Oral evidence: Supply of teachers, HC 199Wednesday 8 June 2016

Oral evidence, 8 June 2016.

Witnesses

Jenni French, Programme Manager in the Education Team, Gatsby Charitable Foundation, Joanna Hall, Deputy Director for Schools, Ofsted, Chris Husbands, Vice-Chancellor, Sheffield Hallam University, Mark Parrett, Audit Manager, National Audit Office, gave evidence.

Chair

Good morning and thank you very much for coming to our session. I think you were all in the audience for the previous one, so you know we have covered CPD. We want to now get a proper handle on how we measure these things and where we go from there. Could you say who you are and what you are representing?

Chris Husbands

I am Chris Husbands. I am now Vice-Chancellor at Sheffield Hallam University, which I was informed this morning is now the largest provider of ITT nationally. Previously, I was Director of the Institute of Education until December 2015.

Jenni French

I am Jenni French and I am representing the Gatsby Foundation.

Mark Parrett

I am Mark Parrett. I am a manager at the National Audit Office responsible for our value for money work with the Department for Education and specifically responsible for our report in February on training new teachers.

Joanna Hall

Good morning. I am Joanna Hall. I am Deputy Director for Schools at Ofsted and I also have initial teacher education in my brief as well.

Chair

Very useful, and thank you all for coming along. The Committee has received evidence that the Government’s data on vacancies is not representative of the real situation. Is that correct, Mark?

Mark Parrett

I think from what we concluded back in February, the Department has been focused on the national picture and has been approaching this from a national perspective. I think what is most important is there is a lot more to be done to look at the local picture and the pressures that schools are facing to fill teaching posts. The national perspective tells you one thing. It does not tell you enough about what is happening locally and I think that is the most important thing.

Chair

Do you think we should break things down into regions?

Mark Parrett

Whether it is regions or local authorities or another definition—if you break things down into regions, there is still significant variation within a region—yes, it should certainly be more local than it is now.

Chris Husbands

Schools are run by very clever people and by and large they cover vacancies, so what the data that comes out on vacancies tells us is simply the very small number of schools who are holding vacancies that they have not covered. That is quite different from whether you are covering vacancies with, as people were saying in the previous session, appropriately qualified, appropriately trained people.

Jenni French

Yes, especially given that the vacancies are collected in November, by which time schools will have done something and there will not be a classroom without a teacher in front of it.

Chair

Yes, because it is a question of live time versus historic, essentially, and that is something that you would want to see perhaps more up to date.

Jenni French

Yes, and also think about collecting the data in a different way. One way might be to look at job adverts, for example, which I know TeachVac already do, but if you monitored the number of times people are advertising for positions, that might give you a good idea whether the vacancies are being filled or just simply covered.

Chair

Yes, but measuring the problem that Chris has alluded to, filling a post with somebody who is not necessarily fully qualified for that post, a temporary measure, is something else we would want to get a handle on. How would we do that, do you think, Mark?

Mark Parrett

There are various measures out there at the moment. We have the vacancy levels, the number of positions filled temporarily. Another significant indicator referred to in the previous session is the number of lessons that are being taught by teachers without a post-A Level qualification in their subject. That reflects, in large part, the choices that schools are having to make, and the extent to which that has happened is increasing. Between 2010 and 2014 it has increased from 14% to 18% for EBacc subjects. To put that in context, for maths, 20% is being taught by individuals without a post-A Level qualification. That is 99,000 hours, so it is quite significant. What we recommended is that the Department should talk more with the profession about pressures they are facing locally and then use that intelligence, together with data, to come up with solutions.

Chris Husbands

I agree with Mark. It is worth putting a footnote on that, which is that the data on no higher than post-A Level qualification normally completely excludes people who have taken subject enhancement programmes after and much later. There is a sense in which some of that data might be self-defeating, so I think you need something that picks that up.

Chair

Joanna, I was going to ask you about the difference between simple data and qualitative data, because obviously with your Ofsted hat on, that is highly relevant. Do you think that is an important question and, if so, how do we set about getting that data?

Joanna Hall

Yes, I think it is. In the last few years, through our “Unseen children” report, Sir Michael’s speech last year in June 2015 and most recently in our annual report, Ofsted has continued to raise concerns about the granularity of regional and local data about supply and demand and also our research in three regions that we looked at last year. When we spoke to head teachers in those regions, there was a very clear message that they were struggling to recruit for particular subjects and the picture on the ground supported some of the findings in the National Audit Office report earlier this year.

