Comparative Fiscal Capacity

A comparative analysis of what each live constitutional path — the current settlement, enhanced devolution and independence — would require Wales to build in fiscal capacity, and how strong the evidence is for each requirement. Working through eleven mechanisms, from territorial funding and borrowing to currency, welfare and payment systems, it argues the paths are not steps on a single ladder of more or less power but different problems of capacity — and recommends none of them. A companion evidence annex carries the full method, mechanism dossiers and source log.

This paper explains what each live constitutional path would require Wales to build, and how strong the evidence is for each requirement. It does not recommend a constitutional option. It does not argue that Wales should pursue or reject the current settlement, enhanced devolution or independence. Where it describes what a path requires or leaves unresolved, that is a statement about capacity and evidence, not a judgment about whether the path should be taken.

Findings at a glance

This paper compares what three constitutional paths would require Wales to build in fiscal capacity, and how well each requirement can be evidenced.

The current settlement gives Wales real but partial fiscal machinery. It administers some taxes, runs a development bank and manages devolved budgets, but the largest fiscal functions, including the main funding system, currency, the core of welfare and financial regulation, sit outside Welsh control. Its defining feature is dependency combined with partial capacity.

Enhanced devolution would change this unevenly. It would strengthen some areas, such as tax, borrowing flexibility and fiscal measurement, but only by building machinery around the powers; and it would leave others, such as currency and payment regulation, largely untouched unless the model is stretched into something else. Its defining feature is selective institutional build, not simply more devolution.

Independence would remove the reserved frame and make Wales responsible for the full fiscal architecture of a state. This does not dissolve the fiscal problem; it converts dependency into design. Most of the required institutions are well diagnosed but not yet specified in Welsh terms; the currency question is well mapped but unchosen; and debt, pensions and inherited liabilities could only be settled by negotiation.

The comparison’s central point is that the paths are not steps on a single ladder of “more” or “less” power. Each asks Wales to solve a different kind of capacity problem: what already exists, what could be built by expanding current institutions, what would need a new system, and what depends on a settlement with another government.

1. The question this paper asks

Debate about Wales’s constitutional future often runs straight to the verdict: which path is better. This paper asks a prior and more answerable question. Whatever path Wales takes, what would it actually have to build, and how well can each requirement be evidenced today?

That question matters because constitutional models are usually compared as destinations, when the more operational question is the machinery each one demands. A power on paper is not the same as the capacity to use it. Wales can hold a competence and still lack the institution, the data, the forecasting, or the delivery system that would make it work. Throughout this paper, then, two things are kept apart: whether a function sits within Welsh control, and whether Wales has the capacity to carry it.¹ Where those two diverge, the divergence is the finding.

The paper compares three live paths: the current settlement, enhanced devolution, and independence. For each, it works through eleven mechanisms of fiscal capacity, from territorial funding and borrowing to currency, welfare, payment systems and the productivity of the wider economy. The findings box above gives the argument in brief. A reader matrix later in the paper summarises the mechanism-by-mechanism comparison after the three paths have been explained. Each row is a mechanism; each column is a path; each cell is a plain-language reading of what that path would require for that mechanism, and, for independence, how strong the evidence is. A fourth model, a federal United Kingdom, was considered during the underlying assessment but is not treated as a live path here; a short note near the end explains why, and the full federal analysis is held in the evidence annex.

A word on what the readings mean. “Reserved” marks a function held at UK level. “Build-required” marks a capacity Wales would have to construct. “Sovereign build” marks a function that independence would place wholly in Welsh hands but that no existing Welsh institution yet performs. And the evidence readings matter as much as the capacity ones: to say a requirement is well diagnosed but its design absent is not to say it is impossible, only that the work of specifying it has not yet been done. The paper is an account of requirements and evidence. It is not, and does not become, an argument for a destination.

2. The current settlement: a partial state

The honest reading of the current settlement is neither that Wales is powerless over its own finances nor that it is autonomous in them. It is a partial state: real machinery in several places, bounded everywhere by functions that are held at UK level or shared. To see why, it helps to look at the actual machinery Wales holds, and the machinery it does not.²

Begin with the money that funds most of Welsh government. The largest single determinant of the Welsh budget, the block grant, is set at UK level through the Barnett formula. The settlement now includes a needs-based factor, derived from the Holtham Commission’s work, which holds long-term Welsh relative funding at 115 per cent of comparable English spending, with a transition factor of 105 per cent applying while Welsh funding sits above that floor. This is real protection, and it matters, but it is protection inside a system Wales does not control: Barnett modified by a needs factor, not a needs-based formula that Wales sets or administers.³

Borrowing and reserves tell the same story in numbers. Under the fiscal framework, Wales can borrow up to £1 billion in total for capital, within an annual limit of £150 million, and up to £200 million a year for resource, within an overall £500 million cap, to manage forecast error. It can hold up to £350 million in the Wales Reserve, drawing down £125 million of resource and £50 million of capital a year. These are genuine budget-management tools, and recent budgets have increased their real value, but every figure is a ceiling agreed with the UK Government, and the Reserve itself is held within the UK Government rather than by Wales.⁴

Tax is where Welsh capacity is most visible and most bounded at once. Wales has its own operating tax authority, the Welsh Revenue Authority, with a workforce of over a hundred, which administers the two fully devolved taxes, Land Transaction Tax and Landfill Disposals Tax. Welsh Rates of Income Tax are set by the Senedd but collected by HMRC, not by Wales. The result is real administrative capacity over a narrow base: the Welsh Revenue Authority proves Welsh tax administration works, but the taxes it runs are a small part of the revenue that funds Welsh public services. Devolved taxes provide about 21 per cent of the Welsh Government’s day-to-day funding; the rest comes through the block grant.⁵

The same shape, real capacity at the margin with the main function held elsewhere, recurs across the other mechanisms. In welfare, the major cash benefits that dominate social-security spending, the state pension, Universal Credit, disability benefits and the Winter Fuel Payment, are reserved and administered by the Department for Work and Pensions. What Wales runs is a layer of devolved assistance around that system: Council Tax Reduction, Free School Meals, the School Essentials Grant, the Education Maintenance Allowance, the Discretionary Assistance Fund, now being drawn together as the Welsh Benefits System. That is genuine delivery capacity, but it sits beside the reserved system, not over it.⁶

In public banking, Wales has the Development Bank of Wales, which lends, invests, guarantees, co-invests and de-risks for Welsh economic purposes, including instruments such as the Sustainable Agriculture Loan Scheme. What it is not, and cannot become without UK regulatory authorisation, is a deposit-taking bank: the development-finance form is within Welsh competence, but the deposit-taking form sits behind the reserved financial-services regime.⁷

In payments, the pattern is sharper still. Wales intervenes in how the public sector pays its suppliers, through procurement practice and Project Bank Accounts that protect payment flows in public delivery chains, but it holds no power over the payment system itself. Payment services are reserved under the financial-services regime, regulated through the Payment Services Regulations 2017, the Financial Conduct Authority and the Payment Systems Regulator. Welsh procurement practice addresses late or unfair payment; it does nothing about dependence on externally owned card and settlement networks.⁸

Debt and pensions complete the picture, and pensions in particular show how a reserved competence can still land as a real Welsh cost. Sovereign debt management is wholly reserved, held by HM Treasury and the Debt Management Office; Wales has no debt-management or gilt-issuance function. Pensions are formally reserved too, but for a narrow firefighters’ exception, with the public-service framework set by the Public Service Pensions Act 2013 and the schemes for NHS staff and teachers run as England-and-Wales schemes with centrally set rules. Yet the costs fall on Welsh bodies. The funded Local Government Pension Scheme is administered by Welsh pension funds, pooled through the Wales Pension Partnership and audited by Audit Wales, with employer and deficit contributions falling on Welsh local authorities. On the unfunded NHS and teachers’ schemes, UK-set increases in employer contributions pass straight through Welsh health and local-government budgets. Competence is reserved; the exposure is Welsh.⁹

The point of this section is not to judge whether the current settlement is adequate. It is to establish, in concrete terms, the starting position from which any change is measured. That position is a partial state: an operating tax authority over a narrow base, real but capped borrowing, a development bank that is not a deposit-taking bank, a welfare layer beside a reserved system, procurement-level payment practice without payment-system power, and reserved liabilities that nonetheless cost Welsh budgets. Every requirement the other two paths impose is a change to this baseline, not to a blank page.

3. Enhanced devolution: the build question

Enhanced devolution is most often pictured as a transfer: powers move from Westminster to Cardiff, and Wales does more. The evidence points to something different, and the difference is structural. Across the mechanisms, enhanced devolution reads less as a legal switch than as a build question. A power can be transferred in law and still not be usable until the institution behind it exists, and for several mechanisms the institution is the hard part.¹⁰

Take the fiscal core first, where the changes are real but bounded. A fuller needs-based settlement, moving beyond the present floor to a more explicit formula, is a technical model and a starting point for negotiation rather than something Wales could adopt unilaterally; it requires UK agreement and legislation. The same is true of wider borrowing flexibility, which the Holtham Commission supported in principle to lengthen planning horizons, but envisaged through UK-linked channels rather than open market borrowing. Further tax devolution is possible too, and Holtham supported it for accountability, but with a warning the paper should carry plainly: a power to vary a tax is not the same as the ability to raise revenue from it, because mobility, behavioural response and the integrated UK economy all bear on what a devolved tax actually yields. In each case the competence question is conditional on Westminster, and the capacity question, the forecasting, the measurement, the management of the resulting risk, is a build.¹¹

Welfare is where the build question is sharpest, and where a distinction has to be made that is easy to blur. “Devolving welfare” can mean two very different things. The first is integrating and strengthening the assistance Wales already runs: drawing Council Tax Reduction, Free School Meals, the assistance grants and the rest into a coherent Welsh Benefits System, improving take-up and access. That is available within the present settlement and is largely an organisational task. The second is devolving major social security benefits, the disability and carer benefits, or more, and that is a different proposition in kind, not degree.¹²

Scotland shows what the second actually involves. Under the Scotland Act 2016, eleven benefits were devolved, and Scotland chose to build its own delivery agency, Social Security Scotland, established in 2018, which by the end of 2023 employed over 3,800 staff and was paying out more than £5 billion a year. Even then the transfer was not clean: years into the programme, a large share of that spending was still being delivered by the Department for Work and Pensions under agency agreements while Scottish systems were stood up, and the new system had to manage continuous interaction with the reserved benefits that remained. The lesson is not that the task is impossible; Scotland did it. The lesson is what the task is: a multi-year institutional build of payment systems, claimant data, assessment, redetermination and appeal routes, fraud and error control, and demand forecasting, carrying real fiscal risk, undertaken before public-facing responsibility can safely move. For Wales, the Independent Commission’s evidence points the same way, judging substantial benefit devolution feasible only alongside a substantial increase in tax and borrowing powers.¹³

The same build logic runs through the other mechanisms, with the added point that enhanced devolution leaves some of them essentially untouched. A Welsh Green Bank in its development-finance form is buildable now, within Welsh competence, using the Development Bank of Wales as host or precedent, but the deposit-taking form stays behind the UK regulatory boundary unless financial services competence is transferred, which would take the model beyond enhanced devolution. Payment resilience is similar: Wales could build a more protected public-sector payment mechanism for its own government, public bodies and grant flows, but it could not regulate payment services or own the rails without the same competence transfer. And currency and monetary policy do not move at all under any recognisable form of enhanced devolution.¹⁴

So the enhanced-devolution finding is twofold. It is a capacity-and-design route rather than a competence switch: most of what it offers has to be built, and the building, not the legislating, is the demanding part.

It is also uneven. It strengthens some mechanisms substantially, leaves others, currency, the core of payments, the deposit-taking boundary, untouched unless the model is stretched into something else, and in the case of welfare requires Wales to construct a delivery institution before the responsibility can be carried safely. That unevenness is not a flaw in the model; it is what the evidence shows the model to be.

4. Independence: the change of form

Independence is where the comparison is most often shortcut, in both directions. It is treated either as the answer to the fiscal problem or as its refutation. The evidence supports neither shortcut. What independence does, across the fiscal settlement, is change the form of the problem. It removes the reserved dependency that defines much of the current settlement and replaces it with a sovereign build requirement.¹⁵

Start with funding and revenue, because that is where the change is most concrete. Today about 21 per cent of the Welsh Government’s day-to-day funding comes from devolved taxes and the rest, the large majority, arrives through the block grant; the Institute for Fiscal Studies put day-to-day resource funding at around £23.1 billion in 2025-26, roughly four-fifths of it from the UK Government. Independence would remove the block grant entirely and require Wales to fund public services from its own revenue, borrowing and fiscal management. That is not a neutral swap, because the Welsh fiscal starting position is not neutral. The PESA Wales estimate puts Wales’s net fiscal deficit at over £21.5 billion in 2022-23. That figure has to be read for what it is: an estimate of the current notional balance under existing UK spending allocations, not a forecast of an independent Wales’s deficit. It attributes to Wales a population share of UK-wide spending that Wales neither controls nor would automatically reproduce, including defence and a share of UK national-debt interest. Defence is the clearest case. Ifan and Siôn note that sustaining Wales’s apportioned per-capita share of UK defence spending would imply defence spending of around 2.6 per cent of Welsh GDP, which illustrates how an allocated UK cost can diverge from the budget choice a small independent state would actually make. Debt interest is similar: the apportioned cost reflects a share of existing UK-wide debt service, while the debt stock and interest burden of an independent Wales would depend on the separation settlement. What the figure does show is the structural gap on the revenue side, which is the more durable finding: successive analyses, from Wales Fiscal Analysis in 2020 through to more recent official and IFS work, link Wales’s persistently low revenue per head to long-standing economic underperformance rather than to a passing shock. The numbers are dated differently and should be read as a consistent picture rather than a single headline, but the direction is stable: the revenue base currently associated with Wales falls well short of current spending levels. Independence does not dissolve that gap; it changes who decides how to manage it, and which of the apportioned costs to carry at all.¹⁶

Borrowing follows directly, and it is more than a higher cap. At present Wales borrows within fixed limits set in the fiscal framework. A sovereign Wales would issue its own debt, and that means building the apparatus that makes sovereign borrowing possible and affordable: a debt-management function, access to capital markets, a credit standing established from a standing start, and the reporting and independent fiscal scrutiny that lenders expect before they lend cheaply. Borrowing cost is not a given; it is earned through credibility, and credibility rests on the fiscal measurement and institutional machinery described below. The conversion here is from a capped facility inside someone else’s framework to a sovereign function that has to be constructed and then continuously demonstrated.¹⁷

Tax and fiscal measurement are the next movement, and they belong together. Wales already has a working tax authority: the Welsh Revenue Authority collects the two devolved taxes and demonstrates that Welsh tax administration functions. But a sovereign tax system would have to absorb the whole of what HMRC currently does for Wales, income tax in full, a consumption tax, corporation tax, the rest, at a scale far beyond the authority’s present remit. Beneath that sits a mechanism that is easy to overlook and impossible to do without: fiscal measurement. National accounts, fiscal statistics, revenue forecasting and independent scrutiny are not back-office optional extras; they are the instruments that make funding, borrowing and taxation credible at all. A sovereign state that cannot measure and forecast its own finances to a recognised standard cannot borrow cheaply or budget honestly. This is why fiscal measurement is a cell almost every other fiscal cell depends on, and why independence makes it a foundational build rather than an afterthought.¹⁸

Welfare and state capacity are the fourth movement, and Section 3 has already shown the shape of the welfare build. At independence it is larger again. Wales would hold the whole social-security system, not a layer of assistance beside it: the state pension and the major working-age and disability benefits, the largest single block of public spending, with all the payment systems, claimant records, assessment, redetermination and appeal routes, fraud and error control and demand-led volatility that go with them. Scotland’s experience of building delivery for a subset of benefits indicates the order of the task; sovereign welfare is a larger version of it. And welfare is only one of the functions a sovereign state must run at once. This is the point of the state-capacity mechanism, which is best understood not as one more item on the list but as the aggregate of all of them: the civil service, the machinery of government, the transition planning and institutional design that every other conversion in this section draws on. Independence is not eleven separate transfers that could be timed independently; it requires the connective capacity that lets them function together.¹⁹

The fifth movement covers mechanisms that each change in a distinct way, and the distinctions matter. Currency is the mapped-but-unchosen case: the options for a newly independent state, continued use of sterling, a formal arrangement, a new currency with its own central bank, are extensively analysed in international evidence and in the Scottish debate, so what is missing is not knowledge of the options but a Welsh choice among them and a transition design.²⁰

Payments and public banking share a different feature: in each, legal competence does not by itself supply the operating infrastructure. Payments are the build-or-join case: sovereignty would give Wales the competence to decide whether to build its own payment infrastructure or join existing systems, but competence alone does not supply rails, and dependence on externally owned card and settlement networks does not end simply because the legal power has moved. Public banking is the deposit-taking-unlocked-but-not-built case: sovereignty removes the UK regulatory ceiling that currently keeps the Development Bank of Wales to development finance, making a deposit-taking public bank legally possible, but the institution itself, with its capital, prudential standing and governance, would still have to be built.²¹

Debt, pensions and inherited liabilities are the last movement, and they are different in kind from everything above. The other conversions are, in essence, design problems: the requirement is known, and what is missing is the specified Welsh institution. This one is not a design gap but a contested settlement. A share of UK debt, the apportionment of accrued public-service pension rights, the division of inherited and contingent liabilities, none of these has a single correct answer that Wales could simply work out and build. They depend on the basis of apportionment chosen, population, output, historic contribution, and on what is agreed in negotiation with the continuing UK state. The honest finding is therefore not a number but a process: a sovereign Wales would require a negotiated liabilities settlement, and the terms of that settlement, not a Welsh design decision, would determine the fiscal starting position. That is why this mechanism reads as genuinely contested rather than merely unbuilt.²²

Taken together, these movements give the independence finding its real shape. Independence removes the dependence that defines the current settlement and replaces it with a requirement to build sovereign fiscal architecture: most of it well diagnosed but not yet designed, the currency question well mapped but unchosen, the liabilities question resolvable only by negotiation. Whether that exchange is worth making is precisely the question this paper does not answer. What it can do, and what this section has tried to do, is set out what the exchange consists of.²³

The comparison by mechanism

Mechanism	Current settlement	Enhanced devolution	Independence
Barnett and territorial funding	Reserved dependency	Build-required, by agreement	Sovereign build (diagnosis evidenced, design absent)
Borrowing and budget management	Reserved, bounded tools	Wider but still bounded	Sovereign build (diagnosis evidenced, design absent)
Devolved taxation and tax administration	Partial Welsh capacity	Expanded partial capacity	Sovereign build (operational base, design absent)
Fiscal measurement and instrumentation	Partial, UK-dependent	Build-required	Sovereign build, dependency cell (design absent)
Currency and monetary policy	Reserved, not held	Reserved, not held	Sovereign build (options mapped, choice unmade)
Debt, pensions and inherited liabilities	Reserved, with Welsh exposure	Reserved, with Welsh exposure	Contested settlement (result varies by basis)
Social security and welfare	Reserved, with partial Welsh delivery	Design-dependent build	Sovereign build (base evidenced, design absent)
State capacity and transition capacity	Real but bounded	Design-dependent build	Sovereign build, dependency cell (design absent)
Payment resilience and payment rails	Reserved, with partial Welsh practice	Build-required, bounded by reserved rails	Sovereign build, build-or-join (design absent)
Green Bank and public finance vehicles	Welsh, development-finance only	Build-required within same ceiling	Sovereign build, deposit-taking unlocked (design absent)
Productivity and economic structure	Shared, partial capacity	Design-dependent build	Sovereign build (diagnosis evidenced, not convergence)

Plain-language summary. The full technical coding of competence, capacity and evidence strength is in the Evidence Annex, Annex A. The federal model was considered but archived; see the note below and Appendix F.

