For more than a decade, England has struggled to recruit enough computing teachers. The problem has always contained an awkward labour-market tension: schools need people with precisely the technical skills that software companies, consultancies and other employers also value.
Now that competition may be changing.
Computing teacher recruitment has risen sharply just as the graduate technology labour market has weakened. This does not prove that computer science graduates unable to secure software jobs are moving into teaching. But the coincidence is striking enough to raise a more interesting question: is the changing technology labour market altering the relative economics of becoming a computing teacher?
If so, one unexpected consequence of the technology jobs slowdown could be to make the classroom a more attractive destination for graduates who might previously have looked almost exclusively towards industry.
Computing teacher recruitment has jumped
The improvement is substantial.
Department for Education figures show that England recruited 723 new postgraduate computing teacher trainees in 2025/26, up from 496 the previous year — an increase of 46 per cent and the highest number since comparable statistics began in 2012/13. The number of computing candidates also rose 16 per cent, from 2,347 to 2,714, while accepted candidates increased 42 per cent to 742.
That growth is considerably stronger than the wider secondary teacher-training market. Across secondary subjects, candidate numbers rose 5 per cent and acceptances 12 per cent over the same period. Computing is therefore not simply moving with a general improvement in teacher recruitment. There is, however, an important qualification.
The government also reduced the postgraduate computing recruitment target from 1,330 in 2024/25 to 895 in 2025/26. In 2024/25, 496 entrants represented just 37 per cent of the target; a year later, 723 represented 81 per cent.
So the dramatic improvement against target partly reflects a smaller denominator. The better measure is the actual number entering training and on that measure the 46 per cent increase remains significant.
At the same time, graduate tech hiring has weakened
The timing becomes more interesting when set against the graduate labour market.
The Institute of Student Employers’ 2025 survey found that graduate hiring among its member employers fell 8 per cent in 2024/25. Digital and IT experienced by far one of the sharpest contractions, falling 46 per cent, with employers surveyed forecasting a further 53 per cent fall in 2025/26. Those projections are not forecasts for the entire UK technology economy, but they offer a useful window into large graduate recruiters. Government-backed analysis published in June 2026 paints a similarly difficult picture at the junior end of the labour market.
Research by the Department for Science, Innovation and Technology and LinkedIn found that UK hiring in April 2026 was 14 per cent lower than a year earlier. Of 38 entry-level occupations tracked, 30 were declining. Entry-level software engineer hiring was down 27 per cent, alongside falls of 15 per cent for data analysts and 11 per cent for data engineers.
The same analysis found that knowledge-intensive sectors such as technology, finance and professional services had become relatively loose labour markets, giving employers greater scope to be selective. It also identified a mismatch between some of the general technical skills candidates possess and the more operational, production-ready skills employers increasingly seek.
The two datasets measure different populations and should not be treated as if one explains the other. But they reveal a notable divergence: computing teacher recruitment is rising while some of the graduate employment routes traditionally associated with computing degrees are becoming harder to enter.
That relationship deserves investigation.
The economics of teaching have changed relatively
For years, one of teaching’s difficulties was the opportunity cost facing a technically capable graduate.
Someone completing a computer science degree could choose among software engineering, data, cybersecurity, technology consulting and a range of related occupations. When technology companies were expanding rapidly, teaching had to compete with an industry offering attractive salaries and significant career mobility.
The calculation looks different when graduate vacancies become scarcer. Competition across the wider graduate market has reached historic levels. ISE members received an average of 140 applications for each graduate vacancy in 2025, compared with 38 applications per vacancy two decades earlier.
Teaching, meanwhile, offers unusually generous support in computing. Eligible postgraduate trainees in 2025/26 could receive a £29,000 bursary or a £31,000 scholarship. From September 2025, the minimum qualified-teacher salary in maintained schools was £32,916 outside London and £40,317 in inner London. Those incentives are not new. Computing bursaries were £27,000 in 2023/24, £28,000 in 2024/25 and £29,000 in 2025/26; scholarships rose over the same period from £29,000 to £30,000 and then £31,000.
That makes the 2025/26 recruitment increase particularly interesting. The bursary rose by only £1,000 from the previous year while entrants increased by 46 per cent. It would therefore be difficult to attribute the change to the higher payment alone. A more plausible possibility is that the value of the incentive has changed relative to the alternatives.
A £29,000 training bursary looks rather different when graduate technology firms are recruiting aggressively than when junior software vacancies are contracting and hundreds of applicants may be pursuing the same opportunities.
Careers, after all, are chosen comparatively. Teaching need not suddenly become dramatically more attractive in absolute terms. Its competing alternatives only need to become harder to access.
But is AI responsible?