Ian Mearns

The Committee has been told that there may be problems with under-reporting of vacancies due to increased competition for teachers. How would it be best to incentivise and reassure leaders to share accurate information? Go on, jump in.

Chris Husbands

It is a really difficult question. Whoever wrote it needs a lot of credit. I am not going to answer your question directly, Ian. I am sorry about that. It is in the interests, for all sorts of reasons, of head teachers to be able to say that all their vacancies are covered and all classes are being taught. Schools are in highly competitive local environments. The consequence of that is that it is probably unlikely that you could devise a sufficiently granular local reporting system that picks this up. The local variables are too great. You would probably have to rely on periodic reports from the National Audit Office to tell you that at a less granular level. I can’t realistically see how you would get to something that tells you what you want there.

Ian Mearns

But Mark has already told us that we know there are significant numbers.

Chris Husbands

Yes, he has.

Ian Mearns

20% of maths lessons are not being taught by subject specialists and a rise from 14% to 18% of EBacc subjects are not being taught by specialist teachers. I think the crucial question is: how do we best incentivise and reassure leaders to share accurate information? Without that accurate information, it is very difficult to do national planning.

Chris Husbands

Agreed. To reframe it slightly, we have an incredibly diversified school system, with the emergence of school autonomy and MATs. It is much more diverse than when the system was administered through 150 local authorities. There is a root question I think for the Department, which is that in a diversified, autonomous system where does data collection sit? Who holds that responsibility and how is that responsibility discharged in a way that for planning purposes of the system intelligence, we have confidence in the data? I think what comes out of the vacancy stuff is tough, technical and very complicated, and we do not have it at the moment. At the moment, nobody is holding that responsibility.

Ian Mearns

For instance, we already know some of the best data available on teacher vacancies is not collected by the Department.

Chris Husbands

Precisely.

Mark Parrett

The data is published, so one way is to put it out there in the public. The second way is to challenge it; I do not know the extent to which it is being challenged. The third aspect could be the definition of what is being reported and I do know there are very precise definitions about some of this that may affect the way it is being reported, but that is a guess.

Ian Mearns

A number of witnesses for the inquiry stated that vacancy data might be the wrong measure to use altogether when estimating and planning teacher supply. What other measures could be used, do you think?

Jenni French

I think if you look at the school workforce census and start looking at how many lessons are taught by non-specialists, that ought to give you an idea of the need for backfilling within the system. Then you need to start looking at teachers leaving the profession. I think that we do not know enough about that at the moment to know how many are leaving and how many, in turn, we need to train.

Ian Mearns

In your written evidence, you propose to use the A Level data to estimate the existing teaching workforce. Would you like to expand on that?

Jenni French

I think what I said was that you can use the A Level data as a proxy of how well a subject is taught within a school. For example, if a school is sending no pupils on to study physics at A Level, you could assume that physics has not been taught that well pre-A Level and wherever they go on to study on A Level, whether they are a—

Chris Husbands

The school workforce census tells you a huge amount. It gives you a snapshot. It doesn’t necessarily give you very good planning data. When I was doing some work for the Department on this a few years ago, one of the questions I asked myself was: how is this handled in other countries? The answer is that nobody does this terribly well. Almost all countries, at that stage, were citing the teacher supply model in England as being a pretty good way of doing it. What we know is that the teacher supply model is flawed in a number of ways, but at a system level it is not bad. It more or less did what it said on the tin. If you want to solve the problem at system level, look at other countries. They do one of two or three things. The first is they carry vacancies and assume that you are going to have non‑qualified people teaching in the system. More or less, that is the sort of direction we may be going in. The second is that there are systems—Ontario is a very good example—where they are very happy to oversupply the market. The consequence of that is that Canadian MPs are perfectly used to their mailbags being full of letters from constituents complaining that their children do not have jobs in teaching, having trained as a teacher. But if you are oversupplying the market, you have to carry that. The third is you go for a very clear central planning model where a very activist ministry does an awful lot of detailed planning, and that is what is done in Singapore and also in Shanghai. Ultimately, you end up with one of those three models: oversupply, carrying undersupply or very tight planning.

Ian Mearns

But given the levels of attrition that we have within the teaching profession anyway, oversupply would seem like a good thing to do, wouldn’t it?