5. What the comparison shows

Having walked through the machinery, the comparison can be stated plainly. The three paths are not steps on a ladder from less power to more power. They are different relationships between competence, capacity and risk.²⁴

The current settlement combines real machinery with dependency: Wales holds genuine capacity in places, but the largest fiscal functions, and the risks attached to them, sit elsewhere. Enhanced devolution does not simply move Wales up a scale; it shifts the balance unevenly, handing Wales more to build around powers it already partly holds, while leaving the reserved core, currency, the payment system, the deposit-taking boundary, in place unless the model is stretched into something else. Independence does not climb higher still; it changes the relationship altogether, taking the reserved frame away and converting dependency into a requirement to design and carry the whole architecture, including the risk, itself.

Seen this way, the recurring question across all three paths is not how much power Wales has but what each path asks of Welsh capacity. For any given function, four answers are possible: the capacity already exists; it could be built by expanding an institution Wales already runs; it would need a wholly new system; or it depends on a settlement reached with another government. Most of what the paper has examined falls into the middle two, capacity that exists in part and must be expanded, or capacity that would have to be built new, which is why the dominant finding is neither powerlessness nor readiness but construction. A few functions sit at the edges: currency, where the options are well understood but unchosen, and inherited liabilities, where the answer is not Wales’s alone to design but must be negotiated.

That is also why the strength of the evidence matters as much as the direction of it. To say that a requirement is well diagnosed but not yet designed is a statement about how far the work has been taken, not about whether it can be done. The paper has tried to keep those two things distinct throughout: what each path would require, and how much is currently known about meeting the requirement. Both are needed to judge a path well, and neither, on its own, decides anything.

The federal model is treated separately because it was considered in the underlying assessment but is not one of the live paths compared in the reader-facing paper.

6. The federal model

Federalism was considered during the underlying assessment but is not treated as a live publication spine in this paper. The reason is not that federalism is analytically irrelevant, but that it has low Welsh-specific evidence, high institutional distance from the present UK settlement, and low present political plausibility as a practical route. The federal column is therefore retained in the evidence annex as archived comparator material, rather than placed beside the current settlement, enhanced devolution and independence as a live path in the reader comparison.²⁵

7. What this paper does and does not settle

This paper has set out what each live constitutional path would require Wales to build in fiscal capacity, and how strong the evidence is for each requirement. The recurring lesson is that competence and capacity are not the same thing, and that the gap between them is where the real work of any constitutional change would sit.

The three paths solve different problems. The current settlement leaves Wales with partial capacity inside a reserved frame. Enhanced devolution would require selective institutional building around powers Wales holds or could receive. Independence would convert dependency into a comprehensive design task, most of it well diagnosed but not yet specified, one part well mapped but unchosen, one part resolvable only by negotiation.

None of that tells Wales which path to take. The strength or weakness of a requirement is a statement about evidence and capacity, not a recommendation. A path that demands a large build is not for that reason a worse path, any more than a path that demands little is for that reason a better one; the costs, benefits and values of any constitutional choice turn on questions this paper does not address. What the paper offers is the ground beneath that choice: a clear account of what each path asks Wales to construct, and how much is known about each requirement today.

The evidence behind the findings is set out in full in the companion evidence annex, which carries the method, the mechanism-by-mechanism dossiers, the complete source log, and the archived federal analysis.²⁶

Evidence Annex

Companion document to the policy paper Comparative Fiscal Capacity.

Status: Evidence base for the companion policy paper.

Purpose. This annex sets out the full evidence base behind the companion policy paper. It contains the method and coding framework, the mechanism dossiers for all eleven fiscal capacity mechanisms across the current settlement, enhanced devolution and independence, the cross-cutting synthesis, the complete source log, and the archived federal column. Where the policy paper states a finding, the corresponding evidence can be traced here by mechanism identifier, for example M04-IND, and source identifier, for example S066.

Non-recommendation boundary. This annex, like the policy paper it supports, explains what the evidence shows about fiscal capacity, dependency and build requirements across constitutional models. It does not recommend a constitutional option, and nothing in it should be read as advocating enhanced devolution, federalism, independence, or the current settlement.

Provenance. This annex was assembled from verified mechanism modules and a consistency-checked source log. The federal column is retained as an archive appendix rather than part of the publication spine. The cross-cutting synthesis was reviewed against the coded cells and source references before being included. The annex preserves the evidence base behind the companion policy paper; it is not itself a recommendation for any constitutional option.

How to read this annex. Annex A sets out the method: the constitutional models compared, the competence and capacity codes, and the evidence bands, with the full technical matrix. Annex B is the heart of the annex: eleven mechanism dossiers, each gathering a mechanism’s variants across the current settlement, enhanced devolution and independence, with a pointer to the archived federal variant. Annex C is the cross-cutting synthesis, which reads the pattern across mechanisms and models. Annex D is the source log. Appendix F holds the archived federal column in full.

Annex A — Method and matrix

A.1 Column codes

Code	Column
B	Baseline: repaired or protected current devolution
ED	Enhanced devolution
FED	Wales in a federal United Kingdom (archived; see Appendix F)
IND	Independence

The baseline corresponds to what the parent economics note called the current settlement, refined to match the Commission’s protected-devolution structure. The unrepaired status quo is not treated as a fourth neutral option.

A.2 Status codes

Competence

Code	Meaning
Welsh	Within Welsh competence
UK	Reserved or UK-controlled
Shared	Shared, negotiated, or jointly governed
Sovereign	Full competence only under independence
Design-dependent	Depends on federal or constitutional design
Uncoded	Not yet coded

Capacity

Code	Meaning
Present	Existing capacity or machinery
Limited	Some capacity exists but is bounded
Partial	Capacity exists for part of the function only
Partial-held	Welsh institutions hold part of the function or adjacent delivery capacity, while the main function remains held elsewhere
Absent	No current Welsh capacity or host identified
Not-held	The function exists within the wider UK state but is not held by Welsh institutions under the model being coded
Build-required	Capacity would have to be built
Build-limited	Some additional Welsh or shared-rule capacity would be required, but the build is partial or supportive rather than a full sovereign institutional function
Design-dependent	Depends on institutional design
Uncoded	Not yet coded

Evidence strength

Code	Meaning
High	Primary source, official document, legislation, fiscal framework, Commission report, OBR, Welsh Government, UK Government, or statistical authority
Medium	Expert report, academic source, think tank analysis, or well-evidenced policy paper with transparent assumptions
Low	Conceptual material, sparse literature, advocacy source, secondary commentary, or inference from adjacent models
Contested	Evidence exists at reasonable strength but produces different results depending on assumptions, definitions, transition model, or constitutional design
Absent	No usable source yet identified

Evidence-band precedents established in the independence column (apply cell by cell, never by analogy):

Medium: Welsh diagnosis directly evidenced but sovereign design unsettled (M01-IND to M04-IND, M07-IND to M11-IND).
High: option space and institutional requirements extensively mapped even though the Welsh choice is unmade (M05-IND).
Contested: serious evidence whose result varies by allocation basis or transition model (M06-IND).

A.3 Mechanism index

ID	Mechanism
M01	Barnett and territorial funding
M02	Borrowing and budget management
M03	Devolved taxation and tax administration
M04	Fiscal measurement and instrumentation
M05	Currency and monetary policy
M06	Debt, pensions, and inherited liabilities
M07	Social security and welfare
M08	State capacity and transition capacity
M09	Payment resilience and payment rails
M10	Green Bank and public finance vehicles
M11	Productivity and economic structure

A.4 Compact index matrix

For navigation only. Full coding sits in the mechanism dossiers in Annex B. The FED column is retained here for completeness of the matrix but its dossiers are archived in Appendix F and are not part of the publication spine.

Mechanism	B	ED	FED (archived)	IND
M01 Barnett and territorial funding	UK / Limited / High / M01-B	UK / Build-required / High / M01-ED	Design-dependent / Build-required / Low / M01-FED	Sovereign / Build-required / Medium / M01-IND
M02 Borrowing and budget management	UK / Limited / High / M02-B	UK / Limited-to-expanded / High / M02-ED	Design-dependent / Build-required / Low / M02-FED	Sovereign / Build-required / Medium / M02-IND
M03 Devolved taxation and tax administration	Shared / Partial / High / M03-B	Shared / Expanded-partial / High / M03-ED	Design-dependent / Build-required / Low / M03-FED	Sovereign / Build-required / Medium / M03-IND
M04 Fiscal measurement and instrumentation	Shared / Partial / High / M04-B	Shared / Build-required / Medium / M04-ED	Design-dependent / Build-required / Low / M04-FED	Sovereign / Build-required / Medium / M04-IND
M05 Currency and monetary policy	UK / Not-held / High / M05-B	UK / Not-held / High / M05-ED	Design-dependent / Build-limited / Low / M05-FED	Sovereign / Build-required / High / M05-IND
M06 Debt, pensions, and inherited liabilities	UK / Not-held / High / M06-B	UK / Not-held / High / M06-ED	Design-dependent / Build-required / Low / M06-FED	Sovereign / Build-required / Contested / M06-IND
M07 Social security and welfare	UK / Partial-held / High / M07-B	Design-dependent / Build-required / High / M07-ED	Design-dependent / Build-required / Low / M07-FED	Sovereign / Build-required / Medium / M07-IND
M08 State capacity and transition capacity	Welsh / Partial-held / High / M08-B	Design-dependent / Build-required / High / M08-ED	Design-dependent / Build-required / Low / M08-FED	Sovereign / Build-required / Medium / M08-IND
M09 Payment resilience and payment rails	UK / Partial-held / High / M09-B	UK / Build-required / High / M09-ED	Design-dependent / Build-required / Low / M09-FED	Sovereign / Build-required / Medium / M09-IND
M10 Green Bank and public finance vehicles	Welsh / Partial-held / High / M10-B	Welsh / Build-required / High / M10-ED	Design-dependent / Build-required / Low / M10-FED	Sovereign / Build-required / Medium / M10-IND
M11 Productivity and economic structure	Shared / Partial-held / High / M11-B	Design-dependent / Build-required / Medium / M11-ED	Design-dependent / Build-required / Low / M11-FED	Sovereign / Build-required / Medium / M11-IND

IND column summary: M01 Medium, M02 Medium, M03 Medium, M04 Medium, M05 High, M06 Contested, M07 Medium, M08 Medium, M09 Medium, M10 Medium, M11 Medium.

A.5 Use rule

A cell should not be treated as available merely because competence exists. It must pass both a competence test and a capacity test. Where competence and capacity diverge, the divergence is the finding, not a caveat.

A cell should not be treated as weak merely because the literature is sparse. Sparse evidence should be coded as sparse evidence.

A cell should not be treated as strong merely because the literature is dense. Dense literature should be coded for contestation, assumption dependence, and source type.

Note on cross-column mechanism notes: the whole-mechanism interpretation notes for M05 through M11 are placed at the head of their respective mechanism dossiers in Annex B, where they introduce the mechanism before its model variants. Mechanisms M01 through M04 have no whole-mechanism note; their model-specific interpretation, where it exists, is federal-specific and is held in Appendix F.

Annex B — Mechanism dossiers

Each dossier gathers one mechanism across the current settlement (baseline), enhanced devolution and independence, with a pointer to the archived federal variant in Appendix F. Mechanisms M05 through M11 open with a whole-mechanism interpretation note; M01 through M04 have no whole-mechanism note (their model-specific interpretation, where it exists, is federal-specific and is held in Appendix F). Cell identifiers (for example M04-IND) and source identifiers (for example S066) are the stable citation anchors used by the companion policy paper.

M01 — Barnett and territorial funding

M01-B. Baseline: repaired or protected current devolution

Competence: UK Government retains control of the block grant and Barnett mechanism.

Capacity: The repaired baseline includes a Holtham-derived needs-based factor in the Barnett calculation. The long-term needs-based factor is 115%, with the transition factor set at 105% while Welsh relative funding remains above the agreed floor.

Evidence strength: High.

Sources: S001, S002.

Caveat: This is not full replacement of Barnett by a needs-based formula. It is Barnett modified by a needs-based factor, so the Welsh Government receives floor protection but does not control the territorial funding system.

M01-ED. Enhanced devolution

Competence: Requires UK agreement and legislation or intergovernmental fiscal agreement.

Capacity: Full enhanced-devolution reform would move beyond a floor toward a more explicit needs-based funding settlement.

Evidence strength: High for Holtham diagnosis and proposed direction; Medium for current political availability.

Sources: S001, S002.

Caveat: Holtham’s needs-based formula is a technical model and starting point for negotiation, not an automatically available settlement.

M01-IND. Independence

Competence: Sovereign. An independent Wales would hold full competence over its own revenue system, public expenditure framework and fiscal policy. The block grant, Barnett formula and UK-controlled needs-based factor would cease to be the main funding mechanism.

Capacity: Build-required. Wales would need to replace the block grant with sovereign fiscal capacity: a functioning tax system generating sufficient revenue, expenditure planning matched to Welsh needs, fiscal forecasting, borrowing strategy and credibility, revenue administration and collection, transition finance and any negotiated settlement terms. The Holtham diagnosis shows that Welsh expenditure needs exceed Welsh revenue capacity relative to the UK average, which means the replacement is not fiscally neutral.

Evidence strength: Medium. The Welsh fiscal baseline is well evidenced through official and academic sources: revenue composition, expenditure levels, net fiscal balance, block-grant dependence and relative need. IFS estimates that Welsh Government day-to-day resource funding in 2025-26 was forecast at £23.1bn, with around 79% provided by the UK Government, largely through the block grant, and devolved taxes making up the remaining 21%. The diagnosis is therefore strong. But the sovereign replacement mechanism, how an independent Wales would close or manage the gap between revenue capacity and expenditure needs, is not settled by the source base. The evidence reaches Medium because Welsh-specific independence fiscal analysis exists, unlike the federal column where evidence rests on comparative inference, but it does not reach High because no source provides an agreed or operational Welsh independence fiscal settlement.

Sources: diagnosis S001 (Holtham, relative need and fiscal-capacity asymmetry only, not independence design), S064 (Welsh fiscal deficit and independence-facing analysis), S065 (Welsh fiscal baseline and sustainability), S066 (current official revenue, expenditure and net fiscal balance), S067 (block-grant dependence and funding composition); baseline contrast S002 (fiscal framework showing the UK territorial-funding machinery independence would replace).

Caveat: Independence removes the Barnett dependence but does not remove the funding problem. It converts it into a sovereign fiscal-capacity problem: revenue base, tax administration, expenditure need, borrowing credibility, transition finance and any negotiated settlement. The diagnosis is stronger than the remedy. Welsh fiscal position, block-grant dependence and relative need are well evidenced; the sovereign replacement design is not settled by the source base. Holtham can be used for diagnosis of Welsh relative need and fiscal-capacity asymmetry, not for the independence design: the diagnosis travels, the remedy does not. On the deficit figure specifically: the PESA Wales net fiscal balance (S066) is relevant to the current fiscal starting position, but its spending side includes apportioned UK-wide costs that a sovereign Wales would not automatically reproduce, notably defence and a share of UK national-debt interest (S064). The figure should therefore be read as a notional current-allocation estimate, not as a forecast of an independent Wales’s deficit; the durable finding is the structural revenue-side gap, not the headline deficit total.

Federal variant: see Appendix F, M01-FED.

M02 — Borrowing and budget management

M02-B. Baseline: repaired or protected current devolution

Competence: Borrowing and reserve powers remain set within the UK fiscal framework.

Capacity: The 2016 framework increased the overall capital borrowing cap to £1 billion and the annual capital borrowing limit to £150 million. Resource borrowing remains available up to £200 million annually, within an overall £500 million cap, for forecast-error management, with repayment within four years. The Wales Reserve can hold up to £350 million, with no annual payment-in limit under the 2016 framework and annual drawdown limits of £125 million resource and £50 million capital.

Evidence strength: High.

Sources: S002.

Caveat: These are real budget-management tools but remain bounded. The Wales Reserve is held within the UK Government, and additional flexibility remains subject to UK consent or later agreement. The 2025 Budget update increases the real value of these tools but does not remove the competence boundary.

M02-ED. Enhanced devolution

Competence: Further borrowing flexibility would require UK agreement.

Capacity: Holtham supports limited capital borrowing powers to improve planning horizons and align capital expenditure with Welsh priorities.

Evidence strength: High for principle; Medium for scope.

Sources: S001, S002.

Caveat: Holtham envisages borrowing through UK-linked channels, not unconstrained market borrowing.

M02-IND. Independence

Competence: Sovereign. An independent Wales would hold full sovereign borrowing competence, including bond issuance, debt management, fiscal-rule design, cash management and reserve policy. The bounded devolved borrowing caps and Wales Reserve would be replaced by sovereign borrowing responsibility.

Capacity: Build-required. Wales would need to build sovereign debt-management capacity from a base that currently holds only bounded devolved tools. The required machinery includes a debt management office or equivalent function, bond issuance and investor-relations capacity, a medium-term debt-management strategy, fiscal rules or institutional guardrails with independent assessment, cash-management and liquidity capacity, risk management including stress testing and contingency planning, credit-rating establishment and maintenance, and government securities market access and, if pursued, domestic market development. The IMF and World Bank guidelines note that building capacity in sovereign debt management can take several years and that country needs are shaped by capital-market constraints, exchange-rate regime, macroeconomic and fiscal policy quality, and institutional capacity. Scotland’s independence proposals envisaged a new Debt Management Office, fiscal rules assessed by the Scottish Fiscal Commission, and a responsibility to establish creditworthiness and minimise borrowing costs.

