This week
The first hard numbers for September 2026 arrived on Friday, when Enroly shared its latest intake data. Across the universities on its platform, acceptances for September are down 12.93% on last year, and CAS issued are down 30.33%. Enroly expects final enrolments to finish 26% to 36% down (source). At the same point last year, both of those numbers were growing by around 40%. The oddest number in the snapshot looks like good news: visa refusals are down 57.65%. Refusals collapsing at the same time as CAS issuance is not a coincidence, it is a mechanism, and it is the thread this issue pulls. The same week, Wonkhe asked whether international education is at an inflection point, and warned that managed growth is tipping into managed decline (source). On these numbers the tipping point has already arrived, and the cut is not coming from where most people are looking.
The numbers
September 2026 is tracking towards a 26% to 36% enrolment shortfall. Year to 3 July, acceptances are down 12.93% and CAS issued down 30.33% across Enroly's university group, against growth of 38.26% and 40.55% at the same point last year (source). The projection is not settled yet: August alone normally accounts for around 37% of a September intake's CAS, and Enroly's own caveat is that intakes have tended to improve slightly from this stage. Takeaway: build your reforecast on the minus 26% and minus 36% scenarios now, and treat August decision capacity as the single biggest lever you still control.
The map is redrawing: South Asia is down by a third, Africa and Europe are growing. By nationality, acceptances are down 55% for Pakistan, 27% for India, 26% for Sri Lanka and 21% for Nepal, while Ghana is up 28%, Nigeria up 26%, Kenya up 17%, China up 13% and the Philippines up 111% from a small base. By region: South Asia down 31%, the Middle East down 39% and Southeast Asia down 30%, against West Africa up 26%, Europe up 18% and Latin America up 15% (source). And the growth is not only Africa: acceptances from the USA are up 7%, South Korea up 24%, Spain up 16% and Egypt up 28%, small markets individually, but together exactly the diversification the sector keeps saying it wants. Every study level is negative, with postgraduate research hit hardest at minus 33% and undergraduate holding up best at minus 6%. Two health warnings before anyone redraws the strategy map. Refusals are rising in exactly these growth markets, up 100% for Ghana and 22% for Nigeria across the Enroly platform (source), and we flagged Nigerian refusal rates quadrupling in Issue 10. And an acceptance is a soft signal: a student can hold acceptances in several countries at once and enrol in none of them. Takeaway: the growth is real but the bases are small and the risk is rising with them. Ghana plus Nigeria plus Kenya cannot replace a third of South Asia; run the replacement maths, and hold the growth story to the same standard of proof as the decline story.
Policy watch
The Immigration and Asylum Bill is in the Commons, and one part of it matters to you. Introduced 30 June, second reading 13 July. Most of it is asylum and appeals machinery, not student policy, and it is not the levy bill (the levy needs its own legislation). The part worth watching is the new power for public authorities to share data with each other (source). The sector keeps asking the Home Office for better, faster data; this is the mechanism that could deliver it, or that could hand your data on without much say from you.
The LLE opens for applications in September, and the sector is looking the other way. Wonkhe argues distance learning is the only growth story left in domestic postgraduate study (on-campus UK PGT has fallen from 302,905 in 2020/21 to 241,240 in 2024/25), yet most universities still treat it as a side project, with Lifelong Learning Entitlement applications opening in September 2026 and the first funded modules from January 2027 (source). The same lesson keeps arriving from different directions: new government export data shows transnational education grew 17% in 2024, to £3.6bn, while growth in total education exports slowed to 2.5% as onshore enrolments fell (source). One caveat before anyone rewrites the strategy: over 80% of that TNE growth came from schools and early years provision, not universities. Flexible and offshore delivery keeps growing while the classic on-campus intake shrinks, but the university share of the prize is not automatic.
The signal
Wonkhe's inflection-point piece puts its finger on the real problem: no single policy did this. Universities could have coped with tougher compliance thresholds, or with more refusals, or with patchy data. It is all three arriving at once that is turning managed growth into managed decline (source). It also describes what that pressure does to behaviour. Universities start using compliance to compete with each other rather than treating it as a shared duty, they watch each other's ratings, and they make big recruitment decisions without ever seeing the evidence the Home Office is acting on.
The Enroly numbers show what that looks like once you add it all up. Each decision not to issue a CAS to a borderline applicant is sensible on its own: the thresholds are real, the ratings will be public, and the licence is the business. Added together across the sector, those sensible decisions become a cut of nearly a third, and no minister had to announce it. In Issue 9 we called the withdrawals surge a bust delivered by stealth. This is the same machine, working earlier in the pipeline.
The planning signal: your September problem is two problems sharing one number. Demand really is down; acceptances have fallen 13%, and Issue 10's affordability squeeze explains a lot of where. But the other 17 points of the CAS fall are happening inside universities, at a desk you control. One of those problems you can do something about this month. The other you cannot.
Spotlight · Sector · The de-risked intake
A declined CAS appears in no statistic. It is not a refusal, not a withdrawal, not a compliance breach. It is simply a student who never arrives, and this September there will be tens of thousands of them, one careful decision at a time.
