6. Back to Basics – World-Class Education 🟢

Education is fully devolved. The last comparable international test of fifteen-year-olds placed Scotland at or near the OECD average in mathematics and science, below England, and well below its own early-2000s scores. Teacher-judged primary literacy and numeracy are at record highs on the domestic scale. Higher-stakes qualifications still show a wide poverty-related gap. The Curriculum Improvement Cycle is rewriting the technical framework for use from August 2028. The constraint is not a missing power. It is what the curriculum specifies, how it is assessed, and whether those choices are treated as settled.

Current reality

Legal and institutional position

School education, the curriculum, national qualifications, inspection, teacher education and the public funding of schools are devolved. The Education (Scotland) Act 1980 and subsequent statutes, including additional-support and qualifications legislation, sit with the Scottish Parliament. Local authorities employ most teachers and own most schools. Ministers set the National Improvement Framework, the Budget and the direction of Education Scotland and Qualifications Scotland. Qualifications Scotland took over from the Scottish Qualifications Authority in 2026. The Scottish Negotiating Committee for Teachers agrees pay and conditions, including class-contact time.

Curriculum for Excellence remains the statutory framework until the new technical materials replace Experiences and Outcomes and Benchmarks from August 2028 for early to third level, with later adoption in the senior phase. The Curriculum Improvement Cycle, led by Education Scotland, is the official vehicle for that change. Ministers have said the new curriculum will clarify the place of knowledge and skills and be benchmarked against international standards.

No one has a reserved veto over what a Scottish school teaches in mathematics, science or literacy. Higher education tuition for Scottish-domiciled students, teacher immigration, and some labour-market regulation sit outside or partly outside this item. They did not cause PISA mathematics to fall.

The latest official numbers

PISA 2022, published by the Scottish Government from the OECD assessment, remains the last completed international cycle. PISA 2025 results are due on 8 September 2026.

Scotland’s mean scores in 2022:

• Mathematics: 471 (OECD average 472). Lower than every Scottish PISA cycle from 2003 to 2018. The 2018 score was 489. England scored 492. 30.7 per cent of Scottish students were below Level 2; 7.7 per cent were at Level 5 or above.
• Science: 483 (OECD 485). Similar to 2018 (490), lower than 2006–2015. England scored 503.
• Reading: 493 (OECD 476). Above the OECD average, lower than 2018 (504) and lower than 2000. England scored 496.

The OECD treats roughly 20 points as about a year of schooling. The mathematics drop from 2018 to 2022 is 18 points; the drop from the mid-2000s is larger. The 2022 cohort was the first assessed almost entirely under Curriculum for Excellence. The decline in mathematics was already visible before 2020. The pandemic does not explain the earlier part of the series.

Achievement of Curriculum for Excellence Levels, 2024-25, published 9 December 2025:

• Primary (P1, P4 and P7 combined): literacy 74.5 per cent at the expected level (highest on record); numeracy 80.3 per cent (equal highest with 2023-24).
• Poverty-related gaps at primary: literacy 19.4 percentage points; numeracy 16.6 — both the narrowest in the series.
• S3, third level or better: literacy 89.8 per cent; numeracy 90.9 per cent — both record highs. Gaps: literacy 10.8 points; numeracy 11.6 — both the lowest on record.

ACEL is teacher judgement against CfE levels. It is an official statistic. It is not the same measurement as PISA or as an external examination.

Qualifications Scotland results, August 2026 (first year under the new awarding body):

• National 5 A–C: 79.2 per cent (78.4 in 2025; 77.2 in 2024; 80.8 in 2022). Entries 339,760. Grade A: 40.3 per cent.
• Higher A–C: 76.7 per cent (75.9; 74.9; 78.9 in 2022). Entries 207,980. Grade A: 30.9 per cent.
• Advanced Higher A–C: 77.6 per cent (76.7; 75.3). Grade A: 30.9 per cent.

Pass rates remain below the immediate pre-pandemic pattern at several levels. SIMD gaps on A–C in 2026: National 5 15.6 points (from 16.4); Higher 15.9 (from 16.8); Advanced Higher 12.1 (from 13.3). The A-grade gap is wider than the A–C gap. School-leaver measures in the 2026 National Improvement Framework still showed, for 2023/24, a gap of 18.1 points for one or more awards at SCQF level 5 and 31.8 points at level 6 — wider than the previous year on those leaver metrics.

Vocational and technical awards through SQA/Qualifications Scotland were reported at record volume in 2025–26 (the 2026 results coverage used a figure of 120,485; the NIF text used “over 100,000” for 2025). The two official counts should not be merged as one year.

Modern Apprenticeships, Skills Development Scotland, 2025/26: 24,849 starts, 2.6 per cent lower than the previous year. Achievement rate: 80.3 per cent, the highest recorded at year-end. Female starts: 41.2 per cent. Non-college providers achieved 81.5 per cent; colleges 74.7 per cent. 82.7 per cent of starts were at SCQF level 6 or above. SDS attributed the fall in starts to weaker employer recruitment, not to a collapse in completer quality.

