Independent studies have found that school choice improves reading and maths by 5-6 percentile points and that it benefits ethnic minorities and the poor most.

A group of Labour backbenchers has published an 'alternative white paper' because they believe that the proposed 'trust' schools will introduce backdoor selection and reinforce the existing pattern of disadvantage. But there is powerful evidence that school choice benefits the least well off, not least from American charter schools.

Charter schools are supported by public funds and may not charge fees. Public authorities pay them a cash amount per pupil, usually lower then the average cost of local state schools. They cannot select their students based on admissions tests, and must obey many public school regulations, including test requirements, although they are often exempt or partially exempt from regulations about teacher certification. To avoid backdoor selection, state laws typically require charter schools to select students by lottery when the number of applicants exceeds the number of available places.

One of the most authoritative studies has been carried out by Caroline Hoxby, Professor of Economics at Harvard University, and Jonah Rockoff, of Columbia Business School. They looked at charter schools in Chicago, where school places are allocated by lottery when a school is over-subscribed. The study compared the achievements of pupils selected by lottery with those who were not (and who, consequently, attended local state schools). This method has the advantage of eliminating the 'selection' effect that statisticians worry about. The results cannot be explained by 'home background' because all the pupils had motivated parents who wanted their children to attend charter schools - some were lucky enough to attend and others were not.

Maths results are reported as 'percentile scores', that is all test results are ranked from 1 to 100. Maths scores were over 6 percentile points higher and reading 5 points higher. When adjusted for gender, ethnicity, participation in the federal free or subsidised lunch programme, and the need for special education, maths results were 6.18 percentile points higher and reading 5.11 points higher.

More than half of the pupils studied were enrolled at schools under the aegis of the Chicago International Charter School, whose students were mainly black or Hispanic and most of whom received free or subsidised lunches.

Check the original source here

A second study compared 99% of charter school pupils in the 4th grade (age 9-10) in charter schools throughout the USA, some 50,000 pupils. Charter schools were compared with schools in the same neighbourhood, and therefore share similar economic conditions and draw from a similar population of parents and pupils. It is fair to conclude that the study compares charter schools with other schools that pupils would have gone to if the charter school were not there. Charter school pupils were found to be 5.2% more proficient in reading and 3.2% more in maths, after deducting charter schools that catered for 'at-risk' or gifted pupils.

Charter schools were especially likely to raise the achievement of pupils who were poor or Hispanic. In highly Hispanic areas the advantage was 7.6% in reading and 4.1% in maths, whereas in a typical charter school the advantage was 4.2% in reading and 2.1% in maths. In high poverty areas the advantage was 6.5% in reading, compared with 2.6% for other charter schools. However, the same advantage was not found in predominantly black areas.

Click here for the original source.

According to Dr David Green, Director of Civitas, "Egalitarians define middle class success as a public policy problem. But middle-class parents whose children are successful should be seen as 'good parents'. We should glad that their children do well and focus public policy on the children who come from a disadvantaged background who are badly served by the existing state system. A pluralistic system that drives up standards by permitting energetic new entrants to shake up the existing schools will benefit them the most. Labour's backbench rebels are acting against the interests of the poor."

School choice and equality: a survey of the evidence by Civitas (December 2005) (PDF)

A useful survey of school choice research in the US can be found here.

The University of Washington has also assessed the evidence about charter schools.

For further information ring Robert Whelan, Civitas, 020 7799 6677.

Or Anastasia de Waal, or David Green, Civitas, 020 7799 6677.