VIEW THIS AS

Auto mode follows the Route Engine until you choose a viewpoint.

YOU ARE HERE

ROUTE CHECK

CONNECTED TO

WHAT NEXT

Use the canonical route for this room, or HELP if you are unsure.

How Education Works | Education Programme Portfolio Reviews & Sunset Decisions — How Systems Decide What to Continue, Scale, Redesign, Merge or Stop

HEW-NODE-0137 · How Education Works · programme portfolio management, programme review, sunset clauses, de-implementation, continuation decisions, scaling, redesign, merger, termination, opportunity cost, spending review, budget reallocation, implementation evidence, impact evidence and institutional learning

Education systems are very good at starting things.

A pilot becomes a programme. A temporary grant becomes a permanent line item. A crisis response acquires staff, reporting templates and office space. A minister announces a new initiative. A later minister announces another. Schools receive new frameworks without old ones disappearing.

Years later, the system may be running dozens or hundreds of overlapping programmes whose original problem has changed, whose evidence is mixed, whose administrative burden is largely invisible and whose funding persists because stopping feels more dangerous than continuing.

A mature education system does not only ask how to launch a programme. It also asks when the programme has earned continuation, when it should change, and when its resources should move somewhere better.

This node sits beside the How Education Works hub, Education Public Expenditure Reviews & Spending Diagnostics, Education Cost-Effectiveness & Benefit-Cost Analysis, Education Impact Evaluation & Causal Inference, Education Policy Pilots & Scaling, Education Budget Formulation & MTEFs, Budget Execution & Public Expenditure Tracking and School Improvement Planning.

Those pages keep their jobs. Public Expenditure Reviews own sector-wide spending diagnosis. Cost-Effectiveness owns comparative value for money. Impact Evaluation owns causal effect. Policy Pilots owns bounded tests and scale-up. Budget Formulation owns fiscal envelopes. This node owns the lifecycle decision across the programme estate: how a system periodically reviews its existing interventions as a portfolio, compares strategic value and evidence, identifies duplication or obsolescence, and makes controlled continue, scale, redesign, merge, pause or stop decisions with explicit transition plans.

The 60-Second Read

  • Every programme consumes money, staff attention, reporting capacity and organisational bandwidth.
  • Continuation should be a decision, not the default state produced by last year’s budget.
  • A programme can be effective and still no longer be the best use of scarce resources.
  • A programme can be strategically important even when its measurable short-run effect is modest.
  • Portfolio review asks how programmes fit together, not only whether each one looks acceptable in isolation.
  • Duplication can occur in purpose, population, delivery channel, data collection, procurement or governance.
  • Sunset clauses create a scheduled decision point; they do not require automatic termination.
  • De-implementation means deliberately reducing or stopping low-value practices, not simply withdrawing support and hoping activity disappears.
  • Stopping has costs: contracts, staff redeployment, learner transition, data migration, public expectations and sunk assets.
  • Sunk cost is not a reason to continue future spending, but sunk assets can still have reuse value.
  • Exit decisions should distinguish programme failure from implementation failure.
  • Evidence should include impact, cost, equity, implementation quality, strategic fit, risk and opportunity cost.
  • Programmes should be compared against realistic alternatives, not against doing nothing when doing nothing is not the real choice.
  • Temporary crisis programmes need explicit normalisation or retirement rules.
  • Merging programmes can reduce duplication but can also create an oversized bureaucracy with blurred accountability.
  • Scaling a programme can change its cost structure and implementation quality.
  • Legacy reporting requirements should disappear when the programme disappears.
  • Decisions need political ownership because programme termination creates visible losers even when system value improves.
  • Reallocation should be traceable: savings should not vanish into an unspecified pool if the stated purpose was to fund a higher priority.
  • The goal is a living portfolio in which old commitments repeatedly compete with current needs.

One-Sentence Definition

Education programme portfolio review is the periodic, structured comparison of existing programmes against evidence, strategic priorities, cost, implementation quality, duplication, risk and alternative uses of resources in order to decide which programmes should continue, scale, redesign, merge, pause or end.

The First Distinction: Programme Review Is Not Audit

An audit can establish whether money was spent lawfully, controls operated and records are reliable. A programme can pass audit and still be ineffective, strategically obsolete or duplicative.