Lindsay Paterson’s 2026 book Lessons from Scottish Schools: Why Knowledge Matters is not an official statistic. It is the principal recent academic statement of the knowledge-curriculum argument that the 2012 and 2022 PISA cohorts book-end Curriculum for Excellence, and that losses have been heaviest among lower attainers.

How it is experienced in daily life

A primary pupil is more likely than a decade ago to be recorded as meeting the expected CfE level. That is what teachers and authorities are measured on. A fifteen-year-old is less likely than in 2006 to match the OECD’s mathematics and science standard. An S4 pupil sitting National 5 enters a system whose overall A–C rate has inched up, whose poverty gap remains in the mid-teens, and whose mathematics pipeline remains a subject of repeated professional concern, even where a single official S4 mathematics rate is not isolated in the 2026 national summary.

A young person who wants a trade enters a Modern Apprenticeship system that completes well when started, and that started fewer people in 2025/26 because employers hired less. The senior phase still presents the academic route as the default prestige path. Employers in construction, engineering, energy and care continue to report shortages. That is a labour-market fact alongside a school-design fact.

Teachers work inside Experiences and Outcomes that specify processes more tightly than content. Secondary departments still report inconsistent knowledge arriving from primary. That complaint is long-standing in inspection and professional evidence. It is not a new discovery.

How it got here

Curriculum for Excellence replaced a more tightly specified subject tradition with four capacities, a broad general education to S3, and Experiences and Outcomes. The OECD’s own 2021 review, the National Discussion and the independent review of qualifications all accepted that the framework needed more clarity on knowledge and progression. The decline in PISA mathematics was already in train before Covid. After 2010, England specified more content, required systematic synthetic phonics, and incentivised a broader academic core. Scotland did not make the same specification choice. Relative positions in 2022 reflect this.

The Attainment Scotland Fund and the Scottish Attainment Challenge directed substantial extra money at poverty-related gaps. ACEL gaps at primary and S3 have narrowed. Leaver and Higher/National 5 gaps have moved more slowly. Money without a specified curriculum can raise teacher-judged levels while leaving the international and exam distributions where they are.

Qualifications reform and the replacement of SQA with Qualifications Scotland changed the awarding body, not the underlying content problem.

What has already been tried, and what that showed

Targeted funding raised some domestic indicators and narrowed some ACEL gaps. It did not reverse the PISA mathematics series.

The 2023 change in fire-alarm policy is not an education lesson; the education analogue is that a clearer rule (what must be known, what will be examined) changes behaviour faster than a capacity statement.

The Curriculum Improvement Cycle is the official admission that specification was too thin. Early 2026 samples use a Know–Do–Understand model intended to replace Experiences and Outcomes from August 2028. A full draft technical framework is due in autumn 2026, with feedback into 2027. Additional in-service days for 2026-27 are under discussion within SNCT. Implementation for S4–S6 follows later; government timelines describe new associated qualifications as becoming available around 2031. PISA 2025 cannot yet judge a cycle that completes in 2028–31, as it will reflect the old framework.

Teacher class-contact reduction, agreed in 2026 for phased introduction, is a workforce measure. It does not specify curriculum content.

What needs to happen

The improvement in measurable terms

• Mathematics and science on PISA, or an equivalent independent test, recovering toward early-2000s Scottish means, with the share below Level 2 falling.
• Primary fluency in decoding, writing and number confirmed by assessment that is not only the teacher’s own CfE judgement.
• National 5 and Higher A–C and A-grade gaps by SIMD narrowing faster than the 2025–26 one-point moves, and leaver gaps at SCQF 5 and 6 falling from the 2023/24 levels.
• A senior-phase mix in which technical and apprenticeship routes have published progression, employer hours and destination rates treated as first-class in inspection and funding.
• Modern Apprenticeship starts recovering when the labour market allows, without trading away the achievement rate.

These are tests of the system, not slogans.

Sequence

Specify the core. Sequence it. Assess it. Train teachers in the specified content. Then widen vocational architecture. Digital, wellbeing and interdisciplinary work come after the floor is secure. Reversing that order is how the last twenty years were spent.

The CIC timetable already exists. The decision that matters is whether Know–Do–Understand names the knowledge, in order, with enough grain that two schools cannot mean different things by “third level mathematics.”

Workforce, capital and standards

Subject knowledge in initial teacher education and in career-long professional learning is the delivery mechanism. Generic initiative time is not. Contact-time reduction only helps if the hours gained are used for planning the specified curriculum, not for more process. Additional support for learning is a parallel capacity constraint: a knowledge-rich sequence fails in practice if unmet additional needs empty the classroom. Inspection must look at what is taught and remembered, not only at how the school describes its capacities.

What would count as success in twelve months

By August 2027:

• a published draft KDU model for each curriculum area that a secondary head of department can turn into a year plan without inventing the core
• national sample assessment or equivalent that can be compared with ACEL, so teacher judgement is not the only official story
• PISA 2025 results absorbed without treating a single cycle as a verdict on a framework not yet replaced
• Modern Apprenticeship and senior-phase technical volumes reported beside academic entries as a standard dashboard
• ITE and Education Scotland professional learning shifted, in published programmes, toward subject content and instructional method

Twelve months will not restore a 2006 PISA score. It can show whether specification is real.

