Summary

  • The Area Director role combines portfolio management, people management, standards review, cross-area review, process judgment, and IESG collective duties. Current NomCom guidance describes a planning load of roughly 10 to 15 working groups, up to 400 pages of drafts every two weeks, and 15 to 25 hours in many non-meeting weeks, while emphasizing that actual allocation varies.
  • Scarcity does not merely lengthen queues. It shapes which proposed groups receive help, which charters are refined, which chairs and reviewers are trusted, which drafts get early attention, which issues are escalated, and which work reaches formal decision. Those choices can be reasonable, but their criteria and effects should be visible.
  • A defensible oversight system needs workload budgets, public queue states, service expectations, auditable delegation, rotating reviewer networks, explicit triage reasons, conflict controls, and outcome measures that reward early correction rather than raw speed or document volume.

Oversight is an allocation system

The visible authority of an IETF Area Director is easy to describe. Area Directors manage work in a technical area, oversee working groups, sit on the Internet Engineering Steering Group, and participate in final review of documents from across the IETF. The less visible fact is that every one of those functions consumes a finite stock of attention. A decision to read one draft carefully today is also a decision not to read another. A difficult chair conversation can displace charter review. A surge of ballot comments can delay errata. A cross-area incident can absorb time previously reserved for early guidance.

Scarcity is not itself misconduct. No governance body has infinite expert time, and the IETF is built substantially on volunteer contribution. The accountability question is what scarcity does to discretion. When the formal portfolio exceeds the attention available, some choices are made before a document reaches the stage where decisions are prominently recorded. Topics are encouraged, redirected, or left without sponsorship. Proposed charters receive more or less iteration. Chairs receive different levels of support. Directorate reviews are assigned to particular people. Drafts wait in states whose labels do not explain priority.

Informal advice can determine whether proponents persist.

These choices may improve the standards process. An Area Director who sees that a proposal lacks a solvable problem should not consume a year of community effort merely to demonstrate openness. A draft with urgent security implications may deserve earlier review than a routine update. A struggling working group may need intensive management while a mature group proceeds largely through its chairs. Triage is unavoidable.

But unavoidable discretion is still governable. The institution can state the criteria used, reveal where work waits, separate expert delegation from decision authority, monitor unequal effects, and adjust capacity when recurring obligations exceed the role. Without that information, entities cannot tell whether a delay reflects technical difficulty, queue pressure, an unrecorded policy preference, unavailable expertise, or simple loss of follow-through.

The central governance risk is therefore not that busy Area Directors make choices. It is that scarcity moves the decisive choice into a less observable part of the process. Formal openness can coexist with practical exclusion if only well-connected entities know how to obtain early attention, recruit a trusted reviewer, or frame work in a way that survives triage. Oversight bandwidth should be treated as a public institutional resource whose allocation requires reasons proportionate to its effect.

The office combines several different jobs

RFC 2418 gives Area Directors a broad mandate. They are responsible for ensuring that working groups in their area produce coherent, coordinated, architecturally consistent, and timely output. Working-group formation requires the advice and consent of the relevant Area Directors. The charter is negotiated primarily between prospective chairs and those directors before IESG approval with IAB advice. Area Directors can intervene in working-group process when circumstances warrant, agree to chairs and consultants, and participate in rechartering or termination.

The current IETF description of roles adds the cross-area dimension. Area Directors manage IETF work within their technical areas and provide review across areas as IESG members. Review directorates gather experienced entities to assist with specialist review. The IESG overview identifies the steering group as responsible for technical management and the Internet standards process, including approval actions associated with the standards track.

This means an Area Director is not just a senior document reviewer. The office has at least six functions. It is a portfolio manager deciding how area work is organized. It is a people manager selecting and supporting chairs and reviewer leads. It is a process guardian handling disputes and checking consensus. It is a technical reviewer for documents in the area. It is a generalist member of the IESG reviewing work from other areas. It is a coordinator of dependencies such as IANA actions, liaisons, errata, and expert input.

