Skip to content

Building the Evidence for Evidence-Based Policing

Jeremy Travis addresses the annual Conference of the American Society for Evidence-Based Policing about the strategy of the Laura and John Arnold Foundation.

An NYPD car
Photo by Scott Roth/Invision/The Associated Press

Jeremy Travis, Executive Vice President of Criminal Justice, gave the following plenary address at the annual Conference of the American Society for Evidence-Based Policing.

I am delighted to be with you today. I look at the topics you are covering, the intellectual firepower you have assembled, the depth of experience represented in this room, the strength of the policing leadership gathered today in the birthplace of our country, and I am truly humbled and deeply impressed. In a very short time, the American Society of Evidence-Based Policing has earned its rightful place as a powerful force for reform in the policing community. By assembling leaders in the profession with policing scholars, you have signaled a willingness to open up the pressing questions facing law enforcement around with country to critical examination and rigorous inquiry.

And look at the topics you have taken on — predictive policing, body-worn cameras, policing and race, de-escalation, violence reduction, officer-involved shootings, officer well-being, and, perhaps the toughest challenge of all, how to become an Evidence-Based Organization. You have made this conference one of those few that earns the title, can’t be missed.”

I want to add a personal note of congratulations to Sgt. Renee Mitchell and Lt. Dan Wagner and the entire Board of the ASEBP for your dedication to the important mission of building evidence-based practice. Throughout my career — at the Vera Institute of Justice, in two positions at the NYPD, during six magnificent years as NIJ Director in the Clinton Administration, to the Urban Institute, then John Jay President, and now at the Laura and John Arnold Foundation — I have shared this commitment to using research, data, and good evidence to advance policing practices that are effective and fair. So I come to you today as an honorary member of your organization.

It is crystal clear to all of us that this is the moment for this movement. Over the past generation, the policing profession has made strides toward evidence-based practices that are truly impressive. We can cite the landmarks on this journey, beginning a generation ago with the Police Foundation’s landmark study on the effectiveness of random preventive patrol, including the Newark Foot Patrol study, the development and testing of hot spots policing, the articulation of the broken windows philosophy, the evolution of focused deterrence, the design of Compstat. The next frontiers are also exciting, with departments harnessing the power of data to improve operations, developing new metrics that include public trust and confidence, committing to the goal of data transparency, working collaboratively with other public agencies to improved outcomes, examining all police practices through the lens of racial justice, providing more data, generating more knowledge about what works, and creating more tools in their toolbox so that the entire organization reflects the commitment to evidence-based practices. It is a very exciting time to work on these issues.

The Criminal Justice Portfolio of the Laura and John Arnold Foundation

This morning I would like to share some information — and insights — from the policing portfolio of the Laura and John Arnold Foundation. Our Foundation is barely a decade old. Laura and John Arnold created this Foundation with a compelling mission: to improve the lives of individuals by strengthening our social, governmental, and economic systems. Their belief, which resonates with the mission of the ASEBP, is that reliance on data and evidence [will] lead us toward promising solutions.”[1] The Laura and John Arnold Foundation — called LJAF for short — has taken on a number of big challenges facing American society — education, health care, pension reform and criminal justice — and has, in a short time, become an important partner, working with government agencies, other organizations, foundations, and individuals committed to reform.

The Criminal Justice Initiative was launched in 2011 under the leadership of my predecessor, Anne Milgram, the former Attorney General of the State of New Jersey. The Foundation made a very wise choice in the first years of the Criminal Justice Initiative to focus on the front end of the criminal justice system, with a special emphasis on bail reform. This decision reflected two conclusions. First, they recognized — or, better, predicted — that the field of practice in this area, from policing to pretrial justice, was ripe for change. Second, Anne and the Board of the Foundation believed that the LJAF tools of data analytics, research, testing new ideas, and building support for policy change could make a difference.