Portfolio review asks whether the programme still deserves to exist in its current form.

The Second Distinction: Spending Review Is Not Programme Lifecycle Management

Spending reviews examine expenditure to identify savings, efficiency and reallocation. Programme lifecycle management goes further into delivery architecture: what happens to staff, beneficiaries, contracts, data, assets, standards, responsibilities and learning when a programme changes state.

A spending review may recommend closing a grant. Portfolio management owns the operational question of how that closure occurs without damaging essential functions.

The Third Distinction: Sunset Is Not Automatic Death

A sunset clause creates an expiry or review point unless the programme is renewed. Its purpose is to force reconsideration rather than make permanence the default.

A strong programme can be renewed. A weak programme can be redesigned. A programme whose job has been absorbed elsewhere can be merged. The useful part is the mandatory decision.

Current OECD Practice Treats Spending Review as Reallocation Machinery

Government at a Glance 2025 reports that spending reviews have become a common instrument for systematic analysis of existing expenditure, identifying savings and reallocations and improving policy effectiveness. In the OECD’s 2023 survey, 34 of 35 responding countries reported having used spending reviews, and more than half were conducting them annually.

OECD’s 2025 Quality Budget Institutions report goes further: regular reviews can examine a programme, an institutional portfolio or a cross-cutting spending area and can eliminate duplication or reallocate resources toward emerging priorities rather than applying indiscriminate cuts.

Education Needs the Same Discipline Because Programme Accumulation Is Real

Schools experience programme accumulation as calendars, forms, meetings, training, data submissions and overlapping messages. A central ministry may see ten separate initiatives. A school sees one staff body trying to execute all ten at once.

Portfolio review therefore measures administrative load and interaction, not only budget.

Start With a Complete Programme Inventory

Systems often cannot review what they cannot see. Build an inventory containing:

  • programme name and identifier;
  • owner;
  • problem statement;
  • target population;
  • legal basis;
  • start date;
  • planned end or review date;
  • annual and multi-year cost;
  • staffing;
  • contracts;
  • delivery organisations;
  • reporting requirements;
  • key outcomes;
  • evaluation evidence;
  • overlapping programmes;
  • dependencies;
  • current decision state.

The inventory often reveals more fragmentation than the budget structure does.

One Programme Can Have Several Names

A central grant, a branded initiative and a training programme may all fund the same underlying mechanism. Conversely, one programme name may contain several unrelated components.

Map the mechanism, not only the brand.

Define the Programme’s Current Job

A programme launched ten years ago to solve low broadband access may now operate in a system where connectivity is nearly universal but device maintenance is the real problem. The original rationale can outlive the original environment.

Every review should restate the problem in current terms and ask whether the programme still addresses it.

Separate Mission Importance From Programme Performance

A low-performing literacy programme does not mean literacy is unimportant. It means this mechanism may not be the best route to the mission.

When programmes become symbols of their mission, stopping them can feel like abandoning the mission. Portfolio review should separate the two explicitly.

Build a Decision Matrix

  • Strategic fit: does the programme address a current priority?
  • Need: is the underlying problem still material?
  • Impact: what causal or outcome evidence exists?
  • Implementation: is the programme delivered as intended?
  • Cost: what resources does it consume?
  • Cost-effectiveness: what outcome is produced for that cost?
  • Equity: who gains and who is missed?
  • Duplication: does another programme perform the same function?
  • Risk: what happens if the programme stops?
  • Opportunity cost: what better use exists for the resources?
  • Adaptability: can redesign plausibly repair weakness?
  • Exit complexity: how difficult is controlled closure?

Do Not Collapse the Matrix Into One Magic Score

A weighted score can support comparison but can hide critical thresholds. A safeguarding programme should not survive simply because low cost compensates for weak safety. A legally mandated function may require continuation even if a particular delivery model performs poorly.

Use hard gates where some dimensions are non-compensable.

Continuation Should Have Reasons

“It already exists” is not a policy rationale. Renewal notes should state what evidence justifies continued funding and what conditions will trigger the next review.

Scale Requires a New Decision

A programme that works in 50 schools has not automatically earned national expansion. Scale changes staffing, supervision, procurement, local adaptation and market conditions.

