The Member Vote on a Budget Already Committed

Summary

• Regional Internet Registry members do not all vote on an annual budget in the same way. RIPE NCC members vote on a charging scheme and discuss the Activity Plan and Budget, while its Executive Board adopts the final plan; APNIC's Executive Council adopts its budget after member feedback, subject to an exceptionally demanding member amendment right; ARIN's Board of Trustees approves the annual budget. Calling every meeting a member budget vote hides the first accountability question: who has legal authority to change the cost?
• Timing can narrow even a lawful decision. RIPE NCC's 2025 sequence placed the member vote on how to raise expected income in May, discussion of the cost plan in October and final Executive Board approval in December. ARIN approved its 2025 budget in October 2024, then in January 2025 authorised its Treasurer to approve reasonable future-year commitments, with a future hotel contract given as an example. Each act is transparent; the combined sequence shows why commitments need dates.
• Published budget figures are plans, not audited outcomes. ARIN's 2025 board-approved cash operating and capital budget was USD 31.638 million, including USD 23.992 million for salaries, benefits and fringe costs for an approved employee count of 106. The personnel line equals about 75.8% of the approved cash budget, but it is not proof that 75.8% was legally irreversible on the approval date; only contract, payroll and termination data could establish that.
• Member denominators require the same caution. RIPE NCC reported 1,039 cast votes and 5.3% eligible-member turnout in May 2025, and 801 cast votes with 4.1% turnout in October. Rounded turnout cannot reconstruct an exact electorate, while an LIR account, legal member, registered vote and cast ballot are different units. ARIN's Service Members and General Members, and APNIC's direct and National Internet Registry relationships, create different authority and incidence populations.
• Every RIR should publish a commitment calendar before the first binding annual decision: proposed cost, board reservation date, tender date, contract signature, cancellation deadline, cash paid, cost to exit, service effect and member decision point. A budget should also show 60%, 75% and 85% pre-commitment sensitivities. The meaningful voting number is the avoidable budget still open to choice, not the gross total projected on the screen.

The meeting begins after the money has acquired a history

The annual financial presentation usually starts with a clean table. Income appears on one side, expenditure on the other. Staff explain changes from the prior year. A treasurer describes reserves and investment performance. Members ask about travel, legal costs, security, headcount or a fee increase. A chair moves to the next item or opens an electronic ballot.

The table makes the budget look as though it is being created in the room. It is not. By the time members see it, much of the money has a history. Employees have been hired under labour law. Office leases have years remaining. Software licences renew unless notice is given months ahead. Meeting hotels require deposits. Security suppliers, auditors and insurance policies run across financial periods. Strategic projects have passed board checkpoints, consumed design work and created dependencies. Grants or contributions may have been promised. A cancellation may save future cash but impose an immediate penalty.

The difference between a proposed amount and an avoidable amount determines whether financial participation is real. Members can reject a USD 10 million line on paper, yet the institution may owe USD 8 million under existing commitments and need another USD 1 million to exit safely. The apparent choice is ten; the practical choice is one. If the commitment history is invisible, management can describe the budget as member-shaped while the member encounters only the cost of reversing decisions already made.

This problem does not require bad faith. Organisations must make commitments before every future detail is known. A registry cannot wait for an annual meeting before renewing every certificate, retaining every engineer or booking every room. Boards have fiduciary duties and executives need delegated authority. The defect arises when ordinary administrative necessity is converted into a one-way ratchet: staff and contracts commit the institution early, then their existence becomes the reason members cannot change scope later.

The correct question at a budget meeting is therefore not only, “Do you approve this total?” It is, “What remains changeable now, who committed the rest, on what date, and what would reversal cost?”

There is no single RIR member budget vote

The five RIRs are often discussed as though they share one constitutional form. Their financial authority differs. Any timing audit must begin with the legal act rather than the meeting label.

