Crime has returned to the centre of the British political debate – and for good reason. It is the most basic test of whether the state still works.
The government’s upcoming white paper on policing is being billed as a defining moment – setting out how neighbourhood policing, standards, structures, leadership and technology will be reformed for a changing world. The early signals from the government seem promising. But words on paper will not be enough.
When people feel unsafe walking home, when high-street shoplifting becomes routine, when phones are snatched in broad daylight and when fraudsters drain bank accounts with little likelihood of being caught, the picture is visceral. It tells people the rules are not being enforced, that no one is really in charge – that Britain is a country in decline.
That judgement is formed not by politicians in Westminster or by the latest stats from the Ministry of Justice, but on streets, in shops, and increasingly on the basis of the videos, posts and alerts people see every day. It is why crime matters so much politically. Not because every statistic is moving in the wrong direction, but because confidence has drained away.
Baroness Louise Casey put it bluntly: “If the public stop reporting crime because they don’t think the police will take it seriously, we are in real difficulty.” That warning, issued two years ago in our paper on policing reform, has proved prescient.
Police Data Doesn’t Tell the Full Story
The latest crime data present a mixed and easily misunderstood picture, shaped by how crime is recorded and reported.
Police-recorded crime in England and Wales over the past year has remained broadly consistent, at around 6.6 million offences. Some traditional high-harm crimes have declined: homicide has fallen, firearms offences are down and robbery has declined modestly overall. Levels of knife crime are broadly stable. That progress should be recognised and celebrated. Lives saved matter.
But the Crime Survey for England and Wales, which captures both reported and unreported crime, shows something else. Drawing on data gathered directly from citizens, it shows what they have experienced. Measured this way, headline crime rose by 7 per cent in the year to March 2025, reaching 9.4 million incidents – that’s 2.8 million more than the police recorded. The increase noted in the Crime Survey was driven almost entirely by fraud, which surged by 31 per cent to an estimated 4.2 million incidences – the highest figure since this crime was first measured.
This divergence matters because the two data sets measure different things. Police-recorded crime reflects what comes into the system and how it is classed. The Crime Survey reflects what people actually experience. The Office for National Statistics is explicit in stating that police-recorded crime is not a reliable indicator of overall crime trends on its own.
In other words, it is entirely possible for a government to say “crime is down” while millions of people experience the opposite. That gap helps explain why public confidence remains fragile. Only around half the population now say the police are doing a good job, a sharp fall from pre-pandemic levels. The problem is not spin. It is that people do not recognise their lived experience in the numbers being cited.
This gives political leaders a choice: fight crime on the basis of unreliable data or fight the crime people feel.
Fraud Is Now the Frontline of Crime
The clearest example of this disconnect is fraud.
Fraud is no longer a marginal or “white-collar” issue. It is now the single most common crime people in Britain experience. Of the estimated 4.2 million fraud incidents in 2025, bank- and credit-account fraud accounted for well over half of those. Incidences of consumer and retail fraud exceeded 1.1 million.
Yet only around one in eight fraud offences is reported to the police or Action Fraud.
This is mass victimisation on a national scale. It destroys trust not just in policing, but in the state’s ability to protect people in a digital economy – and, frankly, between citizens. And it exposes a policing system built for a different era: localised, reactive and uneven in capability.
Fraud doesn’t respect national boundaries, let alone local ones. Britain’s crime problem in 2026 is no longer primarily violent, local or analogue. It is digital, organised and scalable. Policing must catch up.
Visibility Still Matters: Shoplifting, Theft and the Sense of Disorder
Even the limited picture offered by police-recorded data shows increases in the crimes that most shape public confidence.
Between April 2024 and March 2025, shoplifting rose by around 20 per cent to more than 530,000 recorded offences. Theft from the person increased by around 15 per cent. These are not trivial crimes. They affect everyday life and create a sense that low-level disorder carries no consequences.
In an age of low growth and low productivity, we should be more concerned than ever about their direct impact on consumer confidence and business viability. Think of the immediate shock to a business’s bottom line when stock is cleared out by shoplifters and the costs associated with plugging a gap in enforcement are taken up privately. Or take a busy shopping district, where customers are on high alert for phone thieves as they move from shop to shop. Many police forces are doing excellent work on this front, but the reality is that action is uneven.
Even where long-term trends show declines compared with the 1990s or early 2010s, what people experience right now is what matters politically. A shop owner repeatedly targeted or a commuter who no longer feels safe does not take comfort from an overall trend.