EEF’s current scaling framework treats readiness as multidimensional: programme development, feasibility, organisational capacity, policy context, market conditions and scaling strategy all matter.

Redesign Is Often Better Than Binary Continue/Stop

A programme can contain an effective core wrapped in expensive administration. It can serve the right population through the wrong delivery channel. It can have strong uptake but weak targeting.

Redesign asks which mechanism deserves preservation and which architecture can change.

Merging Can Remove Duplicated Interfaces

Two teacher-support programmes may have separate portals, reporting forms, contracts and regional teams while serving the same schools. A merger can reduce administrative burden and create one coherent route.

But merger can also bury distinct expertise inside a large generic programme. Preserve specialised functions where integration would reduce quality.

Pause Is a Legitimate Programme State

When evidence is uncertain, budgets are temporarily constrained or major policy architecture is changing, a pause can preserve options while preventing continued spending. A pause needs rules for staff, contracts, beneficiaries and restart conditions.

Stop Is Sometimes the Correct Improvement Strategy

Systems often speak about improvement as adding. Yet removing ineffective or duplicative work can release more instructional and managerial capacity than adding another intervention.

De-implementation is not failure. It can be evidence-informed improvement.

EEF Treats Sustain as a Decision Phase, Not Eternal Continuation

The Education Endowment Foundation’s 2024 School’s Guide to Implementation describes implementation as an Explore → Prepare → Deliver → Sustain process and explicitly places monitoring, adaptation and decisions about next steps inside the sustain phase.

Its September 2026 guidance on evaluating and sustaining strategy similarly emphasises revisiting intended outcomes and adapting implementation rather than treating launch as completion.

Sunset Clauses Create Memory Against Institutional Inertia

A programme with an explicit review date forces future decision-makers to ask whether assumptions still hold. Without such a date, a temporary intervention can become permanent through administrative momentum.

Set Review Dates According to the Mechanism

A one-year attendance initiative can be reviewed quickly. A teacher-pipeline reform may require several years before downstream outcomes appear. Review timing should reflect the causal chain rather than a universal calendar.

Do Not Wait for Final Outcomes to Review Implementation

If a programme is clearly not reaching schools, waiting five years for final impact data wastes time. Use early implementation gates and later outcome gates.

Define Evidence Thresholds Before the Renewal Decision

If programme owners learn the decision criteria only after results arrive, criteria can shift to defend continuation. Establish expected evidence, minimum implementation, cost ceilings or outcome thresholds earlier where feasible.

Programme Owners Should Not Be Sole Reviewers

Teams build identity, expertise and employment around programmes. That does not make them biased by default, but it creates understandable incentives. Portfolio review should include independent finance, evaluation or policy challenge.

Include Frontline Burden

A programme can cost the central budget only $2 million while consuming thousands of teacher hours in data entry, meetings and training. Measure distributed implementation cost.

Measure the Opportunity Cost of Attention

Leadership time is finite. A school asked to implement six priorities may implement none deeply. Portfolio review should test whether programmes compete for the same people and calendar windows.

Map Programme Interactions

  • complementary programmes that strengthen each other;
  • substitutes competing for the same need;
  • duplicates collecting the same data;
  • contradictory programmes giving different instructions;
  • dependencies where one programme requires another;
  • shared suppliers or staff creating concentration risk.

Duplication Is Not Always Waste

Two programmes can intentionally provide redundancy for resilience or serve different populations using similar mechanisms. The question is whether overlap is deliberate and valuable.

Brand Proliferation Can Hide Mechanism Duplication

Three named initiatives may all fund teacher coaching. A portfolio view groups them by function and asks whether separate brands, contracts and reporting systems are justified.

Temporary Crisis Programmes Need a Return Path

A pandemic response, refugee-support grant or emergency device scheme may be essential during crisis. As conditions change, the system should decide whether to institutionalise the capability, integrate it elsewhere or retire it.

Emergency programmes should not become permanent merely because dismantling is awkward.

Sunk Cost Should Not Dominate Future Choice

Money already spent cannot be recovered. The relevant question is whether the next dollar and next staff hour are justified.

However, existing assets can have salvage value: a platform, training material or regional team may be reusable by another programme.

