Methodology

Pupil exclusion statistics: methodology

Published:: 25 August 2018

1. Overview of exclusion statistics

1.1 Requirements for schools

The headteacher of a school can exclude a pupil on disciplinary grounds only.

Good discipline in schools is essential to ensure all pupils can benefit from the opportunities provided by education.

The decision to exclude a pupil must be lawful, reasonable and fair. In addition, schools have a statutory duty not to discriminate against pupils on the basis of protected characteristics, such as disability or race.

Schools and LAs must arrange alternative provision from the sixth day of the exclusion of pupils of compulsory school age.

For further information: Statutory guidance: school attendance.

1.2 Uses and users of exclusion statistics and data

The data used to publish exclusions statistics is collected via the school census which is used by a variety of companies and organisations including:

mainstream and specialist media companies
housing websites such as Rightmove and Zoopla
specialist publications such as the good schools guide
data analysis organisations such the Fischer Family Trust
academic research and think tank organisations such as Durham University and the Education Policy Institute
central government organisations such as DfE, Ofsted and other government departments

The published data is also used in answers to parliamentary questions and public enquiries - including those made under the Freedom of Information Act.

1.3 Current publications

DfE publishes annual exclusions data via the Permanent exclusions and suspensions in England National Statistics release in July each year. This publication has previously been called ‘Permanent and fixed-period exclusions in England’.

Links to all exclusions publications can be found in Annex C.

1.4 Key exclusion measures

DfE monitors pupil exclusion levels using key measures based on exclusions and suspensions.

Permanent exclusion refers to a pupil who's excluded and who will not come back to that school (unless the exclusion is overturned).

Suspension refers to a pupil who's excluded from a school for a set period of time.

A suspension can involve a part of the school day and does not have to be for a continuous period.

A pupil may be suspended for one or more fixed-periods up to a maximum of 45 school days in a single academic year.

This total includes exclusions from previous schools covered by the exclusion legislation.

Pupils with one or more suspensions refers to pupils who have at least one suspension across the full academic year.

It includes those with repeated suspensions.

For further information: 3.1 Calculation of exclusion rates.

1.5 Cohort used in exclusion measures

Exclusions information relates to all exclusions reported across the full academic year.

However, exclusion rates are calculated as a proportion of all pupils on roll as at the January Census day of the relevant academic year.

1.6 Published geographical and characteristics breakdowns

DfE routinely publishes pupil exclusions information at national and local authority level including breakdowns by pupil characteristics.

School level data is also released but it should be noted that DfE does not hold individual schools to account based purely on their exclusion figures and does not set exclusion thresholds or targets on the use of exclusion.

Ofsted assesses schools' use of exclusion as part of inspection arrangements.

A list of all routinely published breakdowns is available in Annex D.

1.7 Appeals and reviews relating to exclusion

In addition to permanent exclusions and suspension figures, releases also include information on the number of appeals and reviews relating to permanent exclusions reported by local authorities (LAs).

For further information: 3.3 Appeals and reviews relating to permanent exclusions.

1.8 Underlying data provided alongside publications

From the 2008/09 academic year, each release has been accompanied by underlying data, including national, local authority and school level information.

Alongside the underlying data there's an accompanying document (metadata) which provides further information on the contents of these files.

This data is released under the terms of the Open Government License and is intended to meet at least 3 stars for Open Data.

1.9 Suppression of exclusions data

The Code of Practice for Statistics requires reasonable steps are taken to ensure all published or disseminated statistics produced by DfE protect confidentiality.

In this release figures in underlying data files are not rounded or suppressed, except for the figures on Post Looked After Arrangements, which are treated the same way as other statistics on Children Looked After. In the commentary percentages are rounded for brevity.

Symbols used to identify this in published tables:
0	Zero
c	Small number suppressed to preserve confidentiality
z	Not applicable
:	Not available

In old publications different rounding and suppression conventions were used.

To assist users a full time series of data has been made available within the latest published time series tables and underlying data.

1.10 Other related publications

Pupil exclusions information is also available in the following publications:

A profile of pupil exclusions in England - research report published in February 2012 with in-depth analysis of exclusions
other releases published by DfE:

1.11 Devolved administration statistics on absence

DfE collects and reports on exclusions information from schools in England.

