Understanding Internal Exclusion in English Secondary Schools: What the Evidence Tells Us

The use of internal exclusion is the subject of debate in the education sector. While intended as a preventive measure to avoid the need to suspend pupils from school, the reality of its implementation and outcomes is complex. Based on recent studies that draw on the perspectives of pupils, teachers, school leaders, and parents/carers, as well as schools’ administrative data, this blog provides an overview of what we currently know about this approach to behaviour management:

What is internal exclusion?

How many schools use internal exclusion, and how many pupils are internally excluded?

Who gets internally excluded and why?

What is the impact of internal exclusion?

Where do we go from here?

At the outset, it is important to acknowledge that managing diverse behavioural needs in secondary school is a complex and demanding task. Every pupil and staff member deserves a classroom environment where learning is not disrupted and in which everyone feels safe. Balancing the individual needs of pupils and protecting the learning rights of the rest of the class is very challenging. Research on internal exclusion is not a criticism of staff intent, but an opportunity to examine whether a ubiquitous behaviour management tool is delivering the outcomes that schools and pupils deserve.

What is internal exclusion?

Internal exclusion is the practice of removing a disruptive pupil from their regular classroom. Instead of a formal suspension from school, they are sent to work in a separate, supervised space. In the current Department for Education guidance on behaviour in schools, internal exclusion is referred to as ‘removal’; it is a significant sanction, only to be applied for serious disciplinary reasons.

Terminology varies significantly between schools, with names like Reset Rooms, Consequence Booths, or Restart Units. In practice, internal exclusion typically involves independent silent learning. Golding’s survey of staff in nearly 100 schools found 78% reported that silence was mandated and that internal exclusion spaces are most commonly supervised by support staff.

Internal exclusion is also used in the United States, where it is referred to as classroom removal or in-school suspension depending on the length of the exclusionary period. Some of the research findings below are from American studies. This is noted where applicable.

How many schools use internal exclusion, and how many pupils are internally excluded?

Research suggests that internal exclusion is a ubiquitous part of the secondary school landscape:

Data from several large studies (e.g., Golding’s study, the DfE’s National Behaviour Survey of over 300 school leaders, and an Education Policy Institute study of nearly 240 school leaders) indicates that internal exclusion is used in 89% to 99% of secondary schools in England.

Our recent study¹ of more than 34,000 pupils in over 120 schools found that approximately 8% of pupils surveyed reported being placed in internal exclusion in a typical week.

A weighted average derived from Golding’s study suggests schools typically refer approximately 26 pupils per week to internal exclusion. The length of time each pupil spends in internal exclusion following a referral varies, but is most commonly a full school day (60%), followed by half a day or less (28%), 2 days (5%), and finally, 3 days or more (5%).

New findings from The Key Group using administrative data (i.e., actual logs of the use of internal exclusion recorded in schools’ information management systems) from over 1,000 secondary schools (representing nearly 900,000 pupils) substantiate this high frequency. Their analysis revealed that 18.4% of the total pupil population in these schools is internally excluded at least once a year. Furthermore, the volume of referrals is very high. In 2024/25, there were approximately 250 internal exclusion referrals per 100 pupils. As only 18 per 100 are ever internally excluded, this equates to an average of 14 referrals per pupil.

A challenge with the above referral data is the lack of standardisation in how schools record them (i.e., what counts as a single referral can vary from school to school²). An alternative is to calculate the number of schools days in which a pupil was internally excluded for at least part of the day³:

	% of pupil population in schools	Number of full days with at least one internal exclusion (range)	Average (mean) number of full days with at least one internal exclusion	Average (mean) % of full school year (190 days) spent fully or partly in internal exclusion
Rare ↓ Recurrent	10.8	1-3	1.6 days	0.8%
	2.8	4-6	4.8 days	2.5%
	2.2	7-12	9 days	4.7%
	2.2	13-39	21.2 days	11%
	0.4	40-78	53 days	27.9%
	0.07	79 or more⁴	94 days	49.5%

These data indicate that for a significant minority of young people, internal exclusion is not a rare “reset” moment but a regular feature of their educational experience.
The distribution of Key Group data on internal exclusion among the pupil population indicates that the c.8% of pupils in the #BeeWell sample likely spend just over 13 days per year partly or fully internally excluded. This translates to being removed from their classroom approximately every 3 weeks on average

Who gets internally excluded and why?

Research has found that certain groups are significantly more likely to be internally excluded, and the reasons for referral sometimes go beyond disruptive behaviour:

Our recent study¹ found that pupils with special educational needs and those from socio-economically deprived neighbourhoods are twice as likely to experience internal exclusion compared to their peers. These findings mirror those of a major study of in-school suspension among nearly 12,000 high-school students in the United States.

Golding’s study found that while “severe disruptive behaviour” is the primary reason (65%), staff also report referrals to internal exclusion for truancy (26%), uniform or hair violations (15%), and lateness (1%).

What is the impact of internal exclusion?

Views on the impact of internal exclusion are divided. A recent Children and Young People’s Mental Health Coalition study found that 90% of more than 60 young people surveyed and 96% of nearly 250 parents/carers surveyed view internal exclusion as harmful or very harmful. Conversely, in Golding’s study, 47% of school staff surveyed believe it improves pupil behaviour, though 50% reported seeing no change or were unsure of its effect.

Some studies have focused on individual outcomes:

Using causal inference techniques, our recent study¹ found that young people who report being internally excluded experience significantly worse school belonging, relationships with school staff, and (for girls) mental wellbeing than their non-excluded peers.

Large-scale longitudinal studies in the United States have revealed that in-school suspension is associated with worse academic outcomes, increased anxiety and depression, and a greater extent of disruptive behaviour and concentration problems over time.

However, what about children who are not internally excluded? All pupils have the right to learn in a classroom free from disruption and in which they feel safe. However, the only major study on “peer spillover effects” published to date found no consistent positive impact on the academic or other outcomes of non-excluded students in schools in the United States. Essentially, the apparent harm caused to internally excluded young people is not currently proven to be offset by gains for their peers.

Where do we go from here?

The evidence suggests that the current system of internal exclusion is not serving its intended purpose and requires change. To move forward, there is a clear need for the DfE to properly regulate and monitor its use. This includes:

Clearer purpose: Redefining internal exclusion to ensure it remains a proportionate, regulated response to serious disruptive behaviour that does not inadvertently marginalise already vulnerable pupils.

Data standardisation and reporting requirements: Standardisation of pupil-level internal exclusion data and a requirement to submit this to the Department for Education (as is already the case for suspensions and permanent exclusions).

Standards for the internal exclusion environment: Setting maximum durations and minimum learning environment standards.

Improved behaviour management support and resource for schools: To enable more comprehensive professional development opportunities for staff and improve earlier intervention capacity, with the aim of reducing the need for and frequency of internal exclusion.

Endnotes and acknowledgements

¹We are very grateful, as always, for the continued participation of our brilliant #BeeWell schools and young people.

²Consider a pupil who is internally excluded for a whole day. Most schools record this as a single referral, but some record it as two (i.e. morning session, afternoon session), and others might record it as five (based on the number of individual lessons missed). Conversely, some schools may internally exclude a pupil for several days and record this as a single referral.

³We are very grateful to colleagues at The Key Group for providing distribution data for the 18.4% of the pupil population who are internally excluded at least once a year.

⁴To note is that to avoid the need to suppress data cells with small numbers of young people (<5), the highest category in the data underpinning this table is ‘110+’, indicating that pupils had at least one period of internal exclusion (whether for all day or part of the day) on at least 110 days in 24/25. For ease of calculation of averages, these have all been treated as 110.