What the UK's Exam System Could Learn From Abroad

The UK's system of competing, Ofqual-regulated exam boards has real strengths — choice for schools, resilience against a single point of failure — but it isn't the only workable model, and several other countries do specific things better. The Netherlands blends coursework and a national exam so one bad day matters less. Estonia has built exam delivery into a genuinely digital public service. Ireland runs one exam board and one qualification, with none of the coordination overhead the UK's five-board system carries. None of this means the UK's approach is wrong — but a few specific, well-evidenced features are worth taking seriously as UK reform ideas, without pretending the trade-offs disappear.

This is the closing piece of our exam boards cluster, and it's the place to draw the threads together honestly: what plurality buys the UK, what it costs, and which foreign features would genuinely transfer.

What the UK's plural-board system actually gets right

It's worth stating the case for the status quo before critiquing it, because it's a real case, not a null hypothesis.

Choice creates some competitive pressure. AQA, Pearson Edexcel, OCR and WJEC/Eduqas each compete for schools to adopt their specifications, which gives some incentive to invest in teaching resources, exam-board support and clear specification design. A school unhappy with one board's support or resit provision can, in principle, move to another.

No single point of failure. If one board has a serious administrative problem in a given year — a printing error, a security breach in one paper — the entire national cohort isn't affected the way it would be under a single national exam authority. The 2017 Edexcel GCSE Maths paper leak, for instance, was a serious incident, but it was contained to one board's students rather than the whole country's.

Ofqual's comparable-outcomes framework does real work. The regulator's approach to aligning standards across boards (see our piece on why the UK has five boards at all) means the plurality doesn't translate into wildly different grading standards in practice — a genuine achievement of the regulatory model, not an accident.

The cost of all this is coordination: five boards, five sets of specifications per subject, five sets of past papers and mark schemes for parents and teachers to navigate, and the recurring "board shopping" question of whether schools ever choose an easier specification rather than a better-taught one (worth its own separate, careful treatment, and not one this article settles).

What could genuinely transfer: three specific ideas

1. The Dutch dual exam — spreading the risk away from one sitting

In the Netherlands, secondary students sit what's often called a "dual examination": part continuous school-based assessment through the final years, part a shared national exam at the end, with the final result blending both (DUO, the Dutch education executive agency). The effect is that a single difficult exam day, illness, or bad nerves matters less to the overall outcome than it does in a system that rests almost entirely on final written papers.

England has moved in the opposite direction over the past decade, reducing coursework and non-exam assessment in most GCSE subjects in favour of terminal exams, partly over concerns about coursework's vulnerability to unequal support at home. That's a defensible concern — but the Dutch model shows a middle path is possible: a national exam that still anchors standards, alongside meaningfully-weighted school assessment that isn't dismissed purely because it's harder to standardise. It's a genuine trade-off, not a free upgrade, but it's one the UK debate rarely engages with directly.

2. Estonia's digital exam infrastructure — treating delivery as a technology problem

Estonia is frequently cited for strong outcomes in the OECD's PISA rankings, and one distinctive feature of its education system is how far exam administration itself has moved online, as part of a much broader digital-government approach (e-Estonia). This is easier in a small, highly centralised country of roughly 1.3 million people than in England's much larger and more administratively devolved system — that scale difference is real and shouldn't be waved away.

But the underlying principle is scale-independent: treating exam delivery — not just exam content — as something worth serious technological investment, rather than assuming paper scripts and postal logistics are simply how exams work. The UK has made some moves here (on-screen assessment pilots in some subjects), but nothing close to Estonia's default-digital posture.

3. Ireland's single national exam — the case for administrative simplicity

Ireland runs the Leaving Certificate, set by one state body, the State Examinations Commission, with no equivalent of the UK's board-choice layer. For a family, this removes an entire category of decision and confusion: there is no "which board is my child with," no comparing specifications across four alternatives, no board-shopping question at all.

This isn't a claim that Ireland's exams are easier or harder — that's precisely the kind of difficulty comparison Ofqual's comparable-outcomes work exists to avoid making carelessly in the UK's own system, and it applies just as much across borders. It is a claim that the administrative picture is simpler, and that simplicity has a real value to families and schools that the UK's plural system trades away in exchange for the choice and resilience benefits above.

4. Germany's shared task pool — comparability without giving up local control

Germany's Abitur (covered in full in UK exam boards vs Germany) is fragmented in a completely different way to the UK — sixteen federal states each set their own version, rather than competing providers — but the fix Germany has built for the resulting comparability problem is worth the UK's attention on its own terms. Since 2017, the IQB, an institute based at Humboldt University Berlin, has maintained a shared national task pool: a bank of exam questions in core subjects that participating states can draw on when assembling their own Abitur papers, specifically to narrow the gap between what a top grade means in one state versus another. States keep full control over administration and marking, but no longer have to write every question from scratch, and comparability improves precisely where it's needed most.

England doesn't have Germany's geography problem — one Ofqual regulates all four boards already — but it has a milder version of the same comparability question: does a top grade from AQA mean quite the same thing as a top grade from OCR in the same subject? A shared item bank, of the kind Germany built to solve a harder version of that problem, is a specific, evidenced mechanism the UK hasn't tried, rather than a vague appeal to "more consistency."

A side-by-side view

The honest trade-off, stated plainly

Every feature discussed here is a trade-off, not a free improvement:

• More coursework weighting (the Dutch approach) buys resilience against a bad exam day, at the cost of harder-to-standardise assessment.
• Deep exam digitisation (the Estonian approach) buys efficiency and possibly better access provisions, at the cost of significant upfront infrastructure investment and a scale the UK doesn't share with Estonia.
• A single national board (the Irish approach) buys administrative simplicity, at the cost of the competitive pressure and single-point-of-failure resilience the UK's current system provides.
• A shared question bank across boards (the German approach, transplanted from a state-level problem to a board-level one) buys tighter comparability between boards, at the cost of the exact specification differentiation that lets a school choose a board that genuinely suits how it teaches a subject.

The UK doesn't need to import any of these wholesale to learn from them. But a serious look at how other countries balance rigour, fairness and simplicity is worth more than the reflex answer that the UK system, warts and all, is simply how exams are done. It isn't the only way — it's one way, with a specific and inspectable set of costs and benefits, same as everyone else's.