Source: Enroly September'26 Intake Snapshot (July 2026). Year-on-year change in acceptances and CAS issued for the September 2026 intake, at the same date in each cycle, across the UK universities on Enroly's platform. Enroly covers a large share of UK CAS volume but not the whole sector, so read these as a leading indicator, not a census. An acceptance is also a soft measure of demand (students can hold several at once); CAS issued is the harder number, which is why the green bars matter more.
Start with the gap the chart shows: acceptances down 13%, CAS down 30%. Those two numbers should move together; this year there are 17 points of daylight between them, and the gap repeats in every major market. Nigeria: acceptances up 26%, CAS issued down 13%. China: acceptances up 13%, CAS down 16%. India: acceptances down 27%, CAS down 40%. Pakistan: acceptances down 55%, CAS down 67% (source). Students are accepting offers and then not getting CAS at anything like last year's rate. The narrowing is happening at the one stage of the funnel a university fully controls. One honest caveat: some of this gap may be delay rather than refusal, files held for extra checks that will still convert late. August will tell us which it is.
Now add the number that explains it. Visa refusals across the platform are down 57.65% year on year, with Pakistani refusals down 94%. A 94% fall in refusals from the market whose CAS issuance fell 67% is not the Home Office going soft. It is simpler than that: you cannot be refused a visa for a CAS that was never issued. Since 1 June the compliance pass marks demand a refusal rate under 5%, enrolment of at least 95% and completion of at least 90%, with public red, amber, green ratings from summer 2027 and recruitment limits for reds (source). We mapped that regime in Issue 6. This is what it does further down the pipeline: every borderline CAS is now a risk to the licence, so admissions teams interview harder, check finances earlier and turn away anyone who might not enrol, might not complete, or might be refused. The refusal statistics improve because the risk was removed before it could be counted.
The de-risking is not spread evenly, and that is the strategy story. Presenting the September'26 Intake Snapshot, Enroly's Katie Layt noted that around 30% of the university group is up year on year, and that those in growth are averaging roughly 50% up on last year. Strikingly, no type of institution is bucking the trend: every subset contains both growers and decliners, where at this point last year the split within each group was closer to 50/50. The position is worsening across the board, yet a third of institutions are still finding strong growth inside it. The channel mix is moving too: direct applications are now 41% of acceptances, and the share held by the ten biggest agents is shrinking (source). Some universities have found a way to grow inside the rules; most have simply cut. And the growth markets carry the next round of risk: refusals are rising exactly where acceptances are rising, up 100% for Ghana, 75% for India and 22% for Nigeria across the Enroly platform (source). A market growing 28% while its refusals double is a red rating in waiting under a 5% threshold. The tightening has further to run. The Home Office did not cut your September intake; your own risk appetite did. Know which CAS you are declining by policy and which by reflex, because only one of those is a strategy.
Things to think about this week
Rebuild your acceptance-to-CAS conversion rate by market against last September. The sector gap is 17 points; measure yours, then split it into deliberate policy (risk rules somebody wrote down) and drift (caution nobody actually decided on). The drift is enrolment you can still win back.
Treat August as the intake's swing month. Roughly 37% of a September intake's CAS historically issue in August, so this is the last normal week to add decision capacity: interview slots, financial-document turnaround, credibility-assessment staffing. A CAS you cannot process in August is a student you de-risked by accident.
Put the minus 26% and minus 36% scenarios in front of finance this month, with the deficit backdrop attached: roughly 50 providers already sit in the OfS's top two risk categories (source). A July reforecast is a plan; an October one is an excuse.
Jobs · Who's hiring
Liverpool Hope University, Director of International Student Recruitment (source).
University of Bath, Head of Transnational Education (source).
BIMM University, Associate Director of International Recruitment and Partnerships (source).
ESCP Business School London, Senior Business Development Manager, Executive Education (source).
The last word
For two years the sector has asked to be trusted to manage its own risk. The June rules granted the request in the hardest possible form: manage it, in public, to a 5% tolerance, or lose the licence. The Enroly numbers are the sector's first full answer, and the answer is a 30% cut it imposed on itself. Every admissions director in the country is making decisions they can defend; add all those defensible decisions together and you get the contraction Wonkhe is calling an inflection point.
The uncomfortable question is not whether to de-risk. It is whether you know the price of the particular way you are doing it. A declined CAS shows up nowhere: not in refusal statistics, not in a rating, not in a league table. It shows up only in your enrolment number, quietly, in October, mixed in with the genuine demand loss you could not have prevented. Universities that can tell those two apart, and there will not be many, get the chance to grow at exactly the moment most of the sector is pulling back.
So this week's question is blunt: how much refusal-rate headroom do you actually have under the 5% threshold, and how much conversion are you giving up to keep it? If you cannot answer with numbers, your risk appetite is being set by habit, not by strategy.
If this was forwarded to you, subscribe to receive every Tuesday at 7am UK time.
Sources
