Carl Cullinane on the background to today’s grammar schools research.
The months since Theresa May announced her plans to scrap the 1998 legislation banning the establishment of new grammar schools have seen a renewed and impassioned debate on academic selection in England. Sutton Trust research has been at the heart of this debate, both in the media and in parliament, and this week we have published our new Gaps in Grammar research brief, updating the evidence on grammar schools in England, and looking at the question from some new angles.
The government has indicated that new grammar schools are at the centre of their drive for more ‘good school places’ and greater social mobility. However, whether an expansion of selection is likely to lead to either of those goals is more questionable, given the evidence. Just this week, the OECD launched their latest PISA report comparing 15 year old pupils across 75 countries. Ministers have argued that the PISA results show evidence in favour of selection, yet the PISA 2015 England report itself shows that not only is academic selection not associated with higher performance by disadvantaged pupils, the opposite may be true. Selective systems are also shown to have bigger gaps between rich and poor pupils and thus greater inequality. The overall report concludes that all types of pupils benefit from less selection, not more.
In England, the question of whether grammars facilitate social mobility is threefold: how many disadvantaged pupils get into grammars, how do they fare when they do get in, and what effect does this have on non-selective schools.
Studies consistently show that grammar schools are not taking their fair share of disadvantaged pupils. Just 2.5% of grammar school pupils are eligible for Free School Meals (FSM), compared to around 14% in all secondary schools, and close to 20% in grammar school catchment areas. The government has suggested they wish to expand the focus from disadvantaged children, to include ‘just about managing’ families, who, despite largely being in work, find themselves squeezed by increased housing costs, bills and stagnating wages.
Grammar schools are framed as particularly helpful for the children of such groups. However our research indicates that the ‘just about managing’ also struggle to get into grammar schools, and are, in fact, not much more likely to attend grammar schools than the poorest families. Just managing families tend to live in neighbourhoods in the bottom two categories in figure 1 below, which, outside London, shows a steep gradient in access to grammars across the socioeconomic spectrum.
Figure 1. Proportion of Year 7 pupils in grammar schools by deprivation quintile
While the King Edward VI schools in Birmingham have shown it’s possible to admit high proportions of disadvantaged pupils (up to 20%) and Kent County Council has announced a willingness to move in that direction, other grammars have a long way to go in improving their intake of pupils from less advantaged backgrounds.
The likelihood of attending grammars also varies substantially by ethnic background. Disadvantaged white British children enter grammar school at the lowest rate of any major ethnic group. Disadvantaged Indian pupils are four times more likely to attend a grammar than disadvantaged white British pupils, and disadvantaged Chinese pupils fifteen times more. While there have been modest increases in the rate of grammar entry for disadvantaged black children and white non-British over the past five years, the rate of white British entry has not improved.
Figure 2. Proportion of FSM-eligible pupils entering grammars by selected ethnic groups 2012-2016
Parents and policymakers alike tend to be attracted to grammar schools because of their unquestionably high results. However, when you look more closely at the raw scores, this becomes significantly less clear. A variety of studies have shown that much of the advantage of grammar schools is explainable by the fact they admit pupils who are high achieving in the first place. This is true both for disadvantaged pupils and pupils as a whole. Expanding selection therefore has no guarantee that new grammar pupils will start performing better than they otherwise would. Our report also shows that high attaining pupils do just as well at top comprehensives as they do at grammars, with figure 3 showing 97% of high attainers at top comprehensives achieving five good GCSEs, compared to 98% in grammars.
Figure 3. Proportion of high attainers achieving 5 A*-C GCSEs (including English and Maths)
Finally, there is the vexed issue of the effect that grammars have on surrounding schools. The research is mixed on this issue, with a widely cited Sutton Trust report in 2008 indicating that, in the country as a whole, there was no evidence non-selective schools were negatively affected. However, recent indications are that, within the UK, areas with the highest concentrations of grammar schools have lower attainment in surrounding schools. Furthermore, this is particularly true for disadvantaged pupils, suggesting grammars may give with one hand and take away with the other.
It is therefore questionable whether increased selection would lead to more good school places overall, rather than the concentration of previously high attaining students and high quality teachers in a small number of schools, while others suffer in comparison. Ministers rightly want to tackle the issue of low attendance at top universities of disadvantaged children. However, a better way to do this may be to boost the attainment and aspirations of highly able pupils across all comprehensives throughout their secondary education. The best candidates for support are not necessarily those that reveal themselves in the 11 plus.
Taking the evidence as a whole, it’s clear that grammars are no quick fix for England’s social mobility problems.