For information for Wales, Scotland and Northern Ireland:

2. National Statistics badging

DfE’s annual exclusions releases have been badged as National Statistics since the 1999/00 academic year publication.

2.1 National Statistics designation

The United Kingdom Statistics Authority (UKSA) designated these statistics as National Statistics in accordance with theStatistics and Registration Service Act 2007 and signifying compliance with the Code of Practice for Statistics.

Designation can be broadly interpreted to mean that the statistics:

meet identified user needs
are well explained and readily accessible
are produced according to sound methods
are managed impartially and objectively in the public interest

Once statistics have been designated as National Statistics it is a statutory requirement that the Code of Practice shall continue to be observed.

Information on improvements made to these statistics to continue their compliance with the Code of Practice are available our Standards for official statistics published by DfE guidance.

Go to top

3. Methodology

3.1 Calculation of exclusion rates

Within published exclusions statistics both the number of exclusions and the rate of exclusion are presented.

Rates of exclusion are a more appropriate measure for comparisons over time as they take into account changes in the overall number of pupils across different academic years.

As pupils can receive more than one suspension, in some cases the rate of exclusion may be above 100%.

The following sections outline the methodology used to calculate exclusion rates - including a change implemented as of the 2013/14 academic year.

3.1.1 2013/14 releases onwards

This change is an improvement as it more accurately reflects the number of pupils who could be excluded, resulting in more accurate exclusion rates.

This change has a minor effect on exclusion rates, but to assist users recalculated time series data have been made available within national published time series tables and underlying data. Consistent data has been published covering academic year 2006/07 onwards.

Figures on numbers of exclusions are not affected by this change.

Permanent exclusion rate = (Number of permanent exclusions recorded across whole academic year / Number of sole and dual main registered pupils on roll as of January census day) X 100

Suspension rate = (Number of suspensions recorded across whole academic year / Number of sole and dual main registered pupils on roll as of January census day) X 100

3.1.2 Releases before 2013/14

Prior to the 2013/14 annual exclusions release, exclusion rates were calculated as a percentage of the total number of solely registered pupils on roll as at January census day.

Formulae for exclusion rates prior to the 2013/14 release are as follows:

Permanent exclusion rate = (Number of permanent exclusions recorded across whole academic year / Number of sole registered pupils on roll as of January census day) X 100

Suspension rate = (Number of suspensions recorded across whole academic year / Number of sole registered pupils on roll as of January census day) X 100

3.2 Permanent exclusion checking exercise 1997/98 to 2009/10

Due to some known under-reporting and some over-reporting of permanent exclusions, a data checking exercise with local authorities was carried out from 1997/98 to 2009/10 to confirm the number of permanent exclusions.

Therefore, for these years, permanent exclusion figures are estimates based on incomplete pupil level data.

The checking exercise took place at local authority level, where the number of permanent exclusions reported by schools via the schools census were compared to the number of permanent exclusions confirmed by each local authority.

This resulted in a pro-rate variable for each reported exclusion, where each exclusion could account for more or less depending on if schools within the local authority were deemed to have over or under reported their exclusions.

Permanent pro rate = Number of permanent exclusions confirmed by local authority during checking exercise for each school / Number of permanent exclusions reported in school census for each type of school

For example, if 5 permanent exclusions were reported for primary schools by a local authority via the school census but 10 permanent exclusions for primary schools were confirmed by the local authority during the checking exercise, each individual reported permanent exclusion was subsequently counted as 2.

This issue has decreased over time and the checking exercise was discontinued following the 2009/10 academic year.

Following this - ie from the 2010/11 publication onwards - published permanent exclusion figures are as reported by schools.

3.3 Appeals and reviews relating to permanent exclusions

Parents (and pupils if aged over 18 years) are able to request a review of a permanent exclusion.

Data on independent review panels regarding permanent exclusions is collected at local authority level via an annual survey. Data was first collected for the 2012/13 academic year.

Information on the old independent appeal panels has been collected by DfE via an annual survey since 1994/95.

However, figures for 1994/95 to 1996/97 are known to be incomplete. Therefore, in the majority of releases, appeals data has been published from 1997/98 onwards.

