Balancing public safety and individual rights in street policing

Whether viewed from the desk of a police chief, a city mayor, or a citizen in a deprived, high-crime community, maintaining the balance between police effectiveness and fair policing is complicated and difficult to achieve, let alone to sustain over the long term. However, it is at the heart of good policing, for when policing goes out of that balance—as happened in Brixton, London in the 1970s (1) and more recently in Ferguson, Missouri (2)—the outcome can be a major breakdown in law and order, with wider, rippling consequences for our societies.

Tackling this issue in his lecture on “Fair and Effective Policing” to the 2007 Stockholm Symposium, Ben Bowling argued for “good enough” policing (3), in which police aim to be sufficiently effective at preventing and detecting crime, while remaining sufficiently fair to maintain community legitimacy. As an illustration of the complexity and difficulty of the issues, Bowling went on to contend that the British police deployment of “stop and search” powers—the British equivalent of “stop, question, and frisk” (SQF)—was an example of “not good enough” policing (3), failing his benchmark for balanced and proportionate enforcement.

Instead of good enough policing, Manski and Nagin, in their article “Assessing benefits, costs, and disparate racial impacts of confrontational proactive policing” in PNAS (4), use an alternative model of “optimal policing” to explore how conflicts between public safety and community trust might be weighed and resolved. Bowling (3) used Herbert Simon’s (5) idea of “satisficing” as the framework to argue for his balanced approach to policing. In Simon’s model, the goal is not “optimal” but the achievement of a satisfactory sufficiency, which is “good enough, rather than the absolute best” (6). In contrast, Manski and Nagin’s (4) use of the term “optimal” “supposes that the objective of proactive policing policy is to optimize a welfare function that recognizes both the social costs and benefits.” Their definition of “proactive policing” is much broader than SQF, including the use of arrest to deter others, the deployment of officers to hotspots, and “broken windows” policing of minor disorder by crackdowns.

The debate about the appropriate balance in policing—good enough or optimal—is much more than a theoretical one for criminologists. After a significant crime drop over the last two decades, there are signs that violent crimes may be beginning to rise again in some United States and United Kingdom cities (7). As Fig. 1 illustrates, violent crimes involving knives appear to have risen in London since the second half of 2016. Hales’ accompanying analysis for the United Kingdom Police Foundation (8) has been given added weight by the United Kingdom Office of National Statistics (7), which reported that knife crimes recorded by the police in England and Wales had risen by 20% to the highest figure for 7 y. However, as Fig. 1 also shows, the London police use of stop and search had fallen to its lowest level since 2008. The fall in stop and search followed a deliberate political decision by the then Home Secretary, now Prime Minister Theresa May, to restrict the police use of proactive powers on the street (9). Her decision has parallels in the changes to New York Police Department policies, which followed a Federal Court judgment and a subsequent shift in approach to enforcement from New York Mayor de Blasio (10).

Fig. 1.

Knife crime and stop and search in the London metropolitan police area (monthly data for 2008–2016, with three monthly moving averages) (8).

The debate in the United States and United Kingdom over the effectiveness of police stop and search and SQF, together with the relationship between falls in the police use of the powers and changes in crime rates, is certain to persist. Most commentators have tended to adopt a simplistic and one-sided interpretation, focusing either on the benefits of police enforcement to crime reduction or the societal costs of disproportionate exercise of powers. As Brodeur (11) said in setting out his theory of policing, “the tensions that are present in policing have a way of imposing themselves and coming back to haunt one-sided approaches.” The strength of Manski and Nagin’s (4) approach is, first, that it is multisided, proposing a balance that is tipped by a range of costs and benefits rather than a binary scale between crime control and due process, and second, that it recognizes the importance of a detailed attention to the operation of key variables.

One such set of key variables relates to the way that the police target their use of proactive policing, the tactics that they adopt, the way that they are trained, and the attention that they pay to the quality of each interaction with the public. Our knowledge about what works best for each of these has developed significantly in the last two decades since Sherman et al. (12) reviewed the state of knowledge on effectiveness across the whole field of crime prevention.

The Evidence on Fairness and Effectiveness in Policing

There have now been more than 25 randomized controlled field trials testing the effectiveness of targeted policing of microplaces or hotspots. Braga et al.’s (13) Campbell Collaboration systematic review has shown a consistent, significant, positive effect from such targeting on crime and disorder levels. Some have been critical of this evidence and the advocacy of hotspot policing on the basis that it entrenches an aggressive, proactive style of policing characterized by SQF (14). However, such arguments underplay the variety of tactics, including problem-solving and third-party regulation that have been tested and shown to be effective. Furthermore, a recent hotspots test in Peterborough in the United Kingdom (15) has shown that a “soft power” approach, using partly empowered uniformed police and community support officers, can be highly effective in reducing crime and disorder. Manski and Nagin’s (4) optimal model allows us to factor in the degree of cost from the style of policing and weigh it against the benefits. The Peterborough experiment would appear to be a good example of police testing the use of the least-intrusive proactive approach to achieve significant crime-control benefits.

In Peterborough, the crime and antisocial behavior being targeted were relatively low harm, volume (or misdemeanour) offenses of criminal damage and property crime. The knife crime, which features in Fig. 1, is clearly more harmful and less appropriate for a “soft power” approach. Manski and Nagin (4) begin to discuss the implications of this for their model. Standard measures of costs and benefits tend to use volume or prevalence and frequency of offending as the key measure. Weinborn et al. (16) have used the Cambridge Crime Harm Index (17) to demonstrate that crimes weighted by harm rather than volume cluster more tightly in “harm spots.” A harm-focused approach thus offers the police the enforcement equivalent of the advantages of keyhole surgery: intensive targeting of microplaces with the highest harm with a variety of tailored strategies.

Manski and Nagin’s (4) model also has the potential to factor in the more intangible benefits of police conducting their enforcement operations in a style that reduces the costs of proactive policing. For example, in Queensland, Australia, the police tested the use of a procedural justice script in combination with their random breath-testing operations (18). The procedural justice condition yielded higher levels of trust in the police officer conducting the roadside stop, emphasizing the importance of the way that the police conduct proactive operations. However, as Nagin and Telep have argued (19), our knowledge of the most effective approaches to turn positive opinions resulting from contact with the police into positive actions, such as obeying the law, is still weak. There is a long history of well-intentioned police training programs that, at best, have only succeeded in reducing the harm caused by other aspects of the process of socialization into the police and the police organization and culture.

Conclusions: Toward a New Science-Based Professionalism in Policing?

Charlie Beck, the Los Angeles Police Commissioner, and Connie Rice, a civil rights lawyer, writing in the New York Times, have suggested that we should “move to guardian policing, overcome bias and replace the ‘spiral of despair’ in poor neighborhoods with opportunity and justice” (20). Their vision of a new approach to policing relies on community or neighborhood policing as a fundamental building block. This is a key starting point for a new police professionalism.

Manski and Nagin (4) emphasize another important dimension of policing for the future: a scientific approach to thinking through and judging the right balance between proactive policing and community trust. This is a necessary discipline for police leaders who are otherwise driven by events to see-saw between crackdown and community policing, thereby producing neither good enough nor optimal, but the worst of all possible worlds.