RIPE NCC's published Activity Plan and Budget Approval procedure states that the Executive Board prepares and adopts the annual Activity Plan and Budget. Management prepares a draft; the Board reviews it; members receive the draft before the General Meeting and discuss it; the Board considers feedback and signs off the final document. Members exercise real rights elsewhere, including voting on the Charging Scheme and audited Financial Report, electing directors and adopting resolutions. Discussion of a budget is not formal member adoption of that budget.

APNIC describes a similar allocation. Its Executive Council page says the EC determines the activity plan, adopts a draft budget, presents it for member feedback at the AGM and is responsible for final adoption. The APNIC By-laws give members broad authority to determine general policies, examine and approve accounts, elect the EC and review or amend EC decisions. Yet changing an EC decision requires two thirds of the votes of the entire membership, not merely two thirds of votes cast.

ARIN places the annual budget squarely with its Board of Trustees. Its corporate materials say the board-approved annual budget drives the internal work plan. Trustee responsibilities include developing and approving the strategic plan, operating plan and annual budget. On 23 October 2024, the Board unanimously approved the 2025 budget after recommendation from the Finance Committee. Members elect trustees and receive financial reporting, but the public act approving that budget was a board vote.

LACNIC has its own association structure, Member Assembly and Fiscal Commission, while its website publishes annual budgets and member resolutions on fee matters. A member vote on a fee formula, an Assembly's review of accounts and a board's operational budget are related without being interchangeable.

The title “member vote on a budget” should therefore be read as a governance test, not a claim that every RIR holds the same ballot. In some institutions, the member act is a charging decision. In others, it is feedback, approval of completed accounts, election of directors or a high-threshold power to reverse the board. The accountability question remains the same: when the member's legally available lever is exercised, how much prospective cost can it still change?

RIPE NCC separates the revenue vote from the cost decision

RIPE NCC provides the clearest documented sequence because its Board minutes state the distinction directly.

In March 2024, the Executive Board considered charging options for 2025. The 174th Board minutes say members would be asked to vote on how to split the burden needed to provide expected income, rather than on the total income of the budget. All three options were designed to generate EUR 41.1 million. The minutes add that the Board would decide the cost side in December after consultation with members at the October General Meeting.

That is unusually candid governance evidence. It shows three different decisions:

1. Management and the Board identify an expected revenue requirement.
2. Members choose how that requirement is distributed through the Charging Scheme.
3. The Board later adopts the expenditure plan after member discussion.

The member vote was meaningful. Different charging options could place different burdens on accounts and resources. But it did not determine whether total planned spending should be EUR 39 million, EUR 41.1 million or another amount. The expected total had already been built into every option. A member who wanted both a different distribution and a lower institutional cost did not receive that combined choice in one ballot.

The timing reinforced the separation. The Charging Scheme vote occurred at the May General Meeting. A draft Activity Plan and Budget was published for autumn discussion. In October, members discussed the plan. In December, the Executive Board approved the final budget. RIPE NCC's final 2025 document budgeted EUR 41.1 million of income and EUR 40 million of expense and planned 197.2 full-time-equivalent positions.

None of this makes the process illegitimate. A board can reasonably need a revenue envelope before completing every cost allocation. Members may prefer to vote on a fee rule with months of notice. Autumn consultation can change the plan. The institutional weakness is that the public record does not automatically show what portion of the cost remained avoidable at each point.

By May, which 2025 contracts had renewal deadlines before October? Which positions had already been approved or recruited? Which capital projects had entered supplier selection? By October, what could member opposition still cancel without material cost? By December, what was genuinely prospective rather than the formal recognition of commitments made during the year?

A timeline without a commitment ledger answers when documents appeared. It does not answer when financial choice disappeared.

ARIN makes future commitments visible, but not yet measurable in one place

ARIN's public Board minutes expose the issue from the other direction. The board approves the budget, then separately delegates authority for expenses extending beyond the current year.

On 14 January 2025, the ARIN Board authorised the Treasurer to approve expenses beyond the current budget year when they were reasonable and consistent with expenses typically included in ARIN budgets and activity plans. When a trustee asked for an example, the CFO offered a hotel contract for a future Board or members meeting. The minutes describe this as an annual authorisation.

