A Land Value Tax for Wales? The Welsh Government Valuation-Testing Programme (Goodwin-Hawkins/CCRI, 2026)

Overview

In February 2026 the Welsh Government published two working papers by Dr Bryonny Goodwin-Hawkins of the Countryside and Community Research Institute (CCRI, a collaboration between the University of Gloucestershire, the Royal Agricultural University, and Hartpury College), written during a UKRI Policy Fellowship on Welsh local government finance reform: A Land Value Tax for Wales? Claims and contexts (Working Paper, version 3.0, dated January 2026 in the Welsh Government copy) and its companion, A Land Value Tax for Wales? Valuation methodologies (Working Paper 2, version 1.2, February 2026).[1][2] Both are mirrored in full in the University of Gloucestershire's eprints repository; the Welsh Government's own site also hosts a four-page executive summary of the valuation paper.[3] This page is the wiki's dedicated treatment of both — the deep home the wiki's Andy Burnham and TPA "What Would an LVT Actually Do?" pages point to when they mention this programme, per the wiki's one-finding-one-home rule.

What the programme is not, and this matters: despite the Welsh Government's own publication title — "Testing Land Valuation Methods for a Land Value Tax" — neither paper reports the results of an actual field test of any valuation method against Welsh data. Both are desk-based evidence reviews and scoping documents: the first synthesises the international academic literature on LVT's claimed effects, the second identifies five families of valuation method that the author recommends be tested in a future phase. See "What the Programme Actually Found" below for why this distinction matters for how the wiki — and other sources citing this programme — should describe it.

Origins: Wales's Local-Government-Finance-Reform Programme

Welsh interest in LVT is not a 2026 invention. It sits inside a Welsh Government programme on local government finance reform that Goodwin-Hawkins's own paper dates to 2017, prompted by concerns about the fairness and resilience of Council Tax (still banded on 2003 valuations, one of the few respects in which Wales's position is less stale than England's, whose bands still run on 1991 values — see Andy Burnham) and Non-Domestic Rates.[1] A 2021 Welsh Government synthesis of that programme's early findings judged LVT "a credible long-term option" but said further work was needed on data infrastructure, governance, and delivery pathways before any policy decision.[2] An earlier 2019 Welsh Government-commissioned technical assessment (Ap Gwilym et al., A technical assessment of the potential for a local Land Value Tax in Wales) had already scoped the idea, cited in Goodwin-Hawkins's own reference list.[1] Separately, in 2018 Wales selected a vacant land tax as a candidate for its first entirely new devolved tax, aimed at encouraging development of unused land — a narrower, Georgist-adjacent proposal that stalled in negotiations with HM Treasury and remained at an impasse as of 2024.[1]

The programme became publicly live again in mid-2025: on 11 June 2025 the Welsh Government's finance secretary, Mark Drakeford, told the Senedd that LVT remained under active consideration as a long-term replacement for council tax and business rates, citing Denmark and Singapore as comparators, and announced that the Welsh Government had invited tenders to test approaches to valuing land, covering "every aspect of the work."[4] Drakeford framed the near-term goal explicitly as narrowing the gap between theory and feasibility: "I am anxious to see this discussion move beyond the theoretical and into the realms of the practically possible... I want to use the coming months to test the boundaries of what might be possible in the next Senedd term."[4] The two Goodwin-Hawkins working papers, delivered via a UKRI Fellowship rather than directly through that tender process, are the first published outputs of this renewed push — evidence-review and methodology-scoping, ahead of whatever field testing the tendered work eventually produces.

