Special School Teachers: Where do they come from? Where do they go?

In recent years, teachers have tended to be less likely to leave a special school than other types of schools

Turnover measures the percentage of teachers who leave the school they are in from one year to the next (including moving to a different school in the state sector). Turnover will be affected by a wide range of factors. It is not necessarily true that a high turnover rate in a given year is a bad thing (for example, teachers may be leaving a school to take up promotions elsewhere). However, high turnover rate has been linked in previous evidence (from English schools) to poor job satisfaction and puts pressure on a school to recruit more teachers.

The most recent teacher turnover rate (measured from November 2022 to November 2023) was lower in special schools than in primary and secondary schools. Around 15 per cent of teachers in special schools in November 2022 left the school within a year, compared to 17 per cent in primary and 18 per cent in secondary. Indeed, the turnover rate has been consistently lower in special schools over the last eight years. The only exception is the Covid-19 pandemic period, when turnover rates fell across all schools but rates in special schools did not fall by as much. Conversely, average turnover rates have been consistently higher in AP settings than other school types.  

As in our first blog, we also looked at how workforce challenges vary across different types of special school ^[2]. In the latest data, schools that primarily serve pupils with Social, Emotional and Mental Health (SEMH) had a higher rate of teacher turnover, on average, than other special schools. In our previous blog, these schools also had higher vacancy rates, which is likely to be linked to higher turnover rates.

Teachers who leave special schools are more likely to leave the state-funded system than teachers who leave primary or secondary schools, but this could be due to data limitations

When teachers leave special schools, where do they go? The data consistently shows us they are more likely than other teachers to leave the state-funded system entirely. For example, in the latest data, 69 per cent of special school teachers who left their school left the state system. The equivalent rates for primary and secondary were 60 per cent and 56 per cent, and the data from previous years shows very similar trends.

There is an important caveat here. ‘Leaving the state-funded system’ is not the same as ‘stopped teaching’. Teachers in various types of schools, like independent special schools, are not recorded in this data, and any teacher who moves to a school like that would be counted as ‘leaving the state-funded system’. It is possible that teachers in special schools are more likely to move to these types of schools, particularly in areas where a lot of state-funded pupils attend independent special schools. It is difficult to gauge whether, if we could account for these missing schools in the data, it would change the results above.

The chart above also shows that special school teachers (and those in AP) are more likely to join a different sector when leaving their school than teachers in mainstream schools. For special schools, joining a ‘different sector’ would mean joining a primary, secondary or AP setting. Teachers who leave primary and secondary schools tend to join other primary and secondary schools, respectively, if they join another school. However, the same is not true for teachers in special schools. A likely explanation here is that there are far more primary and secondary schools. Therefore, a teacher who is seeking to leave a special school may have more limited opportunities to join another special school within a commutable distance and could therefore seek opportunities in different types of settings. This is less likely to be true for teachers in mainstream schools, where more options within the same sector will be available locally.

This does not mean special schools lose more teachers to mainstream schools each year than come the other way. That is because, in raw numbers, far more teachers leave mainstream schools each year than special schools. Indeed, as the table below shows, more teachers joined special schools from other school types in the most recent data than went the other way. For example, around 270 teachers moved from special schools to primary schools in 2022/23. Nearly three times as many, 750, moved the other way.

The number of teachers that moved between schools between 2022/23 and 2023/24, by setting type of their old and new school: ^[3]

2022/23 to 2023/24			Moved from:
			Special	Primary	Secondary	AP
Moved to		Special	681	753	350	86
		Primary	269	15,351	462	38
		Secondary	198	715	17,842	138
		AP	50	97	219	116

So far, we have seen that teachers are slightly less likely to leave special schools than other schools, but when they do leave special schools they are more likely to leave the state system. Putting this together in the chart below, we find teachers in special schools are slightly more likely than teachers in primary or secondary schools to leave the state-funded system. In the latest data, around 10 per cent of teachers in special schools in 2022/23 had left the system by 2023/24. For primary and secondary schools the most recent rate was only slightly lower, although that gap has been wider in previous years. Again, it should be noted that this could be due to the fact we cannot observe teachers moving to independent special schools.

We also see that the rate at which teachers leave the state system from AP settings is higher than other schools. Theoretically, this could also be due to the data limitations outlined above but given we have previously seen that AP schools have higher turnover and vacancy rates, this should be seen as a sign of the specific workforce challenges in those schools.

Historically, attrition rates have tended to differ substantially by seniority. Classroom teachers are the most likely to leave the state system whilst middle leaders are least likely. The most recent data suggests this pattern is very similar across special, primary and secondary schools, as in the following chart. We also see that attrition rates for middle and senior leaders in AP settings are very high, relative to other sectors.

Finally, we also looked at attrition rates for early career teachers (those in their first five years of teaching), in the following chart ^[4]. We found early career teachers are slightly less likely to leave the state-funded system from one year to the next if they were teaching in a special school. Just under 10 per cent of early career teachers in special schools in 2022/23 had left the system by 2023/24, compared to averages of around 11 per cent for early career teachers in primary or secondary schools. This pattern has been consistent over time.

