What the data shows
The ITT Census reports, for each subject, how many candidates applied and how many were accepted onto a course, alongside how many actually entered training.
For physics: 5,657 candidates, 1,258 accepted – a rate of about 22%. Computing: 2,714 candidates, 742 accepted (27%). Business studies: 1,147 candidates, 219 accepted (19%).
Against target, physics reached 78%, computing 81% and business studies 30%.
A candidate is not a person, and not an application
Two cautions before drawing conclusions, both of which DfE states directly.
A candidate can apply for more than one subject or course. Candidate counts therefore cannot be added across subjects to give a number of unique people, and an acceptance rate calculated from them is not a per-person success rate.
The figures are also best used to compare subjects with each other and with previous years, rather than as an absolute measure of how selective any single subject is.
Where the losses actually occur
The common assumption is that subjects miss targets because too few people want to teach them. The pipeline data complicates that.
Physics attracted a substantial candidate pool relative to the trainees it needed, and still finished below target. That points to attrition through the process – withdrawals, unsuccessful applications, offers declined, accepted candidates who never start – rather than an empty top of the funnel alone.
Business studies looks different again. It finished furthest below target of any named subject, and its estimated postgraduate trainee need for 2026/27 has been revised upward to 1,200 from a 2025/26 target of 900: demand rising as recruitment falls.
Physics has moved the other way. Its 2026/27 trainee need has been revised down to 810 from 1,410 – about 43% lower. A subject can therefore appear to improve against target without a single extra physics teacher being trained.
Why this matters
For policy, it matters a great deal whether a shortfall sits at the top of the funnel or partway down it. Bursaries address the first; application support, selection practice and offer conversion address the second. Physics carries the highest bursary available – £29,000 tax-free – and still finished below target.
For anyone tracking recruitment, percentage-of-target moves for two independent reasons: the number recruited, and the target itself. Both have to be read together.
And the full candidate-to-trainee picture is the more interesting object. Set candidates, accepted candidates, entrants and future trainee need side by side for every subject and you have a supply pipeline rather than a scoreboard – which is where this analysis is heading next.