10/08/2016 13:46 BST | Updated 11/08/2017 06:12 BST

Want Grammar Schools to Work? Then Exclude the Most Well-Off

Pacific Press via Getty Images

You have to feel a little sorry for Justine Greening.

Having seen her predecessor leave office to widespread acrimony over the question of universal academisation, she has within her first month been forced into a debate on perhaps the only issue more toxic than academies - grammar schools. Yet to either blindly champion grammars and overlook their drawbacks, or to wholly discard them as a potential policy tool in pursuit of important social goals, is equally unhelpful.

Let's start with the positives. Evidence suggests that disadvantaged children who get into a grammar school do see an improvement in their academic performance. The education system currently fails to provide for the needs of these high aptitude but low income pupils. Grammar schools, with their focus on broadening pupils' horizons and inculcating an ethos of high expectations, are a powerful vehicle for the most able pupils in deprived areas to receive a high quality education adapted to their needs.

However, the grammar school system has been rightly criticised for two main reasons. Firstly, it creates a two-tier system of education, making insufficient provision for those pupils who fail the entry test, resulting in lower societal expectations of what they can achieve and lower final attainment relative to what those pupils could otherwise have achieved.

Secondly, it can entrench middle class advantages by not having a sufficiently socially diverse intake (in effect, ensuring better-off children's access to high quality education while leaving less well-off children to attend the less well-performing, non-selective local schools). Even once prior attainment is controlled for, research suggests high achieving children on free school meals are considerably less likely to attend grammar schools than a non-free school meal child of similar ability.

So grammar schools can have a positive impact on the educational outcomes of children from poorer backgrounds; yet they have not done so at scale, because too many bright, deprived pupils never make it to the schools in the first place, blocked by other students with advantages such as 'pushy' parents or tutoring in the entry tests. They then end up consigned to a poorer standard of education and, in the worst cases, written off as second-class students.

These obstacles are not however insurmountable. If the right caveats are put in place, past mistakes can be avoided. To be clear: grammars are an undesirable mass solution for the education system's future direction. But if their expansion is properly targeted, and their limitations acknowledged, their beneficial effects can be harnessed for social good.

Firstly, new grammar schools should only be opened in areas of the country which are suffering from high levels of deprivation (as measured by the Income Deprivation Affecting Children Index), and where there are no existing local schools rated Good or Outstanding by Ofsted which can adequately provide for the needs of bright but deprived pupils.

In particular, areas with the best prospects for regeneration and increased economic activity in the short- to medium-term should be prioritised, since these will present the best opportunities for skilled local employment for those gifted students who attend the grammar once they finish school or university.

Secondly, set as a mandatory entrance criterion for any new grammar school that pupils' parents are not higher or additional rate taxpayers (and have not been so for the past five years, save for in exceptional circumstances such as a forced pay cut or redundancy).

Thirdly, the 'Teaching Premium' ResPublica proposed in our Manifesto for the North last month, and which I have discussed elsewhere, should be piloted in the other schools in any borough where one or more new grammar schools are established.

We designed the Teaching Premium to attract high quality teachers to practice in deprived areas: the establishment of a grammar school, teaching exclusively high ability pupils, will itself act as a pull for many teachers.

Yet to ensure the best teachers already in the area are not also all drawn to the new grammar, benefits should be made available to those who choose to remain at existing schools in the deprived area, or who move to the area to teach at one of these schools. This will help to safeguard standards in the local non-selective schools.

Finally, it is also imperative that the non-selective schools are encouraged to recommend pupils to be taken on by the grammar, if those pupils are showing signs of rapid academic progress but did not pass the entry test first time around. This will inhibit a culture of disconnect growing up which can contribute to the diverging quality of education which has been associated with selection in the past.

As emotive and important a topic as selective education is, gut reactions will get us nowhere. A focus on tailoring the policy, to do the most good and least harm, is a better starting point.