The End of Britain's Thought Police? UK Scraps Non-Crime Hate Incident Recording After Decades of Controversy

For more than a decade, police forces across England and Wales recorded tens of thousands of incidents each year that were not crimes. A social media post, a verbal comment in a pub, a child calling a classmate a name — all could be logged in criminal databases, disclosed on background checks, and shadow a person's employment prospects for years. Now, that system is being dismantled.

On December 23, 2025, police commissioners and senior law enforcement officials announced they would end the practice of recording non-crime hate incidents (NCHIs), replacing it with what they described as a "common sense" approach focused on genuine threats [1]. The Crime and Policing Bill 2025-26 formalizes this shift by repealing sections 60 and 61 of the Police, Crime, Sentencing and Courts Act 2022, which had given the Home Secretary authority to issue a statutory code of practice governing NCHI data processing [2].

The decision marks the conclusion of a bitter, years-long fight that pitted free speech advocates, civil liberties groups, and several high-profile legal challengers against hate crime monitoring organizations and community groups who argued the system served as an early warning mechanism.

Origins: From Stephen Lawrence to a Sprawling Surveillance System

The roots of NCHIs trace to one of the most consequential moments in modern British policing. In 1999, Sir William Macpherson's inquiry into the racist murder of teenager Stephen Lawrence found that the Metropolitan Police investigation had been "marred by a combination of professional incompetence, institutional racism and a failure of leadership" [3]. Among Macpherson's 70 recommendations was a principle that would reshape policing for a generation: "A racist incident is any incident which is perceived to be racist by the victim or any other person" [3].

This perception-based recording standard was initially confined to race. Over the following two decades, it expanded to cover religion, disability, sexual orientation, and transgender identity — the five "monitored strands" [4]. Until 2014, police forces applied the system informally; then the College of Policing issued official guidance formalizing the practice. In 2023, the Home Office went further, issuing a statutory Code of Practice [5].

The result was a system in which anyone could report an incident they perceived as motivated by hostility toward a protected characteristic, and police were obligated to record it — regardless of whether any evidence supported the perception, and regardless of whether any law had been broken.

The Scale: 120,000 Records, 60,000 Hours, Zero Crime

The numbers reveal the scope of the system. According to freedom of information requests, police forces in England and Wales recorded approximately 13,200 NCHIs per year as of 2024 [6]. Over the preceding five years, the cumulative total exceeded 120,000 records [7]. Since the system's formalization in 2014, more than 133,000 NCHIs were recorded across 43 police forces [1].

Processing these incidents consumed significant resources. A Policy Exchange report endorsed by Lord Hogan-Howe, the former Metropolitan Police Commissioner, estimated that over 60,000 police officer hours per year were spent on NCHIs [8]. An inspection found a 25% error rate among NCHI case files — one in four records examined failed to meet the system's own standards [8].

Source: Policy Exchange / FOI Responses

Data as of Jan 1, 2024CSV

Recording practices varied wildly between forces. Essex Police logged 21.5 NCHIs per 100 officers annually in 2023 — a rate three times that of the Metropolitan Police, four times that of Greater Manchester, and ten times that of West Yorkshire [8]. This inconsistency raised questions about whether the system measured genuine patterns of hostility or simply reflected different bureaucratic cultures within individual forces.

What Got Recorded: From Social Media to Schoolyard Name-Calling

The breadth of incidents captured by the NCHI system became one of its most criticized features. Reports compiled by Big Brother Watch and other civil liberties organizations documented cases including children writing with chalk on pavements, an 11-year-old calling a peer "shorty," and a man whistling the tune of "Bob the Builder" — the last recorded as potentially discriminatory [7].

Social media posts constituted a significant category. The most prominent case involved Harry Miller, a former police officer who posted 31 tweets expressing gender-critical views between November 2018 and January 2019. A complainant reported these to Humberside Police, who recorded them as an NCHI, visited Miller's workplace, warned him the tweets were "transphobic," and told him that escalation could lead to criminal prosecution. Police classified Miller as a "suspect" and the complainant as a "victim" in crime reports — despite no crime having been committed [9].

Verbal comments, neighbourhood disputes, and written communications also fell within the system's net. The Home Office itself acknowledged that NCHIs had "a chilling effect on free speech and potentially stop people expressing views legally and legitimately" [10].

