Teacher Recruitment in a Shortage Subject: Practical Strategies for Heads
If you’ve spent this year staring at a physics or computing vacancy that’s been re-advertised twice, you’re not imagining the problem. The 2025/26 Initial Teacher Training Census, published by the Department for Education, shows genuine improvement in some shortage subjects alongside persistent gaps in others, and the practical question for heads isn’t whether the shortage is real but what to do about it beyond posting the same advert again and hoping for a different result.
The current picture, subject by subject
The headline is more encouraging than it has been for several years. Across all STEM subjects combined, new teacher trainees rose 21% on the previous year, and the STEM postgraduate initial teacher training (ITT) recruitment target was met for the first time since it was introduced. New entrants to postgraduate secondary training rose in 16 of the 18 subjects tracked in the census, with computing entrants up 44% and physics entrants up more than a third.
Physics is a useful case study in why “improving” and “solved” aren’t the same thing. There were 1,095 new physics entrants in 2025/26, the highest number since comparable records began in 2014/15, but that’s still only 78% of the government’s target of 1,410. Put another way: even after the best physics recruitment year on record, roughly one in five of the physics training places the country needs went unfilled.
Ten of the 18 secondary subjects missed their postgraduate ITT recruitment target outright in 2025/26: business studies, Classics, computing, design and technology, drama, modern foreign languages, music, physics, religious education and a catch-all “other” category. If your gap is in MFL, computing, D&T or the arts, the national training pipeline genuinely isn’t producing enough trainees to fill every vacancy this year, no matter how well an individual school advertises. That’s not a reason to give up on recruitment. It’s a reason to stop treating recruitment as the only tool, and to put real effort into the training pipeline, retention, and flexible ways of covering the timetable, alongside it.
Point candidates to the money, because most won’t have checked
One of the cheapest, most effective things a head can do costs nothing and takes five minutes: tell every promising shortage-subject candidate exactly what financial support exists. A striking number of career-changers and final-year students haven’t looked it up themselves, and a head who can talk fluently about the figures looks more credible than a job advert that leaves candidates to find them.
For trainees starting courses in the 2026/27 academic year, the Get Into Teaching bursary and scholarship scheme offers tax-free bursaries of £29,000 for postgraduate trainees in chemistry, computing, maths and physics, with scholarships of £31,000 available as an alternative route into physics, computing and chemistry, delivered through the relevant subject association rather than the training provider itself. Languages are structured slightly differently: French, German and Spanish trainees can access a £20,000 bursary or a £22,000 scholarship, while trainees in other languages, including ancient languages, get a £20,000 bursary with no scholarship option attached. For the 2026/27 cycle, languages and physics candidates no longer need to meet the usual student finance eligibility rules to qualify for this funding, a detail worth flagging explicitly to older career-changers who assume they’re excluded.
There’s also a route worth knowing about for undergraduate recruitment: trainees starting an undergraduate course leading to QTS in secondary maths or physics can access a £9,000 bursary in their final year. If you have any relationship with a local university’s undergraduate physics or maths department, or sixth-form alumni weighing up teaching as a career, this is worth raising directly rather than assuming they already know.
Don’t forget the retention half of the money
Once someone is hired, the targeted retention incentive is the payment most likely to be missed simply because nobody mentions it. Early-career teachers of chemistry, computing, maths and physics working in disadvantaged schools can claim between £3,000 and £6,000 a year, depending on the school’s disadvantage banding, in each of their first five years in post. It’s opt-in: the teacher has to apply within a claim window themselves. A school that proactively reminds eligible staff and helps them get the paperwork right is quietly buying itself another year of retention in a shortage subject, for the cost of a conversation.
Strategies that work harder than another job advert
Make part-time and job-share the default, not the exception
NFER’s research on flexible working in teaching is unambiguous: there is an unmet demand for part-time work among teachers, particularly in secondary schools, and it’s a demand that drives people out of the profession when it goes unmet. Flexible working is consistently linked to better wellbeing, higher job satisfaction and lower burnout among teachers, all of which show up in the data as staff who stay rather than leave. Because part-time and flexible arrangements are still much less common in secondary than in primary schools, a secondary school that visibly offers a job-share or reduced-hours option for a shortage-subject post is doing something most local competitors aren’t. That reaches a pool of candidates who never apply for a full-time-only advert: parents returning after leave, teachers building a portfolio career alongside industry work, and experienced staff nearing retirement who’d happily stay on for fewer hours.
Grow your own through ITT partnerships
Hosting trainees through a School Direct or university ITT partnership isn’t only a favour to the sector; it’s a low-cost way to see a future colleague teach in your own classrooms for a year before you decide whether to offer them a job. Schools with a genuine, ongoing partnership with a training provider get first look at trainees, rather than competing for them on the open market once they qualify and every other school in the area can see them too. If your shortage-subject gap has been open for more than a term, a conversation with your nearest ITT provider about hosting a trainee next year is arguably a better use of that time than a third round of advertising the current vacancy.
Salaried routes are worth understanding for the same reason. Both School Direct salaried places and Teach First pay trainees a salary while they train, which opens the door to strong career-changers who simply can’t afford a low- or unpaid training year, often exactly the profile of person who ends up teaching physics or computing after a first career in engineering or software. Teach First in particular is built around placing trainees in schools serving disadvantaged communities for a minimum two-year commitment following an intensive initial training period, which makes it a route worth exploring specifically where a shortage-subject vacancy sits in a higher-deprivation school.
