Science1 publisher3 min readPublished
Living in London cuts the public's crime overestimate by only about a quarter
A King's College London study with YouGov measures how far British beliefs about the capital's crime and Muslim population are from the recorded figures, and finds the errors run in opposite directions by party.
The Scientist · Science desk
What happened
- The Policy Institute at King's College London put questions about the capital to two YouGov samples of 2,164 and 3,969 British adults between 23 and 27 March 2026, for its Mistaken Britain series.
- Half the public named London the worst region in England and Wales for violent crime in the year to September 2025, and 37% said the same of sexual offences; London recorded the lowest violent crime rate of any region.
- A third of the public, 34%, agreed there are no-go areas of London where Sharia law dominates and non-Muslims cannot enter, while 36% disagreed and 19% said they did not know.
- Seventy-nine per cent correctly said shoplifting had risen in London and 71% said the same of theft from the person; both have more than doubled since 2019.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- constraint Because the two large errors point in opposite directions, one corrective figure on London's demography moves Reform supporters toward the recorded number and Green and Labour supporters away from it.
- decision Anyone arguing about London crime has to choose between trend claims and level claims, because the public already agrees that specific offences are rising and disagrees that the total has fallen.
- precedent The series measures error and what it travels with, so citing it as proof that facts cannot change minds would claim more than these surveys were built to support.
- exposure Retailers and hospitality in the capital face a belief about theft risk they cannot rebut with a rate, since the study itself says the risk behind the 2,000 reported watch thefts is difficult to calculate.
The closest thing here to a control group is Londoners themselves. Asked what share of Londoners had personally been a victim of crime in the year to September 2025, the public's median guess was 30%, against a recorded 12.2% [3]. Residents of the capital guessed 25% [4]. Living in the city therefore cuts the overestimate from 17.8 percentage points to 12.8, removing about 28% of the error [2]. On fear of crime the correction is larger and still partial. The public put the share of Londoners who do not feel safe walking alone after dark at 60%, well above the recorded 37%, and residents guessed 50% [10]. Residence closes about 43% of that gap [3].
The party pattern runs in both directions. Reform UK supporters sat furthest from the facts on crime and on the size of London's Muslim population. They guessed that as much as half the capital is Muslim [16], more than three times the recorded 15% [5]. Green and Labour supporters erred in the other direction, overestimating how White British London is [17].
What the two surveys did not do is test what correction does. They are cross-sectional, 6,133 adults in total [1], each answering once. The Policy Institute reports that misperceptions about London track how warmly people feel toward the capital and what they think of its direction of travel [18]. That is an association. Whether an inflated crime estimate cools someone on London, or a cool view of London inflates the estimate, this design cannot separate.
Accuracy varies by question. Alongside the accurate readings on shoplifting and theft from the person, 58% of the public correctly said sexual offences and violence against the person had risen [9]. The same public got the aggregates wrong: 62% thought total crime in London rose between 2019 and 2025, when it fell by 1.4% [6], and 40% thought the number of murders had risen, when murders fell by around 15% [7]. Nobody was asked where their estimates came from, so the pattern hints at how people assemble them without showing it.
Two further gaps fall outside crime and demography. London pays more in taxes than it receives back in public spending, and the study finds the public has little idea the capital is the key contributor to public finances [19]. London also tops a global ranking of best cities, and the public underestimates its international standing [20].
The luxury watch question shows what a missing denominator does to a number. Sir Jim Ratcliffe said he "can't wear a watch in London" because it is likely to be stolen [11], and seven in ten of the public agree it is unwise to wear one in the capital [12]. The study sets around 2,000 luxury watches reported stolen in 2023 against that. It says the actual risk is difficult to calculate, and reads the count as evidence such thefts are relatively rare compared with the level of public concern [13].
What to watch
- Whether the Policy Institute follows these descriptive surveys with a randomised correction experiment that measures whether accurate figures move attitudes.
- Publication of full crosstabs showing how far apart Reform, Green and Labour supporters' estimates are on the same question.
- Whether a later count of luxury watch thefts arrives with a denominator that turns the 2,000 figure into a risk per wearer.