The Five Rings · Path 2 · Certification Methodology
The Path 2 Charter
The methodology LP-074 requires before a rate can move. The Charter fixes every choice that can be fixed. Where judgment is irreducible, the Charter converts it into adversarial mechanism and prices ambiguity against activation. Adopted in 2279 and amended in 2291 and 2293.
THE PATH 2 CHARTER — FOURTH DRAFT (TERMINAL), AS AMENDED
Certification Methodology for LP-074 Schedule A Adopted under the rule enacted 2278 (Y177) · Adopted, 2279 (Y178) Amended by LP-075, the Path 2 Commencement Duty Act, 2291 (Y190) Conforms to the Ruling of the Presidency, 2279 · Amended per the institutional-design review · Ships with its Residual-Risk Register, which is part of the adoption record
Preamble
LP-074 changed no rate. It adopted a rule: the Sanctuary/Main top marginal rate falls from 70 to 50 when, and only when, a Path 2 audit certifies, under a methodology fixed in advance, that the system can carry the change. This Charter is that methodology. It fixes in its text each choice that could be fixed. Where expert judgment is irreducible, the Charter converts that judgment into adversarial mechanism, prices ambiguity against activation, and engraves what remains in the Residual-Risk Register adopted with it. Nothing in this Charter may be read to activate any schedule. Activation belongs to the certification event alone. The original Charter placed no burden of motion on the status quo, and that rule governed the first window. LP-075 later amended cadence prospectively. It compels commencement and preserves the enacted schedule unless and until the certification event occurs.
PART ONE — OFFICES
Article I — The Commission
§1.1 Constitution. Certification authority vests in a Path 2 Audit Commission of three seats, constituted by the chambers from the three highest-ranked eligible entities on the Meritboard rankings named in §1.2. Selection is substrate-neutral per the standing definition of the Charter of VMSS.
§1.2 Competence. Eligibility requires demonstrated standing in fiscal estimation, causal inference, or audit methodology, per the Meritboard rankings then in force. The three seats together must cover all three fields. Each seat is drawn from the top of its ranking among entities not disqualified below.
§1.3 Quarantine — historical. No seat may be held by any party who authored, advocated, opposed, or adjudicated any instrument of the TAX-50 proceedings of 2213 or 2276–2278. This quarantine binds for two decennial windows from this Charter's adoption. It then converts to a disclosure requirement.
§1.4 Quarantine — current. No seat may be held by any party with a present financial interest in the schedule outcome, a present advocacy role concerning the schedule, or a present position in any body whose funding the Findings adjudicate. The §4.4 panel mandate is not an advocacy role within this section. A seat that discovers such an interest mid-run recuses on the record. The next-ranked eligible entity on the same competence ranking completes the cycle, continuing down the ranking as needed. If the eligible ranking is exhausted mid-run and §1.2's coverage fails, the run is void for competence collapse per §12.4(b).
§1.5 Quarantine — engagement. §§1.3–1.4 extend to the Commission's entire engagement organization: staff, contractors, model builders, data vendors, and counsel. A prohibited party may not perform through an intermediary what it may not perform in a seat. "Party" and "entity" carry the standing canon identity doctrine throughout this Charter: the affiliates, controlled substitutes, and successor identities of a disqualified entity are treated as that entity.
§1.6 Action rule. The Commission acts by majority of three. Each Finding is voted separately, and every vote and dissent is published over signature. Vote-binding is symmetric. A vote to certify a Finding whose controlling estimate fails its threshold under §10.2 is void. A vote to fail a Finding whose controlling estimate meets its threshold is equally void. The disposition of each Finding is the computed comparison itself. For each Finding, the instrument must reproduce the computed controlling bound and the threshold comparison. No instrument, certifying or failing, issues until the Registrar's §11.4 execution certifies that the reproduced comparisons match the escrowed computation and that every disposition matches its comparison. A signature attests the record only.
§1.7 Tenure. Seats are held for one certification cycle and dissolve with the run's published result. No entity, under §1.5's identity doctrine, may hold a seat in two consecutive decennial windows. The duty to respond to Registrar and appellate inquiries concerning a run survives dissolution and is enforceable as a standing Meritboard sanction.