Use supply-to-permanent conversion deliberately, not as an afterthought
If you already have a supply teacher covering a shortage subject who is proving themselves in front of your pupils, that’s a lower-risk hire than an unknown candidate from an open advert; you’ve already seen them teach your classes, in your building, under your behaviour systems. Some agencies charge a conversion or finder’s fee when a supply placement becomes permanent, so negotiate that fee, or a cap on it, into the contract before the placement starts rather than after you’ve already decided you want to keep them. Treating a longer supply placement as a deliberate trial period for a hard-to-fill post, rather than only ever a stop-gap, turns a cost you’re already paying into part of a genuine recruitment strategy.
Agencies versus direct advertising: what actually pays off
Recruitment cost analyses from sector specialists suggest that advertising and vetting a role yourself typically costs somewhere in the £1,000 to £3,000 range, while a traditional agency placement, once finder’s fees, staff time and onboarding are counted, can run from £6,000 up to £19,000 depending on the seniority of the role, with agency finder’s fees commonly charged at around 15% of first-year salary. On a typical teacher’s salary, that’s several thousand pounds for a single hire, a serious sum for a school already managing a tight budget.
That doesn’t make agencies the wrong choice in every case. For a genuinely hard-to-fill vacancy, very short notice, a rural school with a small local candidate pool, or a subject where direct routes have already been tried without success, an agency’s wider reach can be worth the fee. The practical advice is to negotiate before you’re desperate rather than after: ask for a fixed fee instead of an open-ended percentage, ask what happens if the hire leaves within a term, and be wary of running the same vacancy through several competing agencies at once, which tends to inflate perceived scarcity and price rather than genuinely widen the pool of candidates.
Before reaching for an agency at all, make sure the free channels are being used properly. Teaching Vacancies, the Department for Education’s own national job-listing service for state-funded schools, exists specifically to reduce reliance on paid advertising, and the service’s own figures put the total savings to schools at somewhere between £63.7 million and £84.2 million in advertising costs since it launched in September 2018. Subject associations are worth listing with too: the Institute of Physics, the Association for Language Learning, Computing at School and equivalent bodies in other shortage subjects will often carry vacancies, and they put a job in front of people already engaged with the subject rather than simply browsing job boards.
The most underused channel is usually a school’s own network. A trainee hosted two years ago, a supply teacher who covered well last term, a former colleague on parental leave who might want reduced hours on return: keeping light-touch contact with people like this before a vacancy exists means a school isn’t starting from a blank page every time one opens.
Recruitment and retention are the same problem, seen from two ends
Every specialist a school keeps is a vacancy it never has to fill again, in a subject where filling it took real, sustained effort the first time. That reframes retention from a nice-to-have into arguably the most cost-effective recruitment strategy available to any head.
The particular risk for shortage-subject staff is being spread too thin. A school with a single physics specialist or a single computing specialist can end up leaning on that person to cover classes outside their expertise, take a disproportionate marking load, or plug timetable gaps precisely because they’re rare and valuable, which is one of the fastest ways to lose them. Protecting a shortage-subject specialist’s timetable so it stays genuinely within their specialism, rather than treating them as a general-purpose science or technology teacher for the school’s convenience, is a retention decision as much as a curriculum one.
Workload is the other lever worth pulling directly. Planning and resourcing lessons from scratch every week, on top of marking and the rest of a normal teaching timetable, is one of the most commonly cited reasons teachers give for leaving the profession, and it bites hardest in subjects where a single specialist is covering multiple exam boards or key stages without a department around them to share the load. Giving a shortage-subject teacher access to good, ready-made curriculum resources, free ones like those on Learnaroo Hub can cover some of this ground, won’t fix workload on its own, but it removes a real chunk of from-scratch planning from someone’s evenings. For a candidate weighing up a shortage-subject post against an easier-to-staff subject elsewhere, being able to say that honestly is a genuine selling point rather than a gimmick.
Isolation matters too. Being the only specialist in a subject in a building is genuinely different from teaching in a department of six, and it’s a documented retention risk in shortage subjects specifically, where a school may only employ one or two specialists at all. Protected time and a small budget for external subject networks, the Institute of Physics’s teacher networks, the National Centre for Computing Education, or regional MFL hubs, gives an isolated specialist the peer contact their own building structurally can’t provide. The same logic applies to early-career mentoring: a shortage-subject recruit who came through Teach First, School Direct salaried, or a straightforward career change may have thinner subject-specific support available in-house than a mentor teaching a different subject can genuinely offer, so it’s worth arranging subject-specific mentoring with a neighbouring school if it can’t be provided internally.
Finally, give shortage-subject specialists a visible career path that doesn’t require leaving the classroom to get it. A TLR for leading a genuinely hard-to-recruit subject, involvement in exam board work, or a cross-school subject lead role lets someone build a career as a physics or computing teacher, rather than concluding their only route forward is out of the subject and into generic leadership.
What this looks like in practice
None of this is a single fix, and nothing here replaces having a genuinely good school to work in. But a head who can point a candidate to the real bursary and scholarship figures, who offers a part-time or job-share option by default rather than as a special favour, who has an actual working relationship with an ITT provider, who negotiates agency fees rather than accepting the first quote, and who protects a shortage-subject specialist’s timetable and workload once they’re in post, is doing considerably more with the same budget than a school relying on a single job advert repeated every half-term. In a genuine national shortage, that difference is often what determines whether a physics, computing or languages post gets filled this year, and whether whoever fills it is still there in three.
