Skip to main content

0800 917 1000 Open weekdays 9am to 5pm

IICSA Independent Inquiry into Child Sexual Abuse

Sexual Abuse of Children in Custodial Institutions: 2009-2017 Investigation Report

C.3: The Inquiry’s work on prevalence


13. The Inquiry has carried out extensive work on the prevalence issue:

  • Reviewing the overall prevalence figures for different types of institution in the HMIP Children in Custody series reports.[1]
  • Reviewing HMIP and Ofsted surveys at each institution between 1 January 2009 and 31 December 2017, and other information provided by HMIP.
  • Using its powers under Rule 9 of the Inquiry Rules 2006, obtaining information about disclosures of child sexual abuse received by custodial institutions, local authorities, police forces and the bodies who inspect or visit these institutions, covering incidents said to have occurred between 1 January 2009 and 31 December 2017.[2]

14. This material has now been analysed by the Inquiry and, for the first time, provides a clearer picture of: (i) the total number of reported incidents, including by category; (ii) the rate of incidents per year; (iii) the different types of abuse; and (iv) the alleged perpetrators, each of which can be broken down by type of institution.[3]

15. This analysis, together with the underlying material, was disclosed to the core participants and their observations sought. No core participants disagreed with the analysis.[4]


16. We are aware of the limitations on the data described above. However, despite those limitations, we can conclude that there has been even more sexual abuse of children than is disclosed by the evidence we have seen.

Total number of allegations

17. For the period between 1 January 2009 and 31 December 2017, the Inquiry’s analysis shows:

  • There were 1,070[5] alleged incidents[6] of child sexual abuse within custodial institutions.
  • Of these, 578 alleged incidents were described in terms equating to sexual assault or rape.
  • Around 1,109[7] children were alleged victims of sexual abuse.
  • There were more alleged incidents per year in 2016 and 2017 than in any previous reporting year. Between 2009 and 2015, there were never more than 114 incidents per year. However, in 2016 and 2017 there were 203 and 205 alleged incidents respectively.
  • For reasons that were not always clear, the vast majority of these allegations were not found to be substantiated.
  • In 10 cases where the alleged perpetrator was a member of staff, he or she was issued with a warning, reprimand or had a letter placed on his or her file. In a further nine cases, the alleged perpetrator was dismissed for reasons connected to the alleged incidents. Some other action was taken against the alleged perpetrator in another 120 cases[8] (including suspension, monitoring, words of advice, supervision or being moved to a different location).[9]
  • Only nine alleged incidents resulted in criminal charges, of which four resulted in conviction.[10]

18. We have identified a significant number of complaints of child sexual abuse in custody which have not otherwise come to light. For example, there is no published survey of children at Feltham in 2016, and the January 2017 survey[11] revealed no sexual abuse. However, the evidence provided to the Inquiry refers to five allegations in 2016, and a further allegation in January 2017 before the survey. Again, despite alleged incidents of sexual abuse being recorded elsewhere prior to January 2017, no child took the opportunity to refer to it in the survey. Similarly, in surveys undertaken between October 2012 and February 2017 at Medway, it was in only one year – 2014[12] – that any sexual abuse was reported and only by 2 percent of respondents, which equates to one child. This compares with 44 alleged incidents revealed in the Inquiry’s evidence during the same period of time. Finally, in the October 2016[13] survey of Rainsbrook STC, only one child reported having been sexually abused, whereas the Inquiry’s evidence has revealed 20 alleged incidents of sexual abuse in 2016 prior to the survey.[14]

19. In all of the surveys referred to in which respondents reported no sexual abuse, or where only one respondent reported such abuse, survey questionnaires were offered to the vast majority or all of the children within the institution at that time. While some children may choose not to return questionnaires and some may leave before the next survey is undertaken, it seems children who do complete the surveys are not taking up the opportunity to report sexual abuse that may have been recorded elsewhere.

Trends by institution

20. There were more alleged incidents in 2016 and also 2017 than in any previous period. When the figures are broken down by institution type, on the data currently available, incidents in STCs account for most of the increase: in 2015, there were only 19 alleged incidents but this figure rose to 79 in 2016 and 110 in 2017. This increase is of more concern when the population size of the STC is taken into account; when incidents are expressed as a percentage of the average under-18 population for the year, the increase from 2015 to 2016 rises from 9.1 percent to 54.9 percent.[15]

Types of abuse

21. Based on the information to the Inquiry, the types of abuse alleged for each case[16] of alleged child sexual abuse can be broken down as follows:

Table 1: Type of abuse by institution and in total

Type of institution
Type of abuse (main event) YOI STC SCH* SCH W Other Unknown Total
Rape 28 5 4 3   1 41
Attempted rape 2 1         3
Sexual assault 219 132 91 27 2   471
Attempted sexual assault 5 7 2 1     15
Exposure 12 30 9 4     55
Sexual acts between detainees possibly consensual  8 19 20 3   2 52
Sexual/inappropriate relationship between staff and detainee  15 11 3 4   1 34
Threat of sexual abuse 10 1 2     2 15
Other 96 59 44 10   2 211
Insufficient detail 45 32 14 1 1   93
Total 440 297 189 53 3 8 990§

* ‘SCH’ refers to SCHs holding children on both justice and welfare placements, whereas ‘SCH W’ refers to SCHs only holding children on welfare placements.