Miller v College of Policing: The Ruling That Changed Everything

Harry Miller did not accept the warning. He challenged the College of Policing's guidance through judicial review, and in December 2021, the Court of Appeal handed down a judgment in Miller v College of Policing [2021] EWCA Civ 1926 that fundamentally undermined the legal basis for NCHI recording [9].

The Court found that the 2014 guidance "unlawfully interfered with the right to freedom of expression" under Article 10 of the European Convention on Human Rights. It described the NCHI system as casting an "exceptionally wide" net "designed to capture speech which is perceived to be motivated by hostility against one of the protected strands, regardless of whether there is evidence that the speech is motivated by such hostility" [9].

The judgment emphasized the chilling effect: "Comparatively little official action is needed to constitute an interference" with free expression rights, particularly regarding "political speech and debate of questions of public interest" [9]. The court ordered that additional safeguards be introduced, specifically that complaints that were "trivial, malicious or irrational" must not be recorded [11].

Following the ruling, the College of Policing revised its guidance. But critics argued the revisions were insufficient. Fair Cop, the campaign group Miller founded, and the Free Speech Union continued to press for complete abolition rather than reform [12].

Employment Consequences: Criminal Databases Without Crimes

One of the most consequential aspects of the NCHI system was its intersection with employment screening. NCHIs were recorded on police criminal intelligence systems — CRIS, CRIMINT, IIP, and the Police National Database (PND) [13]. While they would not appear on basic or standard Disclosure and Barring Service (DBS) checks, they could be disclosed on enhanced DBS checks if police judged the information met "relevancy and proportionality criteria" [14].

Enhanced DBS checks are required for anyone working with children or vulnerable adults — teachers, healthcare workers, social workers, care home staff, and many volunteers. A recorded NCHI, even one based on a single uncorroborated complaint, could cost someone their livelihood in these sectors [14].

The Policy Exchange report identified this as one of three primary harms of the system, describing it as "curtailing the employment prospects of individual members of the public through inappropriate disclosures" [8]. There was no formal appeals process for those who had NCHIs recorded against them, and until the Miller ruling and subsequent legal interventions, there was no established route to have records deleted.

Some individuals did achieve deletion. Kevin Mills, a 63-year-old electrician, had an NCHI recorded by Kent Police without his knowledge after a customer complaint. With representation from the Free Speech Union, he secured complete deletion of the record [15]. But such outcomes required legal intervention that most people could not access or afford.

The Case For: Early Warning or Empty Justification?

Defenders of the NCHI system argued it served a legitimate intelligence function. The original Macpherson rationale held that tracking sub-criminal hostility could reveal escalation patterns — identifying individuals or groups whose behaviour was trending toward criminal hate offences before they crossed that threshold [3].

Community organizations, including some monitoring antisemitism, Islamophobia, and anti-LGBT hostility, expressed concern that abolition would leave a gap in the early detection of hate. The West London Equality Centre noted that the Metropolitan Police's October 2025 decision to stop investigating NCHIs removed a channel through which vulnerable people could report targeted harassment that did not meet criminal standards [16].

The government's ECHR memorandum on the Crime and Policing Bill acknowledged that police processing of data about hate-motivated behaviour must remain compliant with human rights standards, even after the repeal [2]. And the statutory code that preceded repeal had introduced a higher threshold, requiring that incidents show "intentional hostility" and a "real risk of escalation causing significant harm or a criminal offence" before personal data could be recorded [5].

However, the empirical case for the system's effectiveness was thin. When asked through freedom of information requests, not a single police force out of 43 could provide a single example of a crime that had been prevented by recording an NCHI [10]. The system generated voluminous records but offered no measurable evidence that it fulfilled its stated purpose of interrupting escalation patterns.

The Case Against: Free Speech, Proportionality, and Resource Allocation

The case against NCHIs unified an unusual coalition. Civil liberties organizations like Big Brother Watch and the Foundation for Individual Rights and Expression (FIRE) argued the system amounted to state surveillance of lawful speech [7][17]. The Free Speech Union campaigned under the hashtag #PoliceOurStreetsNotOurTweets [12]. Policy Exchange, a centre-right think tank, produced detailed analysis of the resource costs [8].