What can be done under current powers

Who holds the lever

Ministers: Framework, Budget, Education Scotland, Qualifications Scotland, the CIC mandate. Local authorities: staffing, additional support, school organisation. Schools: what is taught tomorrow morning. SNCT: time. Universities and ITE: teacher subject knowledge. Employers and SDS: apprenticeship demand and design.

The statute, budget or direction

Education (Scotland) Act 1980 and related Acts; annual Budget; National Improvement Framework 2026; Curriculum Improvement Cycle direction; Qualifications Scotland’s awarding rules; SDS’s MA contracts. All are Scottish.

What is already possible and unused

The CIC can specify content now, in the autumn draft, at a grain the current Experiences and Outcomes refused. Phonics programmes, a national mathematics sequence, and science conceptual progressions do not require a new Act. Inspection frameworks can be rewritten to sample knowledge. Funding and destination measures can treat technical routes as equal without waiting for 2031 qualifications. Schools that have already rebuilt programmes around knowledge can be used as working examples rather than exceptions.

What would still be true if management improved tomorrow

Poverty, language at home, housing and health would still predict attainment. A child who cannot decode at eight would still struggle at fifteen. Employer demand for apprentices would still follow the business cycle. PISA 2025 would still measure the old curriculum. Better management can stop underspecification. It cannot close the SIMD gap by circular alone.

Where the limits are

Law and reserved powers

The curriculum is not reserved. Immigration of teachers and some post-16 benefits rules are. They are second-order here.

Money

The Attainment Challenge and pupil equity funding already exist. Returns depend on what is taught. Teacher numbers, additional-support staffing and contact-time replacement are Budget choices that compete with health and local government. A knowledge-rich curriculum is not primarily a capital programme. It is a specification and training programme. Underfunding additional support will still wreck delivery.

Workforce and capital

Teacher supply in mathematics, physics, technical subjects and Gaelic is already tight. An ageing profile and rural posts compound it. ITE cannot be re-specified in one year. School buildings and devices matter; they are not the cause of a 18-point PISA mathematics fall.

Time, geography, UK-wide systems

A full curriculum generation is a decade. CIC implementation to 2028–31 is already that long. PISA is an OECD instrument; Scotland chooses to sit it. Comparisons with England are possible because both sit it. Rural schools face the same specialist-teacher constraint as rural fire cover: a small labour pool.

Political difficulty, kept separate from legal impossibility

Saying that teacher-judged records can rise while international scores fall is unpopular. Specifying content is described by some as a return to the past. Treating technical routes as equal offends a status hierarchy. Those are political costs. They are not legal bars.

What is not yet known

PISA 2025, due 8 September 2026. The full autumn 2026 KDU drafts. Whether the 2028 framework names content at usable grain. The next school-leaver statistical bulletin for 2024/25 and 2025/26. A consistent official time series for National 5 Mathematics among the S4 cohort, isolated from the all-entry A–C rate. The effect of the 2026 contact-time deal on instructional hours in the core. How Qualifications Scotland’s new qualifications, scheduled later in the decade, will map onto the new framework.

Honest summary

Achievable now. Specify knowledge and progression in the CIC. Assess the core independently of the teacher who taught it. Train to that specification. Publish technical and academic routes on the same dashboard. All of that is already inside Scottish law and the 2026–28 timetable.

Only partly achievable. The poverty-related gap at Higher and in leaver measures is shaped by family resources, early language and health. Schools can narrow it. They cannot legislate it away. Apprenticeship starts follow employers. International recovery takes a cohort, not a press notice.

Would require reserved change. Not the core of this item.

Colour mark: green. The powers to write the curriculum, set qualifications, inspect schools and fund teachers are already Scottish. The long PISA decline and the split between ACEL and external measures are the product of choices made with those powers. Green does not mean the 2028 framework will succeed. It means failure cannot be blamed on Westminster.

What this item does not settle. It does not choose a single textbook. It does not restore selection. It does not claim that wellbeing work is worthless. It does not treat England as a model to copy in every respect. It does not say ACEL gains are fictitious. It says they are a different measure, and that a system which improves only its own judgement while losing ground on the test it shares with other countries is not becoming more demanding.

The working test is simple. Does a pupil leaving primary have fluent reading and number that a stranger could verify? Does a fifteen-year-old have enough mathematics and science to sit with the OECD’s middle, not below Scotland’s own past? Does a young person who will not take three Highers have a technical path that is specified, examined and respected? Those questions are already on the Scottish desk.

Connection

Skills and the economy items in The 28 depend on the same pipeline: if mathematics and technical routes are thin, so are construction, energy and digital. Rural staffing is the same constraint as in fire and health. Child poverty and housing sit behind the SIMD gap. Public finances set teacher numbers and additional-support capacity.

The Independence Debate is not required to change the Experiences and Outcomes. A later Scottish state would inherit this curriculum problem unless it is fixed first.

People’s Future Scotland