Back to Blog

Incorporating Evidence-Based Approaches in Policing Practice

Heather Prince

    We typically fall short of testing many policing practices1.​ At times, we may try new programs and initiatives because they are considered best practices, or because another agency implemented them. However, this type of approach can be costly or harmful if carried out incorrectly. Without testing a new strategy, policy, or technology, confounding variables may keep us from knowing whether it works or is harmful2.​ Agencies’ resistance to testing is due, in part, to the slow nature of research. From start to finish, a research project can take a year or more to complete. Research findings might become dated before they are released — and when completed, research may still be hidden behind paywalls, making it difficult to access. 
In addition, police officers and others who deal directly with law enforcement issues are not typically involved in the research process, so the results produced can be perceived as irrelevant to their agencies. Exacerbating these challenges is the fact that many police departments in the United States still focus on the standard model of policing, typically characterized by the three R’s: random patrol, rapid response, and reactive investigations. This model is utilized because agencies have access to limited resources3.​ Furthermore, agencies may not yet be aware of the effective approach of targeting, testing, and tracking the data4.​ Additional challenges brought on by police unions, such as the fear of changes in working conditions ​and the fear ​that the data will be misinterpreted, often add to the roadblocks associated with implementing police experimentation. 

    How can we, as a field, encourage proper testing of the policies, protocols, and initiatives that we implement in our departments and derive results that are relevant to the goals we are trying to achieve in a timely manner? Here, we address the challenges that many departments face in implementing an evidence-based approach to policing and discuss solutions to these challenges. We also provide examples of studies that have overcome these challenges and produced data that helped agencies update their policies and practices to better meet the needs of officers, departments, and communities.