This is where caution is essential.
It would be tempting to construct a neat narrative: artificial intelligence reduces junior software jobs, computer science graduates struggle to find work and teaching absorbs the surplus.
The evidence does not establish that chain. The DSIT-LinkedIn study explicitly warns against treating the association between AI capability and falling entry-level recruitment as causal. Although some of the largest declines occur in occupations where generative AI has become increasingly capable including software engineering — the report concludes that several forces are operating simultaneously and that considerably more research is needed to isolate AI’s effect.
ISE reaches a similarly cautious conclusion. Its assessment of the graduate market is that conditions are challenging, but employers are not generally reporting that they are replacing entry-level jobs with AI.
The broader labour market matters too. DSIT and LinkedIn note that UK hiring began weakening as the post-pandemic recruitment surge faded in late 2022, and that entry-level hiring is broadly tracking the wider market rather than collapsing independently of it.
AI may therefore be contributing to changes in the composition of junior work, particularly where tools can automate or accelerate information-processing tasks. But it would be premature to declare it the principal cause of the graduate technology slowdown.
A paradox for computing education
There is nevertheless an intriguing paradox here.
A buoyant technology industry can make recruiting computing teachers harder because schools and technology employers draw, at least partly, from the same pool of technical talent. When industry demand weakens, that competitive pressure may ease.
The technology sector therefore does not have to collapse for the relative attractiveness of teaching to improve. Even a reduction in junior vacancies could encourage graduates to consider careers they previously discounted. And that could have consequences beyond headline recruitment numbers.
Recent computer science graduates may bring contemporary exposure to programming, data science, AI, cloud computing and modern software-development practices. That knowledge could become increasingly valuable as England reviews and develops its curriculum and assessment system. The government’s Curriculum and Assessment Review has already prompted a wider reconsideration of what knowledge and skills pupils will need in a rapidly changing technological environment.
But technical knowledge alone does not make an effective teacher. The opportunity for teacher education is to turn strong subject expertise into strong pedagogy. That distinction matters.
Recruitment is only the beginning
There is also a danger in mistaking a favourable recruitment year for the solution to a workforce problem. If a weak technology market encourages more graduates into teacher training, schools gain little in the long term if those graduates return to industry when technology hiring recovers.
The strategic question is therefore not simply how many people begin training in 2025/26. It is how many are still teaching several years later.
Salary progression, workload, school culture, professional development and opportunities to maintain technical expertise will inevitably influence that decision. The challenge for policymakers is to convert a possible recruitment window into a sustainable profession rather than a temporary refuge from a difficult graduate labour market. There is also an opportunity for universities and teacher-training providers.
Computer science departments could introduce teaching as a career before students graduate rather than allowing it to appear only after unsuccessful job applications. School placements, tutoring schemes and structured opportunities to experience the classroom could expose technically capable students to a profession they might otherwise overlook.
Experienced technologists affected by restructuring or redundancy represent another potential source of expertise.
The message should not be that teaching is where people go when technology careers fail. It should be that teaching computing is itself a serious technology career — one in which technical expertise is applied to developing the next generation.
Follow the people, not just the numbers
The biggest gap in the evidence is surprisingly simple: we do not yet know enough about who the new computing trainees are or what they intended to do before applying. That is what needs to be studied next.
Teacher-training providers could ask incoming computing trainees whether they recently completed computer science degrees, what careers they originally intended to pursue, whether they applied for graduate technology roles, and whether changing employment conditions influenced their decision to teach.
Following the same trainees over several years would tell us something even more valuable: whether people who enter teaching during a weak technology market remain when that market recovers. Until then, correlation should remain correlation. But the numbers justify the question.
England recruited 46 per cent more postgraduate computing trainees in a single year, while ISE recorded a 46 per cent contraction in Digital and IT graduate hiring among its surveyed employers, and LinkedIn recorded 27 per cent lower entry-level software-engineer hiring year-on-year by April 2026.
It would be remarkable if these labour-market shifts had no effect at all on graduates’ career choices. For years, policy has largely approached the computing-teacher shortage from one side of the equation: make teaching more attractive through bursaries, scholarships and pay.
The current moment suggests that the other side deserves equal attention what happens to the alternatives outside the classroom.
If technology hiring rebounds strongly, schools may once again face intense competition for computing graduates. If entry-level technology employment remains structurally weaker, teaching could become a more prominent destination for people with computer science degrees.
The irony would be considerable. Technologies increasingly capable of writing code may yet contribute, indirectly, to more computer scientists entering classrooms to teach young people how computing works.
We cannot yet say that is happening.
But we now have enough evidence to start following the people behind the numbers.
Leave a Reply