March 2025: Doing violence to data
Police chiefs and politicians fall for new research that suits their claims about stop and search's effectiveness, while a hospital's knife bin amnesty hints at another reality
Dear StopWatchers,
Another busy month saw much discussion about stop and search’s effectiveness in dealing with knife violence among politicians, while a hospital in south London showed that there are other ways of tackling the problem too.
Meanwhile, as the crime and policing bill continues to made its way through the House of Commons, we submitted written evidence criticising the proposed Respect Orders included in the legislation. We also joined Big Brother Watch, Liberty, INQUEST, Index on Censorship, NETPOL, Open Rights Group and Privacy International in a call opposing clauses 95 and 86 of the bill.
We’d like to flag the upcoming launch of the People’s Tribunal on Police Killings, which brings together families and friends of those who have died at the hands of the police. The event will feature panellists such as Mireille Fanon Mendès-France (Frantz Fanon Foundation), plus expert witnesses such as professor Gus John, professor Gargi Bhattacharyya, and Dr Asim Qureshi (CAGE UK). Grab your tickets for the free event on 5-6 April here before they sell out.
We also have an imminent launch of our own: the publication of our girls and young women’s research project report! You can read a summary of the #GIRLSMATTERTOO findings and via Instagram.
We also made a video (watch below) featuring several of the collaborators explaining how the project came about and what it means to them.
If you are interested in joining us for the online launch of the report (date: Wednesday 16 April, 18:00), we’d love to see you! Email info@stop-watch.org to sign up.
And coinciding with interviewees’ testimonies of interactions with the police, EastEnders aired a storyline in which a teenage girl was strip searched. We published a scene-by-scene analysis of the encounter.
If you want to support the work we’re doing financially, we’d appreciate it! Click the button below to set up a regular or one-off donation.
We hope you enjoy this edition.
Doing violence to data
This month saw some of our elected representatives come together to consider the matter of ‘the use of stop and search’ (Hansard, 12 Mar). The debate took place amidst persistent calls from police chiefs (notably Sir Mark Rowley of the Met) urging greater use of the tactic, as we await the results of an impending police accountability review aiming to give officers ‘the confidence to be able to use their powers’, and ahead of anticipated hikes in the council tax precept for police forces as the result of yet another successful campaign to nab ever more taxpayers’ money to ‘keep the public safe‘.
While we are familiar with the default posture of most parliamentarians – to place on record their gratitude for the officers ‘who work tirelessly to keep us all safe’ – the opening speech from MP Saqib Bhatti (Meriden and Solihull East) who raised the debate was keen to stress the importance of stop and search playing a very important role in tackling knife crime, ‘as backed up by evidence’.
For that ‘evidence’, Bhatti turned to the work of the former Met police chief scientific officer, professor Lawrence Sherman. In a paper written for the Cambridge Centre for Evidence-Based Policing he founded, looking at 15 years of Met stop-search activity (Stop, Search and Knife Injuries in London), Sherman claims to have found a strong causal inverse correlation between stop and searches and knife injuries and homicides, centred around two major policy-driven changes in policing operations. That is, a significant decrease in stop and searches in 2014 led to more knife injuries and homicides, and vice versa following a significant increase in 2018 (watch co-collaborator professor Alex Piquero summarise the findings here).
Along the way, the authors take aim at previous studies attempting to identify associations between stop and search activity and types of crime, criticising them for failing ‘to address the crime type that is, theoretically, most likely to be prevented by SSEs’. We are led to believe that is knife violence.
The problem is, we – and they – ‘cannot be entirely certain that at any point in time the data follow consistent definitions, such as knife crime injuries in public places or non-domestic crimes that exclude violence between family members or intimates’. This seems too vital a detail to let slide in the case of knife homicides, as the annual count / rate is (mercifully) low in England and Wales (globally speaking). At low counts, small fluctuations make for big effects, proportionately. And if the authors can’t even rely on consistent definitions of knife injuries and homicides, how can they claim with confidence that ‘it is unlikely that any sharp changes in measurement occurred simultaneous with sharp changes in either SSEs [stop and search encounters] or knife injuries’?
Authors of previous stop and search studies may well have faced the same dilemma, which may explain why the College of Policing’s own 10-year analysis of Met police stop-searches makes use of ambulance service data on stab-, shot-, and weapon-wound incidents instead. According to the researchers behind that study, police data on weapon-enabled violence ‘were not robust because of the number of boroughs that recorded no offences’.
In fact, police recorded crime statistics for England and Wales in general are unreliable, and have been so for at least a decade: they were stripped of National Statistics accreditation in 2014 ‘following an assessment which found evidence that the quality and consistency of the underlying data may not be reliable’. Which is why you’d think any research using police crime recorded data would be taken with a pinch of salt large enough to dry out a rout of slugs. And you needn’t just take our word for it: Tufton street think-tank Civitas raised these concerns at the time too.