The decision is not suspicious. Booking a large meeting close to its date can be expensive or impossible. Multi-year licences can cost less than annual purchasing. A security or colocation contract should not expire merely because the calendar turns. A treasurer is a plausible control point for recurring future obligations.

The resolution nevertheless proves the analytical point: some future-year spending can be committed before the future-year budget is approved. The authorisation is broad enough to include costs that are reasonable and typical, while the public minutes provide only one example. Members reading the next annual budget cannot tell from the budget alone which lines already contain obligations approved under this authority.

The 2025 budget supplies scale. ARIN's board approved a combined cash operating and capital budget of USD 31.638 million. The budget listed USD 23.992 million for salaries, benefits and fringe costs connected to an approved employee count of 106. The ratio is about 75.8 percent of the approved cash budget.

That percentage is not a commitment ratio. Some employees may leave; vacancies may remain open; benefits vary; labour obligations differ; capitalised software work creates accounting adjustments; and an annual personnel budget includes future services not yet performed. Treating the whole personnel line as legally sunk would be as misleading as treating it as instantly avoidable.

It does, however, show why gross budget votes can exaggerate discretion. Even before leases, software, colocation, insurance, meetings and debt service are considered, most planned cash spending is associated with people. If member or board objection comes after hiring, compensation and organisational design decisions, the available response is not a clean deletion of the line. It is a staffing plan with notice periods, service consequences and possible severance or recruitment loss.

ARIN already publishes valuable components: annual budgets, audited accounts, Board minutes, Finance Committee responsibilities and member-meeting financial reports. The missing bridge is a consolidated future-commitment schedule showing how annual authorisations migrate into the next budget.

APNIC shows the difficulty of a formal override right

APNIC members possess an apparently powerful right: they can review or amend EC decisions by a two-thirds majority of the votes of the entire membership. The denominator changes the practical meaning.

A threshold based on the entire membership treats non-participation as a vote against change. If 20 percent of members cast ballots, even unanimous support among those voters reaches only one fifth of the required base. The right remains legally real and can matter in an extraordinary mobilisation. It is not an ordinary budget-editing tool.

The difficulty grows when commitments precede mobilisation. Members first need to discover the disputed cost, understand it, organise across a large region, satisfy voting and meeting rules, and reach an electorate-wide supermajority. During that time, management continues implementing the approved plan. Contracts may be signed and staff assigned. By the time the threshold is approached, reversal costs become part of the board's defence.

APNIC's planning calendar has also evolved. A December 2024 APNIC publication said the 2025 Activity Plan had been completed, approved by the Executive Council and published before the new year, rather than first appearing at the February AGM as in prior practice. Earlier publication improves operational clarity and gives members a full document before the year begins. It can also mean that the formal AGM discussion follows EC approval unless a separate pre-approval consultation has occurred.

Again, sequence is not outcome. Earlier publication may expand practical influence because members can respond before much annual spending occurs. Later final adoption may allow more feedback but leave the organisation without an approved plan at the start of the year. The quality test is whether comments can change commitments and whether changed lines are identified.

APNIC supplied a useful illustration in 2026 when explaining that software-application savings would become more visible only as licensing contracts expired. Contract expiry is the economic clock. A member can demand immediate savings, yet the institution may be unable to realise them until a licence reaches a notice or renewal date. The correct report should show that date before the budget decision, not introduce it after the saving is requested.

Audited actuals, approved budgets and commitments are three ledgers

Financial debate becomes unreliable when three different records are blended.

The audited ledger describes what happened under an accounting framework. It records recognised revenue, expense, assets, liabilities, commitments disclosed under the applicable standards and the auditor's opinion. It arrives after the year. Members can approve or receive it, but they cannot use that approval to avoid the expense already incurred.

The budget ledger describes what the board plans and authorises. It may use cash, accrual or management views, with adjustments to reconcile to accounting presentation. ARIN explicitly shows a board-approved cash operating and capital budget beside a U.S. GAAP budget. RIPE NCC shows activity costs, full-time equivalents, operating expense and capital expense. LACNIC publishes planned categories and expected financial results. A budget is prospective when approved, but it does not show which parts are already contracted.