These functions have different time patterns. Ballot deadlines create a short, recurrent review cycle. Working-group conflict arrives unpredictably and may demand immediate attention. Chartering benefits from early, exploratory conversation. Chair development requires continuity and trust. Errata can look deferrable until a deployment problem makes them urgent. Strategic area design requires time without a near deadline. A single weekly hour total does not capture the collision among these clocks.

The functions also create role tension. The same Area Director who advised a group early may later judge whether its consensus and document are ready. The person who selected chairs may need to assess their process failures. The director who wants an area program to succeed also sits in a body expected to stop weak specifications. These are manageable tensions, but limited time makes independent re-examination harder. Under pressure, prior confidence and trusted relationships become efficient shortcuts.

Good institutional design should therefore avoid treating "the AD workload" as one undifferentiated quantity. It should show which obligations consume time, which can be delegated, which require personal judgment, which are deadline-bound, and which are repeatedly displaced. Capacity cannot be repaired intelligently if the institution counts only meetings and documents.

The published workload is already substantial

The 2025 NomCom desired-expertise guidance offers unusually concrete planning figures. It says an Area Director should have enough time to manage approximately 10 to 15 working groups, review up to 400 pages of Internet-Drafts every two weeks, follow document-processing tasks, and attend a weekly two-hour IESG telechat, in addition to area business. It says many Area Directors allocate at least 15 hours per week and others as many as 25, while noting that some combine the role with significant non-IETF responsibilities and that delegation patterns vary.

Those figures are not a time-and-motion study and should not be treated as proof that every director works the same hours or receives the same document volume. They are the IETF's own description to prospective candidates of the scale they should be prepared to carry. They establish that bandwidth is not a speculative concern. The institution recruits for a demanding, recurring portfolio whose reading obligation alone can approach a large technical book every fortnight.

The same guidance says the role has a significant people-management component. Area Directors select working-group chairs and work with them to manage groups. They manage directorates or review teams, recruit and rotate experts, judge consensus, process errata, engage with IANA on designated experts and registry actions, and assist with liaison work. It estimates that management may take roughly 30 percent of IESG time, with the remainder spent on technical matters, while recognizing variation among areas.

A simple division shows why queue effects are inevitable. Four hundred pages do not have equal complexity. A short cryptographic draft can require more judgment than a long descriptive document. Reading is not the end of review: comments must be written, responses understood, revised text checked, and cross-area disagreements discussed. Ten working groups do not impose equal management demands. A mature group with experienced chairs may need little intervention; a new or divided group can consume hours in one week.

The published expectations also omit some hidden switching cost. An Area Director may move in one day from a routing mechanism to a charter dispute, then to an applications document, a personnel conversation, an IANA question, and an IESG telechat. Expertise does not eliminate the cognitive cost of changing problem frames. Nor does a calendar capture the background reading needed to recognize when an issue is routine and when it changes Internet architecture.

The relevant conclusion is not that the role is impossible. Many capable Area Directors have performed it. The conclusion is that oversight quality depends on active capacity management rather than personal stamina. A system that assumes every selected expert will privately absorb surges invites uneven delay, burnout, hidden delegation, and dependence on informal networks.

Scarcity moves power upstream

When final review capacity is limited, institutions rationally try to prevent weak work from reaching the expensive stage. In the IETF, that pressure can improve early discipline. RFC 2418 asks whether a proposed working group has a clear and relevant problem, achievable goals, sufficient interest, adequate expertise, manageable overlap, understood related work, and a genuinely open IETF effort. Area Directors use their best judgment in deciding whether to pursue formation through chartering.

The same upstream position carries considerable discretion. A proposal that receives sustained Area Director engagement can become a well-bounded charter, find capable chairs, attract reviewers, and learn how to meet process expectations. A proposal that receives intermittent advice may return repeatedly with unresolved scope. A proposal judged premature may never enter a formal state that produces a visible denial. Entities experience the result as lack of momentum rather than an appealable decision.