Chris Husbands

Oversupply is a good thing, and I read the evidence you took from the Minister just before Christmas. The view of the Treasury and the view of the Department has always been that oversupply is a waste of public resource because there is a cost to public spending. We have always been pretty tight on that and we have also been very uncomfortable with having unemployed teachers. On attrition, there is something that you can build into supply models. This is a largely feminised profession. If you are recruiting significant numbers of 20-somethings into a feminised profession, you are going to be carrying quite a lot of maternity and attrition through that route. There are some things that are built into the demographics of teaching that you have to ride and that is why the teacher supply model ends up being more complicated than pure demographic planning.

Ian Mearns

Do you want to come in, Joanna?

Joanna Hall

I support absolutely all of that about the teacher supply model. Our evidence suggests that if you look at the location of a new accredited school and providers, the SCITT provision, the HEI provision, and understanding where there are those geographically-isolated cold spots across the country, that can then support the teacher supply across different partnerships and different teaching alliances. I think we have particularly picked up in the annual report that that has to be a focus in order that even if we went to a model where there was oversupply, actually understanding the impact of that in those rural areas that are quite isolated, the coastal areas we have reported on, there is another layer there of the picture that would help us maybe tackle some of the vacancies that we might be seeing.

Mark Parrett

We commented quite a lot on the teacher supply model in February, and I would add a couple of things from that. First, the model does contain a number of uncertain inputs, projections, and what we would normally expect with such modelling is that the Department would come up with high and low estimates, estimate the probability of different scenarios occurring and, from that, select its most likely. The model does do that and we presented the results of that in our report, but the Department does not assess the likelihood of those scenarios occurring. It does not choose. It believes that the high and low estimates are unlikely and chooses what it thinks is the most likely. We would like to see more assessment done of the outputs of that model. On oversupply, in terms of recruiting to training, the Department has allowed a couple of subjects to recruit over need in English and history. It does that on the assumption that schools will be able to use teachers flexibly once they become teachers.

Jenni French

On the problem with oversupply, that is great if you can recruit enough in the first place. There are a number of subjects where that is not an option and it has never happened.

Chris Husbands

The shortage of mathematic teachers and science teachers is global.

Chair

I was just about to make that point, that it is all very well talking generically about regional differences but there are significant differences in subjects and the supplies. Mark, do you have any detailed analysis of that or any explanations to help us along?

Mark Parrett

We did present a picture of the 2015-16 intake for teacher training. There was quite a significant range in what the teacher supply model said was required and what we ended up with as trainees, from design and technology, where they filled 41% of training places, up to history, which was 113%, the overfilling I mentioned. We have got to a point now where out of 17 subject groups, only three have managed to fill or exceed the requirement for attracting trainees to start training. That is almost all subjects were not able to recruit enough people to start training.

Chair

On Jenni’s point about a school not having much action, for example in physics A Level, because they simply have not done very well up to A Level—perhaps because of a shortage of teachers—is that something that Ofsted is thinking about looking at in a pretty forensic way across the piece and in individual schools?

Joanna Hall

Obviously, when we revise the leadership and management areas, one of the key questions we might ask particularly of leaders and leaders of teaching is most definitely that about teacher supply, how many subjects might be covered by temporary cover, and the longitudinal picture of that and how it affects the school. In terms of things that we may report on—it is not a requirement to report on that at the present time—that would be a key thread that we might look at under the new common inspection framework.

Chair

That would be quite an interesting area of focus, wouldn’t it?

Joanna Hall

Yes, absolutely. Yes, it would.

Lucy Frazer

Can I talk about that, because I find it very interesting? If the provision is inadequate, you would then say, “Well, it is inadequate but we are going to waive what we would usually say because it is not your fault.”?

Joanna Hall

I think it is about the balance of all the evidence on any inspection in terms of how well do leaders understand any issues they face in their school and their provision, how best are they managing to support the children and the students in that school in order that their progress does not stall. That would have to be balanced, like many other factors on inspection, to look at the impacts on the young people and their progress. Certainly, it is a key part of our judgment for leadership.

Lucy Frazer

I think what you are saying is if they do not understand it, they will be marked down rather than they will be marked up because it was not their fault. Have I understood that correctly?

Joanna Hall

How you manage your own teacher workforce and how you deal with those particular issues, one would hope all leaders and governors have a clear picture of the impacts of whatever they are facing in their school. But certainly in terms of good leadership and good, high quality teaching, one would expect that we can see evidence of that in how those kinds of subjects are taught, even if it be by a non‑specialist.