The commitment ledger describes when the institution became economically or legally constrained. It includes signed contracts, purchase orders, employment obligations, leases, cancellation deadlines, board-designated contributions, non-refundable deposits, project dependencies and termination costs. Some commitments appear in audited notes. Many operational commitments are material to governance without meeting the accounting threshold for separate disclosure.

A member needs all three. Audited actuals reveal whether management spent as reported. The budget reveals intended allocation. The commitment ledger reveals remaining choice.

Without the third record, an executive can make two statements that are both true and jointly evasive: “Members were consulted before the budget was final,” and “We cannot change this line because the organisation is already committed.” The first describes document timing. The second describes economic timing. Accountability requires the dates connecting them.

The committed share needs a sensitivity range

Public data rarely reveal enough contract detail to calculate one exact pre-committed percentage. Pretending otherwise would be false precision. A sensitivity table is more honest and more useful.

Take ARIN's USD 31.638 million 2025 approved cash operating and capital budget as a scale base. This is not a claim about ARIN's actual commitment rate. It asks what member or board discretion would look like under three hypothetical conditions:

Share economically committed before the decisive review	Amount already constrained	Gross amount still open before transition costs
60%	about USD 18.983m	about USD 12.655m
75%	about USD 23.729m	about USD 7.910m
85%	about USD 26.892m	about USD 4.746m

The final column still overstates freedom. Cancelling open work may carry security, legal, service and staff consequences. Some uncommitted costs are essential. Conversely, a signed contract may contain a termination right that makes part of it avoidable. The table is a first approximation, not a result.

Its value is political clarity. If 85 percent is already constrained, a meeting considering the whole USD 31.638 million should not be described as choosing the entire budget. The practical envelope is below USD 4.746 million before exit cost and service priority. If only 60 percent is constrained, members or directors retain much more leverage.

Every RIR should publish its own range using contract data. A central estimate can be accompanied by a lower case that treats cancellable obligations as open and an upper case that treats likely renewal and retention costs as committed. The report should identify the largest assumptions. Precision should improve over time as procurement and staffing records become structured.

Sensitivity is also necessary for member incidence. The population funding the budget may be counted as legal members, resource service accounts, voting members, fee-paying organisations, National Internet Registries or downstream networks. The amount per “member” changes with the unit. A range is preferable to a clean division by the most convenient denominator.

Participation numbers cannot be made more exact than the record

RIPE NCC's 2025 annual reporting provides a useful example of denominator discipline. It records 1,039 cast votes and 5.3 percent eligible-member turnout for the May General Meeting, and 801 cast votes with 4.1 percent turnout in October. Those figures show low participation relative to eligibility. They do not supply an exact electorate through simple division because the percentages are rounded and because registered votes, eligible members, accounts and ballots are separate measures.

The October meeting is especially relevant to budget discussion. Members could give feedback on the 2026 plan, but only a subset registered or cast votes on formal resolutions. Budget comments can come from non-voting entities, mailing lists and open sessions. A cast-vote count is therefore neither the total audience for the budget nor evidence that the silent majority accepted every cost.

Low turnout does not invalidate a lawful decision. Associations need rules that allow action without universal attendance. It does weaken claims that an expenditure has a broad member mandate, especially when members did not vote on the expenditure total at all.

ARIN's membership categories create another denominator boundary. A 2024 member-meeting presentation reported 23,478 Service Members and 1,988 General Members. Those numbers should not be casually added into one electorate because General Member status and service relationships have different meanings and can overlap under the organisation's rules. A budget funded by registration services may affect a larger population than the body with voting rights.

APNIC's direct member base does not capture every operator bearing costs through National Internet Registry arrangements. LACNIC categories reflect resource holdings and fee schedules. Across all five RIRs, “the members approved” is incomplete unless the statement names the eligible body, votes cast, authority exercised and relationship to the payers.