Early filtering is not necessarily exclusion. It can save entities from spending effort on work without a constituency or implementable objective. The accountability problem appears when the criteria are not recorded, comparable proposals receive unexplained differences in assistance, or the route to reconsideration depends on personal access. A soft "not yet" can function as a durable no when no required evidence or next review point is specified.

Bandwidth also shapes charter breadth. A narrowly chartered group may be easier to oversee and more likely to deliver. A broader charter may be necessary where technical decisions cannot be separated without architectural harm. Under capacity pressure, narrowness can become an administrative optimization rather than a technical judgment. Conversely, related work may be combined because there is no director capacity to manage another group, even when a separate forum would produce clearer accountability.

Chair selection is another upstream control. RFC 2418 gives Area Directors authority over the people serving as chairs, and current NomCom guidance emphasizes the management relationship. A strong chair can reduce the director's workload by running open process, documenting consensus, and escalating only genuine problems. This creates a rational preference for people already known to be reliable. If not balanced by rotation and mentorship, the same efficiency can narrow leadership opportunity and make future capacity depend on a small circle.

The institution should therefore record consequential upstream outcomes: proposal received, advice given, criteria not yet met, next evidence requested, charter iteration dates, reasons for combining or separating work, and the status of chair recruitment. Not every exploratory email needs publication. The point is to prevent a practical decision from remaining permanently below the threshold of the formal record.

Queue order is a policy even when nobody calls it one

Backlogs are often described as neutral accumulation. They are not. Once more work is ready than can be handled immediately, some ordering rule applies. The rule may be explicit urgency, document age, dependency, meeting cycle, ease of completion, area priority, reputational pressure, personal reminder, or whichever message is most visible. Silence about the rule does not produce first-come, first-served fairness; it produces an unobserved mix.

Different queue rules distribute benefits and risks differently. Prioritizing old items reduces abandonment but may delay urgent security work. Prioritizing easy reviews improves throughput while complex documents age. Prioritizing documents near publication rewards work that has already received substantial institutional investment. Prioritizing the loudest proponents disadvantages entities who do not know the social expectations of escalation. Prioritizing strategically important protocols requires someone to define strategic importance.

A public queue need not expose confidential personnel issues or tentative security details. It can show the work item, state, age, responsible role, next action, dependency, and a broad priority reason. The Datatracker already provides many document states, including AD Evaluation, Expert Review, Last Call, and IESG Evaluation. The governance opportunity is to make waiting conditions more explanatory. "AD is watching" or "AD follow-up" can cover active analysis, unavailable reviewer, pending author revision, disputed priority, or no scheduled action.

Queue state should identify who can move the item. If authors owe a revision, the Area Director is not the bottleneck. If a specialist review was requested but no reviewer accepted, the issue is network capacity. If the director must decide whether a concern is blocking, the decision sits with the office. If a liaison response is pending, that dependency should not be attributed to working-group delay.

Age thresholds can trigger review without imposing automatic advancement. At a defined interval, the responsible Area Director could confirm the priority, name the next action, delegate assistance, or explain why work remains paused. Repeated aging across a class of items would signal a structural capacity problem rather than a failure of one document.

The aim is not to force equal time for unequal work. Equity requires relevant reasons. An urgent operational fix can move ahead of a speculative extension. A document with downstream dependencies can receive earlier attention. The record should allow entities to understand the reason and allow the IESG to see whether the cumulative pattern systematically deprioritizes certain areas, regions, organizations, or newcomer-led proposals.

Reviewer networks are part of the governing structure

RFC 2418 recognized that many areas form advisory groups or directorates of experienced entities to assist Area Directors, including with specification review. Current IETF material describes directorates and review teams as sources of broad and specialist review. The directorates guidance expressly notes that these reviews can relieve part of the IESG workload while developing expertise and leadership networks.

