Grammar schools: why aren’t rational arguments cutting through?

2 September 2016

Have we entered an era of post-truth? Some commentators think so. Yet few would liken the absurdity of the EU referendum, where unsubstantiated claims were made on both sides, with the current debate on grammar schools. Grammar school opponents have some evidence on their side, whereas their proponents have none. Yet school experts have been hounded on social media by ill-informed rookies, and only one in five members of the public think grammar schools harm social mobility. Rationality just isn’t cutting it. Why?

School selection has many guises
The truth is that school selection is not limited to grammar schools. Many top-performing comprehensive schools, including academies and free schools, have both overt and covert methods of selection. Perfectly legal and common methods (see here), such as the use of catchment areas for selecting pupils when the school is oversubscribed, mean that entry to a popular school is contingent on affording to buy or rent a house in the area. Covert methods, such as complicated school admissions criteria and ‘meet and great’ sessions for parents (a way to circumvent the ban on parent interviews), may discourage parents and pupils from disadvantaged areas from applying to the best schools.

The reality of selective comprehensive schools has given meat to the grammar school enthusiasts. How do we know that new grammar schools would be more selective than new free schools or maintained schools, they say? Some have dismissed this line of reasoning by asserting that the existence of selective state schools does not imply advocating more. Yet this misses the point. Given that we need more school places (roughly 7 per cent more by 2020-21) it is reasonable to ask whether providing these places in grammar schools will make the system more or less selective than new places elsewhere.

Selection constrains choice
A first look at the figures would suggest new grammars are much more selective than the alternative. The 164 grammar schools have an average of 3 per cent of pupils on free school meals, whereas both free schools (the current means of providing new school places) and all state-funded schools have an average of around 15 per cent. Unfortunately, it is not that straightforward: these headline statistics overlook important information about who is applying to the schools in the first place. Grammar schools may dissuade poorer pupils from applying, as research suggests advantaged parents have stronger preferences for academic performance and all parents’ preferences are affected by accessibility of a good school. Ultimately, however, the choice to apply to a school lies with the parent and pupil.

Research, however, suggests admission constraints affect pupil allocation more than parental preferences. One study by the Centre for Market and Public Organisation has examined the preferences of thousands of parents’ choice of primary school (implied by their application forms) with the actual allocation of those pupils to schools. It found that the gap in average school quality attended by the poorest and richest pupils (measured by primary school results) is one third greater when distance and probability of admission was taken into account. That means the process of using proximity and other oversubscription criteria limits the ability of poor pupils to attend good schools, even when they want to.

Another study by the Sutton Trust compared the percentage of disadvantaged pupils in non-selective primary schools to a neighbourhood of likely choosers of that school (“recruitment neighbourhood”). The characteristics of the likely choosers were based on a radius of postcodes from pupils that had, in the five years prior, attended the school. The research found that there were more primary schools with fewer disadvantaged pupils compared to their recruitment neighbourhood than those with more (but only just). The striking finding was that the 100 schools most selective against disadvantaged pupils had complicated admissions arrangements, with as many as 18 different criteria for who is admitted when the school is oversubscribed.

Choice for some, not all
Given this body of evidence, why put further, additional constraints on school choice? Grammar school advocates have a fudged answer. Firstly, they see grammar schools as highly popular among some parents and thus consider their expansion as increasing parental choice. Secondly, despite the persistent correlation between attainment aged 11 and parental income, they still see academic selection as more liberating than constraining.

Both arguments are wrong. It is true that grammar schools are popular with some: evidence suggests that high-attaining pupils travel unusually large distances to attend grammar schools outside the local authority where they live. But extending choice only to those that already have a lot of it means reducing choice for those with less.

The second argument is easier to dismiss in practice than in theory. Imagine a utopia in which pupil attainment at age 11 is equal irrespective of parental income. Would we still be comfortable with academic selection? We might be more comfortable, but we would still lose the benefits of school choice. Allowing pupils to choose schools (and not the other way around) can, in theory, improve allocative efficiency, leading to a better matching of pupils to schools, the migration of pupils to ‘good’ schools (thus the failure of ‘poor’ ones) and higher overall quality of schools through competition.

Can the rationing of choice ever be fair?
The answer to this question depends on what a fair system looks like. Many think a fair system would be more socially cohesive, but that is not necessarily the consequence of increased school choice (many people choose to separate themselves from others). However, if social cohesion is an outcome the Government is trying to achieve, then one idea floated is to require new grammars to reserve a certain proportion of places for pupils on free school meals. The logical extension of this is full admissions ‘banding’, based either on prior ability or parental income, whereby schools that are oversubscribed must prioritise pupils in a way that balances the school intake to represent those living in the area. This is a relatively untested idea, and is not without risks. The theory is that the school’s social fabric would mirror that of its local area.

Those looking to make the system fairer for each individual have other options. As the evidence above highlights, pupils from wealthier backgrounds are more likely than pupils from poorer backgrounds to go to the school they want to. A fairer system would remove this imbalance, making the likelihood of getting into an oversubscribed school equal, irrespective of a pupil’s background. This would require a radical overhaul of the current School Admissions Code, including removing the current priority given to pupils within the catchment area and also removing priority for pupils eligible for free school meals or with special educational needs.

A lottery to avoid all postcode lotteries
Random allocation to popular schools would work in a similar way to lotteries in US charter schools. Once a school becomes oversubscribed, the school would have no say in who gets priority and, as is now, anyone in the country could have applied to the school. This has the potential to better allocate pupils to schools they desire (no longer constrained by postcode), leading to success for popular schools and overall higher quality schools through competition. Pupils may have to travel further, but the grammar school evidence suggests some already do. The opportunity would be there.

If the Government wants to speak for all, not just “the privileged few”, this more liberal system of school choice offers more hope than grammars, free schools or comprehensives. It is not a defence of the status quo, and would require significant change to the way pupils are currently allocated to schools. However, it would send a clear message that every pupil should have equal opportunity to attend a school of their choice. Grammar schools give the illusion of school choice, but risk further compounding unequal access to the best schools.

Amy Finch, Amy Finch, Research Manager and Head of Education, Reform



No comments yet.