Previous version · national priorities, 4 October 2026 · method

How the previous national priorities were chosen

The ranking is a judgement on nine published criteria, shown with its rules and tested under six weightings. It is not a complete ordering of every possible UK policy, and a one-point difference is within the judgement’s error.

Previous version, superseded on 7 October 2026. This table chose the 4 October 2026 priorities. The national action plan replaced them with four outcomes: a reviewed editorial selection with provisional assessments, not a ranking. The table is kept unchanged for reference.

The method in brief

We assessed 68 candidate actions from 34 topic plans (67 published action records, plus drafts considered but not published). Accountability, governance and site-method actions were screened out of the national slate by design. The rest were rated High, Medium or Low on nine published criteria — severity and scale, the benefit added beyond existing law and funding, evidence, cost and funding, scarce capacity, distribution, time to impact, distinct cross-domain outcomes and prerequisites — with the same marginal-change test for funded and unfunded options alike. The six outcome priorities are the top of that table, checked under six weightings; delivery enablers are judged separately because their evidence differs. An editorial judgement with reasons, not government policy.

Six outcome priorities, not seven. The first five are in the top six under all six weightings tested. The sixth place is not stable: early-years hubs (A13), Pension Credit take-up outreach (A25), a coordinated response to one scam route (A06) and, since the restored concerns were added on 4 October 2026, metering homes in water-stressed areas with a social tariff (A53) tie on the base weighting and swap places under others; the published tie-break goes to the stronger evidence, then to distinct cross-domain outcomes — A13 and A53 both have an evaluated effect, and only A13 has evaluated benefits in two areas. Flood-defence allocation by risk (A54) scores lower: its evidence is an audit, not an evaluation. Funded or locally delivered status counts the same way for every candidate: it can help on cost and funding, and the benefit test looks only at what a proposal adds beyond the programme — which is why hot-spot policing (A07), the Youth Guarantee (A08), tutoring (A10), NELI (A11) and surgical hubs (A04) stay in their domain plans.

← Back to the previous decision portfolio

The selection table and sensitivity

The nine criteria and what High, Medium and Low mean

Selection criteria with the anchor for each band.
Criterion	High	Medium	Low
Severity and scale	Severe harm to many people (e.g. deaths, homelessness, long-term worklessness) on a national scale.	Serious harm to a defined group, or wide but moderate harm.	Limited harm or a narrow group.
Benefit added beyond existing law and funding	Nothing like it exists; the whole change is additional.	Builds on an existing programme but adds a specified change the evidence says matters.	Adds little beyond what is already funded, legislated or within existing discretion.
Confidence in the evidence	Randomised trials or a strong controlled evaluation of this intervention in a comparable population.	Strong evidence for a related intervention or population, or observational evidence with a clear mechanism.	Expert recommendation, modelling of a different change, or untested.
Cost and plausible funding	Little public cost, or the incremental cost is small and has a clear route.	A programme route exists, but the incremental cost is not established.	Large, unfunded, or adds recurring spending with no route.
Feasibility and scarce capacity	Uses little scarce workforce or infrastructure.	Needs capacity that is stretched but obtainable.	Competes for a binding scarce resource (builders, installers, therapists).
Distribution and equity	Gains concentrated among poorer or more disadvantaged people, with manageable harms.	Mixed: progressive gains with a material harm to some.	Regressive, or a material harm to a vulnerable group.
Time to first useful result	Measurable change within about a year.	Two to three years.	Many years.
Distinct supported cross-domain outcomes	Evaluated benefits in two or more distinct policy areas.	One supported area plus a measured effect in a related area.	One area, or other links only hypothesised.
Prerequisites	Can start under current powers.	Needs a named decision, allocation or referral route first.	Needs legislation plus major enabling work, or an external agreement.

Rules. A candidate whose benefit beyond existing law and funding is Low is not eligible for an outcome slot, whatever its funding status. Outcome priorities are the top of the base weighting (base: severity, added benefit and evidence double; High = 2, Medium = 1, Low = 0); ties go to stronger evidence, then to distinct cross-domain outcomes. Delivery enablers are rated the same way but chosen by which outcome priorities depend on them.

