Lessons for the field: A checklist for fair and just data-driven policing
Police leaders must use data to improve public safety, officer well-being and perceptions of police legitimacy
By Chief Maris Herold and Tamara D. Herold, Ph.D.
Police agencies engaged in data-driven policing use data to identify and address patterns (e.g., in crime incidents and personnel behaviors). Data-driven policing improves strategic and tactical decision-making by enhancing agency capacity to detect problems and develop efficient and effective solutions to inform deployment and maximize the impact of limited departmental resources.
Some police reform advocates contend that data-driven policing represents a “threat to (society’s) constitutional rights,” and demand the dismantling of biased data systems (e.g., gang databases, prolific offender lists) and defunding of “discriminatory police technology” (e.g., predictive policing software, gunshot detection systems). 
Critics of data-driven policing argue that police data are biased, unreliable and inaccurate, and use leads to targeted harassment of vulnerable individuals and communities. Instances of injustice – calls-for-service from racially biased individuals, police targeting of persons mislabeled as gang-affiliated, unarmed persons shot by officers responding to gun-shot alerts, police cover-ups involving data manipulation (consider early COMPSTAT scandals ) – serve to support the perspective that data-driven policing is dangerous and unethical.
Police executives faced with data-driven policing criticisms should consider two facts:
- Police data are imperfect and subject to misuse or misinterpretation. Inaccurate reports, coding errors, missing information and many other issues compromise data integrity. Policing history is rife with examples of data corruption and misuse, both intentional and unintentional. Yet, these issues plague data use in medicine, education, agriculture, social work and all other professions. Nonetheless, it is difficult to imagine how ignoring data could improve patient outcomes or food safety. Police executives should not avoid using data, but they should work to improve the quality of data, create in-house capacity to interpret data, and promote continuous review of data-driven policy and practice outcomes.
- Failure to collect and incorporate data into police decision-making is simply unethical. Without data, police risk ineffective action (or inaction). Status quo and nonproductive practices, like random police patrols,  remain unchallenged and waste taxpayer dollars. Police risk producing or reinforcing social inequities when stop, arrest, citizen complaint and use of force data are uncollected or neglected. Supervisors need data to monitor officer decision-making, identify training deficiencies and intervene to correct problematic behaviors. Police leaders also need data to ensure officer health and safety. Data can track workplace stressors and identify tactics and strategies that continually place officers in harm’s way.
Data is not the enemy of good policing. But the absence, misreading, or misuse of data by police agencies is.
Fair and just data-driven policing in action
Policing research and leadership experience across three police agencies (major city, campus and mid-sized) highlight the benefits of data-driven policing and pinpoint five steps police executives can take to promote effective, equitable and ethical data-informed policing practices.
The following initiatives show how data-driven approaches can improve public safety while promoting police accountability and building community trust.
Step 1: Enhance data integrity, accessibly and interpretation
In 2020, the Boulder Police Department (BPD) lacked personnel and technology platforms to support data-driven policing. The agency hired an IT innovation specialist to build systems that integrate all city data and allow BPD’s executive team to track key performance indicators, including crime, case management and officer performance. In addition, public-facing dashboards were developed (e.g., police blotter, interactive crime maps) to increase agency transparency. A newly hired data scientist continuously improves the quality and type of data collected to prevent misinformed strategic decision-making. Further, skilled analysis turns data into actionable information. Executives’ budget decisions should prioritize these critical positions.
Step 2: Enable inclusive policing practices
Inclusive policing helps guard against potential unintended consequences of data-driven interventions. It is akin to what Dr. Nancy LaVigne, Director of the National Institute of Justice, calls inclusive research.  Inclusive policing asks those who are most impacted by crime to assist in developing and implementing solutions. It encourages participation by officers, dispatchers, victims, arrestees, social service providers and community members – anyone who has direct experience with the problem being addressed.
In 2015, the Cincinnati Police Department (CPD) observed a significant spike in shootings and developed the PIVOT (Place-based Investigations of Violent Offender Territories) strategy to reduce gun violence. A key component of PIVOT is community partnership. Community members who live and work in the city’s most persistent violent hot spots were involved in PIVOT’s development and implementation. PIVOT’s inclusive policing approach intentionally fostered community dialogue and partnerships resulting in:
- A better understanding of police data, including community observations that explained why gun violence clustered in particular locations
- Suggestions for alternative solutions to traditional criminal justice responses, including intervention by other city agencies like traffic and engineering to prevent drive-by shootings
- Community-driven neighborhood improvements (e.g., walking trails, parks, lighting, community gardens, grocery stores), rather than city-driven redevelopment projects
- Overwhelming community buy-in and improved police-community relations
SEE: CPD’s 2017 Herman Goldstein Award for Excellence in Problem-Oriented Policing Winning Submission
Step 3: Embrace evidence-based interventions
After a University of Cincinnati Police Division (UCPD) officer fatally shot an unarmed black motorist in 2015, UCPD launched an extensive police reform effort. The department subsequently implemented 276 recommendations, including evidence-based and data-driven practices, under voluntary external monitorship. The reform agenda called for proactive crime reduction strategies and extensive officer training.