Evidence strength: Medium. The Welsh fiscal position that would condition borrowing need and credibility is well evidenced through official and academic sources. The institutional requirements for sovereign debt management are well evidenced internationally through the IMF and World Bank framework. Scotland supplies the closest UK sub-state comparator for the transition from bounded devolved borrowing to sovereign borrowing. But no Welsh-specific sovereign borrowing analysis exists: no Welsh borrowing-cost estimate, no Welsh credit-rating assessment, no Welsh debt-management institutional design, and no Welsh fiscal-rule proposal for a sovereign context. The evidence reaches Medium because Welsh fiscal data and international institutional frameworks are available, but it does not reach High because the Welsh-specific sovereign borrowing design is absent from the source base.

Sources: fiscal position S064, S065, S066, S067; baseline contrast S001 (Holtham, diagnosis only), S002 (fiscal framework showing the bounded tools independence would replace); sovereign design S068 (IMF and World Bank debt-management guidelines), S069 (Scottish independence borrowing comparator, not Welsh evidence).

Caveat: Independence converts bounded devolved borrowing and reserve tools into a requirement for sovereign debt-management, fiscal-rule and reserve-capacity machinery. The transition is not an expansion of the existing caps; it is a qualitative change in institutional form. Borrowing credibility and cost are conditional on currency choice (M05-IND) and inherited debt allocation (M06-IND), which are coded separately and should not be pre-resolved here. No Welsh-specific sovereign borrowing framework, credit assessment or borrowing-cost analysis exists in the source base. The Scottish comparator shows what a sub-state independence borrowing design looks like in proposal form but should not be treated as evidence that Wales would or should adopt the same design. Holtham’s borrowing proposals were designed for devolution, not sovereignty, so they do not supply the independence borrowing design. But Holtham remains relevant as a baseline diagnostic source for the limited borrowing and budget-management machinery independence would replace.

Federal variant: see Appendix F, M02-FED.

M03 — Devolved taxation and tax administration

M03-B. Baseline: repaired or protected current devolution

Competence: The repaired baseline includes devolved land transaction tax, landfill disposals tax, and Welsh rates of income tax.

Capacity: Wales has tax-administration capacity for fully devolved taxes and some rate-setting accountability through Welsh rates of income tax. The block grant adjustment model connects devolved tax revenues to changes in the Welsh block grant.

Evidence strength: High.

Sources: S001, S002.

Caveat: The repaired baseline increases accountability and measurement capacity but does not create full tax autonomy. Key income tax architecture, macroeconomic policy, and the wider UK tax base remain outside Welsh control.

M03-ED. Enhanced devolution

Competence: Further tax devolution is possible but conditional on UK agreement and administrative feasibility.

Capacity: Holtham supports tax devolution for accountability, especially limited income tax variation, but recognises constraints from mobility, behavioural effects, and the UK economic union.

Evidence strength: High.

Sources: S001, S002.

Caveat: Rates, base, administration, and behavioural response must be separated. A power to vary tax is not the same as an unconstrained ability to raise revenue.

M03-IND. Independence

Competence: Sovereign. An independent Wales would hold full tax competence over all tax bases, rates, collection, administration, compliance, enforcement and international tax obligations. The current devolved taxes (Land Transaction Tax, Landfill Disposals Tax) and the partially devolved Welsh Rates of Income Tax would become part of a sovereign tax system rather than a devolved component of the UK tax framework.

Capacity: Build-required. Wales has existing tax-administration machinery through the Welsh Revenue Authority, which manages Land Transaction Tax and Landfill Disposals Tax with a workforce of more than 100 staff. Welsh Rates of Income Tax are set by the Senedd but collected and administered by HMRC. This gives Wales real but narrow tax capacity: the WRA proves that Welsh tax administration works, but its scope covers only a small part of the tax system a sovereign Wales would need. A sovereign tax system would require collection, administration and compliance capacity across the full range of tax heads: income tax (full base and collection, replacing HMRC’s role), a consumption tax (VAT equivalent or replacement), corporation tax, customs and excise if applicable, social contributions if any, anti-avoidance, evasion detection, compliance and enforcement, appeals and dispute resolution, taxpayer data systems, revenue forecasting and behavioural modelling, and international tax obligations including treaty networks. The transition from WRA-scale administration to sovereign-scale revenue collection is the core capacity gap.

Evidence strength: Medium. The existing Welsh tax machinery is directly evidenced: the WRA annual report shows a functioning devolved tax authority, and the WRIT annual report shows Welsh rate-setting with HMRC collection. The major tax bases currently administered at UK level are well documented through HMRC and UK fiscal data. The fiscal-position data (S064 to S067) shows the revenue composition and the scale of the replacement problem. But the sovereign tax-system design, including the institutional form of HMRC replacement, transition planning, compliance capacity across all major tax heads, and revenue-risk management at sovereign scale, is not settled by the Welsh source base. The evidence reaches Medium because Welsh tax-administration evidence is stronger here than for any previous IND cell (the WRA is operational machinery, not a design proposal), but it does not reach High because no source provides a Welsh sovereign tax-system design.

Sources: existing machinery S070 (WRA annual report), S071 (WRIT and HMRC collection); fiscal position and revenue composition S064, S065, S066, S067; baseline contrast S001 (Holtham, accountability diagnosis only), S002 (fiscal framework and devolved-tax arrangements); sovereign design comparator S072 (Scottish comparator on tax-system design consequences of independence, not Welsh evidence).

Caveat: Independence converts partial devolved tax responsibility into full sovereign revenue-system responsibility. The existing Welsh machinery (WRA, devolved taxes, Welsh Rates of Income Tax) is a real starting base but covers only a fraction of the tax system a sovereign state would need. The major reserved tax heads, including income tax collection, consumption taxes, corporation tax and customs, would need to be absorbed, administered and enforced at a scale far beyond the current WRA. The tax system is what replaces the block grant: M03-IND and M01-IND are two sides of the same independence fiscal question. Welsh fiscal-position data shows that devolved taxes currently provide around 21% of Welsh Government day-to-day funding, with the remaining 79% provided through the UK block grant; a sovereign revenue system would need to replace that funding function through some combination of taxation, borrowing, fiscal adjustment and transition settlement. Holtham’s tax-devolution analysis was designed for accountability within the UK framework, not for sovereign tax-system design: the diagnosis of Welsh tax capacity travels, the devolution remedy does not.

Federal variant: see Appendix F, M03-FED.

M04 — Fiscal measurement and instrumentation

M04-B. Baseline: repaired or protected current devolution

Competence: Fiscal measurement is partly created by the instruments devolved.

Capacity: The fiscal framework requires forecasting, block grant adjustment, and data arrangements for devolved taxes. The Welsh Government may use OBR forecasts or put in place alternative independent forecasting arrangements. UK Government income tax data access is necessary to support Welsh policy development and comparable forecasts.

Evidence strength: High.

Sources: S001, S002.

Caveat: This is high evidence for the existence of official forecasting and adjustment machinery, but only partial support for the broader fiscal instrumentation argument. The framework improves measurement because devolved taxes require forecasts, reconciliations, and block grant adjustments, but deep fiscal instrumentation remains partial because the most important macro-fiscal instruments and much of the tax base remain UK-controlled.

M04-ED. Enhanced devolution

Competence: More devolved fiscal instruments would increase the need for Welsh measurement capacity.

Capacity: Enhanced devolution would require stronger capacity to forecast, measure, and manage Welsh tax revenues and risks.

Evidence strength: Medium.

Sources: S001, S002.

Caveat: Holtham supports the logic but does not fully develop the later instrumentation argument.

M04-IND. Independence

Competence: Sovereign. An independent Wales would hold full competence over its own fiscal statistics, national accounts, revenue and expenditure measurement, fiscal forecasting, tax-base data, debt reporting and independent fiscal scrutiny. The current framework, under which Welsh fiscal measurement depends partly on UK-produced infrastructure (PESA, OBR, HMRC data, ONS statistics) and partly on Welsh-produced or academic analysis (WRA, Welsh Government budget material, Wales Fiscal Analysis), would be replaced by a requirement for sovereign fiscal measurement and transparency capacity.

Capacity: Build-required. Wales currently produces some fiscal measurement through the Welsh Revenue Authority, Welsh Government budget processes and Welsh academic analysis. But the core fiscal statistics infrastructure is UK-produced: PESA Wales is compiled by the Wales Office from UK Government data, OBR forecasts serve as the baseline for Welsh fiscal planning, HMRC administers and reports on Welsh Rates of Income Tax, and ONS produces the economic statistics on which Welsh fiscal analysis depends. A sovereign Wales would need to build or replace national accounting capacity, an independent fiscal forecasting institution, tax-base measurement across all major tax heads, debt sustainability analysis, fiscal risk assessment, and the institutional architecture for transparent, independently scrutinised public finances. The IMF Fiscal Transparency Code sets out the international standard: fiscal reporting, fiscal forecasting and budgeting, fiscal risk analysis and management, and resource revenue management. Meeting that kind of standard would be central to borrowing credibility, investor confidence and democratic fiscal accountability.

Evidence strength: Medium. The current Welsh measurement landscape is well documented, including its constraints: the OSR has noted limits around timeliness, volatility and the absence of Welsh monthly GDP statistics. The fiscal-position sources (S064 to S067) are themselves products of the measurement infrastructure the cell describes. The international fiscal transparency standard (S073) is clear on the kinds of fiscal reporting, forecasting and risk-management functions expected of a credible sovereign fiscal system. But no Welsh sovereign fiscal-statistics framework, national accounts architecture, independent fiscal institution or fiscal transparency system exists outside the UK framework. The evidence reaches Medium because the need is strongly established (M01 to M03 all depend on it, and international standards are clear) and current Welsh capacity is directly evidenced, but the sovereign design is absent from the source base.

Sources: current measurement S059 (Welsh measurement constraints), S070 (WRA tax data), S071 (WRIT and HMRC collection data); fiscal-position and measurement products S064, S065, S066, S067; baseline arrangements S001 (Holtham, instrumentation logic), S002 (fiscal framework forecasting and data arrangements); sovereign design standard S073 (IMF Fiscal Transparency Code); sovereign design context S068 (fiscal data quality as a condition of borrowing credibility).

Caveat: M04-IND is a dependency cell. Sovereign funding (M01), borrowing (M02) and taxation (M03) cannot operate credibly without fiscal reporting, forecasting, tax-base measurement, debt sustainability analysis and independent scrutiny. This is the same dependency role that M04-FED plays in the federal column, now translated to the independence context. The current partial Welsh measurement capacity (WRA, Welsh Government budget processes, Welsh academic analysis) is a real starting base, but it operates within and depends on UK-produced fiscal infrastructure. Independence would not merely expand Welsh measurement capacity; it would relocate the measurement problem from a devolved supplement to UK data into a sovereign requirement for full fiscal transparency. The Holtham instrumentation logic remains relevant as a diagnostic principle: fiscal powers require fiscal measurement. But Holtham’s measurement arrangements were designed for devolution, not for sovereign fiscal statistics.

Federal variant: see Appendix F, M04-FED.

M05 — Currency and monetary policy

Cross-column mechanism note (M05)

M05 is the first deliberately thin row.

For the baseline and enhanced-devolution columns, thinness is the finding. Currency and monetary policy are reserved and institutionally held at UK level. The repaired baseline can improve Welsh fiscal management, and enhanced devolution can widen fiscal or administrative capacity, but neither changes the currency or monetary-policy architecture unless the option itself is redefined.

For the federal column, the question changes but does not become a Welsh monetary-sovereignty question. In ordinary federal design, currency and monetary policy sit with the federal central bank. Wales’s relevant capacity would be macroeconomic analysis, scrutiny and participation in federal institutions, not independent monetary operation.

The independence column is the point at which the live currency questions arise: sterling use, formal monetary union, unilateral sterlingisation, own currency, central-bank creation, lender-of-last-resort capacity, reserves, credibility and transition risk.

M05-B. Baseline: repaired or protected current devolution

Competence: Reserved. Fiscal, economic and monetary policy, including the issue and circulation of money, sit outside Welsh legislative competence.

Capacity: Not-held by Welsh institutions. Monetary-policy capacity sits at UK level through HM Treasury and the Bank of England.

Evidence strength: High.

Sources: S018, S019, S020.

Caveat: This is a thin cell because the question is structurally outside the Welsh settlement, not because evidence is missing. Repaired or protected devolution can improve fiscal management around the Welsh budget, but it does not give Wales currency or monetary-policy competence.

M05-ED. Enhanced devolution

Competence: Reserved unless the model ceases to be enhanced devolution and becomes a fundamentally different constitutional settlement.

Capacity: Not-held by Welsh institutions. Enhanced devolution can strengthen fiscal administration, budgeting, tax capacity and economic policy within the UK framework, but it does not create a Welsh currency, central bank, lender-of-last-resort function, or monetary-policy instrument.

Evidence strength: High for the current reservation; Low or absent for any enhanced-devolution proposal to alter monetary sovereignty.

Sources: S018, S019, S020.

Caveat: The correct coding is thin-but-settled. Currency and monetary policy do not become a live enhanced-devolution mechanism unless the option is redefined beyond enhanced devolution.

M05-IND. Independence

Competence: Sovereign. An independent Wales would hold sovereign competence over currency choice, monetary institutions, central banking arrangements, financial stability architecture, reserve policy, payments systems, lender-of-last-resort arrangements, and the relationship between domestic banking supervision and the chosen monetary regime.

Capacity: Build-required. No existing Welsh institution currently performs sovereign monetary-policy, central-bank, reserve-management, payments-system, lender-of-last-resort, or currency-issuance functions. Even a transitional or continued sterling-use arrangement would not remove the capacity problem, because it would still require credible fiscal, financial-stability, banking, liquidity, deposit-protection and external-relations arrangements. The capacity gap is not reducible to ordinary administrative build-out: this row concerns the monetary architecture of the state itself, which conditions borrowing credibility (M02), revenue-system operation (M03), fiscal measurement (M04), and the entire economic framework within which the other mechanisms operate.

Evidence strength: High. The evidence is High not because Wales has an agreed sovereign monetary design, but because the institutional requirements and consequences of the main monetary options are extensively documented. Sterling use, formal monetary union, unilateral sterlingisation, a new Welsh currency, reserve-backed transition, central-bank creation, financial-stability arrangements, deposit protection, and lender-of-last-resort capacity are all well mapped in international and UK-adjacent evidence. The Scottish independence debate produced detailed analysis of the currency options available to a newly independent state within the UK monetary context. International monetary economics, IMF and BIS institutional material, and central-banking literature provide extensive evidence on what each monetary option requires and what constraints it carries. The remaining gap is a Welsh policy choice and transition design, not an absence of evidence about what the options require. This distinguishes M05-IND from M01-IND to M04-IND, where the evidence band is Medium because Welsh-specific sovereign fiscal designs are absent. Here, the monetary options themselves are mature, mapped, and institutionally well evidenced, even though Wales has not chosen among them.

Sources: UK monetary architecture S019, monetary policy remit S020, reservation baseline S018, Scottish currency-options comparator S074 and S075, international central-banking and monetary-framework evidence S076 and S077.

Caveat: The High evidence band records the strength of the evidence about what sovereign monetary options require, not the existence of a Welsh monetary design. Wales has not made a currency choice, and the cell does not presuppose one. The strategic and political question of which monetary arrangement to pursue remains entirely open. The capacity requirement is present regardless of which option is chosen: sterling use without a central bank still requires fiscal credibility, deposit-protection arrangements, financial-stability capacity and external monetary relations; a new currency requires all of those plus currency-issuance, reserve-management, exchange-rate and transition capacity. The cell therefore records that the monetary question is well mapped but the Welsh answer is not settled.

Federal variant: see Appendix F, M05-FED.

M06 — Debt, pensions, and inherited liabilities

Cross-column mechanism note (M06)

M06 is one mechanism ID covering three strands that do not code identically.

Debt and inherited liabilities share the clean reserved Not-held shape at baseline and enhanced devolution. Their live problem is apportionment under separation, which is held in the independence column.

Pensions diverges and is the responsibility-capacity asymmetry in live form. Competence is reserved but for the narrow firefighters’ exception, while Welsh-funded bodies carry employer-cost, administration and budget exposure: the funded LGPS through Welsh-administered, Audit Wales-audited funds, and the unfunded schemes through UK-set employer-contribution increases passing into Welsh health and local-government budgets.

The federal column is itself stranded and the most fragile in the grid so far. Ongoing functions may remain federal with only scrutiny capacity, but apportioning the inherited stock is new allocation and assumption architecture, and the evidence is separation-context rather than federal-formation.

The matrix carries the dominant reserved token throughout, UK / Not-held at baseline and enhanced devolution and Design-dependent / Build-required at federal. The divergence lives in the strands and caveats, not in the token.

M06-B. Baseline: repaired or protected current devolution

Debt. Competence: UK. Sovereign debt management is a reserved function held at UK level through HM Treasury and the Debt Management Office. Capacity: Not-held. Wales holds no debt-management or gilt-issuance function; it exists at UK level.

Inherited liabilities. Competence: UK. Accumulated UK public-sector liabilities are recorded and managed at UK level; the apportionment question is not live absent constitutional separation. Capacity: Not-held.

Pensions. Competence: reserved, with a narrow devolved exception for firefighters’ pensions. State pension, social security, and occupational and personal pensions are reserved; the public-service framework sits in the Public Service Pensions Act 2013 with Treasury control of valuations and the cost cap; the NHS and Teachers schemes are England-and-Wales schemes with centrally set rules; Welsh Ministers exercise scheme-related powers only in the firefighters’ field. Capacity and exposure: Welsh-funded bodies carry real exposure despite reserved competence. The funded LGPS is administered by Welsh pension funds, pooled through the Wales Pension Partnership and audited by Audit Wales, with employer contributions and deficit funding falling on Welsh local authorities and Welsh-funded employers. On the unfunded NHS and Teachers schemes, UK-set employer-contribution increases pass through Welsh health and local-government budgets. This is competence-reserved, exposure-present, not Not-held.

Evidence strength: High. The Welsh exposure evidence is stronger on employer-cost and budgetary exposure than on a single consolidated Welsh pensions-risk analysis.

Sources: debt S018, S023, S024; inherited liabilities S025, with apportionment context S027, S028, S029 held for the independence cell; pensions competence S018, S024, S026, S030; pensions exposure S031, S034, S035, S038, S039, S040, S041, S042, S043; fiscal-context background S032, S033.

