Safe For Whom?
Every time crime in the U.K is discussed, especially London, the same pattern emerges. Politicians, commentators and campaigners begin citing falling murder figures. Critics point to knife crime. Yet for many women, the argument often feels as though it is happening in a parallel universe.
In November 2018, Sadiq Khan described London as “one of the safest cities in the world for women”, whilst also acknowledging that women and girls were being raped and sexually assaulted every week. Seven years later, in April 2025, he stated that “it’s simply not right that in 2025, women and girls across the globe still live in fear, in their homes and on our streets”. A city can be safer than many others whilst still failing to be safe. The question is whether the statistics support the narrative that London remains one of the safest cities in the world for women, because once we begin examining the figures, the picture becomes considerably more complex.
In 2025 there were indications that a rape was being reported in London approximately every 54 minutes. Yet the public conversation surrounding rape statistics remains strangely contradictory. We are repeatedly told that recorded rape and sexual offence figures are increasing because victims feel more able to report offences, because awareness has improved and because police recording standards are better than they once were. The Office for National Statistics has repeatedly warned that police-recorded sexual offences should not be viewed as a simple measure of offending because changes in reporting and recording practices have significantly influenced the data. That is an important caveat and any honest discussion of the figures has to acknowledge it. At the same time, it has increasingly become the explanation offered whenever uncomfortable statistics emerge. If rape reports rise, reporting has improved. If victims come forward in greater numbers, confidence has improved. But if the rapes and sexual offences increased too can it not also be said that confidence improved within predators too?
A question put to the Mayor of London referenced a 193 per cent increase in reports of rape over a ten-year period. The response acknowledged that both the Mayor’s Office and the Metropolitan Police recognised the substantial increase in reports of rape and sexual offences. Separately, a London Assembly analysis found that recorded rape offences in London were 140 per cent higher in 2018 than they had been in 2011, whilst total recorded sexual offences had increased by 96 per cent over the same period. These are enormous changes. Even if improved reporting accounts for part of the increase, the scale remains remarkable. The trend has continued in recent years; London recorded 26,803 sexual offences during 2024/25 compared with 24,334 the previous year. Metropolitan Police figures released in early 2025 showed reported sexual offences increasing by a further 7.4 per cent over the preceding twelve months. Whatever explanation is offered, and there are many, the trend toward sexual violence against women and girls (VAWG) remains difficult to ignore.
London is often presented as a success story in crime reduction. In some respects it probably is. Murder rates are lower than many comparable cities around the world. Yet murder is not necessarily how women experience danger. Most women are not walking through a train station late at night calculating their chances of being murdered. They are thinking about whether somebody is following them. They are wondering whether it is worth paying for a taxi rather than waiting on a platform. They are checking who gets into an empty carriage. Walking closer to other women so there is someone to raise the alarm. They are texting somebody to say they got home safely, sharing their location with friends. They are choosing routes based on lighting rather than convenience, moving seats when a stranger’s behaviour becomes uncomfortable. Taking a pair of flats in their handbag so they can change shoes, enabling them to run if needed. They are making hundreds of small calculations that most men never even notice.
This distinction matters because it changes the entire discussion. When politicians talk about crime, they often talk about murder because murder statistics are comparatively straightforward. A murder is generally recorded as a murder. The figures are easier to compare over time and they produce clear headlines. Sexual offences are different because they are underreported, subject to changing definitions, influenced by awareness campaigns and recording practices. Yet they are also far closer to how many women assess risk in their daily lives. A city can become safer in terms of murder whilst simultaneously leaving large numbers of women feeling unsafe in their everyday lives because VAWG is rising.
The issue becomes even more troubling when reporting rates are considered. Sexual offences have always been significantly underreported and continue to be. According to the Office for National Statistics, only around 14.7 per cent of victims who experienced rape or assault by penetration reported the most recent offence to police. Put another way, more than eight out of ten victims did not report. That figure may be the most important statistic in the entire discussion because it means that all the numbers being debated publicly are already incomplete. When London records more than 26,000 sexual offences, it is not recording everything that happened. It is recording what was reported. The majority remains hidden. That raises an uncomfortable question. If only around fifteen per cent of victims report, what confidence should we place in any attempt to describe the true scale of sexual violence? Nobody really knows. We can only estimate, but there remains a substantial gap between what is experienced and what is officially recorded.
That hidden gap becomes easier to understand when we consider the sorts of incidents that often go unreported. Many women will recognise behaviours that never reach a police station. Being followed home, being cornered on public transport, being touched inappropriately in a crowded venue, being photographed or filmed without consent in public or in private. Being harassed at work, being pressured into sex within a relationship. Being subjected to behaviour that is deeply uncomfortable but which seems too minor, too difficult or too exhausting to report. Some of these incidents are criminal offences. Others exist in a grey area between criminality and intimidation. Some happen within relationships. What they have in common is that they contribute to how safe women feel. They shape behaviour, alter routines and influence decisions long before anyone documents anything, describing in their own way how an event is interpreted.