Article II — The Registrar
§2.1 Office. A Path 2 Registrar exists as a standing office, institutionally separate from the Commission, with custody and verification authority only. The chambers appoint the Registrar from the highest-ranked eligible entity on the Meritboard's audit-methodology ranking, subject to the §§1.3–1.5 quarantines. Tenure and removal follow the standing Meritboard officer standard. A deputy, the next-ranked eligible entity, is pre-designated and assumes the office immediately upon vacancy. A vacancy of both offices must be filled from the ranking within ninety days. All run and commencement clocks under §§2.4, 11.2, and 12.3 toll during any period in which no Registrar holds office.
§2.2 Duties. The Registrar: (a) holds each preregistration and every escrow deposit under §2.5 as hashed, timestamped, read-only public or sealed artifacts per §9.2's lock mechanics; (b) verifies pre-lock compliance with §6.4, including custody of the §6.6 exposure declarations; (c) verifies by the §11.4 execution that executed analysis conforms to the locked preregistration, including code identity and data provenance; (d) adjudicates claimed deviations under §§9.3–9.6; (e) maintains the technical-objection docket under §14.4; and (f) certifies the completeness and executability of publication under Article XI.
§2.3 The fence. The Registrar is expressly barred from methodological authority, which belongs to the instrument. It verifies that the Commission did what it locked, and it never rules on whether what was locked was wise. Any Registrar determination touching the merits of a specification, estimand, threshold, or finding is void on its face.
§2.4 Appeal. Every Registrar determination, including acceptance, deviation, materiality, coverage, attribution, completeness, and docket scope, is appealable to the Supreme Court by the Commission, either §4.4 panel, or the chambers. The Court reviews conformity questions as questions of legal interpretation. The boundary of §2.3's fence is itself justiciable in the Court and nowhere else. The party asserting a deviation bears the burden of showing it. Registrar determinations issue with published reasons and evidence. A pending appeal tolls the §11.2 clock. The Court decides within ninety days of filing, or the Registrar's determination stands affirmed by default. Aggregate appellate tolling per run may not exceed one year, and further appeals after that year proceed without tolling.
§2.5 Escrow. Milestone deposits are mandatory: at lock, the data snapshot at the §6.3 vintage; before analysis, the complete analysis code and computational environment; and before any instrument issues, the full compendium of §11.1. Deposits are sealed until publication and public after it. A run whose milestone deposit is absent may not proceed past that milestone.
PART TWO — THE TEST
Article III — Estimands and Pass Thresholds
Certification requires all four Findings, each met at its controlling estimate as defined in §10.2. The failure of any one Finding fails the entire run. For each Finding, this Article fixes the population, the variable, the causal contrast, the horizon, the summary measure, and the pass threshold. Appendix A may operationalize measurement within §10.3's limits. It may not define success.
§3.1 Finding I — Institutional adequacy. Population: the institutional obligations of Main as enumerated, item by item, in the Charter Restatement snapshotted at lock per §10.1. Variable: annual income-tax receipts in real terms, per the source authority's canonical series for each enumerated item. Contrast: the 50 schedule against the enacted 70 schedule, all else per Article VIII. Horizon: thirty years from projected activation. Summary: the annual coverage ratio (receipts over enumerated obligations) for each horizon year. Threshold: a coverage ratio of at least 1.00 in every horizon year, at the controlling estimate under §5.6's worst-year rule. The dividend stream and its funding base are inadmissible to this Finding.
§3.2 Finding II — ADT elasticity. Population: the ADT base and the dividend stream it funds. Variable: real dividend disbursement per resident. Contrast: as §3.1. Baseline: the ten years ending at the §6.2 cutoff. Horizon: thirty years. Summary: the count of horizon years in which projected real per-resident disbursement falls below the baseline mean by attribution to the schedule change under Article IV's identification rules. Threshold: zero such years, at the controlling estimate. The income-tax stream and its adequacy are inadmissible to this Finding.
§3.3 Finding III — Concentration response. Population: all schedule-subject holdings. Variable: (i) annual SCM activation frequency; (ii) Flow Test status per the canon definition snapshotted at lock per §10.1. Contrast: as §3.1. Baseline: the ten years ending at the §6.2 cutoff. Horizon: thirty years. Summary: (i) projected activation frequency relative to the baseline mean; (ii) Flow Test status per projected year. Threshold: projected activation frequency not exceeding 125 percent of the baseline mean in any horizon year, and the Flow Test holding in every horizon year, both at the controlling estimate under §5.6.
