London's Commuter Risk Corridor Gap
Anthony Hegarty MSc DSRM Risk & Crisis Management
(Pre-read Note: VAWG = Violence Against Women & Girls)
.....................................................
Raw crime statistics are available to anyone. What they rarely come with is a shape;
which line?
which station?
which hour?
which category of offence?
That shape is what turns a dataset into something an organisation can actually act on. This piece applies that approach to London's commute:
the Underground
the bus network
and the street
...over the past two to three years.
The Tube
Sexual offences on the London Underground reached a five-year high in 2024/25, with 595 recorded across all lines; still below the 2019/20 peak of 776, but the highest since. Between April and October 2025 alone, 341 sexual assaults were reported. Since April 2015, the Underground has recorded 6,475 such incidents, an average of 1.7 a day.
The risk is not evenly spread. Three stations:
King's Cross St Pancras
Oxford Circus
Tottenham Court Road
...consistently top both the incident tables and the passenger-footfall tables.
The Central and Northern lines, London's two busiest, each exceed 1,000 recorded assaults over the longer period. The Tube accounts for 52% of all crime recorded across the TfL network, despite carrying a shrinking share of its passengers.
The bus network
Buses are the fastest-moving risk environment on the network, not the Tube. Sexual offences against women and girls on buses rose 28.6% in 2024/25, after a temporary dip; the sharpest single-year increase of any TfL mode, and enough to outpace the Underground's own longer-term rise (745 to 856 offences, 2022/23 to 2024/25). The structural read is straightforward: buses are less staffed and less enclosed than a station platform, and the data is consistent with that.
Non-sexual violence on the bus network is a different profile again; 63–70% of it is harassment or common assault without injury, concentrated on weekday afternoons (3–6pm) and linked to younger passengers, with the rarer serious violence clustering late at night near nightlife venues. One improving figure sits in the same dataset: bus robbery fell 26.5% year-on-year.
The peak window for non-sexual bus violence, weekday afternoons, 3 to 6pm, overlaps with the school run and the after-school gap this category is built around. TfL's data doesn't isolate school-age passengers specifically, but the timing sits squarely inside the hours a working parent has no visibility over.
The street
Away from transport, London's street-level risk is a different threat type entirely. Knife crime hit a record 16,344 offences in 2024/25; up 58.5% since 2021; before a subsequent 7% fall. Robbery rose 8.7% to 54,064 offences in the same window. Of recorded knife crime, 62% is robbery-driven; rape and sexual offences make up under 2%. Sexual offences citywide have held near 25,000 a year for three consecutive years.
The concentration here tracks geography rather than mode: Westminster records London's highest overall crime rate; Croydon the highest volume of violence against the person; Newham the highest knife-crime volume; Lambeth and Westminster sit among the top three boroughs for sexual offences. Outer boroughs; Richmond, Kingston, Sutton, sit at the bottom of every one of these tables.
What holds across all three
Reporting does not translate into resolution, in transit or on the street. Only 3% of VAWG offences across the TfL network result in a charge or summons; 58% identify no suspect at all. Street robbery clears at a similar rate; roughly one in twenty. The bottleneck sits downstream of the offence, not at the point of reporting.
The practical implication for anyone responsible for people whose journeys pass through these spaces: transit risk and street risk are not one continuum, and volume alone is a poor proxy for exposure.
A commute through King's Cross at 8am carries a different risk profile to a walk through Croydon at 11pm, different offence types, different actors, different points of intervention. Treating "the commute" as a single undifferentiated hazard obscures the corridor-level detail that actually determines where an organisation's duty of care is being tested.
If you're responsible for people whose journeys pass through spaces you don't control — shouldn't we talk?
Anthony Hegarty MSc Criminal Psychology & Criminology
DSRM Risk - Leader
Sources: British Transport Police, Transport for London, Metropolitan Police Service, Office for National Statistics, London Assembly Police and Crime Committee, Policy Exchange.
Comments