How to reduce police violence?

As revolts break out once again in the United States against racist police violence, researchers are trying to assess the effectiveness of various measures aimed at reducing fatal police interventions. This is notably the ambition of a special issue of the journal ANNALS of the American Academy of Political and Social Science (AAPSS). The file is edited by Lawrence Sherman, the founding father of the evidence-based policingan approach that calls for basing police policies on the results of empirical research.

The main conclusion is that there is little conclusive research on the effectiveness of measures often considered common sense, such as body cameras, de-escalation training, or citizen monitoring bodies (see the article by Robin Engel and colleagues in this special issue). The articles in the special issue focus only on the United States, but they offer interesting lessons for France at a time when police violence has become a central object of political debate and activist demands.

Measuring the impact of post-Ferguson reforms

This special report comes five years after the death of Michael Brown, an African-American killed by the police in 2014 in Ferguson, which sparked the emergence of the movement Black Lives Matter. Spurred on by the movement, President Obama convened a task force of experts who recommended a series of reforms, and several local police departments have restricted the use of weapons (for example, by banning shooting at moving vehicles), expanded the use of body cameras, and increased their de-escalation training (see Laurie Robinson’s article in this issue).

While these measures have not radically reduced the number of fatal police interventions, the authors of the dossier agree that they have brought about significant changes in the management of police action. The dossier seeks to measure the effect of these changes, adopting the empirical approach of evidence-based policingwhich stresses the need to target reforms where they will have the greatest impact, to test new measures, and to monitor their implementation (” targeting, testing, tracking “, see the introduction to the file written by Lawrence Sherman).

What social factors promote the use of force?

In the file of theAAPSStwo articles propose innovative methodologies to study individual characteristics (age, gender, race, seniority) that increase the probability that a police officer will use excessive force, while distinguishing them from other factors such as the context of the intervention or institutional policies.

Greg Ridgeway proposes to analyze cases of fatal police violence where several officers were present, and to compare the characteristics of the shooting officer and those of others who were present but did not use force. His investigation is based on an analysis of 106 interventions by the NYPD where an officer fired his or her gun between 2004 and 2006, and shows that senior officers, as well as those who start their police careers later, shoot less often. The results also show that black officers are more likely to shoot than white officers. With this methodology, Ridgeway offers a solution to one of the recurring problems of investigating violent police officers: it is often impossible to separate the effect of individual characteristics from that of institutional factors (since, for example, black and younger officers are more likely to be placed in units where they are at greater risk). His approach allows for better identification of officers at risk and therefore better targeting of interventions aimed at reducing the use of force.

Linda Zhao and Andrew Papachristos, for their part, conduct a network analysis based on a database containing all complaints filed against Chicago police officers between 2000 and 2016. Their research shows that while officers who use their firearms are a small minority, they occupy a particular position in police networks: they connect teams that are not otherwise connected. In other words, having worked in many different units increases the probability of shooting, even after taking into account individual characteristics. To explain these results, the authors raise the possibility that the policy of transferring officers who “pose a problem” or who are accused of excessive use of force could contribute to spreading violent practices from one unit to another and from one neighborhood to another.

What training is available to reduce police violence?

One of the most frequently proposed solutions to reduce the use of police force is to improve police training. Two articles in the dossier warn that there is little conclusive evidence demonstrating the effectiveness of the most commonly recommended training (implicit bias and de-escalation training), and propose new models.

Scott Wolfe and his colleagues studied an innovative training program that focused on a broader focus on police social interaction skills, focusing on the importance of respect, listening, self-control, and using force only as a last resort. The training was based on the principle of repetitive, deliberate practice, and each officer received one hour of training every two weeks for three to six months. The article suggests that this approach is effective in changing police attitudes, but it does not assess the impact of the training on police behavior in the field.

Harold Pollack and Keith Humphreys propose training police officers in crisis management to improve their ability to deal with individuals in mental health crisis. This type of training emphasizes de-escalation, but also the importance of giving time and distance to the person in crisis. The authors argue that such training would reduce the frequency of the use of lethal force against individuals in crisis, but note that it must also be accompanied by other measures to limit the need for police intervention in the first place, such as better access to social, medical, and housing services for people with behavioral problems.

How to “reconcile” the police and the population?

In their paper, Thomas O’Brien and his colleagues begin by noting that the prevalence of police violence, especially against black people, reinforces the perception in these communities that the police are illegitimate. Any effort at reconciliation and rapprochement between the police and the public, they write, must take into account the historical oppression of African Americans and the central role that the police play in maintaining that oppression.

The authors wanted to measure the impact that a police apology would have following an incident of police violence. To do this, they conducted an experiment that consisted of proposing fictitious scenarios to different groups of respondents, and asking their reaction if this scenario occurred in their neighborhood. Their results demonstrate that only apologies accompanied by an acknowledgement of responsibility on the part of the police have the potential to rebuild trust between the police and oppressed populations. Attempts at partial apologies that do not include an admission of responsibility on the part of the police will tend, on the contrary, to further reduce trust in the police and police-population cooperation.

What about structural changes?

Overall, the dossier does little analysis of the structural conditions that lead to such high rates of fatal police violence in the United States, aside from an article by Franklin Zimring on the American federal model and the resulting blockages to police reform, and one by Daniel Nagin which suggests that there is a link between the rates of gun ownership in the hands of the population and the number of fatal police violence.

Yet there is a wealth of work that analyzes the role that the police play in maintaining the racial and social oppression of certain populations, and that argues that the police cannot be usefully reformed without reducing racial and socioeconomic inequalities and transforming the core of the police mission from social control of marginalized populations to inclusion and the reduction of harms.

Thus, the ambition of this dossier is limited. It is not to radically rethink the way in which police and judicial institutions operate, as demanded by activists Black Lives Matterbut only to measure the effect of certain individual and institutional factors on the number of deaths during police interventions.

What lessons for France?

Although the file of theAAPSS focuses only on the United States, the surveys presented can help to rethink the use of police force in France. This is not about making false equivalences. There are important differences between the two countries: the number of fatal police interventions is much higher in the United States than in France; citizens are allowed to carry weapons there and the police are exposed to much greater risks than in France; and the American police are extremely decentralized. However, there are also similarities that make the results of the American research interesting for a reflection on fatal police interventions in France. In France as in the United States, the police regularly kill people who pose no danger to the police or to others; ethnic minorities are overrepresented among the victims; and social protest against this police violence has become increasingly vocal in recent years.

This file offers two avenues for starting a debate on reducing police violence in France.

Demand greater transparency from the police. The investigations presented here were possible only because the police have made public data on police-citizen interventions, firearms use, and complaints against serving officers. The French police make very little data public, making independent evaluations of reforms difficult (and citizen-captured footage all the more important).

There is no miracle reform. The American example shows that despite a stronger political will in recent years to reduce fatal police interventions, the reforms implemented have had a limited effect and the number of deaths at the hands of law enforcement remains very high. The authors of the dossier conclude that it is necessary to continue to systematically evaluate new measures to better understand their impacts.

Perhaps we should also broaden the scope of our thinking, and go beyond attempts to re-form the existing system, for rethink the police, starting with radically reducing the carrying of weapons by police officers, reviewing laws on the use of force, and re-establishing judicial control of police action with an independent investigative body. We could also aim to reduce, upstream, the need for police interventions, by investing in alternative systems for conflict management such as restorative justice or autonomous mediations, or by reallocating part of the police budget to invest in the most disadvantaged neighborhoods – in education, employment, culture, or mental health services – to reduce, upstream, the need for police interventions.