Tension instead of trust
Camden, New Jersey, isn鈥檛 where you might expect to find the future of policing emerging. It was renowned as one of the most violent cities in the US for decades. Two years ago, struggling with crime and running out of money, the city scrapped its police department entirely.
But now Camden is at the centre of an ambitious new scheme. The White House launched the Police Data Initiative there this week, tasking Camden and 20 other US cities with using data analysis to understand and change police behaviour in an attempt to heal the broken relationship between them and the public.
Advertisement
The move follows a wave of police brutality in the US (see 鈥淎 year of violence鈥�, below) and the establishment of the Task Force on 21st Century Policing, formed by Obama in December. The first report from the task force, published this week, stressed the importance of collecting, sharing and scrutinising such data. It recommended examining department policies, incidents of alleged misconduct, recent stops and arrests and the demographics of officers. Analysing the data, the authors write, is vital 鈥渢o knowing what works and what does not work, which policing practices are effective and which ones have unintended consequences鈥�.
Some police departments are already putting their data to work. This summer, the University of Chicago鈥檚 Data Science for Social Good programme will start building an early warning system for local police departments. The algorithm will scour police data and pick out problematic officers, tagging them for further training or intervention.
Similar systems exist in other parts of the country, but the White House warns that those may not be effective, because there has been little research into which factors most strongly predict a problem. The Chicago pilot will be grounded in such research.
Proactive approach
, who heads up the Data Science for Social Good programme, says systems like this can help prevent problems from happening. 鈥淚t really changes the way a lot of government agencies work. Today, they鈥檙e purely reactive, but now they can be much more proactive,鈥� he says.
Strategies such as predictive policing already use data analytics to try to prevent crime. Now, those same techniques are being applied to the police themselves, trying to pre-empt the possibility of brutality.
On-body cameras may provide a rich source of data. The cameras, which have been shown to reduce the use of force and the number of complaints from citizens, have been adopted by about of police departments across the country. The Department of Justice now plans to million on acquiring more. In Oakland, where officers have been required to wear cameras for the past four years, researchers at Stanford University are developing tools that can automatically detect positive or negative interactions from the audio in a recording.
It鈥檚 clear that the potential is great, but there鈥檚 a long road ahead. William Terrill of Michigan State University in East Lansing sees a parallel with community policing, an approach that attracted attention after the civil unrest of the late 1960s. While the idea was warmly discussed, he says, there is marginal evidence that it was ever implemented effectively. A few decades down the road, we could be saying the same about these data-driven strategies.
鈥淧olice departments are good at making reactionary policies in times of crisis, but not making substantive changes,鈥� Ghani says. 鈥淢y greater fear is that things will calm down and police departments will go back to the same old, same old.鈥�
Understanding the community
To address that problem, cities involved in the Police Data Initiative have promised to release statistics about hot-button topics like vehicle stops and officer-involved shootings. And non-profit Code for America is building software to make it easier for agencies to open up.
The idea is that the community, rather than the police themselves, can analyse that data and hold the officers accountable for their actions. 鈥淧eople want to know and understand what鈥檚 happening in their own communities. They actually want to see the data, to be able to analyse it in their own ways,鈥� says of the University of Cincinnati in Ohio.
This isn鈥檛 the only aspect of policing that needs overhauling. Terrill says the Obama initiative may encourage a long overdue record-keeping upgrade. A typical police department鈥檚 data is disorganised, with separate systems to track things like 911 calls, internal affairs and officer training. This disorganisation means that there is no common metric to compare different police departments, and it can be difficult or impossible to measure performance in areas like reducing use of force.
The ability to compare the performance of different police departments is crucial because it incentivises change and improvement. Terrill says the picture would look very different if existing US police departments operated in a more competitive, transparent environment: 鈥淭hey would have gone out of business years ago.鈥�
A year of violence
The number of deaths of US citizens at the hands of police officers has made for a troubling trend over the past year. Eric Garner died repeating 鈥淚聽can鈥檛 breathe鈥� while an officer kept him in a chokehold. Twelve-year-old Tamir Rice was killed by the police officers while playing with a BB gun in the park. Jason Harrison, a mentally ill man, was shot when he failed to drop a screwdriver. And Michael Brown, 18, was shot and killed by a police officer on the streets of Ferguson, Missouri.
In light of this, Obama formed in聽December, to 鈥渆xamine how to strengthen public trust and foster strong relationships between local law enforcement and the communities that they protect鈥�.
Meanwhile, deaths at the hands of聽police have continued to make the news聽鈥� most notably last month, when 25-year-old Freddie Gray suffered a fatal spinal cord injury while restrained in a police van in Baltimore, Maryland.
This article appeared in print under the headline 鈥淧olicing the police鈥�
