Part XV, Implementation & Transition

A design no one can lawfully reach is a daydream. This Part is the path from today's system to the model, and it must itself satisfy §0.2: you cannot impose a consent-based system without consent. The transition is therefore lawful, consensual, piloted, reversible, and bounded against capture.

XV.1 The governing principle of transition

The means must match the ends. A system whose legitimacy rests on informed consent (§0.2) cannot be installed by decree, emergency, or trickery, that would poison it at the root. Every step is:

• Lawful, using only the powers of the existing constitutional order to reach the ratifying moment, and honest that ratification is itself a constitutional founding, not a pretence that one Parliament bound its successors (§XV.1.1).
• Consensual, confirmed by the people at the decisive moments.
• Deliberated, informed by Citizens' Assemblies, not driven by a single vote in the heat of a moment.
• Piloted, proven at small scale before national rollout.
• Reversible, every stage can be stopped or unwound if it fails.
• Bounded, the transition itself cannot be used to seize power.

XV.1.1 The sovereignty problem, stated honestly

Entrenchment, a Charter that later Parliaments cannot repeal by ordinary majority (§I.1, §I.9), is the whole point of the reform, and under the current constitution it is, strictly, impossible: Parliament is sovereign and cannot bind its successors, and a UK referendum is legally advisory. Pretending otherwise would be exactly the overclaim §0.6.5 forbids. So the model is honest about what ratification actually is:

• Ratification is a lawful founding discontinuity, a new rule of recognition. The two-referendum route (§XV.2) supplies the democratic authority; the enacting legislation, a new Rights & Constitutional Court sworn to the Charter, and the allegiance of every office-holder to the Charter rather than to the Crown-in-Parliament (§I.5.3) together establish a new basis on which UK law is recognised as valid. This is a break, consciously made and consented to, not a seamless continuation.
• The courts' acceptance of the new grundnorm is an explicit gate (Stage 7a below): the settlement takes effect only once the judiciary, on taking the new oath, recognises the Charter as supreme law. A founding the courts do not accept has not happened, and the process returns to deliberation.
• What secures entrenchment is the founding act plus the people, not a legal fiction. Legally, a future Parliament could purport to repeal the Charter; what stops it is that the Charter is now the recognised basis of every official's authority and every court's jurisdiction, and that the people are its guarantor of last resort (§XIV.4). This is how every codified constitution in a formerly-sovereign-parliament tradition actually binds, by a founding moment and its acceptance, not by the old Parliament's permission.

Why say this out loud: a reform whose central promise, an unrepealable core, rested on a doctrine the reform itself contradicts would be built on sand. Naming the discontinuity, and gating it on judicial recognition, is what makes the entrenchment real rather than asserted.

XV.2 The lawful pathway, the two-referendum model

Modelled on New Zealand's successful FPTP→MMP transition (two referenda, deliberation, pilots), strengthened:

Each gate is a genuine stop point: failing a gate halts or returns the process. No stage is skippable.

XV.2a Referendum 2 is disaggregated, and the founding bar is no lower than the amendment bar

Two legitimacy rules govern the ratifying vote, because a permanent settlement adopted by a bare, bundled majority would be less legitimate than a later single-clause amendment:

• Disaggregated, not omnibus. Referendum 2 is put as a small number of separately-voted blocks (e.g. rights & courts; the electoral & citizen layer; the expert and AI machinery; the fiscal & territorial settlement), so a voter can accept electoral reform without being forced to swallow, say, the AI Router, and no contested item rides to permanence on a popular one's coat-tails.
• Founding bar ≥ amendment bar. Any block that enters the super-entrenched core (§I.9.2) must be carried at no less than the threshold required to later change it, a supermajority (≥ 60%) on a high turnout floor, and, where the Union is affected, a majority in each nation. It must never be easier to install the permanent core than to later revise it. All turnout floors are stated numerically in the enacting legislation, never left blank.

