part of the registry's legitimacy.

LACNIC has many of the necessary materials: a public policy list, a policy system with multilingual proposal records, version comparison, public forums, moderators, last call, board ratification, archives and a manual history. The next step is to treat these not as proof that participation is economically meaningful, but as tools for making it more meaningful. The region does not need the registry to become a parliament. It needs the registry's policy procedure to hear more than those who can afford to speak most often.

The thread is where the work begins. The question is who can afford to stay in it.