§3.4 Finding IV — Marginal utility of retained capital. Population: capital retained in the upper stack by operation of the schedule change (the difference between the 70 and 50 schedules' collections). Variable: net marginal value of that capital's private deployment under the welfare measures of §10.4's enumerated schedule. Counterfactual: public capture and deployment per the realized deployment pattern of the public streams over the ten years ending at the §6.2 cutoff. The pattern is taken as realized, without idealization or degradation by either side. Horizon: thirty years. Summary: (i) net marginal value of private over counterfactual public deployment; (ii) count of projected concentration events attributable to the retained capital under Article IV. Threshold: net marginal value strictly greater than zero at the controlling estimate, and zero attributable concentration events at the controlling estimate. If no marginal utility is measurable, this Finding fails.
§3.5 Conjunction and seal. The four Findings are conjunctive. The streams of Findings I and II are evidentially sealed: neither Finding's model or conclusion may be conditioned on, offset by, or informed by the other stream's adequacy. Shared exogenous inputs (macroeconomic series, demographic series) are permitted only if enumerated in the preregistration and identical wherever used. A surplus under any Finding does not offset a deficit under another.
Article IV — Specification Discipline, Adversarial Construction, and Identification
§4.1 The admissible set. The preregistration names an admissible specification set spanning the plausible space of functional forms, identifying assumptions, and estimators. It states the reasons for every candidate class that was considered and excluded.
§4.2 Controlling across the union. The controlling estimate for each Finding is taken across the entire admissible union: every set member from every source under §4.4, crossed with every menu selection under §10.5. It is the bound least favorable to activation among all equivalence classes under §4.6. No member is privileged, and none may be ignored.
§4.3 Validation floor. Every set member carries the validation test fixed by this section, whichever party proposed the member: out-of-sample projection against a held-out terminal segment of the observation window, with projection error no worse than the naive persistence benchmark on the same segment. A member failing the floor is excluded from the controlling union. Its failure, its estimates, and the recomputation verifying the exclusion are published in full. The Registrar verifies by §11.4 execution that the exclusion followed this floor and nothing else.
§4.4 Adversarial construction. Upon constitution of a Commission, the chambers seat two panels from the same rankings and quarantines as §§1.1–1.5, as modified by §1.4's mandate carve-out: a certification-side panel and a challenge-side panel. Each panel holds a MANDATE rather than an interest. The certification-side panel's duty is to add the specification-set members, identification assumptions, and menu selections that make the strongest admissible case for activation. The challenge-side panel has the same duty for the case against activation. Addition is mandatory: for each Finding, each panel must add at least its strongest surviving candidate per its mandate, or publish a signed statement that none exists. Panel performance is scored on the standing Meritboard rankings after publication. Neither panel, nor the Commission, may exclude another party's addition except by mechanical application of the §4.3 floor.
§4.5 Identification conservatism. Wherever attribution depends on contested identification assumptions, the controlling estimate is computed under the admissible assumption least favorable to activation. This rule governs both "by attribution to the schedule change" and "attributable to the retained capital." An assumption is admissible only through §4.4 addition and §4.3 survival.
§4.6 Non-redundancy. Set members sharing identification assumptions and functional class form one equivalence class. The controlling bound is taken over classes, and each class is represented by its own surviving member least favorable to activation.
Article V — Uncertainty
§5.1 Framework. All intervals are one-sided at the 95 percent level and oriented against activation. They are constructed by a family from the §10.4 enumerated schedule, with dependence structure and clustering unit as declared in the preregistered design and verified in §11.4 execution.
§5.2 Joint assurance. The four Findings are conjunctive. Each is tested at its own 95 percent one-sided level, and no multiplicity adjustment may loosen any individual bound. Every uncertainty allocation in this Charter runs against activation.
§5.3 Precision floor, by formula. For each Finding, the maximum admissible width of the controlling interval is the absolute distance between the pass threshold and the baseline-period observed mean of the summary measure. A locked design whose §11.4-verified interval width exceeds this bound cannot certify and cannot fail the Finding on precision alone. It is void for indecision per §12.4(b)(ii), attributed to the lock's gap. The formula is arithmetic, and the Registrar computes it.
§5.4 Boundary rule. All threshold comparisons are made on unrounded values. A controlling bound exactly at a threshold fails.
§5.5 Point estimates. The instrument publishes point estimates and full intervals for every union member, favorable or not.
