DFE: The unequal representation of ethnic minorities for different types of special educational needs: Extent, causes and consequences

Lead Research Organisation: University of Oxford
Department Name: Education

Abstract

Representative research in the US has established that Black students are substantially more likely to be identified with Special Educational Needs (SEN) than the majority White group. The only nationally representative studies in England in the last 25 years (Strand & Lindsay, 2009; 2012) report a more nuanced picture across different types of disability, with for example Black Caribbean pupils being twice as likely as White British pupils to be identified with Behavioural and Emotional and Social Difficulties, but Asian pupils being only half as likely as White British pupils to be identified with Autistic Spectrum Disorders. There are concerns this over-representation reflects social bias in school identification processes and results in inappropriate special education provision, while under-representation reflects cultural or economic barriers to accessing services and results in unmet need. These results have important implications for Government, Local Authorities and schools as the Equalities Act (2010) places a general duty on all public bodies to eliminate unlawful discrimination and advance equal opportunity.

However relatively little is known about the factors which contribute to ethnic disproportionality, largely because of methodological limitations. Studies in the UK have often been small scale and unrepresentative while studies in the US are typically large but based on aggregate data that do not allow the sophisticated analyses possible only with student level data, or based on longitudinal surveys but with samples that are small relative to the incidence of special education. Methodologically, this project will provide three innovations. First it will use national census data on over 6 million students aged 5-16 years to build a nuanced picture across all ethnic groups for every category of SEN. Second it will be the first to utilise multilevel statistical models to estimate the relative influences of student, school and local authority (LA) factors on SEN identification. Third it will move beyond static models of SEN identification at a single point in time to build a dynamic model of the development of SEN by tracking two cohorts, each of 575,000 students, as they progress over time, one through primary school (from Reception to Y6) and the other through secondary school (from Y6 to Y11).

Theoretically, the research will allow the evaluation of competing hypothesises about the possible causes of unequal representation, for example by determining whether socio-economic factors such as poverty and economic deprivation can account for the ethnic disproportionality in different types of SEN. The impact of other factors including age, gender, early educational attainment, attendance and behaviour will be explored, as well as school and LA factors such as size, type and social and ethnic composition. The project addresses published DFE research priorities (2014) around the accurate identification of SEN, the ESRC strategic priority to promote a Vibrant and Fair Society and the Equalities Act (2010) duty to eliminate unlawful discrimination. It is timely given the introduction of a new SEN and Disability Code of Practice in September 2014 and the fact that ethnic minority students now constitute over one-quarter (28%) of the age 5-16 school population.

The project will have wide impact mediated by a strong advisory group comprising practitioners and policy advisers as well as notable academics in the fields of education, psychology and sociology. The outputs will be tailored to maximise impact by: (i) communicating through short accessible summaries and reports; (ii) targeting outputs through practitioner and professional networks; (iii) producing a comprehensive Local Authority (LA) data pack identifying levels of disproportionality for all 150 LAs in England allowing benchmarking and comparative analysis, and; (iv) holding a user-focussed national dissemination event.

Planned Impact

Who will benefit from this research?

The proposed project has a great deal of potential for impact through supporting the DFE to develop more effective policies around appropriate and accurate identification of SEN, supporting Local Authorities in meeting their duties under the Equalities Act (2010) and by informing on-going public discussion on the factors influencing the identification of SEN for different social and ethnic groups. This includes:

* Policy makers at national level in education focussed on supporting appropriate and proportionate identification of SEN (e.g. Department for Education), equity in educational outcomes (e.g. Equalities and Human Rights Commission) and other interest groups (National Association for Special Educational Needs, National Association for Language Development in the Curriculum, British Psychological Society and others).

* Local Authority Services involved in the identification and support of young people with SEN and their families (e.g. Principal Educational Psychologists, SEN Support Services) and those involved in local planning and provision of different kinds of health and educational support (e.g. the provision of Sensory Impairment Services in area with large Pakistani and Bangladeshi populations).

* Other LA Services such as Ethnic Minority Achievement Services (EMAS) and Research and Statistics Services with a focus on raising standards and addressing equity in educational outcomes.

* School leaders, SEN Co-ordinators (SENCOs) and teaching staff through a fuller understanding of the role of school norms and the potential for social biases in the processes of SEN identification.

