There’s a lot of it about.

As I’ve traversed the country this summer, I’ve been hearing endless tales of curriculum transformation – where the fancy phrases and whizzy websites promise courses that are more engaging, more relevant, more connected, and more… belongingy.

They’re usually accompanied by a set of student leaders whose third year was nothing like it was when originally sold to them, and whose stories belie the “nothing to see here guv” responses to OfS’ C6 consultation that have emerged from sector bodies.

Strip away the grandiose language and a pattern emerges – some combination of fewer, larger modules, almost always wedges of 30 credits, an all-core or heavily prescribed first year, reduced optionality in later years, curated pathways instead of individual module choice, fewer and larger assessments, compulsory institutional skills or employability modules, standard semesters and timetables, and minimum-enrolment thresholds for the options left.

The through-line is consolidation. Fragmentation, overload, weak belonging, incoherence, administrative complexity, and financial pressure are all to be relieved by reducing the number of moving parts. And you can see the institutional attraction – one structural change appears to solve several problems simultaneously.

Fewer modules might mean fewer timetable combinations, fewer assessment events, cohorts that stay together, programme outcomes that are easier to map, options that are more financially viable, simpler records and quality assurance, larger teaching groups, and less risk of module cancellation.

But none of those benefits arises automatically from the credit size, and reducing the number of modules is only one of several available responses to those pressures.

Finnish universities offer a useful comparison when the driver is financial rather than pedagogic. They absorbed a €500 million cut to higher education funding in 2015, on top of an earlier €200 million reduction, and the University of Helsinki alone laid off around 400 staff, pushing its academic staff-student ratio out to comfortably higher than anything in the UK.

Rather than shrinking its curriculum into fewer, larger, compulsory blocks, Helsinki kept just 32 undergraduate programmes spread across 11 faculties, built pass/fail credit into study skills, digital literacy, and personal study planning, and let students bank credit earned on other programmes, through work, volunteering, or Erasmus, towards the degree. Staffing fell hard, but the breadth and the choice on offer to students barely moved.

One question that’s been playing on my mind is whether consolidation has been chosen because it’s the best educational response – or because it’s the only response that simultaneously satisfies pedagogic, regulatory, operational, and financial objectives.

Another is why this has become – with little slithers of provider-specific branding – the dominant mimetic response in England, when the rest of Europe seems to be coping with massification in very different ways.

But it’s that study on regret that I’ve been thinking about the most.

What the regret evidence means

Back in April 2025, a team at the University of Bristol – working with HEPI and Advance HE – published The Benefits of Hindsight: Reconsidering higher education choices. On the tin it was a report about student decision-making, careers guidance, and transfer.

Read through a curriculum-design lens, as I have on multiple cross-country trains, it turns out to be a manual for how degrees ought to be scaffolded – and an awkward one for anyone currently stripping the options out of theirs.

Its implicit model is that retention is good, regret is an information problem, and transfer is the corrective valve. The current reforms are only partly compatible with that – they may improve the original decision by making courses easier to explain and by making the options that are advertised more likely to actually run, and common first years may ease internal transfer.

But most frameworks devote far more attention to keeping students on the programme than to making movement between programmes cheap – they strengthen the retention half of the model while neglecting the corrective valve half.

There are at least 10 other ways of reading the same evidence – and the one you believe determines whether the reform wave looks like the cure or the disease.

1. Exit is healthy

The lock-in model says regret is a retention problem and cheap re-entry is the corrective valve – drop-out and swirl are signs of a functioning market in second chances, and the pathology is a system that makes leaving catastrophic. The report’s sunk-cost quotes and the 52 per cent regret rate among completers are its evidence, and it implies that continuation metrics are measuring harm as success.

On this reading the dominant package risks making things worse – large modules create large units of sunk cost, course-specific cores make completed credit hard to reuse elsewhere, and reduced optionality limits the student’s ability to redirect the degree from within. What it wants is portable, divisible, reusable credit.

2. The student as consumer

The mis-selling model says regret is detriment and redress is the valve. The report brushes against it – ASA concerns, “the way that it was sold is not quite exactly how it is now” – without following through – if 35 to 52 per cent of purchasers of a £50,000-plus product wouldn’t buy again, other regulated markets would already be talking about cooling-off periods, refunds, and enforcement, rather than better marketing.