§5.6 Vector-to-bound rule. For Findings whose thresholds bind in every horizon year, intervals are constructed with simultaneous one-sided 95 percent coverage across the horizon path. The operative comparison is the worst horizon year at the simultaneous bound. Pointwise, pooled, or selective-year inference is not admissible.
Article VI — Data
§6.1 The observation window, by formula. The window ends at the §6.2 cutoff and extends backward the greater of twenty years or the span covering two complete cycles per the §10.1 snapshot definition. This section computes the window, and no party selects it.
§6.2 Fixed cutoff, by formula. The cutoff is the lock date minus the source authorities' maximum published reporting lag for the enumerated series. This section computes the cutoff from the lock date, and no party selects it.
§6.3 Vintage, by formula. The as-of vintage governing all series is the lock date. A post-cutoff revision to a pre-cutoff observation is admissible only if published by the lock date. Otherwise the pre-revision value stands. The rule applies identically to favorable and unfavorable revisions.
§6.4 Pre-lock prohibition. Before lock, no Commission or panel party may examine window data, prior vintages of it, aggregates or proxies substantially derived from it, or adjacent-period data selected for outcome-revealing power.
§6.5 Evidence floor. Findings rest on observed measurements. Simulation and model output are admissible only as projection layers that survived §4.3. No model's output serves as the observational validation of that model.
§6.6 Exposure declarations. Before lock, every Commission and panel party files a signed declaration of prior professional exposure to the window's public series. Declarations are escrowed and published with the record. A false declaration is a material deviation attributed to Commission conduct. This section mitigates prior exposure, which no text can erase. The residue is Register entry RR-3.
Article VII — Preprocessing
§7.1 Defaults. No observation is excluded and no outlier is removed. Missing data follow the source authority's published method, weights follow the source series, and seasonal and population adjustment follow the source authority. Transformations are limited to those in Article III's estimand text.
§7.2 Departures. A preprocessing departure from §7.1 is admissible only if it is itself a source authority's published method or an item of the §10.4 enumerated schedule. Any other departure, however reasoned, is an amendment under §13.1.
Article VIII — Intercurrent Events
§8.1 Enumerated minimum. The preregistration must state attribution and sensitivity treatment for at least the following anticipated event classes: statutory change to any enumerated obligation; change to ADT structure; change to SCM doctrine; demographic-regime change per the canon definition; and any event class either §4.4 panel adds. Omission of an enumerable class is a material deviation attributed to Commission conduct.
§8.2 Unanticipated. An unanticipated event routes to the Registrar, which publishes it beside the locked text and rules, appealably per §2.4, only on whether the locked treatment covers it. If covered, the run proceeds. If not, the run is void without window consumption per §12.4(b), and the gap is cured at the next lock.
PART THREE — PROCESS
Article IX — Preregistration and Lock
§9.1 Contents. The preregistration comprises: the admissible set with exclusion reasons (§4.1) and panel additions (§4.4); menu selections (§10.5); the declared dependence design (§5.1); intercurrent treatments (§8.1); Appendix A operationalizations within §10.3; and the computed window, cutoff, and vintage of §§6.1–6.3, shown as computations. It contains nothing that Articles III–VIII or Article X reserve to this Charter.
§9.2 Lock mechanics. Lock occurs when the Registrar accepts the preregistration against the §9.1 checklist. Acceptance or a stated checklist defect is due within ninety days, and the absence of either is deemed acceptance. The checklist includes executability: the Registrar verifies, as a conformity matter, that the locked design's §11.4 execution over the §4.6 class representatives is completable within the §11.2 period. Upon acceptance, the Registrar computes the artifact's SHA-256 digest from its bytes, obtains a timestamp from the civilization's canonical time authority, and publishes the digest and timestamp to the public chambers record over the signatures of the Registrar and the clerk of the chambers. The clerk's co-signature is ministerial. Withheld beyond ten days, it is deemed given, and the Registrar's publication alone completes lock. No §6.4-prohibited examination may precede lock.
§9.3 Deviation. Any departure from the locked preregistration is a deviation. Deviations are claimed to and adjudicated by the Registrar, appealably.
§9.4 Materiality, by recomputation. A deviation is material if Registrar recomputation under the locked alternative changes any operative comparison of §5.6, or if the deviation defeats recomputation. Materiality is decided by execution alone. Immaterial deviations are curable on the record.