Chair

You would be looking at a school and thinking what kind of plan do you have to resolve this particular problem and so on?

Joanna Hall

Absolutely and talking to governors about that as well as leaders, and I think that is a great opportunity. We reported in our annual report that, among the partnership work, one of the big things we have seen in initial teacher education over the last two years is that where those partnerships work well—across teaching schools and the work with HEIs—there is a good shared understanding of how to support those schools where they are sharing subject expertise. That is something that, in our two-part model for inspection, we do clearly look at under that leadership strand for ITE but also in school inspections as well

Stephen Timms

I would like to pick up the issue of more geographically granular data that you have already raised in your evidence. The Department does acknowledge, in its evidence to us, that there is an issue here. They say there are limitations to the value of national statistics since they have the potential to mask regional and local variation. What kind of geographical data do you think ought to be collected?

Jenni French

I think in the first instance it would be helpful to know where people are doing their teacher training. Although the Government do collect data on the locations of teacher training, they do not release that publicly. We do not even know, for example, where people are doing their teacher training, which makes it quite interesting if you are looking externally at what is going on. We could do some work, as external organisations, perhaps matching up vacancies with that data, but it is not being released.

Stephen Timms

The data is there. It is just a question of publishing it.

Jenni French

Yes, it must be there. It is just not published.

Stephen Timms

Do you think that ought to be published by region, by local authority?

Jenni French

To be honest, it could be. There are not that many places where people study teaching training, so it could be published in—

Chris Husbands

Although there is a growing number.

Jenni French

Of course, of schools, but even so, it could be published by region or local authority. I think anything would be—

Chris Husbands

Stephen, once upon a time I was on the board at the TDA, which was the NDPB responsible for teacher education and development. We did an awful lot of work on regional teacher labour markets and subregional structures. The conclusion that the agency came to is that the cold spots are subregional rather than regional and that regional mapping did not do very much because, by and large, you end up with artefacts of somebody’s mapping system and it does not really influence where people move. I think we can connect up the school workforce census. Longitudinally, we know where people train, we know where they teach. We can do some work on attrition rates and longevity rates by provider and by route. That data is there. It tends to be rather detailed and complex data. It is a bit indigestible, but for planning purposes, it does exist. What I would have wanted to see was the Department preserve the teacher supply model, because the National Audit Office do not have many strengths, and supplement it with some of the other datasets that we have now. That is going to be worth quite a lot of sufficiently granular data to tell us where we need to be going.

Stephen Timms

Is it generally the case that you think the Government have the data and it is simply a question of presenting it, or do you think there is some additional data that needs to be collected?

Chris Husbands

I think there are sufficient datasets out there. Not all of them are collected by Government vacancy datasets. What the Government have rather lost over the last five years is the capacity to analyse and make use of those datasets with sufficient granularity.

Mark Parrett

One of the things that the Department has told us is that it is working on linking the datasets between training and then out into the profession, so you can track individuals or cohorts and see how they are doing. We looked at the training being provided by region and we know there are links between where people live and where they train and then when they work as teachers. You have quite a big disparity between where training is taking place around the country, which is lowest in the east of England and highest in the north-west. We also know, and presented in our report, that not all types of training are offered on an equal basis across the country, so if you want to pursue a salaried teacher training route, for example, a lot of that is more available in London and the south-east. But Chris is right, there is a lot of data out there. A lot of it is already published down to school level. It certainly is in the census and there are other sources of data. The other thing that the Department told us it is concentrating on is engaging at a more direct level with schools to understand the challenges they are facing and how best to measure and deal with these issues. The caution we have given there is that a lot of the conversation has happened with schools involved in School Direct, which is less than half of the school population and is not representative, necessarily, of primary schools or schools in rural areas or in areas of high deprivation. That would be the caution around that sort of engagement. All of that is good.

Stephen Timms

You said the data that is needed is already being gathered. Is it being gathered by the Department from individual schools? Is it being collected by local authorities and handed on to the Government?

Mark Parrett

The census is gathered from schools directly and from local authorities that may submit on behalf of maintained schools. That was introduced in 2010 and combined a whole series of datasets. It is a statutory responsibility to fill it in; 99.7% of schools completed it last year. It is a national statistic. It has particular quality controls around it. There is an awful lot of data in there and it is published, so that is why.