    The benefits of data-driven approaches to policing should seem obvious. Yet we still resistthose approaches due to a focus on the standard model of policing5.​ The goal for 
forward-thinking policing professionals is to have evidence-based policing be more accessible to the 18,000 police departments and 800,000 police officers in the United States. Instead of a concept championed by a few practitioners or academics, evidence-based policing might become part of the fabric of daily policing. Some believe that this could be achieved with reliable partnerships between crime analysts and researchers. Herman Goldstein, the father of problem-oriented policing, suggested in a conversation with Lawrence Sherman that for every 6​ 
100 police officers, there should be at least two crime analysts6.​ However, this would require an increased budget and a mandate to focus on data-driven/smart policing. Many smaller agencies could not afford this expense, which limits their ability to implement evidence-based approaches. Another challenge derives from agencies’ interest in and ability to implement evidence-based practices. Effective police efforts to reduce crime have included various strategies: hot spots policing, crime prevention through environmental design, situational crime prevention, “Koper” directed patrols7​ and various problem-oriented policing practices8.​ Despite these positive examples, many police departments throughout the United States still focus on random patrol. Any data collection or analysis that occurs is typically out of necessity and presumably rooted in hiring, staffing, and retention burdens. 
    Evidence-based policing is intended to enhance strategy, operational deployments, initiatives, and policy while still maintaining an awareness of policing’s inherent uncertainty. Unfortunately, the policing profession continues to resist research and data. This may be due to a variety of reasons, including the constant requirement to respond to critical situations, which leaves little time to evaluate those responses.​ ​There are also important stakeholders who, frustrated by the lack of much-needed resources for law enforcement, see calls for increased police efficiency as attempts to divert attention from the deeper need for more funding and personnel. Nonetheless, there are additional reasons for resistance that seem to confound advocates of evidence-based policing — perhaps a lack of trust in police, the many cultures that police leaders must navigate, or the decentralization of U.S. policing compared to law enforcement in other industrialized countries. There are 18,000 individual police departments in the United States, and the prevailing practices in many of them are based on tradition, 9​ culture, politics, law, agency-specific values, and public opinion9.​ Together, these challenges limit the use of evidence-based approaches in policing. 
    As with any challenge, solutions should be individualized based on an agency’s needs. Agencies where resistance to evidence-based policing is due to limited resources may require different solutions than agencies where resistance focuses on the years-long lag time between initiating research and generating actionable results, or even those where the main obstacle is a hesitancy to change the policing culture by implementing new practices. It is important to consider the needs and barriers of each individual agency and use an approach tailored to the unique culture of the agency and the community it serves. 
    First, it is important to recognize that evidence-based practices have been deeply embedded in policing for centuries — we just did not refer to them as such. Crime reduction is an excellent example of how strategies can be individualized to promote evidence-based policing in diverse communities. In 1829, crime in London was deterred by “preventative policing.” Criminals were severely punished as an example to others and hangings were prevalent. As the head of the London Metropolitan Police Force, Robert Peel introduced proactive policing strategies. Peel wanted police to manage criminals using swiftness and certainty10.​ He reduced the use of severe punishment and leveraged cooperative relationships with the community to address crime. The evidence in favor of this less severe approach, some may argue, was that crime did not increase after hangings were abolished11.​ Of course, ​deterrence of this type did not function as a crime-reduction strategy for all types of violent crime even back then, as crime is complex and highly nuanced. 
    A different approach to crime reduction might be the perception of increased surveillance. For example, the French philosopher Michel Foucault believed that people conform to norms and 
expectations when they believe they are being watched — ​a phenomenon he termed panopticism12.​ When looked at from a cost-benefit perspective, perhaps a constant police watch through visible patrol is most effective at reducing crime; this principle is where contemporary policing is deeply rooted13.​ Nagin, for example, suggests that police officers act 
as sentinels and avert crime with little need for apprehension14.​ We see this strategy put into practice through hot spots policing, in which police officers focus on the places, people, property, and times most closely associated with the commission of crimes in a data-driven approach to crime reduction15.​
    These examples demonstrate how deeply evidence-based policing is rooted in the field. Since Peel tested new approaches to reducing crime 200 years ago, law enforcement has consistently used evaluation and evidence to support policing strategies. Foucault’s theories have supported surveillance-based policing for decades. Those in law enforcement who balk at the “new” approach of evidence-based policing might be assuaged by the fact that they have always used evidence to support their work — now, it just has a name. 
    Next, the timeliness of data requires discussion. A major challenge of conducting research in the field is that policing is a swift profession — current challenges may lose relevance by the time a formal study is completed. This issue can be mitigated by our approach to the data. Traditional research prioritizes extensive analysis supported by the collection of large amounts of data over extended periods of time. If law enforcement agencies focus instead on collecting just enough data to answer their own narrow research questions (e.g., where and when does crime happen in my local jurisdiction, and who is committing it?), they can analyze results much more quickly and respond to existing issues with greater agility. 
    For example, trends suggest that as much as half of all crime in any jurisdiction occurs at just 16​ 3% to 5% of locations16.​ What if we could reduce crime at these hot spots through proactive deterrence instead of reactive enforcement? Hot spots policing is effective at reducing crime because it analyzes data specific to the time and location of crime, excludes any analyses that are not directly relevant to those parameters, and produces analyses quickly enough for law enforcement to address crime in hot spots as it arises. This approach, though narrow in its scope, demonstrates that with adequate focus and resources, collecting and analyzing evidence can benefit police agencies in a timely manner. It is up to agencies themselves to determine which problems to focus on and what resources are necessary to craft effective solutions. 
    If we look inward at our roles in a democratic society and restructure our measurements and reward systems to focus on deterrence and prevention where appropriate, then we may align 
our actions to desired outcomes17.​ The result might be the institutionalization of evidence-based approaches to policing based partly on analyzing and assessing data with more crime analysts. If police leaders through the sergeant ranks can build a receptivity to scientific research in policing, then we might better understand the impact of our interventions and be able to say definitively that they are working as intended, rather than causing more harm. By reviewing data, conducting rigorous evaluations, and using the best available evidence and research, we can inform our long-term decisions, policies, and practices and place officers at the times and locations where crime is likely to occur, making us more effective in improving 
public safety18.​ These solutions may follow the adage “what gets measured gets done” — in other words, setting measurable goals to spur improvement — or solutions may come from incentivizing commonsense outcomes and creating reward structures that focus on deterrence when appropriate. 
    The most effective way to incorporate evidence-based approaches into policing is to deploy experts on the topic within agencies. Anyone — officers, sergeants, or chiefs — can be these experts. They can encourage their agencies to take a deeper look at their policies through an evidence-based lens. Highlighting the role of individuals does not mean we cannot discount the limitation that a lack of other resources brings. Reduced time, finances, and equipment can all hinder the viability of an evidence-based approach. However, personnel who champion this model are still the most essential component to ensuring the success of evidence-based policing within agencies. These are people who can train others in the department, network with external colleagues to bring in new knowledge, and — most importantly — lead the agency’s efforts to test and evaluate the benefits and drawbacks of both new and existing policies. 
    When incorporating evidence-based policing into current practices, especially in agencies that resist change, the overall strategy should be to identify situations in which evidence-based approaches are already being utilized and build on those successes. The greatest advantage of evidence-based policing is that it already exists within most policing institutions. Increasing the visibility and use of innovative crime reduction models depends on champions within those institutions who recognize their use and advocate for educating colleagues and expanding the influence of evidence across policies and protocols. 
Recent Examples of Success Using Evidence-Based Policing 
    Despite the challenges of implementing evidence-based approaches in policing, there has been a recent surge in interest in and practice of these approaches in departments across the nation. Law enforcement agencies are embracing evidence-based policing and advocating for the testing and evaluation of policing practices. Programs like the National Institute of Justice’s Law Enforcement Advancing Data and Science Initiative provide an opportunity for law enforcement personnel interested in evidence-based policing to network and share ideas. Groups such as George Mason University’s Center for Evidence-Based Crime Policy and the American Society of Evidence-Based Policing provide similar networking opportunities and disseminate information that can help agencies build on their own evidence-based policies and practices. 
    The topic of crime reduction and deterrence serves as an excellent example of the effect of evidence-based practices in policing. One method of crime deterrence uses increased police 
presence to try to reduce crime19.​ This strategy evolved from data and anecdotal experiences of police showing that the behavior of community members changes when law enforcement officers are present (which reflects Foucault’s theory of panopticism). When police are watching, a driver who is speeding may slow down, a frequent jaywalker may choose to cross at a designated crosswalk, a car thief may be deterred, and a drug user may refrain from pulling out drug paraphernalia. In this way, police presence can prevent crime on a day-to-day basis by making the environmental conditions unsuitable for crime20.​
    Although promising, results from increasing police presence to deter crime have been mixed. For example, random policing patrols have not proven successful21,​ whereas increasing patrols​ in high-crime areas (hot spots policing) has been found to reduce crime22.​ In the Philadelphia Foot Patrol Experiment, hot spots were patrolled for 16 hours a day, five days a week, by new police academy recruits23.​ As a result, the hot spots documented 23% fewer crimes than did control areas. A recent trial addressing deterrence and police presence, the Philadelphia Predictive Policing Experiment, included both marked and unmarked police vehicles and demonstrated that marked vehicles had a deterrent effect on property crimes in predicted hot spots, while unmarked vehicles did not24.​ Putting a “cop on the dot” in areas where crime takes place is probably not realistic for most resource-strapped police departments whose officers are running from call to call. However, if police leaders can place an officer in that hot spot for 15 minutes at a time, unpredictably and sporadically, then research shows a resulting decrease in crime that lasts for two hours25.​ Suddenly, this form of deterrence becomes much more realistic and manageable.