* Parents and children with SEN through longer-term improvements in the equitability of the process of SEN identification.

How will they benefit from this research?

An important part of the project will be to tailor and modify the evidence found into easily usable forms for the different audiences at the levels of national policymakers, local authorities, schools and families. We have recruited an Expert Advisory Group (EAG) of professional users, policy advisers and notable academics in the area from the multi-disciplinary fields of education, sociology and psychology who are already involved in shaping the research and its direction (see attached eight letters of support from Directors of Education and other senior managers). This group will co-produce and focus our outputs for each of our target audiences. These forms will include evidence summaries, press releases, media pieces as well as conferences and other project publications (e.g. see academic beneficiaries).

A particularly significant output from the project is the LA Data Pack. The Equalities Act (2010) place a general duty on all public bodies to eliminate unlawful discrimination and advance equality of opportunity. However while LAs have the statutory responsibility to identify and address SEN in their areas there are no statutory requirements on LAs with regard to collecting, examining and reporting on disproportionality in SEN, no published indicators or accountability measures, and so no evidence as to the extent of ethnic disproportionality at local level. There is a gap between the general duties of the Equalities Act (2010) and the specific evidence needed to identify and address issues of ethnic disproportionality. The LA Data Pack will be aimed at LA service Leads and will allow LAs to identify the extent of disproportionality for each type of SEN in their area, to examine and question this in the light of their knowledge of local factors, to compare raw and adjusted figures after taking account of socio-economic and other factors, and to interrogate data across LAs to provide a basis for comparison, benchmarking and focussed questions.

What will be done to ensure that they benefit from this research?
This is fully detailed in the 'Pathways to Impact' attachment.

Publications

10 25 50

 
Description Compared to White British pupils, there is marked disproportionality for the following ethnic groups and Special Educational Needs (SEN):
• Black Caribbean and Pakistani pupils are over-represented for Moderate Learning Difficulties (MLD), while Indian and Chinese pupils are under-represented;
• Black Caribbean and Mixed White & Black Caribbean pupils are substantially over-represented for Social Emotional and Mental Health (SEMH) needs;
• All Asian Groups (Indian, Pakistani, Bangladeshi and Other Asian) are substantially under-represented for SEMH and for Autistic Spectrum Disorders (ASD).

The over-representation for MLD can be accounted for by socio-economic factors, but the ethnic disproportionalities for SEMH and ASD remain substantial even after pupil background controls for age, sex and socio-economic deprivation. This is not because of the limited socio-economic measures available in the NPD, as we found the same results after control for parental social class, parental education and family income using the Second Longitudinal Study of Young People in England (LSYPE2).

Prior attainment/development also does not account for the ethnic disproportionality in SEMH and ASD. Literacy and mathematics measures from the Early Years Foundation Stage Profile at age 5 were strongly predictive of the likelihood of subsequent identification of MLD, and the Personal, Social and Emotional Development (PSED) measure was highly predictive of subsequent identification of SEMH and ASD. However, this did not remove the ethnic disproportionality for SEMH and ASD which remained substantial. The findings for the secondary cohort, accounting for age 11 English and mathematics attainment on-entry to secondary school, led to the same conclusion.

Local Authorities (LAs) account for little (2%-6%) of the variation in the identification of SEN. Patterns of disproportionality vary little in direction across LAs e.g. of 113 LAs with sufficient data for SEMH calculations, 84 show over-representation of Black Caribbean/Mixed White & Black Caribbean pupils, none show under-representation. Similarly, of 94 LAs with sufficient data for ASD calculations, 79 show under-representation of Asian pupils, only three show over-representation. This consistency suggests that variation in LA policy and practice plays a limited role in the over-representation of Black Caribbean/Mixed White & Black Caribbean pupils with SEMH or the under-representation of Asian pupils with ASD.

There is variation between schools in the frequency with which they identify SEN, but schools play a limited role in accounting for ethnic disproportionality, with the notable exception of identification of SEMH in secondary school. In null models, around one-fifth of the variance in MLD is between schools (22%-25%) somewhat less for SEMH (13%-15%) and much less for ASD (11%-12%). Some of this variation can be explained by the socio-economic composition of the pupil intake, and by factors like school size and type (e.g. Grammar schools had very few SEN pupils). However, differences between schools played little role in accounting for ethnic disproportionality, with the notable exception of SEMH in secondary schools. Differences between secondary schools account for a substantial part of the over-representation of Black Caribbean and Mixed White and Black Caribbean pupils with SEMH. i.e. their over-representation occurs much more in some secondary schools than it does in others.