Delegation is therefore not an exception. It is an operating model. A security, operations, general, Internet, transport, routing, applications, or other specialist review can identify issues earlier and give the Area Director evidence beyond personal expertise. Review teams distribute labor and expose documents to perspectives that the originating group may lack.

The accountability risk is that access to the review network can become equivalent to access to oversight. Who is invited to review? Which reviews are requested early rather than at Last Call? Which comments are treated as major? Whose non-response leads to reassignment, and whose leads to delay? Which experts are repeatedly trusted on a subject? These choices can shape whether a document advances even though the formal decision remains with the Area Director.

A reviewer network can also reproduce concentration. Experienced reviewers are chosen because they have demonstrated judgment. Frequent assignment gives them more experience, visibility, and paths to leadership. New reviewers receive fewer consequential assignments because they are less proven. Without deliberate rotation and mentorship, a capacity tool becomes a closed loop: the institution relies on the people it has already given opportunities to prove reliable.

That does not justify random assignment to high-risk documents. It supports a tiered model. A newer reviewer can work with an experienced reviewer; a review can receive quality feedback; assignment data can show whether the pool is broadening; and area leadership can identify expertise gaps before they become emergencies. Rotation should preserve continuity where a document requires sustained context, but continuity should not become indefinite ownership of a technical domain.

Most importantly, the Area Director must own the consequence. A directorate review is advice. If it leads to a block, charter change, or return to the working group, the Area Director should state the adopted reason and be able to defend it. If the Area Director declines a major review finding, a short disposition can explain why. This prevents outsourced analysis from becoming outsourced authority.

Informal triage is efficient and difficult to contest

Standards work depends on conversation. Entities ask an Area Director whether an idea belongs in an existing group, needs a Birds of a Feather session, conflicts with another effort, or is mature enough for chartering. Chairs ask whether a milestone change is sufficient or rechartering is required. Authors test whether a draft is likely to be sponsored. These exchanges save time and often improve proposals before formal action.

Informality becomes consequential when advice determines access but leaves no stable reason. A entity with a long relationship may receive several iterations of candid guidance. A newcomer may receive a short answer without knowing whether it is a tentative view, a capacity constraint, a technical objection, or an instruction to stop. The difference may arise from time rather than bias, but its effect is still unequal.

The IETF's open participation model makes this especially important. Formal membership does not allocate rights or votes. Participation is by individual contribution. Practical equality therefore depends heavily on whether people can discover the path from proposal to considered work. If the path is encoded in relationships, meeting-room conversations, or unwritten expectations, nominal openness does not guarantee comparable access.

The solution is not to prohibit private or synchronous advice. That would reduce the institution's ability to mentor and solve problems. It is to publish the consequential output. After material pre-chartering discussion, the proponent should be able to obtain a short statement of the problem definition, unmet formation criteria, relevant existing venue, requested evidence, and next checkpoint. Similar notes can accompany a decision not to sponsor an individual submission.

The note should distinguish technical readiness from capacity. "The proposal lacks evidence of implementation interest" is different from "the area cannot support another chartering effort this quarter." Both may justify delay; only one is a defect in the proposal. Conflating them pressures proponents to rewrite technical content when the actual issue is institutional bandwidth.

Capacity-based deferral should lead to a capacity response. The IESG can reassign work, add co-management, recruit a consultant, use a directorate, adjust area boundaries, or state when attention will become available. A technical rejection should lead to specified evidence or an appealable conclusion. Accurate classification makes the remedy match the cause.

Three clocks compete inside the role

Area Director work is governed by at least three clocks. The publication clock is short and visible: Last Calls, telechat agendas, ballot positions, and document revisions create dates. The working-group clock is medium-term: milestones, chair performance, consensus disputes, and rechartering unfold over months. The institutional clock is long: cultivating reviewers, reorganizing areas, improving process, and developing future leaders take years.