Every rated candidate with its marginal action, existing status, nine criterion bands (H high, M medium, L low), base points and the outcome.
Candidate and what it adds	Existing status	Severity and scale	Benefit added beyond existing law and funding	Confidence in the evidence	Cost and plausible funding	Feasibility and scarce capacity	Distribution and equity	Time to first useful result	Distinct supported cross-domain outcomes	Prerequisites	Base points	Outcome
Set a minimum price per unit of alcohol in England (A49)An English minimum unit price, by primary legislation.	No English policy; legislation needed; not a spending programme.	High	High	High	High	High	Medium	Medium	Low	Medium	19	Selected: Outcome priority 1Within-UK controlled evaluation of deaths; fully additional; harm to some dependent drinkers recorded.
Offer evidence-based employment support alongside healthcare (A09)Fidelity reviews, NHS integration and an impact evaluation inside Connect to Work.	Funded programme (committed to 2028–29); delivered by local areas.	High	Medium	Medium	Medium	Medium	High	High	Low	High	16	Selected: Outcome priority 2UK RCTs with modest, place-varying effects; the funded programme does not pay for the integration proposed.
Fit home adaptations faster for older people at high risk of falling, after an occupational-therapist hazard check (A52)Falls-risk targeting, an OT check and a speed standard within the Disabled Facilities Grant.	Existing grant (£723m, 2026–27); delivered by housing authorities.	Medium	Medium	High	Medium	Low	High	High	Low	Medium	14	Selected: Outcome priority 3High-certainty Cochrane evidence; targeting is what the evidence says matters; OT capacity scarce.
Fix cold and inefficient homes with quality-assured work (A18)Health-linked selection, independent inspection before payment and an evaluation inside the Warm Homes grants.	Funded programme (£4.4bn low-income grants); delivered by councils.	High	Medium	Medium	Medium	Low	High	Medium	Low	Medium	13	Selected: Outcome priority 4Observational energy savings; health evidence from a New Zealand trial in different homes; installers scarce.
Use social-housing funding to reduce temporary accommodation (A02)A buy/build/lease comparison against temporary accommodation, and awards tied to lettings.	Funded programme (£39bn envelope, bidding); delivered by councils and housing associations.	High	Medium	Medium	Medium	Low	High	Medium	Low	Medium	13	Selected: Outcome priority 5Severe harm; mechanism supported by a US trial of subsidies for families; construction capacity scarce.
Rebuild integrated early-years help where access is weakest (A13)Evaluated Sure Start features, weakest-access areas first, and a phased, evaluable roll-out of the family hubs.	Funded programme (over £500m); delivered by councils.	Medium	Medium	Medium	Medium	Low	High	Low	High	Medium	12	Selected: Outcome priority 6Only candidate with evaluated benefits in two areas; transfer uncertain; Flying Start found no early child effect.
Meter homes in water-stressed areas, with a social tariff that protects low-income households (A53)Meter every home in seriously water-stressed areas, with a protected social tariff.	Smart-meter roll-out committed by companies (50% by 2030); compulsory metering allowed in designated areas; bill-funded.	Medium	Medium	Medium	Medium	Medium	Medium	High	Low	Medium	12	Tied at the boundary; lost the evidence tie-breakUK quasi-experimental evidence of an 18–22% fall in use; less affluent households lose without a social tariff; benefits are in one area (water), not a second.
Help eligible people claim support they already qualify for (A25)Assisted applications for people likely to be entitled to Pension Credit.	Campaigns already run; DWP and councils.	Medium	Medium	Low	Low	High	High	High	Low	High	12	Tied at the boundary; lost the evidence tie-breakUntested outreach; higher take-up adds unbudgeted benefit spending.
Stop scams across banks, platforms and telecoms (A06)A coordinated bank–telecoms–police response to one scam route, with an evaluation.	Fraud Strategy funding (over £250m); national and private delivery.	High	Medium	Low	Medium	Medium	Medium	High	Low	Medium	12	Tied at the boundary; lost the evidence tie-breakLargest crime by volume; no evaluated effect; Singapore reports activity, not prevention.
Reform property taxes without hiding the losers (A22)Council tax revaluation and reform, revenue-neutral.	Not government policy; national legislation.	Medium	High	Medium	Medium	Medium	Medium	Low	Low	Low	11	Not selectedModelled distributional gains with concentrated losers; years to implement.
Improve pension saving while protecting low take-home pay (A24)Higher default pension saving with an employer floor.	External proposal; awaiting the Pensions Commission.	Medium	High	Medium	Medium	Medium	Low	Low	Low	Low	10	Not selectedModelled; take-home pay harm now; benefits over decades.
Offer reablement to more older people leaving hospital, and judge it by whether they are still at home three months later (A51)A defined reablement cohort with a 91-day comparison.	Established service, Better Care Fund; local.	Medium	Medium	Low	Medium	Low	High	High	Low	Medium	10	Not selectedCochrane could neither support nor refute an effect.
Allocate flood-defence money by the new risk map, keep existing defences in condition, and report homes better protected every year (A54)Allocate the 2026–29 flood programme by the 2024 risk map, protect maintenance and report homes better protected each year.	Funded programme (£4.2bn, 2026–27 to 2028–29); Environment Agency and partners.	High	Medium	Low	Medium	Medium	Medium	Low	Low	Medium	10	Not selectedLarge mapped exposure (6.3 million properties); the evidence is an audit, not an evaluation; defences take years.
Measure what school mental health teams achieve as they reach every school, and close the gap for children with greater needs (A50)A comparison-based evaluation and a route for greater needs.	Funded national roll-out to 2029–30.	Medium	Medium	Low	Medium	Medium	Medium	Medium	Low	Medium	9	Not selectedNo outcome evaluation yet.
Reduce food-trade friction through a clearly scoped agreement (A29)Conclude the SPS agreement.	Negotiation under way.	Medium	Medium	Medium	Medium	Low	Medium	Low	Low	Low	8	Not selectedDepends on an external agreement.
Fund SEND capacity and track provision actually delivered (A12)A costed specialist-capacity plan for SEND.	Reform announced; not costed.	High	Medium	Low	Low	Low	High	Low	Low	Low	8	Not selectedSerious harm; unevaluated statutory reform; specialists take years to train.
Set an explicit test for whether benefits cover essentials (A16)An essentials benchmark for benefit rates.	Not government policy.	High	Medium	Low	Low	Low	High	Low	Low	Low	8	Not selectedA benchmark, not a delivery programme; rate decisions uncosted.
Make electricity pricing support cheaper clean heating (A17)Rebalance levies between electricity and gas.	Partly done in Budget 2025.	Medium	Medium	Low	Low	Medium	Low	Medium	Low	Medium	7	Not selectedModelled on an overtaken baseline; losers among gas households.
Target repeat violence at places where harm clusters (A07)Target repeat violence at hot spots.	Hot Spot Action Fund pays all 43 forces.	Medium	Low	Medium	Medium	High	Low	High	Low	High	—	Not eligible: adds little beyond existing law or fundingAdds little beyond the funded programme; English results mixed.
Build paid routes into work for excluded young people (A08)Paid routes into work for young people.	Youth Guarantee funded (£820m).	High	Low	Medium	Medium	Medium	High	High	Low	High	—	Not eligible: adds little beyond existing law or fundingAdds little beyond the Youth Guarantee as funded.
Target tutoring at the gaps pupils actually have (A10)Target tutoring at measured gaps.	Within schools’ pupil-premium discretion.	Medium	Low	Medium	Medium	Medium	High	High	Low	High	—	Not eligible: adds little beyond existing law or fundingSchools can already do this.
Meet early language needs before children fall further behind (A11)Protected time to deliver all 20 weeks of NELI.	Funded free for every state school to 2028–29.	Medium	Low	High	High	Medium	High	High	Low	High	—	Not eligible: adds little beyond existing law or fundingAlready offered nationally; full delivery is a school decision.
Protect elective surgery from emergency disruption (A04)Protect elective surgery in ring-fenced hubs.	Established NHS policy.	High	Low	Medium	Low	Low	Medium	Medium	Low	Medium	—	Not eligible: adds little beyond existing law or fundingEstablished policy; evaluated on activity at hub trusts, not waits.
Clear asylum cases without sacrificing decision quality (A14)Faster asylum casework with quality checks.	Reform money (£200m) allocated.	Medium	Low	Medium	Medium	Low	Medium	Medium	Low	Medium	—	Not eligible: adds little beyond existing law or fundingAdds little beyond the allocated reform; trade-offs recorded.
Fund military readiness before counting new announcements (A26)Close readiness gaps.	Defence Investment Plan funded.	Medium	Low	Low	Low	Low	Medium	Low	Low	Low	—	Not eligible: adds little beyond existing law or fundingReadiness data withheld; impact cannot be assessed publicly.
Give productive investment a stable delivery pipeline (A03, enabler)Approved-versus-aspirational status, milestones and completed assets in the infrastructure pipeline.	Pipeline exists (£725bn envelope).	High	Medium	Low	Medium	Medium	Medium	Medium	Low	Medium	—	Selected: Delivery enabler 1Sequences the builders and connections NP1 and NP4 need.
Turn permissions into homes people can live in (A01, enabler)A blocker, owner and infrastructure cost for each priority housing site.	Planning reforms in force; not costed.	High	Medium	Low	Low	Low	Medium	Medium	Low	Medium	—	Selected: Delivery enabler 1Planning discipline that NP1’s new-build route depends on.
Buy measurable results and manage contracts through delivery (A23, enabler)Outcome-and-milestone baselines and contract management for named renewing contracts.	Procurement Act duties in force.	Medium	Medium	Low	Medium	Low	Medium	Medium	Low	High	—	Selected: Delivery enabler 2Commercial capacity that NP1, NP4 and NP5 all draw on.
Tie spending decisions to realistic outcomes and costs (A20, enabler)A published cost-of-current-standards baseline per service.	Green Book and Spending Review practice.	Medium	Medium	Low	Medium	Medium	Medium	Medium	Low	High	—	Selected: Delivery enabler 2Realistic baselines before money is committed.

Sensitivity check: is the choice stable?

Rank of each eligible outcome candidate under six weightings (ties share a rank). “In the top 6” counts the weightings in which it would make the slate. 5 of the 6 selected outcome priorities are in the top 6 under every weighting. Not stable: Rebuild integrated early-years help where access is weakest — its place is shared and changes with the weighting.

Ranks of eligible outcome candidates under six weightings.
Candidate	Base: severity, added benefit and evidence double	All nine criteria equal	Evidence triple	Cross-domain outcomes ignored	Cost and funding ignored (funded status cannot help)	Severity triple	In the top 6
Set a minimum price per unit of alcohol in England (A49, selected)	1 (19)	1 (13)	1 (21)	1 (19)	1 (17)	1 (21)	6 of 6
Offer evidence-based employment support alongside healthcare (A09, selected)	2 (16)	2 (12)	2 (17)	2 (16)	2 (15)	2 (18)	6 of 6
Fit home adaptations faster for older people at high risk of falling, after an occupational-therapist hazard check (A52, selected)	3 (14)	3 (10)	3 (16)	3 (14)	3 (13)	3 (15)	6 of 6
Fix cold and inefficient homes with quality-assured work (A18, selected)	4 (13)	5 (9)	4 (14)	4 (13)	4 (12)	3 (15)	6 of 6
Use social-housing funding to reduce temporary accommodation (A02, selected)	4 (13)	5 (9)	4 (14)	4 (13)	4 (12)	3 (15)	6 of 6
Rebuild integrated early-years help where access is weakest (A13, selected)	6 (12)	5 (9)	6 (13)	10 (10)	7 (11)	7 (13)	3 of 6
Meter homes in water-stressed areas, with a social tariff that protects low-income households (A53)	6 (12)	5 (9)	6 (13)	6 (12)	7 (11)	7 (13)	4 of 6
Help eligible people claim support they already qualify for (A25)	6 (12)	3 (10)	8 (12)	6 (12)	4 (12)	7 (13)	4 of 6
Stop scams across banks, platforms and telecoms (A06)	6 (12)	5 (9)	8 (12)	6 (12)	7 (11)	6 (14)	4 of 6
Reform property taxes without hiding the losers (A22)	10 (11)	11 (7)	8 (12)	9 (11)	10 (10)	10 (12)	0 of 6
Improve pension saving while protecting low take-home pay (A24)	11 (10)	14 (6)	11 (11)	10 (10)	11 (9)	12 (11)	0 of 6
Offer reablement to more older people leaving hospital, and judge it by whether they are still at home three months later (A51)	11 (10)	10 (8)	12 (10)	10 (10)	11 (9)	12 (11)	0 of 6
Allocate flood-defence money by the new risk map, keep existing defences in condition, and report homes better protected every year (A54)	11 (10)	11 (7)	12 (10)	10 (10)	11 (9)	10 (12)	0 of 6
Measure what school mental health teams achieve as they reach every school, and close the gap for children with greater needs (A50)	14 (9)	11 (7)	14 (9)	14 (9)	14 (8)	14 (10)	0 of 6
Reduce food-trade friction through a clearly scoped agreement (A29)	15 (8)	15 (5)	14 (9)	15 (8)	17 (7)	17 (9)	0 of 6
Fund SEND capacity and track provision actually delivered (A12)	15 (8)	15 (5)	16 (8)	15 (8)	14 (8)	14 (10)	0 of 6
Set an explicit test for whether benefits cover essentials (A16)	15 (8)	15 (5)	16 (8)	15 (8)	14 (8)	14 (10)	0 of 6
Make electricity pricing support cheaper clean heating (A17)	18 (7)	15 (5)	18 (7)	18 (7)	17 (7)	18 (8)	0 of 6

All 35 topic routes: how each relates to the national plan

A contribution means an action on that page is part of a priority. It never adds to a priority’s evidence count, which comes only from effect records.

