Who is teaching the teachers?

9th January 2019 by Timo Hannay [link]

Much has been written recently about school funding and teacher retention¹. This post looks at an under-appreciated topic that lies at the intersection between these two perennial media favourites: teacher professional development. After all, what is the point of hiring teachers if you don't motivate and develop them? At best they will be less effective, at worst they will leave and create yet another hard-to-fill vacancy².

But how much do schools spend on developing their staff, have recent funding pressures affected this, and what if any evidence is there for an impact on school effectiveness? Working with the Teacher Development Trust³, we have used school finance data from the Department for Education (DfE) to try answer these and other questions. Our main findings are:

• In 2017, state primary schools in England spent an average of 0.66% of their budgets on staff development and training, equivalent to £685 per teacher. State secondary schools in England spent 0.40%, or £378 per teacher. Only 19% of primary schools and 6% of secondary schools spent at least 1% of their budgets on staff development.
• Although staff development spending has risen in nominal terms since 2012, in 2017 it declined by 7.5% in primary schools and 12% in secondary schools. This reduction is broad based: a majority of state schools in England spent less on staff development in 2017 than in 2016.
• Spending on learning resources, another important but somewhat discretionary line, has also been falling, both in absolute terms and as a proportion of school budgets. This suggests that recent changes in school spending patterns may reflect general funding pressures rather than specific reprioritisations.
• Schools with higher proportions of deprived pupils tend to spend more on staff development. This may be at least partly because they receive higher levels of funding, though the fact that they also tend to have staff who are younger and less likely to be qualified could also be having an effect.
• Because of the bias in staff development spending towards more deprived schools it is difficult to determine the effects of staff development in isolation from other factors. However, we find some evidence that higher levels of staff development spending correlate with lower staff absence, at least in secondary schools.
• In addition to this blog post, we have collaborated with the Teacher Development Trust to make available an interactive map showing geographical variations in spending and a benchmarking tool that allows individual schools to compare their staff development spend with those of other schools locally and nationally.

For a more detailed analysis, read on.

Spending suspended?

In this section we examine recent trends in school spending, with a particular focus on staff development⁴, but we will also look at certain other types of spending as well as teacher and pupil numbers. This analysis relies mainly on DfE data covering school financials, staff and pupils. These sources – particularly the financial data – have various complexities and shortcomings that we have taken steps to address; for details, see Footnote 5.

As shown in Figure 1 total spending by primary and secondary schools rose between 2012 and 2017, but per-pupil spending showed a more modest rise and has been flat since 2015. (These and other figures presented here are in nominal terms, so make no adjustments for inflation.)

From 2012 to 2017, pupil numbers grew in primary schools (+15%) while remaining virtually constant in secondary schools. Over the same period, staff and teacher numbers grew in primary schools by +21% and +14%, respectively, leaving the pupil:teacher ratio very slightly higher at 20.8. Staff and teacher numbers in secondary schools fell very slightly (both -1%) and the pupil:teacher ratio stayed flat at 15.6, albeit with a slight dip to 2014 followed by a rise since then.

Spending by primary schools on staff development and training rose strongly (+79%) between 2012 and 2016, but fell back (-7.5%) in 2017. As a proportion of total primary school spend it fell to 0.66%, slightly above the level in 2013.

Staff development spending by secondary schools started much lower and grew less strongly during 2012-16 (+60%), but still declined in 2017 (-12%).

This recent decline in staff development spending could be a symptom of general pressure on school budgets, but might instead reflect a deprioritisation by schools. To help distinguish between these possibilities, we also analysed spending on learning resources, which, like staff development, represents a potentially important form of investment, but a discretionary one that is therefore vulnerable to cuts when budgets are tight. This form of spending has also declined in absolute terms, as a percentage of total school spending and on a per-pupil basis. Whilst this isn't conclusive, it suggests that general budgetary pressures are a more likely cause of cuts in staff development than specific deprioritisation.

