Summary
- AFRINIC reported 1,666 active members at the end of 2018, 1,760 in 2019, 1,836 in 2020, 1,989 in 2021 and 2,506 in 2025. The public record reviewed here does not provide the same clearly labelled year-end active-member count for 2022, 2023 and 2024, so a precise annual per-member series cannot honestly be presented for those years.
- On the two years with matching disclosed denominators, legal fees were about USD 319 per active member in 2021 and USD 350 per active member in 2025. These are accounting allocations, not extra invoices and not evidence that every member paid equally or received an equal benefit.
- Using the 2021 and 2025 active-member counts only as denominator endpoints, the USD 1,250,527 legal bill for 2022 implies roughly USD 499 to USD 629 per member, the USD 1,133,630 bill for 2023 implies USD 452 to USD 570, and the USD 27,322 bill for 2024 implies USD 11 to USD 14. The ranges are sensitivity tests, not estimates of the missing counts.
- Legal fees were 5.23 times the combined published remote-site and computer-expense lines in 2022, 5.90 times them in 2023 and 2.65 times them in 2025. Those two operating lines are not the whole registry-service budget, but the comparison shows why members need legal expenditure translated into service and governance outcomes rather than defended by an aggregate total.
- The appropriate control is a quarterly member-unit legal account: cost per active member and per paying cohort, authority to instruct, matter objective, budget and forecast, operational effect, rights protected, recoveries, outcome and durable control change. Privileged advice can remain protected; the economic mandate cannot remain invisible.
A denominator turns a large number into a governance question
AFRINIC's published legal figures are large enough to attract accusation and abstract enough to resist proof. USD 1.25 million can be described as extraordinary, necessary, wasteful or prudent without saying who bore it, what scale of institution carried it, or what result the money bought. Dividing the amount by active members does not settle those questions. It makes them harder to evade.
The member is an appropriate starting unit because AFRINIC is a membership-based registry. Members pay annual fees, rely on the registry record, elect directors when ordinary governance functions, and bear the consequences when litigation reaches accounts, contracts, staff authority or elections. The unit asks a direct question: if the institution recognised a given legal expense during a year, what was the average amount associated with each organisation in the active membership base?
That amount is not a levy added to every invoice. AFRINIC funds expenditure through a mix of membership revenue, allocation fees, reserves, investment or finance results and timing differences. Members occupy different fee categories. A large Local Internet Registry can pay much more than a small end site. Some litigation may protect the institution as a whole; some may arise from a decision that affected only a subset. A per-member figure is therefore a lens on scale, not a statement of legal liability.
The lens is still useful because it restores the principal to the account. “Legal fees” is an expense class. “USD 500 per active member” is an institutional choice large enough to compare with member fees, technical operating lines and alternative forms of risk reduction. It invites a second question: what did each member receive in legal protection, operational continuity, restored rights or reduced future exposure?
The calculation also reveals a basic disclosure defect. AFRINIC published active-member counts consistently in annual reports through 2021 and reports a 2025 count in its later consolidated material. The same clearly labelled public denominator is not evident for the intervening years in the reviewed record. When spending becomes controversial, a missing denominator is not a minor statistical inconvenience. It prevents members from measuring the unit burden while decisions are still open to influence.
The active-member baseline grew before the institutional vacuum
AFRINIC's annual reports provide a straightforward pre-crisis membership series. The 2018 report states that the organisation had 1,666 active members at year-end. The 2019 report gives 1,760. The 2020 report gives 1,836. The 2021 report gives 1,989. Later consolidated reporting states that the count reached 2,506 at year-end 2025.
| Year-end | Reported active members | Change from prior disclosed year |
|---|---|---|
| 2018 | 1,666 | baseline |
| 2019 | 1,760 | 94, or 5.6% |
| 2020 | 1,836 | 76, or 4.3% |
| 2021 | 1,989 | 153, or 8.3% |
| 2025 | 2,506 | 517 across four years, or 26.0% |
The endpoint growth is institutionally important. A growing membership can spread a fixed central cost across more organisations. It can also create more service demand, billing records, disputes and governance complexity. No conclusion follows from the count alone about whether AFRINIC became more efficient.