For a budget timing audit, four denominators should be published: fee-paying accounts, legal member organisations, voting-eligible members and votes cast. Where downstream organisations finance the institution indirectly, a fifth estimate should be labelled separately. Missing or overlapping data should remain visible.

Staffing is a commitment curve, not a fixed block

Personnel is often the largest RIR expense. It is also the easiest line to discuss badly.

Critics can treat every position as removable overhead. Management can treat every position as essential and legally fixed. Neither view reflects the curve of commitment.

Some roles protect essential services and require scarce knowledge. Removing them would increase operational risk or replacement cost. Some are attached to time-limited projects. Some vacancies can be frozen immediately. Contractors may have shorter notice than employees. Permanent staff may have statutory protections, accrued leave, pension or severance obligations. Attrition can reduce headcount without dismissal but only slowly and unpredictably. Reorganisation can save money after an initial cost.

The budget should therefore divide personnel into service-critical floor, currently filled adjustable roles, vacant positions, approved new hires, fixed-term roles and externally contracted labour. It should show the earliest cash-saving date under ordinary attrition, hiring freeze, redeployment and formal reduction scenarios. Salary privacy does not prevent publication of aggregate bands.

ARIN's 106 approved employees and USD 23.992 million personnel line establish a planning scale, not a finding that any job is unnecessary. RIPE NCC's planned 197.2 FTEs establish another scale under a different currency and activity model. LACNIC's 2026 budget places USD 6.043 million of salaries and personnel expense against USD 12.156 million of operating expense, about half. APNIC reports its own staffing and cost structure.

These figures should not be compared as productivity rankings without service scope, geography, salary markets, outsourced labour and accounting treatment. They show why timing matters. Once the institution hires to a planned level, next year's member discussion begins from an inherited payroll. A credible budget cycle must expose proposed positions before recruitment, not only the aggregate personnel line after contracts are signed.

Leases, licences and venues create silent vetoes

Long-term commitments can exercise a veto without anyone casting a ballot.

An office lease can run for years. The annual budget contains only one year's rent, while the decision was made earlier and exit may require a sublease, buyout or move. A cloud, cybersecurity or software licence can renew automatically. A meeting venue can impose escalating cancellation charges. A data-centre contract can be commercially cancellable yet operationally difficult to move because systems and security controls depend on it.

These obligations are not inherently wasteful. Long contracts can lower price, preserve service and secure capacity. The governance failure is presenting the annual instalment as though it remains an annual choice.

ARIN's future-commitment resolution makes hotel contracting explicit. APNIC's software explanation makes licence expiry explicit. RIPE NCC's board minutes and budgets discuss office and infrastructure investment across years. Each example points toward a common register:

Commitment	Required public timing field
Employment plan	approval, recruitment start, contract date, earliest saving date
Office or data-centre lease	signature, term, break date, exit cost, migration dependency
Software or security licence	renewal date, notice date, alternatives, transition period
Meeting venue	reservation, deposit, cancellation ladder, expected attendance
Multi-year project	board gate, vendor award, milestone, remaining cancellable spend
Contribution or grant	pledge date, payment schedule, conditions and revocation right

The register need not publish vendor secrets or individual employment terms. It should show when governance choices close. Members can then entity before the organisation incurs a penalty and can distinguish a genuine sunk cost from an executive preference to avoid inconvenience.

Sunk cost is not a mandate

Once money has been spent, it should not determine future spending merely because abandoning the project would make the past expense look wasted. That is the classic sunk-cost problem.

RIR governance adds a legitimacy twist. Management may argue that a project received authority because previous budgets funded it. Yet previous budgets may have been board decisions after limited member feedback, and each year's commitment can be cited to justify the next. Historical expenditure then manufactures its own mandate.

A project should instead face a continuation test. What future outcome remains? What further amount is required? What is the cost of stopping? Which service depends on completion? Could a smaller alternative preserve the benefit? The money already spent belongs in the explanation, not in the decision criterion.