The reforms cut both ways – a smaller set of modules that reliably runs is more honest than a catalogue that’s routinely cancelled, but removing options that students applied on the basis of is itself a form of mis-selling. What it wants is a clear, enforceable bargain.

3. Regret means it worked

The development model says identity change is the point and the curriculum itself is the valve. The report half-states it – “higher education is designed to prompt exactly that sort of critical reflection that leads to changed views” – and then treats regret as a problem anyway.

The reforms are compatible where they provide broad foundations and later pathway choices, and incompatible where students are expected to develop but not allowed to act on that development – celebrating intellectual transformation while requiring students to follow the route chosen before the transformation occurred. What it wants is branch points built into the award itself.

4. The chooser doesn’t exist yet

The premature-specialisation model says choice at 17 is the defect and late specialisation is the valve – nobody’s information is good enough at application because the person doing the choosing doesn’t exist yet. The Scotland gap – 74 per cent against 64 per cent – is its evidence. It flatters the common-first-year standardisers – but only where the first year is shared across related subjects.

A student entering a broad biosciences family and later choosing biochemistry is delaying specialisation, while a student completing a wholly prescribed Biochemistry year has never had the option, and reduced outside-subject optionality removes the chances to discover adjacent fields. What it wants is subject-family entry and provisional pathways.

5. Fewer, better choices

The choice-architecture model is the standardisers’ best defence – regret is a decision-overload problem, and fewer, better-structured choices are the valve. One university’s claim that “a good curriculum does not necessarily offer more choice; it offers academically coherent choice” is the model verbatim, and the reforms’ central mechanisms are exactly those it predicts.

The trouble is that the report never shows that module choice causes regret – students regret courses, institutions, career consequences, and fit. An entire reform wave is resting on an assumption. The better version is soft architecture – defaults rather than prohibitions, guidance rather than removal.

Helsinki’s own practice backs that up. Its students keep real freedom to build their own study path within a programme, but every one of them completes a credit-bearing personal study plan, revisited with a named tutor at fixed points across the degree through the university’s student records system.

That suggests the overload the choice-architecture model worries about is as much a planning-support problem as a menu-size problem – structured guidance seems to convert options into commitments students actually act on, rather than a pile of choices they freeze in front of. Soft architecture, in other words, includes planning support as well as defaults.

6. Blame the market

The labour-market model says the degree is fine and regret is displaced blame for a brutal graduate market. Graduate regret is overwhelmingly career-coded – accounting for that 56 per cent figure – concentrates in lower-skilled jobs, and fluctuates with employment. As one respondent put it, “When I landed a job, I almost started forgetting about it.” The same courses would generate half the regret in a stronger market.

Its strongest version distinguishes three deficits.

• A capability deficit – students lack what available work requires – warrants applied learning and work experience.
• A translation deficit – students possess capabilities they can’t recognise, evidence, or communicate – warrants recognition rather than more content – AGCAS and GFI research found students discounting learning not formally connected to their degree, and much of what employers say they can’t find is developed through representation, volunteering, and real responsibility – precisely the settings that accelerated, consolidated curricula squeeze out before reinserting them as a simulated employability module.
• And a demand or utilisation deficit – the economy doesn’t generate or properly use graduate-level roles – warrants valves that sit outside the curriculum entirely – regional development, employer investment, job redesign.

A reform framework that answers all three with “embed employability” is analytically weak.

7. Fit is the product

The belonging model says regret is relational and community design is the valve. The single most common reason undergraduates give for regret has nothing to do with academic content or career prospects – at 31 per cent, “I would have been happier or fitted in better,” tops the list.

The disability findings sit naturally alongside it – 40 per cent of disabled students report regret against 33 per cent overall, and 82 per cent of disabled graduates wish they’d transferred.

Stable cohorts and common modules can support community – but they risk turning belonging into a lock-in retention mechanism, where a student stays because changing course means losing friends, tutors, and identity. What it wants is belonging that survives a change of route – group social mentoring and associative activity at a broader subject level.