Legacy Systems Need Migration Plans

Stopping a programme can strand data, user accounts, contracts or reporting history. Define archival ownership, data retention, migration and access before closure.

Contracts Create Exit Constraints

Multi-year technology licences, leases and service contracts can make rapid termination costly. Portfolio review should know break clauses, renewal dates and termination liabilities before deciding timing.

Staff Are Not Line Items

Programme closure affects people with accumulated expertise. Redeployment, retraining, consultation and employment law can determine whether useful capability is preserved or lost.

Beneficiary Transition Is Part of Ethical Closure

Students receiving a scholarship, therapy support or transport service cannot simply disappear from the programme because a budget line ends. Define grandfathering, handover and alternative support.

Grandfathering Can Protect Current Learners

A programme can close to new entrants while continuing for existing cohorts until a natural transition point. This reduces disruption while still ending future commitments.

Closure Can Be Phased

Stop new contracts first, then new enrolments, then delivery, then data operations. A phased plan can preserve service continuity and prevent rushed disposal.

Communicate the Mission Continuity

If a programme supporting disadvantaged readers closes because a better mechanism is replacing it, explain that the mission remains. Otherwise beneficiaries may interpret closure as withdrawal of concern.

Do Not Call Every Closure an Efficiency

Stopping a programme can reduce cost and worsen outcomes. Savings are not efficiency unless similar or better value is preserved.

Do Not Call Every Continuation Stability

Continuing low-value programmes can create instability elsewhere by consuming money and attention needed for core functions.

Opportunity Cost Makes Portfolio Review Necessary

New priorities often arrive without new fiscal space. If every existing programme is protected, the only way to fund innovation is higher spending or thinner funding across everything.

Portfolio review creates explicit competition between old commitments and new needs.

Link Review to the Budget Calendar

OECD guidance repeatedly emphasises that spending review influences decisions only when findings arrive before budget formulation. The same rule applies to programme review: a beautifully reasoned report released after budgets are locked becomes archival knowledge.

Create a Reallocation Ledger

If a programme closes and releases $15 million, record where the fiscal space goes. Reallocation transparency strengthens trust and tests whether portfolio review actually shifts resources toward stated priorities.

Use Review Tiers

  • Light annual check: need, cost, major implementation issue.
  • Periodic strategic review: evidence, duplication, strategic fit, alternatives.
  • Deep review: causal evaluation, cost-effectiveness, organisational redesign and exit planning.
  • Trigger review: major cost growth, legal change, evidence of harm, system shock or technological obsolescence.

Not Every Programme Needs the Same Evidence

A statutory safeguarding function may be justified primarily by legal duty and risk reduction. A voluntary enrichment programme may need stronger comparative value evidence. Match evidence requirements to programme purpose.

Portfolio Balance Matters

A system can overinvest in remediation while underinvesting in prevention, or in pilots while underinvesting in routine operations. Review the mix of spending across prevention, core service, support, innovation and recovery.

Innovation Needs Space Created by Retirement

Without stopping anything, “innovation” often means adding another layer. A healthy portfolio deliberately releases resources from lower-value activity to fund new tests.

But Novelty Is Not a Priority by Itself

Old programmes can be excellent. New programmes can be fashionable and weak. Portfolio review should favour evidence and need, not age.

Case Study: The Three Literacy Initiatives

Invented example: a ministry funds three literacy programmes created under different governments. All train teachers to use structured small-group reading support, but each has its own portal, consultant network and reporting form.

Review finds similar mechanisms and overlapping schools. The system merges professional development and data collection while preserving one specialist component for multilingual learners. Administrative cost falls without deleting the underlying mission.

Case Study: The Programme That Worked but Was Still Stopped

Invented example: a summer mathematics camp produces modest positive effects at high cost. A school-year tutoring programme produces similar gains for one-third the cost and reaches more disadvantaged students.

The camp is not labelled a failure. It is phased out because a better alternative has emerged.

Case Study: The “Failed” Programme That Needed Redesign

Invented example: a career-guidance programme shows low student uptake. Interviews reveal students cannot access appointments during school hours and the digital booking system requires parental login.

The mechanism remains plausible. Delivery is redesigned before a stop decision is made.