This is particularly important for broad institutional activities that are useful but not indispensable to number uniqueness or record integrity. A conference series, measurement platform, office renovation, training programme or outreach initiative may create value. Its existence and past cost do not make it equivalent to RDAP, reverse DNS or RPKI continuity. If members reject further compulsory funding, the board should examine sponsorship, user funding, partnership, transfer or closure rather than treating history as constitutional protection.

The same rule applies to staff. Institutional knowledge is valuable, and abrupt removal can be destructive. But hiring a team around a programme should not permanently remove the programme from member choice. The commitment calendar gives enough lead time to handle people fairly while preserving governance authority.

A vote needs alternatives that are executable

A binary approve-or-reject motion can be coercive when rejection has no prepared path.

If a board presents one budget days before a meeting, members know that rejection could leave payroll, security and vendor authority uncertain. Responsible voters may approve despite opposing parts of the plan. The institution can then cite the result as endorsement of every line.

Meaningful choice requires executable alternatives. At minimum, the budget should include a base plan, a core-continuity plan and a reduced-scope plan. Each should state service outcomes, staffing changes, contract consequences, reserve use and implementation dates. Members or directors can then decide among real operating states rather than between the management plan and chaos.

The alternatives should be published before commitment deadlines. A reduced plan introduced after the office lease renews or new staff are hired is cosmetic. The calendar must work backward from the member or board decision to the last dates on which major costs can still change.

This does not turn members into line managers. Members decide the envelope and institutional scope. Boards remain responsible for fiduciary judgment. Executives implement. The alternatives simply stop management from using implementation complexity to erase the higher-level choice.

Where members do not formally adopt the budget, the same discipline improves board accountability. Trustees should record which member-supported alternative they considered, why they rejected it and what commitments constrained them. Elections can then evaluate a decision rather than a general promise of prudence.

The annual calendar should run backward

Most budget calendars run forward: management prepares a plan, the board reviews it, members comment, and the final version is adopted. A commitment-aware calendar begins with the last meaningful member or board decision and runs backward from contract deadlines.

Assume final adoption occurs in December. A software contract requiring six months' notice must be reviewed by June. A venue deposit due in May needs member-facing justification before that date. Recruitment for January positions may begin in September, so the positions should appear in a preliminary plan by July. A three-year strategic project entering procurement in March should not wait for the autumn budget document.

The result is a rolling sequence:

1. Publish the current multi-year commitment register at the start of the year.
2. Flag all renewal and cancellation deadlines occurring before the annual meeting.
3. Present early decision papers for commitments that would constrain the next budget.
4. Record board reservations separately from signed contracts.
5. Publish the draft budget with committed, cancellable and open amounts.
6. Provide executable alternatives before members comment or vote.
7. Adopt the final plan and explain changes caused by member input.
8. Reconcile actual spending, commitment changes and audited results after year end.

This sequence accepts that some financial decisions are continuous. It also ensures that annual governance remains connected to them. The institution no longer waits until a polished budget exists to reveal that its important components were decided months earlier.

Commitments can be transparent without exposing security

Registries have legitimate confidentiality concerns. Publishing the exact termination clause of a security supplier or the salary of a named engineer could harm the institution. Legal advice may be privileged. Vendor pricing can be commercially sensitive.

These concerns do not justify an empty commitment record. Aggregate disclosure can show category, amount range, approval date, term, earliest exit, authority and service dependency. Auditors or an independent finance committee can verify protected details. A public report can say that three critical security contracts worth a combined range renew in the second quarter with six-month notice, without naming systems or prices precisely.

Materiality thresholds can prevent noise. A contract below a defined amount may appear only in aggregate. All multi-year commitments affecting essential services, all obligations above a percentage of annual expense and all decisions that remove a planned member choice should receive individual entries.

The same approach works for legal cost. Members do not need privileged strategy to know the approved budget, amount committed, matter category, decision authority and effect on operational reserves. For personnel, they need aggregate role groups and change costs, not personal files.

Transparency should focus on authority and timing. Who could bind the institution? When did they do so? Which later body could still change the decision? What cost had already become unavoidable? Those facts are rarely security secrets.