The shortest clock tends to win. A document on the next telechat agenda creates an immediate obligation. A prospective reviewer who needs mentoring can wait. A working group whose milestones are stale but whose entities are not complaining can wait. A strategic review of overlapping charters can wait. Over time, the institution may complete documents while weakening the conditions that make future completion possible.

This is a classic oversight trap. Urgent transactions crowd out maintenance of the oversight system. The effect appears later as a smaller reviewer pool, recurring late defects, exhausted chairs, unclear area boundaries, and dependence on incumbent expertise. The Area Director then faces even more transaction work because preventive capacity was not built.

Protected time is one response. The IESG can define a portion of area capacity for leadership development, charter health, and process improvement rather than allowing every hour to be consumed by the publication queue. The allocation need not be identical across areas. It should be visible enough that chronic displacement can be detected.

Another response is to separate service expectations by clock. Ballot reviews need a response interval. Charter proposals need a checkpoint cadence. Chair and reviewer development need periodic portfolio review. Errata need severity-based handling. A single statement that Area Directors should act "timely" does not help when obligations collide.

The clocks should also inform surge support. When one area faces an unusual volume of documents or a major conflict, peer Area Directors can cover cross-area reviews, a directorate can expand, or an additional advisor can be assigned. Current NomCom guidance already expects Area Directors to support peers during surges or absences. The institution should make that support operational before emergency redistribution depends on personal willingness.

More directors can help, but headcount is not a complete design

RFC 7475 changed the definition of an IETF area so that it may be managed by one or more Area Directors rather than being limited to one or two. The RFC explains that the variation in director numbers historically reflected workload and personal preferences and that flexibility allows the IESG to assign enough selected directors to manage working groups effectively. It also warns that if several areas require more than two directors, the IESG should consider alternative ways of organizing the work rather than simply expanding significantly.

This is a sensible caution. Additional directors can reduce portfolios, provide coverage, and broaden expertise. They can also increase coordination costs, blur which director owns an issue, expand the number of cross-area reviewers, and create more positions that the NomCom must fill with credible candidates. If duties remain undefined, adding a person may produce shared ambiguity rather than usable capacity.

Before changing headcount, the IESG should diagnose the load. Is the constraint document reading, chair management, chartering, conflict resolution, directorate leadership, errata, or cross-area review? Is it a sustained area characteristic or a temporary surge? Can work be delegated without transferring authority? Are current area boundaries combining domains that require different expertise? Would administrative support free technical judgment, or is the shortage itself expert judgment?

Where multiple directors share an area, the allocation should be public. Working groups need to know the responsible director, backup, and escalation route. Cross-cutting area decisions should identify whether one director leads or collective agreement is required. Leave coverage should not require entities to restart context. Shared portfolios should be reviewed for hidden imbalance, where one director carries high-conflict groups and another carries routine work.

Headcount also interacts with selection. RFC 8713, the current core of BCP 10 for NomCom operation, provides that approximately half of IESG positions are reviewed each year and that the IESG supplies desired expertise for open positions. Increasing or reorganizing positions changes succession, continuity, and the candidate pool. Capacity planning should therefore occur early enough for the NomCom to recruit against the actual job rather than an inherited title.

The right principle is function before seat. Define the obligations, measure the recurring load, identify what requires office-holder judgment, and then decide whether to add directors, reorganize areas, strengthen reviewer teams, or remove low-value process. Personal endurance should not be the default capacity strategy.

Selection must evaluate allocation judgment, not only technical stature

Area Directors need technical credibility, but the office is not an award for technical contribution. Current desired-expertise guidance emphasizes people management, consensus judgment, prioritization, communication, follow-through, ability to recruit experts, and capacity to review work outside a narrow specialty. Those requirements reflect the real job more accurately than a model of the Area Director as the most senior engineer in an area.