“Active” also needs precision. AFRINIC's annual reports use the term for the membership overview, and its billing material has referred to active Resource Members in relation to annual invoices. Election rules add the separate concept of an eligible Resource Member in good standing: current membership, completed formalities and paid fees by the applicable cutoff. A designated voter register is narrower again because an eligible organisation must complete representation and identity steps.
These populations should not be combined. Active membership is suitable for measuring the broad institutional burden. Paying members are suitable for testing fee incidence and collection. Eligible members are relevant to governance rights. Designated voters measure participation in a particular election. A legal dispute may affect all four groups differently.
The public account should publish each denominator with a definition and date. Otherwise a favourable denominator can be selected after the event. Dividing by all active members produces a lower unit figure than dividing by members actually invoiced and collected. Dividing by the roughly 550 organisations appearing on the final 2025 designated-voter register would produce a much higher number, but that would misstate the cost base: failure to designate a voter does not mean an active member ceased funding or depending on the institution.
The calculable legal series begins in 2021, not 2018
The assigned period begins in 2018 because the active-member baseline and pre-crisis institutional scale matter. It does not follow that a comparable legal-cost series exists from that date. AFRINIC's earlier annual reports do not present the same later, separately reconciled legal-fee disclosures in a form that can be joined without restating the accounts.
The later public finance pages identify USD 633,807 for 2021, USD 1,250,527 for 2022, USD 1,133,630 for 2023, USD 27,322 for 2024 and USD 877,929 for 2025. The total is USD 3,923,215. AFRINIC's audited 2022 statements place the 2021 and 2022 legal figures in the expense note, while the newer finance pages give the later payee-level totals.
The gap between the 2018 start date and the 2021 calculable point should remain visible. A zero inserted for 2018, 2019 or 2020 would be false. Ordinary corporate advice may have been classified within professional or administrative expenses. The absence of a separate line does not mean no legal work occurred. A proper long series would require ledger restatement under common categories.
This limitation protects the unit analysis from becoming a morality chart in which a calm earlier era is manufactured by missing data. What can be said is narrower: AFRINIC had a growing active-member base before the later legal-fee series became material and separately disclosed. From 2021 onward, the public figures support a bounded member-unit calculation.
Two years allow exact matching; three require a sensitivity range
For 2021, the operands match. The annual report gives 1,989 active members at year-end, and the audited statements give USD 633,807 in legal fees. Dividing the latter by the former produces USD 318.66 per active member.
For 2025, the later consolidated report gives 2,506 active members, and the finance page gives USD 877,929 in legal costs. The result is USD 350.33 per active member.
These calculations use year-end counts rather than average membership during the year. A member joining in December receives the same denominator weight as one active all year. A more exact measure would use active member-months, matching each month's legal expense or accrual with the population then active. AFRINIC has the billing records needed to produce that measure; the public does not.
For 2022 through 2024, substituting an invented count would create false precision. A transparent alternative is to divide each expense by both disclosed endpoint counts: 1,989 and 2,506. The result shows how sensitive the unit cost is to a plausible population scale without claiming that either endpoint was the actual count in the intervening year.
| Year | Published legal fees | Unit result using 2,506 members | Unit result using 1,989 members | Status of denominator |
|---|---|---|---|---|
| 2021 | USD 633,807 | USD 252.92 | USD 318.66 | Exact year-end count is 1,989; USD 318.66 is the matching result. |
| 2022 | USD 1,250,527 | USD 499.01 | USD 628.72 | Sensitivity range only. |
| 2023 | USD 1,133,630 | USD 452.37 | USD 569.95 | Sensitivity range only. |
| 2024 | USD 27,322 | USD 10.90 | USD 13.74 | Sensitivity range only. |
| 2025 | USD 877,929 | USD 350.33 | USD 441.39 | Exact year-end count is 2,506; USD 350.33 is the matching result. |
The table deliberately displays the non-matching endpoint calculations in the exact years too. That makes denominator sensitivity visible. It also prevents the false impression that every row is equally certain.
