Crime mappers are doing it for themselves
Today in the Guardian’s Technology section Heather Brooke – who was one of the key drivers behind getting access to MP’s expenses – writes (as part of the Free Our Data campaign) in Met keeps crime stats under lock and key about how the Metropolitan Police insist that (a) they’re not going to release data for crime mapping (b) even if they did, they keep it amalgamated on such a level that it wouldn’t be any use to anyone.
The Met also cites privacy as a reason not to release location specific crime data. Yet the Data Protection Act does not prohibit personal information being disclosed, even if one considers anonymised crime reports “personal”; and Boris Johnson’s pledge was only ever to publish crime data by street level, not by exact address. The law’s purpose is to ensure that disclosure is for a legitimate purpose. State-mandated ignorance benefits no one.
Crimes are not a great secret, particularly not violent crimes – such as the spate of stabbings in the UK in recent months – though without access to the raw data, how can we know how and where it’s rising? [Richard] Pope [of planningalerts.com] thinks the main problem is that the police are not technically savvy, citing an encounter at a meeting between locals, the council and the police where the Met admitted it couldn’t provide incident detail broken down by area – so the council ended up paying the Met just to get this information.
But people aren’t necessarily waiting for the police. Take this mashup generated by MapMan which looks at that topic du jour, knife crime.
Via the Digital Urban blog, here’s London Teenage Murders 2007, Knife Assaults and Regeneration Areas: Mapped – A Clear Pattern Emerges:
Created using Google MyMaps the list has been compiled via various websites (such as http://www.capitalradio.co.uk/article.asp?id=532062) with street names identified in related press articles and plotted on the map. Actual position within the street will not be accurate, but the street names themselves should be. Note the map relates to all murders, not just knife related incidents.
Using MapTube [URL corrected] the map can be overlaid with other data sets, such as a map uploaded detailing assault using a knife or sharp objects extracted from all hospital admissions (2007). The map is based on data with a cause code of ICD-10 X99 (assault by sharp object) and excludes all codes that may indicate accidental injury (ICD10 – W25, W26), self inflicted (ICD10 – X78) and undetermined intent (ICD10 Y28).
Figures are directly age standardised per 100,000 population with CI’s – Actual counts were excluded in the map due to disclosure surrounding low numbers. By overlaying the two maps you begin to get a picture of the extent of knife crime and the number of murders in London.
Each link is clickable for more information. Such data should really be available via either the http://www.london.gov.uk/ or http://www.met.police.uk/ along with other locations of crime in the city. It may be alarming to see such incidents mapped but this is the city we live in and the public should have a right to view exact locations of crime in their neighbourhoods.
There’s plenty more: they then overlay urban deprivation and find an interesting correlation with the number of teenage murders.
OK, so you might find that obvious. But it also tells you where the energies need to be focussed – and whether parents in Hampstead or Notting Hill really need to worry about the possibility of their child being a victim.
(One other thing: the gender of the victims. I suspect it’s overwhelmingly male too, isn’t it?)
Anyhow, this is all stuff that’s been done at zero cost to the police. Maybe if they think they’re overcome with data, we could help them out some more. Make the data available for free, and we’ll help you for free.
(crossposted with the Guardian Technology blog)
- The following posts may be related...(the database guesses):
- BBC's iPM looks at crime mapping in Chicago (17 June 2008; score: 33.16%)
- Home Office responds re OS and crime maps (21 November 2008; score: 28.49%)
- Is the Information Commissioner against crime mapping? Actually, no (2 June 2008; score: 27.56%)
- Crime mapping coming more widely as government gets on board (18 June 2008; score: 27.24%)
- Ordnance Survey says Met Police crime maps break its licence. Does Jacqui Smith know? Or Gordon Brown? (19 November 2008; score: 26.68%)




July 18th, 2008 at 7:54 am
Maptube url should be http://www.maptube.ORG maybe?
July 18th, 2008 at 8:00 am
@David – you’re right. Corrected. Thanks.
July 18th, 2008 at 3:58 pm
(One other thing: the gender of the victims. I suspect it’s overwhelmingly male too, isn’t it?)
As I completed the analysis on hospital admissions for knife assaults and uploaded the thematic maps to CASA – I can positively confirm that Males = 92% / Females = 8% of all hospital inpatient admissions for 2006/07. These represent the most serious of cases whereby a patient is seen by a consultant rather than treated in A & E -
July 19th, 2008 at 6:08 pm
[...] Guardian’s Free our Data campaign — everyone should have one — praises a crowdsourced effort to collect the data that, for some odd reason, London police won’t [...]
July 23rd, 2008 at 4:15 pm
I think there is a danger of over simplifying the issues around the delivery of highly granular crime statistics.
See dialogue at http://giscussions.blogspot.com/2008/07/pinpointing-crme.html
July 28th, 2008 at 9:11 am
Heather Brookes claims that the use of GIS in crime fighting hasn’t made its way into the UK. She should check the wide range of case studies from around the country being presented at the 6th National Crime Mapping Conference in Manchester this week – http://www.uk-cmc.org/.
There is a marked difference between the police using GIS internally to map crimes, and the provision of similar information to the public via the Internet. Publishing crime statistics to the public is a complicated process, and the suppression of victims’ identities has to be of paramount importance. Heather’s implication that street level mapping is easily available in the US is misleading, as it is available for Chicago but not everywhere. Indeed, street level mapping is clearly inappropriate for sparsely populated rural areas where there are few residents per street or even per postcode, risking the identity of the victim.
If Heather had looked beyond London, where she noted the absence of online crime mapping, to other areas, she would have found that incidents other than property crimes are indeed mapped and published. The residents of the West Midlands, for example, are able to view crimes by category, such as assault or robbery, at http://www.myneighbourhood.info.
It is essential that the police communicate accurate statistics to the public in an easy to understand format. In many cases, the public’s perception of crime is worse than the reality and an online map can offer reassurance. This will not always be the case, which is why crime statistics should always be displayed alongside supplementary information about preventing crime and building safer communities.
Ben Allan – MD – Dotted Eyes Ltd
August 21st, 2008 at 2:51 pm
I thought your readers might be interested in the NI Crime Map we created http://banjax.com/sandbox/crime