by Julian Goldsmith

Police analytics may get a boost from commissioners

Oct 10, 20124 mins
IT LeadershipIT Strategy

Although shows like CSI and Numbers would have you believe using analytical technology is commonplace in crime fighting, the practice is still in its infancy within many law enforcement organisations across the world.

The UK is in no way a leader in this field and the regional constabularies that make up law enforcement in this country have a patchwork record of using technology to uphold the law and detect crime more efficiently.

Much of this is to do with the federated way law enforcement is organised in the UK and the culture within the police forces in terms of attitude to the usefulness of technology.

Analytics is viewed more or less as dark art that has no place in day-to-day police work.

Retired Greater Manchester Police superintendant Keith Bentley, who was speaking at an IBM briefing on crime analytics, summed the situation up neatly by describing it as a choice between local policing and centrally-enforced standards on fighting crime.

There are still no consistent collaboration policies between police forces and with other agencies, such as local authorities. Initiatives which are lucky enough to find influential sponsors rarely flourish after that sponsor has moved on.

More than this though, Bentley suggests that the resistance to direct-appointments at senior levels throughout UK policing has led to an aversion to novel ways of detection, in favour of protecting budgets allocated to maintaining staff levels.

All this could change though, with the election of new police commissioners next month. These people may not necessarily have come through the ranks and may be more willing to give 21st century methods a try. They will also be keen to implement processes that quickly bring crime rates under control. As elected officials, they are much more accountable to the local citizenry than career police officers, and will need to show voters rapid results to be sure of re-election.

According to Bentley and Ron Fellows, Global Lead, IBM Crime Information Solutions, there are a growing number of examples where the use of crime analytics has provided tangible benefits.

Fellows cited an IBM project with Hertfordshire police where collaboration and data integration led to a 10 per cent increase in crime detection. But this is the lowest level of using data to fight crime. Other forces he said, have gone further to use data to predict where crime is likely to occur.

For instance, Fellows worked with police in Tucson, Arizona to analyse historic crime data, integrated with geographical information to more effectively place officers where crime is likely to be committed. It resulted in a 60 per cent reduction in drug-related crime in one community over 12 months.

The system draws from a variety of data sources to match individuals to reported crimes with a degree of certainty.

Bentley insists that there are some notable IT successes in UK policing, such as the HOLMES2 major-crime investigation management system used by all constabularies.

He pushed through an initiative to use analytics in Manchester in the middle of the last decade. The initiative strove to pin-point likely crime hotspots in Oldham, by integrating historical crime data, established theories on criminal behaviour and geographical analysis. It greatly improved the areas already stretched police presence by distinguishing public perceptions of their safety from statistical likelihoods of crime occurring in a particular area.

The system drew on connections between suspects and localities to highlight the reasons why they might be worthy of police attention.

Bentley noted that many constabularies already have the software and systems to analyse historical data, but not the ability to pull reliable predictions from them. Predictive analysts on the police payroll are few and far between and the prospect of recruiting more of these highly trained people is viewed with distaste by senior police officers, who would prefer to see the money spent on more bobbies wearing out shoe-leather on the street.

There is also the background issue of the culture of policing in the UK, which is firmly wedded to the notion of local policing. A greater reliance on analytical tools and data distributed in real time to officers in the field would require a command-and-control approach that would fundamentally alter current policing processes at the root.

The few instances where crime analytics have been deployed demonstrate that real improvements in law enforcement are achievable by using IT to fight crime. The technology is proven. However, the political will to sign off the necessary investment and drive through the necessary collaboration between agencies is not there.

Perhaps with the advent of police commissioners, this will change.