One of my formative experiences as a postgraduate criminology student 20 years ago was spending part of a night shift observing a police custody suite in Edinburgh, where I was studying. I was wholly unprepared for the levels of vulnerability that presented among the drug users, alcoholics and drink drivers in the cells that night, and quickly came to appreciate the high level of responsibility held by the custody staff. I still remember their preoccupation with alcoholics ‘drying out’ in the cells, and the risks of seizures and other health complications that could arise.

That small vignette provides the backcloth to a more recent interest in the use of strip search in custody, not least in the context of some high-profile cases reported in the media, and – inevitably – questions about whether decisions to strip search detainees are necessary, proportionate and consistent (fair). Although the case of Child Q was not a custody strip search but done under stop and search powers away from a police station, it did nevertheless cause the use of strip searches by police, especially by the Metropolitan Police (Met), to be given renewed scrutiny, with a particular focus on the treatment of children and the question of whether race features in decision making (whether consciously or otherwise). This recent House of Commons briefing on strip searches by police gives a good overview.

Earlier this year, I sought to examine the question of racial and gender disproportionality in strip searches of children in police custody by the Met using a combination of published (FOI) data on numbers of custody strip searches by age, gender and ethnicity, and published Home Office data on arrests. My analysis suggested that, overall, Black boys were 1.8x as likely to be strip searched as White, and that Black girls were 0.7x as likely (so somewhat less likely) to be strip searched than White girls. But my analysis was limited by having to use two mis-matched datasets, because one was based on calendar years and the other on financial years, and by the lack of detail on reasons for arrests in the strip search data. In the former case, there was a risk that in the gap errors would have crept in, especially for girls (being relatively few in number), and in the latter case that the lack of detail prevented more like-for-like comparisons being made.

In order to address those gaps, I submitted a Freedom of Information (FOI) request to the Met for five years of custody data on detentions by gender, age, ethnicity, offence type and whether strip searched, plus subject demeanour on arrival at custody and PNC warning markers. This was to allow me to calculate comparative strip search rates for a variety of different sub-groups: my hope was to try and provide a more robust picture of strip search practices, including more like-for-like comparisons that controlled for things like differences in offence type. I heard back recently, and although I didn’t get everything I asked for – in particular I didn’t get the PNC warning markers – the data that I did receive have provided what I hope are some very useful insights into the use of strip search in custody, including how it has changed over time, and how it varies by the personal characteristics of those being detained. The full analysis is available in the form of a slide pack.

Before highlighting some of the key findings, it is worth saying that the data I received did not – and perhaps could not – cover the full range of factors considered by Custody Sergeants when deciding whether to authorise strip searches, which are primarily concerned with ensuring the safety of detainees and staff. Offence type is clearly relevant, not least if there is a need to mitigate any risks presented by the possibility of concealed drugs or weapons (and it must be said that the data shared by the Met were very limited in terms of offence details). But other factors will include, for example, any information recorded about a detained person on police systems (have they previously concealed drugs or weapons or self-harmed?), the answers (or not) the detained person gives to direct questions about their mental state, their behaviour, and any other concerns officers may have. So the data I received can only take an examination of strip search decisions so far.

With those caveats in mind, I think there are five key findings:

1. Overall strip search rates in custody increased from 16.5% to 23.2% of detainees between 2017/18 and 2020/21, before falling slightly to 19.9% in 2021/22. The varying proportion of arrests linked to drugs and/or weapons seems largely to explain these headline changes, because of the high rates of strip search associated with those offences: when a higher proportion of arrests relate to drugs or weapons, overall strip search rates for all detentions are higher.
2. Strip search rates are higher for arrests relating to weapons and especially drugs than other arrest reasons (combined), higher for adults (18+ years) than youths (10-17 years), higher for males than females, and have increased more rapidly for youths than adults. The large gap in strip search rates between male and female youths seems notable, even when controlling for offence type (for example, just looking at arrests for drugs offences only).
3. Differences are observed in strip search rates by ethnicity when controlling for age (10-17 vs. 18+), gender and offence type. Overall, Black* males are strip searched at 1.5x (adults) and 1.8x (youths) the rate of their White counterparts – and at a higher rate for all offence groups. By contrast, Black females are strip searched less often overall (0.9x and 0.8x the White rates respectively), for all offence groups in the case of adults. My original estimates for boys and girls held up very well.  (*Note that Black includes those with Black/White mixed heritage in the analysis.)
4. Rates of ethnic disparity have changed over time, including having notably fallen for male youths arrested only for drugs offences (a reasonably good like-for-like comparison). But this is partly a function of related strip search rates having increased rapidly for all male youths, while the size of the ethnicity gap has remained largely unchanged.
5. The profile of arrest offences varies between custody suites, but that doesn’t appear sufficient to explain differences in strip search rates between them. For example, large differences are seen between strip search rates for arrests related only to drugs offences.  

In the slide pack, I set out seven key questions that I think arise from the data. I won’t repeat them in full here, but an overarching theme is to wonder (a) whether sufficient data are collected on Custody Sergeant decision making to allow equality of treatment to be fully examined by the Met (and other forces) and those holding them to account, and related (b) to what extent more detailed data would explain the disparities seen in the more limited FOI data made available to me.  

My main takeaway from the data is that the Met appears to have been taking an increasingly risk-averse approach to younger detainees in particular, and to wonder why that is and where the line should be drawn regarding the necessity and proportionality of strip searches – and how to know if it has been crossed?

Ever since my experience in Edinburgh 20 years ago I have been aware that police custody can be a place of great vulnerability. That can relate to the condition of detainees when they arrive in custody, for example due to existing health or substance use issues, but is also related to the austere conditions in custody and the inherent nature of being detained. Some detainees can also be very determined to harm themselves or others, or to evade detection (for example by swallowing evidence at great risk to their own health), and the responsibility to keep everyone safe must weigh heavily on the shoulders of Custody Sergeants. They need to get their decisions right every time, and they may reasonably err well on the side of caution in doing so.

Having acknowledged all of that, however, I think we can still ask whether it is possible that risk-based strip search decisions – taken well away from the public gaze – may have drifted towards giving insufficient regard to the possibility that strip search may itself be traumatic and involve risks to detainees’ mental wellbeing, especially among the under-18s. In particular, I wonder about the large gap in strip search rates between boys (much higher) and girls (much lower), and whether the vulnerability of boys to emotional harms that may be caused by strip searches may have been given progressively less, and perhaps ultimately too little, consideration?  

There will, of course, be no easy answers. Indeed, there never are in policing, which constantly requires officers and staff to balance competing risks with limited information and without the benefit of hindsight, and with what can be catastrophic consequences if things go wrong. I hope my analysis can at least help support a more informed debate.

The FOI data have been published on the MPS website and are available at https://www.met.police.uk/SysSiteAssets/foi-media/metropolitan-police/disclosure_2022/october_2022/data-for-custody-detentions-and-related-strip-searches-from-april-2017-to-march-20222

You can read Gavin’s full analysis in this slide pack.