Case Study: The Emergency Platform That Never Left

Invented example: a rapid remote-learning platform launched during school closures remains funded five years later even though most schools have returned to their main systems. The platform now duplicates identity, messaging and content functions.

Review preserves its emergency broadcast capability but retires the duplicated learning-management functions and archives old data under a controlled migration plan.

Case Study: The Grant With No Natural End

Invented example: a temporary grant supports schools during a curriculum transition. The curriculum becomes routine, but the grant persists because schools depend on it for ordinary staffing.

The system decides the need is real but the programme architecture is wrong. Funding is absorbed into the regular school formula and the temporary grant closes.

Failure Mode 1: Every Programme Reviewed Alone

Repair: compare overlapping functions and shared resource demands across the portfolio.

Failure Mode 2: Audit Pass = Continue

Repair: separate lawful execution from strategic value and impact.

Failure Mode 3: Sunk Cost Defends Future Spending

Repair: decide using future cost and value while identifying reusable assets.

Failure Mode 4: Stop Without Transition

Repair: build beneficiary, staff, contract, data and asset exit plans.

Failure Mode 5: Programme Owner Reviews Itself

Repair: include independent challenge and cross-portfolio comparison.

Failure Mode 6: Only Budget Cost Is Counted

Repair: include teacher time, leadership attention, reporting and compliance burden.

Failure Mode 7: Mission and Programme Are Confused

Repair: preserve the mission while allowing mechanisms to compete.

Failure Mode 8: Crisis Programme Becomes Permanent by Accident

Repair: include normalisation, integration or retirement checkpoints from launch.

Failure Mode 9: Merger Means One Giant Programme

Repair: merge duplicated interfaces while preserving specialised functions and accountable owners.

Failure Mode 10: Savings Disappear

Repair: maintain a reallocation ledger linking released resources to new priorities or fiscal targets.

The Programme Portfolio Review Operating Chain

  1. Build a complete programme inventory.
  2. Group programmes by function and target population.
  3. Map legal and policy obligations.
  4. Restate each programme’s current problem.
  5. Map theory of change.
  6. Collect implementation evidence.
  7. Collect impact and outcome evidence.
  8. Collect full financial cost.
  9. Estimate distributed administrative burden.
  10. Assess equity and reach.
  11. Identify duplicates and complements.
  12. Map dependencies.
  13. Compare cost-effectiveness where possible.
  14. Identify realistic alternatives.
  15. Assess strategic fit.
  16. Assess risk of stopping.
  17. Assess opportunity cost of continuing.
  18. Classify decision: continue, scale, redesign, merge, pause or stop.
  19. Obtain independent challenge.
  20. Make political and budget decision.
  21. Define conditions for renewal where continuing.
  22. Define scale conditions where expanding.
  23. Define redesign milestones where repairing.
  24. Define merger ownership where integrating.
  25. Define pause and restart triggers where holding.
  26. Build staff and beneficiary transition where stopping.
  27. Resolve contracts and assets.
  28. Archive or migrate data.
  29. Remove obsolete reporting requirements.
  30. Record resource reallocation.
  31. Schedule the next portfolio review.

A Programme Portfolio Dashboard

  • programme owner;
  • mission;
  • target population;
  • legal basis;
  • annual cost;
  • full implementation burden;
  • evidence rating;
  • impact estimate;
  • cost-effectiveness;
  • equity performance;
  • implementation health;
  • duplication flags;
  • dependency flags;
  • contract end dates;
  • sunset or review date;
  • decision state;
  • exit complexity;
  • resources released;
  • reallocation destination;
  • next review.

Canonical Owner Boundaries

This node owns the programme lifecycle across the existing portfolio: periodic continuation decisions, sunset review, de-implementation, redesign, merger, controlled closure and explicit reallocation of organisational capacity.

The Return Path

Return to a school calendar crowded with initiatives.

Every programme probably began with a reason. That is not the same as saying every programme still has the best reason today.

Education systems cannot create capacity only by hiring more people and raising budgets. They can also create capacity by retiring work that no longer earns its place.

That is the deeper purpose of portfolio review: not cutting for its own sake, but making yesterday’s commitments compete honestly with today’s learners.

A programme should become permanent because its function remains necessary and its design still earns the resources — not because nobody scheduled the moment when the system would have to decide again.

Return to the How Education Works hub.