Economy and cost of livingContributes to priority: NP3, NP5A09 is NP3; A01 and A03 form the delivery enabler NP5, whose final aim is living standards.
HousingContributes to priority: NP1, NP5A02 is NP1 (buy or build social-rent homes for households in temporary accommodation, decision brief on this page); A01 (priority-site blockers) is part of the delivery enabler NP5. Flood-defence allocation by risk (A54, restored climate-adaptation concern) is retained here.
NHSContributes to priority: NP3A09 forms NP3. Surgical hubs (A04) stay in the NHS plan: hubs are established policy, so the proposal adds little beyond it, and the evaluated effect is activity at hub trusts, not total activity or waits.
CrimeImportant domain actions retained locallyHot-spot violence prevention (A07) is well evidenced, but the Hot Spot Action Fund already pays for it in every force, so the proposal adds little; scam prevention (A06) is a published near-miss — the largest crime by volume, but no evaluated effect, so a bounded test comes first.
Jobs and payContributes to priority: NP3A09 forms NP3 (decision brief on this page). Paid youth routes (A08) stay in the jobs plan: the proposal adds little beyond the Youth Guarantee as funded.
EducationContributes to priority: NP2A13 is NP2 (decision brief on this page). NELI (A11) is now funded for every state school to 2028–29 and tutoring (A10) is already within schools’ pupil-premium discretion, so neither adds much as a national decision.
SENDImportant domain actions retained locallySEND capacity (A12) is a serious harm but an unevaluated, uncosted statutory reform, and NELI (A11) is already funded nationally; both stay in the SEND plan. No national priority claims to reduce SEND demand.
ImmigrationImportant domain actions retained locallyAsylum casework (A14) adds little beyond the reform money already allocated and carries recorded trade-offs (appeals, move-on homelessness); A15 redistributes a fixed settlement.
InequalityContributes to priority: NP2, NP4A13 (NP2) and A18 (NP4) are on this page; the benefits-adequacy test (A16) stays here as an untested proposal.
InflationFurther evidence neededElectricity pricing (A17) needs re-modelling on current tariffs, and A16 is untested; the Bank of England’s remit is not a priority lever.
Energy and billsContributes to priority: NP4A18 forms NP4; A17 stays here until it is re-modelled.
Public financesContributes to priority: NP6A20 is part of NP6; A21 and A47 support every priority’s funding decisions.
Follow the moneyContributes to priority: NP6A23 is part of NP6. Tax-compliance yields (A19) cannot be assumed to fund any priority. Water metering with a social tariff (A53, restored water-security concern) is retained here and was a near-miss for the sixth outcome place.
Revenue optionsImportant domain actions retained locallyTax choices (A19, A21, A22) decide who pays; a revenue-neutral property reform cannot fund other priorities, and no receipt is counted as committed.
Public procurementContributes to priority: NP6A23 is part of NP6; public-buyer innovation trials (A42) need evaluation first.
BenefitsContributes to priority: NP3A09 is part of NP3; benefit effects are not separately evidenced, so they are not counted. Take-up (A25) and adequacy (A16) stay here.
PensionsImportant domain actions retained locallyThe costed saving package (A24) helps future pensioners but squeezes take-home pay now and awaits the Pensions Commission. Pension Credit take-up outreach (A25) is a published near-miss: quick, cheap to run and well targeted, but untested, and the extra benefit spending is not budgeted.
DefenceContributes to priority: NP6A23 (defence equipment contracts) is part of the delivery enabler NP6. Readiness (A26) stays here: readiness data are withheld, so its impact cannot be assessed publicly.
MonarchySupporting governance or evidence functionCost transparency and the Grant formula (A27, A28) are accountability actions, not national outcome programmes.
BrexitImportant domain actions retained locallyThe SPS agreement (A29) depends on a negotiation already under way and trades access against regulatory autonomy.
Media and informationSupporting governance or evidence functionPrebunking and local-reporting pilots (A31, A32) concern informed choice; UK effects are untested.
Votes and seatsSupporting governance or evidence functionA democratic choice: the actions (A33, A34) inform it without implying one correct system.
Political fundingSupporting governance or evidence functionDisclosure and donation limits (A35, A36) are accountability actions.
Government deliveryContributes to priority: NP6A20 is part of NP6; A37 (owner, budget and measure for every commitment) is the delivery discipline the national plan uses.
Manifesto trackerSupporting governance or evidence functionA37 and A38 make commitments checkable; they are scrutiny, not programmes.
Party comparisonsSupporting governance or evidence functionOne costing template for every party (A38, A37) supports informed choice, not a national programme.
PollingSupporting governance or evidence functionPoll provenance and back-testing (A39, A40) are evidence-quality actions.
Growth and productivityContributes to priority: NP5A03 and A01 form the delivery enabler NP5 (decision brief on this page); A45 (sequencing shared capacity) supports it in “What has to happen first”.
InnovationFurther evidence neededSpin-out terms (A41) and public-buyer trials (A42) rest on expert recommendation; adoption and useful outcomes are not yet measured.
AI capability and resilienceFurther evidence neededCloud exit (A43) and AI adoption (A44) are untested; adoption rates are not outcomes.
UK SWOT and priority action planSupporting governance or evidence functionThis page: the synthesis itself, never counted as an outcome area.
Policy scenariosSupporting governance or evidence functionModel validation (A46) and the funding ledger (A47) support the plan; the national funding table uses the same ledger.
HampshireContributes to priority: NP1, NP4Local delivery of A02 and A18 through Hampshire allocations; a geography, never counted as an extra policy area. The local plan (A48) stays here.
HealthContributes to priority: NP2, NP7Alcohol minimum unit pricing (A49) is NP7, with its decision brief on this page; early-years help (A13, NP2) is here for its evaluated effect on children’s hospital admissions. School mental health teams (A50) need outcome evidence first.
Social careContributes to priority: NP8Falls-risk home adaptations (A52) are NP8, with a decision brief on this page: high-certainty trial evidence, and the targeting is the change. Reablement (A51) lacks trial evidence. Cost protection (a cap or the means test) changes who pays, not how much care is delivered, and is assessed separately from capacity.

How every candidate was assessed in words (68 candidates: 67 published action records and 1 unpublished draft (A05))

