"Ideally, we would do research on grammar schools from the older data when grammars were about and we saw them abolished to look at the discontinuities across time and between regions. The 1958 and 1970 cohort studies would be perfect. They are, alas, no use for us. (A. Manning, J. Pischke, Comprehensive versus Selective Schooling in England & Wales: What do we know?: "We conclude that we probably do not know very much about the effect of comprehensive schooling in Britain, or elsewhere for that matter").
We can, therefore, only look at current grammars.
- First, grammar schools are socially selective
This is confirmed by the aggregate data on grammar schools.
The CPS critique claims that this selection is all because it is only in leafy areas that these schools exist.
This is wrong. As the CPS later says, the Sutton Trust has shown how the catchment areas for grammars are indeed leafier (12% FSM as against a national average of 14%). However, the grammar schools themselves only admit 2%. The selection is internal to the schools.
Dr Leon Feinstein has researched this effect. At the age of 11, a child is not tabula rasa. High ability kids from low income families will be beaten in tests by low ability high income kids. The intervention comes in too late. (L. Feinstein,"Inequality in the Early Cognitive Development of British Children in the 1970 Cohort" . Economica, 70, (2003) pp 73-97). This is supported by work by the Nobel laureate James Heckman which shows that interventions can only be cost-effective if they take place in the early years. By that point, it is too late.
- Second, areas with grammar schools do no better than those without
The CPS say that kids in grammar schools do better. This is true,
but areas with grammar schools do not do better as a whole. The
Northern Irish evidence they cite is dramatic, but to pretend that the
only difference between Northern Ireland and the rest of the UK is its
educational system is not credible. A different religious structure,
employment rates, employment patterns, income distributions, gender
balances and Government spending rates make such comparisons useless.
Looking at directly comparable regions in England and making
econometric evaluations, the academic research concludes that aggregate
change in standards for kids is "not substantive". (A. Atkinson, P.
Gregg, B. McConnell, ‘The result of 11 plus selection; An investigation
into opportunities and outcomes for pupils in selective LEAs’)
- Third, the only way to raise social mobility is to lift the success of all schools
As the CPS says, the academic comprehensives are also socially
selective. They are right. We have cited a paper by Briggs and
Burgess which shows that FSM kid are half as likely as non-FSMs to go
to academic schools The CPS claim this is simply due to the
house-price premium. They should have read the report. House prices
are of course important (location accounts for a large part of the
difference). However. it shows that whilst location matters, one
seventh of the reduced likelihood of good schooling comes from poverty
alone. Adjusting for test scores and other factors, a poor kid living
in the same postcode (a area with just 15 houses in it) is still 7%
less likely to go to a good academic school than his richer neighbour.
(A 2% reduction from 29%) (S. Burgess, A. Briggs, School Assignment,
School Choice and Social Mobility, CMPO Working Paper No. 0.16/157,
The bad schools tend to be full of poor kids and the good schools have
almost none. However, the CPS see this as an argument for increasing
and embedding social selection. In truth, it supports a case for
raising the standard of all schools – especially those at the bottom.
How can we do this without dragging the top down? By contrast to the
evidence on grammar schools, Professor Caroline Hoxby at Harvard has
followed on from Friedman’s research and now leads the field in
research on improving education through the use of choice.
The research is fairly clear-cut and has overcome great hostility from
the left. It shows that school choice is ‘a rising tide that lifts all
boats". We need a system which is open, fair and allows independence
within schools. (C. Hoxby, ‘School Choice and School Productivity’,
2001) This is a schooling system with a strong track record of
success in Holland, Sweden and certain US states. Hoxby’s model
schools use a number of systems for allocating places, but not academic
We want to allow people to set up schools easily, to allow competition,
to foster better results and improve social mobility; the evidence
unambiguously suggests that this is the way. Per-capita funding and
liberalising the supply side of the educational sector are the way
forwards for genuine equality of opportunity."