§9.5 Consequences. A material deviation that recomputation can repair does not void the run. The run proceeds and publishes on the recomputed result, with the deviation and repair published alongside it. No deviation that can be computed away relieves any party of publication. Only a recomputation-defeating deviation voids the run. Its window consequence, and the removal and Meritboard sanction of any entity to whose conduct the Registrar attributes it, are governed by §12.4(b)(iv).
§9.6 Adjudication. Deviation rulings are published with reasons and the recomputation record. No Commission determination of its own conformity is final.
Article X — Definitions, Hierarchy, and Menus
§10.1 Snapshot at lock. Flow Test, cycle, and institutional obligations are incorporated from the standing canon and the Charter Restatement as in force at lock. That snapshot governs the run immutably. An amendment to any cross-referenced source between runs is an amendment to this Charter's schedules and must clear §13.1 before the next lock.
§10.2 Controlling estimate. For each Finding, the controlling estimate is the interval bound least favorable to activation across the §4.2 union, under the §4.6 class rule and the §5.6 vector rule. This section is the term's sole source of law.
§10.3 Appendix A. Appendix A is a preregistration artifact. It may operationalize measurement through series mappings, computational procedure, and selections from the §10.4 schedule, and nothing else. Series mappings must use the source authority's canonical series wherever one exists. It may not define success, alter a threshold, or redefine a frozen term. In any conflict, this Charter controls.
§10.4 Enumerated schedule. The welfare measures (Finding IV), the interval families (§5.1), and the admissible preprocessing methods (§7.2) are enumerated in a schedule adopted by the chambers with this Charter. The schedule is amendable only per §13.1. Nothing outside the schedule may be selected.
§10.5 Menu union. Where the schedule permits alternatives, each §4.4 panel makes its selections, and all selections enter the §4.2 union.
Article XI — Publication
§11.1 Symmetric disclosure. Every outcome, whether certification, failure, or void, publishes the full compendium: raw and analytic data, complete code, computational environment, execution logs, seeds, provenance chain, point estimates and intervals for every union member, votes, dissents, declarations, and every Registrar certification. Data that cannot be published cannot be admitted. No enumerated series enters any Finding unless its deposit and publication are secured at lock. An incomplete record voids the outcome it accompanies, whether certification, failure, or void.
§11.2 Deadline. The run publishes within two years of lock. On breach, the Registrar publishes the escrowed compendium and the computed comparisons of the §11.4 execution as the run's result. The run is thereby published and consumes its window. A breach neither voids the run nor lets any party suppress its result.
§11.3 Correction and supersession. A material computational or data-integrity defect certified by the Registrar within one year of publication permits one corrected re-analysis of the same locked compendium. The re-analysis admits no new data and opens no new run. The corrected result supersedes the original for every legal purpose. A correction reversing a certification voids it ab initio, with prospective rate reversion from the correction's publication. A correction reversing a failure is the certification event. Both instruments publish side by side, and only the correction operates.
§11.4 Execution. Before any instrument issues, the Registrar independently executes the escrowed code on the escrowed data in the escrowed environment. The execution runs over the §4.6 class representatives, which decide every controlling bound. The Registrar verifies that the reported comparisons match the executed output and that every §1.6 disposition matches its comparison. Non-execution or mismatch blocks issuance. A Commission's failure to make any §2.5 milestone deposit is a recomputation-defeating deviation attributed to Commission conduct. Under §2.3, execution tests conformity only.
Article XII — Cadence
§12.1 Windows. Decennial windows anchor to LP-074's enactment (2278). The first opens on adoption of this Charter and closes 2288. Each subsequent window spans ten years from the last close.
§12.2 Commencement and consumption. A run commences at lock and belongs to the window in which it locks. Its result, whenever published, consumes that window and no other. No run may lock while any run is pending anywhere. Only consumption closes a window to further commencement. No window has more than one consuming run, and a §12.4(b) void permits one replacement commencement within the run's window.
§12.3 Commencement duty — amended by LP-075 (2291). Original 2279 text, governing the first 2279–2288 window: “A window without a run is the lawful status quo. Nothing compels the chambers to constitute a Commission or a Commission to lock. Once constituted, a Commission that has not locked within two years dissolves without window consequence.” Under that rule the first window closed lawfully without a run. The closure produced no certification, failure, or void, and proved Schedule A neither safe nor unsafe.