A01 Turn permissions into homes people can live in — Selected (NP5)Part of the delivery enabler NP5 with A03: a site-by-site blocker list is a planning discipline, not an outcome intervention, so it no longer sits inside the homes priority.
Scale of benefit
High: a typical home costs 7.6 times earnings in England and supply is below the 300,000 a year ambition.
Cross-domain effects
Housing; links to work and poverty plausible but untested.
Evidence
Modelled (OBR, for reforms already in law); this site package is unevaluated.
Additionality
Adds a site-by-site blocker list to reforms already in force.
Cost and funding
Not yet costed; infrastructure costs fall on government, councils, developers and bill-payers.
Deliverability
Needs planning staff, connections and builders; powers exist.
Risk
Counting permissions as homes; local opposition; environmental harm.
A02 Use social-housing funding to reduce temporary accommodation — Selected (NP1)NP1: the direct route from committed money to a severe harm — households in temporary accommodation — framed as a buy-or-build choice against the cost of temporary accommodation.
Scale of benefit
High: 135,580 households, including 177,530 children, in temporary accommodation.
Cross-domain effects
Housing; poverty and children’s stability plausible, untested.
Evidence
Observational: the last programme delivered homes but below plan; the temporary-accommodation link is unevaluated.
Additionality
Ties an existing £39bn programme to temporary-accommodation outcomes.
Cost and funding
Programme envelope exists; scheme costs need awards.
Deliverability
Needs awards, sites and builders; no new law.
Risk
Cost inflation; homes not additional locally.
A03 Give productive investment a stable delivery pipeline — Selected (NP5)The delivery enabler NP5: low investment is a persistent weakness and NP1 and NP4 depend on the builders and connections it would sequence. Judged as an enabler, not an outcome intervention.
Scale of benefit
High: investment the lowest in the G7 in 23 of 33 years.
Cross-domain effects
Living standards (modelled, limited confidence); other areas depend on the projects chosen.
Evidence
Modelled (OBR general investment model); no study of a pipeline.
Additionality
Pipeline exists; adds approved-versus-aspirational status and completion measures.
Cost and funding
£725bn programme envelope; project costs published one by one.
Deliverability
Workforce and grid constrained.
Risk
Locking in poor projects; cost escalation; interest costs.
A04 Protect elective surgery from emergency disruption — Kept in its domain planAdds little beyond established policy (benefit added: Low), and its evaluated effect is activity at hub trusts, not total activity or waits; kept on /nhs.
Scale of benefit
High: 65.4% of pathways within 18 weeks against a 92% standard.
Cross-domain effects
Health only.
Evidence
Quasi-experimental: +21.9% routine operations at hub trusts; total elective activity unchanged; waits not measured.
Additionality
Hubs are established NHS policy; the proposal adds little beyond it.
Cost and funding
Not yet costed; 2026/27 capital finishes existing schemes.
Deliverability
Staff shared with emergency care.
Risk
Moves work rather than adds it.
A05 Connect prevention to the services that benefit — Further evidence neededA candidate considered but never published: the NHS plan did not shortlist it because there is no specific intervention, cost or effect record, and on /health it is replaced by specific prevention actions with their own evidence (A49, A13, A50).
Scale of benefit
Potentially wide (prevention).
Cross-domain effects
Not specified.
Evidence
Brief draft only; not reconciled.
Additionality
Unclear.
Cost and funding
Not costed.
Deliverability
Not specified.
Risk
Vague scope.
A06 Stop scams across banks, platforms and telecoms — Further evidence neededFlagged near-missThe crime candidate with the largest scale, but nothing yet shows cross-sector sharing cuts victimisation; kept on /crime with an evaluation-first step.Flagged for the next selection: Ties with NP2 on the base weighting: the largest crime by volume (47% of Crime Survey incidents), quick to test and with funding routes, but no evaluated effect of cross-sector signal sharing; Singapore’s police–bank model reports activity, not demonstrated prevention. Recommended as a bounded, evaluated test of one scam route on /crime.
Scale of benefit
High volume: fraud is 47% of Crime Survey incidents.
Cross-domain effects
Crime only; wrongly blocked payments are a recorded harm.
Evidence
Expert workshops; no evaluation of cross-sector signal sharing.
Additionality
Fraud Strategy 2026–2029 funded (over £250m).
Cost and funding
Programme funding exists; the pilot is not costed.
Deliverability
Needs lawful data flows and firms’ cooperation.
Risk
False positives and privacy intrusion.
A07 Target repeat violence at places where harm clusters — Kept in its domain planBenefit added beyond the existing fund is Low — the same test applied to selected priorities, each of which adds a specified change — and English results were mixed (−7% then no significant effect); kept on /crime.
Scale of benefit
Serious harm concentrated in few places.
Cross-domain effects
Crime (with a trust risk); sharp-object admissions are the same incidents, not a second area.
Evidence
Synthesis of 65 studies; English results −7.0% (2021–22) then +0.9% (2022–23).
Additionality
The Hot Spot Action Fund already pays all 43 forces for this; the proposal adds little beyond it.
Cost and funding
Programme envelope; local costs not costed.
Deliverability
Police and partners can act now.
Risk
Community trust and disproportionate stop and search.
A08 Build paid routes into work for excluded young people — Kept in its domain planBenefit added beyond the Youth Guarantee is Low; a single-area effect measured for a different group. Kept on /employment and named as NP3’s strongest alternative.
Scale of benefit
981,000 young people not in education, employment or training.
Cross-domain effects
Work only.
Evidence
Quasi-experimental: Kickstart left 11 more in 100 in unsubsidised work at two years.
Additionality
The Youth Guarantee already funds paid placements; the proposal adds little beyond it.
Cost and funding
Programme envelope; support costs not costed.
Deliverability
Depends on employer vacancies.
Risk
Deadweight and poor placements.
A09 Offer evidence-based employment support alongside healthcare — Selected (NP3)NP3: one of the few tested responses to health-related inactivity; the funded programme covers delivery, not the integration and evaluation proposed.
Scale of benefit
High: 2.78 million inactive through long-term sickness.
Cross-domain effects
Work (supported); health (limited confidence, needs testing).
Evidence
Randomised trials of 9,785 people: modest, place-varying effects.
Additionality
Connect to Work funded; adds fidelity, NHS integration and evaluation.
Cost and funding
Committed £1,092m to 2028–29; NHS side and later years not committed.
Deliverability
Can start within current powers.
Risk
Unsuitable jobs; over-claiming benefit savings.
A10 Target tutoring at the gaps pupils actually have — Kept in its domain planStrong general evidence, but the proposal adds little beyond what schools can already do (benefit added: Low); kept on /education.
Scale of benefit
Disadvantage gap of 27.3 points at GCSE.
Cross-domain effects
Education only.
Evidence
Synthesis: about four months’ progress.
Additionality
Schools can already buy tutoring through the pupil premium; the proposal is targeting advice.
Cost and funding
Not costed nationally.
Deliverability
Schools can act now.
Risk
Untargeted tutoring dilutes effects.
A11 Meet early language needs before children fall further behind — Kept in its domain planRemoved from NP2 at the re-selection: NELI is now funded nationally, so a national decision adds little (benefit added: Low); full 20-week delivery is a school decision. Kept on /education and /education/send.
Scale of benefit
Children starting school with weak language fall behind.
Cross-domain effects
Education (high confidence).
Evidence
Randomised trial in 193 schools: effect size 0.26.
Additionality
DfE funds NELI for every state school with a reception class to 2028–29; the remaining step is full delivery in each school.
Cost and funding
Historical price only; staff time not costed.
Deliverability
Needs trained teaching assistants.
Risk
Delaying statutory assessment; diluted delivery.
A12 Fund SEND capacity and track provision actually delivered — Kept in its domain planA serious harm, but the remedy is a statutory-system reform with no evaluated effect; NP2 does not claim to reduce SEND demand. Kept on /education/send.
Scale of benefit
High: 46.1% of new EHC plans within 20 weeks; council deficits about £8bn.
Cross-domain effects
Education; shares specialist health capacity.
Evidence
Audit diagnosis; package unevaluated.
Additionality
SEND reform announced; the 2026 legislative baseline is unverified.
Cost and funding
Not costed.
Deliverability
Specialists take years to train.
Risk
Restricting plans without improving support.
A13 Rebuild integrated early-years help where access is weakest — Selected (NP2)NP2: the only candidate with evaluated benefits in two distinct areas; sixth on the table, ahead of tied candidates without evaluated effects.
Scale of benefit
Gaps open before school; services have thinned.
Cross-domain effects
Education and health (both supported); crime mixed.
Evidence
Quasi-experimental (IFS 2025): GCSEs +0.9 points, admissions −8.5% by 15.
Additionality
Best Start Family Hubs are funded (over £500m); adds the evaluated features, weakest-access areas first and a phased, evaluable roll-out.
Cost and funding
Programme envelope (over £500m) for the hubs; the evaluated service mix is not costed; a historical appraisal (£2.7bn cost, £5.5bn discounted benefits) is not cash.
