Education-line Home Page

Reducing exclusion from school: what really works

Susan Hallam and Frances Castle

Institute of Education, University of London

Paper presented at the European Conference on Educational Research, Edinburgh, 20-23 September 2000


In recent years in the UK, behaviour problems leading to the exclusion of pupils from school have increased dramatically. To address this problem, the Department for Education and Employment set up of three types of project: Multidisciplinary behaviour support teams; In-school Centres; Secondment of teachers to Pupil Referral Units. This paper describes an evaluation of these projects. Questionnaires were sent to project leaders; follow up telephone calls were made to selected projects; seven projects were selected for in depth study. The findings suggest that setting up projects alone does not guarantee change. Success depended on the projects being implemented with the full commitment of school management; involving the whole school; including parents; and placing the responsibility for their behaviour on pupils. Where these conditions were met the projects had cost benefits; promoted a more positive school ethos; and generated change. These findings have implications for interventions worldwide aiming to improve behaviour in school.


Between 1993 and 1996 the number of pupils excluded from school increased dramatically (Parsons, 1999). By 1995/96, there were 12,500 permanent exclusions from primary, secondary and special schools, an increase of 13% on the previous year (DfEE, 1997). During 1996/97 the rate of increase in the numbers of pupils excluded reduced to 2% suggesting a slowing down (DfEE, 1998). This pattern was confirmed by the figures for 1997/98, 12,700, representing a reduction of 3% on the previous year (DfEE, 1999). However, there were wide regional variations. The pupils most highly at risk of exclusion are Afro-Caribbean boys (Commission for Racial Equality, 1985; Gillborn and Gipps, 1996).

The reasons for pupils being excluded from school tend to be related to general disobedience, or physical aggression against staff and other pupils, e.g. constant refusal to comply with school rules, verbal abuse, insolence to teachers and disruption and defiance (Galloway, 1982; DFE, 1992; NIAS, 1989). Physical aggression accounts for approximately 25% of all exclusions. Most exclusions occur after a series of minor incidents rather than a single major one.

The cause of the increase in exclusions up to 1996/7 is not clear. It is likely that it resulted from a range of complex factors interacting. Certainly, schools have been subject to a period of great change and have been under pressure to adopt inclusive policies in relation to pupils with Special Educational Needs. Some have suggested that the increase was due to the effects of the Education Reform Act on school management (Stirling, 1993; Lloyd-Smith, 1993). However, the evidence does not support this conclusively. Others have suggested that the ethos and organisation of the school is the key factor in the level of exclusion (Galloway, Mortimore and Tutt, 1989; Garner, 1993). This might be mediated through teacher stress and the paucity of training for teachers in coping with exceptionally difficult children (DES 1985, 1989). Financial constraints on schools, leading to larger classes and lack of resources coupled with the demands of the National Curriculum and Office for Standards in Education (OFSTED) inspections may have been important, as may have been the increasing pressure to raise standards. The market philosophy introduced into education has also increased competition between schools leading them to wish to present a more favourable image to education consumers. Schools may therefore have tightened up on thresholds of acceptable behaviour to present an attractive image to parents. Increased parent power may also have contributed in that parents themselves have exerted pressure on schools to exclude children who, by their behaviour, interfered with the education of the remaining children.

The UK has also undergone substantial social change. According to European Union figures, 20% of the UK population live in poverty. Such factors are likely to influence children's behaviour and performance in school, the links between poverty, ill-health and educational under-achievement being well documented (Kumar, 1993). The complex interactions between changes in schools, families, communities and society which have led to increased levels of disruption, non-attendance and delinquency are likely to underpin the increases in exclusions (Galloway, 1985a; 1985b; Hallam and Roaf, 1995; Hallam, 1996).

Consequences of exclusion for the child and their family

The consequences of exclusion for the child are serious. A sizeable proportion of permanently excluded pupils simply disappear from the educational system (Blyth and Milner, 1993). Many appear to be out of school for long periods of time with little or no educational input and as a result suffer educationally and personally (Cohen et al., 1994; Parsons et al., 1994). Many of those excluded are already among the most vulnerable in the community. Exclusion therefore serves to exacerbate what are already difficult circumstances for the child.