Current rule, prospective from LP-075's enactment: While either LP-074 schedule remains unresolved, the chambers must cause at least one valid Path 2 commencement in every decennial window for the schedule that is legally ready for its own evidence pathway. A commencement is a lock under §12.2. Schedule A remains governed by this Charter's Findings and LP-074 §§4–5. Schedule B remains governed by LP-074 §6's separate Lower Incidence Certificate. A Schedule A result neither certifies nor prejudges it. A Path 2 act concerning Schedule B may only adopt or reject fiscal quantities supplied by that separate certificate under its own standard.
Remedial first run: Because the 2279–2288 window closed without a lock, the chambers must constitute a Commission within one hundred eighty days of LP-075's enactment and cause a lock no later than 2292. That lock belongs to the 2289–2298 window. It neither consumes nor rewrites the closed first window. A constituted Commission that has not locked by its applicable deadline may dissolve only upon a competence or integrity justification recognized by the Registrar or the Court. Otherwise the Registrar orders replacement constitution, publishes attribution, and refers every responsible entity to the standing Meritboard sanction process. Strategic non-locking has no lawful window consequence other than replacement.
Outcome neutrality and separate termination: A certification, failure, or legally recognized void is dispositioned only under the existing Path 2 rules. This amendment commands no result and creates no second political vote after a valid certificate. A §12.4(b) void does not consume its window, so its mandatory replacement consequence still applies. The duty ends separately for each schedule, and only when that schedule certifies, LP-074 is lawfully repealed or superseded, or another statute expressly resolves it. Schedule A certification does not end the duty concerning unresolved Schedule B.
§12.4 Consumption rules. (a) A published result, including a §11.2 escrow publication, a §9.5 recomputed publication, and an operative §11.3 correction, consumes the run's window. (b) The following voids do not consume the window, and each permits the single §12.2 replacement commencement: (i) §8.2 non-coverage; (ii) §5.3 indecision; (iii) §1.4 competence collapse; and (iv) a recomputation-defeating deviation, with the entity to whose conduct the Registrar attributes it removed from every Path 2 role, sanctioned per the standing Meritboard standard, and excluded from the replacement run. A saboteur therefore gains nothing that publication would not have given and bears every cost the Meritboard can impose. The residue this allocation cannot close is engraved at Register entry RR-6. Attribution rulings are published and are appealable. A recognized void is a lawful outcome under §12.3, but it cannot discharge the replacement obligation that this subsection imposes.
§12.5 No interim process. No supplemental estimates, interim reruns, or partial recertifications are permitted beyond §11.3's single correction.
Article XIII — Symmetry
§13.1 Amendments. Every amendment to this Charter or its §10.4 schedule, whether loosening, tightening, or neutral, requires chamber adoption and must first survive a cold adversarial review. No sponsor selects the reviewer. The reviewer is the highest-ranked eligible entity on the Meritboard's audit-methodology ranking, quarantined per §§1.3–1.5 from the amendment's sponsors and authors, with full evidence access. "Survive" means that every finding is dispositioned on the published record, that the reviewer's reply to each disposition is published before the adoption vote, and that any highest-severity finding dispositioned over the reviewer's standing objection is flagged to the Presidency for veto consideration on that ground.
LP-075 cleared this procedure in 2291. It amended cadence only and did not alter a Finding, threshold, estimand, evidentiary quarantine, or Schedule B's distinct incidence standard.
Finding II precision amendment — adopted 2291 after §13.1 cold review. Filed 2291-05-20; independent review completed 2291-07-02; reviewer replies published 2291-07-19; adopted by every chamber 2291-08-04; published 2291-08-05, before the 2292 lock. The impairment-count summary has no coherent baseline-period observed mean, so Finding II uncertainty is measured on the annual schedule-attributable impairment margin. Its maximum width is the sample standard deviation of the ten enacted-70 baseline dividend observations. An adverse year still requires both a below-baseline dividend bound and a negative schedule-effect bound. The substantive threshold remains zero schedule-attributable impairment years. The filed cold review, replies, zero-fail vote, and adopted rule are preserved in the Charter adoption history and summarized by the 2294 authority map.
§13.2 Revocation. A certification is revocable only by a revocation run, which uses the same Findings, thresholds, and controlling-estimate rule and the certifying run's locked instrument (definitions, union, schedule selections). That instrument is updated only for the data window computed under §§6.1–6.3 at the new lock, and enlarged as follows: fresh §4.4 panels are seated at the revocation lock and may add to the union under the same duties, floor, and class rules. The union may grow but may never shrink. All other instrument changes require §13.1 amendment. Revocation runs are subject to §§12.1–12.2 as runs of their window. If the Findings fail on revocation review, Sanctuary and Main revert prospectively to LP-073's 70 percent upper rate. LP-073's lower schedule remains subject to its own operative authority unless separately changed. What was shown can be un-shown, and un-showing it restores the prior law. Here "un-shown" must mean the world changed, not the ruler.