† The ‘Other’ column includes the three incidents of abuse which occurred outside a custodial institution, referred to above.

‡ This includes, for example, sexual comments, voyeurism, grooming behaviour and sexual gestures.

§ This refers to the 990 cases of abuses: see explanation for ‘case’.


22. Based on the information to the Inquiry, the perpetrators for each case of alleged child sexual abuse can be summarised as follows:

Table 2: Type of abuse by perpetrator and in total

Type of institution
Perpetrator YOI STC SCH* SCH W Other Unknown Total
Detainee 113 143 98 23   8 385
Staff 240 112 74 30 3   459
Staff member and detainee together 1 1         2
Religious figure 4           4
Teacher 34 5 5       44
Ex-teacher 1           1
Other type of perpetrator 9 10 4       23
Not stated 38 26 8       72
Total 440 297 189 53 3 8 990

* ‘SCH’ refers to SCHs holding children on both justice and welfare placements, whereas ‘SCH W’ refers to SCHs only holding children on welfare placements.

† The ‘Other’ column includes the three incidents of abuse which occurred outside a custodial institution, referred to above.

‡ This refers to the 990 cases of abuses: see explanation for ‘case’.

23. Of the 990 cases, 385 (38.9 percent) related to other detainees and 461 in total (46.6 percent) involved a staff member acting alone or with a detainee.

24. There were 121 cases involving abuse allegedly perpetrated by detainees in SCHs (including those on welfare placements), 143 in STCs and only 113 in YOIs. This is despite the overall population of YOIs being several times higher than the populations of SCHs and STCs at any given time. It raises the question of whether the higher ratio of staff per detainee in the SCHs and STCs compared with YOIs may make it more likely that abuse between detainees is identified, recorded and reported to agencies by staff.

Circumstances of abuse

25. Some patterns have emerged in the circumstances of the reported abuse:

  • 92 incidents were alleged to have taken place during search or restraint. The percentage was highest in YOIs. Of the 469 alleged incidents occurring in YOIs, 57 (11.1 percent) occurred during search or restraint.
  • 53 alleged incidents happened in the showers or bathrooms. Again, the percentage was highest in YOIs. Of the 469 alleged incidents occurring in YOIs, 40 (8.5 percent) happened in the showers or bathroom.
  • 27 incidents were alleged to have happened in a cell. This raises the question of whether the absence of CCTV in these areas may increase risk of sexual abuse.[17]

26. The figures which the Inquiry’s analysis gives for incidents in each category may be lower than the actual numbers. This is because the precise details of each allegation were not always apparent from the material provided, and so there may be further incidents falling within the above categories which have not been counted above.

How incidents were reported and responded to

27. Many allegations (where such information is available) were disclosed to someone within the institution by the alleged victim, a witness or fellow detainee. A small number of incidents (19) were not reported at the institution where the alleged abuse took place, but rather at another custodial institution by a witness or victim. Some were reported after release.

28. In 166 allegations, it was recorded that the alleged victim withdrew, retracted or was unwilling to pursue the allegation. Investigators described 49 allegations as malicious or similar. Many other complaints were found not to be substantiated for other reasons, though those reasons were not always clear.

Limitations and issues

29. The Inquiry has identified a number of limitations in relation to the data obtained from the published surveys:

  • The sample group included 18-year-olds, ie adult not child victims.
  • The survey results give the percentages of respondents who have been sexually abused, but not the total number of incidents of abuse.
  • Prior to October 2012, children in STCs did not take part in surveys in which a question about sexual abuse was asked.
  • Children in SCHs have never been asked specifically about sexual abuse.[18]

30.  The further data obtained by the Inquiry may also have limitations:

  • A number of respondents, including the PPO, the Youth Justice Board, HM Prison and Probation Service (HMPPS), Ofsted, police forces and at least six local authorities, highlighted shortcomings in the data provided or the methods used to search for it, which may mean those bodies received more allegations of sexual abuse than those they have provided to the Inquiry.
  • Similarly, there are significant levels of incomplete records held by the various bodies, including some YOIs, some STCs (and some STCs’ records have not been made available to the Inquiry), and some local authorities, although we note there might be various reasons for this.
  • Using existing systems, it is not possible to categorise an incident as sexual abuse of a child in custody, so as to allow records to be retrieved easily and thus to distinguish custody-related abuse from other categories.
  • We have also seen incidents recorded in one place but not in others as we would have expected.[19] There were 130 occasions on which police were said to have been informed about an incident which the relevant police force did not refer to when responding to requests. Similarly, there were 211 occasions on which a local authority was informed about incidents, but which were not included in local authority evidence.
  • There were also issues with inaccurate recording or missing details.[20]