Lord Herbert of South Downs, chair of the College of Policing, delivered the decisive institutional verdict. He confirmed that if proposals were approved, "non-crime hate incidents will no longer be recorded. They will go." He emphasized: "This will not be a mere rebranding exercise. The threshold of an incident will be significantly increased. Common-sense professional judgment will guide decisions and only where there is a genuine risk of harm and a clear policing purpose will incidents continue to be recorded" [1].

Home Office Minister Lord Hanson of Flint told the House of Lords that NCHIs were "not fit for purpose" and would be "replaced with a clearer, more proportionate model" [18]. The College of Policing and National Police Chiefs' Council jointly concluded that "recording information about non-crime incidents on a crime system is not the right solution" [18].

The resource argument carried particular weight at a time when police forces faced significant operational pressures. With over 60,000 officer hours consumed annually — equivalent to roughly 30 full-time officers doing nothing else — and recorded hate crimes running at 137,550 per year in the year ending March 2025, the trade-off between processing non-crimes and investigating actual offences became increasingly difficult to justify [8][19].

Source: Home Office / GOV.UK

Data as of Oct 1, 2025CSV

What Happens Now: Existing Records and Replacement Systems

The fate of existing NCHI records remains one of the most consequential unresolved questions. The government's ECHR memorandum states that "personal data relating to non-crime hate incidents continues to be processed after the repeal comes into force" but does not mandate deletion [2]. The repeal removes the statutory code of practice but does not establish a framework for expunging historical records.

Individuals who wish to challenge existing records can pursue data protection-based deletion requests, potentially supported by the precedent set in the Miller ruling and cases like Kevin Mills's [15]. But there is no automatic expungement mechanism, and the process requires legal knowledge and often legal representation that many affected individuals lack.

The replacement framework remains deliberately sparse. Officials have described a system in which "fewer, more serious incidents" will be recorded, with a "significantly increased" threshold and a requirement for "genuine risk of harm and a clear policing purpose" [1]. What this means in practice — what the new threshold is, who decides whether it is met, what data is retained and for how long — has not been specified in detail.

This vagueness concerns observers on both sides. Journalist Sarah Phillimore warned the abolition could be a "slow death" given the institutional interests invested in the existing system [12]. Fair Cop questioned whether police might "just quietly keep on recording 'crimes' that they don't tell us about" [12]. Hate crime monitoring organizations worry that without any sub-criminal recording mechanism, serial harassers who repeatedly target individuals while remaining below criminal thresholds will go undetected.

International Context

The UK's approach to sub-criminal hate incidents has no direct parallel among comparable democracies. Canada tracks hate crimes through Statistics Canada and the Canadian Centre for Justice and Community Safety Statistics, but recording is limited to criminal offences [20]. Germany requires police to record specific forms of hate crime by law, with civil society organizations publishing supplementary data from their own counselling activities — a model that keeps non-criminal monitoring outside the police system [21]. Australia channels non-criminal complaints through National Human Rights Institutions and state-level anti-discrimination commissions rather than through police [20].

In each case, the common thread is a separation between criminal enforcement and civil monitoring that the UK's NCHI system had blurred. The replacement "common sense" model appears to move closer to this international norm, but without the established civil society infrastructure that other countries rely on.

The Gap That Remains

The abolition of NCHIs resolves a genuine problem — a system that recorded lawful speech in criminal databases, harmed employment prospects without due process, consumed significant police resources, and produced no measurable crime prevention benefit. The Miller ruling, the Policy Exchange report, and the institutional conclusions of the College of Policing and NPCC all pointed in the same direction.

But it also raises a question that no one has fully answered: what replaces the monitoring function for behaviour that is targeted, persistent, and hate-motivated but falls short of criminal conduct? Stalking law evolved from a similar recognition that a pattern of individually lawful acts could constitute criminal behaviour in aggregate. The NCHI system, for all its flaws, was at least an attempt to capture analogous patterns in the hate context.

The government's position is that "common-sense professional judgment" will fill this gap [1]. Whether that judgment proves sufficient — or whether the abolition simply shifts the problem from over-recording to under-recording — will depend on implementation details that remain unwritten. The 137,550 recorded hate crimes in the year ending March 2025 [19] suggest that hate-motivated behaviour in Britain is not declining. The question is whether the state can track the trajectory from hostility to harm without recording the thoughts of citizens who have broken no law.

Reporting based on parliamentary records, government publications, court rulings, and freedom of information data compiled through March 2026.