Deliverability
Workforce-limited.
Risk
Transfer to today’s services uncertain.
A14 Clear asylum cases without sacrificing decision quality — Kept in its domain planBenefit added beyond the allocated reform money is Low, and the evidence records trade-offs (appeals, move-on homelessness), not a cross-domain gain.
Scale of benefit
Asylum backlog and accommodation costs.
Cross-domain effects
Migration, with recorded harms (appeals, move-on homelessness).
Evidence
Observational: initial backlog fell while quality fell to 52%.
Additionality
Reform money (£200m) already allocated; the proposal adds little beyond it.
Cost and funding
Programme funding; casework not costed.
Deliverability
Home Office and tribunal capacity.
Risk
Queue moves to appeals.
A15 Match local service funding to changing population needs — Further evidence neededRedistributes a fixed settlement; nothing links formula responsiveness to outcomes.
Scale of benefit
Pressure on services in fast-changing areas.
Cross-domain effects
Public services, plausible.
Evidence
Audit lessons; no outcome evidence.
Additionality
Formula reform under way.
Cost and funding
A reallocation: gains are other areas’ losses.
Deliverability
MHCLG data.
Risk
Zero-sum.
A16 Set an explicit test for whether benefits cover essentials — Further evidence neededA benchmark, not a delivery programme; the rate decision it informs is uncosted.
Scale of benefit
4.0 million children in poverty after housing costs.
Cross-domain effects
Poverty plausible; recurring fiscal cost recorded.
Evidence
Untested proposal (JRF estimates).
Additionality
Universal Credit Act 2025 baseline.
Cost and funding
Not costed.
Deliverability
Needs a remit; possibly legislation.
Risk
Recurring cost; work incentives.
A17 Make electricity pricing support cheaper clean heating — Further evidence neededThe energy-pricing candidate: needs re-modelling on current tariffs; named as NP4’s strongest alternative.
Scale of benefit
Electricity costs 3.3 times gas per unit.
Cross-domain effects
Energy, with losers among gas households.
Evidence
Modelled on a 2025 baseline now overtaken.
Additionality
Budget 2025 moved 75% of the Renewables Obligation to taxation.
Cost and funding
Not costed.
Deliverability
DESNZ and Ofgem.
Risk
Higher bills for households that cannot switch.
A18 Fix cold and inefficient homes with quality-assured work — Selected (NP4)NP4: measured energy savings and a funded programme; the change is health-linked selection, independent inspection before payment and an evaluation.
Scale of benefit
About 2.36 million households in fuel poverty (England).
Cross-domain effects
Energy (supported); health and poverty need testing; damp is a recorded risk.
Evidence
Observational (NEED): median gas use −17% after solid wall insulation.
Additionality
Adds targeting and quality assurance inside the £4.4bn programme.
Cost and funding
Programme envelope; quality assurance not costed.
Deliverability
Installer-limited.
Risk
Poor installation causing damp; rent rises.
A19 Collect more tax through specific, auditable compliance work — Kept in its domain planThe taxation candidate: yields cannot be assumed to fund other priorities.
Scale of benefit
Tax gap £59.2bn (6.4% of tax due).
Cross-domain effects
Public finances only.
Evidence
Audit; no marginal return available.
Additionality
A £6.5bn-a-year package is already in forecasts.
Cost and funding
Not costed; yield is not cash.
Deliverability
HMRC.
Risk
Banking recoveries that never arrive.
A20 Tie spending decisions to realistic outcomes and costs — Selected (NP6)Part of the delivery enabler NP6 with A23: realistic baselines before money is committed.
Scale of benefit
Spending decisions short-term and reactive (NAO).
Cross-domain effects
Public service delivery; no direct outcome effect recorded.
Evidence
Audit and appraisal guidance; no evaluated saving.
Additionality
Adds a published cost-of-current-standards baseline.
Cost and funding
Analytical staff time, not costed.
Deliverability
HM Treasury can act.
Risk
Paperwork; gaming.
A21 Show tax choices with their real trade-offs — Kept in its domain planInforms the funding decisions every priority needs; not an outcome programme.
Scale of benefit
Who pays for every priority.
Cross-domain effects
Distributional trade-offs recorded as harms.
Evidence
Modelled (HMRC, IFS).
Additionality
Method, not a service.
Cost and funding
Not a direct spending programme.
Deliverability
HM Treasury and Parliament.
Risk
Packages presented as additive.
A22 Reform property taxes without hiding the losers — Kept in its domain planRevenue-neutral, so it cannot finance other priorities; a strong distributional option kept on /revenue-opportunity.
Scale of benefit
Council tax still on 1991 values.
Cross-domain effects
Poverty and inequality: supported benefit and concentrated losers.
Evidence
Modelled (IFS, 2019 values).
Additionality
Revenue-neutral reform.
Cost and funding
Revaluation not costed.
Deliverability
Revaluation takes years.
Risk
Concentrated losers; transition relief.
A23 Buy measurable results and manage contracts through delivery — Selected (NP6)The delivery enabler NP6: the outcome priorities that spend money do so through contracts that need the same scarce commercial staff.
Scale of benefit
£461.3bn of public procurement; 34 of 172 major projects Red.
Cross-domain effects
Public service delivery and defence equipment, plausible.
Evidence
Audit and expert recommendation; no evaluated saving.
Additionality
Applies Procurement Act duties to named contracts.
Cost and funding
Commercial staff not costed.
Deliverability
Capability-limited.
Risk
Box-ticking; supplier withdrawal.
A24 Improve pension saving while protecting low take-home pay — Kept in its domain planThe pensions candidate: benefits arrive over decades and it carries a recorded harm now.
Scale of benefit
Future retirement incomes.
Cross-domain effects
Pensions (supported) with a take-home-pay harm.
Evidence
Modelled IFS package.
Additionality
External proposal; the Pensions Commission has not reported.
Cost and funding
Costed: £5bn employers and £6bn employees a year; relief up to £3.7bn.
Deliverability
Legislation and phasing.
Risk
Opt-outs; squeezes pay now.
A25 Help eligible people claim support they already qualify for — Kept in its domain planFlagged near-missWorthwhile and quick to start, but untested and single-area.Flagged for the next selection: Ties with NP2 on the base weighting: quick, well targeted and within DWP’s powers, but outreach is untested and higher take-up raises benefit spending that is not budgeted. Recommended as a bounded test on /pensions and /benefits.
Scale of benefit
Up to £2.5bn of Pension Credit went unclaimed in 2023–24 (one year’s estimate).
Cross-domain effects
Pensioners only.
Evidence
Untested outreach.
Additionality
Campaigns already run.
Cost and funding
Up to £2.5bn a year if all claimed; outreach not costed.
Deliverability
DWP and councils.
Risk
Treating estimates as individual eligibility.
A26 Fund military readiness before counting new announcements — Kept in its domain planThe resilience candidate: impact cannot be assessed publicly because readiness data are withheld.
Scale of benefit
Readiness in doubt at the last public assessment (2024).
Cross-domain effects
Defence only.
Evidence
Expert; readiness data classified.
Additionality
The Defence Investment Plan is funded; readiness gaps are not public.
Cost and funding
Within MOD’s settlement; gaps not costed.
Deliverability
MOD.
Risk
Cost growth; long lead times.
A27 Publish a fuller, auditable monarchy cost statement — Supports governance or deliveryAccountability action on monarchy costs.
Scale of benefit
Security costs undisclosed.
Cross-domain effects
Improves how decisions are made and scrutinised; no outcome area is claimed.
Evidence
Untested proposal.
Additionality
Transparency or process change, not a service.
Cost and funding
Mostly staff time (not zero) or not yet costed; not a public-spending programme.
Deliverability
Within the named institution’s current powers, or needs legislation.
Risk
Low direct risk; misreading or capture of the process.
A28 Test the royal grant against an explicit public purpose — Supports governance or deliveryAccountability action; the Grant reset is already announced.
Scale of benefit
Grant formula linked to Crown Estate profits.
Cross-domain effects
Improves how decisions are made and scrutinised; no outcome area is claimed.
Evidence
Expert or audit recommendation.
Additionality
Transparency or process change, not a service.
Cost and funding
Mostly staff time (not zero) or not yet costed; not a public-spending programme.
Deliverability
Within the named institution’s current powers, or needs legislation.
Risk
Low direct risk; misreading or capture of the process.
A29 Reduce food-trade friction through a clearly scoped agreement — Kept in its domain planDepends on a negotiation already under way; the outcome is not in UK hands alone.
Scale of benefit
Goods exports to the EU −14% in volume since 2019.
Cross-domain effects
Trade, with an autonomy trade-off.
Evidence
Observational (CEP).
Additionality
Government policy under negotiation.
Cost and funding
Partial FSA funding; implementation not costed.
Deliverability
Depends on an EU agreement.
Risk
Not concluded; exceptions.
A30 Make the remaining border process usable for small exporters — Further evidence neededUntested; sequenced after the SPS outcome.
Scale of benefit
Border costs for small exporters.
Cross-domain effects
Trade.
Evidence
Untested proposal.
Additionality
Pilot.
Cost and funding
Not costed.
Deliverability
DBT and HMRC.
Risk
Little change without SPS.
A31 Teach people to recognise manipulation, then test retention — Supports governance or deliveryAccountability and informed-choice actions: the brief bars a manufactured national spending programme for a page about a democratic choice, and none claims an outcome in another area.