The Claims and Contexts Paper: A Systematic Review, Not a Verdict

The first paper undertakes what it calls "an adapted systematic literature review" of international, high- and upper-middle-income-country studies published 1999–2024, screening down to 70 studies meeting its empirical-evidence inclusion criteria (observed outcomes, evaluations of actual LVT implementation, or simulations calibrated to real data), assessed with a weight-of-evidence framework.[1] It tests seven commonly cited claims for LVT against that evidence base — three efficiency claims (reduces distortionary effects on investment; promotes efficient land use; incentivises public investment) and four equity claims (discourages land speculation; captures unearned gains; is more progressive; reduces socio-economic inequality).[1]

Headline finding: "many claims made for LVT are plausible in theory but not strongly supported by direct empirical evidence. Where effects are observed, these are typically localised and context-dependent... and appear to depend more on tax rates, valuation, planning systems, and market conditions than on the tax base in land itself."[1] Broken down:

• Efficiency. Claims that LVT reduces distortion on investment are the "most frequently cited" but "the evidence base is limited" — most studies assume non-distortionary land taxation rather than testing it. There is comparatively stronger evidence that LVT encourages development of under-used land specifically in already-viable, high-demand urban locations; the review found no support for LVT generating development where underlying demand is low. Claims that LVT incentivises public investment are "the weakest empirically" — there is no robust evidence LVT causes governments to invest more, only that investment raises land values.[1]
• Equity. The claim that LVT discourages speculation is "plausible but weakly demonstrated." There is "some support" that LVT shifts tax burdens toward higher-value land under revenue-neutral reforms, which is formally more progressive — but evidence that this translates into reduced socio-economic inequality is "limited," since most studies measure changes in tax liability rather than broader welfare outcomes.[1]
• Incidence — whether an LVT passes through to tenants or is absorbed by landowners — is explicitly flagged as an open empirical question the review could not resolve, a gap the wiki's own transition wealth-shock page documents from the opposite direction via the conflicting Danish DØRS and Nielsson–Wroblewski–Yding results.

Conclusion: the paper "does not rule out LVT as a potential component of future Welsh tax reform," but stresses this is "not a standalone or transformative solution" and recommends further Wales-specific analysis, "clear policy objectives," and technical work on valuation before any reform proceeds.[1]

Implications for Wales Specifically

The paper devotes a substantial discussion section to why Wales's own characteristics complicate a straightforward read-across from the international evidence:

• Geography. Wales has roughly 154 people per km² against 450 in England, and nearly half its population is concentrated in the south-east (Cardiff, Newport, the Valleys) on only about 14% of its land area; Powys alone covers roughly a quarter of Wales at 26 people per km².[1] The paper's point: LVT's most consistently observed effects worldwide depend on high-demand, dense urban markets — conditions that hold for a small share of Welsh land and population, "most notably Cardiff and a small number of other urban centres."[1]
• Land use. Between 80% and 90% of Welsh land area is agricultural, raising separate design questions around farmland treatment that the wiki's LVT hurts farmers objection page addresses generically.
• No existing valuation infrastructure. "Wales does not currently operate a valuation system designed to isolate unimproved land value at scale" — Council Tax and NDR use different valuation logics entirely, so an LVT would need new data infrastructure, methodology, and professional capacity built essentially from scratch, on top of primary and secondary legislation.[1][2]
• Devolution constraints. Wales gained primary legislative competence over Council Tax and NDR only in 2011, and its 2018 attempt at a new vacant land tax stalled on UK Treasury approval — a precedent the paper cites directly as a feasibility risk for any land-value-based reform requiring similar sign-off.[1]
• Redistribution risk. Because Welsh land values vary sharply between local authorities, a locally retained LVT "would likely replicate existing spatial inequalities ... rather than mitigate them" absent an equalisation mechanism similar to NDR's pooled-and-redistributed model.[1]