Across all five disclosed legal-cost years, applying 1,989 members to every year produces USD 394.49 per member-year. Applying 2,506 to every year produces USD 313.11. This USD 313 to USD 394 interval is not the actual five-year average because the real annual counts and member-months are absent. It is an endpoint stress test showing the order of magnitude. A published annual denominator would narrow it at once.
The average did not arrive as an equal bill
AFRINIC's fee structure makes equal burden unlikely. Its published schedule charges LIR annual fees by aggregate IPv4 category, ranging from USD 1,000 for Micro and USD 1,400 for Extra Small through USD 38,400 for Extra Large. End-site annual fees range from USD 200 for the smallest IPv4 category through USD 2,500 for the largest, with an ASN-only annual amount of USD 50. Academic and research institutions may qualify for discounts.
The 2021 average legal allocation of USD 319 exceeded the published USD 200 and USD 300 annual end-site tiers. The 2025 average of USD 350 did too. That does not mean those members' fees were spent entirely on lawyers or that AFRINIC invoiced a legal surcharge. It shows that the average legal expense was on the same scale as, or larger than, the annual amount charged to some small end-site categories.
For a Micro LIR paying USD 1,000, a USD 350 institutional average represents 35% of the headline annual membership fee. For an Extra Large LIR paying USD 38,400, the same average is less than 1%. But equal division is not how revenue is raised. Larger members contribute more, new allocations generate other fees, and the institution may draw reserves. A true incidence study would allocate legal cost according to actual revenue collected from each cohort.
That study should publish at least three views. The equal-member view shows institutional scale. The fee-weighted view shows how much of each cohort's contribution was absorbed. The resource-weighted view tests whether organisations holding more addresses effectively subsidised disputes that affected all members. None should be treated as the single moral truth.
The distinction matters because a uniform “member interest” can hide conflict. A case concerning a large resource decision may expose one cohort more directly. A governance case about board authority affects every member's ability to control the company. A bank-restraint application can threaten payroll and services across the base. An election dispute affects voting members but also prolongs extraordinary administration funded by all.
Members need to see which cost pool served which institutional purpose. Without that mapping, the largest payers can claim they financed everyone else, while small members can claim they paid for disputes beyond their influence. Both positions may contain part of the truth and neither can be tested from an aggregate payee table.
Legal cost per member is not legal value per member
A lower unit cost is not necessarily better. If legal advice releases restrained operating funds, preserves the authoritative record, prevents an unlawful transfer or restores a valid board, an expensive engagement can create value far exceeding its fee. If an institution refuses necessary advice to keep the average low, members may suffer much larger losses.
A high unit cost is not self-proving waste either. Litigation cost depends on the number of matters, urgency, forum, appeals, opposing conduct, evidence volume, specialist counsel and the consequence of losing. AFRINIC did not control every filing, order or timetable in the disputes surrounding it.
The governance failure is the absence of a matching outcome unit. Members can see totals and some payee names. They cannot consistently see the objective authorised, the phase completed, the exposure reduced, the right preserved, the recovery obtained or the institutional control changed. Cost has a denominator; value does not.
A member-unit account should therefore pair every material matter with an outcome statement. If the objective was to restore access to funds, report the amount or payment capacity restored and the continuity effect. If it was to defend authority over a registry decision, report the legal issue resolved, current finality and operational implication. If it was to organise a lawful election, report the milestone achieved and reusable control produced. Privileged merits advice can remain private.
Value should also be assessed over time. An interim order may protect the institution for three months but generate an appeal. A legal opinion may clarify authority yet fail to prevent recurrence because the bylaws remain unchanged. A settlement may reduce immediate cost but leave no public precedent. Members need both the immediate result and the durable repair.
The correct question is not “Did the lawyers win?” It is “What member-level risk was reduced, for how long, under whose authority, and at what total cost?”
The 2021 and 2022 audited totals show material budget share
The audited 2022 financial statements provide a broader denominator. Total expenses by nature were USD 4,119,317 in 2021 and USD 5,657,583 in 2022. Legal fees were therefore about 15.4% of the 2021 total and 22.1% of the 2022 total.