Teachers in special schools are generally as experienced as those in primary and secondary schools, and this is true at different grades

On average, teachers in special schools appear to have similar amounts of teaching experience (13 years) to teachers in primary and secondary schools ^[5]. The average tenure is currently a little lower at special schools, despite the lower turnover rate we saw earlier. This appears to be explained by composition effects: there has been an increase in the number of teachers in the special school sector over this period ^[6]. We found relatively limited differences in average levels of experience and tenure across types of special school.

Teachers in AP settings appear to have considerably more teaching experience on average, at just under 16 years. Despite this, the average tenure of an existing teacher in these settings is five and a half years, substantially under the national average of seven years. This is not surprising given the higher turnover rates in these settings, but it once again emphasises the recruitment and retention challenges that the AP sector faces.

The experience of leaders does not appear to vary strongly across special schools, compared to primary and secondary schools. On average, a middle leader has around 15 years' experience, and a senior leader around 20 years. The average leader in a secondary school has slightly less experience than senior grades in other types of schools. Overall, this probably suggests that career progression occurs at a similar rate across school types, at least on average ^[7].

Special schools rely less on ITT for their annual recruitment than primary and secondary schools. Special schools recruit more from primary schools than secondary schools

Among teachers joining special schools in 2022/23, just under 20 per cent joined from a recent ITT route ^[8], as shown in the following chart. This was lower than in primary and secondary schools, where the average was about 30 per cent. Special schools appear to rely less on flows directly from ITT for recruits.

This may be explained by a lack of a focused route on ITT courses for teaching pupils in a specialist setting, which may be confusing for a trainee interested in doing so. Currently, trainees have to choose between either a primary or secondary-focused course ^[9]. A small proportion of primary and secondary courses offer a ‘SEND specialism’, whilst the ITT core content framework requires providers to offer training in educating pupils with SEND to all trainees. Some ITT providers do offer placements in special schools (recent example here), with a view to helping trainees work with children with SEND in both mainstream and special settings.

More could be done to increase the flow of trainees to special schools. A relatively radical option would be to set up an ITT route with courses specifically for trainees who want to teach children in specialist settings. Whilst this would have the advantage of being a clear route to special schools for these trainees, there is a risk it gives the impression that teaching children with SEND is somehow fundamentally different to educating other children. It may even reduce the potential for teachers to move between special schools and other schools. Indeed, the data above suggests many teachers move to special schools having previously worked in different types of settings, and vice-versa.

Instead of a new route, there are options within the existing system. For example, clearly ensuring that placements in special schools can fulfil trainees’ qualification requirements and communicating that clearly. We have been told that some training providers felt they needed more reassurance about the extent to which they can use special schools for placements. The reliance on teachers moving from mainstream schools into special schools suggests more could be done to emphasise SEND expertise in the Early Career Framework too. This would help ensure new teachers across the system are developing these skills and are able to take up opportunities in special schools if they want to.

Finally, the chart above also shows that special schools and AP source their recruitment more from other locations not immediately trackable in the data. This could include recruits who did not pass through ITT in the last three years, returners to the teaching profession, teachers joining from overseas, or teachers from other schools not covered in the data (such as independent special schools). This data reinforces the impression that special schools use different routes to mainstream schools to recruit, with fewer teachers coming directly from ITT and fewer teachers in special schools having QTS, perhaps because of a shortage of candidates with QTS. More work should be done to understand special schools’ recruitment decisions.

Trainees who join special schools tend to have done a primary course, but ITT targets do not account for this

We looked at the profile of teachers who joined special schools directly from or whilst on an ITT route and compared the routes they took to those joining other schools ^[10]. First, we look at the different training routes joiners took, via three main routes:

1. Higher education routes, which will typically mean a postgraduate teacher training course, but also includes courses for undergraduates
2. Employment-based routes, which are where teachers receive a salary while training (including Teach First, School Direct – salaried and apprenticeships).
3. School-centred ITT routes, which are fee-paying ITT provision run by schools.

As the chart above shows, there are some differences between special schools and other schools in this data. Special schools appear to rely more than other schools on employment-based routes, with just under half of all recently trained teachers in special schools using this route.

The next chart shows the ITT specialism taken by recent or current trainees joining different types of schools in 2022/23. Due to statistical limitations, it was not possible to extract data about the individual secondary subjects taken for Special and AP trainees, so we simply classify the data into ‘Primary’ and ‘Secondary subjects’. Here, we can see that recent or current trainees joining special schools tend to have taken a primary course, probably reflecting the curriculum typically taught in special schools. The opposite is true for AP, where most pupils are secondary age.

The reliance on special schools for teachers trained in primary ITT could cause problems in future. ITT recruitment into primary subjects has generally been healthy over the past decade, but the postgraduate recruitment target has been missed for the last three years. Given special schools also appear to rely on these recruits, this could be a sign that special schools are going to find the labour market tighter in future.