    In addition to increasing the presence of foot and vehicle patrols, it may be that simply the appearance of a police presence is enough to reduce crime26.​ In a recent trial conducted in a 
suburban area of Connecticut, the use of police lights was tested to see whether having the lights on would reduce crime. Over an approximately four-month period, night shifts were randomly assigned to having police lights on or off. Reductions were seen in burglaries from motor vehicles, stolen vehicles, pedestrian stops, traffic stops, motor vehicle accidents, and criminal arrests, though these reductions did not reach statistical significance27.​
    In Vallejo, California, a municipality located between Oakland and Sacramento, our police department employed a randomized controlled trial to investigate the effectiveness of utilizing code-2 flashing police lights (blue, red, and amber) for reducing crime in a specified shopping area. (See the related article, “How Do We Know It Works? Conducting a Rapid Research Police Experiment to Test the Effectiveness of Flashing Police Lights on Total Auto Crime,” in this volume.) This trial was conducted in the vein of rapid, practitioner-led research. Officers patrolled with code-2 lights on (intervention) or off (control) during their entire shift for a 34-day trial period. Our findings revealed that running code-2 lights during a shift reduced motor vehicle crimes by 50% (​p =​ 0.07). The findings further suggested that increased visibility of law enforcement in high-crime areas has a significant impact on reducing crime without using excessive resources. These results may have broad implications for agencies seeking to increase effectiveness while limiting cost. 
    When taken together, these studies provide insight into how the evidence-based approach works. Existing literature provides evidence on the efficacy of common policing tactics. Police departments apply the evidence from the literature to test specific, local practices that may be unique to their agency. Some tests may successfully demonstrate that a policy or protocol is effective. Others may demonstrate the opposite, or have mixed results. As more police departments test their unique practices, a greater evidence base forms. Conclusions about a broad range of practices and the theories underlying them can be formed from this new, robust evidence base. Departments can share results with one other. Over time, every community and agency benefits from testing and evaluating new and established policies, protocols, and strategies.
    There is no doubt that implementing evidence-based approaches in policing can be a challenging endeavor. However, past and present examples demonstrate that these challenges need not prevent testing and evaluation to improve policing practices. Understanding and addressing challenges such as the timeliness of research, changing cultures, and resource limitations are key to the successful implementation of evidence-based policing across the United States. Ultimately, champions in every department are needed to build evidence-based approaches that will continue to enhance the policing profession. However, it is important to remember that officers’ skills, experience, and instincts remain critically important — especially when coupled with the best available data, science, and research. 

[1] Lawrence W. Sherman, “The Rise of Evidence-Based Policing: Targeting, Testing, and Tracking,” 
Crime and Justice​ 42 no. 1 (2013): 377-451. 
[2] Ibid
[3] Ibid
[4] Ibid
[5] David Weisburd and John E. Eck, “What Can Police Do To Reduce Crime, Disorder, and Fear?” The Annals of the American Academy of Political and Social Science 5​ 93 no. 1 (2004):​ ​42-65, doi:10.1177/0002716203262548. 
[6] ​Cambridge Centre for Evidence-Based Policing, “A Conversation with Herman Goldstein (Full Interview),” retrieved from YouTube. 
[7] Research shows that intermittent and sporadic random patrols of street-segment hot spots for 10 to 16 minutes every two hours extends deterrence. Departments can utilize computer-aided design software, GPS, body-worn camera GPS, and automated vehicle locators to monitor, deploy, and ultimately test the effectiveness of these interventions. 
[8] ​Weisburd and Eck, “What Can Police Do.” 
[9] ​Gary Cordner and Geoffrey Alpert, “Striking a Balance: Research, Science, and Policing,” Research in Brief, ​Police Chief​ 8 (2018): 14-15. 
[10] Lawrence W. Sherman, “Al Capone, the Sword of Damocles, and the Police-Corrections Budget Ratio,” ​Criminology & Public Policy​ 10 no. 1 (2011): 195-206, doi:10.1111/j.1745-9133.2010.00700.x. 
[11] Ibid
[12] Kitty Calavita, ​Invitation to Law and Society: An Introduction to the Study of Real Law,​ 2nd Ed. (Chicago: University of Chicago Press, 2016). 
[13] Jerry H. Ratcliffe, Travis Taniguchi, Elizabeth H. Groff, and Jennifer D. Wood, “The Philadelphia Foot Patrol Experiment: A Randomized Controlled Trial of Police Patrol Effectiveness in Violent Crime Hotspots,” ​Criminology ​49 no. 3 (2011): 795-831, doi:10.1111/j.1745-9125.2011.00240.x. 
[14] Daniel S. Nagin, “Deterrence: A Review of the Evidence by a Criminologist for Economists,” Annual Review of Economics​ 5 (2013): 83-105, doi:10.1146/annurev-economics-072412-131310.
[15] ​Ibid.; and Sherman, ​“The Rise of Evidence-Based Policing.” 
[16] Lawrence W. ​Sherman, Patrick R. Gartin, and Michael E. Buerger, “Hot Spots of Predatory Crime: Routine Activities and the Criminology of Place,” ​Criminology​ 27 no. 1 (1989): 27-56, doi:10.1111/j.1745-9125.1989.tb00862.x. 
[17] Cynthia Lum and Christopher S. Koper, ​Evidence-Based Policing: Translating Research into Practice​ (Oxford: Oxford University Press, 2017). 
[18] Ibid
[19] ​Nagin, “Deterrence.” 
[20] Robert J. Sampson, Stephen W. Raudenbush, and Felton Earls, “Neighborhoods and Violent Crime: A Multilevel Study of Collective Efficacy,” ​Science 2​ 77 no. 5328 (1997): 918-924, doi:10.1126/science.277.5328.918​; and Weisburd and Eck, “What Can Police Do.” 
[21] George L. Kelling, Tony Pate, Duane Dieckman, and Charles E. Brown, ​The Kansas City Preventive Patrol Experiment,​ Washington, DC: Police Foundation, 1974, NCJ 150523. 
[22] Martin B. Short, Jeffrey Brantingham, Andrea L. Bertozzi, and George E. Tita, “​Dissipation and Displacement of Hotspots in Reaction-Diffusion Models of Crime​,” ​Proceedings of the National Academy of Sciences of the United States of America 1​ 07 no. 9 (2010): 3961-3965. 
[23] Ratcliffe, Taniguchi, Groff, and Wood, “The Philadelphia Foot Patrol Experiment.” 
[24] Jerry H. ​Ratcliffe et al., “The Philadelphia Predictive Policing Experiment,” ​Journal of Experimental Criminology​ (2020), doi:10.1007/s11292-019-09400-2. 
[25] Christopher S. Koper, “Just Enough Police Presence: Reducing Crime and Disorderly Behavior by Optimizing Patrol Time in Crime Hot Spots,” ​Justice Quarterly​ 12 no. 4 (1995): 649-672, doi:10.1080/07418829500096231. 
[26] Nagin, “Deterrence.” 
[27] ​Jeremiah P. Johnson, “Testing Cruise Lights on Patrol: An Illuminating Experience,” Translational Criminology​ (Fall 2019): 20-21.