8. Price creates regret

The stakes model says regret is proportional to sunk cost – English students regret more than Scottish ones, and fee-and-debt anguish saturates the free text. Two things follow.

First, larger modules raise the stakes of every assessment event – a failed 30-credit module is a quarter of the year, and recovery typically means unfunded time, delayed progression, maintenance and visa complications, capped marks, and credit that can’t move. Module enlargement reduces the frequency of assessment risk while increasing the size of each risk.

Second, most reforms rearrange the same 120-credit intensity into fewer containers rather than reducing what students must do or extending the time available – flexibility that never touches the pace, chosen partly because it doesn’t challenge the three-year norm or the funding model.

What it wants is programme-level assessment reform and recoverable credit, rather than bigger blocks.

9. Same regret, different damage

The stratification model reads regret as unequally distributed harm. POLAR quintile one students report far higher consequence from wrong choices – 47 per cent say “a lot of difference” against 36 per cent overall – and 76 per cent of free-school-meals graduates wish they’d transferred, while first-in-family students, in their own words, had “no one to ask.”

Same regret rates, very different damage – so universal fixes miss the point, and some reform elements increase the inequality – intensive blocks are difficult with fluctuating health, large assessment units magnify disruption, placements may be unaffordable, and transfer costs fall hardest on students without family resources.

What it wants is targeted scaffolding for the students with least slack.

In the 2026 Student Academic Experience Survey from HEPI and Advance HE only half of care-experienced students and students with a trans identity or history would make the same choice again, against 52 per cent of students who reported harassment, 56 per cent of international students, 59 per cent of students with a mental health condition, and 60 per cent of disabled or estranged students.

The sector can note that only 7 per cent would avoid higher education altogether and call that reassuring, but that misses the point – the students most likely to regret their choice are already carrying the greatest financial, personal, and institutional risk, and a heavier, less forgiving curriculum structure is the last thing that helps them.

10. Maybe it’s mostly noise

And the artefact model – the null hypothesis. Hindsight is cheap – 37 per cent of regretful undergraduates said nothing realistically available at the time would have changed their decision, and 10 per cent of graduates report their regret resolved itself over time.

This model’s answer favours interventions that are reversible, targeted, and testable over a wholesale redesign triggered by regret – switching windows and credit banking can be undone, but permanently eliminating modules and pathways cannot. It counsels against redesigning whole institutions around a retrospective survey response – which is, on at least one reading, exactly what’s happening.

Whose degree is it anyway?

Underneath all of that sits the debate’s missing organising principle – a curriculum is, above all, an allocation of control over 360 credits and roughly 3,600 notional hours of someone’s life, and curriculum reform is partly a struggle over who decides what they’re for.

Control of credit determines which departments receive workload, which modules and posts remain viable, what students must do rather than may do, and whose conception of a graduate prevails.

There are five claimants.

1. Under disciplinary sovereignty, the subject owns the degree – 360 credits in Biochemistry presumptively belong to Biochemistry, giving credit away is experienced as dilution, and every agenda gets “embedded” until a small subject module is expected to deliver employability, sustainability, EDI, wellbeing, and citizenship simultaneously.
2. Under institutional trusteeship, the university owns it – hence graduate-attribute modules, common first years, and standard architectures, though this moves credit from the department to the centre rather than to the student.
3. Under student proprietorship, the student owns a substantial portion – able to allocate credit to an adjacent discipline, a placement, a volunteering role, a language, or a project.
4. Under regulatory ownership, the regulator indirectly owns it – continuation, completion, and progression requirements shape which structures institutions dare to maintain, and experimental arrangements like over-credit schemes or experiential recognition tend to be the first casualties of compliance anxiety.
5. And under social formation, purposes beyond student and subject – employers, professions, communities – claim their share, each individually persuasive and collectively overloading.

Seen this way, optionality is an allocation right over part of the degree. Moving credit into the core transfers that right from students to programme teams or the institution, and many 30-credit reforms look like a settlement between the first two claimants – departments keep a large prescribed core while the institution takes a common block or two.

The student’s share shrinks, and it shrinks just as the student becomes better able to make informed choices – the development model’s logic run in reverse.