Instead, Sherman and Piquero chose to treat the data with ‘the most advanced methodology available’ while their ‘best information’ is that ‘it is unlikely that any sharp changes in measurement occurred simultaneous with sharp changes in either SSEs or knife injuries’, as though manipulated crime figures could be made pure through quasi-experimental testing, regression modelling and inshallah.
So when professor Piquero claims to have ‘never seen data like this or results of this strength and calibre’ before, it comes as no surprise to us why this is: it is built on a house of cards. To be clear: while police forces record crime more accurately now – still not to standard, mind – Sherman and Piquero’s period of analysis focusses on a period when ‘lax compliance’ with crime recording standards was commonplace among forces. This fatally undermines the validity of their findings.
Shaky foundations aside, the fact remains that drugs are by far the most common reason for searches, with the annual volume of weapons searches lagging behind by roughly 40 percentage points. And for all officers’ suspicions about a potential link between drug and knife possession, only 1% of drugs searches find offensive weapons, according to Home Office stop-search records (as per table SS_42 of the 2023/24 dataset).
But these inconvenient facts have not stopped some police chiefs and politicians – looking for a simple winning excuse to justify stop-searches of marginalised communities at scale – from heralding the study. Where other comprehensive analyses into the complexities of stop and search and violent crime found insignificantly small, non-causal associations between both variables at best, for the likes of Sir Rowley and MP Bhatti, any chance to show an inverse link between stop-search volumes and knife injuries / homicides – even using poor quality data – provides them with ‘the clearest Utilitarian calculus for justifying the trauma imposed by SSEs’. Sherman and Piquero’s paper legitimises their preconceptions about stop and search and gives licence to press ahead with policing operations that ride roughshod over the sensitivities behind persistent Black / white ethnic disparities in the figures, while underplaying the concerns of racialised communities in the process.
Meanwhile, the collective trauma of overpolicing continues to be felt by those racialised neighbourhoods, brewing resentment amongst those inhabitants who will wonder why it is that they must be treated as collateral damage to save that proverbial life. The answer to this is obvious to anyone who has experienced or seen stop-searches in ‘hot spot crime zones’: token nods to conducting stop and search in a ‘procedurally just’ manner will simply be ignored by units of officers who exhibit racist behaviours. After all, as Sherman and Piquero are at pains to stress in their paper, operational independence is a matter of conscience for each and every constable.
For all the scientism behind the presentation, the study’s authors cannot get around three things: 1) what the Campbell Systematic Review into stop and search termed ‘backfire effects’; 2) the fact that the data available is of moderate quality at best, according to the Youth Endowment Fund’s evaluation of stop-search powers; and 3) the possibility of non-policing alternatives having a significant impact on levels of knife violence. If it were possible to employ a reductive approach to discovering the nature of the complex relationship between stop and search and knife violence, researchers would have done so years ago.
Besides, something about a bivariate analysis seems rather incomplete when the presence of a knife amnesty bin outside a south London hospital has arguably made as much difference to getting weapons off the streets as stop and search patrols (Sky News, 12 Mar). That the collection of discarded knives in the six months to the end of February was comparable with the number of weapons searches resulting in ‘positive outcomes’ conducted by Met police officers in the entire surrounding borough suggests that there are viable alternative solutions to tackling the problem of knife violence.
A more nuanced view of stop and search’s utility from the likes of Sir Mark Rowley and Saqib Bhatti would be a fine thing. But they would prefer Sherman and Piquero’s flawed experimental statistics over the conclusions of the Campbell Review’s meta-analysis, which concludes: ‘results suggest that pedestrian stops can be an effective crime control strategy, but one that comes with considerable drawbacks’ (emphasis added). Or the idea that ‘stop and search is a tool of social control’ rather than a weapons-finding tactic.
And away from all of the argument over its effectiveness, we must remember that this was never the legal justification for the tactic in the first place. If the reasoning for each and every stop was actually a priority, the tactic would likely become more effective, and maybe even less racist(!). But it would likely result in fewer stop-searches too, which Sir Rowley ruled out at a February London Assembly police and crime committee meeting ahead of the stop and search charter. Not content with doing violence to young people under the pretence that it is the most effective way to deter knife violence, the police seem to want to do violence to data too.
Deaths from police contact, cases old and new
Oladeji Omishore
An inquest into the death of Oladeji Omishore found that failures in the Met Police’s control room may have played a role in his death. Deji, who was experiencing a mental health crisis, was tasered three times by officers on Chelsea Bridge before falling into the River Thames in 2022. Officers made no attempt to de-escalate the situation. Despite this, the police watchdog ruled that the officers involved would not face misconduct proceedings (INQUEST, 25 Mar).