Those shares are more informative than a legal figure in isolation. In 2022, more than one dollar in five of reported expense by nature sat in the legal line. Employee benefit expense was larger, and many operational activities remained funded, so the figure does not show that litigation replaced the registry. It shows that legal work had become one of the institution's largest identifiable cost classes.
On a 1,989-member denominator, total 2022 expenses represented about USD 2,844 per active member while legal fees represented about USD 629. On a 2,506-member denominator, the corresponding figures are about USD 2,258 and USD 499. Again, neither denominator is established as the actual 2022 count. The comparison demonstrates that member-count uncertainty changes the unit cost but does not make the legal share immaterial.
The budget-share view also corrects a weakness in comparisons with one technical line. Registry services depend on people, facilities, security, contracts, communications and depreciation, not only “computer expenses”. A legal-versus-server narrative can understate the full cost of technical continuity and exaggerate how many machines a legal bill could have bought. Total-expense share avoids that category error while still showing priority.
The member-unit and budget-share measures should be published together. One shows how spending relates to the principal population; the other shows how it competed within the institution. Neither proves causation. A delayed technical project may have been blocked by governance authority rather than money. The disclosure should state the actual funding source and any deferred item.
The narrow operating-line comparison is still a useful warning
AFRINIC's later finance pages separately list remote-site and computer expenses within “other expenses”. Combined, those lines were USD 239,224 in 2022, USD 192,273 in 2023, USD 219,989 in 2024 and USD 330,740 in 2025.
Legal fees were 5.23 times that combination in 2022, 5.90 times it in 2023, 0.12 times it in 2024 and 2.65 times it in 2025. On the 2025 active-member count, the combined remote-site and computer lines equalled about USD 132 per member, while legal cost equalled about USD 350.
These ratios must be read narrowly. Remote-site and computer expenses are not AFRINIC's whole core registry budget. Employee costs, telecommunications, insurance, bank charges, depreciation, professional services and other functions support operations. Some legal work may itself be necessary for service continuity. The lines can also change classification across years.
Why use them at all? Because they are concrete and consistently published in the newer disclosure. They demonstrate the scale at which legal expense competed with visible technical-operating categories. They also reveal volatility: legal fees fell to USD 27,322 in 2024 while the two operating lines continued. A unit-cost system should explain whether the fall reflected reduced activity, a different payer, timing, accrual treatment or a narrow classification. The total alone does not.
The comparison becomes governance evidence only when linked to decisions. If a legal overrun was financed by delaying a remote-site upgrade, state that. If both were fully funded and no operational item moved, state that instead. Opportunity cost should be traced through the budget, not invented from two large numbers.
The 2024 trough makes cash timing and classification impossible to ignore
The legal series is highly uneven: USD 633,807, USD 1,250,527, USD 1,133,630, USD 27,322 and USD 877,929. A fall of more than 97% from 2023 to 2024, followed by a large rise, is not an ordinary stable professional-services pattern.
Several explanations are possible. Activity may genuinely have fallen. A receiver or another party may have paid some cost. Bills may have been recognised in a different period. Work may have shifted into election, receiver, professional or other categories. A dispute may have paused. AFRINIC's public totals do not allow the article to select among these possibilities.
This is why annual cash expense alone is a poor measure of legal commitment. Members need accrued cost, cash paid, unpaid invoices, approved commitments, adverse-cost exposure, recoveries and third-party funding. A year can look cheap while commitments accumulate outside the recognised line. Another can look expensive because prior work is finally invoiced.
The member-unit report should distinguish all of these. “Cost per member” can refer to cash, recognised expense or committed budget; each answers a different question. Cash affects liquidity. Expense affects the accounts. Commitment affects future discretion. Exposure affects downside risk. Publishing one without the others invites a false story.
The same discipline applies to recoveries. A costs order in AFRINIC's favour is not cash until assessed and collected. A costs order against AFRINIC is not a known amount merely because a judgment says “with costs”. The register should show status without pretending that an undetermined order is a final invoice.
Election cost must remain separate from legal cost
AFRINIC's 2025 finance page reports USD 1,043,425 associated with two election efforts. The list includes receiver fees, legal-service providers, election vendors, identity checks, connectivity, travel and logistics. The same page separately reports USD 877,929 in legal costs.
