How big data is changing the nature of policing from reactive to proactive

Dwindling budgets, the changing nature of crime and the contemporary demands of communities are forcing police across the country to reevaluate their approach to law enforcement. Adding data-driven insights to their traditional policing expertise could provide a solution. It has already found a home in numerous areas of policing, from fraud detection to the identification of crime hotspots, and is transforming methods of policing from reactive to preventative.

“We strive to be more preventative,” Detective Chief Superintendent of West Midlands Police Andy Hill told the audience at the Reform Big Data in Government Conference on Tuesday. “That isn’t a new concept for policing. If we look back to Sir Robert Peel’s 1829 Principles of Policing, it’s number one on that list and then it recurs again down at number nine [the final entry on the list].”

Data analytics is how he hopes his force, the second largest in the country, can become more effective at realising that principle. Dr Peter Langmead-Jones, the head of better inspection at Her Majesty’s Inspectorate of Constabulary (HMIC), has been pursuing the potential applications in criminal justice.

Putting big data into practice

Langmead-Jones’s team have been using big data to predict the locations of future crimes in order to more effectively deploy resources to prevent them. They evaluated the efficiency of police resources by analysing seven years of police records across 14 different forces alongside more than 1,200 variables from ONS Output Areas, small geographical areas designed specifically for statistical purposes.

There are 43 police forces in the UK, all with varying needs that mean they need to be judged within that specific context. The findings revealed that 40 percent of the time police community support officers were not in the locations where they were most needed. The difference in crime could be explained with more than 80 percent accuracy, and then analysed in relation to the location of officers to determine whether police could be deployed more efficiently.

“The city centre of Manchester has got precious little in common with the Pennine wards of Rochdale, and very little to do with the footballer’s wives territory in the south of Greater Manchester,” says Dr Langmead-Jones, who’s also head of external relations and performance at Greater Manchester Police.

“They’re very different, highly localised contexts, which have an impact on the services that area requested and the judgements of those services.”

The contexts of these localities isn’t solely dependent on their residential populations. The high density of people associating with a high demand for policing varies for a number of reasons, whether that’s social, professional, for transportation or as a function as deprivation.

Using details like the location of cashpoints painted a fuller picture of where population density was concentrated in a way residential records could not. Their work has also been used to identify where and when women are at high risk of domestic abuse, through statistical analysis of the impact of football results and temperature changes, and to establish better practices on how to support different absences from work.

Barriers to implementation

Changing the organisational culture will be fraught with difficulties if the police are not fully informed about the process of accumulating and analysing the data that is acted on. They need to understand the entire pathway in order to justify any actions that result from them, both to the public and to themselves, as James Slessor, the managing director of global public safety at Accenture explained.

“The human user can be neglected and that person must receive the right information in the right format at the right time for it to have any value.”

Data silos and interoperability issues are other recurring limitations to public sector innovation, and the police forces are no exception.

“We have 43 different police forces in England and Wales,” says Hill. “Each police force, it will not surprise you, has lots of legacy silo systems that we’ve built up over a number of years.

“It will probably not surprise you that most of those don’t talk to each other, and if they do, the exchange of data can be challenging from a security perspective on our existing national infrastructure.

“So we do need to modernise. We need to modernise infrastructures and maximise the value and insight we get from that data.”

Predictive policing has already been implemented around the world, but it is not without its critics. Data analysis has been accused of placing a disproportionate focus on only those crimes that are most commonly reported of reinforcing stigma among certain groups that quickly leads to racial profiling. Public support is critical, so their trust is essential.

“We need to be able to demonstrate a clear public gain and a clear public benefit to work with and do things [with data] across all of the security, the ethics, the legitimacy and the insights, because ultimately we want to deliver fundamental policing bringing that data back to 1829 in a fundamentally data-driven way, that puts data as an asset in the heart of our decision-making,” says Hill.

There’s work to be done, but the evidence suggests that data analytics has the potential to finally realise the overriding ambition for policing that Sir Peel made official almost 200 years ago.