Reinventing American Policing


This is a tumultuous time for policing in the United States. Deadly use of force by the police in Ferguson, MO, New York City, North Charleston, South Carolina, and Baltimore, MD have led to protests, heated debates, riots, and questioning of police tactics.

Citizens and politicians have called for many reforms to make the police more accountable and transparent. Indeed, President Barack Obama convened a special task force to make recommendations for the reform of policing in America.

This is not the first time citizens have strongly questioned police tactics in the U.S. and other democratic societies. The recurring themes in public discourse on policing concern the core question: How can police effectively prevent crime and keep citizens safe, while at the same time, maintain their trust, confidence and satisfaction?

Both objectives form the bedrock of effective policing in a democratic society. In difficult times, however, discourse often focuses on one objective with the other receding into the background. Our country's present focus is on citizens' confidence in and trust of the police. At other times, especially when crime is on the rise or the threat of terrorism looms, the emphasis is on crime control.

The Two Principles of Policing

In our above-titled essay, we argue that both crime prevention and community trust are important and must be consciously kept on the front burner in deliberations about police reform even when public sentiment is focused on only one of these objectives. Neither should be allowed to trump the other. When that does happen bad policy choices can follow. Accordingly our blueprint is guided by two principles that have coequal standing:

Principle 1: Crime prevention—not arrests—is paramount. Crimes averted, not arrests made, should be the primary metric for judging police success in meeting their objective to prevent crime and disorder.

Principle 2: Citizen reaction matters. Citizen response to the police and their tactics for preventing crime and improving public order matter independent of police effectiveness in these functions.

Principle 1 follows from decades of research that reaches a provocative bottom line: we can't arrest our way out of crime. This does not mean that police should stop making arrests. Arrest serves the important functions of bringing perpetrators of crime to justice, redeeming victims, and reinforcing the capacity of the police to deter crime via the threat of apprehension.

Nonetheless, an arrest also signals a failure of prevention. We now have compelling research that would-be offenders are not deterred by harsh punishment or by rapid response and clearing cases. Further, more than 80 percent of arrests made in the U.S. are for misdemeanors and ordinance violations, often with no identifiable victim. Arrests for minor infractions outside of the Part I crime index do not come without costs to the police and thus to victims of crimes not prevented, to society in its payment for over-use of prisons and jails, and to the misdemeanants themselves.

What we do know is that deterrence and prevention of crime is best achieved by increasing the perceived risk of apprehension, which is created in an officer's “sentinel” role, not in his or her role as “apprehension agent”.

Much research supports the crime prevention effectiveness of proactive, sentinel-like deployment strategies rather than the traditional react, investigate, and arrest approach that characterizes the apprehension agent role. (Proactive policing activities focus police efforts on those people, places, times, and situations that are at high risk of offending. It also takes a problem-oriented approach, which analyzes context to devise tailored and specific prevention tactics. )

Two types of proactive policing that challenge Principle 1 and also Principle 2 are “broken windows” or “zero tolerance” policing, and the use of stop-question-and-frisk. Broken windows policing involves making large numbers of arrests usually for minor crimes such as disorderly conduct or drug possession with the ultimate goal of preventing serious crime. Stop-question-and-frisk (SQF), also known as field stops, focuses on getting police out of their vehicles and having them question suspicious persons. Field stops are believed to create deterrence, not only by increasing interactions between police and potential offenders but also by facilitating broken-windows arrests.

Our review of research on these tactics brings us to three conclusions. First, there is no good evidence for the theory underlying broken windows policing that arrests for minor legal infractions are effective in preventing more serious crime.

Second, results on the effectiveness of SQF are mixed and often based on deeply flawed analysis. The most convincing evidence of effectiveness is limited to circumstances where SQF is used for its legally intended purpose—to prevent gun carrying in violent crime places. Third, both are controversial and have been the subject of much criticism, particularly in Black and Hispanic communities, who distrust the police.

The controversy about aggressive policing tactics such as zero-tolerance policing and SQF challenges thinking about how police can achieve crime prevention goals and at the same time maintain and improve trust and confidence, especially within minority communities that may be most impacted by police deployment decisions. In some circumstances, achievement of these principles may be complementary.

As already emphasized, the research evidence shows that other non-arrest based approaches can effectively prevent crime. Thus replacement of zero-tolerance policing tactics with tactics that place less emphasis on arrest but are comparably effective in preventing crime will likely be less noxious to citizens, particularly in minority communities.

Similarly, tactics aimed at improving police legitimacy that are grounded in procedural justice concepts or community oriented policing may advance both principles (although the research is not as strong with regard to their fulfilling Principle 1). Notwithstanding, tactics such SQF and zero tolerance policing may not be reconcilable as promoting both principles. How then should Principles 1 and 2 be balanced?

The first step in achieving balance is acknowledgement by all parties that both principles are important in their own right. In the heated debate about the New York Police Department's use of SQF and broken windows policing, this important and seemingly obvious point was lost.

