Optional modules are not as optional as they look

The underlying law – Part 2 of the Consumer Rights Act 2015 – hasn’t changed. What the CMA thinks it means in practice very much has.

One caveat before I get going. The revised document remains guidance rather than new law or a definitive interpretation of it. It sets out the factors the CMA says it will consider when assessing fairness, including when using its new enforcement powers – and fairness itself remains a fact-sensitive question, ultimately one for a court, or for the CMA using its statutory powers, to determine.

The overall shift matters for higher education because the CMA’s dedicated advice for the sector – CMA182 – was last properly updated in May 2023, and now reads like a period piece.

It predates the unfair commercial practices provisions of the Digital Markets, Competition and Consumers Act 2024 coming into force, it predates the CMA acquiring the power to determine breaches itself and impose penalties of up to the higher of £300,000 or ten per cent of global turnover, and it now predates a version of CMA37 that is longer, considerably more interventionist, and much less impressed by the way large organisations tend to do contracting.

I’ll skip lightly over the enforcement machinery here – the interesting story isn’t the stick, it’s how far the CMA’s account of what fairness involves has moved while the sector’s terms and conditions have stood still.

A tick box is not a contract pack

The student contract, as practised, is an offer letter, a set of central terms, academic regulations, assessment regulations, a fees policy, a deposits policy, a module catalogue, a portal announcement, some FAQs and whatever the agent said at the recruitment fair – scattered across several sites, applicable “as amended from time to time”, with a tick box at enrolment declaring that the student has read and understood the lot.

The revised guidance offers little comfort for any of that. On the CMA’s analysis, a declaration proves nothing about whether the student had a real opportunity to read the terms, whether the important ones were prominent, or whether anyone could work out their practical effect – and an expectation that students periodically check a webpage for amendments would be very hard to defend.

The guidance’s emphasis on early access, prominence and retainability points towards a complete and retainable set of contractual documents being made available before acceptance – and for universities, it also makes preserving the exact version that applied to each cohort increasingly difficult to avoid.

There’s much more on digital presentation too – whether documents work on a phone, whether formats such as PDFs are usable with assistive technology, whether key terms surface at the point they matter, whether the student can save a copy.

Transparency isn’t achieved because the text is technically somewhere online, and a summary or FAQ can aid comprehension but can’t contradict the contract it summarises.

“Operational requirements”

The biggest movement is on variation clauses – and given the state of sector finances, that’s also where the biggest exposure sits.

CMA182 already recognised that changing university is not like changing broadband provider – a student who walks faces lost credit, a housing contract, maybe a visa, maybe caring responsibilities that mean there’s no realistic alternative within reach.

What the revised CMA37 adds is a framework for testing the clause itself – how broad the power is, whether the reasons for using it are objective and sufficiently specific, whether the triggering event is external or arises from the university’s own commercial and operational choices, how important the affected feature is, what notice and practical protection accompany the change – including whether the student can avoid it without loss or serious inconvenience – and, running through all of it, the cumulative effect of repeated changes.

The point universities will least enjoy is that specifying a reason doesn’t, by itself, make a clause fair. “Operational requirements”, “staff availability”, “financial sustainability” and “business needs” all appear in real variation clauses across the sector, and all of them hand the university close to unlimited discretion while sounding reassuringly procedural.

Applied to higher education, the CMA’s analysis suggests a distinction between changes compelled by external law, events nobody could have foreseen, educational changes backed by academic judgement, and ordinary failures of staffing, planning or cost control.

The guidance doesn’t create closed categories of permissible and impermissible change – everything remains a case-by-case assessment – but those four are unlikely to be treated alike, and only some of them look capable of justifying rewriting the deal mid-course.

And CMA182’s current suggestion that advance warning may allow a broad variation provision to operate – because the student can decide whether to continue – now needs heavy qualification.

Warning someone you’re about to exercise an unfair power doesn’t make the power fair, and an apparent option to leave is not a real one if leaving means starting again somewhere else, minus the credit, the deposit and the tenancy.

Twenty options in the prospectus, five that run

Optional modules are where that framework hits hardest. The standard institutional line is that changes to them are minor because no individual module was ever guaranteed – “optional modules may vary” is carrying a remarkable amount of weight in terms and conditions across the country.

Applied to optional modules, though, the analysis gets uncomfortable – because the thing that was sold may not have been any particular module. It may have been the breadth – the “tailor your degree” claim, the specialist pathway, the spread of disciplinary areas, the balance of theory and practice, the range of assessment.

A course advertised with twenty options is not obviously the same service when five remain, even if every one of the twenty was, individually, “indicative”. The legal mechanics matter here – a reduction in breadth may change the main characteristics of the service that informed the student’s decision to enrol, while the difficulty of transferring bears on whether a right to withdraw amounts to meaningful protection at all.

On that analysis, the cumulative point matters just as much. A university that would never dare cut half its options in one go can reach the same place across three academic years of individually “minor” closures, each approved on its own merits by a programme board that never sees the running total.

The relevant comparison may well be between the course presented when the student decided to enrol and the course actually available when they pick their modules – and it is hard to see how a large reduction becomes fair by being divided into separate decisions.

“Actually available” is carrying weight there too. A catalogue can list twelve options and offer very little real choice, once you account for the two that don’t run this year, the three that clash, the two that are capped, the one whose prerequisite wasn’t offered, and the allocation system that assigns students to what’s left. The practical effect is likely to matter more than the webpage – what students can in fact take, not what the catalogue says.

None of which means low demand or staffing problems are illegitimate reasons. But they explain a decision rather than justify it – and the questions that follow are uncomfortable.