Caveat: The dominant token records debt and inherited liabilities, which are cleanly UK-held. The pensions strand diverges: competence is reserved but for the firefighters’ sliver, while Welsh-funded bodies carry employer-cost, administration and budget exposure. The token compresses that split; the divergence is the finding.

M06-ED. Enhanced devolution

Debt. Competence: UK. Capacity: Not-held. Enhanced devolution does not transfer the debt-management or gilt function.

Inherited liabilities. Competence: UK. Capacity: Not-held. The apportionment question does not become live under enhanced devolution.

Pensions. Competence: reserved, with the same narrow devolved exception for firefighters’ pensions. Capacity and exposure: the same Welsh exposure as baseline. Enhanced devolution may give marginally more fiscal room to absorb pension-cost exposure but does not change the pension framework or scheme competence.

Evidence strength: High.

Sources: as M06-B.

Caveat: As M06-B; the divergence persists unchanged. Enhanced devolution changes neither the debt function nor the pension framework.

M06-IND. Independence

Competence: Sovereign. An independent Wales would hold sovereign responsibility for managing its own public debt, liabilities and pension system, including any share of UK public-sector debt, inherited or contingent liabilities, and accrued pension obligations allocated through the separation settlement. The apportionment questions held throughout the baseline, enhanced-devolution and federal cells would become active: negotiation of a share of UK public-sector debt, allocation of inherited and contingent liabilities, apportionment of accrued pension rights and ongoing pension obligations, and the design of a Welsh state pension and public-service pension framework.

Capacity: Build-required. Wales would need sovereign debt-management capacity (linked to M02-IND), liability-management and contingent-liability capacity, a state pension system or negotiated continuation arrangement, public-service pension scheme governance, LGPS transition or continuation arrangements, actuarial and fiscal-risk capacity, and negotiation capacity for the separation settlement itself. The pension strand carries an additional structural complication: the NHS and Teachers schemes are currently England-and-Wales structures, not Wales-only, unlike the Scottish schemes which were already administratively separate before the independence debate. Separation would therefore require scheme division or apportionment of accrued rights across a shared-scheme boundary, not merely the transfer of an already-separate Welsh scheme.

Evidence strength: Contested. The evidence base is substantial, but the outcome is highly sensitive to the apportionment basis and transition model. Debt continuity principles are well established: S027 records the UK Government’s position that the continuing state would honour UK-issued debt while an independent successor state would negotiate a fair and proportionate share. The Scottish independence debate provides structured comparator evidence on debt and pension allocation through S028 and S029. However, population share, GDP share, historic contribution, geographic allocation and negotiated settlement can produce materially different results. Pension liabilities are further complicated by the structure of England-and-Wales schemes (NHS and Teachers are not Wales-only), and inherited or contingent liabilities are especially negotiation-dependent. The evidence therefore supports the conclusion that an independent Wales would require a negotiated liabilities settlement, not a determinate estimate of what that settlement would be.

Sources: debt continuity S027, pension apportionment S028, debt and liability apportionment logic S029, balance-sheet context S025; pension exposure baseline S031, S034, S035 (carried forward from M06-B for the Welsh pension landscape).

Caveat: M06-IND is the first Contested cell in the independence column. The evidence band records that the source base is real and serious but that the outcome varies depending on the allocation principle chosen, the negotiation model, the scheme structure and the transition design. This is not a Welsh design gap (which would be Medium) or an extensively mapped option space (which would be High). It is a genuine contestation in the evidence itself. The three strands do not code identically even at independence: debt continuity principles are the most settled, pension apportionment is well-documented but structurally complicated by the England-and-Wales scheme boundary, and inherited or contingent liabilities are the most negotiation-dependent. The cell should be read alongside M02-IND (sovereign borrowing) because the debt share and inherited-liability allocation directly affect borrowing credibility and fiscal starting position.

Federal variant: see Appendix F, M06-FED.

M07 — Social security and welfare

Cross-column mechanism note (M07)

M07 is a split-responsibility row.

The baseline is not simply reserved. The main social-security system is reserved to the UK, but Wales already carries devolved welfare and assistance functions through Welsh benefits, local-authority delivery, discretionary assistance and schemes tied to education, local taxation and poverty mitigation. The baseline therefore combines UK competence over the main cash-benefit system with partial Welsh capacity at the margins.

Enhanced devolution has two possible meanings and they must not be collapsed. Strengthening the Welsh Benefits System means streamlining and improving access to benefits and grants already governed by Welsh Government or local authorities. Devolving major social-security benefits would be a different proposition: it would transfer fiscal volatility, administrative burden and claimant risk, and would require much larger tax, borrowing, digital, data and delivery capacity.

The federal column is design-dependent. A federal UK could retain social security at federal level, which is the Commission’s assumption, or allocate parts of it to Wales. Either design still requires Welsh capacity. If social security stays federal, Wales needs intergovernmental and analytical capacity to manage interactions with devolved services. If social-security powers are allocated to Wales, Wales needs delivery capacity and fiscal-risk capacity.

The independence column makes social security a full sovereign responsibility (M07-IND, Medium): the Welsh starting base is evidenced but no sovereign welfare design exists. It should not be allowed to set the baseline or enhanced-devolution reading.

M07-B. Baseline: repaired or protected current devolution

Competence: UK for the main social-security system, with limited devolved welfare and assistance schemes in Wales. Major cash benefits including the state pension, Universal Credit, disability benefits and winter heating support remain reserved and administered by DWP. Limited Welsh schemes include Council Tax Reduction, Free School Meals, the School Essentials Grant, Education Maintenance Allowance, and the Discretionary Assistance Fund or adjacent devolved assistance.

Capacity: Partial-held. Wales holds administrative and policy capacity for Welsh benefits and devolved grants, including local-authority delivery and current Welsh Benefits System streamlining work, but does not hold the main social-security system, DWP administration, or the fiscal risk of major cash benefits.

Evidence strength: High.

Sources: S044, S045, S046, S047.

Caveat: This is not a simple reserved row. The main system is reserved, but Wales has real capacity around devolved benefits, take-up, eligibility alignment, local delivery and discretionary assistance. The baseline finding is split responsibility: UK competence over the main system, Welsh exposure and limited Welsh assistance capacity at the margins.

M07-ED. Enhanced devolution

Competence: Design-dependent. Enhanced devolution could mean either strengthening the Welsh Benefits System within existing devolved schemes, or a more substantial transfer of social-security powers. The first is available within the present settlement; the second would require UK legislation and a redesigned fiscal settlement.

Capacity: Build-required. A larger Welsh welfare role would require policy, administrative, digital, appeals, fraud and error, forecasting, data-sharing, payment, delivery and claimant-support capacity. It would also require fiscal capacity to carry volatility, demand risk and liabilities.

Evidence strength: High for the current Welsh and Scottish evidence base; Medium for any specific Welsh enhanced-devolution design.

Sources: S044, S045, S046, S047, S048.

Caveat: The Commission’s Welsh-specific evidence supports feasibility only with a substantial increase in tax and borrowing powers. Scotland shows that partial devolution creates major institutional and administrative costs, interaction problems with reserved UK benefits, and the need for a dedicated delivery agency. The cell therefore records enhanced welfare devolution as possible but capacity-intensive, not as a simple transfer of policy choice.

M07-IND. Independence

Competence: Sovereign. An independent Wales would hold full competence over the whole social-security system: the state pension, working-age benefits, disability and carer benefits, housing support, child-related payments, and the devolved assistance schemes that already exist. The reserved DWP-administered system and the limited Welsh schemes would become a single sovereign welfare responsibility.

Capacity: Build-required. Wales holds real but partial machinery: the Welsh Benefits System, local-authority delivery, discretionary assistance, and the streamlining work underway. But the core working-age and pensioner cash-benefit system, the largest single item of public expenditure, is administered by DWP, with payment systems, claimant data, assessment, appeals, and fraud-and-error control all at UK scale. A sovereign Wales would need a full benefits agency or agencies, payment infrastructure, claimant data systems, assessment and decision-making capacity, an appeals tribunal route, fraud-and-error control, forecasting for demand-led volatility, and a transition arrangement from DWP systems. The Scottish experience shows that even partial devolution required a dedicated agency and carried major administrative cost and reserved-system interaction problems; full sovereign welfare is a larger jump again.

Evidence strength: Medium. The Welsh starting base is directly evidenced (S044, S045, S046, S047), and the Scottish comparator evidences the institutional cost and design of building benefit-delivery capacity (S048). The scale of the reserved system is documented through DWP expenditure and caseload data (S078). But no Welsh sovereign social-security design exists: no Welsh benefits-agency design, no transition plan from DWP, no Welsh demand forecast or fiscal-risk model for a sovereign welfare budget. This is the M01-M04 pattern: diagnosis and starting base evidenced, sovereign design absent. Medium, not High (no mapped sovereign option space comparable to M05’s currency literature), and not Contested (the gap is design-absence, not basis-sensitivity).

Sources: existing Welsh machinery S044 (Commission welfare analysis and capacity warning), S045 (Welsh Benefits System), S046 (Welsh Benefits Charter), S047 (streamlining route map); comparator for benefit-devolution institutional cost S048 (Scotland Act 2016 implementation, comparator only, not Welsh evidence); reserved-system scale S078 (DWP Benefit expenditure and caseload tables 2026, expenditure and caseload context for the UK benefit system).

Caveat: Independence converts a split-responsibility row into a full sovereign welfare responsibility. The Welsh Benefits System is a real starting base but covers a small fraction of the cash-benefit system; the state pension and working-age benefits are the largest expenditure transfer and carry demand-led volatility. The Commission’s warning (S044) that even enhanced welfare devolution would require substantial tax and borrowing powers applies with greater force to sovereignty: M07-IND is fiscally bound to M01-IND (revenue), M02-IND (borrowing for volatility) and M06-IND (the state-pension apportionment strand). The Scottish comparator shows the institutional cost of building delivery capacity but is not evidence of a Welsh design.

Federal variant: see Appendix F, M07-FED.

M08 — State capacity and transition capacity

Cross-column mechanism note (M08)

M08 is the capacity row rather than another reserved-function row.

The baseline finding is that Wales already has real state capacity for devolved government. It has institutions, civil-service machinery, delivery systems, public finance processes, audit and risk arrangements, and devolved policy administration. But that capacity is bounded by the functions Wales currently carries. It should not be mistaken for transition capacity for reserved or sovereign functions.

Enhanced devolution is therefore not only a competence question. Additional powers require additional machinery. The Scottish social-security example is the warning: transferring functions means building delivery institutions, systems, staff, data, governance and risk capacity, and doing so before public-facing responsibility is moved.

The federal column is design-dependent but build-required. A federal Wales would need capacity not only to administer its own functions but to participate in shared-rule machinery: fiscal rules, equalisation, tax administration, borrowing oversight, dispute resolution and intergovernmental coordination.

The independence column makes state capacity full sovereign capacity (M08-IND, Medium): it is the aggregate dependency cell, drawing on every other independence build requirement, with current capacity evidenced but no sovereign transition design in the source base.

M08-B. Baseline: repaired or protected current devolution

Competence: Welsh for devolved machinery, UK for reserved functions. Wales holds competence over the machinery of devolved government and the administration of devolved functions; reserved functions and the state-wide systems Wales does not administer remain UK.

Capacity: Partial-held. Wales has real governing capacity: ministerial departments, civil-service administration, public finance management, audit, risk, internal control, devolved delivery bodies and local delivery networks. But it does not hold full transition capacity for major reserved functions, sovereign institutions, macro-fiscal management, large-scale benefit transfer, currency, defence, foreign affairs or full state administration.

Evidence strength: High.

Sources: S044, S045, S046, S047, S049, S050.

Caveat: This is an inversion of the reserved rows. The issue is not absence of Welsh machinery, but the boundary between ordinary devolved-state capacity and transition capacity for larger constitutional functions. Baseline capacity is real but bounded.

M08-ED. Enhanced devolution

Competence: Design-dependent. Enhanced devolution could require Wales to assume additional functions within a still-UK constitutional framework. The competence position depends on which functions transfer and which remain reserved.

Capacity: Build-required. Further devolution would require programme design, legal implementation, staffing, digital systems, data-sharing, finance, audit, appeals or complaints routes, risk management and delivery readiness before functions move.

Evidence strength: High.

Sources: S044, S048, S049, S050.

Caveat: The evidence is strong for the capacity condition, not for any specific package of enhanced devolution. The Commission and the Scottish social-security example support the principle that capacity must be built ahead of delivery, and that capability is gained through experience. Enhanced devolution is therefore not just a legal transfer question; it is an institutional build question.

M08-IND. Independence

Competence: Sovereign. An independent Wales would hold the full range of sovereign state functions: not only the devolved machinery it already runs, but defence, foreign affairs, borders, full fiscal and monetary administration, the reserved regulatory state, and the central coordinating capacity of a sovereign government.

Capacity: Build-required. This is the row where the capacity gap is widest and most cross-cutting. Wales has real devolved-government capacity (S049, S050): ministries, civil service, public finance management, audit, delivery bodies. But sovereign statehood requires building or acquiring the entire reserved-function state, plus the transition capacity to stand it up: machinery-of-government design, a sovereign civil service at scale, the institutions every other independence cell depends on (M01 revenue, M02 debt management, M03 tax administration, M04 fiscal statistics, M05 monetary, M06 liabilities, M07 welfare, M09 payments, M10 financial regulation), and the negotiation and transition-management capacity to sequence it all.

Evidence strength: Medium. The existing Welsh state capacity is directly evidenced (S049, S050), and the Commission (S044) and Scottish comparator (S048) evidence the principle that capacity must be built ahead of delivery. But M08-IND is a dependency cell: it aggregates the build requirements of every other independence cell, and like them it has an evidenced diagnosis (current capacity is real but bounded) and an absent sovereign design (no Welsh transition plan, no machinery-of-government blueprint, no sequencing or costing). Medium fits the same way it fits M04-IND, which is also a dependency cell: the need is strongly established through the cells that depend on it, current capacity is evidenced, the sovereign design is absent. Not High (no mapped option space), not Contested (not basis-sensitive; it is design-absent).

Sources: existing capacity S049 (Welsh Government consolidated accounts and machinery), S050 (civil-service workforce and organisational capacity); capacity-build principle S044 (Commission), S048 (Scotland Act 2016 implementation, comparator only); dependency on M01-IND to M07-IND, M09-IND, M10-IND.

Caveat: M08-IND is the aggregate dependency cell of the independence column. Its evidence band reflects the same structure as M04-IND: the need is established through the cells that rely on it, current Welsh capacity is real but bounded, and no sovereign transition design exists in the source base. The cell should not be read as a verdict on whether Wales could build sovereign state capacity, only that the capacity is build-required and the design is not evidenced. Read alongside every other independence cell.

Federal variant: see Appendix F, M08-FED.

M09 — Payment resilience and payment rails

Cross-column mechanism note (M09)

M09 is a competence-capacity split like the reserved rows, but with a live margin the pure reserved rows do not have.

The core payment-services regime is reserved at baseline and under enhanced devolution. Schedule 7A reserves financial services, and the operative regime sits with the Payment Services Regulations 2017, the FCA and the Payment Systems Regulator. The competence token is therefore UK at both baseline and enhanced devolution, and it does not become design-dependent until a federal settlement could relocate it.

The capacity finding is the partial inversion of a pure reserved row. Wales holds adjacent delivery capacity over public-sector payment flows through procurement practice and Project Bank Accounts, which is why the baseline is Partial-held rather than Not-held. But that capacity addresses poor payment practice and counterparty risk, not rail-level exclusion or denial, so it should not be read as payment-rail resilience.

The enhanced-devolution finding is that resilience work does not have to wait for constitutional change, but its scope is bounded. Wales can build a protected public-sector payment mechanism within current competence; it cannot own or regulate the core regime without a transfer of financial-services competence. The federal column is design-dependent, and the independence column (M09-IND, Medium) is held with the France caution carried forward: constitutional location determines how far payment resilience can be owned, regulated and operationalised, but it does not determine whether resilience work must wait, and sovereign allocation does not by itself guarantee resilience.

M09-B. Baseline: repaired or protected current devolution

Competence: UK. Payment services and payment systems sit within reserved financial services and financial markets. Schedule 7A reserves financial services; the operative regime is the Payment Services Regulations 2017, FCA authorisation and registration, and the Payment Systems Regulator, all at UK level. Wales does not hold competence over payment-services regulation, payment rails, or payment-system sovereignty.

Capacity: Partial-held. Wales holds adjacent delivery capacity over public-sector payment flows, not over the payment regime itself. Project Bank Accounts show Wales already using procurement levers to protect payment flows in public delivery chains, and the Development Bank of Wales is a live financing host. But the held part addresses public-sector payment practice, not the rails: it does not reduce dependence on external card and payment networks, and it does not address exclusion, suspension or denial of access by an external payment-infrastructure controller.

Evidence strength: High.

Sources: competence S018, S051, S052; adjacent Welsh capacity S055, S056.

Caveat: The Partial-held token records real but narrow capacity. Project Bank Accounts and procurement payment practice address poor payment practice, delayed payment and counterparty risk, which is a different threat model from payment-rail exclusion or denial. The landscape contains a financing host and procurement-side payment-flow precedents but no dedicated payment-resilience vehicle and no Welsh deposit-taking public banking host. The held part should not be read as rail-level resilience.

M09-ED. Enhanced devolution

Competence: UK. The core payment-services regime remains reserved under enhanced devolution. Wales cannot regulate payment services, own payment rails, or create payment-system sovereignty unless financial-services competence is transferred, which would take the model beyond enhanced devolution.

Capacity: Build-required. Wales can strengthen public-sector payment resilience at the margins: procurement, grants, continuity planning and authorised-provider arrangements. A protected banking and payment mechanism for Welsh Government, public bodies, procurement flows and grant payments is achievable within current competence and would give the public sector more control over its own operational payment flows.

Evidence strength: High.

Sources: competence S018, S051, S052; Welsh capacity S055, S056.

Caveat: The available build is real but structurally bounded by the UK payment-services regime: Wales can strengthen public-sector payment resilience around procurement, grants, continuity planning and authorised-provider arrangements, but cannot regulate payment services, own payment rails, or create payment-system sovereignty under enhanced devolution unless financial-services competence is transferred. The value of the available build is institutional rather than constitutional; it could expand if competence changed later.

M09-IND. Independence

Competence: Sovereign. An independent Wales would hold sovereign competence over financial-services regulation, payment-system oversight, and the decision whether to build or join payment infrastructure.