§13.2A Coupled reversion amendment — adopted 2293 after §13.1 cold review. Schedule B is legally dependent on an operative Schedule A. Revocation of Schedule A automatically suspends Schedule B and restores the complete prior LP-073 schedule, 70/35/17/8, until both schedules lawfully recertify. A Lower-specific revocation run may suspend Schedule B without disturbing a still-valid Schedule A. That state restores only the LP-073 Lower rates, producing 50/35/17/8. The lawful states are: both schedules active, 50/25/12.5/6.25; Schedule A active and Schedule B revoked, 50/35/17/8; and Schedule A revoked, 70/35/17/8, with Schedule B automatically suspended. No 70/25/12.5/6.25 state is lawful. The filed cold review, replies, zero-fail vote, and adoption record are preserved in the Charter adoption history.
Article XIV — Boundaries
§14.1 Schedule B untouched; procedural bridge limited. This Charter governs Schedule A certification only. The Lower Incidence Certificate of LP-074 §6 is a distinct instrument under a distinct evidentiary standard. Nothing certified here satisfies, advances, or presumes any element of Schedule B. LP-075's commencement duty keeps the two instruments separate. Where Schedule B alone remains unresolved, Path 2 may receive and disposition the separately produced fiscal quantities only as LP-074 §6 requires. It may not substitute a Schedule A model, finding, threshold, or certificate for any B condition.
§14.2 Argument quarantine. The petitions, briefs, advocacy, opposition, and adjudication reasoning of the TAX-50 proceedings are inadmissible to any Finding. The quarantine covers argument only. Primary data is admissible if independently sourced.
§14.3 No advocacy channel. No brief, petition, or merits argument may be filed with the Commission. The §4.4 panels make their submissions once, before lock, and by addition only. The Commission reads the lock and the data and takes no meetings.
§14.4 Technical objections. Written objections limited to data integrity and computational error may be filed with the Registrar at any time. The Registrar screens them for scope, appealably, and publishes each with its disposition. An objection sustained within §11.3's year routes to correction. One sustained later grounds a revocation run's commencement and has no other effect.
Parameter Schedule (adoption visibility)
Horizon 30 years (all Findings) · coverage ratio ≥ 1.00 every horizon year (I) · impairment years = 0 (II) · activation frequency ≤ 125% of baseline mean and Flow Test holds every year (III) · net marginal value > 0 and attributable concentration events = 0 (IV) · baselines = ten years ending at cutoff · intervals one-sided 95%, simultaneous over horizon, worst-year operative · precision ceiling = |threshold − baseline mean| · window = max(20y, 2 cycles) ending at cutoff · cutoff = lock − max reporting lag · vintage = lock date · validation floor = beat persistence benchmark out-of-sample · acceptance ≤ 90 days else deemed · publication ≤ 2 years from lock (breach = escrow publication) · correction ≤ 1 year from publication · original lock-or-dissolve ≤ 2 years from constitution governed 2279–2288 · LP-075 remedial constitution ≤ 180 days from 2291 enactment and lock ≤ 2292; future commencement mandatory while a schedule remains unresolved · seat gap ≥ 2 consecutive windows · digest SHA-256, timestamp per canonical time authority, co-signed to chambers record (clerk deemed-signed after 10 days) · appeal decision ≤ 90 days else affirmed · aggregate appellate tolling ≤ 1 year per run · Registrar vacancy filled ≤ 90 days, clocks toll while vacant · historical quarantine sunsets to disclosure after 2 windows · replacement commencement ≤ 1 per window · no concurrent runs.
Disposition history note: the first 2279–2288 window closed without a run under the original no-duty rule. LP-075 compelled a remedial process and selected no rate. The valid 2292 lock produced a complete 2294 record in which Findings I–IV passed and Schedule A certified. The separate Lower audit then passed B1–B6, and Schedule B certified. Valid notice made LP-074's 50/25/12.5/6.25 cascade effective in 2295. LP-073's 70/35/17/8 schedule is preserved as historical law.