Alan Smithers: American research shows Kelly the way

Another year, another Schools Bill. All credit to the Government for acknowledging that secondary education in England is very uneven and unfair. The prospectus accompanying the Secretary of State's speech at the North of England Conference last week could not have been clearer: "A child's achievement is still too closely linked to his or her parents' background and the gap between high and low performing schools is too large."

What is less clear is how its preferred solution of parental choice across independent trust schools would actually work. The hope seems to be that, through measures such as choice advisers and free school transport, many parents who do not now make choices will be enabled to do so, and that sponsorships and partnerships will raise the number of good school places. The Government appears to be taking its inspiration from such schemes as the Chicago Public Schools Choice Programme, where there is solid evidence that parental preference has had a major impact on reducing racial segregation.

Research from Sweden, however, is not so encouraging. There, the National Agency for Education conducted a review of reforms that have seen that country moving towards something like what is envisaged for England. This revealed that, while school development could be enhanced, "school choice reinforces segregation both ethnically and in terms of performance".

Why the diametrically opposed findings? In part, the difference stems from the starting points. In Chicago, the sharp racial division between schools broke down as ambitious black parents sought places in high performing "white" schools. But, in Sweden, where schools were initially more homogenous, the highly educated exercised their right to choose to a greater extent and converged on the same schools. As a result, the agency said, "a clear division not only of pupils but also schools into 'better' and 'worse' is today a reality".

There is another important difference between Chicago and Sweden, and that is how they have dealt with the inevitability that the supply of school places will not exactly match the wishes of parents and, therefore, some schools will have a lot more applicants than they can take. In Sweden, academic selection is ruled out, but it has crept in - for example, schools may offer diagnostic tests and counsel prospective parents on the likelihood of their child's success and/or encourage special tutoring as a condition of getting a place. So the way is open for schools and parents to manipulate the system. In contrast, Chicago decided that the places in oversubscribed schools should be settled by ballot, contributing to a more even spread of pupils.

The debate in England has tended to centre on the potential for increased academic selection by trust schools controlling their own admissions. Less attention has been given to what the National Union of Teachers has memorably called "all sorts of fancy footwork". An admissions code, whether mandatory or not, will be vulnerable to those lightest on their feet.

If the Government's objective is social justice, the transparent fairness of settling places by ballot has much to commend it. Even this, however, would not be enough, because there is nothing in the proposals to encourage schools actively to seek out disadvantaged pupils so that they get into any ballot. Without this, places in the best schools would still be shared out among parents like those who apply for them now.

The current policy is to close failing schools and replace them with expensive new academies. The investment is in structures. Over time, a splendid new school is likely to become popular with parents and the children from its immediate neighbourhood are likely to be squeezed out.

The Policy Exchange think tank has come up with the very interesting suggestion that it would be better to attach the money as a large premium to the pupils from the failing school so that other schools would actively seek to recruit them. This idea could be broadened to differential funding for the disadvantaged.

Whatever their other merits, parental preference and self-governance for schools will not, of themselves, do much to uncouple achievement from background or close the gap between schools. To bring this about, there would also have to be genuinely fair admissions and, probably, recruitment incentives. The passage of the Schools Bill would be a lot smoother if the Government could persuade its supporters that the proposed reforms would benefit most of those being let down by today's system.

The writer is Professor of Education and the director of the Centre for Education and Employment Research at the University of Buckingham

education@independent.co.uk