Capacity: Build-required. Wales holds adjacent public-sector payment-flow capacity (Project Bank Accounts, DBW as financing host) but no payment-rail capacity. Sovereignty removes the reserved-competence ceiling but does not supply rails: a sovereign Wales would still depend on externally owned card and settlement networks unless it built or joined sovereign payment infrastructure, and would need payment-system regulation, oversight, and operational resilience capacity.

Evidence strength: Medium. The Welsh starting base (public-sector payment-flow capacity) is evidenced (S055, S056), the reservation baseline is documented (S018, S051, S052), and the feasibility of sovereign payment infrastructure is evidenced by international precedent (S053, S054). The France caution (S057) shows that sovereign monetary and central-bank allocation does not by itself remove exposure to externally owned payment rails. But the evidence points consistently to a single conclusion: payment resilience requires a build-or-join decision, and no Welsh sovereign payment design exists. This is the M01-M04 pattern: feasibility and caution both evidenced, Welsh sovereign design absent. The band is Medium, not Contested: unlike M06, the evidence does not produce materially different results depending on allocation basis; it points one way, to a design gap. The France caution is a caveat within a Medium cell, not a driver to Contested.

Sources: reservation baseline S018, S051, S052 (removed under sovereignty but recording what is replaced); Welsh public-sector payment-flow base S055 (DBW), S056 (Project Bank Accounts); feasibility precedents S053 (UPI), S054 (Pix); France caution S057 (producer confirmed European Central Bank; comparator and context only).

Caveat: Independence makes payment-system competence sovereign and removes the reserved ceiling, but it does not by itself deliver payment-rail resilience. France (S057) shows that full sovereign monetary and central-bank allocation leaves exposure to externally owned rails unless dedicated infrastructure is built or joined. The feasibility precedents (UPI, Pix) show sovereign rails are buildable at scale but were built by far larger states, and feasibility is not a Welsh build plan. M09-IND is bound to M05-IND (monetary architecture) and M10-IND (financial regulation).

Federal variant: see Appendix F, M09-FED.

M10 — Green Bank and public finance vehicles

Cross-column mechanism note (M10)

M10 is a form-dependent row.

The baseline finding is that Wales already has development-finance capacity through the Development Bank of Wales. DBW lends, invests, guarantees, co-invests and de-risks. That capacity is real and sits within Welsh competence. But the row covers a range from development lending to deposit-taking public banking, and Welsh capacity covers only the first form.

The three-form distinction is the row’s analytical structure. A Green Bank as development lender is DBW-adjacent and available within existing Welsh competence and institutional capacity. Green savings products delivered through an authorised partner could create a public-facing savings-to-investment link without the Welsh vehicle itself becoming a deposit-taking institution. A deposit-taking Green public bank is the strongest form and the largest institutional jump: it would require UK regulatory authorisation under the reserved financial-services regime, along with prudential structure, governance, capital and liquidity capacity. The third form belongs in the contestable category, not in present capacity.

Enhanced devolution does not change this structure. The buildable Green Bank form is development finance within devolved purposes. The deposit-taking boundary is a UK regulatory constraint that sits within the financial-services reservation, and it does not move under enhanced devolution unless the model is redefined to transfer that competence.

The federal column is design-dependent but the federal question is narrow. The development-finance form does not depend on federalisation. The only question federalisation adds is whether it changes the regulatory allocation for deposit-taking, and that is not evidenced for Wales.

The independence column (M10-IND, Medium) removes the UK regulatory constraint on deposit-taking, making the strongest form available, but sovereign competence does not supply the institutional capacity.

M10-B. Baseline: repaired or protected current devolution

Competence: Welsh. Development finance within devolved purposes sits within Welsh competence. The Development Bank of Wales operates as a development-finance host: lending, equity investment, guarantees, co-investment, de-risking and targeted public finance for Welsh economic and policy purposes. The constraint is institutional form, not competence for the development-finance model: a deposit-taking bank would require UK regulatory authorisation under the reserved financial-services regime.

Capacity: Partial-held. Wales has real development-finance capacity through the Development Bank of Wales, including the Sustainable Agriculture Loan Scheme and related repayable public-finance instruments. But the row covers the full range from development lending through green savings to deposit-taking public banking, and Welsh capacity covers only the development-finance form. DBW lends, invests, guarantees and de-risks, but it is not a deposit-taking institution and cannot host a deposit-funded Green Bank without restructuring or partnership.

Evidence strength: High.

Sources: S055; competence ceiling S018.

Caveat: The Partial-held token records the gap between available development-finance capacity and the full Green Bank range. Three forms should be distinguished. First, a Green Bank as development lender: DBW-adjacent, available in form, buildable within existing Welsh competence and institutional capacity. Second, green savings products delivered through an already-authorised partner: the Welsh vehicle does not itself take deposits but uses the proceeds or associated finance for eligible green investment. Third, a deposit-taking Green public bank: the strongest form and the largest institutional jump, requiring UK regulatory authorisation, prudential structure, governance, capital and liquidity capacity. The third form belongs in the contestable institutional-form category, not in present capacity.

M10-ED. Enhanced devolution

Competence: Welsh. The buildable Green Bank form is development finance within devolved purposes. Enhanced devolution does not need to transfer financial-services competence to enable a development-finance Green Bank, because that form sits within existing Welsh competence. The deposit-taking boundary remains a UK regulatory constraint unless financial-services competence is transferred.

Capacity: Build-required. Moving from existing DBW development-finance capacity to a functioning Green Bank would require policy design, product design, governance, staffing, risk management, pipeline development, monitoring, evaluation and possibly wholesale or bond-like fundraising. The build is real but within the development-finance family, not the banking family.

Evidence strength: High.

Sources: S055; competence ceiling S018.

Caveat: The Green Bank build under enhanced devolution is bounded by form. Development-finance instruments, including lending, guarantees, co-investment, de-risking and repayable finance for green projects, are buildable within Welsh competence and can use DBW as a host or precedent. Green savings via an authorised partner may be achievable without deposit-taking authorisation. A deposit-taking Green public bank would require UK regulatory authorisation and a different institutional form, and should not be treated as available under enhanced devolution unless financial-services competence is transferred.

M10-IND. Independence

Competence: Sovereign. An independent Wales would hold sovereign financial-services regulation, removing the UK regulatory ceiling that currently blocks a deposit-taking public bank.

Capacity: Build-required. Wales has development-finance capacity (DBW, S055) but no deposit-taking capacity. Sovereignty makes the deposit-taking Green public bank form available for the first time, but building it would still require prudential regulation, authorisation machinery, governance, capital adequacy, liquidity management, and depositor protection: a sovereign regulatory and institutional build, not just a remit extension for DBW.

Evidence strength: Medium. The Welsh starting base (DBW development finance) is directly evidenced (S055), and the competence ceiling that sovereignty removes is documented (S018). The three-form structure (development lender, green savings via partner, deposit-taking public bank) is the established analytical frame. But no Welsh sovereign deposit-taking bank design exists, and no sovereign financial-regulation architecture is specified. This is the M01-M04 pattern again: evidenced base, absent sovereign design. Not High (no mapped option space; the deposit-taking-bank design space is not an extensively mapped literature in the way currency options are), not Contested (not basis-sensitive).

Sources: Welsh development-finance base S055 (DBW); competence ceiling removed under sovereignty S018; three-form structure carried from M10-B and M10-ED.

Caveat: Sovereign competence removes the UK regulatory constraint on deposit-taking, making the strongest Green Bank form available, but it does not supply the capacity. A deposit-taking Green public bank would require a sovereign prudential-regulation regime (bound to M09-IND financial regulation), authorisation, capital, liquidity and depositor protection. The development-finance form (DBW-adjacent) remains available as it was under devolution; sovereignty changes only the deposit-taking ceiling. The cell records availability of the form, not a Welsh design for it.

Federal variant: see Appendix F, M10-FED.

M11 — Productivity and economic structure

Cross-column mechanism note (M11)

M11 is the structural-economy row.

Unlike a tax, pension, payment or banking row, productivity is not a single institutional lever. It is the outcome of many systems: skills, firm investment, innovation, infrastructure, transport, public services, sector composition, labour markets, finance, trade exposure, management practice, population structure, regional geography and macroeconomic context. That is why the baseline competence token is Shared and the capacity token is Partial-held.

The baseline finding is not absence. Wales has real economic-development capacity and live productivity policy. It also has official statistics, development finance, business support, skills policy, devolved public-service systems and regional delivery machinery. But the productivity gap is persistent, and some of the strongest levers that shape economic structure remain UK-level or externally conditioned.

Enhanced devolution should be treated as a capacity-build route rather than as a guaranteed productivity route. The evidence supports the need for long-term Welsh productivity strategy and stronger institutional machinery. It does not prove that any one enhanced-devolution settlement would close the gap. The evidence band is therefore Medium: strong diagnosis, live Welsh policy direction, but no settled constitutional productivity model.

The federal column remains design-dependent and Low. Federalism changes the allocation of economic levers and shared-rule machinery, but it does not itself provide a Welsh productivity settlement. It would require Welsh capacity to operate within federal fiscal, labour, market, investment and intergovernmental systems.

The independence column (M11-IND, Medium) makes the full set of economic levers sovereign, but sovereignty is not convergence. The productivity gap is structural and persistent, and acquiring the levers does not by itself raise productivity: that remains a long-term capacity, investment, sectoral and institutional question.

M11-B. Baseline: repaired or protected current devolution

Competence: Shared. Wales holds broad economic-development competence and important adjacent levers over skills, education, transport within devolved competence, planning, business support, development finance, local delivery and public-service capacity. The UK retains major structural levers including fiscal, economic and monetary policy, currency, financial services, financial markets, competition, import and export control, employment rights, industrial-relations law and important labour-market functions.

Capacity: Partial-held. Wales has real economic-development machinery, measurement capacity, development-finance capacity, policy programmes and live productivity policy. But it does not hold the full macroeconomic, fiscal, monetary, financial, trade, labour-market or infrastructure system that shapes productivity and economic structure.

Evidence strength: High.

Sources: competence S018, S058; measurement S059; diagnosis S060, S061; analysis S062.

Caveat: The baseline evidence is strongest for the diagnosis and for the competence split, not for any claim that current devolution can or cannot close the productivity gap. Wales’ productivity gap is persistent but not static: Welsh productivity remains well below the UK average, while some periods show stronger relative growth. The cell therefore records a structural problem and a partial capacity base, not a fatalistic conclusion.

M11-ED. Enhanced devolution

Competence: Design-dependent. Enhanced devolution could mean a stronger Welsh economic-development state within existing powers, or a wider transfer of powers around tax, borrowing, transport, energy, Crown Estate, skills, business support, innovation, employment support, regional policy and public investment. The competence position depends on the package.

Capacity: Build-required. Wales would need a more durable productivity state: long-term strategy, stronger economic intelligence, sector policy, SME scaling support, innovation capacity, skills alignment, development-finance depth, infrastructure coordination, public-service productivity, regional delivery and evaluation capacity.

Evidence strength: Medium.

Sources: S061, S062, S063.

Caveat: The evidence base is strong that productivity is a long-term structural problem and that Wales has live policy machinery. It is not yet strong evidence that any specific enhanced-devolution package would close the gap. The Welsh Government’s productivity mission is evidence of policy direction, while the Productivity Institute evidence supports the need for sustained, coordinated action. The cell should therefore treat enhanced devolution as a capacity-build route, not as a proven convergence mechanism.

M11-IND. Independence

Competence: Sovereign. An independent Wales would hold the full set of economic levers: fiscal, trade, industrial, competition, labour-market, financial, and the macroeconomic framework, alongside the economic-development competence it already holds.

Capacity: Build-required. Wales has real economic-development machinery and live productivity policy (S060, S061, S062, S063) but not the full macroeconomic, fiscal, monetary, trade and labour-market system. Sovereignty would make all those levers allocable by Wales, but productivity is not a lever: it is the long-run outcome of many systems, and acquiring the levers does not by itself change the outcome.

Evidence strength: Medium. The productivity diagnosis is strongly evidenced (S060 official productivity data, S061 economic and fiscal report, S062 Productivity Institute, S064 fiscal-deficit-and-underperformance link). The competence map is clear (S018, S058). But no Welsh sovereign economic strategy exists, and crucially the evidence does not establish that sovereignty would close the productivity gap: S062 and S064 show the gap is structural and persistent, and sovereignty is not convergence. Medium fits: diagnosis evidenced, sovereign design and outcome absent. Not High, not Contested. This is the closest of the M07-M11 batch to a pure M01-IND-style Medium.

Sources: productivity diagnosis S060 (sub-regional productivity), S061 (Wales Economic and Fiscal Report), S062 (Productivity Institute); fiscal-underperformance link S064 (Ifan and Sion); competence map S018, S058.

Caveat: Independence makes the full set of economic levers sovereign, but sovereignty is not convergence. The productivity gap is structural and persistent (S062, S064), and acquiring fiscal, trade, industrial and labour-market competence does not by itself raise productivity: that remains a long-term capacity, investment, sectoral and institutional question. M11-IND is the structural-economy row and depends on the whole grid: revenue (M01), borrowing (M02), the monetary framework (M05) and state capacity (M08) all condition what an independent economic strategy could achieve. The cell records sovereign competence over the levers, not an evidenced productivity outcome.

Federal variant: see Appendix F, M11-FED.

Annex C — Cross-cutting synthesis

This section reads the pattern across the coded mechanisms and constitutional models. It is included as an audit-facing synthesis: claims remain tagged to mechanism cells, source-log entries, or preserved cross-column mechanism notes so that the route from finding to evidence remains visible.

The synthesis does not recommend a constitutional option. It explains what the grid shows about fiscal capacity, dependency, build requirements and evidence strength across the current settlement, enhanced devolution and independence. Federal material is used only as archived comparator material where relevant.

The inline tags are retained in this annex because this is the evidence base. In the companion policy paper, those tags can be converted into lighter footnotes or annex references.

1. What the grid shows under the current settlement

The baseline reading is a partial-state pattern: Wales is neither empty of fiscal capacity nor autonomous in it. The current settlement gives Wales real administrative, policy and delivery capacity in several areas, but that capacity is bounded by reserved functions, UK fiscal architecture, UK monetary institutions, UK welfare administration, UK financial regulation, and shared or externally produced data systems [M04-B; M05-B; M06-B; M07-B; M09-B].

The recurring shape across mechanisms is adjacent or partial capacity held by Wales while the main function sits elsewhere. Wales runs devolved welfare and assistance schemes, local-authority delivery and the Welsh Benefits System, but the main cash-benefit system is reserved and administered by DWP [M07-B; S044, S045, S046, S047; S078]. It holds development-finance capacity through the Development Bank of Wales but not a deposit-taking form, because that sits within the reserved financial-services regime [M10-B; S055; S018]. It intervenes in public-sector payment flows through procurement practice and Project Bank Accounts, but does not own or regulate the payment rails [M09-B; S055, S056; S018, S051, S052]. It has real devolved-government machinery, civil service, public-finance management and audit, but that machinery is bounded by the functions Wales currently carries [M08-B; S049, S050].

In the core fiscal mechanisms the boundary is sharper still. Territorial funding is UK-controlled through the block grant and Barnett, with a needs-based factor inside that system rather than Welsh control of it [M01-B; S001, S002]. Borrowing and reserve tools are real but bounded within the UK fiscal framework [M02-B; S002, S003]. Devolved taxation gives Wales partial tax capacity and rate-setting accountability but not full tax autonomy [M03-B; S001, S002]. Currency and monetary policy are not held at all: this is thin because the question sits structurally outside the Welsh settlement, not because evidence is missing [M05-B; S018, S019, S020]. Debt and inherited liabilities are cleanly UK-held, while the pensions strand diverges as a responsibility-capacity asymmetry: competence is reserved but for the firefighters’ exception, yet Welsh-funded bodies carry employer-cost, administration and budget exposure [M06-B; S018, S023, S024, S031, S034, S035; cross-column note M06].

The baseline finding is therefore not a verdict on adequacy. It is a structural observation: the current settlement produces a partial state, with genuine capacity at the margins and the main fiscal, monetary, welfare and regulatory functions located elsewhere or shared. Where competence and capacity diverge, the divergence is the finding, not a caveat [00 use rule].

2. What enhanced devolution would actually change

Enhanced devolution reads mainly as an institutional-build question, not a legal switch. The enhanced-devolution column repeatedly codes Design-dependent or Build-required rather than a simple transfer of capacity: the mechanisms become possible or stronger only if institutional design, fiscal headroom, administrative capacity and intergovernmental arrangements are built alongside any competence transfer.

The pattern is visible across the column. A fuller needs-based settlement is a technical model and a negotiation starting point, not an automatically available outcome [M01-ED; S001, S002]. Further borrowing flexibility and tax devolution are possible but conditional on UK agreement and on the capacity to forecast, measure and manage the resulting risks [M02-ED; M03-ED; M04-ED; S001, S002]. Welfare carries two distinct meanings that must not be collapsed: strengthening the Welsh Benefits System within existing devolved schemes is available now, but devolving major social-security benefits would transfer fiscal volatility, administrative burden and claimant risk and would require much larger capacity, on the evidence of the Scottish experience [M07-ED; S044, S048; cross-column note M07]. State capacity is the same lesson generalised: additional powers require additional machinery built ahead of delivery [M08-ED; S044, S048, S049, S050; cross-column note M08].

The synthesis must also state what enhanced devolution does not move. Currency and monetary policy remain outside it unless the option is redefined into a different constitutional settlement [M05-ED; S018, S019, S020; cross-column note M05]. The core payment-services regime does not move either: Wales can strengthen public-sector payment resilience at the margins, but cannot own or regulate the rails without a transfer of financial-services competence, which would take the model beyond ordinary enhanced devolution [M09-ED; S018, S051, S052, S055, S056; cross-column note M09]. The deposit-taking Green Bank boundary is the same constraint in another form [M10-ED; S055; S018; cross-column note M10].

The enhanced-devolution finding is therefore that the column is a capacity-and-design route, not a competence switch, and that it is uneven: it strengthens some mechanisms substantially, leaves others untouched, and in several cases requires building the institution before the power is usable.

3. What independence converts from dependency into build-requirement

This is the central conversion finding. Independence does not make the fiscal problem disappear; it changes its form, converting dependency into construction.

