Schools and Inequality, Part 253

Simon Burgess

Recent debates about inequality have been dominated by discussion of pay at the very top of the income distribution. Whilst this is undoubtedly important, we surely also need to be concerned about the earnings gap between the middle and the bottom, and between the bottom and the very bottom.

This is where education comes in. How can school reform reduce earnings inequality? For all that special factors do matter, earnings are hugely affected by qualifications, and so the chances for pupils to get good qualifications matter for earnings inequality.  The earnings penalty for not getting at least 5 good grades at GCSE is estimated to be over 25%.

There are obviously a lot of facets to this, many of which have been discussed previously in this blog.

A key issue is your chance of getting into a high-performing school, bringing school admissions to centre stage. A report last week from the BBC’s Sean Coughlan  discusses an innovative admissions policy from a new academy in Birmingham.

Innovation is surely needed.

At the moment we decide the question of who gets into high-performing schools by working out who lives nearest. Unsurprisingly, the housing market reflects this advantage in substantial premia for houses close enough to guarantee entry to the most popular schools. And as reported here, estate agents are providing ever more, and more localised, information on nearby schools.

We showed here (scientific paper) and here (blog) that this schools’ admission policy has a major effect on pupils’ chances of getting into a top school. We showed that the gap in average school quality between the richest and poorest fifth of families increases by over 30% when factoring in the probability of admission based on proximity to the school.

If we want to do something to reduce inequality and to raise social mobility, one place to start is to even up the chances of attending the best schools.

Maybe we should just abandon the proximity criterion altogether? There are pros and cons to the idea of having tight local catchment areas around schools. The major con is as just mentioned, it has a very regressive impact. The pros are that it can generate a sense of neighbourhood – kids can play together, study together and form friendships. The Mayor of Boston explains clearly the downside of having more fragmented school catchment areas: “Pick any street. A dozen children probably attend a dozen different schools. Parents might not know each other; children might not play together. They can’t carpool, or study for the same tests.”

So it’s not an easy problem.

The school in Birmingham, University of Birmingham School, has a very interesting solution to this problem: “It is basing its admissions for 11-year-olds on how close families live to four “nodes” across the city. After siblings and looked-after children, half the intake will be based on closeness to the school, with the other half of pupils shared between closeness to the other three points on the map.” The idea is explicitly to try to secure a diverse intake to the school, “mixed in terms of socio-economics, ethnicity and academic ability.”

Mossbourne has a different approach to the same aim, combining ability banding and geographical rings round the school. Whilst living very close to the school helps, the school reserves places for those living further away.

In essence, both of these solutions are like this: reserve a fraction of places for ‘local’ children; the remainder of school places are open to anyone and if those are over-subscribed, use something other than distance as a tie-breaker, such as random draw.

The policy question is: what is the optimal fraction to reserve for those living further away? 25%? 50%? Research to be done, but its surely not 0%.

One way to make progress is to study what happens in practice when schools innovate.

More, please.