Longitudinal studies, even with large representative samples, can often be under-powered to detect relatively low incidence outcomes like type of SEN for ethnic minority groups. Results from sample studies need to be interpreted with caution and more population level studies, like those reported here, are required.

We outline the detailed findings in the Executive Summary (see URL given in the next section) and the implications for policy and practice in a subsequent section.
Exploitation Route Five key implications for policy/practice:

1. LAs, multi-academy trusts (MATs) and schools must have due regard to the Public Sector Equality Duty (PSED) requirements, and should monitor ethnic disproportionality in the identification of SEN. LAs or MATs with high levels of disproportionality should further investigate practices in their areas/trusts. Schools should identify priorities for the partnerships within which they work, so they can pool resources and develop effective responses.

2. OFSTED should incorporate data on ethnic disproportionality into pre-inspection reports for LA SEND inspections, and include the issue of ethnic disproportionality within the LA SEND inspection. The PI will take this forward by offering briefing sessions to staff within the DFE and OFSTED to consider how data could be presented and what questions could be asked.

3. The original detailed guidance on data collection by type of SEN (DFE, 2005) no longer exists following the new SEND Code of Practice. The DFE should consider new guidance on definitions and criteria for defining different types of SEN.

4. Secondary schools in particular should review their processes around the identification of SEMH needs, given variability between schools is a strong component of ethnic disproportionality in this domain. Teachers should also be aware of the significant over-identification of summer born pupils for MLD and to consider carefully whether they are making sufficient allowance for the age of the child when forming their judgements.

5. LAs and schools need to raise awareness of ASD among Asian communities, improve outreach and review the extent to which the services are configured appropriately for access by ethnic minority groups.

The PI will seek to advance all of the above through dissemination of the research findings through the media, focussed communications using channels such as DFE, NASEN etc), and continued engagement with policymakers.
Sectors Education,Healthcare

URL http://www.education.ox.ac.uk/research/the-unequal-representation-of-ethnic-minorities-in-special-education-in-england-extent-causes-and-consequences
 
