Exclusions could be cut by a quarter by putting social workers into schools, a Home Office study has found.
A three-year pilot project in Yorkshire comprehensives found that employing home-school support workers reduced both exclusions and truancy - and was highly cost-effective.
In one case, a support worker managed to prevent the permanent exclusion of a girl in local authority care who would otherwise have been sent to a residential school costing more than pound;45,000 a year.
But the fall in truancy made fixed-term exclusions rise in some cases because truants returning to school found it hard to adjust to discipline.
The findings of the pound;270,000 project, conducted between 1996 and 1999, are reported by Graham Vulliamy and Rosemary Webb in the latest issue of the Oxford Review of Education.
But, despite the importance of its findings for current government policy, the Home Office has still not published the final report, which the researchers submitted nearly three years ago.
Most education authorities allocate education welfare officers to large secondary schools. But they are not usually members of the school staff, some are untrained and many are confined to chasing up truants.
In the Yorkshire project, five support staff - all trained social workers - were placed in seven relatively disadvantaged schools in North Yorkshire and the City of York.
Each was given a caseload of up to 10 pupils at a time, selected by senior management mostly because of their behaviour. Three-quarters had caused severe disruption in class and half had offended. The support workers befriended the pupils, taught them to manage their anger and tried to improve their self-esteem and relationships with others.
They also supported their families and stepped in immediately to help with crises in school that could lead to exclusion.
Over the three years, they helped 208 challenging pupils, nearly two-thirds of them boys. Half were in Years 9 and 10.
Senior managers at the schools estimated that they saved 26 pupils from permanent exclusion, representing a 25 per cent reduction in the exclusion rate over the three-year period.
They also said the intervention of support staff had led to sharp reductions in the number of fixed-term exclusions for some pupils.
But the number rose in the case of others who were coaxed back after long-term truancy and found it hard to re-adjust. As one head remarked:
"Some of these children have never attended school regularly, or not for two or three years, and they're used to doing exactly as they like; they're completely out of control and so when they come into school they're not used to conforming."
Oxford Review of Education, Vol. 29. No.1, March 2003, www.tandf.co.uk