Police executives used evidence-based resources to reduce numerous crime problems. For example, police reduced off-campus student burglary victimization by 30% the first year and over 70% the second year, using a combination of evidence-based strategies (e.g., Koper Curve directed patrols, improving natural surveillance, target hardening, awareness campaigns). When evidence did not exist concerning the effects of de-escalation training, UCPD partnered with researchers to collect data and ensure that trainings improved outcomes without sacrificing officer safety.
SEE: UCPD’s 2022 Herman Goldstein Award for Excellence in Problem-Oriented Policing Winning Submission and Evaluation of Police Use of Force De-escalation Training: Assessing the Impact of the Integrating Communications, Assessment, and Tactics (ICAT) Training Program for the University of Cincinnati, OH Police Division (UCPD)
Step 4: Empower data-driven decision-making
The Cincinnati Collaborative Agreement resulted from a class-action lawsuit alleging police brutality and biased policing. It has become an international model for improving police-community relations.
Problem-oriented policing (POP) is the main component of the historic agreement. Backed by strong research evidence, the POP process requires police to use data to identify and analyze crime problems and develop solutions in partnership with key stakeholders. It also requires police to collect data and evaluate their efforts. It provides the framework to engage in data-driven decision-making and inclusive policing. Further, POP improves officer wellness and safety. It eliminates long-standing problems that require a continuous police response, reducing stressors and risk of injury associated with repeated exposure to high-risk situations.
Under the Collaborative Agreement, Cincinnati experienced significant crime reductions, improved police-community relations, and reduced reliance on traditional criminal justice mechanisms, including arrests and incarceration.
SEE: CPD’s Collaborative Agreement and Center for Problem-Oriented Policing
Step 5: Ensure accountability
Traditional police organizational structures are built to be reactive (i.e., respond to calls for service and conduct investigations). Police executives must develop accountability models to create agency incentives and capacity to engage in proactive data-driven decision-making.
UCPD developed a Tactical and Strategic Investigations Policy and BPD adopted the Stratified Policing Model to align agency practices with problem-solving. Both models assign crime reduction roles and responsibilities to every rank within the department. They also mandate the implementation of evidence-based interventions and external partnerships. Regular meetings between executives and command staff ensure that personnel are engaging in problem-solving, meeting crime reduction goals and viewing crime prevention as a primary policing function.
SEE: UCPD’s Tactical and Strategic Investigations Policy and Stratified policing: An organizational model for proactive crime reduction
The five following action items will help executives to build a fair and just data-driven police agency:
- Hire personnel to enhance data integrity, accessibility, and interpretation
- Promote partnerships to enable inclusive policing practices
- Create a learning culture that embraces evidence-based policing interventions
- Use problem-solving to empower data-driven decision-making
- Adopt organizational models to ensure accountability
Executives who embrace these steps toward best practices in data-driven policing will promote agency efficiency and effectiveness and preemptively address critics who argue that police use of data negatively impacts equitable practices and ethical decision-making.
1. Díaz A. (September 13, 2021.) Data-driven policing’s threat to our constitutional rights. The Brookings Institution, Washington, D.C.
2. Francescani C. (March 9, 2021.) NYPD report confirms manipulation of crime stats. Reuters.
3. Sherman LW, Weisburd D. (1995.) General deterrent effects of police patrol in crime “hot spots”: A randomized, controlled trial. Justice Quarterly, 12(4):625-648.
4. La Vigne N. (December 7, 2022.) From the Director: Harnessing the Power of Data-Driven, Inclusive Research. NIJ.
About the authors
Police Chief Maris Herold (Boulder, Colorado) currently serves as the Co-Chair of IACP’s Research Advisory Committee and was recently inducted into George Mason University’s Evidence-Based Policing Hall of Fame. Chief Herold has led reform across major city, mid-sized and campus police departments during her 30-year policing career and is an internationally recognized crime reduction strategist and problem-solving expert. Connect with Chief Herold on LinkedIn.
Dr. Tamara D. Herold is a crime scientist and currently serves as a National Institute of Justice Senior Advisor and Associate Professor at the University of Nevada, Las Vegas. She studies crime and crowd dynamics and helps police agencies prevent crime and improve policing practices. Connect with Dr. Herold on LinkedIn.