Stephen Timms

One of the issues that has been raised is that local authorities perhaps do not fully know what is happening in academies in their area. Is that an issue? Is there any data that is gathered for maintained schools that is not being gathered for academies or is that not an issue?

Mark Parrett

It is not an issue because academies will submit it directly and the response rate is as high for them as it is for local authorities.

Chris Husbands

I wanted to say something that I guarantee is unhelpful, but it is a non-data point. It is worth remembering that the way the teacher labour market operates in England is almost unique internationally. We very unusually now, and academies explicitly, employ teachers at the level of the school, not at the level of the school board or school district, and teachers decide whether or not they want to apply for jobs, so it is a free market on both sides. If you look at most public education systems—and the situation that we have got into there is a consequence of policy decisions taken since the 1981 Education Act, the 1986 Act and beyond—in most jurisdictions somebody is intervening to move things around. In Ontario school principals do not even know whether their leadership team this year will be the leadership team they have next year because the school board might shuffle people around to places where they are needed. In France, teachers will be moved around regionally at least. I do not think that is a helpful thing to say because I do not think it is desirable for us to row back on school autonomy. I certainly think the way the teacher labour market operates works sufficiently well. But it is worth remembering that this is a global problem. How do you get enough teachers and how do you get the right teachers in the right places? We are trying to do it with a set of tools that almost nobody else is deploying globally and that puts us into an odd position.

Catherine McKinnell

We have talked about the regional variations but also we know there are significant variations in attrition rates in respect of subject areas. If the Government had better data on that, how would you see that data being used to good effect? Let us go back a step. Do you think we should be selecting data on attrition rates by subject and how would that information be used?

Jenni French

I think that we should be collecting that information. In the teacher supply model the subjects are grouped rather than individualised and I think that hides quite a lot of differentiation between subjects, so it is not predicting accurately in different individual subjects how the attrition might fall. But also it could be done, as Mark mentioned earlier, if we had the linked teacher training profile with the school workforce census. We would start to be able to see if there are some subjects with higher attrition rates and then we might be able to direct early intervention. We talked about those first three years but we could direct it specifically at those subjects with the higher attrition rates and help to look at keeping some more of those teachers in the profession.

Catherine McKinnell

It has been done to some extent in the past, particularly where there has been a glaring shortage in the sciences, for example. It would be interesting to hear how more could be done.

Joanna Hall

That would be a useful piece of information in order that where schools are struggling, whatever partnership they may be working in, whether SCITT or an HEI, to understand what that picture is by subjects. We reported last year that we continue to be concerned about the level of understanding, particularly at subregional level, what is the big picture in region and subregion by subject attrition and how can we best support and use that data then, particularly for schools that are in challenging circumstances. They report to us that it is very difficult to attract and then retain the very good quality teachers that they need in order to improve their schools. I think that kind of information would help us to tackle some of those issues we hear as well.

Mark Parrett

We are particularly focused on initial teacher training and that aspect, but this comes on to something around evaluation of what works as well, which we have not discussed yet. We said to the Department that you need to be evaluating the various initiatives that have been put in place to attract new teachers but also to set those against the relative cost effectiveness of measures to retain teachers. It has not been done, so that sort of information would be useful in the evaluation.

Catherine McKinnell

Are there other areas that we are not collecting data on that you think would be helpful? You have mentioned the route into teaching and the attrition rate and cost effectiveness. Other potential factors could be where bursaries are being provided, the social background of teachers and what the attrition rates are. Is there anything else that you think of that we should be collecting that would give us a better picture?

Mark Parrett

On bursaries, the Department did some qualitative analysis in 2012-13 and some quantitative analysis in 2013-14 that showed some link between bursaries and people wanting to apply to train as teachers and that did show some positive relationship. However, they have not extended that to look at whether people went on to apply, completed the training and then went on into teaching and how well they did, which comes back to these data-linking points. The Department has a general understanding of the impact of bursaries but needs to do a lot more. By the end of 2016-17 they will have spent nearly £1 billion on bursaries so we would expect a greater degree of evaluation around whether that is working.

Chris Husbands

My recollection is the TDA did look at the cost effectiveness of bursaries about 10 years ago and concluded that it was very easy to overspend.

