Market forces have changed the fortunes of schools in England and Wales in unexpected ways, according to new research.

The failing and improving schools could not always have been predicted when the market began to operate. At the end of the 1980s the average percentage of places filled among schools with a consistently declining market share was just below the average for the sample, at 76 per cent.

By last year, this figure had dropped to 63 per cent, and four of the schools - out of a total sample of 300 - had closed.

At the beginning of the study the consistently improving schools filled only 65 per cent of their intake places - dramatically fewer than the sample mean - but rose to 92 per cent by last year. The explanation for this unexpected pattern will be sought in the next phase of research by the Open University team, funded by the Economic and Social Research Council.

It found that 10 per cent of schools were suffering from declining rolls which sometimes forced closure, while 10 per cent consistently improved their annual recruitment. More than half of these took in maximum pupil numbers or above.

So far, the Impact of Competition on Secondary Schools (ICOSS) research has discovered that the market forces have tended to reinforce existing differences between schools in the proportion of intake places filled. Almost 40 per cent of secondaries were recruiting at or above capacity when competition was introduced, and the vast majority continued to do so.

The “middling” schools in the sample - 45 per cent of the total - did increase the percentage of places filled, but only as a reflection of population growth which saw the average number of new pupils rising from 80 to 88 per cent. Two-thirds of the schools which consistently failed to fill places showed no underlying trend in annual recruitment, once changes in pupil numbers were taken into account.

This was reinforced by analysing groups of closely located schools. While there was usually at least one which was regularly full or oversubscribed, in only a quarter was there a school showing either consistent decline or improvement or popularity over the full six years of research.

ICOSS also revealed a tendency for popular schools to be associated with both a high number of GCSE passes and low proportions of pupils eligible for free meals.

Jason Hardman, one of the researchers, said: “While the association between recruitment success and GCSE results is consistent with the hypothesis that competition promotes an increase in educational standards, the link with socio-economic status raises questions about the social cost of striving for excellence in education.”