Scale of benefit
Manipulation online.
Cross-domain effects
Improves how decisions are made and scrutinised; no outcome area is claimed.
Evidence
US experiments; UK effect untested.
Additionality
Transparency or process change, not a service.
Cost and funding
Mostly staff time (not zero) or not yet costed; not a public-spending programme.
Deliverability
Within the named institution’s current powers, or needs legislation.
Risk
Low direct risk; misreading or capture of the process.
A32 Test independent support for public-interest local reporting — Supports governance or deliveryAccountability and informed-choice actions: the brief bars a manufactured national spending programme for a page about a democratic choice, and none claims an outcome in another area.
Scale of benefit
Local reporting gaps.
Cross-domain effects
Improves how decisions are made and scrutinised; no outcome area is claimed.
Evidence
Untested proposal.
Additionality
Transparency or process change, not a service.
Cost and funding
Mostly staff time (not zero) or not yet costed; not a public-spending programme.
Deliverability
Within the named institution’s current powers, or needs legislation.
Risk
Low direct risk; misreading or capture of the process.
A33 Give electoral reform an informed public decision process — Supports governance or deliveryAccountability and informed-choice actions: the brief bars a manufactured national spending programme for a page about a democratic choice, and none claims an outcome in another area.
Scale of benefit
Votes and seats diverge.
Cross-domain effects
Improves how decisions are made and scrutinised; no outcome area is claimed.
Evidence
Expert recommendation on process only.
Additionality
Transparency or process change, not a service.
Cost and funding
Mostly staff time (not zero) or not yet costed; not a public-spending programme.
Deliverability
Within the named institution’s current powers, or needs legislation.
Risk
Low direct risk; misreading or capture of the process.
A34 Compare voting systems without pretending votes stay unchanged — Supports governance or deliveryAccountability and informed-choice actions: the brief bars a manufactured national spending programme for a page about a democratic choice, and none claims an outcome in another area.
Scale of benefit
Readers misread fixed-vote comparisons.
Cross-domain effects
Improves how decisions are made and scrutinised; no outcome area is claimed.
Evidence
Untested editorial method.
Additionality
Transparency or process change, not a service.
Cost and funding
Mostly staff time (not zero) or not yet costed; not a public-spending programme.
Deliverability
Within the named institution’s current powers, or needs legislation.
Risk
Low direct risk; misreading or capture of the process.
A35 Make the political-money trail timely and enforceable — Supports governance or deliveryAccountability and informed-choice actions: the brief bars a manufactured national spending programme for a page about a democratic choice, and none claims an outcome in another area.
Scale of benefit
Slow, weakly enforced disclosure.
Cross-domain effects
Improves how decisions are made and scrutinised; no outcome area is claimed.
Evidence
Expert or audit recommendation.
Additionality
Transparency or process change, not a service.
Cost and funding
Mostly staff time (not zero) or not yet costed; not a public-spending programme.
Deliverability
Within the named institution’s current powers, or needs legislation.
Risk
Low direct risk; misreading or capture of the process.
A36 Cost donation limits together with replacement party finance — Supports governance or deliveryAccountability and informed-choice actions: the brief bars a manufactured national spending programme for a page about a democratic choice, and none claims an outcome in another area.
Scale of benefit
Donor concentration.
Cross-domain effects
Improves how decisions are made and scrutinised; no outcome area is claimed.
Evidence
Expert recommendation; caps uncosted.
Additionality
Transparency or process change, not a service.
Cost and funding
Mostly staff time (not zero) or not yet costed; not a public-spending programme.
Deliverability
Within the named institution’s current powers, or needs legislation.
Risk
Low direct risk; misreading or capture of the process.
A37 Give every commitment an owner, budget and measurable result — Supports governance or deliveryUsed as the plan’s delivery discipline: each priority names an owner, a first decision, a milestone and a measure.
Scale of benefit
Commitments without owners or measures.
Cross-domain effects
Improves how decisions are made and scrutinised; no outcome area is claimed.
Evidence
NAO recommendation.
Additionality
Transparency or process change, not a service.
Cost and funding
Mostly staff time (not zero) or not yet costed; not a public-spending programme.
Deliverability
Within the named institution’s current powers, or needs legislation.
Risk
Low direct risk; misreading or capture of the process.
A38 Apply the same independent costing test to every manifesto — Supports governance or deliveryAccountability and informed-choice actions: the brief bars a manufactured national spending programme for a page about a democratic choice, and none claims an outcome in another area.
Scale of benefit
Manifestos costed on different bases.
Cross-domain effects
Improves how decisions are made and scrutinised; no outcome area is claimed.
Evidence
Expert recommendation (IFS).
Additionality
Transparency or process change, not a service.
Cost and funding
Mostly staff time (not zero) or not yet costed; not a public-spending programme.
Deliverability
Within the named institution’s current powers, or needs legislation.
Risk
Low direct risk; misreading or capture of the process.
A39 Make every poll traceable to its methods — Supports governance or deliveryAccountability and informed-choice actions: the brief bars a manufactured national spending programme for a page about a democratic choice, and none claims an outcome in another area.
Scale of benefit
Poll provenance.
Cross-domain effects
Improves how decisions are made and scrutinised; no outcome area is claimed.
Evidence
BPC disclosure rules.
Additionality
Transparency or process change, not a service.
Cost and funding
Mostly staff time (not zero) or not yet costed; not a public-spending programme.
Deliverability
Within the named institution’s current powers, or needs legislation.
Risk
Low direct risk; misreading or capture of the process.
A40 Test polling summaries against past errors before publishing them — Supports governance or deliveryAccountability and informed-choice actions: the brief bars a manufactured national spending programme for a page about a democratic choice, and none claims an outcome in another area.
Scale of benefit
Shared polling error.
Cross-domain effects
Improves how decisions are made and scrutinised; no outcome area is claimed.
Evidence
Observational (BPC review).
Additionality
Transparency or process change, not a service.
Cost and funding
Mostly staff time (not zero) or not yet costed; not a public-spending programme.
Deliverability
Within the named institution’s current powers, or needs legislation.
Risk
Low direct risk; misreading or capture of the process.
A41 Make university spin-out deals clearer and quicker — Further evidence neededAdoption of recommended terms is not yet measured.
Scale of benefit
Research reaching use.
Cross-domain effects
Innovation only.
Evidence
Expert review.
Additionality
Recommended terms partly adopted; £20m proof-of-concept money.
Cost and funding
Programme funding; process not costed.
Deliverability
Universities and DSIT.
Risk
Little effect on scale-up finance.
A42 Use public buying to test useful innovations — Further evidence neededConsidered for NP5 and NP6 but has no evaluation; kept on /public-procurement and /innovation.
Scale of benefit
Public buyers as first customers.
Cross-domain effects
Innovation and service delivery, unevaluated.
Evidence
Expert recommendation.
Additionality
Bounded trials.
Cost and funding
Not costed.
Deliverability
Commissioning bodies.
Risk
Pilots without buyers.
A43 Make cloud exit and continuity work in practice — Further evidence neededThe technology-resilience candidate: owned by individual organisations and untested at national level.
Scale of benefit
Dependence on two cloud providers.
Cross-domain effects
Resilience across sectors, unevaluated.
Evidence
CMA market investigation.
Additionality
CMA chose provider commitments.
Cost and funding
Not costed.
Deliverability
Each organisation.
Risk
Duplicated capacity costs.
A44 Measure useful AI adoption before scaling it — Further evidence neededAdoption rates are not outcomes.
Scale of benefit
AI adoption 34.9% of firms.
Cross-domain effects
Service delivery, with a workload risk.
Evidence
Untested; the main adoption scheme found no productivity effect.
Additionality
Pilots.
Cost and funding
Not costed.
Deliverability
Service owners.
Risk
Hidden review work.
A45 Sequence reforms that depend on the same people and infrastructure — Supports governance or deliverySupports delivery: its shared-capacity mapping is used in “What has to happen first”, not presented as a priority.
Scale of benefit
Projects competing for the same workers and grid.
Cross-domain effects
Improves how decisions are made and scrutinised; no outcome area is claimed.
Evidence
Untested proposal.
Additionality
Transparency or process change, not a service.
Cost and funding
Mostly staff time (not zero) or not yet costed; not a public-spending programme.
Deliverability
Within the named institution’s current powers, or needs legislation.
Risk
Low direct risk; misreading or capture of the process.