These bets have paid off. Our signature contribution to the burgeoning pretrial reform movement has been the development and testing of a risk assessment instrument — called the Public Safety Assessment, or PSA — that can aid judges in making decisions on the release or detention of individuals facing criminal charges. The PSA has been implemented in 40 jurisdictions across the country with intensive technical assistance provided by LJAF. Those jurisdictions include six states, two municipalities, 32 counties, undoubtedly including some represented in this room. But we soon learned that this success had generated a demand we could not accommodate. Over 600 jurisdictions across the country approached LJAF to ask for assistance. So last month we released a Request for Proposals (RFP) to select a National Provider to work as our partner to meet this demand. Working together, we will select ten cities that are committed to implementing risk assessment as a component of their pretrial reform efforts. These cities will become our national demonstration sites. The National Provider will also provide technical assistance to another 200 jurisdictions over the next five years. And the fundamentals of the PSA will be made available to all jurisdictions on a web platform that will include guidance on the principles of effective implementation. At the same time, reflecting our commitment to research, we released a second RFP calling for implementation research in the ten demonstration sites, research on the next generation of risk assessment, and a broader examination of the impact of pretrial detention. Through these efforts, and others yet to be announced, we hope to make a significant contribution to the pretrial reform movement that will move the country toward the goal of reducing unnecessary pretrial detention.

The Policing Portfolio of the Laura and John Arnold Foundation

We are very proud of the fact that the Arnold Foundation has made a significant investment in studying the implementation of body-worn cameras. Our Board recognized that this practice would grow rapidly, boosted by the availability of new technology and a public demand for more transparency and accountability. To date, we have invested $5.3 million in this research portfolio, partnering with nine police departments across the country, including LA County, Albany, Prince George’s County, Anaheim, Pittsburg, Spokane, Tempe, Arlington, and Washington, D.C. Most of these projects involved RCTs, which will make it possible for the public and policy community to answer a fundamental question: beyond the evidentiary value in an individual case, what difference does it make for police officers to use cameras to record their interactions with residents? What are the costs? What are the benefits? It is far too early to have answers — much less definitive answers — to these questions. And some of the early trials have yielded conflicting findings. But we believe that this work will help shape the trajectory of this rapidly expanding practice to reflect the best available evidence on effectiveness.

Our Foundation has also fostered partnerships between universities and police agencies, most notably the Center for Policy Research and Policy, a collaboration between the IACP and the University of Cincinnati. Launched in 2016, the Center facilitates the exchange of data and research between the policing community and academia. (As a side note, let me express my gratitude to Dr. Robin Engel of the University of Cincinnati and Vince Talucci, Executive Director of the IACP, for generously hosting a day-long brainstorming session that helped shape the policing strategy we are announcing today.) LJAF has also invested in the Misdemeanor Justice Project at John Jay College. It began as a partnership with the NYPD, expanded to a data analysis project involving courts and corrections, and has now been expanded to university-police department partnerships in seven other cities — Durham, NC; Los Angeles, CA; Louisville, KY; Meridian, MS; Prince George’s County, MD; Seattle, WA; and St. Louis MO. We have also supported the work of the Proactive Policing Lab at George Mason University, the Center for Evidence-based Crime Policy at George Mason, the Policing Project at NYU, and many other research centers around the country. We believe we have made a difference working with these partners and look forward to continuing our collaboration.

Addressing Frequent Utilizers

At our one-day strategy session at the IACP, we put an issue on the table that resonates with all police agencies, namely, How best to exercise the power to arrest?” Our conversation was guided by a brilliant literature review prepared by Dr. Robin Engel, with assistance from Dr. Rob Worden and a team of graduate students. (Her paper will be made public shortly). The response to this topic by police chiefs and academia alike was very gratifying. There is a genuine hunger to look at this essential exercise of police discretion more systematically.

With this wind in our sails, we have started to lay out a more detailed set of plans, which now fall under the overall heading of Alternative to Arrest, or the A2A initiative, with three sub-headings. As a first venture into this terrain, we will test new interventions to address a problem known to every jurisdiction — the challenge of Frequent Utilizers,” those individuals who cycle between emergency rooms, homeless shelters, drug treatment facilities, police lock ups, and local jails. Last month, my colleague Lynn Overmann and the Sheriff of Middlesex County, Peter Koutoujian, announced a $1.6 million three-county demonstration project to use data integration to identify these frequent utilizers. (The other two counties are Johnson County, IA and the City of Long Beach, CA.) We also announced three RCTs to build the evidence base about what actually works to address the complex needs of Frequent Utilizers.[2] The goal? To develop interventions that will result in better life outcomes for these individuals, save money, and allow public agencies to use their resources more effectively.

Story of Joe J.

Let me give you one example, provided by Lt. Dan Wagner of Cambridge Police Department (also Vice President of ASEBP), that was discussed at the press conference launching this initiative. Joe J., a homeless man with severe alcoholism and health problems, is well known to Cambridge police. Of the 116 incidents that officers have logged in their case notes, many follow a familiar pattern: Over and over again, officers are called to assist Joe J. in getting to a shelter or hospital when he is no longer able to get there himself. From 2013 to 2018, case notes indicate that Joe J. has been arrested only once — part of a concentrated and strategic effort by the Cambridge PD to connect him to treatment.