Jenni French

I think it is also worth thinking about if you are applying, particularly in physics, to do a School Direct course you have a much lower chance of success than if you apply to an HEI. It would be interesting to know, by looking at retention, is that because the schools do a better job of picking the people more likely to be successful or are they turning away lots of people who could go on to be good teachers? I think if we knew a bit more about how long people stayed in the classroom it would help us hone the interview techniques and think about the type of people who could go on to become teachers.

Joanna Hall

You asked about any other kinds of data. Having heard the first panel and this notion of a continuum—as you join the profession and progress through your NQT year and receive, one would hope, high quality CPD—how do we track the profile of those students who become teachers and move on to middle and senior leadership positions? You asked earlier about the new College of Teaching. Our evidence would suggest that understanding how we get those leaders through into our schools who can work with our young people to drive improvement would be beneficial and that is something that would be very helpful as a longitudinal picture.

Catherine McKinnell

A lot of what we are talking about here is measuring quantity, but the other key factor is quality. What do you think would be the best way to monitor the quality of the teaching workforce?

Joanna Hall

I will start with that one. The key message from us at the moment in terms of initial teacher training inspection outcomes, as you will have seen in the NAO report, is well over 95% good or better providing strong initial teacher education. I think our change in June 2014 to a two-stage inspection model particularly wanted to recognise that end stage point for trainee teachers at the end of their training and then the progression into the NQT year is our second stage in the autumn term, working with their employing schools to see on the ground how ready they are to join the profession. In the quality coming through we have seen, particularly over the last five years, a real marked improvement in that partnership working between the employing school and also the transition from ITE into the NQT year and beyond. Colleagues this morning said there is still more to do and I think that is absolutely true when we look at CPD on our school inspection: is it piecemeal; is it well structured; does it develop people’s subject knowledge to the level we would want it to? We have some positive quality indicators from inspections, definitely.

Chris Husbands

I agree with everything Joanna has said. Let us be clear that of the in-school factors that we know about, quality of teaching is the biggest driver of pupil attainment. Against that, it is worth getting a little bit real about the capacity of the initial teacher training to impact on that in the short term. A very quick bit of mental arithmetic: there are 500,000 teachers in the profession and about 30,000 trained each year. Let us imagine that we could identify the weakest 5,000 of those and not recruit them and let us also imagine we could somehow recruit 5,000 a year who are definitely going to be better than the remaining 25,000. These are incredibly heroic assumptions. If we did that it would take about 20 years to have more than one standard deviation impact on pupil attainment because the impact of the flow on the stock is relatively small. We should have high quality teacher education but we should be a bit realistic about its capacity to influence the system in the short term. CPD is a much bigger lever, and certainly in schools facing challenging circumstances, CPD is an absolutely massive key to bringing about change.

Mark Parrett

Another aspect I will bring out is that the approach that has been taken is to increase schools’ inter-training and school-led provision so that schools are more involved in selection and training and that has expanded a lot, as we said in our report. There is a risk we brought out in the report that the Department needs to assess, that if you have schools involved in School Direct are recruiting individuals to training, they will cherry pick the best ones and keep them. Given you had 57% of schools at the time we reported not participating in School Direct and only 29% of primary schools in rural areas, and there are other statistics around that, there is a risk that any quality benefits from this approach are not evenly spread. We have highlighted that as a risk that needs to be assessed and addressed.

Chair

I want to probe the position about the subject. Sir John in the first panel was taking about the importance of the subject and you all mentioned—Jenni in particular and Joanna as well—the importance of understanding the subject in terms of data, both quantitatively and qualitatively. I wonder how we might explore that in more detail and come up with something that is going to be helpful. Mark, could you talk us around that? It is important to get a clearer understanding of each individual subject in terms of supply and of those people expressing interest in entering the training arena. That is just as important as the number of people in training because then you get a clearer idea of flow.

Mark Parrett

I go back to what I said earlier about the huge variation in the extent to which training places are being filled by subject and bursaries are being differentiated by subject. There are higher bursaries available for physics, for example. Bursaries in other subjects where there is less of a need have reduced a bit, so there is some differentiation there. We need to know whether they are working or not and we need to see the evidence on that.

Jenni French

Particularly focusing on physics again, there are about 3,000 graduates a year and we need to recruit about 1,000 teachers a year. It is not going to happen if we go for appointing only physics graduates. I think the subject knowledge enhancement has been key for pre-initial teacher training, recruiting people from a wider background, and unfortunately that is drifting away a bit with the changing system and that is not going to help until we get it sorted out. We are never going to get anywhere close to the targets with some of the more difficult to recruit subjects.