The conversions run mechanism by mechanism. Block-grant dependency becomes a requirement to build sovereign fiscal capacity, because the Welsh fiscal diagnosis shows expenditure needs exceeding revenue capacity relative to the UK average, so the replacement is not fiscally neutral [M01-IND; S001 (diagnosis only), S064, S065, S066, S067]. Bounded devolved borrowing becomes sovereign debt management, a qualitative change in institutional form rather than an expansion of the existing caps [M02-IND; S068, S069]. Partial tax administration becomes a full sovereign revenue system: the WRA proves Welsh tax administration works, but covers only a fraction of what a sovereign system would collect [M03-IND; S070, S071]. UK-produced fiscal statistics become a sovereign fiscal-measurement and transparency requirement [M04-IND; S073, S068, S059]. Reserved welfare becomes sovereign welfare delivery, the largest single expenditure transfer, carrying demand-led volatility [M07-IND; S044, S048, S078]. Reserved monetary architecture becomes a currency and central-bank settlement [M05-IND; S074, S075, S076, S077]. Inherited debt, pensions and liabilities become a negotiated separation settlement [M06-IND; S027, S028, S029, S025].

M04 and M08 are the cells that show the conversion is systemic, not additive. Sovereign funding, borrowing and taxation cannot operate credibly without fiscal reporting, forecasting, tax-base measurement and independent scrutiny, which is the dependency role M04 plays [M04-IND caveat; M04-FED dependency-row note]. And every other independence build draws on aggregate state and transition capacity: M08 aggregates the build requirements of the whole column, which is why it is a dependency cell in the same structural sense as M04 [M08-IND; S049, S050; cross-column note M08]. Independence does not merely add powers mechanism by mechanism; it requires the infrastructure that lets the mechanisms work together.

The conversion finding is therefore that independence relocates the fiscal question rather than resolving it: from dependency on UK architecture into a requirement to construct sovereign architecture, with the systemic cells (M04, M08) conditioning whether the rest can function.

4. Administrative, systemic, contested: a stress test, not a taxonomy

This section applies the administrative / systemic / contested cut as a stress test, not as a completed typology. The cut is a lens held against the coded cells; its value lies partly in where it fails. Some mechanisms fit cleanly, some straddle categories, and some expose the limits of the taxonomy. The section shows the cut being tested, not imposed.

Clean fits. Some mechanisms sit reasonably within one group. As an administrative-scale build, M03 has a real adjacent base in the WRA and the central issue is scaling from that operational machinery to sovereign-scale revenue collection, though it carries large fiscal consequences [M03-IND; S070, S071]. As systemic builds, mechanisms that condition the others fit the middle group: M04 fiscal measurement, on which sovereign funding, borrowing and taxation all depend [M04-IND]; M05 monetary architecture, which conditions borrowing credibility, revenue operation and the whole economic framework [M05-IND; cross-column note M05]; and M08 state capacity, the aggregate dependency cell [M08-IND; cross-column note M08].

Straddlers. M07 welfare does not sit cleanly in one box. It is administrative in its delivery build, the construction of a benefits agency, payment systems, assessment, appeals and fraud-and-error control [M07-IND; S044, S048]. But it is systemic in its fiscal exposure, because it is the largest expenditure transfer and is bound to revenue (M01), borrowing for volatility (M02) and the state-pension apportionment strand (M06) [M07-IND caveat; M06-IND]. It crosses the administrative/systemic line and should be named as a straddler rather than forced into either group.

Resistant cases. M09 payments resists the cut altogether. It is a build-or-join resilience decision, not a pure scaling build and not a system-conditioning architecture in the way M04, M05 and M08 are. The France caution shows sovereignty alone does not resolve it: full sovereign monetary and central-bank allocation still leaves exposure to externally owned payment rails unless dedicated infrastructure is built or joined [M09-IND; S057, S053, S054]. M09 exposes the limit of a taxonomy built around scaling and system-conditioning, because its central question is a strategic build-or-join choice.

The stress test therefore returns a partial lens. A cut that cannot hold M07 and M09 without forcing them is a lens that mostly works and has named exceptions, not a finished classification. The partiality is the finding: the grid is being read, not bent to fit the categories. Any later use of the administrative / systemic / contested grouping should carry M07 and M09 as the cases that mark its edges.

(The contested group is treated in section 5 rather than here, because the single clear case, M06, is best read through the evidence-band pattern: its contestation is evidential, embedded in apportionment logic, not a feature of build type.)

5. What the evidence-band pattern says about what is known, buildable, and politically unresolved

The independence column has a distinctive pattern: M01 to M04 Medium, M05 High, M06 Contested, M07 to M11 Medium [00 compact index matrix; 03 coding state].

The dominant Medium pattern should not be read as weak evidence. Medium here means the Welsh starting position or diagnosis is directly evidenced, but the sovereign design is absent [00 evidence-band precedents]. The work is therefore buildable in principle but not yet specified in Welsh institutional form. This holds across the fiscal core (M01 to M04) and across the later mechanisms (M07 to M11): in each, the diagnosis or operational base is evidenced and the sovereign design is the gap [M01-IND; M02-IND; M03-IND; M04-IND; M07-IND; M08-IND; M09-IND; M10-IND; M11-IND]. The pattern maps to the known/buildable distinction: what is known is the Welsh position and the institutional requirements; what is unbuilt is the specific Welsh design, on paper, not in principle.

M05 is High because the monetary option space is extensively mapped internationally and in UK-adjacent evidence, even though Wales has not made a currency choice [M05-IND; S074, S075, S076, S077; cross-column note M05]. High records the maturity of the evidence about what the options require, not the existence of a Welsh design or a Welsh decision. The remaining gap at M05 is therefore a political choice and a transition design, not an absence of evidence: it sits in the politically-unresolved category rather than the not-yet-known one.

M06 is Contested because the allocation basis changes the evidential result itself. Population share, GDP share, historic contribution, geographic allocation and negotiated settlement produce materially different outcomes, and the pension strand is further complicated by the England-and-Wales scheme boundary [M06-IND; S027, S028, S029; cross-column note M06]. The uncertainty is not merely political; it is embedded in the apportionment logic. M06 is the one independence cell where the evidence supports the conclusion that a negotiated settlement would be required rather than a determinate estimate of what it would be.

The band pattern therefore separates three things cleanly. What is known: the Welsh diagnosis, the operational starting bases, and the institutional requirements of each mechanism. What is buildable but unspecified: the sovereign designs across the Medium cells, buildable in principle but absent in Welsh institutional form. What is politically or evidentially unresolved: the monetary choice at M05 (politically open, evidentially mapped) and the apportionment outcome at M06 (evidentially contested, settlement-dependent). The pattern is a read of the state of knowledge, not a ranking of options.

6. Boundary discipline and stopping line

The synthesis must not recommend a constitutional option.

It may say:

what each model makes possible
what each model leaves unresolved
what each model requires Wales to build
where the evidence is strong, medium, low or contested
where the design is absent
where the result depends on political choice or negotiated settlement

It must not say:

Wales should pursue enhanced devolution
Wales should pursue independence
Wales should reject one model
a particular party or institution should act next
a constitutional route is preferable because of the grid

The stopping line:

The synthesis explains what the grid shows about fiscal capacity, dependency and build requirements. It does not choose the constitutional future.

Annex D — Source log (78 sources)

Scope: S001 through S078. S036 and S037 intentionally unallocated. This source log is the reference target for the mechanism cells in Annex B and Appendix F, and for the footnotes in the companion policy paper.

S001 — Fairness and accountability: a new funding settlement for Wales. Independent Commission on Funding and Finance for Wales, Holtham Commission, July 2010. Scope: Enhanced devolution, fiscal repair, and independence diagnosis; Mechanisms: M01, M02, M03, M04. Primary source for the diagnosis of Barnett, the case for needs-based funding, tax devolution for accountability, and limited borrowing powers. For the independence cell, used as diagnosis only: relative need and fiscal-capacity asymmetry. The Holtham remedy (needs-based floor inside the block-grant system) does not travel to the independence design.

S002 — The agreement between the Welsh Government and the United Kingdom Government on the Welsh Government’s fiscal framework. Welsh Government and UK Government, 19 December 2016. Scope: Enhanced devolution, fiscal repair, and independence baseline contrast; Mechanisms: M01, M02, M03, M04. Operative implementation source for the repaired baseline. Adds machinery and limits to the Holtham diagnosis: needs-based factor, Welsh rates of income tax, block grant adjustment, Wales Reserve, capital borrowing, resource borrowing, and forecasting arrangements. For the independence cell, used as baseline-contrast source showing the UK territorial-funding machinery that independence would replace.

S003 — Welsh Government response to the UK Autumn Budget 2025. Welsh Government, 26 November 2025. Scope: Enhanced devolution and fiscal repair; Mechanisms: M01, M02. Current-position update. Records 10% uplift to Welsh capital borrowing and reserve limits in 2026-27, further increases with inflation from 2027-28, and unlimited access to the Wales Reserve in 2026-27. This updates the baseline flexibilities but does not alter the underlying competence structure.

S004 — Equalization Program. Government of Canada, Department of Finance Canada, Current official page checked June 2026. Scope: Federal or intermediate models; Mechanisms: M01, M04. Comparative source for a constitutionally entrenched federal equalisation principle based on provincial fiscal capacity and comparable public services at comparable taxation.

S005 — Approach to horizontal fiscal equalisation. Commonwealth Grants Commission, Australia, 2025 methodology material checked June 2026. Scope: Federal or intermediate models; Mechanisms: M01, M04. Comparative source for independent statutory advice on distributing GST revenue among states according to horizontal fiscal equalisation.

S006 — Basic Law for the Federal Republic of Germany, Article 107. Federal Republic of Germany, Current official text checked June 2026. Scope: Federal or intermediate models; Mechanisms: M01, M04. Comparative constitutional source showing financial equalisation among the Laender embedded in the federal finance architecture.

S007 — Intergovernmental fiscal transfers and fiscal equalisation in a time of consolidation, OECD Working Papers on Fiscal Federalism No. 50. OECD; Dougherty, Montes Nebreda and Urrutia, 30 June 2025. Scope: Federal or intermediate models; Mechanisms: M01, M04. Comparative design source for intergovernmental transfers and equalisation. Useful for design principles and safeguards, but not Wales-specific.

S008 — Australian Loan Council. Australian Government Directory, Department of the Treasury, Current official page checked June 2026. Scope: Federal or intermediate models; Mechanisms: M02, M04. Comparative source for shared scrutiny of public-sector borrowing. The current arrangements operate voluntarily and emphasise transparency, market scrutiny, and informed judgments about each government’s financial performance.

S009 — Basic Law for the Federal Republic of Germany, Articles 109 and 109a. Federal Republic of Germany, Current official text checked June 2026. Scope: Federal or intermediate models; Mechanisms: M02, M04. Comparative constitutional source showing budget autonomy for Federation and Laender combined with balanced-budget borrowing rules and Stability Council oversight.

S010 — Fiscal rules for subnational governments. OECD, 2021. Scope: Federal or intermediate models; Mechanisms: M02, M04, M06. Comparative design source for subnational fiscal rules, including budget balance objectives, borrowing and debt rules, expenditure limits, moral hazard, bailout risk, and market contagion.

S011 — How to Design Subnational Fiscal Rules: A Primer. International Monetary Fund, February 2020. Scope: Federal or intermediate models; Mechanisms: M02, M04, M06. Comparative design source for the rationale and limits of subnational borrowing, including investment finance, stabilisation limits, coordination costs, and macroeconomic risk.

S012 — Ontario’s Credit Rating: Fall 2025 Update. Financial Accountability Office of Ontario, 8 December 2025. Scope: Federal or intermediate models; Mechanisms: M02. Comparative example of a sub-state borrower with credit ratings, market access, liquidity management, fiscal flexibility, federal transfers, and debt-risk exposure.

S013 — Core Activities, Canada Revenue Agency ministerial transition briefing. Canada Revenue Agency, 9 September 2025. Scope: Federal or intermediate models; Mechanisms: M03, M04. Comparative source showing a federal revenue agency administering federal and most provincial and territorial taxes, with exceptions such as Quebec and Alberta corporate tax administration.

S014 — Federal-Provincial Fiscal Arrangements Act. Government of Canada, Current official text checked June 2026. Scope: Federal or intermediate models; Mechanisms: M03, M04. Comparative statutory source defining tax collection agreements under which the federal government collects provincial personal or corporate income taxes and makes payments to provinces.

S015 — Basic Law for the Federal Republic of Germany, Articles 105, 106 and 108. Federal Republic of Germany, Current official text checked June 2026. Scope: Federal or intermediate models; Mechanisms: M03, M04. Comparative constitutional source showing tax legislative powers, revenue apportionment and tax administration distributed through the federal finance constitution.

S016 — OECD Fiscal Decentralisation Database, tax autonomy indicators. OECD, Current dataset page checked June 2026. Scope: Federal or intermediate models; Mechanisms: M03, M04. Comparative source for subnational tax autonomy as a design variable, distinguishing control over tax rates, bases and revenue from mere receipt of funds.

S017 — Fiscal Federalism 2022, Chapter 3, Twenty years of tax autonomy across levels of government. OECD, 2022. Scope: Federal or intermediate models; Mechanisms: M03, M04, M06. Comparative design source on subnational tax autonomy across levels of government. Useful for the general design principle, not Wales-specific.

S018 — Government of Wales Act 2006, Schedule 7A, reserved matters. UK Parliament, Current official text checked June 2026. Scope: Baseline and enhanced devolution; Mechanisms: M05, M06, M09, M10, M11. Legal source for reservation of fiscal, economic and monetary policy, including the issue and circulation of money; reserves social security and occupational and personal pensions, the competence anchor for M06; reserves financial services and financial markets, the competence anchor for M09; the deposit-taking ceiling for M10; and the macroeconomic reservation constraining M11 productivity and economic structure.

S019 — Monetary policy. Bank of England, Current official page checked June 2026. Scope: Baseline and enhanced devolution; Mechanisms: M05. Official source for the UK monetary-policy function and the Bank of England’s role as the UK central bank.

S020 — Monetary policy remit, Budget 2025. HM Treasury, 26 November 2025. Scope: Baseline and enhanced devolution; Mechanisms: M05. Current remit source for the Monetary Policy Committee and the Treasury’s role in defining price stability under the Bank of England Act framework.

S021 — Federal Reserve Board, Monetary Policy. Federal Reserve Board, Current official page checked June 2026. Scope: Federal or intermediate models; Mechanisms: M05. Comparative federal source showing monetary policy located in the federal central bank. Useful for the adjacent federal principle, not Wales-specific.

S022 — Monetary policy. Bank of Canada, Current official page checked June 2026. Scope: Federal or intermediate models; Mechanisms: M05. Comparative federal source showing national monetary policy conducted by the central bank through an inflation-control framework. Useful for the adjacent federal principle, not Wales-specific.

S023 — UK Debt Management Office, homepage and Who we are. UK Debt Management Office, executive agency of HM Treasury, Current official pages checked June 2026. Scope: Baseline and enhanced devolution; Mechanisms: M06. Institutional location of UK sovereign debt management. The DMO is an executive agency of HM Treasury, legally and constitutionally part of HM Treasury, responsible for UK Government debt and cash management.

S024 — Senedd, Powers, reserved-powers model. Senedd Cymru, Current official page checked June 2026. Scope: Baseline and enhanced devolution; Mechanisms: M06. Reader-friendly explainer of the reserved-powers model; lists Schedule 7A headings including fiscal, economic and monetary policy, social security, and occupational and personal pensions. Companion to the Schedule 7A legal anchor S018.

S025 — Whole of Government Accounts 2023-24. HM Treasury, Year ended 31 March 2024; checked June 2026. Scope: Baseline and enhanced devolution; Mechanisms: M06. Consolidated view of the UK public sector’s financial position. Balance-sheet and liability context, not a Welsh allocation source.

S026 — Fiscal risks from public service pensions (box in Fiscal risks and sustainability report). Office for Budget Responsibility, July 2025. Scope: Baseline and enhanced devolution; Mechanisms: M06. Distinction between unfunded pension liabilities, a large government liability and mostly pay-as-you-go, and the funded LGPS, with different fiscal treatment. The box sits within the OBR’s July 2025 Fiscal risks and sustainability report and is based on ONS and OBR data from July 2025.

S027 — UK debt and the Scotland independence referendum. HM Treasury, 13 January 2014. Scope: Independence and inherited liabilities, separation-context; Mechanisms: M06. Continuing UK Government would honour UK-issued debt; an independent successor state would take on a fair and proportionate share of liabilities, with a separate contract and negotiated terms. Feeds the independence cell and informs federal design reasoning.

S028 — Scotland referendum 2014: the impact of independence on pensions. House of Commons Library, 2014 (Standard Note SN06914, last updated during referendum period). Scope: Independence and inherited liabilities; Mechanisms: M06. State pension rights accrued before independence and public-service pension liabilities would need apportionment between successor states.

S029 — The economic implications for the United Kingdom of Scottish independence, Chapter 4 on Scotland’s fiscal position. House of Lords Economic Affairs Committee, April 2013 (HL Paper 152). Scope: Independence and inherited liabilities; Mechanisms: M06. Recommended population-based apportionment for existing UK public-sector debt and known future liabilities such as public-service pensions; noted that transferring debt would not be straightforward.

S030 — Public service pensions: facts and figures (Briefing Paper CBP 8479). House of Commons Library, December 2019. Scope: Baseline and enhanced devolution; Mechanisms: M06. Distinguishes pay-as-you-go schemes from the funded LGPS; NHS and Teachers are England-and-Wales schemes; nationally set rules with local administration; Wales has powers only in relation to firefighters’ pensions within the 2013 Act framework.

S031 — The Public Sector in Wales: an analysis of recent trends in the public sector workforce and pay bill. Wales Fiscal Analysis, Cardiff University (Guto Ifan and Cian Sion), June 2019. Scope: Baseline and enhanced devolution; Mechanisms: M06. Pension contributions a source of pressure on the public-sector pay bill following UK employer-contribution increases; records Welsh Government funding to local authorities for pension contribution costs for teachers, firefighters and other public-sector employees.

S032 — Final Budget 2026-27. Welsh Government, 2026-27. Scope: Baseline and enhanced devolution; Mechanisms: M06. Records block grant adjustments, fiscal flexibilities, capital borrowing limits and Wales Reserve access. Broader fiscal-exposure context, not a pensions-specific source.

S033 — Draft Budget 2025-26 committee scrutiny. Senedd Finance Committee, 2024-25. Scope: Baseline and enhanced devolution; Mechanisms: M06. Senedd scrutiny of employer-cost pressures on Welsh-funded bodies; the cited passage concerns employer National Insurance rather than pensions, so it triangulates exposure logic only.

S034 — NHS Pension Scheme and Teachers’ Pensions employer contribution material. NHS Business Services Authority; Teachers’ Pensions, Checked June 2026. Scope: Baseline and enhanced devolution; Mechanisms: M06. NHS Pension Scheme employer contribution approach and the teachers’ employer contribution rate and administration levy. Scheme mechanics, not constitutional competence.