Description During the project period The Advisory and User Engagement Group for the project brought together a wide range of experts, including the Special Adviser for Special Educational Needs and Disability (SEND) from the England Department for Education (DFE), and Principal Educational Psychologists, leaders of SEN Services and leaders of Ethnic Minority Achievement (EMA) Services from several Local Authorities (LAs), including both large counties (Cambridgeshire and Warwickshire) and urban contexts (Lambeth, Merton, Wandsworth and Wolverhampton). Also represented were members from a range of stakeholder groups and acknowledged academic experts in the area of SEN (Professors Geoff Lindsay, Kathy Sylva, Judy Sebba and Harry Daniels). The group met twice and was a valuable two-way exchange, both in the members taking learning from the project back into their organisations, and also in the project team updating and honing their analyses in the light of the group feedback. At the launch stage and dissemination An expert seminar was held at St. Cross College, Oxford in December 2018 to bring together 20 key stakeholders to consider the implication of the project findings for policy, for practice and for further research. This included the Department for Education (DFE) lead for SEND research, the national lead for equalities from the Office for Standards in Education (OFSTED), the Equalities and Human Rights Commission (EHRC), the Chair of the National Association of Principal Educational Psychologists (NAPEP) and two other Principal Educational Psychologists, A representative of the National Association for Special Educational Needs (NASEN), the President of the London Educational Research Network (LERN) and several independent researchers/stakeholders. We believe this maximised the impact of the projects findings with key users. The launch (see Press Release in Portfolio) was covered in the press and media. For example the story was covered by the BBC News (Black pupils' schooling 'dumbed down over special needs' 15 February, 2019) and in Children and Young People Now (Black pupils 'more likely to be diagnosed with special needs'. 15 February 2019). Professor Strand was interviewed by Andrew Pierce on LBC Radio (15/02/19) (listen here, scroll down to 'Related News'). The Department for Education (DFE) included the key findings from the research in their SEND newsletter sent to all primary, secondary and special schools in April 2019 (copy available). The DFE also requested that the authors present the results to the National Pupil Database (NPD) User Group, which we did both in September 2018 and in February 2020. The National Association for Special Educational Needs (NASEN) included an article on the research in their membership magazine Connect (Sep 2019, copy available). NASEN is a charitable membership organisation that supports education practitioners by providing relevant training, resources, advice and information to enable staff to meet the needs of all pupils including pupils with learning differences (www.nasen.org.uk). In terms of academic outputs, there is one paper under review, one in preparation, and seven papers presented at conferences including the American Educational Research Association (AERA), the European Conference on Educational Research (ECER) and the British Educational Research Association (BERA). Subsequent use and impact The project website received 2,128 views in 2019. The LA report page was visited 580 times with 342 downloads of individual LA reports, the highest being for Manchester (18), Hackney (17), Wandsworth (17), Ealing (12) and Islington (10). In 2020 the project website received a further 2,006 views and the LA reports page was visited 350 times. Bristol, Hackney, Oxfordshire, Salford, Sheffield and Southwark LAs all contacted us between April-June 2019. All were using the LA reports produced by the project, and had been asked by their Director of Childrens' Services to analyse their latest data in the same way. These were just the LAs who contacted us because they were seeking clarification on methodology, the level of use of the data was much wider than just these LAs. Attached are letters from Hackney LA and Sheffield LA detailing how and why they used our reports and analysis. Ben Peters, Senior Education Policy and Data Officer, Education & Youth, Greater London Authority (GLA) wanted to produce an analysis of the data at a London (regional) level of aggregation, and Prof. Strand worked with him to provide this. The Gypsy Roma Traveller (GRT) Northern LA Co-ordinators Networking group used the report in their response to the OFSTED consultation on the new Educational Inspection Framework (The group includes co-ordinators from the following LAs: Blackburn, Bury, Bradford, Chester, Cheshire West, Darlington, Doncaster, Kirklees, Lancashire, North Yorkshire, Rochdale, St. Helens, Salford and Sheffield). The research informed Prof. Strand's own submission to national consultations by the DFE on the "Think Autism Strategy" (April 2019) and by OFSTED on the new Education Inspection Framework (April 2019). The DFE appointed Professor Strand to the Advisory Group for their SEND Futures Longitudinal Study in May 2019. The SEND Futures Longitudinal Study aims to provide robust evidence on response and retention among the SEND population in order to inform a potential future longitudinal study of children with SEND in England. It will also provide timely policy-relevant evidence on the outcomes and experiences of education of children with SEND. Professor Strand gave oral evidence on two occasions to the Government Commission on Racial and Ethnic Disparities (CRED) on 18/12/20 and 17/01/21 on the subject of ethnic inequalities in educational outcomes, including the identification of special educational needs (SEN). The Race Disparity Unit at the Cabinet Office commissioned further research from Professor Strand on ethnic, socio-economic and gender inequalities in educational achievement (£19,725).
Sector Education
Impact Types Societal

 
Description Letter from Hackney Local Authority
Geographic Reach Local/Municipal/Regional 
Policy Influence Type Implementation circular/rapid advice/letter to e.g. Ministry of Health
Impact 1st May 2019 To Professor Steve Strand Steve, I am writing to thank you for your recent co-operation in sharing the findings of your research project into the unequal representation of ethnic minorities in special education in England: extent causes and consequences. The report yourself and your team produced on this subject late last year was brought to my attention by Annie Gammon, Director of Education here at Hackney Learning Trust (Education Dept. for London Borough of Hackney). This was of great interest to us here at HLT as both SEND and vulnerable pupils, as defined by contextual factors such as ethnicity, are two of our main priorities here, as we move beyond an organisational priority of raising standards and achievement for all pupils in Hackney. The LA-specific data you passed to me has enabled a detailed analysis into where our potential areas of focus should be as we utilise your research as key management information for senior-level decision makers into how we direct our activity and allocate our resources. Again, many thanks for your help with this - your work and findings are immensely useful to us. Best regards Simon Utting Head of Management Information, Systems and Analysis (MISA) Hackney Learning Trust
 
Description Presented oral evidence twice to the Government Commission on Racial and Ethnic Disparities (CRED) and was awarded a direct commission for further research and analysis (£19,725)
Geographic Reach National 
Policy Influence Type Participation in a guidance/advisory committee
URL https://www.gov.uk/government/organisations/commission-on-race-and-ethnic-disparities
 