Chris Husbands

I agree with that. We may come on and talk about School Direct and the impact on higher education, but one of the things that you need if you want good quality subject training is a critical mass of scale that enables you to recruit a cohort of physics teachers, a cohort of mathematics teachers, and then you can put some high quality subject training around them. Once you fracture this across an entire schools system you abuse that.

Chair

Joanna, would you concur with that?

Joanna Hall

I do, and I think the point that was made earlier about learning the craft to become a teacher educator is a very different path from just in the classroom as a practitioner. Subject expertise has to be a really important focus to take forward because it is a very different set of skills and knowledge.

Chair

You would like to see this Committee think very carefully about how we can enhance the focus on the subject?

Joanna Hall

I think it would be of great benefit and it would also reflect some of the concerns we have raised about the lack of real integrity in some of the subject knowledge work. I think that would be really helpful.

Chair

Do you all agree with that? Yes.

Ian Mearns

It begs the question in my mind—it might be a barmy question—are there enough places in universities for physics undergraduates? Are the university courses fully subscribed or could we create more places and help solve the problem that way?

Chris Husbands

I am absolutely delighted to say for your sake that that is a BIS issue not a DfE issue. But by and large we have moved away from what we used to call manpower planning in relation to higher education development and we have gone for demand-led growth and that has happened. It costs a lot of money to train a physics undergraduate. You need a lot of kit to do it appropriately. There are stats on the number of physics departments there are in UK universities now and there were 25 years ago and I did not do my homework on that because I did not think I would be asked it today. But it is a different sort of question. How do we make sure that at graduate level we are producing enough scientists and technicians? Jenni’s figure is dead on. It is 3,000 physics graduates a year. The school system could take 1,000 of them and you would have just about enough to meet the supply need. You are not going to get 35% of physics graduates coming into teaching, so that is a big set of issues there for national strategy in relation to science and technology.

Lucy Frazer

How is the data for how many teachers we need affected by the independent sector? When you were talking earlier about the maintained sector, do you chop out those from national teacher training who go into the independent sector to count your figures?

Mark Parrett

The teacher supply model is forecasting the requirement for the publicly-funded sector.

Lucy Frazer

I presume people who go and teach in independent schools go through similar training.

Mark Parrett

Yes.

Lucy Frazer

How do you know which sector they are going to go into?

Mark Parrett

First, there is some movement from having completed training. Some go on to the independent sector. Some also come back from the independent sector to the maintained sector. There is not good quality data available on the extent to which that is happening.

Chris Husbands

I think there is a reasonable hypothesis that in the context of the teacher supply equation it is a relatively constant number year by year.

Lucy Frazer

Is that taken into account?

Chris Husbands

You can account for it and more or less it looks after itself.

Jenni French

Although I think there may be some subject variation and I am not sure that is—

Chris Husbands

Especially around physics.

Jenni French

I think particularly around physics where more teachers might feel that in the independent sector they can teach just their subject so be attracted to the independent sector, so you may find it is disproportionate.

Mark Parrett

Although the sort of numbers we are talking about are that of the numbers recruited to a teaching place in England within six months, about 7% went to the independent sector in 2013-14.

Lucy Frazer

When you are working out how many we need, chop off 7%?

Chris Husbands

That is roughly consistent with the number of pupils in independent schools, which is around 7% or 8%.

Mark Parrett

The calculation is an assumption around other numbers entering the workforce. The Department assumes that a number will come back from the independent sector so they are not forecasting for the independent sector.

Joanna Hall

In Sir Michael’s commentary in February this year, the teacher brain drain commentary, Sir Michael called for a better understanding of those who do exit and go to work in independent schools abroad, so that we would have a tighter dataset of what the profile of those qualified teachers looks like—it could be by subject—and how many of those do come back and work in the UK. I think it was with caution that we need more analysis of that, and that was very much a key part of Sir Michael’s commentary this year.

Chris Husbands

That is not a constant. The phenomenal growth in international Anglophone independent schools over the last 15 years has created a pretty large globally footloose workforce.

Lucy Frazer

The White Paper suggests, I think quite rightly, that we should have a national database of vacancies. If that goes ahead, will it give the Government any more of a handle on vacancies than the data? It is real adverts that are going in in real time for real places. Would that help inform the data they are putting together or is it too unstructured?