S035 — Local Government Pension Scheme funds for England and Wales statistics; Wales Pension Partnership. Ministry of Housing, Communities and Local Government; Wales Pension Partnership, LGPS statistics 2024-25 updated 13 May 2026; Wales Pension Partnership checked June 2026. Scope: Baseline and enhanced devolution; Mechanisms: M06. LGPS England-and-Wales employer contributions, scheme income and fund value; the Wales Pension Partnership identifies the eight Welsh LGPS funds as a Welsh investment pool. Keeps the funded LGPS distinct from the unfunded central schemes.

S036 — Intentionally unallocated.. n/a, n/a. Scope: n/a; Mechanisms: n/a. The OECD and IMF subnational-fiscal-rules source this would have covered is already logged as S010, S011 and S017.

S037 — Intentionally unallocated.. n/a, n/a. Scope: n/a; Mechanisms: n/a. The OECD and IMF subnational-fiscal-rules source this would have covered is already logged as S010, S011 and S017.

S038 — Supporting information for the Estimate of Income and Expenses 2024-25. Audit Wales, 2024-25. Scope: Baseline and enhanced devolution; Mechanisms: M06. A Welsh audited public body making provision for increased employer contributions to the Civil Service Pension Scheme. Narrow but clean example of employer pension-cost exposure inside a Welsh institutional budget.

S039 — Welsh Government oversight of further-education colleges’ finances and delivery. Auditor General for Wales, Wales Audit Office (now Audit Wales), 2017. Scope: Baseline and enhanced devolution; Mechanisms: M06. Records rising pension contributions and pension deficits as pressures on FE college finances, including large deficit increases after actuarial valuation. Supports the exposure mechanism rather than a current claim alone. Older report (financial analysis based on 2015/16 audited figures, funding information to 2016/17).

S040 — Health board allocations 2025 to 2026. Welsh Government, 2025-26. Scope: Baseline and enhanced devolution; Mechanisms: M06. Funding to cover the increased employer contribution for the NHS Pension Scheme held centrally in 2025-26; UK-level scheme contribution changes pass through Welsh health-budget machinery.

S041 — Supplementary Budget 2024-25. Welsh Government, 2024-25. Scope: Baseline and enhanced devolution; Mechanisms: M06. Explains SCAPE as the process for setting employer contributions in unfunded public-service pension schemes; records additional UK Government funding for increased SCAPE-related costs.

S042 — Provisional Local Government Settlement 2024-25: summary of responses. Welsh Government, 2024-25. Scope: Baseline and enhanced devolution; Mechanisms: M06. Records concern about permanent funding for teachers’ pension employer contributions and the desire to manage that funding through the settlement.

S043 — NHS Pension Scheme employer contribution rates 2025-26. NHS Business Services Authority, 2025-26. Scope: Baseline and enhanced devolution; Mechanisms: M06. Confirms the 2025-26 transitional employer contribution approach and the rate employers apply. Sits behind S040 rather than replacing Welsh-budget evidence.

S044 — Independent Commission on the Constitutional Future of Wales, final report, welfare benefits section. Independent Commission on the Constitutional Future of Wales, January 2024, updated July 2024. Scope: Baseline, enhanced devolution, federal or intermediate models; Mechanisms: M07, M08. Welsh-specific source for the split between limited devolved welfare benefits and reserved cash benefits; also source for the capacity warning that further devolution of benefits would require substantial tax and borrowing powers to carry risks and liabilities.

S045 — The Welsh Benefits System: what is it, and what are the latest developments?. Senedd Research, 24 January 2025. Scope: Baseline and enhanced devolution; Mechanisms: M07, M08. Source for Welsh Benefits System definition, current Welsh benefits, Welsh Government streamlining work, take-up concerns and administrative development.

S046 — Welsh Benefits Charter. Welsh Government, Streamlining Welsh Benefits Steering Group, 22 January 2024, last updated 20 August 2024. Scope: Baseline and enhanced devolution; Mechanisms: M07, M08. Official source for guiding principles for the design and roll-out of payments and grants devolved to the Welsh Government.

S047 — Streamlining Welsh benefits: route map phase 1, 2025 to 2026. Welsh Government, 14 January 2025. Scope: Baseline and enhanced devolution; Mechanisms: M07, M08. Official source for the Welsh Government’s current administrative streamlining programme for grants and payments governed by Welsh Government and delivered by local authorities and other organisations.

S048 — Implementation of the Scotland Act 2016: annual reports, social security section. Scottish Government, Ninth annual report 20 May 2025 (with eighth annual report figures). Scope: Enhanced devolution and federal or intermediate models; Mechanisms: M07, M08. Comparative source for partial social-security devolution, new agency creation, benefit transfer, delivery costs and interaction with reserved UK benefits. Records that eleven benefits were devolved under the Scotland Act 2016; that Social Security Scotland was established in 2018; that the agency employed over 3,800 full-time-equivalent staff as at 31 December 2023 (and 3,909 as at 30 September 2024) and delivered over £5 billion in benefit payments in 2023-24, of which £3.2 billion was delivered through DWP agency agreements and around £2 billion directly. Evidences the institutional cost and the parallel-delivery transition of building benefit-delivery capacity.

S049 — Welsh Government consolidated annual accounts 2024 to 2025. Welsh Government, 15 December 2025, page last updated 14 January 2026. Scope: Baseline and enhanced devolution; Mechanisms: M08. Official source for existing Welsh Government administrative machinery, governance, risk management, internal control, audit and accountability systems.

S050 — Welsh Government People Survey 2025. Welsh Government, 2 March 2026. Scope: Baseline and enhanced devolution; Mechanisms: M08. Official source for Welsh Government civil-service workforce engagement and organisational capacity evidence. Useful as capacity context, not as a constitutional capacity measure.

S051 — Payment Services Regulations 2017. UK Government, Current official text checked June 2026. Scope: Baseline and enhanced devolution; Mechanisms: M09. Operative UK regulatory framework for payment services; locates payment-institution authorisation and conduct requirements at UK level.

S052 — FCA payment-services authorisation and the Payment Systems Regulator. Financial Conduct Authority; Payment Systems Regulator, Current official pages checked June 2026. Scope: Baseline and enhanced devolution; Mechanisms: M09. The operative UK payment regulators: FCA authorisation and registration of payment institutions, and the Payment Systems Regulator’s oversight of payment systems.

S053 — Unified Payments Interface. National Payments Corporation of India, Current official material checked June 2026. Scope: Federal or intermediate models, and independence; Mechanisms: M09. Sovereign instant-payment infrastructure operating independently of the global card networks at very large scale. Feasibility precedent, not a Welsh source.

S054 — Pix instant payment system. Central Bank of Brazil, Operational since November 2020; checked June 2026. Scope: Federal or intermediate models, and independence; Mechanisms: M09. National instant-payment system built by the central bank, now a dominant payment method. Feasibility precedent for sovereign payment infrastructure.

S055 — Development Bank of Wales, institutional form and remit. Development Bank of Wales, Current official material checked June 2026. Scope: Baseline and enhanced devolution; Mechanisms: M09, M10. Strongest existing Welsh development-finance host: lending, guarantees, co-investment and de-risking. Not a deposit-taking bank, so it cannot host a public-sector payments mechanism or a deposit-funded Green Bank without restructuring or partnership.

S056 — Project Bank Accounts in Welsh public procurement. Welsh Government, Current official guidance checked June 2026. Scope: Baseline and enhanced devolution; Mechanisms: M09. Closest operational precedent for intervening in public-sector payment flows through procurement practice. Addresses poor and delayed payment and counterparty risk, not payment-rail exclusion or denial.

S057 — Report on card schemes and processors, with 2026 ECB payment-sovereignty material and the France/Cartes Bancaires case as live context. European Central Bank, 28 February 2025, with 2026 ECB payment-sovereignty material as live context. Scope: Cross-cutting mechanisms; Mechanisms: M09. Used for M09 as the caution that monetary sovereignty or full sovereign state capacity does not by itself remove dependence on externally owned payment rails. The ECB records the euro area’s reliance on international card schemes and frames that reliance as relevant to payment resilience and autonomy. The France/Cartes Bancaires case illustrates the same concern in a national setting, where a domestic scheme seeks to reduce dependence on Visa and Mastercard. Do not treat as Welsh evidence.

S058 — Economic development: what is devolved?. Law Wales, 10 March 2020, checked June 2026. Scope: Baseline and enhanced devolution; Mechanisms: M11. Source for the competence split: broad Welsh competence over economic development, with reservations around fiscal, economic and monetary policy, currency, financial services, financial markets, competition, import and export control, consumer protection, employment and job-search functions.

S059 — Economic Growth Wales. Office for Statistics Regulation, 10 March 2026. Scope: Baseline; Mechanisms: M04, M11. Source for Welsh economic measurement constraints: Welsh growth is measured through ONS regional GDP and GVA data, with no Welsh monthly GDP statistics and limits around timeliness, volatility and revisions.

S060 — Sub-regional productivity: 2002 to 2023. Welsh Government, 20 June 2025. Scope: Baseline; Mechanisms: M11. Official Welsh source for the current productivity diagnosis: Wales at 84.9% of UK GVA per hour worked in 2023, lowest of the twelve UK countries and English regions; GVA per job filled at 82.3% of UK figure.

S061 — Wales Economic and Fiscal Report 2025. Welsh Government, 14 October 2025, updated 3 November 2025. Scope: Baseline and enhanced devolution; Mechanisms: M11. Official Welsh source for productivity and economic structure: low relative productivity, stronger 2009 to 2023 productivity growth than the UK as a whole, post-2019 relative weakening, production-sector exposure and Brexit-linked vulnerability.

S062 — Wales’ Productivity Challenge: A Focus on the Future. The Productivity Institute, Melanie Jones, January 2025. Scope: Baseline, enhanced devolution and federal or intermediate models; Mechanisms: M11. Expert source for the long-term productivity challenge, absence of convergence, drivers of the gap and need for long-term coordinated strategy across government, firms, workers, regions and public services.

S063 — New economic mission to close Wales’ productivity gap with UK and boost pay. Welsh Government, 2 June 2026. Scope: Enhanced devolution; Mechanisms: M11. Live policy source for Welsh Government’s productivity mission: goal to halve the productivity gap with the rest of the UK within ten years, make Wales easier for businesses to start, grow and invest, and create a new agency to drive economic development.

S064 — Devolution, independence and Wales’ fiscal deficit. Ifan and Sion, National Institute Economic Review, 2022. Scope: Cross-cutting and independence phase; Mechanisms: M01, M02, M03, M04, M11. Academic source linking Wales’ fiscal position to long-standing economic underperformance, primary-income weakness and lack of convergence under current arrangements. Records that the spending apportioned to Wales in the net-fiscal-balance figure includes non-identifiable spending, a population share of spending incurred on behalf of the UK as a whole rather than a specific region, such as defence and interest on the UK national debt. Notes that sustaining the apportioned per-capita share of UK defence spending would imply around 2.6 per cent of Welsh GDP on defence, illustrating that an apportioned UK cost is not the same as the budget choice an independent state would make; and that the apportioned national-debt interest is a share of existing UK-wide debt service rather than a settled liability of an independent Wales. Use carefully: relevant to cross-cutting interpretation and independence phase, not a baseline verdict on constitutional choice.

S065 — Wales’ Fiscal Future: A path to sustainability?. Wales Fiscal Analysis, Cardiff University (Guto Ifan, Cian Sion and Ed Gareth Poole), March 2020. Scope: Independence and cross-cutting; Mechanisms: M01, M02, M03, M04. Core Welsh fiscal baseline and sustainability source for the independence fiscal-capacity question. Revenue, expenditure, fiscal deficit and sustainability analysis, including closing the fiscal gap both as part of the UK and as an independent country.

S066 — Public Sector Expenditure and Revenue Analysis, Wales. Wales Office, drawing on Office for National Statistics country and regional public sector finance data; classified as an ONS official statistic, Most recent edition released June 2024 (financial year ending 2023). Scope: Independence; Mechanisms: M01, M02, M03, M04. Current official statistical anchor for Welsh public-sector revenues, expenditures and net fiscal balance. The June 2024 edition records a net fiscal deficit for Wales of over £21.5 billion in 2022-23. Caveat: this is a notional estimate of the current fiscal balance under existing UK spending allocations, not a forecast of an independent Wales’s fiscal position. The spending side apportions to Wales a population share of UK-wide costs that a sovereign Wales would not automatically reproduce, notably defence and a share of UK national-debt interest (see S064). Note: the underlying ONS country and regional public sector finances release is moving from annual to biennial.

S067 — Recent changes and the future outlook for Welsh Government funding. Institute for Fiscal Studies, 24 February 2026. Scope: Independence; Mechanisms: M01, M02, M03, M04. Block-grant dependence and current funding composition. Source for the 2025-26 £23.1bn day-to-day funding figure, the 79% UK Government and block-grant share, and the 21% devolved-tax share used in the M01-IND evidence narrative.

S068 — Revised Guidelines for Public Debt Management. International Monetary Fund and World Bank, April 2014. Scope: Independence and cross-cutting; Mechanisms: M02, M04. Authoritative international framework for sovereign debt management: objectives, institutional framework, strategy, risk management, cash buffers, market development. Source for the institutional capacity a sovereign debt-management function requires, and for fiscal data quality as a condition of borrowing credibility (M04-IND). Not Welsh-specific.

S069 — Building a New Scotland: A stronger economy with independence, borrowing and fiscal framework section. Scottish Government, October 2022. Scope: Independence; Mechanisms: M02. Closest UK sub-state comparator for transition from bounded devolved borrowing to sovereign borrowing: sovereign debt issuance, fiscal rules, Debt Management Office, credit establishment, investor credibility. Not Welsh evidence; Scottish comparator only.

S070 — Welsh Revenue Authority Annual Report and Accounts 2024 to 2025. Welsh Revenue Authority, Welsh Government, 23 July 2025. Scope: Independence; Mechanisms: M03, M04. Institutional source for current Welsh tax-administration capacity. The WRA manages Land Transaction Tax and Landfill Disposals Tax, has more than 100 staff, and operates as a real but narrow devolved tax authority. Also feeds M04-IND as current Welsh tax-data capacity.

S071 — Welsh Rates of Income Tax Annual Report 2025. HM Revenue and Customs, 11 September 2025. Scope: Independence; Mechanisms: M03, M04. Source for Welsh Rates of Income Tax being set for Wales but administered through HMRC, with 2023 to 2024 outturn and taxpayer numbers. Supports the distinction between Welsh rate-setting and UK collection machinery. Also feeds M04-IND as current Welsh tax-data capacity.

S072 — Taxing an independent Scotland. Stuart Adam, Oxford Review of Economic Policy, 2014. Scope: Independence comparator; Mechanisms: M03. Comparator source on the tax-system design consequences of independence for a smaller nation. Scottish comparator only, not Welsh evidence.

S073 — Fiscal Transparency Code. International Monetary Fund, 2019. Scope: Independence and cross-cutting; Mechanisms: M04. International standard for fiscal transparency: fiscal reporting, fiscal forecasting and budgeting, fiscal risk analysis and management, and resource revenue management. Source for the sovereign fiscal-transparency standard in M04-IND. Not Welsh-specific.

S074 — Fiscal Commission Working Group, First Report: Macroeconomic Framework, including Annex: Assessment of Key Currency Options. Fiscal Commission Working Group, Scottish Government, February 2013. Scope: Independence comparator; Mechanisms: M05. Scottish comparator for sovereign currency options. Examined four options (formal sterling union, sterlingisation, euro, new currency) and recommended retaining sterling in a formal monetary union. Strongest analytical Scottish comparator for the currency-options question. Not Welsh evidence; Scottish comparator only.

S075 — Building a New Scotland: A stronger economy with independence, currency and financial stability section. Scottish Government, 17 October 2022. Scope: Independence comparator; Mechanisms: M05. Later Scottish policy comparator for the currency question. Proposes continued sterling use from independence day, followed by transition to a Scottish pound guided by criteria rather than a fixed timetable, with an independent Scottish Central Bank established during the sterling phase. Same document as S069 (borrowing section) but used for a different analytical purpose. Not Welsh evidence; Scottish comparator only.

S076 — Monetary Policy and Central Banking. International Monetary Fund, Current factsheet checked June 2026. Scope: Independence and cross-cutting; Mechanisms: M05. International institutional source for the role of central banks in monetary policy, financial stability, macroprudential policy, and the IMF’s role in promoting effective central bank frameworks. Source for the general institutional requirements of sovereign monetary-policy capacity. Not Welsh-specific.

S077 — Re-thinking the lender of last resort. Bank for International Settlements, BIS Papers No. 79, 2014. Scope: Independence and cross-cutting; Mechanisms: M05. International institutional source for the lender-of-last-resort function as a core central bank responsibility: the unique ability to create liquid assets, the central bank’s central position within the payment system, the macroeconomic stabilisation objective, and the design of LOLR frameworks including governance, operational arrangements and international dimensions. Not Welsh-specific.

S078 — Benefit expenditure and caseload tables 2026. Department for Work and Pensions, Published 14 April 2026. Scope: Independence; Mechanisms: M07, M06. Reserved-system scale for social security and pensions: historical and forecast expenditure and caseload data for the UK benefit system that an independent Wales would need either to replace, negotiate transition from, or interface with. Outturn and forecast tables are consistent with the OBR Economic and fiscal outlook of March 2026. Source for the M07-IND reserved-system-scale point; also relevant to the M06 state-pension strand. Not Welsh-specific.

Note on source numbering: S036 and S037 are intentionally unallocated. The OECD and IMF subnational-fiscal-rules sources they would have covered are already logged as S010, S011 and S017, which carry M06 in their mechanism tags and are reused for the M06 federal cell.

Source-log consistency pass complete: S057 producer confirmed; S026, S030, S031, S039, S065 and S066 dates confirmed. No “Date to confirm” or producer-to-confirm flags remain.

Appendix F — Federal column archive

The federal column is archived, not part of the publication spine. It is retained here in full for analytical completeness and as the resolution target for the “Federal variant” pointers in the Annex B dossiers. The companion policy paper carries only a short note explaining why federalism was archived (low Welsh-specific evidence, high institutional distance, low present political plausibility); the full federal cells and federal-specific interpretation notes are held here.

Status note

ARCHIVED from the publication spine.
Retained for analytical completeness.
NOT part of the main publication path unless deliberately restored later by explicit instruction.
Analytical rationale for archiving: the federal column codes Low across all eleven mechanisms; its evidence rests on comparative inference from other federations rather than Welsh-specific evidence; and UK federation is not a realistic near-term scenario.