XV.3 Sequencing, what to build first, and why

1. Electoral reform first (Part III). The mandate must be honest before anything is built on it (§0.5), so STV+ and verifiable elections come first. This is also BIG's founding work and the most evidence-ready component. (Honest caveat: changing the Commons voting system is itself a constitutional act requiring the same legitimacy machinery, so "first" means first-in-sequence within the settlement, not a step severable from it.)
2. Citizen layer & identity (Part II / VIII), the secure, private, inclusive foundation everything else needs, built and audited in parallel.
3. Integrity & transparency machinery (Part VI / VIII), stood up early, so the rest is watched from birth, never retrofitted.
4. Expert layer & decision machinery (Part IV / V), introduced once mandate, identity, and integrity exist to bound and watch it.
5. Institutional architecture (Part IX), sortition chamber, full bicameral legislature, constitutional court, ratified by Referendum 2.
6. The wider constitution (Parts X-XIV), the fiscal constitution, the territorial settlement, justice/security oversight, citizenship, and party/defence rules are codified into the Charter as part of the settlement, not bolted on later.
7. Crisis & resilience doctrine (Part VII), defined before it is ever needed.

The tech build (Part VIII) runs throughout: open-source, independently audited, reproducibly built, piloted before trusted.

XV.4 Pilots, de-risk before national rollout

Prove each component small before betting the country on it:

• STV+ in local elections (precedent exists: Scotland since 2007) → measure the Gallagher index improvement and public comprehension.
• Digital identity pilot in a region, test sybil-resistance, privacy, inclusion, and the offline path.
• Citizens' Assemblies at local level, test sortition deliberation and its interaction with elected bodies.
• Transparency ledger + procurement forensics for a single department, test verifiability and anti-corruption.
• Devolved-model pilot, a nation or region runs the fuller model (Parts X-XI) at its scale first.
• Outcomes ledger, publish the wellbeing composite before the reform, to establish a baseline to be judged against.

A pilot that fails its evaluation stops or redesigns that component, pilots are real gates, not theatre (§XV.6).

What cannot be piloted, stated honestly. The highest-stakes elements are irreducibly UK-wide and cannot be run in one region: the codified Charter and super-entrenched core, the head-of-state settlement, the national fiscal constitution, and replacing the Lords with the Sortition Chamber. For these the assurance is not a pilot but deliberation + Referendum 2 plus strengthened reversibility — the extraordinary route (§I.9.3) and manifest-error correction (§I.9.4) keep even the core revisable by a sustained supermajority. The design names which elements carry the largest irreducible bet rather than implying pilots de-risk everything (docs/COSTING.md §5).

XV.5 Migrating the existing institutions

Continuity, not chaotic rupture:

• House of Commons → Representative Assembly elected by STV+.
• House of Lords → Citizens' Sortition Chamber, replacing an unelected chamber with a lot-selected, representative one (a clean resolution of a centuries-old anomaly).
• Civil service → Expert Execution Layer, professionalised, scope-bounded, integrity-watched; building on existing competence rather than discarding it. Impartiality (Northcote–Trevelyan, the Civil Service Code) is constitutionalised; board delivery-accountability plus named owners replace individual ministerial responsibility (no accountability gap); and the minister↔Board interface is piloted before national rollout (§XV.4; full plan in docs/COSTING.md §4).
• Supreme Court → Rights & Constitutional Court with explicit Charter-enforcement power (the codified Charter gives it the written instrument it currently lacks).
• Existing watchdogs (NAO, Electoral Commission, ICO, OBR, etc.) → consolidated, strengthened Integrity & Safeguard Assembly + Independent Fiscal Authority with guaranteed independence and funding (Parts VI, X).
• Devolution → entrenched territorial settlement (Part XI), converting revocable devolution into a codified, consented division of powers, including an English tier.
• Security & intelligence oversight → the cleared oversight regime of Part XII; war powers moved from prerogative to the Assembly.
• Citizenship & immigration → the codified, rights-bound regime of Part XIII.
• The monarchy → reduced to a strictly ceremonial head of state, or replaced by an elected non-executive presidency (§IX.3, §XVIII D2), with zero governing power either way. The transition mechanics are named, not glossed: converting or abolishing the Crown engages (a) royal assent, which is sought as a matter of continuity and legitimacy but is not constitutive: under the founding discontinuity (§XV.1.1) the settlement's authority derives from Referendum 2 and judicial recognition (Stage 7a), not from the monarch's consent — so a withheld assent cannot veto the founding (were it constitutive, the new order would rest on the old sovereign's permission, the very fiction §XV.1.1 disclaims); (b) the Perth Agreement and the 14 other Commonwealth realms that share the Crown, whose own arrangements are unaffected (the UK decides only its own head of state); (c) the disestablishment knock-ons — the Supreme Governor role, the coronation, the Lords Spiritual (§I.5.2); and (d) the honours and residual prerogative functions, reassigned to the relevant constitutional office-holders. These are high-friction steps, named honestly as such (§XV.7), not a Referendum-2 tick-box.