Crime does not have an even impact across society. The evidence shows a clear socioeconomic gradient in who bears the brunt of everyday insecurity. Analysis using Office for National Statistics data shows that around one in four people living in the most deprived 10 per cent of neighbourhoods live in high-crime areas, compared with around one in 30 in the least deprived 10 per cent. That gap is driven largely by visible, street-level offences – shoplifting, theft from the person and other acquisitive crimes – which are far more common in poorer communities and far more corrosive of trust when they go unchecked.
The same pattern holds for serious violence. Knife crime is heavily concentrated in urban areas facing entrenched social and economic disadvantage, particularly among younger people, reflecting repeat harm and limited opportunity rather than affluence. Fraud looks different but reinforces the same inequality. While this crime cuts across income groups, the damage is greatest where financial resilience and digital access are weakest. Weak policing does not just undermine confidence – it entrenches inequality in practice.
This is compounded by the online environment. We are all just one scroll of a personalised newsfeed away from watching a shop being robbed with no visible response. That scene becomes part of people’s reality whether or not it unfolds on their own street.
The test for policing is not just whether crime falls on paper but whether people “feel” this fall. It is whether order is visible in practice. This requires significant structural reform alongside a new approach to engaging with the public.
Trust Is Built When Crime Is Taken Seriously
Crime also hits people in profoundly different ways.
Around one in ten adults experience domestic abuse, stalking or sexual assault each year. For these victims, confidence in policing is not shaped by headline crime figures but by whether cases are taken seriously. Huge question marks remain around the competent gathering and analysis of evidence, especially from digital devices, and whether investigations progress at all. This is compounded by a justice system that is struggling under its own weight.
Here, capability matters as much as presence. Digital forensics and data handling are no longer niche specialisms. They sit at the heart of justice for victims of the most serious crimes.
When cases collapse because evidence is delayed, lost or mishandled, trust does not just fall. It evaporates. We need visible local policing backed by best-in-class national capability.
Neighbourhood Policing Must Be Guaranteed, Not Optional
This is why a renewed focus on neighbourhood policing is so important.
In our 2023 paper Rebuilding Trust and Delivering Safer Communities: A Plan for Reforming UK Policing we argued that prevention had been hollowed out of the policing model and that neighbourhood policing needed to be restored as a universal, protected service. We recommended a “neighbourhood-policing guarantee”, backed by legislation, to ensure safer-neighbourhood teams could not simply be redeployed elsewhere when pressure mounted.
The government’s recent announcement that almost 2,400 additional officers have been brought into neighbourhood-police roles in just six months is a welcome step towards delivering that guarantee. Visible, accessible neighbourhood policing is how trust is rebuilt. When people know who their local officers are and see them regularly, confidence returns.
But this only works if it is a deliberate, lasting change. For too long, policing drifted towards a model where trained officers were pulled into back-office roles and endless administration. That was not a single ideological decision. It was the predictable outcome of incentives that prioritised headcount over where this headcount was actually being deployed. The number of trained uniformed police officers in back-office roles like HR and IT support has surged by 40 per cent in the past six years to more than 12,600, while the number of visible frontline officers has fallen to around 67,000 – down from more than 70,000 a decade ago.
The priority now must be locked in: fewer officers behind desks, more officers on streets – and all supported by tools that let them do their jobs properly.
Modern Crime Requires National Capability – and Faster Rollout of Technology
The biggest unresolved gap in British policing remains capability.
Fraud, cybercrime, child sexual exploitation, and serious and organised crime do not respect force boundaries. Yet the system still expects 43 forces in England and Wales to procure separately, train separately and build specialist capabilities in parallel.
This is inefficient and unrealistic. Nearly half of all crime now crosses force boundaries or happens online. Reducing the number of forces by strengthening the regional tier and creating a single national police force to lead on serious and organised crime, cybercrime and counterterrorism is the way forward. This is what we set out in 2023 and what must now be delivered at pace. Criminals do not respect force boundaries. Neither should core policing capability.
We also recommend the creation of a national digital forensics agency, because digital evidence now sits at the heart of most investigations and capacity varies wildly across the country.
The same logic applies to technology more broadly. Take live facial-recognition technology: only 13 police forces currently use it. The technology has rolled out at a glacial pace in the past two years and yet, where it is used, the results are striking.
In London, the Metropolitan Police has made more than 1,300 arrests using live facial recognition over the past two years, including arrests for rape, domestic abuse and serious violence. Around a quarter of those arrests in 2025 related to offences involving violence against women and girls. The technology has also identified more than 100 registered sex offenders breaching licence conditions.