A46 Make policy models reproducible before expanding the workbench — Supports governance or deliverySupports the plan’s evidence: models must reproduce known results before use.
Scale of benefit
Unvalidated policy models.
Cross-domain effects
Improves how decisions are made and scrutinised; no outcome area is claimed.
Evidence
Green Book and Magenta Book guidance.
Additionality
Transparency or process change, not a service.
Cost and funding
Mostly staff time (not zero) or not yet costed; not a public-spending programme.
Deliverability
Within the named institution’s current powers, or needs legislation.
Risk
Low direct risk; misreading or capture of the process.
A47 Publish one funding ledger across all action plans — Supports governance or deliverySupports delivery: the national funding table reads the same ledger, counting each cost once.
Scale of benefit
Costs on different bases.
Cross-domain effects
Improves how decisions are made and scrutinised; no outcome area is claimed.
Evidence
Audit and appraisal guidance.
Additionality
Transparency or process change, not a service.
Cost and funding
Mostly staff time (not zero) or not yet costed; not a public-spending programme.
Deliverability
Within the named institution’s current powers, or needs legislation.
Risk
Low direct risk; misreading or capture of the process.
A48 Build a Hampshire plan around specific local needs and powers — Kept in its domain planHampshire delivers NP1 and NP4 locally through A02 and A18 allocations; its own pilot stays on /hampshire.
Scale of benefit
A 5.7-year life-expectancy gap inside Hampshire.
Cross-domain effects
Local: a geography, never an extra outcome area.
Evidence
Untested.
Additionality
Local plan.
Cost and funding
Not costed.
Deliverability
Split local powers; reorganisation under review.
Risk
Small samples.
A49 Set a minimum price per unit of alcohol in England — Selected (NP7)NP7: first on the selection table under every weighting — a within-UK controlled evaluation of deaths, fully additional in England and not a spending programme.
Scale of benefit
Alcohol deaths and admissions, concentrated in deprived areas: alcohol-specific deaths in England were 13.8 per 100,000 in 2024 (ONS).
Cross-domain effects
Health (a within-UK evaluation supports it); a recorded harm for some dependent drinkers on low incomes; no consistent effect on crime.
Evidence
Controlled interrupted time series, Scotland against England: deaths wholly attributable to alcohol −13.4% (95% CI −18.4% to −8.3%); admissions −4.1%, not significant.
Additionality
Fully additional: England has no floor price and ministers say they have no plans for one.
Cost and funding
Not a public-spending programme; administration not costed; drinkers pay more for the cheapest drinks, mostly to retailers.
Deliverability
Needs primary legislation and a government decision; enforced through existing licensing.
Risk
Financial strain for some dependent drinkers on low incomes; the floor eroding with inflation.
A50 Measure what school mental health teams achieve as they reach every school, and close the gap for children with greater needs — Further evidence neededA funded national service with no measured outcome yet: the action is to make it evaluable, not to scale an untested effect.
Scale of benefit
One in four 16–24-year-olds in England has a common mental health condition; teams cover 60% of pupils.
Cross-domain effects
Health only; education outcomes not measured.
Evidence
Early evaluation without a comparison group (perceived benefits only); classroom-wide programmes tested in trials mostly found no benefit.
Additionality
The roll-out is funded to 2029–30; adds a comparison-based evaluation and a route for greater needs.
Cost and funding
Programme commitment without a published amount; evaluation and specialist route not costed.
Deliverability
NHS England, DfE and schools can phase it within current plans.
Risk
Measuring contacts not outcomes; teams absorbing cases beyond their training.
A51 Offer reablement to more older people leaving hospital, and judge it by whether they are still at home three months later — Further evidence neededIts effect is not established, so a local, evaluated extension comes before any national claim.
Scale of benefit
5.7% of over-65s leaving hospital receive reablement.
Cross-domain effects
Health and care (one area); a staffing trade-off with long-term home care.
Evidence
NICE recommendation; Cochrane evidence too weak to support or refute an effect.
Additionality
Reablement is established and free for six weeks; adds a defined cohort and a 91-day comparison.
Cost and funding
Programme envelope (Better Care Fund); extra capacity not costed.
Deliverability
Workforce-limited; local plans.
Risk
Drawing staff from long-term home care; selecting people likely to recover anyway.
A52 Fit home adaptations faster for older people at high risk of falling, after an occupational-therapist hazard check — Selected (NP8)NP8: high-certainty trial evidence, and the targeting the evidence requires is exactly what the proposal adds. Its earlier exclusion as a locally delivered grant was withdrawn: the same is true of NP4.
Scale of benefit
Falls among older people at higher risk; adaptations took over six months in the last national review.
Cross-domain effects
Health and care (one area); housing adaptations are the means, not a second outcome.
Evidence
Cochrane 2023: rate of falls −38% for people at higher risk (high certainty); no effect when untargeted.
Additionality
The Disabled Facilities Grant exists (£723m in 2026–27); adds targeting, an OT check and a speed standard.
Cost and funding
Programme envelope; OT capacity not costed.
Deliverability
Housing authorities and OTs can act now; contractor and OT capacity limit pace.
Risk
Other applicants waiting longer within a fixed grant.
A53 Meter homes in water-stressed areas, with a social tariff that protects low-income households — Kept in its domain planFlagged near-missEvaluated in England and quick to show results, but benefits are in one area and the distributional cost needs a social tariff; it ties for sixth place on the base weighting and loses the published tie-break on cross-domain outcomes. Kept on /follow-the-money, threaded to the water findings on /growing-the-economy and /hampshire.Flagged for the next selection: Ties with NP2 on the base weighting and is in the top six under four of six weightings: UK quasi-experimental evidence of lower use within months, at little public cost. The tie-break (evidence, then distinct cross-domain outcomes) keeps NP2, the only candidate with evaluated benefits in two areas. Recommended for the next review once the new water regulator’s first price review is set.
Scale of benefit
A public water supply deficit of up to 5 billion litres a day by 2055 in England; ten of fourteen Environment Agency areas in drought and temporary use bans affecting around 30 million people in September 2026.
Cross-domain effects
Water supply (no separate policy area is counted); drought links to growth and Hampshire are the same outcome, not a second one.
Evidence
A UK household panel analysis of Southern Water’s universal metering (2010–15): use fell 18–22%; less affluent households about £10 a year worse off without protection.
Additionality
Companies have committed to smart meters for 50% of homes and businesses by 2030; adds metering of every home in designated areas and a protected social tariff.
Cost and funding
Bill-funded through regulated allowances; not costed here.
Deliverability
Existing powers in designated areas; needs the regulator’s decision and installer capacity.
Risk
Hardship for large low-income households; effects fading.
A54 Allocate flood-defence money by the new risk map, keep existing defences in condition, and report homes better protected every year — Kept in its domain planA real and growing risk, and allocation by the risk map adds a test the last programme lacked, but the evidence is an audit, not an evaluation, and defences take years; it does not beat the current outcome priorities on the published criteria. Kept on /housing as the adaptation decision with a data thread.
Scale of benefit
About 6.3 million properties in England in areas at risk of flooding, around 8 million by mid-century; the last capital programme expected to protect 40% fewer properties than planned.
Cross-domain effects
Housing (homes protected); adaptation and Hampshire links are the same outcome.
Evidence
NAO audit (2023) of the previous programme; no evaluation of risk-based allocation or reporting.
Additionality
The £4.2bn 2026–29 programme is funded and funding rules were consulted on; adds allocation by the 2024 risk map, protected maintenance and annual reporting of homes better protected.
Cost and funding
Programme envelope; the change itself needs staff time, not new capital.
Deliverability
Defra and the Environment Agency can decide under current powers; construction capacity limits delivery.
Risk
Counting homes “better protected” rather than flooding avoided; schemes delayed by costs.
A66 Commission childcare places where councils can show parents cannot use the funded hours — Kept in its domain planStandalone addition (NG1) placed on /education with the childcare chapter. Assessed in the 7 October 2026 review (shared assessment model); not part of the superseded 4 October slate, which is kept unchanged.
Scale of benefit
80% of eligible under-threes of working parents registered for funded hours in January 2026; 71.6% of eligible disadvantaged two-year-olds.
Cross-domain effects
Education (early years) and employment (parents’ work); the work link is indirect.
Evidence
Regression-discontinuity evidence on free hours and mothers’ work (indirect); a self-reported DfE process evaluation.
Additionality
Adds places in demonstrated local gaps, not more hours for everyone.
Cost and funding
Not costed: no published cost per commissioned place.