  2. The requests asked for, in respect of any allegation or report of child sexual abuse occurring at a custodial institution: (a) a brief description of the alleged incident; (b) the number of victims; (c) the date(s) of the incident(s); (d) the type of investigation undertaken; (e) the outcome of the investigation; and (f) whether the alleged perpetrator was a member of staff, detainee or someone else. The requests also asked, in respect of incidents in SCHs, whether the victim had a justice or a welfare placement at the institution.
  3. See Annex 2 to Inquiry Counsel’s analysis, summarised at 10 July 2018 168–180
  4. See Note from Inquiry Counsel dated 11 June 2018 at INQ001709 and Addendum to Note from Inquiry Counsel date5 July 2018 at INQ001769 and accompanying Annexes, as summarised at 10 July 2018 168–180 and 20 July 2018 4–7
  5. There were, in addition, two alleged sexual assaults of children resident in custodial institutions where the incident happened at court, and one by a manager of a custodial institution while the victim was on licence. In any case where it can be inferred from the date, description and/or action taken that different records are likely to be referring to the same incident, the Inquiry has counted it as a single incident for the purpose of the analysis. In most cases, the position is reasonably clear. However, in relation to 26 incidents, it is unclear whether they are duplicated in other records. In relation to those cases the analysis has rounded ‘down’ rather than ‘up’. It is possible therefore that there are 26 more alleged incidents of abuse, above those recorded here.
  6. Each act of abuse has been counted as an incident. Sometimes several incidents of abuse were recorded as part of the same case, ie in the same allegation, complaint or report. This was usually where a group of incidents occurred on the same date and close in time, involving the same parties, but sometimes where there were series of similar incidents for which only the date of the last incident was recorded in material provided to the Inquiry. Each case is recorded in a single row of data, and the number of incidents is reflected in the ‘incidents’ column. There were in total 990 such cases of abuse, the explanation for the difference between this figure and the total number of incidents being that some cases were recorded as a group of incidents of the types described.
  7. This is an estimate because although the figures have been adjusted where it is clear that the same victim was involved in two or more incidents, the material did not indicate in every case whether two or more incidents involved the same victim, or the exact number of victims for some incidents involving multiple victims.
  8. The figure of 139 for ‘Count of action against perpetrator’ in the third sheet of Annex 2 includes the 19 cases where action as specified in this paragraph was taken.
  9. However, in many cases it is unclear whether the action was taken only temporarily during the investigation, or at the end of it.
  10. Albeit that one of the incidents which resulted in conviction, and possibly a further incident which resulted in charge only, occurred outside the relevant custodial institution: the perpetrators paid for sexual activity with a child, the child having been moved from the institution to other premises.
  11. INQ001125
  12. INQ001481
  13. INQ001571
  14. This pattern is reflected in other establishments. For example, the July 2014 survey at Hassockfield STC disclosed no complaints of sexual abuse but the Inquiry’s evidence shows that there were a total of five alleged incidents in 2014. Similarly, the June 2014 and August 2017 surveys of Cookham Wood revealed no reports of sexual abuse. However, the Inquiry’s evidence found 12 alleged incidents recorded or occuring between May 2013 (when the previous survey of Cookham Wood had been undertaken) and June 2014; and 10 alleged incidents of sexual abuse between September 2016 (when a survey was undertaken) and August 2017 (when a survey revealed no sexual abuse).
  15. It is important to note that the percentages show rate of incidents (which, for the avoidance of doubt, is the number of incidents expressed as a percentage of the average population of the institution for the same year) and not the percentage of detainees who have been sexually abused. The latter is impossible to ascertain from data available because in many cases no victim identifier was used by evidence providers.
  16. Here, ‘case’ refers to an incident or group of incidents, within the same allegation, complaint or report, either occurring on the same date at around the same time between the same parties, or forming part of a series of similar incidents where only the last incident date is recorded. The total number of such ‘cases’, 990, is lower than the total number of incidents, 1,070, because some cases involved a group of similar incidents.
  17. Where figures are higher than the number of incidents in the third sheet of Annex 2, this is to take account of where more than one incident was recorded in a case.
  18. See Annex 1 to Note from Inquiry Counsel at INQ001710
  19. For example: (i) In February 2015, Ofsted received from the Youth Justice Board a spreadsheet of alleged incidents at Rainsbrook recorded on the Youth Justice Board’s IARMS system. This spreadsheet has been compared to IARMS records provided directly to the Inquiry investigation team from the Youth Justice Board. There are four incidents referred to on the former but not on the latter. (ii) Several allegations of sexual abuse were not recorded on child protection logs but instead in other places such as on security information reports or complaint forms. (iii) The NSPCC’s Childline service received eight disclosures of sexual abuse from detainees in 2015 regarding children in YOI Werrington which do not appear to have been recorded by any state body (INQ001709; INQ001769)
  20. For example (i) the date of incident is not always given; (ii) the location of abuse may not always have been recorded accurately, such as where different custodial institutions are geographically close to each other; (iii) the categorisation of abuse may not always be consistent, such as in respect of incidents of alleged sexual abuse being recorded by an evidence provider as physical abuse, and so the relevant information was not provided until further queries by the investigation were made; and (iv) in just over 10 percent of the PPO’s records, the age of the victim is missing and so it is impossible to know whether the alleged victim is a child or adult (INQ001709; INQ001769)
Back to top