NomCom evaluation should therefore examine how candidates allocate scarce attention. Can they distinguish urgent from merely prominent work? Can they delegate while retaining responsibility? Do they document reasons? Can they support chairs without taking over the working group? Do they invite expert challenge? Can they tell a familiar contributor no and an unfamiliar contributor what evidence would change the answer? Can they stop low-quality work without turning early advice into opaque exclusion?

These qualities are difficult to assess because NomCom deliberations about individuals are confidential, appropriately so. The IESG can help by making role outcomes more measurable and by providing desired expertise that reflects current load. Community feedback can then address observable conduct: response reliability, chair development, review quality, conflict handling, queue management, and treatment of dissent.

Selection is only one control. BCP 10 also includes recall procedures, but recall is a severe instrument and should not be the routine answer to structural overload. If several capable directors struggle with the same volume or role collision, replacing individuals will reproduce the problem. Performance review should separate personal failure from portfolio design, support gaps, or unrealistic expectations.

Terms and succession matter as well. Approximately half-turnover preserves continuity while allowing renewal. A good handover should include open documents, aging issues, working-group health, chartering prospects, reviewer capacity, conflict disclosures, and inherited DISCUSS positions. Personal notes may contain sensitive material, but the institutional state should not vanish with the office holder.

A successor should be free to reassess priorities rather than inheriting every informal commitment. That freedom is more legitimate when previous decisions are documented. Otherwise reconsideration can appear arbitrary, while unquestioned continuity can preserve an unexplained filter. A visible record protects both incoming directors and entities.

Conflicts and networks require procedural distance

The IETF draws leaders from the same community that develops and deploys its protocols. Area Directors inevitably know authors, chairs, employers, implementers, and reviewers. This embedded expertise is valuable. It also means that workload shortcuts can follow trust networks: send the difficult review to the person whose judgment is known, accept the chair's summary from a long relationship, answer the message from a familiar contributor first.

Trust is not corruption. Institutions need it to operate. The governance concern is whether consequential choices can be explained without relying on the relationship itself. Reviewer selection should be tied to expertise, availability, diversity of perspective, and conflict context. Chair selection should use published role requirements and development plans. Queue priority should use the needs of the work, not ease of communication with its proponents.

The active IESG ballot procedure includes Recuse for an Area Director who is an author, working-group chair, or otherwise an interested party in a document. Comparable conflict awareness should extend upstream. An Area Director involved deeply in developing a proposal may need a peer to assess charter readiness. A director whose employer has a direct commercial stake may still provide expertise but should not be the only person making a disputed gate decision.

Procedural distance can be modest. A second Area Director can review a consequential non-sponsorship decision. A directorate assignment can include a reviewer without prior design involvement. A charter dispute can receive an independent area or IESG check. The purpose is not to imply bad faith. It is to reduce the need for entities to prove motive before obtaining review of a decision.

Network concentration should be measured in aggregate. How many review assignments go to the most frequently used reviewers? How often do new reviewers receive paired assignments? How diverse are chair appointments across employer, geography, experience, and pathway into the IETF? These measures cannot decide individual suitability, but they can reveal whether efficiency is narrowing the institution's future capacity.

The strongest reviewer network is not the largest address list. It is one in which expertise is discoverable, assignments are reasoned, conflicts are handled, newer entities can become trusted through supervised work, and the Area Director remains answerable for the formal consequence.

Service levels should reveal waiting, not reward haste

A useful oversight dashboard would not rank Area Directors by how many documents they approve. Volume is shaped by area size and document complexity, and approval is not the only valuable outcome. Nor should it impose a short universal deadline. Some technical and consensus issues require time.

Service levels should instead focus on information and handoff. A new charter proposal receives acknowledgment and a named next step. A draft in AD Evaluation has a visible responsible director and status. A requested expert review has an assignment or a recorded capacity gap. A substantive author response receives acknowledgment. An aging item receives periodic status confirmation. A leadership or process concern has an escalation route.

