IT SOUNDS elementary, but for the first time, UK police officers will be running their own scientific trials to discover which practices actually reduce crime.
In the past 50 years, just 110 randomised controlled trials of police practice have been conducted around the world. Now, the UK police and forces from several other countries are starting to overthrow convention and move forward with more evidence-based approaches.
Advertisement
For thousands of years, law enforcers have trusted their intuition and instincts. But one in every 26 police efforts to reduce crime actually have the effect of increasing crime, says , director of the Institute of Criminology at the University of Cambridge. 鈥淭o do things without knowing the consequences is to act unethically,鈥� he says.
Now, new police officers across the UK will be taught how to understand and implement scientific evidence in policing, and how to run their own experiments and trials 鈥� made possible in part by a 拢10 million injection of cash from the UK government and the Higher Education Funding Council for England. The US has also begun to push for a more scientific approach to policing with the recent launch of the (ASEBP).
Early results from several trials show that some common police practices are ineffective, or even harmful, while others highlight ways to ensure that people are all treated fairly.
Police practice is littered with examples of good intentions gone wrong. Take, for instance, the 鈥渟cared straight鈥� programmes established in the US in the 1970s. The idea sounds simple: identify young people headed for a future in crime, take them to prisons and scare them into good behaviour.
鈥淩esearch shows this actually increases the likelihood of young people offending in the future,鈥� says , director of research and education at the UK College of Policing. While the UK has phased these programmes out, they are still popular in some US states.
鈥淭here鈥檚 a lot of police practice that runs against the evidence,鈥� says Sherman. For example, he says, 鈥減rosecution of juveniles leads to an increase in repeat offending鈥�. Research suggests that restorative justice 鈥� in which an offender has to explain their actions to the victim 鈥� seems to be more effective. 鈥淎rrest kids, perhaps, but don鈥檛 take them through the criminal justice system,鈥� Sherman says.
鈥淲e鈥檝e prosecuted people for thousands of years without testing our approaches鈥�
Even arresting someone can have unrealised effects. Take, for example, police responses to call-outs for domestic violence. In one trial, police either warned people about their behaviour or arrested them. Sherman and his colleague Heather Harris studied the resulting impact. They found that the partners of suspects that had been arrested rather than warned were 23 years later.
鈥淪tress-related diseases appeared to be accelerated,鈥� says , a former detective superintendent now based at the University of Cambridge鈥檚 Institute of Criminology. 鈥淚t鈥檚 controversial. Police are being asked to do more about domestic violence, but 鈥榙oing something鈥� is not as straightforward as it sounds.鈥�
Trailblazers in evidence-based policing also face another problem 鈥� how they and their colleagues learned their jobs. 鈥淥fficers are experience-based rather than evidence-based, although that is changing,鈥� says Neyroud. At least, it is in countries like the UK. The US is further behind, says Ren茅e Mitchell, a police sergeant at Sacramento Police Department in California, and founder of the ASEBP.
Cultural shift
Part of the problem in the US is that responsibility for policing is a local issue. The lack of federal control means that little progress is made, says Sherman.
鈥淭here are 18,000 police departments in the US, and some of those consist of one man and a dog,鈥� says Neyroud.
鈥淭here are 18,000 police departments in the US, and some of those consist of one man and a dog鈥�
Mitchell hopes the ASEBP will help change that. 鈥淭here鈥檚 a cultural shift that needs to happen,鈥� she says. For instance, many police officers have yet to learn that they can expect better outcomes when they treat suspects with respect. 鈥淚鈥檝e trained cops in communication skills 鈥� they say they鈥檒l use them on citizens but not suspects because it鈥檚 鈥榥ot safe鈥�,鈥� she says. 鈥淲e need to show cops they are safer when they treat people well 鈥� that punishment doesn鈥檛 begin with them.鈥�
Mitchell will start by making existing research more accessible to police officers in the US. 鈥淢y dream is to take every piece of police research and translate it into police-speak,鈥� she says. She also hopes to encourage officers to team up with academics to test whether what they are doing actually works.
Randomised, controlled trials are the best way to test police practice, says Sherman. That means randomly giving suspects one of two different types of treatment, and observing the effects. It鈥檚 a controversial idea, says Neyroud. 鈥淧eople think it鈥檚 ethically wrong to randomise the treatment of offenders, because everyone should be treated equally,鈥� he says. 鈥淏ut we鈥檝e prosecuted people for thousands of years without testing our approaches 鈥� the time has come to challenge them.鈥�
At any rate, given the record of prejudice and within many police forces, randomising treatment may turn out to be the fairest thing to do, says Sherman.
Time for change
Research results are already starting to transform police practice. One highly researched idea that is becoming widespread is 鈥渉otspot policing鈥� 鈥� targeting specific areas that are known to be at the greatest risk of crime. 鈥淔our to 5 per cent of locations account for 50 per cent of crime,鈥� says Neyroud. Mathematicians have been developing equations to help police forces, such as the LAPD, identify these hotspots. 鈥淧olice were just policing hotspots selected on opinion,鈥� says Sherman. 鈥淚t turned out they were completely different to those identified by technology.鈥� He says these equations have been put to use in Trinidad over the past two years, and they鈥檝e seen a 45 per cent drop in homicide as a result.
Research has also changed the way police conduct interviews. In the UK, police no longer do interrogations, says Neyroud, because they are more likely to elicit false confessions. Instead, they perform cognitive interviews, which draw on psychology studies to enhance the recall of a victim or suspect. For example, interviewees are encouraged to recount their story in reverse order, because they are more likely to remember the most recent aspects. They are also encouraged to report every detail, no matter how trivial, since this is known to help trigger other more relevant memories.
Many aspects of policing are long-overdue a scientific assessment. For instance, stop-and-search practices are often in the spotlight, with police accused of targeting ethnic minorities. London鈥檚 Metropolitan Police Service recently attributed a rise in knife crime to the scaling back of stop-and-search programmes. There is no evidence that is the reason, says Neyroud. 鈥淏ut it would be nice to know.鈥�
Last month, the UK College of Policing launched the first controlled trial to find out if training helps officers overcome any unconscious biases they might have, and and search people fairer and more effective.
Another controversial issue in need of analysis is the use of tasers. 鈥淎re tasers useful at reducing the use of force or not?鈥� says Neyroud. 鈥淎 taser can be lethal under some circumstances.鈥�
Other projects are assessing the use of body-worn cameras, how to improve police responses to cases of child exploitation and how to .
It is still early days, but proponents of evidence-based policing are hopeful. 鈥淭he ethics of policing haven鈥檛 yet developed to the point where you have to prove that something doesn鈥檛 cause harm before you do it,鈥� says Neyroud. 鈥淲e鈥檙e not quite there yet. But give us a chance.鈥�
Leader: 鈥�Counter-intelligence: When politicians make law a scientific ass鈥�
(Images:Andrew Testa/Panos, Adam Patterson/Panos Pictures)
This article appeared in print under the headline 鈥淧olice academy: Back in training鈥�
