Why Pupil Premium Spending Usually Fails: A Systems Diagnosis

Schools receive pupil premium funding with a clear mandate: close the attainment gap.

But across England, the gap persists.

Not because schools don’t try. Schools spend billions on pupil premium provision. The problem is that what schools spend it on doesn’t match what the evidence shows actually works.

The typical approach:

• More tutoring (expensive, inconsistent delivery)
• Intervention programmes (often short-term, disconnected from classroom teaching)
• Technology solutions (high cost, low adoption)
• Additional staffing (creates dependency, not independence)

Schools expect these to move the needle. They rarely do.

The research tells a different story about what actually closes gaps.

What the Evidence Actually Shows

The Education Endowment Foundation (EEF) has spent over a decade studying what works in real schools. Their conclusions are stark.

The strategies with the largest effect sizes are not expensive. They do not require specialist providers. They require consistency.

The EEF Teaching and Learning Toolkit identifies the following as among the highest-impact approaches for closing attainment gaps:

Metacognition and self-regulated learning — +7 months progress Teaching pupils to understand how they learn: to plan their work, attempt it, and monitor their own progress. Cost: low. Training: moderate. Dependency created: none — pupils own the skill.

Feedback and formative assessment — +8 months progress Specific, actionable feedback given regularly. Not summative grades. Not generic praise. Feedback that tells pupils what they did well, what didn’t work, and what to try next. Cost: low. Dependency created: pupils use feedback to improve.

Collaborative learning and peer tutoring — +5 months progress Pupils learning from each other in structured ways. Cost: minimal. Dependency created: none — pupils become the resource.

Reading and language interventions — +4–6 months progress Targeted reading and comprehension support in the early years. Addressing reading gaps early removes one of the biggest barriers to progress across every subject.

Parental engagement — +4 months progress Clear, consistent communication between school and home about what pupils are learning and how families can support it.

What these five strategies share: they are not add-ons. They are changes to how teaching and learning work, every day, for every pupil.

Why Schools Get This Wrong

The typical pupil premium response — tutoring, intervention programmes, technology — makes intuitive sense. You identify a gap. You add a resource to fill it. The gap closes.

Except it usually doesn’t. Here is why.

Most pupil premium interventions run as separate programmes:

• Intervention time three times per week
• A specialist or teaching assistant delivers it
• It is disconnected from what happens in the classroom
• When the programme ends, the skill does not transfer

For something very specific — a speech and language programme targeting a particular need, for example — this can work. But for closing the general attainment gap, isolated interventions face two problems the evidence is consistent on.

Transfer is unreliable. A pupil develops a skill during intervention time but does not apply it in lessons, because the context is different and the habit has not been built in the place where it matters.

Dependency is structural. The pupil improves while the support is running. When it ends, progress stalls. The school has not built a system — it has built a reliance on a resource that is finite and expensive.

Meanwhile, the most advantaged pupils in the same school are developing metacognitive habits, receiving regular feedback, practising structured collaboration, and building reading fluency — as part of the ordinary flow of lessons. The gap is being maintained by the difference in daily experience, not closed by occasional intensive support.

The Mechanism: Why System-Level Strategies Work

Understanding why these strategies close gaps helps explain why they have to be embedded, not added.

Metacognition works because disadvantaged pupils often lack the self-monitoring habits that more advantaged pupils develop at home — the instinct to check their own understanding, to recognise when they are stuck, to adjust their approach. When explicitly taught and practised in every lesson, those habits become automatic. The gap narrows because the daily experience of learning changes.

Feedback works because it transfers responsibility for improvement from the teacher to the pupil. Instead of waiting for an adult to identify and address a gap, the pupil can see the gap and act. This scales in a way that adding adult support does not.

Collaborative learning works because it removes the dependency on adult expertise. Pupils learn from each other. A disadvantaged pupil who may lack academic peer models at home encounters them in the classroom — every day, structured, with clear roles and expectations.

Reading interventions work because reading fluency is a leverage point. A pupil who cannot read fluently faces compounding difficulty across every subject. Addressing this early — systematically, for all pupils — removes a root cause rather than managing its symptoms.

Parental engagement works because it extends the learning environment. Families who know what is being taught and what they can do to support it create more practice opportunities at home. This does not require parents to be experts. It requires the school to communicate clearly and consistently.

Why Systems Matter More Than Individual Interventions

Schools that close gaps do not run better intervention programmes. They embed these strategies into how teaching works for every pupil, every day.

That means:

• Lessons are structured to include explicit metacognitive practice — planning, doing, reviewing
• Feedback is regular and specific, not concentrated in summative assessments
• Pupils work in structured groups with clear roles, not in loosely organised pairs
• Reading is taught systematically from the beginning, so gaps do not compound
• Communication with families is consistent and curriculum-linked, not reactive

When these are part of the system — used with all pupils, in all lessons — disadvantaged pupils access them daily rather than in isolated intervention slots.

That is when gaps close.

The HOD or SLT lead reviewing book scrutiny data is often looking at the output of this problem without seeing the cause. The gap in outcomes between disadvantaged pupils and their peers frequently reflects a gap in the quality and consistency of daily teaching experience — not a lack of additional support.

The Problem With Adding Resources

Most schools cannot justify the shift to system-level change. Not because the evidence is unclear — it is not — but because changing how teaching works requires changing what teachers do in every lesson. Adding a tutor is an administrative decision. Changing systems is a CPD and culture question.