Time should be measured by state. Waiting for author revision is different from waiting for Area Director review. Waiting for a directorate volunteer differs from waiting for an external liaison. Active technical discussion differs from an item with no recorded next action. State-based measures protect office holders from being blamed for dependencies they do not control and reveal where institutional investment is needed.

Quality measures should look at outcomes. How often does early review remove a serious issue before IETF Last Call? How many final-review concerns repeat classes that could have been found upstream? Do rechartered groups improve milestone delivery? Are chair vacancies filled through a broad pool? Do expert reviews receive dispositions? Are appeals and escalations concentrated around opaque early decisions?

The dashboard should allow narrative context. A severe security event may dominate one quarter. A new area may spend more time chartering. A mature area may process many routine documents. Numbers identify questions; they do not answer them automatically.

Public reporting can be periodic and aggregated, with document-specific status remaining in the Datatracker. Sensitive personnel matters should stay protected. The standard is enough visibility to understand how oversight capacity is allocated, not exposure of every conversation.

Service levels also create a basis for resource requests. If data shows that charter review repeatedly waits while ballot obligations consume capacity, the IESG can add support or protect planning time. If expert-review assignment is the bottleneck, the remedy is reviewer recruitment and tooling. If author revisions dominate, clearer initial feedback may help. Measurement should lead to a matched intervention.

An accountable delegation chain

Delegation should be recorded as a chain from task to advice to decision. The Area Director defines the question. A reviewer or directorate provides analysis. The working group or authors respond. The Area Director adopts, modifies, or declines the recommendation and records the consequence. Each stage has a different owner.

This chain prevents two common failures. The first is anonymous authority: entities are told that "the directorate" requires a change even though the review was advisory and no Area Director has adopted the point. The second is invisible labor: a reviewer supplies substantial analysis, but the final record neither credits nor disposes of it, reducing incentives and making repeated findings hard to trace.

Assignments should state scope and expected timing. A security review of one mechanism is not an invitation to redesign the whole document. An operations review should identify whether it is early advice, Last Call review, or investigation of a particular concern. Scope protects both the reviewer and the authors from unbounded work.

The Area Director should provide disposition at the right level. Minor comments can be left to author judgment. Major findings should receive a reasoned response. A finding that becomes a DISCUSS or prevents advancement needs a clear adopted claim and release condition. A declined finding should not require a long rebuttal, but the record should show that it was considered rather than lost.

Delegation data can support capacity planning. Which teams have unfilled assignments? Which reviewers carry repeated urgent work? Which areas rely on the same cross-area experts? Where do reviews arrive too late to be useful? The answer may support recruitment, pairing, revised timing, or a change in document checklists.

Accountable delegation also makes absence easier to manage. If an Area Director becomes unavailable, a peer can see the question, evidence, and pending decision rather than reconstructing private correspondence. Continuity is a governance benefit, not merely administrative convenience.

A bandwidth budget for each area

Each area should maintain a lightweight capacity budget across recurring functions: working-group management, chartering and new work, document shepherding and AD Evaluation, cross-area ballots, directorate leadership, errata and IANA actions, liaison support, conflict resolution, and leadership development. The budget need not record personal timesheets. It should estimate demand, identify surge risk, and reveal obligations that are consistently displaced.

The budget should be reviewed at least around IETF meeting cycles and before the IESG submits desired expertise and positions to the NomCom. Inputs can include active groups, chair vacancies, planned milestones, drafts approaching publication, aging AD Evaluation items, expected recharters, review-team coverage, and known cross-area initiatives.

When demand exceeds capacity, the response should be explicit. Options include temporary peer coverage, reassignment of working groups, additional Area Directors where structurally justified, a consultant, stronger directorate support, administrative assistance, schedule adjustment, area reorganization, or a transparent deferral of lower-priority work. The response should not default to unrecorded silence.