Reserve use can hide a late vote

Large reserves make commitment timing more consequential. If the annual fee or income decision does not cover the chosen scope, the board can use reserves to honour prior commitments. That may preserve services. It can also neutralise member resistance.

Suppose members reject a fee increase after contracts and hiring have occurred. Management can say that immediate cuts would be harmful and draw reserves for a year. At the next meeting, the depleted reserve becomes evidence that fees must rise. The institution has converted a late member objection into a temporary financing problem and returned with a stronger stability argument.

The answer is not to forbid reserve use. It is to connect a draw caused by rejected revenue to a scope decision. If the board uses reserves to bridge commitments, it should publish which obligations predated the member act, who approved them, when they expire and what change will occur before the next cycle. Replenishment should not be automatic.

This rule separates external shock from governance mismatch. A sanctions-related payment delay may justify replenishing a buffer once funds arrive. A cyber incident may justify rebuilding after costs are verified. Persistent refusal to fund an optional programme should lead to programme review, not a claim that the reserve minimum requires higher fees.

The commitment ledger and reserve policy are therefore one system. Commitments reveal why a draw is necessary. Reserve triggers prevent the draw from preserving those commitments indefinitely. Together they keep continuity finance from becoming a substitute for timely consent.

A budget-rights statement for members

Every RIR should give members a short financial rights statement before the annual cycle.

Members should know which body adopts the budget, which body sets fees, which accounts they approve, what petition or amendment rights exist, and the applicable thresholds. The statement should distinguish consultation from consent. It should say whether silence has any legal effect.

Members should receive the commitment register in time to affect renewals. They should be able to propose a reduced plan using the institution's cost assumptions. Management should respond to material alternatives with reasons and transition costs.

Members should see turnout and incidence denominators separately. A low-turnout vote may remain valid under the bylaws, but reporting should not inflate it into universal approval. Fee-paying accounts, eligible voters, registered votes and cast ballots should not be merged.

Members should receive a post-decision change log. If feedback altered a line, the document should identify the amount and reason. If no material change occurred, the board should say so. Consultation is more credible when its effect can be measured.

Finally, members should see the next opportunity for change. A rejected proposal should not disappear into another annual cycle after the relevant contracts renew. The institution should identify the date by which the issue must return to preserve practical choice.

These rights are modest. They do not give a meeting authority to interrupt live registry systems. They make financial authority legible before it hardens.

The strongest defence of early commitment

Management has a serious counterargument. RIRs deliver critical services in specialised labour markets. They need stable staff, secure facilities and reliable suppliers. Publishing every tentative option can create uncertainty, raise vendor prices and make employees fear that annual politics will end their work. Boards are elected or otherwise accountable and are legally responsible for the organisation. They should not be paralysed by plebiscites.

That defence supports delegation. It does not support invisible timing.

A board can receive authority to commit within a member-approved multi-year envelope. It can protect sensitive negotiations. It can sign urgent security contracts. The condition is that the envelope has boundaries, exceptions are reported and recurring future-year commitments return for periodic review.

Stability for staff also improves when the institution distinguishes core roles from discretionary expansion. Engineers protecting records should not be made bargaining chips in every fee debate. Programmes outside the service floor should not borrow those engineers' indispensability. A multi-year workforce plan can show which capabilities are protected and which growth depends on future authorisation.

Vendors benefit from clarity as well. A published procurement calendar and decision authority can reduce uncertainty. Competition can occur before commitment. Confidential bids remain protected while contract duration and exit rights receive oversight.

The choice is not annual chaos or executive freedom. It is structured delegation with a visible commitment frontier.

The metric is the avoidable budget

Gross expenditure measures organisational scale. Audited expenditure measures recognised history. The avoidable budget measures current governance power.

For each annual plan, the board should publish five numbers:

• Total planned expense.
• Amount already paid.
• Amount contractually committed but unpaid.
• Amount cancellable with stated exit cost and timing.
• Amount open to decision without material penalty.