Dividing the election total by 2,506 active members gives about USD 416 per member. That arithmetic is useful for scale and dangerous for classification. The election total is not an additional legal-fee line. Some providers may appear in both contexts, and the accounts must determine how amounts are classified. Adding every item associated with a lawyer to the legal total without reconciliation could double count.
The separate unit figures reveal a wider point: institutional failure creates a professional-cost perimeter larger than litigation. Restoring authority can require a receiver, court directions, nomination and election committees, identity verification, voting technology and repeated logistics. A failed election can make prior spending non-recoverable even when no invoice was improper.
Members should receive an election-cost-per-eligible-member measure alongside cost per active member. The active denominator shows who sustained the institution. The eligible denominator shows the scale of restoring a vote. A third measure, cost per successfully onboarded voter, can test administrative efficiency but must not be used to blame members who did not participate without examining notice and access.
Keeping these classes separate would improve both debates. Lawyers would not be made responsible for every cost of governance collapse. Election administrators would not hide legal work inside a broad event total. Members could judge whether the restoration design produced reusable voter verification, data, procedures and authority clarity.
Active members were principals, not merely cost absorbers
Calling the result “per active member” can sound as though members are customers receiving a service package. They are also corporate principals. In ordinary conditions they elect directors, approve financial statements through the applicable member mechanisms and hold rights under the bylaws and company law. Legal expenditure can protect or displace those rights.
That dual status changes the duty of explanation. A supplier can tell customers that legal cost is embedded in prices. A membership institution must also show who authorised the dispute strategy and how members could review it. The question is not only whether the service remained online. It is whether the expenditure preserved a lawful route back to member control.
During a board vacuum or receivership, the authority chain becomes exceptional. Courts may authorise a receiver. The receiver may instruct counsel within the appointment. Staff may hold records and operational knowledge. Former directors, claimants and committees may have separate representation. A payment from AFRINIC's accounts does not answer who lawfully selected the objective.
Each material engagement should therefore identify the instructing authority, source of power, conflict check and approval limit. If authority changed during the matter, the handover should be recorded. A future board should be able to see which obligations it inherited and which require fresh approval.
This protects counsel as much as members. Lawyers should know who speaks for the client. Auditors should be able to distinguish company expense from a director's or other party's expense. Members should not have to infer authority from a payee name years later.
A member-level account can protect privilege
The standard objection to legal-spend transparency is that litigation strategy and privileged advice cannot be published. That is correct and incomplete. Privilege protects confidential communications and legal analysis. It does not require secrecy about every budget, objective, authority, phase or outcome.
A public matter entry can identify an issue class rather than a sensitive theory. It can state the forum, date opened, institutional objective, authority to instruct, approved budget, cost to date, forecast to the next milestone, funding source, operational risk, current phase, recoveries and closure status. It can describe whether member rights, funds, registry records or governance authority are affected.
The member-unit fields can then be generated without exposing advice: expense per active member, committed budget per active member, recovered cost per active member and net cost per member. Where a matter affects a narrower cohort, a second denominator can be shown with reasons.
An independent audit or finance reviewer should reconcile the public entry to invoices and the general ledger. The reviewer can confirm that privilege-sensitive detail exists and that procurement, authority and conflict controls were followed. Members do not need hourly narratives to know whether a USD 500-per-member matter stayed within budget.
Publication should be quarterly during constitutional crisis and at least annual otherwise. A material overrun should be disclosed before the next irreversible phase where possible. Reporting after final judgment is too late to change counsel, narrow scope, pursue settlement or reconsider an appeal.
The threshold should be set per member before a crisis
Aggregate approval limits become stale as membership grows. A USD 100,000 mandate means USD 60 per member at a base of 1,666 and USD 40 at a base of 2,506. The same nominal threshold can therefore become less sensitive over time even as the institution changes.
AFRINIC could define escalation in member units as well as dollars. For example, a matter budget above a stated amount per active member could require independent review; a forecast overrun above a lower amount could require reauthorisation; and cumulative legal commitments above a percentage of core-service expenditure could trigger a member notice. The precise numbers belong to a valid governance decision, not this article.