Other news
Outcry as UK’s Black police leader faces misconduct probe over tweet: Chief inspector Andy George, president of the National Black Police Association, is under investigation for a tweet criticising disparities in how Black and Asian officers are treated. His comments followed the controversial promotion of a firearms officer after being cleared of murder. George warns that the investigation sends a ‘chilling message’ and stifles free speech on racial disparities (Guardian, 04 Mar).
George has also raised alarm over the Met’s move to immediately pause budget allocation for leadership of Met Black Police Association chair, effectively blocking newly elected Diane Vincent from taking up the role. He told the London Evening Standard: ‘It has become glaringly evident that the Metropolitan Police Service is engaged in a systematic effort to dismantle the Met BPA’ (12 Mar).
New inspection framework to shaping the future of policing: His Majesty’s Inspectorate of Constabulary and Fire & Rescue Services (HMICFRS) has outlined its inspection programme for 2025–29, which will focus on neighbourhood policing, reducing knife crime, and addressing violence against women and girls (HMICFRS, 04 Mar).
Undercover police engaged in long-term sexual relationships with the women they spied on: ITV’s documentary The Undercover Police Scandal: Love and Lies Exposed recounts the harrowing stories of women deceived into long-term relationships by undercover officers. The series highlights the systemic abuse behind these operations and the relentless pursuit of justice by the women affected, underscoring the ongoing struggle for accountability (Guardian, 06 Mar).
Police 100 times more likely to be convicted of cybercrime: An analysis by Novara Media reveals that police officers and staff are 100 times more likely to be convicted of cybercrime than the average person. Misuse of sensitive police databases accounted for 42% of all cybercrime convictions in 2024, raising alarming concerns about data security within law enforcement (Novara Media, 07 Mar).
Met Police suggest planting roses to deter burglars: The Met police have advised homeowners to plant thorny shrubs like climbing roses to prevent burglaries. The advice, included in leaflets distributed to victims, suggests that dense, mature plants near walls and fences can act as a deterrent. Critics have questioned this approach, given that 82% of burglaries went unsolved by the Met in 2022-2023, making it one of the worst-performing forces in the country (LBC, 15 Mar).
Consultation on changes to codes of practice under terrorism legislation: The government has opened a consultation over the codes of practice for schedule 7 of the Terrorism Act 2000 and achedule 3 of the Counter-Terrorism and Border Security Act 2019. A new addition under paragraph 25 clarifies that an examining officer may stop and question a person to determine their involvement in terrorism or to assess whether their presence in a border area is related to entering or leaving Northern Ireland. The revised guidance emphasises that officers should only exercise examination powers where engagement with the individual has been unsuccessful or deemed inappropriate (Gov.uk, 17 Mar).
Mass ANPR camera deployment planned in Manchester: Manchester councillors plan to install automatic number plate recognition (ANPR) cameras at seven junctions to enforce ‘moving traffic offences’. These offences include ignoring yellow box markings, making prohibited turns, and driving the wrong way down one-way streets (Manchester Evening News, 18 Mar).
Survey reports financial struggles and retention crisis in UK police service: A Metfriendly survey found that 20% of respondents plan to leave policing for another job, citing high stress and financial difficulties. 55% of police officers and 51% of police staff reported experiencing high levels of stress (Emergency Services Times, 19 Mar).
Record rise in misconduct cases against West Midlands police officers: Misconduct cases involving West Midlands police officers have surged, with 138 cases recorded in 2021-22, rising to 240 the following year, and reaching 343 in 2023-24 (Express & Star, 19 Mar).
Campaign fights back against deadly police pursuits: The Northern Police Monitoring Project has launched the 'End Police Pursuits' campaign, arguing that police pursuits are disproportionate and frequently result in police-induced fatalities. The campaign highlights the dangers associated with high-speed chases and has initiated a petition calling for an end to the practice (The Canary, 23 Mar).
Permanent facial recognition cameras installed in Croydon: Live facial recognition (LFR) cameras will be installed on two roads in Croydon city centre as part of a pilot project. The Met states that the cameras, mounted on existing lampposts and buildings, will only be activated when officers are using the technology in the area (The Register, 27 Mar).
The news comes as the MP for Liverpool Wavertree (Paula Barker) raised concerns about the risk of facial recognition misidentifications in parliament, and called for safeguards to protect the public. The Home Office acknowledged the concerns and stated that police algorithms funded by the department undergo independent testing. However, the government states that it is still reviewing the legal framework to ensure facial recognition is used responsibly and equitably.
Section 60 watch*
London
Lambeth (05 Mar)
Thames Valley
Castlefield, High Wycombe (2-3, 22 Mar), Cowley, Oxfordshire (14 Mar)
West Midlands
Walsall (17-19 Mar)
West Yorkshire
Wakefield city centre (15 Mar)
Greater Manchester
Cheetham & Crumpsall (11-12 Mar)
Merseyside
Walton (13-14 Mar), Kirkby (15-16 Mar), Bootle and Netherton (19 Mar)
* This is not a comprehensive list
Stay safe,
StopWatch.