Because of his substance use issue — and what officers suspect is an undiagnosed mental illness — efforts to get Joe J. a neurological evaluation have been fruitless so far. Joe J. also has many physical health problems. Officers have picked him up after he’s fractured his leg in three places, and when he’s suffering from hypothermia and frostbite. His case notes detail countless trips to local emergency departments for these conditions and for extreme intoxication. In just six weeks in 2017, Joe J. was taken to the emergency department 38 times. Over the course of two years — 2016 and 2017 — he had 371 EMS transports, with an average of three EMS transports every four days.

Officers, hospital staff, and shelter staff want to find long-term treatment for Joe J., but they’re struggling to figure out what the solution looks like. He has been Section 35ed — an involuntary hospitalization for substance use disorder — more than eight times in five years, but those experiences in treatment programs have been too short to provide long-term help. We know the most likely solutions to Joe’s challenges: long term, effective health and mental health care, and supportive housing. But until those solutions are identified and deployed, police will remain on the front lines of response to these complex social service issues.

We are committed to funding research to help us find those long-term, community-based solutions. But we know we also need to address the status quo. How can we provide better information to police and other first responders so they can identify people who may need an alternative to arrest or a trip to the hospital?

Story of 26-Year-Old Female

This next slide, also provided to us by Lt. Dan Wagner, is much simpler. It represents a young woman’s interaction with emergency police and EMT response. You can see someone who is on path that likely has a bad outcome. But if you look at just the police dispatch, or just the EMT dispatch, you get two very different viewpoints on the challenges this young woman is facing. Imagine if we could create a system that would give you the information you need when you next come in contact with her so you can connect her to the services she needs?

There is an exciting level of innovation in this arena — we are seeing diversion programs for individuals facing specific challenges such opioid use disorder and mental illness, including co-responder models that involve both police and mental health providers, crisis intervention teams, mobile crisis units, as well as deflection policies that bypass the arrest decision entirely. We are also seeing increased interest in crisis stabilization sites, as well as efforts to engage individuals in long-term treatment and services to address their underlying needs, such as permanent supportive housing. We are also interested in learning more about alternatives to custodial arrests, such as summonses, desk appearance tickets, and stationhouse releases. Please watch our website, or sign up for our email list, to be informed about upcoming RFPs and program announcements.

Analyzing Arrests for Low Level Offenses

A second area of interest will be tracking the use of the arrest power overall, particularly for low-level offenses. Two examples will illustrate the point. The first involves a natural experiment in Cincinnati. According to a recent thoughtful article by Dr. Robin Engel in Criminology and Public Policy[3] the City of Cincinnati was under a court order forcing it to reduce available jail space by 36 percent. Faced with this crisis, the Cincinnati Police Department realized it had to reduce the volume of low-level arrests so that there were fewer cases flowing into the county jail. By using strategies such as problem‐​oriented policing, place‐​based policing, and focused deterrence, Cincinnati Police were able to narrow their focus to the repeat places, problems, and groups of individuals that were driving crime within the city. The most fundamental challenge was to shift their officers’ view of the use of arrest power. They had to see an arrest as a limited commodity rather than as a standard response. The findings are impressive. Contrary to the predictions of some skeptics, the closure of the jail corresponded with a reduction in police-driven arrests for minor offenses and, more importantly, a continued decline in violent crime.

The second example comes from my hometown, New York City, and uses data from the Misdemeanor Justice Project, a research program I formerly co-directed with my John Jay colleague, Prof. Preeti Chauhan. The MJP documented significant shifts in enforcement levels over the past 36 years. As you will note, the rate of felony arrests has remained relatively constant over this time period.

Overall, the enforcement by the NYPD — combining all four trends — have dropped dramatically by 72 percent between 2011 and 2017. The net effect of these significant shifts is stunning: this past year, there were approximately one million fewer enforcement interactions between the NYPD and residents of New York than in 2011. And this trend occurred at the same time that crime rates in New York City continued to drop.[4]

In the coming months and years, we hope to engage with police departments and researchers around the country on their trends in the use of arrests for low-level offenses, as well as pedestrian stops. As this audience knows well, these high-volume interactions between police and citizens carry enormous risk, often fray relationships between police and the public, and frequently fail to resolve the problem that led to the call for service. In this connection, we are also interested in exploring the ways that arrests are used as a measure of productivity and hope to evaluate other ways to measure performance at the officer and departmental levels.