Per-member thresholds should never create an absolute prohibition on emergency advice. A court deadline may require immediate action. The rule can permit urgent authority for a short phase, followed by prompt review and a published explanation. Emergency power should buy time to govern, not remove the matter from governance.
The thresholds should also distinguish defensive and affirmative action. Responding to an injunction threatening accounts may require faster authority than initiating a broad enforcement claim or discretionary appeal. A higher burden of alternatives and expected-value analysis should attach when AFRINIC chooses escalation.
Finally, the denominator must be locked to a published date and definition. Decision-makers should not switch from active members to all resource records, or from paying members to registered voters, simply to produce a more favourable number.
Outcomes should be scored against five member interests
A legal mandate can be evaluated without reducing it to wins and losses. Five dimensions are sufficient to make member value visible.
The first is legal protection: rights preserved, liabilities avoided, compliance achieved, assets released, claims resolved and finality obtained. The second is operational continuity: records, payment capacity, staff, publication services and member support protected. The third is governance restoration: progress toward lawful directors, valid member decisions and a clear authority chain. The fourth is economic control: budget accuracy, procurement, recoveries, adverse costs and avoided expense. The fifth is institutional repair: contract, bylaw, treasury, election or dispute controls changed so the same problem is less likely to recur.
Each phase can receive a short evidence-based assessment. A matter may perform well in one dimension and poorly in another. Counsel can obtain urgent access to funds while the treasury remains vulnerable. A receiver can conduct an election while voter-record controls remain disputed. A settlement can control cost while leaving members without a reusable public rule.
The assessment should identify uncertainty. Interim orders are not final wins. Avoided liability may be a range. Operational harm can be prevented without being observable. The reviewer should state what evidence supports each conclusion and what remains unknown.
The per-member figure then becomes a disciplined question: for USD 350, USD 500 or USD 600 per active member, which dimensions improved, by how much, and for how long? If the answer is unknown, that is not proof of waste. It is proof that the institution did not build an outcome record adequate to the expenditure.
Cost concentration should be visible before another appeal or claim
Payee tables show that large portions of annual legal cost can be concentrated with one firm. Concentration may be justified by expertise, continuity, urgency or a connected set of proceedings. It also creates dependency, weakens price comparison and makes scope drift harder to detect.
The member-unit account should therefore show concentration without implying misconduct. It can state the percentage of total legal spend with the largest provider, the procurement route, the reason competition was impracticable if applicable, the approved rate or fee structure, and whether an independent cost review occurred. Sensitive rates can be reviewed confidentially while aggregate assurance is published.
Fee structures allocate risk differently. Hourly billing gives flexibility and transfers duration risk to AFRINIC. Fixed-phase fees improve predictability and can generate change disputes. Caps control exposure but may encourage narrow scope. The institution should explain why the chosen structure matched the matter.
Matter concentration is as important as provider concentration. Several firms may work on one constitutional dispute, making payee diversification look healthier than the underlying risk. Conversely, one firm may handle routine corporate work and several urgent cases. Reporting must roll payees up to matters and matters up to institutional objectives.
An appeal or new claim should come with the cumulative member-unit cost of the dispute, not only the next retainer. Sunk cost is not a reason to continue. The decision should compare expected legal, operational and governance value of continuation, settlement, narrower relief or no action.
The missing 2022-2024 denominator should be repaired, not estimated away
The easiest response to a missing annual active-member count is to interpolate between 1,989 and 2,506. That would create a neat chart and unsupported facts. Membership may have grown unevenly. Closures, arrears, suspensions, new allocations and classification changes could alter the count. Institutional disruption may have affected record maintenance.
AFRINIC can publish the authoritative figures from its membership and billing records. Each count should include the definition of active, snapshot date, Resource Member and Associate Member treatment, accounts in recovery, terminated accounts and any restatement. If historical states cannot be reproduced exactly, the institution should publish a range and explain why.
The repair should go beyond three numbers. Monthly active-member and member-in-good-standing counts would allow member-month denominators. Fee revenue by category would permit incidence analysis. A reconciliation from the membership register to the financial statements would let auditors confirm that the principal population and revenue base align.