The second question is who owns the time the credit represents. Credit is supposed to denote student effort, yet institutions treat curricular and departmental allocations as fixed while treating the student’s time as expandable – expecting students simultaneously to complete a full 120-credit year, undertake placements, join societies, volunteer, engage in representation, and earn enough to live – while still completing the work, for a disabled student given extra time they don’t actually have.

Student time is the residual resource – the thing that absorbs every unresolved curricular demand. And when learning becomes more efficient, through standardisation or AI-assisted work, the reform wave assumes the released capacity belongs to the institution – enabling more outcomes to be inserted or staffing to be reduced, rather than lower intensity or more student-controlled credit.

It reframes reduced optionality as enclosure – previously discretionary credit brought under programme control. Some common learning creates entitlement and coherence. But the burden of justification should reverse – rather than asking why students should be allowed to choose this credit, the institution should have to explain why it’s necessary to remove the student’s authority over it.

The LLE and the 300-hour unit

There’s obviously an environmental driver that helps explain why the magic number is 30. The Lifelong Learning Entitlement doesn’t require 30-credit modules – DfE confirms smaller combinations are fine – but separately financed modular study normally has to total at least 30 credits.

A single 30-credit module therefore needs no bundling, wrapping, or combined transcript – one enrolment decision, one fee calculation, one addition to the modular catalogue – where twenties, fifteens, and tens all require packaging.

Universities have a incentive to make ordinary degree modules directly compatible with the emerging funding architecture – under four-by-30, every module is already a separately fundable product, with no parallel modular curriculum required – and even institutions with little immediate modular activity can present the change as future-proofing.

The paradox is that a funding reform intended to make higher education more modular may make degrees less internally granular. Under twelve-by-10, a new claim on credit can be accommodated in small increments. Under four-by-30, it means taking a quarter of the year from an existing claimant.

A department might tolerate 10 credits of language learning and reject a 30-credit language module – so experiential learning, languages, civic activity, service, and student-designed projects find fewer insertion points, and institutional purposes end up crammed into one generic block, “embedded” unevenly across reluctant modules, or displaced into co-curricular time students no longer have.

The floor matters after failure too. One failed module in a four-by-30 degree creates an administratively neat, finance-eligible repeat package – 30 credits becomes both the convenient purchasing unit and the convenient failure unit, with students repeating and paying for 300 hours even where only part of the module caused difficulty.

And the wider regulatory environment pulls the same way. “Quality” is defined partly through the NSS, which doesn’t ask about regret, and through continuation and completion metrics that reward institutions for keeping students where they are – so a framework that deepens lock-in can register, on the official instruments, as an improvement.

Europe

Looking at how our European cousins are coping with massification, it’s clear that external packaging and internal granularity are separable, where large administrative containers routinely hold smaller taught and assessed components, with stakes pooled by arithmetic at the boundary, and no correlation anywhere between module count and cost.

Recoverability is structural rather than financial – examination-based systems can rerun assessment cheaply through guaranteed calendared resits, automatic re-registration and quick-turnaround retakes, offered as routine service rather than concession, while coursework-based systems must ration recovery through capped resits, capped marks and board discretion. England has shifted to the costlier mode without buying the recoverability that students really wanted – recoverability the cheaper one affords elsewhere.

Compensation can also be a published arithmetic entitlement rather than a discretionary judgement – automatic weighted-average compensation, mechanical credit ledgers, and rules letting later success retroactively discharge earlier failure are all cheaper than individualised administrative decisions or curricula built on dozens of discretion points – automatic generosity is cheap, discretionary restriction is expensive.

Breadth is frequently mandated rather than merely permitted, with compulsory off-discipline credit reaching up to a third of the award, and can reduce cost, especially when the breadth is other subjects’ cores. Placement supervision and student support costs are routinely transferred to students and external actors. And when European systems actually credit the learning hours involved in becoming a student and becoming a graduate, pressure is relieved elsewhere.

Reform redistributes regret

None of this is a marginal concern. The 2026 Student Academic Experience Survey found that one in three students would rewrite a decision that came with years of work, a mountain of debt, and consequences that could follow them for decades – rising to more than four in 10 by the third year.