Jenni French

I think it could do. There are people already doing similar things. TeachVac is already doing a database so that is worth looking at first. But also it would help if there was a number associated to each post so we could track a job: do we have a constant churn within a school or is it the same post not being filled that we have to re-advertise? At the moment, the vacancy for a science teacher in a large school does not tell you, very easily, whether or not it is the same teacher.

Chris Husbands

I agree with that. There are ways in which the teacher labour market is structured. Beyond initial training, the market for classroom primary teachers is local. The market for secondary heads of department and primary deputies is probably subregional. Only the secondary head market is probably genuinely national. There are variations around that, so there is a question about if you can run a Government-led free vacancy service but you cannot necessarily guarantee that we would require all head teachers to use it.

Lucy Frazer

You mentioned TeachVac and the National Association of School-Based Teacher Trainers gave us evidence and praised the work they have done in relation to the data vacuum, as they say. Do you think that TeachVac should be centrally funded to deliver data or do you think there should be a centrally-funded independent body for that role?

Chris Husbands

I don’t know the answer to that. I do know that for the last 100 years the economics of running The Times Educational Supplement has depended on its job advertising, so if we are going to have a public intervention in the market we do need to understand why we are doing it, who we are funding to do it and on what basis. I think there are some quite tough policy issues there for a country that has largely become pro-market in its public policy.

Mark Parrett

We do not have a view on who and how it should be funded. Particularly with regard to initial teacher education, we believe that the information available in the market should be improved. Just about everyone told us that the routes into teaching are confusing and applicants need much better information on where they can train, how much it is going to cost and what sort of quality of provider they might go for. Equally, on the other side, on the supply side, it is much longer-term information on what the demand might be and from me as a provider over time rather than an allocation model that changes each year and gives no certainty. It is not just advertising of teaching vacancies in schools, which one would think is a good thing, but also the broader information that is available to enable the market to operate.

Ian Mearns

The problem is when we have significant specialist teacher shortages in certain subjects is not the whole idea of spending huge amounts of money to advertise teacher vacancies just inherently wasteful? It has to be, because I am aware of schools who have had to re-advertise the same position two or three times and that becomes quite expensive on a unit cost basis. There has to be a more rational way of doing this, surely. I do not know what the editorial independence of The Times is, for instance.

Chair

I think you have made your point.

Chris Husbands

Yes, is the answer.

Chair

Can we move on to the final question?

Marion Fellows

Various bits of data are already available from different public, private and not-for-profit sources. Can the answer be to create a single portal for sharing all the data collected across all these areas on teacher supply?

Mark Parrett

As I said, the census did that to a certain extent in 2010, when it was brought in, in bringing together sets of data. One of the drivers behind that was also to reduce the burden on schools in having to produce different sets of data at different times of the year. That is available online. The results of the census are available online, as are various other bits of information. Bringing them all together in one way where you can access them in a consistent way in theory would be a good thing, yes.

Michelle Donelan

Does that mean that all the data collectors should work together collecting that data?

Mark Parrett

We have said the Department should work with the sector on this to improve the information available.

Jenni French

That needs to be linked. At the moment there are datasets but it is impossible to draw them together, whereas they could be drawn together and published in their entirety so that people are able to analyse them and not subsections of the data.

Chris Husbands

The data linkage is complex but can be done. In my previous job, we ran a series of big projects on administrative data linkage. There are techniques. It takes resource, it can be done, but the benefits for policymaking are considerable.

Joanna Hall

From Ofsted’s perspective, it is the accuracy of the data. No matter who pulls it together and how they do so, it is the level of accuracy within that data that informs us about what is happening across the teacher workforce but also clearly then in terms of our inspection frameworks and provides a piece of evidence that we can look at and track and report on. It does not matter who does it and how it is done but the real accuracy is important.

Mark Parrett

The Department has said a number of times to us in the Public Accounts Committee how that data linking is going to help with some of these issues. You could, for example, look at the results of that work when it is completed, which I think will be quite beneficial.

Chair

We literally got to the bell and we have asked all our questions and had some absolutely excellent answers and some excellent panellists to give those answers, so thank you very much indeed. The whole morning has been incredibly instructive and I think we have some interesting ideas, not least this issue about seeing what is going on in schools in responding to the challenges of teacher recruitment and supply of teachers. It is all excellent stuff, as well as this focus on subjects as opposed to just simply generic training. Thank you all very much indeed. Oral evidence: Supply of teachers, HC 199 21