Scope: M01-FED through M11-FED full cells, plus the federal-specific interpretation notes for M01-FED through M04-FED. No whole-mechanism cross-column notes are held here; those are placed at the head of their mechanism dossiers in Annex B.

M01-FED. Federal UK

Competence: Design-dependent. A federal UK would require a formal federal finance settlement rather than simple continuation of the present Barnett mechanism. A needs-based or fiscal-capacity-based territorial funding system could be part of that design, but the competence would sit in the federal constitution, federal legislation, or shared intergovernmental finance rules rather than in unilateral UK Treasury discretion.

Capacity: Build-required. Wales would need capacity to participate in, scrutinise, and challenge a rule-based equalisation system. The relevant capacity is not merely receipt of funds, but measurement capacity, fiscal data capacity, institutional capacity, and dispute capacity.

Evidence strength: Low.

Sources: S004, S005, S006, S007.

Caveat: The comparative principle is strongly supported by adjacent federal models, but the Welsh application is inferential. Canada shows constitutional equalisation by fiscal capacity; Australia shows independent allocation advice under horizontal fiscal equalisation; Germany shows financial equalisation inside constitutional finance architecture; the OECD supports the general design concern. None supplies a Welsh federal formula. The cell therefore records new architecture required, not an evidenced Welsh funding outcome.

Interpretation note (M01-FED)

The federal column should not treat the Holtham floor as though it automatically becomes federal equalisation. The floor can be read as a design precursor because it introduces needs-based protection into the repaired baseline. But federal equalisation would require a different institutional form: codified rules, shared measurement, agreed fiscal indicators, dispute management, and protection from unilateral adjustment.

The finding is neither that federalism improves the Welsh funding position nor that it weakens it. The finding is that federalism changes the question. Under repaired devolution, Wales receives needs-based protection inside a UK-controlled block grant system. Under a federal UK, territorial funding would need to become part of a shared-rule finance constitution, with Wales participating as a constituent unit rather than only as a devolved recipient.

M02-FED. Federal UK

Competence: Design-dependent. A federal UK could allocate borrowing powers to Wales as a constituent unit, but the scope would depend on the federal finance constitution: whether Wales could borrow independently, whether borrowing would be subject to shared fiscal rules, whether there would be aggregate borrowing coordination, and whether any borrowing would be limited to capital investment, forecast error, stabilisation, or wider fiscal autonomy.

Capacity: Build-required. Wales would need debt-management capacity, liquidity management, forecasting capacity, investor-relations capacity, credit-risk management, and the institutional ability to operate within shared fiscal rules.

Evidence strength: Low.

Sources: S008, S009, S010, S011, S012.

Caveat: The comparative principle is strongly supported by adjacent federal models, but the Welsh application is inferential. Australia shows shared transparency and market-scrutiny machinery through Loan Council arrangements; Germany shows budget autonomy combined with constitutional borrowing rules and Stability Council oversight; OECD and IMF sources show why subnational borrowing is usually rule-bound because of bailout, contagion, coordination, and macroeconomic risks; Ontario shows that a sub-state borrower’s market access depends on credit standing, liquidity, debt management, transfers, and fiscal flexibility. None supplies a Welsh federal borrowing settlement. The cell therefore records borrowing as new shared-rule fiscal architecture, not simply larger borrowing headroom.

Interpretation note (M02-FED)

The federal column should not treat the present Welsh borrowing framework as though it merely scales upward. The repaired baseline gives Wales bounded borrowing and reserve tools inside a UK fiscal framework. A federal UK would change the institutional question: borrowing would need to be allocated between the federal state and the Welsh constituent unit, with rules for fiscal discipline, market access, reporting, debt sustainability, and macroeconomic coordination.

The finding is neither that federalism gives Wales unconstrained borrowing power nor that it leaves Wales in the present position. The finding is that federal borrowing capacity is form-dependent. A federal settlement could give Wales more meaningful borrowing competence, but only inside a shared fiscal-risk architecture. The added competence would therefore create matching obligations: debt management, credit credibility, liquidity protection, forecasting, and compliance with shared rules.

M03-FED. Federal UK

Competence: Design-dependent. A federal UK could allocate tax powers to Wales in several ways: exclusive Welsh taxes, shared tax bases, assigned revenues, tax-rate variation, surcharges, or jointly administered taxes. The scope would depend on the federal finance constitution and on which taxes remained federal, which became Welsh, and which were shared.

Capacity: Build-required. Wales would need tax-policy capacity, forecasting capacity, tax-base measurement, behavioural-response modelling, compliance oversight, data access, revenue-risk management, and the ability to operate either its own tax administration or a shared administration agreement.

Evidence strength: Low.

Sources: S013, S014, S015, S016, S017.

Caveat: The comparative principle is strongly supported by adjacent federal models, but the Welsh application is inferential. Canada shows that provincial tax powers can coexist with federal collection agreements and federal administration for most provinces and territories. Germany shows tax legislation, revenue apportionment and administration embedded in the constitutional finance architecture. OECD sources show that tax autonomy varies by design and must be distinguished from revenue assignment. None supplies a Welsh federal tax settlement. The cell therefore records shared tax-base architecture required, not simply more devolved tax variation.

Interpretation note (M03-FED)

The federal column should not treat the present devolved-tax settlement as though it simply scales upward. The repaired baseline gives Wales some devolved and partly devolved taxes, with corresponding administrative and measurement capacity, but key tax architecture remains UK-controlled.

A federal UK would change the institutional question. Tax powers would need to be allocated across federal and Welsh levels, and the settlement would have to decide which level controls tax bases, rates, collection, forecasting, data, compliance and revenue risk. A shared collection model could reduce the need for a wholly separate Welsh revenue agency, but it would not remove the need for Welsh tax-policy, forecasting and scrutiny capacity.

The finding is neither that federalism gives Wales full tax autonomy nor that it leaves the present settlement unchanged. The finding is that federal tax capacity is form-dependent. A federal settlement could give Wales more meaningful tax competence, but only through a shared tax-base and tax-administration architecture. Added competence would create matching obligations: revenue forecasting, tax-base measurement, behavioural modelling, compliance capacity and accountability for revenue risk.

M04-FED. Federal UK

Competence: Design-dependent. A federal UK would need to allocate authority over fiscal data, tax-base measurement, revenue forecasting, equalisation indicators, debt reporting, budget surveillance and intergovernmental fiscal statistics. The competence could sit in a federal statistical or fiscal body, shared intergovernmental machinery, independent advisory institutions, Welsh institutions, or some combination of these.

Capacity: Build-required. Wales would need the ability to measure its tax base, forecast devolved or federalised revenues, scrutinise equalisation formulae, assess borrowing and reserve positions, model behavioural effects, understand fiscal transfers, and contest or verify data used in shared-rule fiscal machinery.

Evidence strength: Low.

Sources: S004 to S017 (already logged for M01-FED, M02-FED and M03-FED; each carries M04 as a secondary mechanism). No new source log entries required.

Caveat: The comparative principle is strongly supported by the adjacent federal sources already logged, but the Welsh application is inferential. Canada shows equalisation depending on fiscal-capacity measurement; Australia shows independent allocation advice under horizontal fiscal equalisation; Germany shows budget surveillance and financial architecture embedded in federal institutions; OECD sources show that tax autonomy and fiscal decentralisation require comparable indicators. None supplies a Welsh federal instrumentation model. The cell therefore records fiscal instrumentation as a precondition of federal fiscal capacity, not as an evidenced Welsh institutional settlement.

Interpretation note (M04-FED)

M04-FED is the dependency row for the first three federal mechanisms. Territorial funding, borrowing and taxation cannot operate as federal powers unless the system can measure the fiscal facts on which they rest.

Under repaired devolution, Welsh fiscal measurement expands where instruments are devolved: devolved taxes create forecasting, reconciliation, block grant adjustment and tax-administration requirements. Under a federal UK, the measurement problem would change scale and form. Wales would not only need to measure its own devolved instruments. It would need capacity to participate in shared-rule fiscal machinery: equalisation, borrowing surveillance, debt reporting, tax-base comparison, transfer calculation and dispute resolution.

The finding is not that federalism automatically gives Wales better fiscal data. The finding is that federal fiscal competence would be hollow without instrumentation. A federal Wales would need measurement capacity as a condition of meaningful participation in the finance constitution. That capacity is build-required even where the formal design remains unsettled.

This keeps M04-FED aligned with M01 to M03, but makes its role explicit: M04 is the precondition row, not the afterthought row.

M05-FED. Federal UK

Competence: Design-dependent, but likely federal-central in any conventional federal model. A federal UK could retain a single currency and central bank, with monetary policy conducted at federal level rather than by Wales as a constituent unit.

Capacity: Build-limited. Wales would need macroeconomic analysis and scrutiny capacity to understand and challenge federal monetary effects, but not full central-bank capacity unless the federal design gave constituent units an unusual monetary role.

Evidence strength: Low.

Sources: S021, S022.

Caveat: The adjacent federal principle is strongly supported by comparative examples, but the Welsh application is inferential. The Federal Reserve and Bank of Canada show monetary policy located in federal or national central-bank institutions. They do not supply a Welsh federal monetary settlement. The cell therefore records currency and monetary policy as a federal-central function in the ordinary design case, not as an upward scaling of Welsh fiscal devolution.

M06-FED. Federal UK

Debt. Competence: Design-dependent. Ongoing sovereign debt management would likely remain a federal-central function, but a federal settlement would have to decide whether and how to apportion the inherited debt stock at formation. Capacity: ongoing management Build-limited; apportionment of legacy debt Build-required.

Inherited liabilities. Competence: Design-dependent, and this is the strand most likely to require new architecture: federalisation would need explicit allocation and assumption rules for legacy liabilities. Capacity: Build-required.

Pensions. Competence: Design-dependent. A federal settlement could leave the schemes federal with Welsh scrutiny, or reassign scheme responsibility and apportion accrued rights. Capacity: stranded, Build-limited if schemes stay federal, Build-required if accrued rights are apportioned.

Evidence strength: Low. The Welsh-specific federal evidence is thinner than for M01 to M04, because the comparative material is separation-context rather than federal-formation. The inherited-liabilities apportionment basis is contested or basis-sensitive, since results depend on the chosen basis of population, spending or territoriality, but the band is recorded as Low because the weakness is the absence of Welsh federal-formation evidence, not only disagreement.

Sources: comparative design reasoning S010, S011, S017 (reused); apportionment logic S027, S028, S029 (separation-context).

Caveat: The federal cell is itself stranded and the most fragile in the row. Ongoing debt management and scheme operation could remain federal with only scrutiny capacity required (Build-limited), but the binding federal problem is apportionment of the inherited stock, legacy debt, inherited liabilities and accrued pension rights, which is new allocation and assumption architecture (Build-required). That is why the dominant capacity token is Build-required, not Build-limited, and not Design-dependent.

M07-FED. Federal UK

Competence: Design-dependent. A federal UK could leave social security primarily federal, share it, or allocate parts to Wales as a constituent unit. The Commission’s own federal definition assumes social security policy would remain reserved in its federal model, so any broader Welsh welfare competence would be a design choice rather than an automatic feature of federalism.

Capacity: Build-required. If social security remained federal, Wales would still need scrutiny, data and intergovernmental capacity to manage the interaction between federal benefits and devolved services. If parts were allocated to Wales, it would need the same administrative and fiscal-risk capacity required under enhanced devolution, at larger or more formal scale.

Evidence strength: Low.

Sources: S044, S048.

Caveat: The Welsh federal application is thin. The Commission points toward social security remaining federal in its federal model, while Scotland and Northern Ireland show that partial or fuller devolution of benefits is constitutionally possible inside the UK but operationally complex. No source supplies a Welsh federal social-security formula. The cell therefore records the issue as design-dependent and capacity-building-heavy, not as a federal automatic.

M08-FED. Federal UK

Competence: Design-dependent. A federal UK would require Welsh state capacity as a constituent unit, but the required functions would depend on the federal design: which powers are federal, which are Welsh, which are shared, and how intergovernmental machinery operates.

Capacity: Build-required. Wales would need the capacity to participate in federal institutions, scrutinise shared rules, manage intergovernmental disputes, operate devolved or federalised functions, handle fiscal data, and coordinate across Welsh, federal and local systems.

Evidence strength: Low.

Sources: S044, and the federal dependency recorded at M01-FED to M04-FED.

Caveat: The comparative and Commission evidence supports the general capacity requirement, but not a Welsh federal state-capacity model. M01 to M04 already show the dependency: federal territorial funding, borrowing, tax and fiscal measurement all require Welsh capacity to engage with shared-rule machinery. This cell records federal capacity as a build requirement, not as an evidenced Welsh institutional settlement.

M09-FED. Federal UK

Competence: Design-dependent. A federal settlement would have to decide where payment-services competence sits: it could remain reserved, be shared, or be allocated through joint institutions. It should not be assumed to follow automatically from federalisation.

Capacity: Build-required. If competence were allocated to or shared with Wales, the capacity to operate or co-govern a payment system would have to be built. India’s UPI and Brazil’s Pix show that sovereign payment infrastructure outside the global card networks is technically feasible at scale, though both were built by states considerably larger than Wales, and feasibility is not the same as allocation.

Evidence strength: Low.

Sources: competence baseline S018, S051, S052; feasibility precedents S053, S054.

Caveat: The Welsh federal application is thin and rests on institutional reasoning rather than Welsh-specific evidence. Federalisation makes payment competence an allocation question, not an automatic transfer. The feasibility precedents show a sovereign system can be built but not that a federal Wales would hold or need one.

M10-FED. Federal UK

Competence: Design-dependent. A federal settlement could allocate financial-services regulation differently, which would affect whether a deposit-taking Welsh institution requires federal or Welsh-level authorisation. The development-finance form would likely remain within Welsh competence under any conventional federal design.

Capacity: Build-required. If deposit-taking competence were allocated to or shared with Wales, the capacity to operate a deposit-taking Green public bank would have to be built: regulatory compliance, prudential structure, governance, capital adequacy, liquidity management and depositor protection. The development-finance form would require the same build as under enhanced devolution.

Evidence strength: Low.

Sources: S055 for baseline Welsh capacity; institutional reasoning for the federal design question.

Caveat: The Welsh federal application is thin and rests on institutional reasoning. The development-finance Green Bank form does not depend on federalisation; it is available within the present settlement. The federal question is therefore narrow: does federalisation change the deposit-taking constraint? That depends on the federal design of financial-services regulation, which is not evidenced for Wales.

M11-FED. Federal UK

Competence: Design-dependent. A federal UK could allocate more economic-development, tax, borrowing, infrastructure, skills, labour-market or investment powers to Wales, or retain key levers federally. The productivity effect would depend on the federal economic constitution and the allocation of macroeconomic, fiscal, market, labour and investment powers.

Capacity: Build-required. Wales would need the same productivity-state capacity required under enhanced devolution, plus federal participation capacity: shared-rule institutions, fiscal equalisation engagement, intergovernmental economic strategy, federal funding negotiation, labour-market coordination, investment governance and economic data capacity.

Evidence strength: Low.

Sources: S062 for the productivity challenge; institutional reasoning for the federal design question.

Caveat: The evidence for the Welsh federal application is thin. The federal column changes the allocation question but does not itself solve the productivity problem. The grid has already shown that federal funding, borrowing, taxation, fiscal measurement, welfare, state capacity, payment resilience and public-finance vehicles would all require new shared-rule architecture. M11 records that productivity would sit across those systems rather than follow automatically from federal form.

See Evidence Annex, Annex A, for the competence, capacity and evidence-band coding method. ↩︎
See Evidence Annex, Annex B, M01-B through M11-B. ↩︎
See Evidence Annex, Annex B, M01-B; Annex D, S001, S002. ↩︎
See Evidence Annex, Annex B, M02-B; Annex D, S002, S003. ↩︎
See Evidence Annex, Annex B, M03-B; Annex D, S070, S071, S067. ↩︎
See Evidence Annex, Annex B, M07-B; Annex D, S044, S045, S046, S047. ↩︎
See Evidence Annex, Annex B, M10-B; Annex D, S055, S018. ↩︎
See Evidence Annex, Annex B, M09-B; Annex D, S055, S056, S051, S052, S018. ↩︎
See Evidence Annex, Annex B, M06-B; Annex D, S018, S023, S030, S031, S035. ↩︎
See Evidence Annex, Annex B, M01-ED through M11-ED. ↩︎
See Evidence Annex, Annex B, M01-ED, M02-ED, M03-ED; Annex D, S001, S002. ↩︎
See Evidence Annex, Annex B, M07-ED; Annex D, S044, S045, S046, S047. ↩︎
See Evidence Annex, Annex B, M07-ED, M08-ED; Annex D, S044, S048. ↩︎
See Evidence Annex, Annex B, M09-ED, M10-ED, M05-ED; Annex D, S055, S018, S051, S052. ↩︎
See Evidence Annex, Annex B, M01-IND through M11-IND. ↩︎
See Evidence Annex, Annex B, M01-IND; Annex D, S064, S065, S066, S067. ↩︎
See Evidence Annex, Annex B, M02-IND; Annex D, S068, S069. ↩︎
See Evidence Annex, Annex B, M03-IND, M04-IND; Annex D, S070, S071, S073. ↩︎
See Evidence Annex, Annex B, M07-IND, M08-IND; Annex D, S044, S048, S049, S050. ↩︎
See Evidence Annex, Annex B, M05-IND; Annex D, S074, S075. ↩︎
See Evidence Annex, Annex B, M09-IND, M10-IND; Annex D, S055, S057, S018. ↩︎
See Evidence Annex, Annex B, M06-IND; Annex D, S027, S028, S029. ↩︎
See Evidence Annex, Annex A; Annex B, M01-IND through M11-IND. ↩︎
See Evidence Annex, Annex C; Annex A. ↩︎
See Evidence Annex, Appendix F. ↩︎
See the companion document, Comparative Fiscal Capacity: Evidence Annex. ↩︎

Findings at a glance

1. The question this paper asks

2. The current settlement: a partial state

3. Enhanced devolution: the build question

4. Independence: the change of form

The comparison by mechanism

5. What the comparison shows

6. The federal model

7. What this paper does and does not settle

Evidence Annex

A.1 Column codes

A.2 Status codes

A.3 Mechanism index

A.4 Compact index matrix

A.5 Use rule

Annex B — Mechanism dossiers

1. What the grid shows under the current settlement

2. What enhanced devolution would actually change

3. What independence converts from dependency into build-requirement

4. Administrative, systemic, contested: a stress test, not a taxonomy

5. What the evidence-band pattern says about what is known, buildable, and politically unresolved

6. Boundary discipline and stopping line

Status note

Comments