More generally, the health of ITT recruitment into special schools (and AP) is not directly measured by DfE in the way it is for primary and secondary teachers. There is no specific ITT route with its own reporting and target for special schools, as there is for primary teaching and every secondary subject. The number of trainees doing a primary or secondary course with a SEND specialism is also not reported.

Furthermore, the target for primary (and secondary) trainees does not explicitly account for the system’s requirement for special school teachers, the recruitment of trainees into special schools, and the flow of teachers from mainstream schools into special schools. This means these targets are almost certainly underestimating the total number of new trainees that the system really needs. This problem may have been masked in the past by a healthy amount of over-recruitment into primary ITT, but may become more pressing as primary recruitment falls repeatedly below target.

The lack of targets and reporting here could mean future shortages of teachers in special schools come to the attention of policy makers and the wider sector less quickly than shortages in other areas. Given that teacher vacancy rates in special schools are as high as those in secondary schools and the education of SEND pupils in specialist settings remains a key priority for the Department, the DfE should directly incorporate the demand for special schools' teachers into its workforce planning. That would mean specifically tracking recruitment into specialist settings from ITT routes, and explicitly and transparently accounting for special school demand for teachers within existing (or new) ITT targets.

Conclusions and Recommendations

In addition to the recommendations in the first blog (link):

• Given the relatively high teacher vacancy rate in special schools and the lack of a SEND or special school focused target in ITT recruitment statistics, DfE should conduct more deliberate workforce planning for special schools (and AP), in line with existing processes for primary and secondary school teachers. On the demand side, this means explicitly including special schools when determining overall teacher needs. On the supply side, DfE should account for flows of teachers from different ITT courses into special schools when setting ITT targets.
• The data suggests special schools do not recruit their teaching staff from ITT routes as much as primary and secondary schools. Policy makers and sector leaders should assess the extent to which this is due to shortcomings in the existing ITT system and consider options to improve the flow of trainees into special schools.
• The wider system seems to rely on teachers developing experience with SEND pupils in the mainstream sector. Further consideration should be given to this issue, including the emphasis on SEND in the Early Career Framework and the role nearby special schools can play in helping build up experience across the teacher workforce.
• The heightened turnover rate and attrition rate in AP settings is clearly a cause for concern. It also reinforces our previous recommendation that DfE should publish statistics for special schools and AP settings separately.
• The data tentatively suggests existing teachers in special schools are more likely to leave the state-funded sector each year. Given the current policy emphasis on helping mainstream schools become more inclusive and capable of educating pupils with SEND, more could be done to help teachers in special schools who would otherwise leave teaching to stay. Achieving this by helping them move into jobs where they can share their expertise in mainstream schools could be a useful way of both tackling shortages and improving provision across the system.

Finally, more research should assess how challenges differ across special schools, rather than treating the sector as a homogenous block. For example, in our analysis, schools where a majority of pupils have SEMH as a primary need faced significant teacher recruitment and retention challenges, which were often more similar to AP settings than other special schools. Further work can help sector leaders understand how challenges vary across special schools.

This project has been funded by the Nuffield Foundation, but the views expressed are those of the authors and not necessarily the Foundation. Visit nuffieldfoundation.org.

It was undertaken in the Office for National Statistics Secure Research Service using data from ONS and other owners and does not imply the endorsement of the ONS or other data owners. Thanks to all those who reviewed these blogs and offered helpful comments before publication, and to NFER staff for helping prepare the data.

References

[1] As in our first blog, non-maintained special schools (which are not academies) are not included in this analysis. Independent special schools, which often have publicly-funded pupils, are also not included.

[2] See our first blog (link) for how we define these categories and the limitations of this experimental approach.

[3] Teachers who left a school but did not join a new school in the data are omitted from this table.

[4] The number of early career teachers in AP settings is low. This means the attrition rate is influenced by small numbers of teaching leaving, which is why the series here is more volatile than the other school types.

[5] We measure experience here by measuring the time since the teacher’s recorded first year of teaching. We do not account for any intervening years when someone was not teaching.

[6] School workforce statistics suggest 28,200 teachers (FTE) were working in ‘State-funded special or PRU’ in England in 2023/24, sharply up from 21,600 in 2015/16. At the same time, school statistics show the number of special schools increased whereas the number of AP schools fell over the same period. As such, it is almost certainly the case that state-funded special school teacher numbers have increased over this time.

[7] More work could test this finding, by controlling for the effect leaders leaving teaching have on these averages, more accurately counting years of experience, or even measuring experience in different types of schools and/or grades.

[8] Joiners from same/different sectors refers to teachers who were at a different school in the year before. We define ITT joiners as those who joined the school and have an ITT record within the previous three years. Trainees who join one school for their training year and then join a different school in the following year are not counted as ITT joiners in that following year. Trainees who were at their school (e.g. as an unqualified teacher) are counted as an ‘ITT joiner’ in the year they appear in ITT performance profile data.

[9] There is also a further education ITT pathway.

[10] We define newly qualified teachers here as teachers in the first year of teaching during or after their final year of ITT .