Relatively few excluded pupils return to mainstream school. Department for Education figures (DFE, 1995) indicated that while 27% of primary pupils returned to mainstream school only 15% of secondary pupils did so. A number of Local Education Authorities (LEAs) reported that it was becoming increasingly difficult to return pupils to mainstream school particularly in Years 10 and 11. When exclusion occurs at primary school there is particular cause for concern.

Reports on juvenile crime (Utting, Bright & Henricson, 1993; Crime Concern, 1993) suggest that the risk of crime is increased by a number of factors including poverty, unemployment, homelessness, social isolation, poor parenting, family conflict and breakdown, and school failure. Children most at risk are those where several of these factors occur together. Lack of achievement, behavioural problems, truancy and exclusion from school are predictive of subsequent involvement in crime (West, 1982; Hagell & Newburn, 1994; Ouston, 1984). It would therefore seem important to attempt to maintain all children in full-time mainstream education. While this may lead to an increase in educational costs in the short term, there may be long term cost benefits.

The pilot projects

Acknowledging the arguments, educational, social and financial for attempting to reduce the number of exclusions from school, the Department for Education and Employment (DfEE) set up a series of pilot projects which had the reduction of exclusion and indiscipline as their principle aim. The projects were of three kinds: In-school centres to support pupils at risk of exclusion from school; Multi-disciplinary behaviour support teams; secondment of mainstream school teachers to Pupil Referral Units (PRUs). Forty four Local Education Authorities (LEAs) were successful in securing funds for 67 projects. Seven Grant Maintained schools were also successful in their bids for funds to establish In-school centres. The funded LEA projects included 24 Multi-disciplinary behaviour support teams, 34 In-school centres and 9 secondments to PRUs.

An evaluation of the projects was commissioned by the DfEE to establish how the projects had been implemented in different schools and LEAs; monitor the effects on exclusions from school; examine the strengths and weaknesses of different projects and their implementation; consider cost effectiveness; and identify examples of good practice in each type of project.

The research

118 questionnaires were distributed to the LEAs involved and to each project. 91 responses were received, a response rate of 77%.

Considerable differences in the implementation of the projects became apparent. Of the 17 Multi-disciplinary teams that responded, 3 worked only in primary schools and 3 only in secondary schools. The majority operated cross phase. The number of schools which teams supported varied hugely. One secondary team worked with only 3 schools, spending two terms in each of them; one cross-phase team worked with one secondary school and 6 'feeder' primary schools, while four of the teams indicated that they worked with more than 50 schools each.

The 27 responding In-school centres were located in 19 LEAs. At primary level there were three projects, 2 operating at KS2 and one at KS1 and KS2. At secondary level 24 projects responded. Six operated at KS3, two at KS4 and 16 at both KS3 and KS4. Two of the primary In-school centres functioned independently of the schools where they were sited and took KS2 pupils, on a part-time basis, from other schools.

Differences were identified in the setting-up and functioning of secondments to PRUs. In one case, teachers had been seconded for periods of two terms in each of the three years, in others several teachers were seconded on a part-time basis. In only one case had there been year long, full-time secondments. These differences were, in some cases, associated with recruitment difficulties. In some projects, several teachers were seconded for short periods of time to enable the delivery of as broad a curriculum as possible in the PRU. To this end, one project had seconded six teachers for periods of six weeks, each of them spending three hours per week in the PRU.

After initial analysis of the data arising from the questionnaires, in-depth telephone interviews were undertaken with personnel from a sample of projects. These were selected to reflect identified differences in implementation and also apparent differences in success. The selection was based on the questionnaire responses received at that time. Interviews were undertaken with personnel from 5 Multi-disciplinary behaviour support teams, 8 In-school centres and 2 secondments to PRUs.

Where possible, two interviews were undertaken for each project. In the case of the Multi-disciplinary behaviour support teams, the interviews were with Head Teachers of client schools. In the case of the In-school centres, they were usually with the Special Educational Needs Co-ordinator (SENCO) and either a Head of Year, Head Teacher or Deputy, although this varied. For the secondments to PRUs, follow-up interviews were undertaken with the teacher in charge and the secondee.