It should then show an essential-service adjustment. Some open spending cannot prudently be removed because it protects records or security. Some committed spending can be exited because the service is no longer justified. The adjusted avoidable amount is the practical decision envelope.

The ratio of adjusted avoidable expense to total planned expense can become a governance indicator. A ratio of 35 percent months before member review may be healthy for an organisation with substantial recurring obligations. A ratio of 5 percent would suggest that participation arrives after decisions. The target will vary, but the trend should be visible.

The second indicator is pre-authorisation lead time: the median number of days between member-facing disclosure and the date a material commitment becomes expensive to reverse. The third is consultation effect: the value of changes made after member input. The fourth is rollover: recurring costs continued without a fresh scope decision. The fifth is exception use: future commitments approved outside the normal calendar.

These measures turn budget accountability from ceremony into operating evidence. A board can show that early commitments were necessary, that members saw them in time, and that alternatives remained executable. Critics can identify where choice vanished. Auditors can verify the dates without deciding policy.

Financial accountability should happen before sunk cost

The member vote on a budget already committed is not primarily a problem of ballot design. It is a problem of sequence.

RIPE NCC's public record shows a member charging vote, an autumn budget discussion and a later Board cost decision. ARIN's record shows Board budget approval and a separate annual delegation for future-year commitments. APNIC shows member feedback, EC adoption and a formal override threshold based on the entire membership. LACNIC shows how fee resolutions, published budgets, fiscal oversight and association decisions occupy different layers.

These systems contain meaningful accountability. They also permit financial choices to harden before dispersed members can organise around them. Staff, contracts and strategic projects move continuously; member governance appears episodically. Without a commitment calendar, the continuous layer wins by default.

The reform is straightforward. Name the deciding body. Publish the first reservation and final signature dates. Separate audited actuals from budget plans. Label member denominators. Show 60, 75 and 85 percent commitment sensitivities until exact data improve. Present alternatives before cancellation deadlines. Tie reserve draws to the commitments that caused them. Measure how much changed after members spoke.

A lawful vote can still be weak. A well-attended meeting can still arrive late. A transparent budget can still conceal the moment of commitment. The decisive evidence is the amount members or their elected board can avoid without breaking essential services when the choice is formally offered.

If that amount is small, the institution should stop calling the gross total a choice. It should move governance earlier.

Evidence and limits

RIPE NCC's authority allocation comes from its Activity Plan and Budget Approval procedure. The 174th Executive Board minutes establish that 2025 charging options shared a EUR 41.1 million expected-income total, that members were asked how to divide the burden and that the Board would decide the cost side after autumn consultation. The 2025 Activity Plan and Budget and later publication establish planned income, expense, staffing and the calendar; planned figures are not audited outcomes. RIPE NCC's 2025 Annual Report supplies the reported cast-vote and turnout figures, whose rounded percentages are not reverse-engineered into a false exact electorate.

ARIN's 23 October 2024 Board minutes establish Board approval of the 2025 budget. The 14 January 2025 minutes establish the annual future-commitment delegation and hotel-contract example. The 2025 Budget supplies the USD 31.638 million approved cash operating and capital total, USD 23.992 million personnel plan and approved employee count of 106. These are approved planning figures; ARIN's annual-report archive separately identifies audited outcomes.

APNIC's authority and threshold come from its Executive Council description and By-laws. The December 2024 Activity Plan publication establishes the change to pre-year completion and approval. APNIC's later explanation that software savings emerge as licensing contracts expire is used only to demonstrate commitment timing, not to judge whether any licence was excessive.

LACNIC's 2026 Budget supplies the planned personnel and total-expense comparison. It is not treated as an audited result or as proof of a specific member voting sequence.

No exact RIR-wide pre-commitment percentage is asserted. The 60, 75 and 85 percent table is a labelled sensitivity using ARIN's published budget as a scale base. A complete calculation would require contract signatures, renewal and cancellation dates, employment obligations, leases, purchase orders, project gates, exit costs, protected supplier terms and the exact authority attached to each member or board act.