For further information on independent review panels: Statutory guidance: school exclusion

3.3.1 Reviews of permanent exclusions: 2012/13 onwards

From 1 September 2012, independent appeal panels were replaced with independent review panels - a new system by which the governing body's decision to uphold a permanent exclusion decision can be reviewed.

The role of the independent review panel is to review the governing body's decision not to reinstate a permanently excluded pupil.

In reviewing the decision the panel must consider the interests and circumstances of the excluded pupil, including the circumstances in which the pupil was excluded, and have regard to the interests of other pupils and people working at the school.

Where a panel decides that a governing body's decision is flawed in the light of the principles applicable on an application for judicial review, it can direct the governing body to reconsider its decision.

If the governing body does not subsequently offer to reinstate a pupil, the panel will be expected to order that the school makes an additional payment of £4,000.

This payment will go to the local authority. In other circumstances a panel can recommend, rather than direct, that a governing body reconsiders its decision.

No additional payment is required if reinstatement is not offered following a recommendation to reconsider.

Whether or not a school recognises that a pupil has special educational needs (SEN), all parents (or pupils if aged 18 or over) have the right to request the presence of a SEN expert at an independent review panel.

The SEN expert's role is to provide impartial advice to the panel about how SEN could be relevant to the exclusion.

For example, whether the school acted reasonably in relation to its SEN duties when excluding the pupil.

The data does not show whether any reinstatement offer has been accepted by parents.

3.3.2 Appeals against permanent exclusions: prior to 2012/13

Prior to 1 September 2012, independent appeal panels were used to appeal the decision to permanently exclude a pupil. Panels could either:

uphold the exclusion
overturn the exclusion. In this case they could either:
- direct immediate reinstatement or reinstatement at some future date
- decide that because of exceptional circumstances or other reasons it was not practical to give a direction requiring reinstatement, but that it would otherwise have been appropriate to give such a direction

Go to top

4. Data collection

4.1 The collection process

The school census collects exclusions information two terms in arrears.

For example, the summer school census collects information on exclusions occurring during the autumn term.

Data is collected two terms in arrears to allow time for any independent reviews panels and governing body reconsiderations to be concluded before supplying information to DfE.

Phase	Census	Schedule
State-funded primary,secondary, special schools and pupil referral units	Summer census	Collects autumn term exclusions - occurring between 1 August and 31 December inclusive
	Autumn census	Collects spring term exclusions – occurring between 1 January and Easter Sunday inclusive
	Spring census	Collects summer term exclusions - occurring between Easter Monday to 31 July inclusive

Schools submit their school census returns via DfE's data collection software COLLECT. For further information on the school census - including exclusions: School census: closed collections.

4.2 Background of absence data collection

The following sections outline how exclusions data collection has changed since it was first collected in 1994/95. Consistent data is available from academic year 2006/07 onwards.

4.2.1 School census: 2005/06 to present

In summer 2006 the school census started to collect exclusions data on a termly basis (spring, summer and autumn collections) from maintained secondary schools, 14 city technology colleges and academies to cover the 2005/06 academic year.

The collection amalgamated the previous annual school census and termly exclusion survey, covering both permanent exclusion and suspension (fixed period exclusions) as well as reason for exclusion – all collected at a pupil level.

There is a data gap for suspensions occurring within primary and special schools for the 2005/06 academic year.

This is because these schools did not move to the termly census collection until the following year, and the termly exclusion survey was discontinued in 2004/05.

Permanent exclusions in 2005/06 continued to be collected for primary and special schools via the annual school census.

In the summer 2007 school census the scope of the collection was extended to include primary and special schools, collecting exclusions data for the 2006/07 academic year.

In the spring 2014 school census the scope of the collection was extended again to include termly pupil referral unit (PRU) exclusion data.

Previously this information had been collected annually via the PRU census which ran from 2010 to 2013.

The first full academic year of PRU exclusions data was collected for the 2013/14 academic year.

4.2.2 Termly exclusions survey: 2002/03 to 2004/05

The termly exclusions survey was first carried out in summer 2003, collecting fixed-period exclusions for the first time for the 2003 summer term.

There was evidence of under reporting from schools and this term of data was published as experimental statistics.