Selection of projects for fieldwork

An interim report was prepared on the basis of the documentation provided by the project leaders in their initial bids for funding and subsequent progress reports; data from questionnaires sent to LEA representatives; data from questionnaires sent to directors of In-school centres, heads of Multi-disciplinary behaviour support teams, and leaders of projects seconding mainstream teachers to PRUs; exclusion data from the DfEE data base for the years 1994/5 and 1995/6; and telephone interviews with those in a position to provide further evaluative information about the projects, e.g. head teachers, SENCOs, teachers. A Quality Assurance Group, made up of representatives from schools, LEAs and educational researchers, in conjunction with the DfEE steering group selected the projects which would be subjected to in depth analysis. Although the initial intention had been that two contrasting examples of good practice would be selected from each of the three types of projects, it became clear that the secondment of mainstream teachers to PRUs had not been widely taken up by LEAs, the projects were therefore selected from the In-school centres and the Multi-disciplinary behaviour support teams.

Criteria for success and issues considered by the selection team were whether changes had been made at the level of the whole school which were likely to be long lasting; whether the impact on individual pupils had been substantial and was likely to be long lasting; whether there were likely to be long-term changes in approaches to children's behaviour, discipline and attitudes towards school in the family.

The seven projects selected for in-depth study by the Quality Assurance Group included three Multi-disciplinary behaviour support teams and four In-school centres. One of the In-school centres was at primary level.

Field Work

During the field visits in-depth interviews were undertaken with project leaders and focus interviews with groups of pupils and teachers. Detailed case studies of particular pupils were considered and interviews undertaken with their parents.

Analysis of data

The analysis of the data took place at a number of different levels. Overall comparisons of the different types of project were undertaken taking account of their effectiveness in relation to exclusion from school; how they were implemented; their relative costs; and the evaluations of those involved.

Comparisons were made within project types to consider how the projects were implemented differently in different locations and how these differences affected the reduction in exclusions from school; cost effectiveness and the evaluations of those involved.

Detailed comparisons were made of the case study projects taking account of the reduction in exclusions from school; interviews with head teachers and project directors; the information derived from interviews with the pupils and teachers; cases studies of selected pupils; and cost analysis.


A full account of the findings can be found in Hallam and Castle (1999). A summary of the key points is provided here.

Project implementation - There was wide variation in the way that the projects were implemented. This variation depended, in part, on the phase of education within which the projects were operating, existing provision, geographical factors, funding, resources and staffing. The success of the projects did not depend on any particular type of implementation but rather on the quality with which they were enacted and the extent to which they were able to engender long-term change in the school.

Secondment of teachers to Pupil Referral Units - This was not widely taken up by LEAs. Where it was implemented it was regarded as successful by those involved particularly for expanding the curriculum within the PRUs; developing the skills of secondees in the management of behaviour; and improving links between PRUs and mainstream schools.

Reduction in exclusions - Reliable data regarding exclusions was only available for the first year of the programme. The data suggested that In-school centres and Multi-disciplinary behaviour support teams, depending on how they were implemented, could both be successful in reducing exclusions from school. While in some cases it was extremely difficult to establish the effects of the Multi-disciplinary behaviour support teams because of the way they operated, such evidence as was available suggested a 20% reduction in exclusions in the schools within which they worked. While not all In-school centres reported a reduction in permanent exclusions, the overall reduction for all centres for the year 1996/97 was 4.3%. These figures compared favourably with the national rise in that year of 2% (DfEE, 1998).

Cost effectiveness - Where costings based on accurate figures were possible, intervention by a Multi-disciplinary behaviour support team cost on average 40% less and an In-school centre on average, 30% less than the potential cost to the education system of educating out of school those pupils who might have been excluded. This was irrespective of the social costs to pupils, families and society.

Successful projects - Successful projects of both types shared many features in common. These are outlined in the sections below. In relation to particular types of project, the success of Multi-disciplinary behaviour support teams did not depend on the professional make up of the team, although having a member supporting home-school links was valued. The quality of the staff and their recent and relevant experience working in a school or with difficult young people emerged as crucial, particularly in establishing their credibility in schools. Successful In-school centres were characterised by the existence of an appropriately staffed physical centre; the adoption of an approach which combined withdrawal of pupils with support for them within normal classes and flexibility in the provision of support.

The context of the projects - Successful projects involved interventions at three levels: whole school development work; class-based work; and work with individual pupils. Each was necessary, although different emphases and combinations were employed in different projects. The particular combinations of geographical, social and educational contexts within which individual projects were embedded were important in determining the type of intervention necessary to complement existing provision and satisfy local needs.