One side argued that these tactics are effective in preventing crime and applied in a non-discriminatory way and the other side argued they were ineffective and discriminatory. However, two points were strikingly absent in this debate. The first is that citizen reaction per se was not emphasized as being important in its own right independent of whether these tactics were being conducted in fashion that was legal. Such an acknowledgement would have opened the door to a discussion of the implications of the possibility that SQF or other aggressive policing tactics might be effective in preventing crime, but were also being used in a fashion that was deeply alienating of minority communities.

The second point absent in these debates was a lack of acknowledgement that there are viable alternatives to these tactics to prevent crime. The strength of the belief in the crime prevention returns of tactics such as SQF and broken windows policing reflect the lack of development and use of a broader policing toolkit that will advance Principle 1, but at less or no cost to Principle 2. The origin of this limited toolkit comes from how police have chosen to define their craft and how they have built their institutional and organizations norms around a definition that emphasizes arrest as the key metric of success.

A Blueprint for Reinvention

Only with the recognition that citizen reaction is independently important and also that there are viable and multiple alternatives to crime control that do not emphasize arrest can police, in consultation with the community, begin to think in terms of devising policies that are effective in preventing crime and are also less alienating of minority communities.

We summarize our seven suggestions below predicated on turning that recognition into action:

1) Prioritize crime prevention over arrest

Arrests are costly to all involved—society, the police, and the person arrested. Even for arrests for serious crimes it is important that police broaden the organizational response to asking themselves the question: Is there anything that we could have done to prevent this crime from happening in the first place? Accordingly, we recommend that police focus their efforts, reforms, and resources on what we call sentinel-like activities that prevent crime and in so doing avert the need for arrest and all its ensuing costs. A great deal of research demonstrates the crime prevention effectiveness of policing that focuses on high-risk places and people and uses problem-solving approaches tailored to the specific circumstances of such places or people.

Translating this research into tangible deployment strategies that target high-crime places, times, or people without alienating communities will require significant changes in how patrol officers, detectives, and specialized units are daily deployed. It will require de-emphasizing the importance placed on traditional response-investigate-and-arrest approaches as well as raising the status and importance of patrol officers who are in the best position to carry out sentinel-like activities. It will require supervisors carefully monitoring what types of strategies and tactics officers and specialized units are using at hot spots and among repeat offenders.

It will also require providing training and guidance (see item 3) to expand officers' toolkits to include opportunity mitigation techniques, problem-solving, guardianship activities, and third party policing approaches.

2) Create and install systems that monitor citizen reactions to the police and routinely report results back to the public and also managing and line officers

This blueprint item involves two important components, both in support of Principle 2. The first component is that police should routinely and rigorously survey citizens on their reactions to the police in general and also to specific tactics they use or might use in the future. The second component is that the results of such surveys, as well as the actions taken because of survey results, be regularly reported back to both citizens and officers. We note that this recommendation echoes Recommendation 1.7 in the report of the President's Taskforce on 21st Century Policing (2015).

While the first component is not entirely novel, the practice of systematically understanding citizen reaction to the police continues to be an underdeveloped area of police operations. While some departments survey or interview citizens on an ad hoc basis, typically the surveys are not representative of the targeted population. As a result the data may give a very misleading impression of the public's perceptions of the effectiveness and credibility of the police and just as important do not provide the basis for tracking how citizen perceptions are changing over time.

Either full population or targeted surveys (i.e., specific groups or neighborhoods or people who receive services from the police) should regularly be administered. Only in this way will the police have the data for evaluating their success in achieving Principle 2.

The second component of this recommendation—that the police regularly report back to both citizens and officers the results of the polls—is novel. Further, the purpose of feedback should not just be informational. The feedback should also include changes in police strategies and tactics made in light of the polling information that are developed in conjunction with officers and citizens. Providing officers and managers with results also creates the feedback loops that Sherman (1998) argues are crucial to implementation of evidence-based policing. Systematizing and investing in this feedback loop can improve openness, transparency, and legitimacy, which are all hallmarks of democratic policing.

3) Reform training and redefine the “craft” of policing

Perhaps the most fundamental change to achieve 1 and 2 above is a change in training, not only at the academy but also throughout officers' careers. While it would be naïve to believe that training is the panacea for reform efforts, formal and informal training is an important part of the socialization of officers and shapes and challenges their beliefs about the goals and functions of policing.

The content of police training depends on what agencies, trainers, supervisors, and fellow officers define as the “craft” of policing, which in turn is shaped by beliefs and expectations about the functions, purposes, and methods of good policing.

Thus, to achieve Principles 1 and 2, training must be reoriented to inculcate officers with the belief that these principles form the bedrock of the craft of policing. Training must also provide them with the tools required to make the reduction of calls for service as coequal to the reaction to calls for service. Such training requires instructing officers on how to structure their discretion during their non-committed time, especially in ways that we know are effective in achieving prevention and improving citizen trust.