Description Project Website use
Geographic Reach National 
Policy Influence Type Contribution to new or Improved professional practice
 
Description Race Disparity Unit, Cabinet Office (direct commission)
Amount £19,725 (GBP)
Organisation Cabinet Office 
Sector Public
Country United Kingdom
Start 11/2020 
End 02/2021
 
Description Advisory and User Engagement Group 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact 15/03/18 - Meeting of the project Advisory & User Engagement Group. Included the Special Adviser for SEND from the England Department for Education; the Principal Educational Psychologist (PEP) from three Local Authorities (LA) (Warwickshire, Wandsworth and Cambridgeshire); the Head of the Cambridgeshire Race Equality and Diversity Service; the Learning & Teaching Adviser (Equalities) Merton LA; the Head of Research, Lambeth LA; the Ethic Minority Achievement Adviser, Wandsworth LA; the Advisory Teacher for EAL Wolverhampton LA; the Special School Improvement Adviser, Wolverhampton LA; two Professors of Education from Oxford University and Warwick University.
Year(s) Of Engagement Activity 2018
 
Description Advisory and User Engagement Group 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact 27/03/17 - Meeting of the project Advisory & User Engagement Group. Included the Special Adviser for SEND from the England Department for Education; the Principal Educational Psychologist (PEP) from three Local Authorities (LA) (Warwickshire, Wandsworth and Cambridgeshire); the Head of the Cambridgeshire Race Equality and Diversity Service; the Learning & Teaching Adviser (Equalities) Merton LA; the Head of Research, Lambeth LA; the Ethic Minority Achievement Adviser, Wandsworth LA; the Advisory Teacher for EAL Wolverhampton LA; the Special School Improvement Adviser, Wolverhampton LA; two Professors of Education from Oxford University and Warwick University.
Year(s) Of Engagement Activity 2017
 
Description Expert Seminar at St. Cross College, Oxford 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Policymakers/politicians
Results and Impact This was an Expert Seminar funded by the project and held at St. Cross College, Oxford on 6 December 2018. It brought together 20 key stakeholders to consider the implication of the project findings for policy, for practice and for further research. There were Equalities and Human Rights Commission, OFSTED (national leads for SEND and for Equalities), Department for Education, The Chair of the National Association of Principal Educational Psychologist, two Principal Educational Psychologists, the President of the Local Authority (LA) London Educational Research Network, representative of LA Ethnic Minority Achievement Services, School Improvement Services and Research teams, independent researchers and leading Professors in the field (from Oxford, Warwick and Exeter). The meeting helped shape the recommendations of the report, and suggested opportunities for engagement with the DFE and OFSTED to build ethnic disproportionality into the OFSTED LA inspection framework.
Year(s) Of Engagement Activity 2018
 
Description Oral evidence to the England Government Commission on Ethnic and Racial Disparities (CRED) 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Other audiences
Results and Impact I included the research finding in my oral evidence session with the Government Commission on Ethnic and Racial Disparities (CRED) on 18 December 2020. The Commission are supported by the Cabinet Office who organised the meeting. I reported in the context of ethnic differences in on educational outcomes (including identification of SEN).
Year(s) Of Engagement Activity 2020
URL https://www.gov.uk/government/organisations/commission-on-race-and-ethnic-disparities
 
Description Paper presented to the National Pupil Database (NPD) User Group Conference, Abbey Centre, Great Smith Street, London, 23 January 2020. 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact Lindorff, A., & Strand, S. (2020). Ethnic Disproportionality in Special Educational Needs Identification In England: A Multilevel Analysis. Paper presented to the National Pupil Database User Group Conference, Abbey Centre, Great Smith Street, London, 23 January 2020.
Year(s) Of Engagement Activity 2020
 
Description Presentation of Research findings - ethnicity and SEMH Difficulties 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Postgraduate students
Results and Impact Presentation to the University of Durham, School of Education, Public Research Seminar Series, 13/03/19
Year(s) Of Engagement Activity 2019
 
Description Research presentation at the University of Southampton 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Postgraduate students
Results and Impact This was a presentation made in the Public Research Seminars of the University of Southampton on 1 December 2020
Year(s) Of Engagement Activity 2020