In Croydon alone, live facial recognition has led to more than 70 arrests, including individuals wanted by police for years. This is not speculative innovation. It is effective, and results in dangerous people being taken off the streets – to deny this technology purely through reasons of competence and geography is a social injustice.
The same rings true with routine tasks. Approximately 97 per cent of police technology budgets are spent on maintaining existing legacy systems rather than investing in innovation or new technology. Police should not be spending time doing anything that AI could do – from taking notes to filing reports. A minute wasted on a task like this is a minute not spent on the front line.
This is not about replacing policing with technology. It is about using technology to make policing more effective, freeing officers’ time, improving detection and ensuring that where you live does not determine the level of protection you receive.
Crime does not stop at the M25, and neither should the tools used to tackle it. Tackling crime requires a whole-of-country effort. Facial-recognition technology must urgently be rolled out UK-wide, and other innovative tech must be utilised to stop criminals before they strike.
Combined with national standards that ensure privacy is protected and fundamental rights are respected, a national police force with centralised procurement will enable economies of scale. This will allow the faster rollout of proven technology, taking tools that work and deploying them consistently and fairly across the country.
Standards, Accountability and Visible Competence
Trust will not be rebuilt if failure is simply observed and left to be repeated, rather than addressed and fixed for good.
In 2023, we argued that the standards regime in policing was too weak. Inspectorate findings lacked teeth. Underperformance lingered. The public saw problems identified but not resolved.
That is changing. The Crime and Policing Bill that is progressing through Parliament will give police chiefs enhanced authority to dismiss officers who fail vetting or are unfit for duty. This has to be combined with stronger inspection, and backed by consequences for failure and transparent performance data that people can understand. Far from technocratic fixes, these are examples of measures that will build trust and legitimacy in the police.
Right now, His Majesty’s Inspectorate of Constabulary and Fire & Rescue Services (HMICFRS) inspects and reports on police effectiveness and efficiency, and forces are expected to respond to recommendations. But HMICFRS does not currently have the statutory regulatory powers to compel forces to act on its recommendations or requests for leadership change as a matter of course. As we argued before, a new standards regime should be introduced, with real regulatory powers for HMICFRS so that forces are required to act on its recommendations.
Policing as the Test Case for Public-Service Reform
There is a bigger point here.
Across public services, Britain is struggling with an outdated operating model: reactive, labour-intensive and fragmented. Expectations have moved on. The state has not kept up.
At the Tony Blair Institute for Global Change, we have argued for a new model of public services that are powered by data to become personalised, preventative and always on. Policing is where this theory must meet reality.
This should start with the citizen and a simple question: what is their front door to policing today?
If you report fraud, do you know what happens next? Can you track it? Do you get updates? Or do you disappear into a void?
What is needed is a unified national policing app – a single digital front door for citizens in need of help from the police. It should provide comprehensive victim support and allow people to track the status of reported crimes in real time. This is not about replacing hands-on policing with an app, but about making the system visible, navigable and accountable. It is ultimately about respecting victims and their right – their need – for information. In a digital economy, citizens should not disappear into a void when they report a crime.
A modern policing system should feel more like a high-quality, responsive urgent-care system: a clear point of entry, transparent triage, real-time updates and early intervention that stops the next crime, not just records the last one.
This is not about flashy innovation. It is about applying the same logic we expect to be applied everywhere else: use data properly, intervene early, free professionals from pointless administration and focus human judgement where it matters most.
The Test Ahead
The reforms now being debated matter because they point in the right direction: restoring neighbourhood policing, rebalancing priorities, strengthening capability and using technology properly. The question is whether this ambition succeeds.
The tests are simple – and the answers will be evident on the ground:
Is neighbourhood policing genuinely protected and visible?
Is national capability for combatting fraud, cyber-crime, and serious and organised crime built at pace?
Are technologies proven to catch criminals rolled out faster and more consistently?
Are standards enforced and failure dealt with?
Does every citizen get the protection they need, regardless of geography?
And above all, does policing feel more present, more responsive and more effective in people’s lives?
If the answer to all of these questions is yes, confidence will return and the most important measure of them all – the level of crime people actually feel and experience – will improve.
If the system settles for incrementalism, it will not.
Crime is fast becoming a signal of national decline. Backed by robust reform, it can become the clearest signal of a state that works. Fix policing, and you begin to restore trust. Fail, and the consequences will be felt far beyond policing itself.