Use the menu below to explore these and other metrics; click on the legend to show or hide different school types.

Never mind the quantity, feel the width

Whilst it is clear that staff development makes up a small proportion (well under 1%) of total school spend, and that total staff development spending declined between 2016 and 2017, such aggregate figures can hide a lot of detail. For example, the reduction in overall spending might be the result of a few previously high-spending schools suddenly deciding to cut back. Figure 2 shows that this is not the case. On the contrary, decreases in spending appear to have been broad based.

The green columns in Figure 2 show the proportions of schools that increased their staff development spend compared to the year before while the red columns indicate the proportions reducing it. The majority of primary schools showed increased staff development spending between 2013 and 2016; it was only in 2017 that the balance tipped the other way, with a majority spending less than in the year before. Among secondary schools there is a similar pattern, though the spending reductions began earlier: a majority of secondary schools cut back in 2015, and in 2016 the proportions showing increases and decreases were virtually the same. By 2017 a clear majority showed lower spending than the year before.

To put it another way, the median change in staff development spend between 2016 and 2017 was -6.9% for primary schools and -12.1% for secondary schools. (As a reminder, all of these figures are in nominal terms and make no adjustments for inflation, so even constant spending represents a cut in real terms.)

Use the menu below to switch between primary and secondary schools. You can also hover over each column to see corresponding values.

This is all very well, but whether staff development spending is rising or falling, a more important question is what its absolute level ought to be. This is largely a matter of opinion, so we won't try answer it directly here. Instead we'll examine the range of current spending levels and allow you to draw your own conclusions. As we saw above in Figure 1, during 2017 staff development spend per teacher amounted to about £685 per primary school teacher and £378 per secondary school teacher (respectively, about 0.66% and 0.40% of total school spending). But these average values conceal a lot of variation, as can be seen below in Figure 3, which shows the distribution of staff development spend as a proportion of total school spend.

In 2017, very few schools – primary or secondary – spent more than 1% of their budgets on staff development. 4.5% of primary schools and 10.5% of secondary schools reported spending nothing at all. The overall distributions haven't changed much from year to year, though (as we already saw in Figure 1), staff development spend as a proportion of the total has increased a bit since 2012. (Use the menu above Figure 3 to see results for other years. You can also click on the figure legend to see either primary or secondary schools alone, or both together.)

For the sake of argument, if we believe that schools ought to spend around 0.5% of their budgets on staff development then in 2017 63% of primary schools, but only 25% of secondary schools, would have met that threshold (representing aggregate shortfalls of £19.7m and £36.7m, respectively). If we were to say that it ought to be 1% of total spend then only 19% of primary schools and a mere 6% of secondary schools would have made the grade (implying aggregate shortfalls of £97.4m and £121.8m, respectively).

To look at it another way, 92,577 primary school teachers (40% of the total) work in schools that spend less than £500 annually per teacher on staff development. The corresponding number for secondary school teachers is 158,646 (78% of the total).

Development and deprivation

Clearly staff development spending varies widely by school, but what, if anything, can we say about the types of schools that are likely to spend more or less than the average? One of the most significant factors is poverty. This is shown in Figure 4, which groups schools into five equally sized 'deprivation quintiles' based on the proportions of pupils who are eligible for free school meals (FSM). Note that in this case we have aggregated data from across all years (2012-2017 inclusive) because this gives larger sample sizes, but the patterns don't change much from year to year.

At both primary and secondary schools, overall spending per teacher is higher at schools with greater levels of deprivation – unsurprising since these schools usually receive more funding. The same schools also show higher average levels of staff development spending.

It's tempting to speculate that higher total funding causes schools to spend more money on staff development, but this isn't necessarily the case. For example, schools with greater deprivation also have slightly lower proportions of qualified teachers and considerably lower proportions of older teachers, so they might feel the need to make up for this relative lack of experience by spending more on staff development. Note that schools with higher deprivation also tend to show lower Ofsted ratings and higher staff absence – something we will return to below.