Deliverability
Councils hold the sufficiency duty; staff supply limits delivery.
Risk
Displacing existing places; paying for places that would open anyway.
A70 Close the gap in effective depression treatment where completion and recovery are lowest — Kept in its domain planStandalone addition (NW1) placed on /health. Assessed in the 7 October 2026 review (shared assessment model); not part of the superseded 4 October slate, which is kept unchanged.
Scale of benefit
674,683 finished Talking Therapies courses in 2025–26; reliable recovery 40.7% in the most deprived areas against 52.7% in the least.
Cross-domain effects
Health; work effects not found in the trial.
Evidence
One randomised trial (Norway) and uncontrolled English service data; NICE guidance.
Additionality
Targets extra capacity at the areas with the lowest completion and recovery.
Cost and funding
Not costed: no cost per course published.
Deliverability
NHS England and boards can act; therapist supply limits it.
Risk
Longer waits between sessions.
A72 Bring effective stop-smoking treatment to routine and manual workers — Kept in its domain planStandalone addition (NW3) placed on /health. Assessed in the 7 October 2026 review (shared assessment model); not part of the superseded 4 October slate, which is kept unchanged.
Scale of benefit
16.6% of routine and manual workers in England smoked in 2025, against 5.9% in managerial and professional jobs.
Cross-domain effects
Health; inequality in healthy years plausible.
Evidence
High-certainty trial evidence on treatments; outreach itself untested.
Additionality
Adds targeted outreach to routine and manual smokers.
Cost and funding
Not costed.
Deliverability
Councils can act within the ring-fenced grant.
Risk
Counting self-reported quits.
A67 Appraise one bus corridor where cost, hours or reliability stop people reaching jobs — Kept in its domain planRelated option (NG2) within city-region growth, placed on /growing-the-economy. Assessed in the 7 October 2026 review (shared assessment model); not part of the superseded 4 October slate, which is kept unchanged.
Scale of benefit
Bus journeys in England at 90% of their 2019–20 level; fares up 7% against 3% inflation in the year to March 2025.
Cross-domain effects
Growth and employment; association only.
Evidence
Observational British studies; the fare-cap evaluation measured patronage, not jobs.
Additionality
Appraises one corridor against a comparison.
Cost and funding
Not costed.
Deliverability
Combined authorities hold the powers.
Risk
Displacement between areas.
A69 Test management and adoption support for existing firms against a comparison — Kept in its domain planRelated option (NG4) within research-to-adoption, placed on /innovation. Assessed in the 7 October 2026 review (shared assessment model); not part of the superseded 4 October slate, which is kept unchanged.
Scale of benefit
Mean management practices score 0.55 in 2023; national adoption schemes found no productivity effect.
Cross-domain effects
Productivity (growth).
Evidence
A randomised trial abroad; UK evaluations finding no effect.
Additionality
Adds a comparison and a productivity outcome to business support.
Cost and funding
Not costed.
Deliverability
BIST and mayoral authorities.
Risk
Another null result.
A73 Get Access to Work decisions made in weeks for people at risk of leaving work — Kept in its domain planRelated option (NE2) within work that pays, placed on /employment. Assessed in the 7 October 2026 review (shared assessment model); not part of the superseded 4 October slate, which is kept unchanged.
Scale of benefit
Access to Work decisions averaged 109 working days in November 2025 against a 25-day target.
Cross-domain effects
Employment; disability employment gap 29.7 points.
Evidence
Audit and qualitative research; no counterfactual estimate.
Additionality
Adds a decision standard for in-work cases with a comparison.
Cost and funding
Not costed.
Deliverability
DWP can act; backlog clearance by September 2027.
Risk
Longer waits for other applicants.
A74 Offer early arrears and legal advice to renters before a notice becomes homelessness — Kept in its domain planRelated option (NE3) within settled homes for families, placed on /housing. Assessed in the 7 October 2026 review (shared assessment model); not part of the superseded 4 October slate, which is kept unchanged.
Scale of benefit
54.7% of prevention duties ended with a secure home in January–March 2026; 135,580 households in temporary accommodation.
Cross-domain effects
Housing; child poverty plausible.
Evidence
An English quasi-experimental evaluation and US trials.
Additionality
Adds early advice for renters served notice or in arrears, with comparison councils.
Cost and funding
Not costed.
Deliverability
Councils and the Ministry of Justice.
Risk
Delaying rather than preventing homelessness.
A56 Hold network companies to the reformed connection timetable and publish time to connect — Kept in its domain planGrowth member and resilience queue item (G06), placed on /energy. Assessed in the 7 October 2026 review (shared assessment model); not part of the superseded 4 October slate, which is kept unchanged.
Scale of benefit
58% of the projects needed for 2030 had connection offers by June 2026; deadlines extended twice.
Cross-domain effects
Energy and growth.
Evidence
Ofgem’s ex-ante impact assessment only.
Additionality
Adds published time-to-connect and enforcement of the reformed timetable.
Cost and funding
Bill-funded; not costed.
Deliverability
Ofgem and NESO powers exist; network build limits it.
Risk
Slipping deadlines.
A60 Publish a costed care workforce and capacity plan alongside the Fair Pay Agreement — Kept in its domain planWellbeing member (W08), placed on /social-care. Assessed in the 7 October 2026 review (shared assessment model); not part of the superseded 4 October slate, which is kept unchanged.
Scale of benefit
111,000 vacant care posts and 23.1% turnover in 2024/25; 470,000 more posts needed by 2040.
Cross-domain effects
Social care and health.
Evidence
Econometric estimates of pay and employment (an overstatement by the authors’ account).
Additionality
Adds a costed plan linking pay to capacity.
Cost and funding
Not costed beyond a £500m first-year envelope.
Deliverability
DHSC and HM Treasury.
Risk
A plan without funding.
A61 Pick the specialties where diagnostics and theatre time hold waits back, and add capacity there — Kept in its domain planWellbeing member (W10), placed on /nhs. Assessed in the 7 October 2026 review (shared assessment model); not part of the superseded 4 October slate, which is kept unchanged.
Scale of benefit
65.4% of pathways within 18 weeks and 23.1% waiting six weeks or more for a test in July 2026.
Cross-domain effects
Health and work.
Evidence
A synthetic-control study of surgical hubs; descriptive productivity analysis.
Additionality
Adds capacity where one constraint binds, by specialty.
Cost and funding
Not costed.
Deliverability
NHS England and trusts.
Risk
Displacement between specialties.
A77 Assess overheating in care homes and fund building-specific cooling where risk is highest — Kept in its domain planResilience queue item (C39), placed on /social-care. Assessed in the 7 October 2026 review (shared assessment model); not part of the superseded 4 October slate, which is kept unchanged.
Scale of benefit
677 heat-associated deaths in care homes in summer 2025.
Cross-domain effects
Health and social care.
Evidence
Modelling of three care homes; no trial.
Additionality
Adds assessment and building measures in existing care homes.
Cost and funding
Not costed.
Deliverability
DHSC, UKHSA and providers.
Risk
Measures unsuited to buildings.
A62 Close capital-tax reliefs before raising rates, with an official costing — Kept in its domain planEquality member (E03), placed on /revenue-opportunity. Assessed in the 7 October 2026 review (shared assessment model); not part of the superseded 4 October slate, which is kept unchanged.
Scale of benefit
Business asset disposal relief about £1.0bn (2023–24); agricultural and business property relief about £1.9bn (2022–23).
Cross-domain effects
Public finances and inequality.
Evidence
Official costings of rate rises; independent costings of packages.
Additionality
Adds official costings of closing reliefs before rate rises.
Cost and funding
A revenue measure; yield not costed officially.
Deliverability
HM Treasury and Parliament.
Risk
Behaviour reducing yield.
A55 Raise completion in construction apprenticeships in the shortage trades — Kept in its domain planGrowth member (G02), placed on /education. Assessed in the 7 October 2026 review (shared assessment model); not part of the superseded 4 October slate, which is kept unchanged.
Scale of benefit
Construction apprenticeship achievement rate 58.3% in 2024/25, the lowest of any subject area.
Cross-domain effects
Education, housing and growth.
Evidence
DfE returns comparing completers with non-completers.
Additionality
Adds completion support in shortage trades.
Cost and funding
Not costed.
Deliverability
DfE and Skills England.
Risk
Paying for completions that would happen anyway.
A65 Target minimum-wage enforcement where underpayment is highest, and test whether it works — Kept in its domain planEquality member (E10), placed on /employment. Assessed in the 7 October 2026 review (shared assessment model); not part of the superseded 4 October slate, which is kept unchanged.
Scale of benefit
About 445,000 jobs paid below the minimum in 2025; arrears identified for 25,200 workers in 2024/25.
Cross-domain effects
Employment and inequality.
Evidence
Survey estimates and enforcement statistics; no evaluation of deterrence.
Additionality
Adds targeting and an evaluation of enforcement.
Cost and funding
Not costed.
Deliverability
The Fair Work Agency.
Risk
Shifting underpayment between sectors.