The legal landscape itself has changed dramatically over the period many people are trying to compare. Marital rape was not recognised in English law until the landmark R v R decision in 1991. The Sexual Offences Act 2003 significantly modernised sexual offence legislation and placed consent at the centre of the legal framework. Recording practices have also changed. In 2016, rape offences began to be recorded differently, with greater emphasis on counting offences on a per-offender basis. These changes matter because they make comparisons across decades more difficult than many commentators acknowledge. Yet they do not make the figures meaningless. If anything, they reinforce the need to approach simplistic political claims with caution.
The issue becomes even more complicated when we ask a deceptively simple question: what exactly is London? Many people assume London is a fixed geographical area. In reality there are several overlapping definitions. There is Greater London. There is the Metropolitan Police District. There is the City of London, which operates its own police force. Boundaries have changed repeatedly over the decades. Some datasets include the City, others do not. Some compare boroughs or policing areas. The Metropolitan Police District itself covered different geographical areas before being aligned with Greater London in 2000. Crime figures are often presented as though they are directly comparable across decades when, in reality, the population being measured, the area being measured and sometimes even the way offences are counted have changed. This does not make the statistics worthless. It simply means that caution is required whenever somebody claims the figures provide a simple story.
What is perhaps more interesting is that caution cuts both ways. If changing recording practices can be used to explain increases in crime, they can also be used to explain decreases. If changing boundaries make rises difficult to interpret, they also make falls difficult to interpret. The wider national picture provides useful context. The Office for National Statistics estimates that around 739,000 women experienced sexual assault during the year ending March 2025. More significantly, prevalence appears to be increasing. In 2015, around 1.7 per cent of adults aged 16 to 59 reported experiencing sexual assault during the previous year. By 2025 that figure had risen to 2.4 per cent. This matters because it is not measuring police reports, it is measuring victimisation. People are being asked directly about their experiences whether they reported them or not. If reporting changes were the sole explanation for rising figures, we would not necessarily expect victimisation surveys to show the same trend.
Nor is London unique. Some of the highest rates of recorded sexual offences in England and Wales are found outside the capital. Cleveland has consistently recorded some of the highest rates in the country. West Yorkshire, Northumbria, Greater Manchester and Avon & Somerset have all recorded rates that rival or exceed those seen in London. These are very different places with very different demographics, policing challenges and populations, yet they are all grappling with similar concerns. The significance of this is that it changes the nature of the argument. If the same pattern is appearing across large cities, smaller cities, mixed urban-rural areas and entirely different police force regions, then the explanation is unlikely to be unique to London and the claim that the style of reporting or boundaries or some other excuse does not hold water.
South Yorkshire is not exempt. Sheffield is frequently regarded as one of England’s friendlier and more liveable cities. It performs well in quality-of-life surveys and is often praised for its green spaces, affordability and sense of community. Yet sexual offences remain a significant concern there too. The same pattern emerges repeatedly: underreporting, rising recorded offences and continuing concerns about what happens after victims engage with the criminal justice system. The events of the last two decades have reduced women’s faith in the legal system to bring predators to justice.
Scotland presents a similar picture. Despite operating under a separate legal system, violence against women and girls remains a major policy concern. Sexual crime has increased significantly compared with historic levels and much like England and Wales, explanations frequently centre on improved reporting and increased confidence amongst victims. Yet the same unanswered questions remain. If reporting is improving, why do so many victims still remain silent? If confidence is improving, why do so many cases fail to result in meaningful outcomes?
This brings us to one of the most uncomfortable statistics in the entire debate. In March 2026, a London Assembly member asked why the positive detection rate for rape had fallen from 27.6 per cent in 2015 to 11.6 per cent in 2026. At the time of writing, a formal answer had not yet been published. Regardless of the eventual explanation, the question itself is revealing. If the figures cited are correct, then whilst reported rape has increased substantially, positive outcomes appear to have fallen dramatically. The implication is not necessarily that more rape is occurring. The evidence does not allow us to make that claim with confidence. What it does suggest is that reports have increased whilst successful outcomes have become more difficult to achieve. For victims considering whether to come forward, that matters. For confidence in the criminal justice system, it matters even more.
Perhaps the truth is that several things are happening simultaneously. Reporting, awareness and recording is improving. However, sexual violence may also be far more common than many people would like to believe. This is why the debate about crime so often feels unsatisfactory because we are frequently asked to choose between believing crime is rising and believing crime is falling. In reality, both can be true depending on which crimes are being measured and how they are measured. Politicians point to murder statistics and declare success. Critics point to certain sexual offence figures and argue the opposite. Both sides can produce evidence supporting their position. The real question is which statistics matter most to the people living with them.
Perhaps the question is not whether London is one of the safest cities in the world for women. Perhaps the question is how we would know? Because when recorded rape has increased by more than 140 per cent over a relatively short period, when more than eight out of ten victims still do not report, when sexual offences continue reaching record highs, when victimisation surveys suggest the trend extends far beyond London and when detection rates appear to be moving in the wrong direction, many women may reasonably conclude that the statistics politicians choose to talk about are not the statistics that shape their lives.