Universities are bound together more tightly than ministers like to admit. They share credit lines, pension schemes, suppliers, and reputations. Contagion, once started, moves faster than policy can catch it.
The question up until recently was the wrong one: could a university fail. The grown-up question is what happens next: to students who haven’t applied yet, to local communities, and to neighbouring universities. We have to begin with the obvious: students come first.
Failure at one institution produces contagion effects, the magnitude of which depends on regional centrality and clustering. Government should focus on keeping that transmission reproduction number (or R number – remember that from Covid?) below one. This piece maps some of those transmission channels – I’ve modelled small changes to the bottom line of an average neighbouring university, based on student spend, pensions, interest rates and group buying.
Our patient zero – that is, the first to fall – is a provider that is OfS-registered and regionally significant, and I have estimated the price shock for its financially average neighbour. Of course, I have to assume no rescue package from government (they have signalled as such).
The calculations below are based on the average university using 2023–24 HESA data, excluding FE colleges with HE provision. Each percentage change is illustrative rather than predictive and actual outcomes will depend on local factors.
Stay at home
Student demand runs on policy signals and vibes as much as price. In 2024 we saw sponsored study visas fall year-on-year and dependants drop sharply, while PGT overseas remained twitchy.
Throw a closure into that salad and you start to see conversion eroding. Bursary spend and fee waivers will rise to keep offers attractive. A percentage point here or there looks insignificant but adds up across multiple providers.
International students will start looking elsewhere: Australia, Canada, Germany. Or they’ll just stay at home. Better to choose a sure thing than risk having your course disrupted halfway through. Home students may be similarly spooked – but have fewer alternatives.
There are a few antidotes: multiple guaranteed transfer corridors, decent student protection plans, and teach-out clarity. And most importantly, comms that make sense to agents and parents.
Illustrative hits to an average university elsewhere:
- Additional bursary spend on international: £80m × 0.5% = £400k
- Reduced international demand: £80m × 0.5% = £400k
Protect the USS
Multi-employer pension schemes, like USS and LGPS, can go very squiffy when a member exits. In that case, the rules force the member to pay a large exit bill called “Section 75”, and the sums can be eye-watering. It’s a standard expectation of a “last man standing” scheme.
Trinity College, Cambridge wrote a cheque for about £30m to leave USS in 2019. USS has suggested that, for a sample of employers (mainly Oxbridge colleges), a crystallised bill could represent anywhere from 4 to 97 per cent of their cash and long-term investment balances, averaging around 26 per cent.
In practice, an insolvent provider wouldn’t cough up, so other universities would absorb the orphan liability. But there isn’t a mechanical “spread the S75 bill this year” formula; it would show up, if at all, via the valuation and rate-setting process. The scheme is currently in surplus, so additional contribution costs are uncertain. Of course, not all universities are enrolled in USS, but the vast majority are enrolled in multi-employer schemes.
Illustrative hit to a USS-enrolled university elsewhere:
- Salary base: £181m × 72% = £130m
- USS proportion: £130m × 70% (say) = £91m
- 1 pp rate bump: £91m × 1% = £910k
Save livelihoods
Universities drive jobs, rents, transport and culture. Liverpool estimates £2.2bn GVA and 26,630 jobs supported nationwide, roughly one in fifty locally. Northampton reports £823m GVA and 10,610 jobs. National estimates put the sector above £116bn.
Remove the local provider and the GVA virtuous circle turns vicious. Cafés lose footfall, landlords lose tenants (poor them), and pubs are no longer full of students. The extent depends on how rooted the provider is in its community.
Government will find itself paying anyway. Either pre-emptively with small civic grants to keep key services alive, or retrospectively with bigger cheques after the rot sets in. Maybe it will finally put a stop to town and gown tensions.
Illustrative hit to an average university elsewhere:
- No direct cost to other universities
- Material GDP and tax impacts for government
- Likely need for community grants.
Flatten the yield curve
Lenders rarely treat a closure as an isolated blip; being hawkish, they would probably reprice the entire university category.
Add 50 basis points to a £90m facility and you’ve created a recurring £450k drag until you refinance. All in all, that’s not a huge bite out of your cash flow, but it will certainly make you more cautious.
To fix this, listen to your finance directors: stagger your maturities and fix your rates well in advance. Or, radical thought – stop yanking at your credit lines and make do with what you have.
Illustrative hit to an average university elsewhere:
- Additional interest costs: £90m × 0.50% = £450k
Herd immunity
Group buying is one of the few places with cash on the table. In 2023–24, the UK Universities Procurement Consortia (UKUPC) members put about £2.4bn through frameworks and reported roughly £116.1m (4.84%) in cashable savings. The Southern Universities Procurement Consortium (SUPC) talks about £575m of member spend and average levy rebates of around £30,000 per full member.
If fewer universities use those routes, frameworks lose clout, and with that, discounts and rebates. The more volume that stays in the collective pot, the better the prices – but for critical services, it’s still wise to have a backup supplier in case one fails.
Another group issue is shared services. Up until recently, they were seen as a poisoned chalice, but are now growing out of necessity. The usual worries are well-rehearsed: loss of control, infighting and VAT jitters. Still, some experiments, like Janet and UCAS, have been tremendously successful, although pricing relies on throughput.
Shared IT, payroll, procurement or estates often come with joint and several obligations. If one partner hits trouble, you start to see real governance friction.
The practical fixes are contractual. Ringfence any arrears so they do not spill onto everyone else, and rebalance charges on a published, defensible formula.
Illustrative hit to an average university elsewhere:
- Frameworked spend: £131m (total non-staff) × 60% (frameworked, say) x 4.84% (cashable savings) x 10% (diminution) = £380k.
- Shared services: impossible to quantify.
What ministers can do without a podium
I’ve modelled small changes to the bottom line (again, illustratively) – in this example one university going under could cost others £2.5m, or 50 per cent of the average university’s 2023–24 surplus. This number isn’t rigorous or comprehensive, but serves as an interesting thought experiment.
The rational response is a resolution regime that protects students and research, temporary liquidity for solvent neighbours, clear transfer routes when the worst happens, and deployment of short, targeted grants for civic programmes.
A single collapse could probably be absorbed; a string of them could set off an irreversible domino effect with far-reaching consequences. Ministers need to plan for this now – or else risk a very hefty civic bailout.