This is not clerical perfectionism. During governance crisis, membership data determine voting, notice, fee collection and the legitimacy of the route back to ordinary control. The same record that should answer who could vote should also answer how many organisations bore the institution's costs.
Publication can preserve confidentiality. Counts and category totals do not require exposing account names, resource holdings or arrears. Any small-cell privacy risk can be handled by grouping.
A legal-spend constitution would make the unit durable
AFRINIC should adopt a standing legal-spend policy before the next crisis. It should define matter classes, authority to instruct, emergency limits, procurement expectations, conflicts, privilege handling, budget phases, member-unit thresholds, forecast frequency, settlement review, appeal review, adverse-cost accounting, recoveries and closure assessment.
The policy should protect a core-service floor. Legal spending necessary to preserve records, lawful payment capacity and compliance can sit within that floor when properly authorised. Broader merits litigation should not consume funds or attention needed to maintain the verified registry state and a route to member government without explicit review.
It should also require a reserve distinction. A legal reserve estimates known and contingent case exposure. A continuity reserve protects defined registry functions. Combining them allows the first dispute to consume the money represented to members as operational runway.
Quarterly reporting should show both in months and per member. How many months of core service remain after committed legal budgets? What is the legal commitment per active member? How much is immediately available if accounts are restrained? What recoveries are expected, and with what confidence?
No rule can guarantee cheap litigation. It can make authority, cost and trade-offs visible while they remain reversible. That is the purpose of the unit.
The number is not a verdict; it is a right to an account
AFRINIC's disclosed legal total of USD 3,923,215 across 2021 through 2025 does not prove that its lawyers charged improperly, that every case was avoidable or that members received no value. The public record is limited public evidence for those conclusions. It does show that legal expense became material relative to the institution, its active membership and identifiable operating lines.
The two matching annual calculations are clear: about USD 319 per active member in 2021 and USD 350 in 2025. The intervening endpoint sensitivity ranges rise to roughly USD 499 to USD 629 in 2022 and USD 452 to USD 570 in 2023 before collapsing in 2024. Across five years, the endpoint stress test is roughly USD 313 to USD 394 per active member-year. These figures describe scale, not guilt.
The unanswered question is what institutional asset the member can point to in return. A clearer contract, released bank account, preserved registration, valid board, reliable electorate, bounded receiver mandate, recovered cost or durable dispute rule can justify legal expense. Another hearing date, unexplained payee total or authority dispute cannot justify itself.
Active members should not need access to privileged advice to know what they collectively financed. They are entitled to a reconciled denominator, lawful authority chain, budget, outcome and remaining exposure. They are also entitled to know when legal strategy displaced a service investment and when it did not.
The per-member number does not turn AFRINIC's legal bill into an equal invoice. It turns the bill into an accountability unit. In a membership institution, that unit should have existed before the first million-dollar year.
Sources and analytical limits
AFRINIC's 2018, 2019, 2020 and 2021 annual reports provide the disclosed active-member counts of 1,666, 1,760, 1,836 and 1,989. AFRINIC's later consolidated annual and financial report provides the 2,506 active-member count for year-end 2025. The reviewed public record does not provide an equivalently clear active-member snapshot for 2022 through 2024; no interpolated count is presented as fact.
The official finance pages for 2022, 2023, 2024 and 2025 provide the cited legal totals and the remote-site, computer and election-expense lines. AFRINIC's 2026 special general meeting material and audited 2022 statements supply the 2021 legal reconciliation and the 2021-2022 total-expense comparison.
The membership fee page establishes the published LIR and end-site categories used only for scale comparison. It does not reveal the actual category distribution, fee revenue by cohort or source of funds used for each legal payment.
All per-member figures are arithmetic allocations. Year-end active members are not average member-months. The 2022-2024 ranges use the disclosed 2021 and 2025 counts solely as sensitivity endpoints and are not estimates of those years' membership. The article does not authenticate invoices, assess professional rates, allocate fees to individual cases, infer wrongdoing, or conclude that a named provider, director, receiver, employee, litigant or member acted improperly.