In 2003/04 the first full academic year of fixed period exclusions data was collected via the termly exclusion survey. The survey also collected termly permanent exclusions for the first time.

In addition, reason for exclusion was collected via the termly exclusions survey for the first time as of the 2003/04 academic year.

In published statistics for 2003/04 and 2004/05 a combination of annual schools census and termly exclusion survey figures are published.

Overall, permanent exclusion figures are sourced from the annual school census, with term and reason breakdowns coming from the termly exclusion survey.

All fixed-period exclusion figures are sourced from the termly exclusion survey.

The termly exclusions survey was discontinued after the collection of data for the 2004/05 academic year.

4.2.3 Annual school census: 1994/95 to 2001/02

In 1996 permanent exclusions relating to the 1994/95 academic year were first collected by DfE for primary, secondary and special schools via the annual school census. 1995/96 is deemed to be the first reliable year of exclusions data.

In 2002 the annual school census changed to an annual pupil level school census, collecting pupil level exclusions for the 2000/01 academic year.

Permanent exclusions from City Technology Colleges (CTCs) and academies were also first collected for the 2000/01 academic year.

4.3 Data coverage

4.3.1 Coverage for 2006/07 to present

Schools provide, via the school census, individual exclusion data for all pupils receiving a permanent exclusion or suspension (fixed-period exclusion) within the collection period.

Within published exclusions statistics, schools are categorised into the following phases.

Phase	Types of school
State-funded primary	Local authority maintained schools Middle schools as deemed Sponsored academies Converter academies Free schools
State-funded secondary	Local authority maintained schools Middle schools as deemed City technology colleges Sponsored academies Converter academies Free schools University technical colleges Studio schools
Special	Local authority maintained special schools Non-maintained special schools Sponsored academies Converter academies Free schools

4.3.2 Coverage for 2005/06

Exclusions information was collected on a termly basis, at enrolment level, via the school census for the first time to cover the 2005/06 academic year. In this first year of collection, information was collected for secondary schools only.

4.3.3 Coverage from 1994/95 to 2004/05

Permanent exclusion information was collected via the annual school census, at school level, for all pupils in primary, secondary and special schools. From 2000/01 the annual schools census moved to a pupil level collection and permanent exclusions from CTCs and academies were collected for the first time, though these were not included in published statistics until the 2006/07 academic year.

4.4 What exclusions information is collected

Schools are expected to supply the following via their school census returns:

4.4.1 Category of exclusion

For each exclusion, this indicates the type of exclusion:

suspension
permanent

4.4.2 Reason for exclusion or suspension

For each exclusion, schools are required to report the main reason for exclusion.

The following table provides descriptions of reasons for exclusions.

The categories should cover the main reasons for exclusions and the 'other' category should be used sparingly.

The descriptions should be used as a guide and are not intended to be used as a tick list for exclusion decisions.

Exclusion reason	Description
Bullying	Cyber bullying Homophobic bullying Physical Verbal
Damage	Arson Graffiti Vandalism
Drug and alcohol related	Alcohol abuse Drug dealing Inappropriate use of prescribed drugs Possession of illegal drugs Smoking Substance abuse
Persistent disruptive behaviour	Challenging behaviour Disobedience Persistent violation of school rules
Physical assault against adult	Obstruction and jostling Violent behaviour Wounding
Physical assault against pupil	Fighting Obstruction and jostling Violent behaviour Wounding
Racist abuse	Derogatory racist statements Racist bullying Racist graffiti Racist taunting and harassment Swearing that can be attributed to racist characteristics
Sexual misconduct	Lewd behaviour Sexual abuse Sexual assault Sexual bullying Sexual graffiti Sexual harassment
Theft	Selling and dealing in stolen property Stealing from local shops on a school outing Stealing personal property (adult or pupil) Stealing school property
Verbal abuse / threatening behaviour against adult	Aggressive behaviour Carrying an offensive weapon Homophobic abuse and harassment Swearing Threatened violence Verbal intimidation
Verbal abuse / threatening behaviour against pupil	Aggressive behaviour Carrying an offensive weapon Homophobic abuse and harassment Swearing Threatened violence Verbal intimidation

4.4.3 Start date of exclusion/suspension

Each exclusion is reported with a start date. This start date should reflect the date the exclusion starts - ie the date the pupil was asked to leave the school by the headteacher.