The Valuation Methodologies Paper: Five Untested Families, Not Five Verdicts

The companion paper addresses the practical question the first paper's own evidence review surfaces as decisive: how would Wales actually value unimproved land?[2] Its framing is blunt about the underlying difficulty: "there is little land in the UK that is truly 'unimproved,' and land and buildings are commonly sold together at a single price," so estimating a land value "requires some form of analytical separation or administrative rule," and "the 'true' land value can never be known with certainty."[3] Drawing on a review of the empirical LVT literature and a scoping review of 84 property taxes with a land component across the EU27 plus 11 other high-income democracies (a review the paper is explicit "did not assess accuracy, appeal rates, administrative costs, or downstream outcomes" — it catalogues design choices, not performance),[2] the paper concludes "no single optimal method exists" and instead groups viable approaches into five methodology families it recommends testing in Wales:[2][3]

1. Market-based statistical methods — estimating land value from transaction data; closest to a market-value logic, but dependent on rich datasets and liable to struggle in thin rural markets.
2. Algorithmic and machine-learning approaches — large linked datasets and modern data-science techniques predicting land values at scale; capable of capturing complex interactions but data-intensive and potentially opaque (the same transparency trade-off the wiki's mass appraisal methods page discusses for CAMA/ML approaches generally).
3. Formula-based approaches — explicit rules (e.g. a value per square metre) producing administrative values; simple, transparent, and stable, but insensitive to nuance or market change.
4. Conventional valuation approaches — professional appraisal practice and standards, requiring adaptation since "the UK valuation profession does not regularly separate the value of land from improvements."
5. Innovative and experimental approaches — self-assessment, participatory valuation, multi-criteria decision tools; the least tested at scale of the five.

The paper's own framing of the policy question is worth quoting directly, because it is the clearest statement in either document of what has and has not been established: "the overarching policy question is not whether land value can be estimated, but which approach (or combination of approaches) would be sufficiently robust, transparent, feasible, scalable, and aligned with Welsh policy objectives in practice."[3] That is a statement that assessment is an open, testable design question — not a finding that any tested method failed an accuracy bar.

What the Programme Actually Found — and How It Has Been Cited Elsewhere

[VERIFY]: Two pages already on this wiki — research/tpa-what-would-lvt-do and people/andy-burnham — cite this Welsh Government programme for the claim that it found "no [assessment] methodology... yet accurate enough for live implementation" / "none 'accurate enough to be deployed in a live land value tax.'" Having now fetched and read both Goodwin-Hawkins working papers in full, this wiki page's own scan finds no passage in either document reporting that any specific valuation method was tested and found inaccurate. Both papers are dated before any field test: the Claims and contexts paper is a literature synthesis, and the Valuation methodologies paper explicitly proposes five method families for future testing rather than reporting results against them. The closest the papers come to the "not yet accurate enough" framing is a more general, and materially weaker, statement: that Wales currently has no valuation system built to isolate unimproved land value, that "infrequent or insufficiently accurate valuations limit the share of land value uplift that can be captured" in general (citing the international literature on valuation practice, not a Welsh test), and that "whether defensible valuations can be delivered at a national scale remains an open question" for Wales specifically.[1]

This is not a small distinction for a wiki built on claim-level sourcing. "A government scoping review found no method has yet been tested for Wales, and recommends testing five candidate families" is an honest, if less dramatic, description of these two documents. "Welsh Government research found no methodology accurate enough for live implementation" over-states what has been shown — it reads as though a completed test failed, when the actual sequence is: evidence review (this programme, Feb 2026) → tendered field testing (announced by Drakeford in June 2025, not yet reported) → a future accuracy verdict that, as of this page's last review, has not been published. The wiki should correct or soften the two citing pages to reflect this once this finding is reviewed; per this task's file scope this page flags the issue but does not itself rewrite those two pages' substantive claims.

What This Means for the LVT Feasibility Debate

Read honestly, the Welsh programme is a useful, government-adjacent example of the wiki's own land-cannot-be-assessed objection playing out in real institutional time, but on the "not yet attempted" side of that objection rather than the "attempted and failed" side. The objection page's own "Limits and Caveats" section already concedes assessment quality is uneven internationally and that at least one real mass-valuation system (Lithuania's) has badly mistracked market value; Wales adds a different data point — a serious, publicly funded evidence review concluding that the prerequisite work (building a valuation system capable of isolating land value at national scale, then testing candidate methods against it) has not yet been done, not that it was tried and failed. That is consistent with, and arguably softer than, Dan Neidle's UK-wide modelling, where the 55% land-share calibration is at least a working, if unverifiable, assumption rather than an untested proposal.