The effect on effectiveness

What if any impact does staff development spending have on school effectiveness? This is a tricky question to answer because, as we have seen, such spending tends to be higher at schools with larger proportions of deprived pupils, and these same schools also tend to perform relatively poorly on metrics such as staff absence. We can see this in Figure 5 by comparing the 'High staff development' schools (ie, those in the top quintile for staff development spend) in the left-hand group with the 'Low staff development' schools (bottom quintile) in the centre group. (We will come to the right-hand group soon.) Of course, staff development spending is much higher in the 'High staff development' group than in the 'Low staff development' group – because that's how these schools were selected. But these groups also differ greatly in the proportions of FSM-eligible pupils, so comparing them directly could be misleading.

For this reason, we also created a group of 'Similar low staff development' schools (on the the right-hand side of Figure 5). These are a subset of the 'Low staff development' group selected to match as closely as possible the proportions of FSM pupils seen in the 'High staff development' group. This is a more sensible group against which to compare the 'High staff development' schools. Doing so reveals some positive effects, at least among secondary schools: average Ofsted ratings are slightly higher and staff absence is substantially lower. Such results shouldn't be considered conclusive, but they hint at potential benefits and would warrant further research.

Poorly educated?

In management circles, staff training and development has long been considered an important means to attract, retain and motivate talented people, as well as an obvious way to improve the effectiveness of individuals and organisations. It therefore seems ironic that the education system spends a such a relatively small – and currently declining – proportion of its budgets on coaching its own staff. If there was one sector you might expect to understand the value of training, this would surely be it, but apparently not.

To see national trends in staff development spending, and to compare the spending of individual schools with those of other schools locally and nationally, use the interactive map and benchmarking tool provided on the TDT website.

As ever, we welcome your feedback: [email protected]

Footnotes:

1. For the DfE's own analysis of teacher recruitment and retention, see here. SchoolDash also continues to track secondary-school teacher recruitment activity in our Jobs section.
2. See our recent analysis of rising secondary school recruitment activity.
3. The Teacher Development Trust is a charity that promotes continuing professional development of teachers. SchoolDash also believes that this is an important area but takes no formal position on how much there should be or how it should be conducted. Our job here is to crunch the numbers as objectively and usefully as possible in order to allow you to make up your own mind.
4. In the school financial data released by the DfE this is usually referred to as 'Staff development and training' or something similar.
5. School financial data is published separately for local authority-maintained schools and academies (which receive funding direct from central government). The analysis presented here combines both data sets, allowing for various differences in data formats. It also includes 'part returns' (ie, financials for recently established schools than cover less than a full academic year). For academies that are members of a multi-academy trust, we have included any central expenditure (which usually amounts to only a few percent of the total) and allocated this equally to each school in the trust. Some financial figures are suppressed by the DfE, especially where they might compromise privacy (eg, salary costs relating to a single member of staff). Moreover, this policy seems to have varied over time such that data suppression in some years is much more aggressive than in others. (In particular, a very large proportion of primary school teacher costs were suppressed in 2014.) In order to enable reasonable year-on-year comparisons, we have therefore estimated some suppressed values by interpolating from adjacent years where possible. Tests on data where we have values for all years suggests that this is a reasonably accurate approach: 65-70% of values are within ±30% of the true value, and the mean error for each year is very close to 0%. In any case, this approach is almost certainly much more accurate than simply assuming zero values for suppressed data. Note that even after applying this method some suppressed values remain. As a result, total levels of spend across all schools are likely to be underestimates, but we believe any discrepancies to be small and they should not vary much by year. Finally, some DfE data sources give numbers rounded to the nearest £1,000 while others provide more accurate figures. This should make little difference for aggregate analyses, but it may sometimes be significant when looking at individual schools.