4.4.4 Length of exclusion/suspension in sessions

For each suspension , the actual number of sessions for which the exclusion is in effect is recorded. Each session counts as half a school day. Exclusion only covers the number of sessions the pupil is expected to be at the school.

A suspension can be discontinuous which means there can be days between the start and finish dates of the exclusion that are not included as part of the exclusion.

Where a pupil is dually registered, exclusion sessions are not recorded for the days when they would have been attending the other establishment.

For example, a pupil is excluded for six sessions starting on a Monday. However, the pupil is registered with another provider on the Wednesday from which they have not been excluded.

Therefore, the pupil is excluded from the school for sessions on Monday, Tuesday and Thursday. The pupil's attendance on the Wednesday would be recorded separately by the other provider.

Exclusions sessions are not recorded for the days when the pupil is attending an approved educational activity.

For example, a year 10 pupil has been excluded for six sessions starting on a Wednesday. However, the pupil attends work experience on a Thursday from which they're not excluded.

The pupil in this situation would be excluded from the school for sessions on Wednesday, Friday and Monday.

Days on which there's a planned school closure do not count towards an exclusion.

For example, a pupil has been excluded from a school for four sessions starting on a Friday but the school is planned to be closed on the Monday for an INSET day.

Therefore the pupil's exclusion should include the two sessions on the following Tuesday, rather than the Monday.

4.4.5 In care and SEN indicators

For each exclusion, the pupil's in care status and SEN provision at the time of exclusion is recorded. This may be different to their current provision.

4.5 What exclusion information is not collected

DfE does not collect the following information:

4.5.1 Permanent exclusions overturned at governing body consideration stage and at governing body reconsideration stage

DfE does not collect information on permanent exclusions which were overturned at the governing body consideration stage and at governing body reconsideration stage.

Those do not count as exclusions and the exclusion decision has effectively been overturned.

By collecting the data two terms after that in which the exclusion started, it's anticipated that all the exclusion reviews and reconsiderations procedures would have been completed by the time the data is collected and so only exclusions which have been upheld will be included in school census returns.

4.5.2 Specific details of exclusion incident

Exclusions information is collected with a corresponding reason for exclusion.

For further information on reasons collected: 4.4.2 Reason for exclusion.

Specific details of the incident resulting in the exclusion are not collected via the school census.

For example, the location of the incident, members involved in the incident, if weapons were used etc.

4.5.3 Exclusions data for independent schools

DfE does not collect exclusions data from independent and nursery schools.

Go to top

5. Data processing

The following sections outline how exclusions data goes from raw school census returns to a finalised dataset used to prepare National Statistics releases.

5.1 Data linking

Exclusions data is collected at exclusion level and therefore aggregated to annual figures by simply summing the number of termly exclusions together.

Each exclusion is reported alongside a unique pupil identifier.

Pupil level exclusions data is derived by linking exclusion records collected in each termly census to unique pupil identifiers, giving the total number of exclusions received by each pupil across a given period (such as term or academic year).

Exclusions are then linked to the relevant pupil's characteristics data (also collected via the school census) at the term of the exclusion.

Where characteristics data is missing, information collected in the census term reporting the exclusion is used.

If, again, data is also missing here a range of other relevant censuses will be used to attempt to find a characteristic (except for ethnicity as this information is only collected in the spring school census) as per Table 4: Characteristics linking for exclusion data.

Exclusion period	Source of exclusion data	First choice for characteristic (same as exclusion period)	Second choice for characteristic (following census)
Autumn term	Summer school census	Autumn school census	Summer school census
Spring term	Autumn school census	Spring school census	Autumn school census
Summer term	Spring school census	Summer school census	Spring school census

5.2 Data removed

To prepare the exclusions data for publication routine checks are applied to the linked data and where necessary data is removed as follows.

Lunch time exclusions are not reported in published exclusions statistics.

Any exclusions from sponsored academies with a start date before the academy opened are removed before analysis. They're removed because when a sponsored academy opens it's a new school and not responsible for exclusions accrued within the predecessor school.