XV.6 Reversibility and safeguards during transition

The transition is the most dangerous moment (old checks weakening, new ones not yet load-bearing), so it is bound hard:

• Every stage is reversible until Referendum 2; failed pilots stop their component.
• The existing constitutional protections remain in force until the new Charter is ratified, there is never a gap with no rules.
• The transition cannot be used to grab power: no actor gains unchecked authority during it; the Commission is integrity-vetted and time-limited; emergency powers cannot be used to drive the transition (§I.6.4).
• A defined resting state if Referendum 2 fails. Because electoral, identity, and integrity machinery is built before Ref 2 (§XV.3), a failed confirmatory vote must not leave a half-migrated limbo with weakened old checks and unratified new ones. The resting state is pre-legislated for each pre-Ref-2 component: which changes persist (e.g. STV+ for local elections, where already piloted and separately mandated), which roll back, and under whose authority — so "the process returns to deliberation" (§XV.2) is a defined state, not a vacuum.
• Independent monitoring of the transition itself, reporting publicly.
• Reversibility does not end at ratification. Phased rollout (Stage 8) keeps component-level rollback gates tied to the live outcomes score (§VI.7): a component that measurably worsens outcomes is paused and reviewed under §V.7, not pressed on with. And because the amendment routes (§I.9) and the public score are permanent, the people keep a lawful path to revise the settlement, including, through the extraordinary route (§I.9.3), the super-entrenched core itself. Nothing is placed beyond the reach of a sustained, cross-generational supermajority: "unwinding" the core is deliberately hard but never impossible (the near-eternity design, §I.9.2), which is exactly what makes "adopted by consent and stays only by consent" literally true rather than rhetorical.

XV.7 Cost, capability, and honesty

Stated plainly (Axiom 6; §0.6.5): this is expensive, slow, and hard. The honest, order-of-magnitude assessment (full annex: docs/COSTING.md) is a one-off transition of single-digit to low-tens of £ billions over a decade, and a *steady-state net additional recurring cost of a few hundred £ million a year — under 0.1% of the ~£1.2 trillion the state already spends — with headcount broadly flat, because the model reorganises the civil service and consolidates watchdogs rather than adding a new bureaucracy. Each added institution passes an explicit marginal test (§IX.1): the capture-resistance it adds must exceed the cost, attack-surface, and complexity it creates. Anyone claiming the reform is quick or free is wrong; equally, the cost of the status quo* — distorted mandates, low trust, short-termism, poor allocation of that £1.2 trillion — is large, recurring, and paid silently. The authoritative, independently peer-reviewed costing is produced by the Commission at Stage 2, before Referendum 1 (§XV.2).

XV.8 The international dimension, the UK as exemplar

• The design is published openly (the verifiability ethos applied to the project itself), so it can be scrutinised, improved, and adopted.
• Parallel adopters (Canada, Australia, and other reform-seeking democracies) can pilot components and share learning.
• Success in the UK is the proof-of-concept that makes the model exportable, the mission's end state (§0.1).

XV.9 Failure modes and safeguards

Part XV ends. Next: Part XVI, the full adversarial analysis: every named attack mapped to the safeguard that answers it.