London High Street Crime Data May Be Distorting Risk Scores
Insurers pricing risk for London high streets may be relying on data distorted by changes in police recording methods rather than actual fluctuations in street-level safety. A study by Seenity, a provider of risk intelligence software, analyzed 488 London high streets across two periods: 2016–2019 and 2023–2026. The findings indicate that 29% of these locations shifted into different risk bands between the two timeframes.
Seenity’s research highlights a discrepancy between crime volume and crime severity. While burglary—a heavily weighted metric in the firm's index—dropped by 24%, other crimes saw significant increases. Shoplifting rose by 74%, and theft from the person increased by 53%. Overall crime harm, which accounts for severity, rose by 7%. Other metrics also showed declines: non-residential fires fell by 17%, injury road collisions decreased by 16%, and nitrogen dioxide levels dropped by approximately 34%.
The study identifies three specific administrative changes that influenced crime records. In July 2022, police.uk updated anonymized crime locations across the capital. In February 2024, the Metropolitan Police transitioned to the CONNECT records management system, which Seenity estimates shifted 1,500 offenses monthly between theft categories. Additionally, updated guidance on recording shop theft involving threats or violence has led to these incidents being classified as robbery rather than theft.
Oren Atia, CEO of Seenity, noted that such administrative adjustments can create statistical trends that do not reflect reality on the ground. For the analysis, Seenity utilized 13 million police.uk street-level crime records, 287,000 road collision records, London Fire Brigade incident data, and the Greater London Authority’s atmospheric emissions inventory. To ensure accuracy, the firm tested its findings against various catchment definitions and hazard weightings, removing streets that did not withstand these validation checks.
The findings suggest that underwriters relying on pre-pandemic location data may be operating on outdated assumptions. A pricing model based on incident volume rather than severity may overstate risk in areas where burglary has declined while simultaneously understating risk in areas experiencing surges in personal theft.
For commercial property and retail underwriters, this study underscores the necessity of scrutinizing the underlying methodology of location-based risk scores. As police recording practices evolve, insurers should assess whether their pricing models are capturing genuine shifts in risk or merely reflecting changes in administrative data classification.
Why it matters
Inaccurate risk scoring based on shifting police data could lead insurers to misprice commercial property premiums and coverage on London high streets.
FAQ
What is the main finding regarding London high street risk?
The study suggests that changes in how police record offenses are distorting risk scores, rather than reflecting actual changes in street-level safety.
How did crime metrics change between the two study periods?
Burglary fell by 24%, while shoplifting rose by 74% and theft from the person increased by 53%.
What administrative factors influenced the crime data?
Factors included updates to anonymized crime locations by police.uk, the Metropolitan Police's move to the CONNECT records management system, and revised guidance on recording shop theft involving violence.
Source: Insurance Business UK. See the original source for full details.
Scan with your phone
Scan with your phone