Year-on-year comparisons of local authority (LA) data may be affected by schools converting to academies.

5.3 Variables added

Most of the variables needed to produce national level exclusions statistics can be calculated directly from the fields collected via the school census.

However, to prepare the data for publication the following variables are also added to the dataset.

5.3.1 Pupils with one or more suspension

Pupils with one or more suspension across the academic year are identified as such and are then used in the one or more suspension measure.

5.3.2 School and academy type

School type information - including school type, academy type and academy open date - is added to our underlying data prior to producing any analysis to allow us to produce the school- and academy-type breakdowns in our statistical releases.

These variables are derived using a combination of Get Information About Schools data and the academies and academy projects in development data.

Within exclusion National Statistics, academies are only indicated as academies if they were open as of 12 September.

Go to top

6. Data quality

Data quality

The following should be taken into account when reviewing published exclusions statistics.

The exclusions information reported in published releases is based on data returned by schools as part of the school census.

It does not include data which has been submitted by local authorities (LAs) or schools outside of the school census.

It's a school's responsibility to record exclusions data correctly in their school census return.

As exclusions information is collected two terms in arrears, when a school closes and is replaced by a sponsored academy data is not available for the last two terms that the predecessor school was open.

Caution is recommended when interpreting the data for Traveller of Irish Heritage and Gypsy/Roma children due to small population sizes.

6.1.1 Data checking

Prior to the release of exclusion National Statistics the raw data goes through a thorough quality assurance process to ensure it is fit for purpose.

The main areas of this checking are:

matching derived data back to raw sources
comparing year-on-year trends across different geographies and pupil groups to look for and investigate substantial changes
comparing figures and trends to other published sources
cross-checking data within the release to ensure totals are consistent

Go to top

Annexes

Annex A - Calculations

The following calculations are used to produce absence National Statistics:

Permanent exclusion or suspension rate

(Number of permanent exclusions/suspensions recorded across the whole academic year / Number of sole and dual main registered pupils on roll as at January census data) X 100

Pupils with one or more suspension rate

(Number of pupils with one of more suspension in full academic year / Number of sole and dual main registered pupils on roll as at January census day) X 100

Proportion of all exclusions

(Number of exclusions / Overall total number of exclusions) X 100

Average exclusions per day

Number of exclusions / 190

Average suspensions per pupil

Number of suspensions / Number of pupils with one or more suspensions

Average length of suspension

Total number of suspension sessions / Total number of suspensions

Go to top

Annex B - Exclusion by reason codes

The following codes are used by schools when taking registers, further details are available in the Department’s School Attendance advice:

Code	Reason
PP	Physical assault against a pupil
PA	Physical assault against an adult
VP	Verbal abuse / threatening behaviour against a pupil
VA	Verbal abuse / threatening behaviour against a adult
BU	Bullying
RA	Racist abuse
SM	Sexual misconduct
DA	Drug and alcohol related
DM	Damage
TH	Theft
DB	Persistent disruptive behaviour
OT	Other

Go to top

Annex C - Links to pupil exclusions statistics and data

Published exclusions statistics are available at the following links

Go to top

Annex D - Standard breakdowns

The following breakdowns are currently available in published exclusions statistics.

National level
Exclusion numbers and rates		For permanent exclusion and suspensions broken down by school type. Time series
Exclusions information by characteristic		Permanent exclusion and suspensions information by age, gender, free school meal eligibility, national curriculum year group, SEN provision, ethnic group and level of deprivation
Duration of fixed-period exclusions		The duration of suspensions, broken down by school type
Number of fixed-period exclusions		The number of suspensions broken down by school type
Exclusions by reason		The number and proportion of permanent exclusion and suspensions broken down by school type
Exclusion review panels		Information on the number of exclusion review panels in maintained primary, secondary, special schools and academies

Local authority level
Exclusion numbers and rates		For permanent exclusion and suspensions as well as those with one or more suspensions broken down by school type
Exclusions by reason		For suspensions and permanent exclusions
Exclusions by ethnicity		For suspensions and permanent exclusions. A school type split is provided for suspensions
School level
Exclusion numbers and rates		For permanent exclusion and suspensions as well as those with one or more suspension broken down by school type

Go to top