There is also a budget logic that points the wrong way. Pupil premium funding arrives as a per-pupil allocation. Spending it on identifiable provision for identifiable pupils feels accountable. Spending it on improving the quality of teaching for all pupils feels harder to justify — even though the latter is what the evidence supports.

The DfE and Ofsted are both clear on this. The pupil premium is not ring-fenced for separate provision. It is intended to improve outcomes. Schools have flexibility. The accountability question is whether it worked — not whether it was spent on a named intervention.

What This Means for Your School

If the attainment gap in your school is not closing despite consistent pupil premium spending, the diagnostic question is this: what is the daily experience of a disadvantaged pupil in your classrooms?

Do all pupils — not just those identified as pupil premium — experience:

• Explicit metacognitive practice in every lesson (planning, reflection, self-monitoring)?
• Regular, specific, actionable feedback on their work?
• Structured peer collaboration with clear roles?
• Systematic reading instruction (primary) or comprehension support (secondary)?
• Clear communication between school and home about what is being taught?

If the answer to any of these is “not consistently,” that is where the pupil premium strategy should focus — not on adding provision, but on building the systems that make these things happen reliably for every pupil, every day.

Tools that make consistency easier to achieve are part of that system. Pupil planners structured around metacognitive habits — with prompts for planning, reflection, and self-assessment built into every lesson — give pupils the scaffold that makes daily practice possible without adding to teacher workload.

The goal is not more support. It is a better system that disadvantaged pupils access as a matter of course.

That is how gaps close.

Frequently Asked Questions

Does this mean we should stop running intervention programmes entirely?

Not necessarily — but it does mean interrogating what each programme is actually for. A well-designed reading intervention in Year 1 or a targeted speech and language programme addresses a specific, bounded need. That is different from running general booster groups for pupils who are behind because their daily classroom experience is not giving them enough. The question to ask about any intervention is: what happens when it ends? If the answer is “the pupil is on their own again,” the intervention is managing a gap rather than closing it.

Our pupil premium strategy is reviewed by Ofsted. Won't they expect to see named provision?

Ofsted’s current framework does not require schools to demonstrate named provision for pupil premium pupils. Inspectors are looking for evidence that the funding is improving outcomes — and they are increasingly sceptical of provision maps that list activities without evidence of impact. A school that can show how it has embedded high-quality teaching practices across the whole curriculum, with evidence that disadvantaged pupils are making stronger progress as a result, is in a better position than one with an extensive provision map and a persistent gap. The Ofsted school inspection handbook is explicit that inspectors consider the quality of education for all pupils, not the volume of additional provision.

How do we justify spending pupil premium on whole-school CPD rather than direct pupil support?

The DfE guidance on pupil premium use is clear that funding can be spent on CPD that improves the quality of teaching, provided the school can demonstrate the impact on disadvantaged pupils’ outcomes. The strongest justification is the simplest one: if metacognition, feedback, and collaborative learning are the highest-impact strategies, and if teachers need CPD to deliver them well, then CPD is the most direct route to improving outcomes. Document the rationale, the training delivered, and the evidence of change in teaching practice and pupil progress. That is a more defensible spending decision than another round of tutoring with inconsistent attendance and inconclusive impact data.

What does "structured peer collaboration" actually look like in a lesson?

The key word is structured. Random groupwork — “discuss this with the person next to you” — produces variable results and tends to amplify existing differences, with more confident pupils dominating and quieter ones disengaging. Structured collaboration means defined roles (questioner, explainer, checker), a clear task with a concrete output, and an expectation that every pupil contributes and can be asked to report back. In practice this might be reciprocal reading pairs in English, think-pair-share with individual written responses in science, or designated group roles in a problem-solving task in maths. The mechanism is that every pupil is required to articulate their thinking — not just the ones who volunteer.

We're a primary school. Does this apply to us in the same way?

The principles apply, but the emphasis shifts. In primary, systematic phonics and early reading instruction are the single highest-leverage intervention — addressing reading fluency before it becomes a compounding barrier. Metacognitive strategies are appropriate from Key Stage 1 upwards when taught concretely (simple planning and reflection prompts rather than abstract frameworks). Parental engagement tends to have a stronger effect in primary, where the home learning environment has a more direct relationship with school progress, and where clear communication about what families can do is more immediately actionable. The mistake to avoid in primary is the same as in secondary: identifying a group of pupils as needing support and extracting them from quality whole-class teaching to deliver it.

What if teachers resist changing how they teach?

This is the real implementation question, and the post’s argument about systems is partly an answer to it. If the expectation of metacognitive practice, structured feedback, and collaborative learning lives only in the school’s pupil premium strategy document, teachers will continue doing what they have always done. Resistance to change is not primarily a motivation problem — it is a design problem. If the lesson structure, the exercise books, and the planning tools don’t make these practices the default, they will not happen consistently regardless of how much CPD has been delivered. The question for SLT is not “how do we persuade teachers to change?” but “what do we need to put in place so that the right practice is also the easiest practice?”

How long before we see the gap closing?

Honestly, longer than most schools expect — and the timescale matters for how you measure progress. A pupil premium strategy built around system-level change will not show dramatic results in a single term’s data. Metacognitive habits take time to build. Feedback loops compound over months. The appropriate measure is whether disadvantaged pupils are making stronger relative progress over a year, not whether the gap has closed in a half-termly data drop. Schools that abandon system-level approaches because they don’t show quick results are often returning to intervention programmes that produce short-term data movement without lasting change. Set the expectation internally before you start: this is a two-to-three year commitment, not a quick fix.