4) Recalibrate organizational incentives

Rewards, promotions, incentives, and informal “pats on the back” shape the expectations and tendencies of both leaders and the rank and file and can institutionalize Principles 1 and 2 within organizational culture.

Toward this goal, the metrics used to judge performance and suitability for promotion and rewards should measure the officer's knowledge of evidence-based strategies known to reduce crime and improve community trust and confidence.

Candidates for promotion should also be evaluated on how well they translate this knowledge into practice by how they use their non-committed time to prevent crime and/or improve community relations, and their documentation of the use of problem-solving techniques, crime analysis and surveys of community reaction in their daily activities.

In addition, the basis for awarding medals, citations, and commendations needs to be recalibrated. Such recognition is rarely given for preventing crime or improving citizen-officer interactions but instead usually praise success and bravery in apprehensions. This imbalance needs to be corrected.

5) Strengthen Accountability with More Transparency

Police accountability encompasses a vast and complex array of legal, procedural, and organizational issues that go far beyond the scope of this essay. We therefore limit our focus to one viral dimension of accountability—transparency. Improving systems of transparency toward both prevention and improved trust and confidence requires increasing the availability of data and policies related to police-citizen interactions particularly those involving the use of force; doing a better job of communicating back to the public the outcomes of investigations into allegations of police misconduct as well as crime prevention efforts; reassessing systems of discipline and review that may impede the ability of agency's to learn from mistakes; and using analysis for better supervision and management.

6) Incorporate the analysis of crime and citizen reaction into managerial practice

Advancement of the above blueprint items will require that officers, supervisors and leaders shift from reactive and procedures-based decision making to more critical thinking and analytic problem solving. This requires that they have access to high quality analysis of both crime and citizen reactions and are able to manage that information to obtain specific outcomes.

With regard to Principle 1, the types of targeted interventions that we advocate require that officers at all levels have access to and understand analysis that locates concentrations of crime, identifies high risk people, and better illuminates underlying issues that contribute to crime problems. Such analysis is required to carry out Sherman's (2013) “triple T” (targeting, testing, and tracking the success of tailored tactics). Analysis can also serve to create greater transparency and accountability with regard to choice of strategies and tactics and their outcomes.

With regard to Principle 2, the charge of analytic units must be broadened to include measuring and tracking citizen reactions and interventions designed to improve those reactions. The importance of accurately measuring citizen reaction through rigorous surveying, interviewing, and other quantitative and qualitative approaches requires staff with expertise in research methods and in constructing and conducting surveys and analyzing survey data.

Incorporating analysis into managerial practices does not stop at beefing up resources for analysis in agencies. It requires adjusting supervisory and managerial decision making from reacting to crime with standard operating procedures to using analysis as a basis for proactive initiation of strategies for reducing crime and improving trust and confidence of citizens. Without the backing of strong analysis of crime and community reaction, police will be flying blind in managing their efforts to advance all of the above blueprint items.

7) Strengthening national level research and evaluation

The approach we are espousing requires a strong and continually developing knowledge base on how police can prevent crime as well as maintain trust and confidence with citizens. Although much research has been completed on these topics, important gaps remain. While we generally understand that targeting hot places and people can yield benefits, recent research indicates that certain approaches may be more effective than others. Further, while some approaches may be more effective in preventing crime than others, we have little knowledge of the community reactions they engender and how negative reactions can be mitigated by modification of tactics.

Finally it will be important to invest in mechanisms that translate knowledge into operational forms, including helping agencies build the capacity to develop their own knowledge, as well as building exchanges between researchers and practitioners to facilitate receptivity, translation, and institutionalization of research and scientific processes .

Important activities are already underway for creating the research infrastructure we advocate. National-level programs include the National Police Platform, the Smart Policing Initiative by the Bureau of Justice Assistance, the Matrix Demonstration Projects, and the recent President's Task Force on 21st Century Policing. International efforts include the College of Policing in the United Kingdom and the Scottish Institute for Policing Research.

Still much more is needed.

We recognize that appeals such as this, particularly from representatives of the research community such as ourselves, are routine and just as routinely ignored. Notwithstanding, just as advancement in medical practice depends on a robust system of medical research and dissemination with institutions such as the National Institutes of Health in the U.S. and the Medical Research Council in the U.K, a comparably robust infrastructure of research and dissemination on what works in policing is required to advance our blueprint.

A commitment to—and investment in—continuing to build this knowledge and the capacity to develop future knowledge is required to advance our blueprint.

Editors Note: The authors will discuss their proposal at the Vera Institute of Justice in New York today. An extended version of their paper will appear in Crime and Justice: A Review of Research, to be published next year by the University of Chicago Press.

Cynthia Lum is associate professor of the Department of Criminology, Law and Society at George Mason University. Daniel Nagin is the Teresa and H. John Heinz III University Professor of Public Policy and Statistics in the Heinz College at Carnegie Mellon University. The complete essay upon which this feature is based, along with full citations and references, is available from the authors. Your comments are welcome.

Comments are closed.