Capacity budgets can also protect quality. If final-review volume rises beyond the range for meaningful reading, the IESG should not simply celebrate throughput. It should decide which reviews can be delegated, which documents require full personal attention, and whether telechat scheduling should change. Trust-based No Objection positions are part of the present ballot model, but the institution should know when trust is a deliberate review allocation rather than an emergency habit.

The budget must preserve flexibility. Protocol crises and personnel events cannot be forecast precisely. A rigid quota could discourage Area Directors from spending necessary time on a hard problem. The purpose is to show tradeoffs and mobilize support, not to punish variance.

Over several cycles, the budget becomes evidence for institutional design. If one area repeatedly needs extraordinary support, its boundary or staffing may be wrong. If all areas struggle with cross-area ballots, the collective review model may need adjustment. If leadership development is always displaced, the reviewer and chair pool will eventually confirm the cost.

Seven safeguards for scarce oversight

First, classify every delay. Distinguish technical unreadiness, missing consensus, external dependency, reviewer shortage, Area Director queue, and institutional capacity. Different causes require different remedies.

Second, publish consequential triage reasons. Chartering, sponsorship, chair selection, and document priority should produce a stable reason and a condition for reconsideration when they materially affect access to the process.

Third, make queue states actionable. Each active item should show the next action, owner, and checkpoint. Age triggers status review, not automatic approval.

Fourth, delegate analysis without delegating accountability. Reviewer provenance, scope, conflicts, and disposition should be visible. The Area Director owns any formal consequence drawn from advice.

Fifth, broaden the reviewer and chair pool. Pair newer entities with experienced leaders, rotate assignments where continuity permits, and monitor concentration. Capacity should reproduce itself rather than consuming the same small network.

Sixth, budget by function and clock. Protect time for working-group health, new work, and leadership development as well as the immediate publication queue. Use surge support before obligations fall into silence.

Seventh, adjust structure before demanding heroics. Use evidence to consider area boundaries, number of directors, review-team support, tooling, and low-value obligations. Selection should test prioritization and management judgment, but no candidate should be expected to cure a structural overload through private sacrifice.

These safeguards do not remove discretion. They place discretion where it can be evaluated. Technical standards cannot be managed by strict first-come queues or automatic formulas. But departures from ordinary order can carry reasons, and delegated expertise can have a traceable relationship to official authority.

Attention should be answerable

Area Directors occupy a difficult and necessary position. They must see enough detail to stop a technical defect, enough architecture to identify cross-area harm, enough process to protect rough consensus, and enough people to sustain volunteer groups. They must do this while reading a recurring document queue and serving in a collective body whose decisions affect the global Internet.

The IETF's core documents recognize both authority and flexibility. RFC 2418 gives Area Directors substantial responsibilities in formation, chartering, leadership, intervention, and output. RFC 7475 permits the institution to vary director numbers in response to workload while warning that headcount alone may not be the long-term answer. BCP 10 supplies a recurring selection and confirmation process. Current NomCom guidance states the time and management demands with unusual candor.

The next step is to govern the allocation of attention as seriously as the formal vote. A proposal delayed before chartering, a draft waiting for AD review, a chair without support, or a reviewer assignment circulating inside a familiar network can shape standards outcomes before the IESG records a ballot. Those are not peripheral administrative details. They are the operating surface of oversight.

Limited bandwidth will always require judgment. The institution should know what is waiting, why it is waiting, who can move it, what assistance is available, and whether the same entities repeatedly receive or lack access to attention. It should reward the early discovery of defects, not just the speed of final approval. It should cultivate enough reviewers and chairs that delegation expands the community rather than closing it.

Authority is visible in charters, RFCs, ballots, and appeals. Capacity is visible only when the institution chooses to measure it. Once measured, the question becomes tractable: not whether every item receives equal time, but whether scarce oversight is allocated through relevant reasons, reviewable decisions, and a structure capable of sustaining the work. Area Directors need room to judge. The community needs to see where that judgment enters the queue.

Sources