For advocates, the honest reading is double-edged: the systematic review's finding that LVT's efficiency and equity claims are "plausible in theory but not strongly supported by direct empirical evidence" is a genuine limits-and-caveats result the wiki should not soften, but it also does not "rule out LVT" — its authors explicitly frame gaps as research priorities, not disqualifying findings, and the valuation-methods paper treats assessability as a design and testing problem, structurally the same framing the wiki's own assessment objection page takes ("a solvable engineering problem, not a fundamental barrier"). For policymakers' staffers, the paper's Wales-specific findings — no existing valuation infrastructure, stalled precedent from the 2018 vacant land tax, and redistribution risk from uneven Welsh land values absent an NDR-style pooling mechanism — are exactly the kind of implementation-agenda detail the wiki's reader priorities call for.

See Also

• Objection: Land value can't be assessed accurately — the general objection this programme's "not yet tested" status bears directly on
• What Would a Land Value Tax Actually Do? (Neidle/TPA, 2026) — the UK-wide modelling exercise that cites this Welsh programme as cautionary precedent
• Andy Burnham — UK Prime Minister whose 2026 LVT campaign cited this Welsh finding secondhand
• United Kingdom — the wider UK land-tax record this Welsh programme sits within
• Objection: if LVT is so good, why has almost no one adopted it? — Wales's decade-long study-without-adoption pattern as a live example of the adoption gap this objection describes
• Mass Appraisal Methods — the CAMA/hedonic/ML techniques among the "algorithmic and machine-learning" family this programme proposes testing
• Denmark — one of the two international comparators (with Singapore) Mark Drakeford cited in his June 2025 Senedd statement
• Wales — the place profile covering Wales's wider devolved-tax history and the 2025 political context this programme sits within

Sources

1. Bryonny Goodwin-Hawkins (2026), A Land Value Tax for Wales? Claims and contexts, Working Paper version 3.0, Countryside and Community Research Institute. Welsh Government copy (PDF) · University of Gloucestershire eprints copy (both fetched and read in full, 74 pp., this session) — used for the systematic-review methodology, the seven-claims framework, all efficiency/equity findings, the Wales-specific discussion (geography, devolved tax system, redistribution risk), and the paper's own conclusions. The two copies carry a one-year date discrepancy on their title page ("January 2026" in the Welsh Government copy vs. "January 2025" in the eprints copy) that this page cannot resolve from the primary texts alone; both are otherwise the same version-3.0 text.
2. Bryonny Goodwin-Hawkins (2026), A Land Value Tax for Wales? Valuation methodologies, Working Paper 2, version 1.2, Countryside and Community Research Institute. University of Gloucestershire eprints copy, full text (41 pp.) (fetched and read in full this session) — used for the five-methodology-group framework, the international-practice scoping review (84 taxes, EU27 plus 11 other democracies), the glossary definitions, and the paper's own framing of the "which approach" policy question.
3. Bryonny Goodwin-Hawkins (2026), A Land Value Tax for Wales? Valuation methodologies — four-page executive-summary edition, version 1.1, February
4. Welsh Government copy (PDF) (fetched and read in full) — used only to confirm the executive-summary wording quoted above and cross-check it against the full working paper (source 2); a condensed public-facing summary of the same report.
5. Chris Haines, "Land value tax in Wales could replace council tax and rates," South Wales Guardian, 11 June 2025. southwalesguardian.co.uk (fetched and read in full) — used for Mark Drakeford's Senedd statement, the Denmark/Singapore comparators, the tender-invitation announcement, and Peter Fox's opposition response on the local-government funding formula.