The new education secretary has announced qualification reform for England that goes beyond the recommendations of the Francis review, arguing that we need “pathways for all” to combat the alarming numbers of NEET young people. Meanwhile, publication of the latest PISA results has prompted a fresh round of policy comparisons. However, before indulging in any misjudged policy-borrowing, we need to understand how the system came to be the way it is. This involves applying the principle of “Chesterton’s fence”. In other words, before making changes to policy infrastructure, it’s worth checking why it was put there in the first place. To that end, new research from the Centre for Education Systems (CES) suggests that when it comes to assessment and qualifications, there are three areas in which the UK nations stand out from the pack. But it also points to how we ended up here. Most countries run national exams for qualifications at the end of the upper secondary phase. However, there is much greater variation in approaches to assessment at the end of lower secondary education. Exam-based assessment at 16 The first stand-out feature of the UK, and probably the most famous, is the amount of exam-based assessment experienced by 16-year-olds in GCSE-based systems. Scotland’s National 5s come in slightly lower. Even once exam content is cut back by 10 per cent, as was proposed by the curriculum and assessment review, England’s lower secondary school pupils will still spend more time in exams than their counterparts in many other countries. It’s not necessarily that pupils elsewhere are assessed less, although they certainly are in some places. What is striking is greater reliance on teacher-assessed grades. Many countries assess a small core of subjects through centralised exams, but leave the rest of to be assessed in schools. In other cases, a defined percentage of each subject is assessed through exams and the rest within school, much as England did until the 2015 reforms. Those changes followed concerns about the vulnerability of non-examined assessment to cheating, and to the skewing effects of home background and parental support. These concerns will only have grown in the era of AI. This is our first Chesterton’s fence. A second stand-out feature of all UK education systems – not just England – is the unusual drop-off in breadth of study following key stage 4. Freedom to specialise Seventeen-year-olds in the home nations (Scotland slightly less so), have unparalleled freedom to specialise to an extent that goes beyond what many jurisdictions allow even for university students. Of the 15 countries reviewed in CES’s recent study, only the UK home nations and New Zealand allow pupils to drop maths and their national language while still in education. By international standards, having no compulsory subjects or cross-curricular components after age 16 is incredibly unusual. Such extreme specialisation has wash-back effects, since national qualifications almost invariably take place just before pupils specialise. It’s almost like saving your progress on a video game so that if your character dies, they respawn at the last save. The predominant reason for this specialisation – our next fence – flows from the structure of the UK’s university system. British undergraduate degrees are unusual in (typically) focusing on a single subject. Where other countries’ degrees begin with a broad, entry level of content, the UK’s rely on A-levels or Highers to lay that groundwork. That’s why we ask pupils at upper secondary to study a small number of subjects in more depth, where other places prioritise shallower breadth. Multiple exam boards A third unusual feature of UK, and in particular English qualifications, is the multiple private exam boards that compete for candidates in a quasi-free market. Other than Japan, all other researched countries have a single organisation with responsibility for national assessments for all subjects. Combined with the high choice that comes with specialisation, this makes our qualifications system vulnerable to fluctuations in the market. If a subject or qualification becomes unfashionable or loses credibility, the demand drops, and if the demand drops it becomes financially inviable for the exam boards to provide as an option. Policymakers in the UK therefore have fewer levers of control when designing qualifications reform. We can build it, but they might not come. Herein lies the third fence. This isn’t to say that our system is, or should be, set in stone. The challenges facing young people in the UK today are severe and unprecedented, and if we don’t provide qualifications that marry young people’s needs with our economy’s demands, we all suffer. England, like the rest of the world, will have to learn how to deal with AI and other unfair advantages in assessment. Universities and exam boards alike could be asked to make changes to adapt to the needs of a changing world. But when planning any redesign of qualifications, policymakers have to manage these interdependencies. The best laid plans will otherwise join the rapidly growing pile of failed assessment system reforms.
Barrie Hunt 11 October 2026 This article offers an important and interesting perspective on English qualifications and I agree with the symptoms it identifies. However, I should like to develop the author’s explanations, from the perspective of a former teacher and Director of an exam board, and will add ‘excessive emphasis on ill-defined standards’. Assessment by teachers could be the subject of a separate article. Sufficient here to note that all such discussions should be informed by two simple principles: • Many teachers are heavily influenced by their surroundings and have no access to the necessary overview of how their standards match national performance. • However, teachers are generally much better than examinations at placing their classes in rank order. It would be easier to introduce ‘mathematics’ into the 16-19 curriculum if the subject were to be divided into ‘numeracy’ – motivated by relevance to lifelong learning – and ‘traditional mathematics’ – based on algebra, trigonometry and calculus that largely serves specialist needs. The author’s conclusion that three exam boards make English qualifications “vulnerable to fluctuations in the market” must be challenged… When exam boards were rationalised 30 years ago, they argued that competition would improve service to centres and, having worked solidly for a month between 6am and 2am to ensure that schools received their new GCSE specifications before the start of December, I can assure your readers that competition was built on service. External regulators have always taken alignment of standards very seriously. Ofqual builds the regulatory framework around the National Curriculum through its “Subject Level Conditions” that every exam board qualification must meet. The result is that it can approve three diverse GCSE qualifications, each meeting fundamental criteria, but enabling a degree of choice that centres would not get from a single awarding body. Of course market forces apply but, with a strongly centralised National Curriculum, there is unlikely to be space for additional subjects. Whilst the article identifies three areas where the UK stands out from pack, I would add a fourth. In creating Ofqual to avoid the apparent conflict of interest between the Curriculum and Assessment aspects of QCA, the government paid insufficient attention to the impact on the curriculum. It is telling that Ofqual does not use the internationally recognised Validity, Reliability, Impact on the curriculum and Practicality (the VRIP principles) for quality assessment. Its statutory objectives emphasise standards, without defining them, and reliability – impossible to achieve in a six-level grading system of 4-9, since grade boundaries are so close together. It is worth noting that my submission to the CAR demonstrated how GCSE failed every VRIP criteria. With Ofqual so far removed from the classroom any innovation from Ministers is bound to fail…
Bob Moon 16 October 2026 Interesting article, excellent on comparative data. I would add however, two crucial points. First we are the only European country to have a 16+ normative exam system. All children are ranked against each other. The final total of GCSE’s is always a bell curve. A third of young people fail. Exam board statisticians adjust the outcomes (grades) to ensure this is constant. Those students on the wrong side of the bell curve predominantly come from poorer homes. Second our GCSE is the most expensive exam system in Europe. In the final year of GCSE study schools must spend more on the exam entry than the resources to study for the exam. Alice in Wonderland?! Bob Moon