Among students whose experience was worse than expected, 67 per cent would choose differently. Among those who had considered withdrawing, 66 per cent would. Among those who thought their course was poor value for money, 58 per cent would. Regret deepens the longer a student stays inside the system, which is exactly the point in a degree at which the current reform wave is proposing to close off the last of the exits.

Put the models together and a verdict forms. The dominant package is likely to reduce regret expressed as “the course felt fragmented”, “there were too many deadlines”, “the options I was promised didn’t run”, and “I never knew who was actually on my course”.

It may increase regret expressed as “I had to choose my field too early”, “I discovered another interest but couldn’t pursue it”, “one difficult module cost me 30 credits”, “changing course would have meant discarding too much completed work”, and “the course was redesigned for efficiency after I enrolled”. It exchanges some risk of fragmentation and overload for greater risk of lock-in, homogenisation, and high-stakes failure.

The better alternative is a different bundle of valves – none of them exotic, and most of them already operating somewhere in the sector – organised around six principles.

1. Delay irreversible choice – 60 to 90 credits of shared cognate learning in the first year with published routes into several related degrees, 15 to 30 credits of protected pass/fail exploratory credit, and formal reorientation points at which interest, challenge, belonging, and career direction are reviewed before final pathway selection.
2. Plan the choice as well as offering it – credit-bearing personal study plans revisited with a named tutor at fixed points across the year, built into peer tutoring rather than bolted on as a one-off induction task, and integrated into student records systems so a change of plan is easy to register rather than easy to lose.
3. Make movement survivable – portable and divisible credit, accredited annual exits, credit banking and re-entry guarantees, funded support for switching, independent course-change advice – since students considering transfer shouldn’t have to rely solely on staff whose programme loses income when they leave – and tutorial and peer communities that survive a change of route.
4. Make failure recoverable – recognition of partial achievement within large modules, compensation where outcomes have substantially been met, in-year reassessment with continued teaching, and flexible progression with small amounts of outstanding credit – alongside programme-level limits on assessment volume and deadline bunching, with no assumption that every module must be 30 credits.
5. Guarantee experience without prescribing one identical experience – every student completes a placement, research experience, community project, client brief, or enterprise activity within the existing credit total, with space in the 180 for recognition of suitable paid work and volunteering, supported placements reserved for students without professional networks, and the capabilities developed formally recorded on the transcript.
6. Make the bargain clear – a course-content contract publishing the compulsory curriculum, likely options and rotation schedules, viability rules, transfer destinations, and the expected assessment pattern, with funded correction where the institution materially changes what was sold.

A better machine, wrong journey

The universities I’ve seen have pretty much all selected a powerful and administratively convenient central valve – reduce the curricular moving parts. That valve is well suited to problems of fragmentation, timetable complexity, unreliable choice, weak programme identity, and assessment proliferation.

It’s much less well suited to premature commitment, changing identity, transfer, sunk costs, unequal capacity to absorb a wrong decision, and the need to leave and return. And the environment isn’t neutral – universities are adapting the curriculum to a system that insists on rapid full-intensity progression, penalises recovery, measures quality in ways that reward lock-in, funds 300-hour blocks most cleanly, and holds them responsible for labour-market outcomes they can influence but don’t control.

The alternative is progressive, recoverable, relational, and increasingly student-owned commitment. A programme should begin with common and transferable foundations, expose students to adjacent fields and possible destinations, permit real pathway choice once that experience has been had, support that choice with structured and ongoing planning rather than leaving students to work it out alone, preserve time for real responsibility, and recognise an exit or return at each stage.

It should combine a stable subject-level home with freedom over some credit, use frequent formative practice, visible accountability, and bounded repair, and, where the year is built from four 30-credit blocks or aligned to the LLE, make at least one block properly composable rather than using the funding transaction size as the pedagogical unit.

A student-centred curriculum keeps curation, coherence, disciplinary formation, community, and employability, but pairs them with reversibility, temporal flexibility, funded recovery, recognition, subject-level associative life, and an honest allocation of credit, labour, and time.

The question I can’t shake isn’t really about what 360 credits contains. It is who had the authority to decide, which work actually disappeared, who surrendered something to make room, and whether the student was treated as the proprietor and active member of any real part of the degree. It is theirs, after all.