Reforming Emergency Response

Our third area of exploration, flowing directly from our brainstorming session at the IACP, is the complex issue of emergency response. The vast majority of the approximately 240 million 911 calls per year are unrelated to emergency events or crimes in progress, yet police are often the de facto responders.

We see opportunities for potential projects focused on improving and leveraging the emergency response to drug overdoses, people in mental health crises, or other issues such as homelessness, where the conduct at issue is more related to health or service needs than to criminal intent. There are further indications that the burden of calls for service has become greater for some agencies due to budget cuts and to the increasing complexity of issues that police are being asked to handle. It is critical that we increase law enforcement’s ability and capacity to safely respond to emergencies that involve mental health crises or substance use overdoses. Further, as communities begin to explore different models of emergency response, there is a growing interest in the field to better understand the use of 911 to initiate the emergency response. This includes understanding what kinds of calls come into 911, how they are currently analyzed and routed, what the outcomes of those calls are, and whether there may be opportunities to improve police efficiency by more effectively categorizing and routing 911 calls for service. As a first step, we are funding a few projects to begin this exploration of 911 calls for service as an important first step to identifying promising and feasible ways to increase police effectiveness and efficiency and improve outcomes for individuals and the community.

We will shortly begin a systematic exploration of the options available to police departments and the potential for collaboration with health, mental health, and other non-enforcement responses. We have been impressed with the intense interest in this topic. We are eager to engage with new partners to explore technological advances, innovative multi-disciplinary responses, and intensive problem-solving approaches to the complex neighborhood problems that result in calls to 911. We hope that our approach to research and development can produce evidence of effectiveness of these new approaches and that this evidence will result in the more effective allocation of public resources.

These are the policing issues on our front burner under the heading of Alternatives to Arrest. We will announce new grants under this heading and release RFPs in the coming months. So stay tuned. Beyond our front burner, we also have a number of ambitious ideas on our many back burners. In the near future, we plan to take a close look at ideas to improve the police response to violence — both community violence and domestic violence. We are also engaged in a cross-Foundation exploration of strategies that will improve the government’s response to the opioid epidemic. We are looking at law enforcement and criminal justice contributions to the national strategy on this crisis. Please stay tuned for future announcements on these topics.

In the coming years, we are looking forward to working closely with our partners in the policing profession and the research community, many of which are in this room. We want to collaborate with you to test ideas, promote a robust public discourse about new strategies, and develop a stronger evidence base of effectiveness. We can only do our work with your active collaboration. I thank you for the invitation to speak with you today, for your ongoing support, and for your work on behalf of communities across the country.

[1] Laura and John Arnold Foundation (2017), 2017: Year in Review.
[2] Rochester Forensic Assertive Community Treatment (R‑FACT): A research team based in the University of Rochester Medical Center is heading up a study in collaboration with Minnesota’s Department of Corrections to conduct a randomized controlled trial of a replication of R‑FACT. R‑FACT is a standardized model of FACT, an intervention that employs a multi-disciplinary team to address individuals’ with complex physical and mental health needs. The trial will measure the impact of R‑FACT on healthcare system use and recidivism rates.

Forensic Assertive Community Treatment (FACT): The City of Long Beach, in partnership with the Long Beach Criminal Justice Lab and California State University-Long Beach, will conduct a feasibility study followed by a full-scale randomized controlled trial to evaluate FACT-Wraparound, an expanded version of FACT. The model aims to address a broad range of mental illnesses that commonly occur among justice-involved individuals. The full-scale RCT will measure re-arrest rates, patients’ mental health and physical health, and recidivism rates.

Housing First combined with Shared Medical Appointments: MDRC is launching a large-scale feasibility study across three sites to determine the potential for a multi-site evaluation of Housing First combined with Shared Medical Appointments. It is based on a standardized version of Housing First that refers homeless patients to services, provides supportive housing, and matches peers across housing placements to attend medical appointments. If successful, this study will contribute to establishing a first-of-its-kind trial in the United States of an adjusted Housing First model that aims to decrease systems use among frequent utilizers with substance use disorders.
[3] Engel, R. S., Corsaro, N. and Ozer, M. M. (2017), The Impact of Police on Criminal Justice Reform. Criminology & Public Policy, 16: 375 – 402.
[4] For reports from the Misdemeanor Justice Project, click here.