The development of projects over time - The evidence suggested that successful projects evolved throughout their duration. Projects which consulted and established needs from the outset and developed effective communication and collaboration established the basis for the long-term development of good practice which was embedded in the life of the school. Successful projects assisted individuals in schools and those associated with them in a variety of ways; pupils with the responsibility for their own behaviour; teachers with the development of a wider range of skills and the confidence to use them and parents with understanding and an active role to play. This took time and, over time, needs and priorities changed. Successful projects adapted accordingly.

Consultation and defining need from the outset - Successful projects consulted widely with those involved at the outset and from this established programmes of work which were appropriate for the particular school environment. Some Multi-disciplinary behaviour support teams were able to offer schools a choice of intervention to suit their needs.

Effective communication and collaboration in the school - For projects to be successful good communication, between all concerned parties, senior management, teachers, parents and pupils was vital. Establishing systems within schools which facilitated the identification of pupils causing concern, monitored behaviour and progress and allowed this information to be regularly disseminated was very important. Such systems facilitated the effective co-ordination of targets, the focusing of provision and raised awareness of the difficulties which were common to particular subjects or groups of teachers. Once pupils were identified as giving cause for concern the holding of regular review meetings attended by all concerned facilitated communication.

Effecting long term change - Successful projects changed practice, embedding the changes within the ethos of the school so that they became the norm. This was achieved by working with teachers enabling them to perpetuate and replicate interventions, providing materials to be used after the completion of the project and changing understandings of and attitudes towards pupils at risk of exclusion.

The involvement of the whole school staff - The commitment of the head teacher and the senior management team were crucial to success. They could facilitate the involvement of other staff by making time available; could initiate, support and endorse the introduction of projects as well as the provision of rewards and sanctions; and publicly value the work that was going on. However, this alone was not sufficient for success. The commitment and involvement of the whole staff was needed, including subject teachers and ancillary staff. Projects were successful where they offered a wide range of professional development for all. To be effective training needed to be linked with strategies already employed within the school and involve the full spectrum of staff from management to ancillary staff. If the whole school staff were not involved, projects tended to be seen as separate and exclusively concerned with Special Educational Needs.

Effects on teachers - In the successful projects, there was evidence that teachers exhibited greater confidence in their ability to manage challenging behaviour; felt that they had acquired more strategies for working with difficult pupils; and believed that pupils had been retained in school who might otherwise have been excluded. There was evidence from both types of project that in the region of 80% of pupils supported had been retained in mainstream education. Teachers felt supported by the projects and exhibited an increased understanding of children at risk of exclusion acknowledging their individuality and the need for teaching and classroom management strategies to take account of it.

Developing responsibility in pupils for their own behaviour - Successful projects encouraged pupils to take responsibility for their own behaviour and gave them strategies to assist in self-monitoring and developing self reliance. They were consulted and actively involved in identifying targets for themselves. Their self-development was further enhanced by encouraging them to act as mentors to younger pupils, provide mutual support for each other and adopt the positive self-management strategies they had acquired in a variety of situations.

The role of parents - Successful projects actively involved parents. This was achieved in a variety of ways: providing them with support; giving them opportunities to support each other and their children; involving them in the identification of the needs of their children, decisions concerning strategies for dealing with them and assistance in acquiring the necessary skills; and in some cases providing them with opportunities for accredited training. Successful projects held regular review meetings where parents and pupils were actively involved in decision making.

The development of understanding - Where projects involved all concerned parties, school management, teachers, support staff, parents and pupils, a shared understanding of the issues and a language for discussing and communicating about them evolved.


The research suggested that the setting up of Multi-disciplinary behaviour support teams and In-school centres was not of itself a guarantee of a reduction in exclusions from school. Projects effective in reducing exclusions were implemented with the full commitment of school management; involved the whole school; included parents; and placed the responsibility for their behaviour on pupils. No single intervention, e.g. anger management, peer support, counseling, appeared to be effective unless these criteria were satisfied. Where they were satisfied the projects had cost benefits in the short and long term; promoted a more positive school ethos; and generated change which was likely to continue in the long-term.


BLYTH, E., & MILNER, J. (1993) Exclusion from school: A first step in exclusion from society? Children and Society, 7(3), pp. 255-268.

