16 October 2006 1 Randomised Control Trial on the London Buses Sgt Ben Linton (MPS) Tim Herbert (TfL)
Dec 24, 2015
16 October 2006 1
Randomised Control Trial on the London Buses
Sgt Ben Linton (MPS)
Tim Herbert (TfL)
Why a trial
Introduce the concept of Evidence Based Policing to the Safer Transport Command
Demonstrate that research can take place in operational environments without compromising performance
Build on partnership working between the MPS and TfL
Gain experience to assist future research
Demonstrate that the tactic works
Why an RCT
Relatively easy to explain
Strength of findings
Rigour required demanding commitment to design
Why a hotspot RCT
16 October 2006 4
19500 bus stops in London – cannot police them all
Builds on Op Beck conducted on the London Underground
Hotspot policing works
5
• Chris Koper, Minneapolis, 1995. Observed 17,000 instances of police presence at hotspots, and 4000 instances of disorderly or criminal behaviour.
• Measured impact of police presence on survival time – how long between police leaving and the next occurrence of crime/disorderly behaviour.
• Less than 10 minutes, no impact. More than 15 minutes, no added benefit
• 15 minutes – longest "survival" time – optimal time to police hotspots
Koper
6
Renee Mitchell Police Sergeant Sacramento (California) Feb-May 2011
• Crime reduction of 25% at targeted hotspots.
• Officer productivity increased.• Cost benefit of approx $300,000.• No identified negative impacts of
tactic.
Practical Implementation
London’s bus network
• 700 bus routes
• 19,500 bus stops
• Approximately 7,500 buses • More than six million
passengers each weekday
• 2014 year of the bus http://www.ltmuseum.co.uk/whats-on/year-of-the-bus
What do we know about bus-related crime?
High crime bus routes pass through high crime areas (Pearlstein & Wachs 1982; Newton 2008)
Bus stops are crime generators (Roman 2005; Weisburd et al. 2012)
Crime is concentrated at a small number of bus stops (Loukaitou-Sideris 1999, Newton & Bowers 2007)• 0.05% (n=10) of bus stops in Los Angeles accounted for 18% of bus stop crime. • 20% of all shelter damage in the Wirral occurred at 2.5% of all shelters.
Crime is more common at bus stops than on board buses (Levine 1986)
Bus-related crime and disorder in London
Driver Incident Reports Bus stops
% Count % Count
10 2,322 0.3 6420 4,645 0.9 18550 11,615 5.2 1,027100 23,245 37.7 7,391
Spatially concentrated
• 30 bus routes accounted for approximately 20% of DIRs in 2013.
• 300 bus stops were associated 27% of nearby DIRs* (n=6,204, see map) in 2013.
* DIRs recorded during 2013 (n=45,324) that were matched with iBus route run data 59.5% (n=26,954) and were within 200m of a corresponding bus stop (n=23,245).
12 or more DIRs at or near a bus stop during 2013
Bus-related crime and disorder in London
Site identification
• ‘Community Safety’ Driver Incident Reports enriched with iBus route run information
• Use F.M.E. to pass attribute data between DIRs and the nearest bus stop with a corresponding route run
• Identify ‘hot’ bus stops by the frequency of DIRs, their temporal stability and CAD reports
General hypothesis
Directed police patrols at high crime bus stops between 12-1900 hours will reduce crime levels compared to business as usual
Research design
• Multisite randomised controlled trial across London• Random assignment of sites to 3 blocks to Treatment and Control• Intervention delivered by pairs of PCs and PCSOs • 3 x 15-minute patrols, Monday-Friday 12-1900 hours at treatment sites
with on board patrols along contiguous bus route runs• “Business as usual” at the control sites• Duration: 3–6 mths
16 October 2006 14
Bexley DA7 4Take route 229 to Forresters Crescent (74278)
Supervision, monitoring and motivation
• 1 Inspector, 2 sergeants• GPS devices issued to each pair of officers• Detailed patrol patterns• Weekly briefings
GPS data
16
Things We Learned
• Timescales. Almost 12 months from inception to data gathering.• Data sharing – legal aspects and security• Data processing – CPU• Dedicated team – ensure consistent treatment and commitment• Testing phase – longer than expected. Assuming things about data. Combined with
testing phase.• Biting off more than you can chew – pan-London deployments – time, radio channels,
travelling, local issues. Would have been much easier to be smaller scale. • Practicality – research in the real world.• Partnership – with academics or partner agencies. Benefits far outweigh the legal or
contractual issues.
17
Thank you
[email protected]@tfl.gov.uk