Was the problem foreseeable? Did the university keep advertising modules it knew were unlikely to survive the restructure? As I noted when the DMCC provisions came in, recruiting applicants onto a portfolio you already suspect won’t exist once the redundancy round completes is not a strong position under any of this law.

The RPI example is no longer self-evidently safe

CMA182 currently offers fee increases linked to the Retail Prices Index as an example of how to do it acceptably. The CMA hasn’t withdrawn that example – but the revised guidance means it can no longer be treated as self-evidently safe. It would now have to be tested against a much fuller set of questions.

An index-linked clause has to answer for itself – whether the index bears any relation to the university’s actual costs, whether the calculation is predictable, what the base figure and reference date are, whether a cap is needed, whether the price comes down when the index does, and whether the student can realistically avoid the increase at all.

I’ve written before about how badly the sector’s fee promises perform against tests like these – and, by way of analogy rather than authority, about Ofcom concluding, for telecoms contracts a fraction of the length and value of a degree, that inflation-linked mid-contract rises were unfair enough to ban outright.

Two adjacent points. A statutory fee cap is a maximum a university is permitted to charge, not an instruction to charge it – so “fees will rise to the government maximum” is a policy choice dressed up as compliance. And the theoretical right to abandon your degree rather than pay the increase is unlikely, on its own, to amount to much of a safeguard.

Deposits and withdrawal get similar treatment. Labelling a deposit “non-refundable” settles nothing, and a staged fee-liability schedule can be clearly drafted and still be at substantial risk of being found unfair if it makes no allowance for costs saved or losses avoided following withdrawal – including whether the place could realistically be filled – or for circumstances in which the university’s own breach contributed to the student leaving. That’s not to say every institutional failing dissolves fee liability. But a schedule that ignores those questions entirely is exposed.

University-side cancellation – the under-recruited course pulled in August, the closed campus, the terminated franchise – needs grounds that are serious and specific, appropriate notice where possible, and refunds for what wasn’t supplied. In higher education, avoiding serious detriment is also likely to require practical help finding an alternative course or provider – a conclusion the guidance points towards rather than spells out.

And an event can be entirely outside a university’s control and still leave the student entitled to their money back for services not delivered – the excuse and the remedy are separate questions.

The guidance is explicit that disputes involving a trader’s own workforce may well be within its control – and foreseeable partner failure or thin contingency planning would also be difficult to defend as genuinely external, although that assessment would depend on the facts.

Stricter in the headlines, thinner in the substance

All of which, for England, reframes the argument about the Office for Students’ proposed condition C6 – the “treating students fairly” condition that would replace C1 and C3. The consultation closed on 9 July, David Kernohan covered the proposals here, and final decisions are due in the autumn.

Chunks of the sector pushback amounted to the argument that C6 goes beyond consumer law, and much of that case was built by comparing the proposals with the May 2023 higher education advice and the old CMA37. That comparison is now out of date.

A useful way of reading C6 against the new guidance is to sort its provisions into three piles – restatements of the law, applications of CMA guidance, and deliberate OfS additions. The condition’s strengths and weaknesses look rather different depending on which pile you’re standing in.

On hidden and linked terms, on intelligibility, on digital presentation, on responsibility for agents and delivery partners, the revised CMA37 has largely caught up with C6 – and on digital contracting it’s more detailed than the condition it was being unfavourably compared with.

C6 does go further than the law in places. An overarching duty to treat every student fairly across everything a provider does, selected grey-list terms treated as flatly prohibited, a requirement for free and independent complaints advice, the single webpage of documents – these sit in the third pile, regulatory choices rather than restatements, and OfS should defend them as such rather than presenting the whole condition as consumer law with a lanyard on. Whether students can actually use any of it is a separate problem Mack Marshall has set out on the site.

It’s the middle pile where the trouble sits. C6 is an independent regulatory condition, not implementation guidance for CMA37, so there’s no formal obligation to keep pace with the CMA – but where OfS has chosen to encode consumer-law concepts, it sometimes encodes them less completely than the CMA now does, and in precisely the areas most likely to generate the actual casework. C6 would prohibit unilateral changes “where no valid reason is specified in the contract” – a formulation that comes close to implying that specifying a reason is enough, which is the safe harbour the CMA has just spent several thousand words closing.

It recognises changes to “one or more modules” without requiring anyone to aggregate them. Its expectation that providers deliver all advertised and core modules as planned is simultaneously too absolute for provisional individual options and silent on the breadth point that matters more.

Its price provisions reproduce the grey-list concern about increases without a right to cancel, with none of the CMA’s analysis of whether cancelling a degree is a right anyone can really exercise. Its refund provisions are about whether policies clearly explain entitlement and calculation – and a clearly explained unfair policy is still unfair. And “circumstances reasonably outside the provider’s control” is an over-broad force majeure defence waiting to happen.

The risk is a parallel code that is stricter than consumer law in the headline areas and materially weaker in the ones where students lose out – with providers able to claim, in perfect good faith, that they comply with the regulator’s fairness condition while operating terms at substantial risk of being found unfair under the Consumer Rights Act.

From a regulator with form on running two tiers of fairness at once, that would be quite the achievement – two regimes, two answers, and the weaker one with the word “fairly” in its title.

The fix isn’t complicated. Before finalising C6 in the autumn, OfS should run a concordance exercise against the final July 2026 CMA37 – identifying which provisions restate the law, which apply CMA guidance, and which are deliberate regulatory additions. It should retain the additions openly, on their own merits.

And it should strengthen the provisions on variation, refunds, cancellation and practical exit to at least the level of the analysis the CMA has just published.