COHEN, R., HUGHES, M., ASHWORTH, L. & BLAIR, M. (1994) School's out: The family perspective on school exclusion (London: Family Service Units/Barnados).

COMMISSION FOR RACIAL EQUALITY (1985) Birmingham LEA and schools: Referral and suspension of pupils (London: CRE).

CRIME CONCERN (1993) Home affairs committee inquiry into juvenile crime: Memorandum of evidence. (London: HMSO).

DEPARTMENT FOR EDUCATION AND SCIENCE (1985) Education observed: Good teachers (London: HMSO).

DEPARTMENT FOR EDUCATION AND SCIENCE (1989) Discipline in Schools: Report of the Committee of Enquiry chaired by Lord Elton, "The Elton Report" (London: HMSO).

DEPARTMENT FOR EDUCATION (1992) Exclusions: a discussion paper (London: HMSO).

DEPARTMENT FOR EDUCATION (1995) Final report to the Department for Education: National Survey of local education authorities' and provision for, children who are out of school by reason of exclusion or otherwise (London: HMSO).

DEPARTMENT FOR EDUCATION AND EMPLOYMENT, (1997) Permanent Exclusions from School in England 1995/96 (London: HMSO).

DEPARTMENT FOR EDUCATION AND EMPLOYMENT, (1998) Permanent Exclusions from School in England 1996/97 and Exclusion Appeals Lodged by Parents in England 1996/97 (London: HMSO).

DEPARTMENT FOR EDUCATION AND EMPLOYMENT, (1999) Permanent Exclusions from School in England 1997/98 and Exclusion Appeals Lodged by Parents in England 1997/98 (London: HMSO).

GALLOWAY, D (1982) A study of pupils suspended from school, British Journal of Educational Psychology, 52, pp. 205-12.

GALLOWAY, D. (1985a) Schools and Persistent Absentees (Oxford, Pergamon).

GALLOWAY, D. (1985b) Meeting special educational needs in the ordinary school? Or creating them? Maladjustment and Therapeutic Education, 3(iii), pp. 3-10.

GALLOWAY, D, MORTIMORE, P. & TUTT, N. (1989) Enquiry into discipline in schools: University of Lancaster, in: N. Jones (ed) School Management and Pupil Behaviour (London: Falmer Press).

GARNER, P. (1993) Exclusions: the challenge to schools. Support for learning, 8(3), pp. 99-103. .

GILLBORN, D. & GIPPS, C. (1996) Recent Research on the Achievements of Ethnic Minority Pupils (London: OFSTED/HMSO).

HAGELL, A. & NEWBURN, T. (1994) Persistent Young Offenders. (London: Policy Studies Institute).

HALLAM, S. (1996) Improving School Attendance (Oxford: Heinemann).

HALLAM, S. & ROAF, C. (1995) Here Today, Here Tomorrow: Helping Schools to Promote Attendance (London: Calouste Gulbenkian Foundation).

HALLAM, S. & CASTLE, F. (1999) Evaluation of the Behaviour and Discipline Pilot Projects (1996-99) supported under the Standards Fund Programme. Research Report RR163. (London: Department for Education and Employment).

KUMAR, V., (1993) Poverty and inequality in the UK: the effects on children. (London: National Childrens Bureau).

LLOYD-SMITH, M. (1993) Problem behaviour, exclusions and the policy vacuum, Pastoral Care, 11(4), pp. 19-24.

NOTTINGHAM ADVISORY AND INSPECTION SERVICE (1989) Pupil exclusions from Nottingham Secondary schools (Nottingham: Nottinghamshire County council)

OUSTON, J. (1984) Delinquency, family background and educational attainment. British Journal of Criminology, 24, pp. 2-26.

PARSONS, C. BENNS, L. HAILES, J. & HOWLETT, K. (1994) Excluding primary school children (London: Family Policy Studies Centre).

PARSONS, C. (1999) Education, Exclusion and Citizenship (London: Routledge).

STIRLING, M. (1993) How many children are being excluded? British Journal of Special Education, 19(4), pp. 128-130.

UTTING, D. BRIGHT, J. & HENRICSON, C. (1993) Crime and the family: Improving child-rearing and preventing delinquency. (London: Family Policy Studies Centre).

WEST. D.J. (1982). Delinquency: its roots, careers and prospects. (Oxford: Heinemann).

This document was added to the Education-line database on 06 December 2000