By Mélanie Seabrook, Vardan Gupta, Alexander Luscombe, and Andrew Pinto

KEY TAKEAWAYS

Municipal police in Ontario are overburdened – they are called upon to respond to issues beyond their scope and training. Investing in a range of municipal services – including housing, public health, long-term care, social assistance, paramedics, and children’s services – would promote community health and wellbeing and reduce demand for police through preventing crime and other crises from occurring, freeing police capacity for core functions. Ontario municipal funding for services promoting health and wellbeing hasn’t kept up with police funding over the past 12 years. Despite public support, municipalities face challenges in de-prioritizing police budgets to reinvest in other services, mainly due to the influential role of police boards in budget-setting. Community Safety and Wellbeing Plans present an opportunity to better engage local communities in municipal priority setting and could support the reprioritization of resources in future budget-setting.

Toronto: University of Toronto, School of Cities, 2025. 21p.

By Rebecca Gambler

Customs and Border Protection (CBP) has seen a significant increase in the number of individuals they apprehend between U.S. ports of entry along the southwest border. This has resulted in overcrowding and difficult humanitarian conditions in its facilities.

CBP monitors the care of individuals in its custody in various ways. For example, staff are required to conduct and document welfare checks every 15 minutes for individuals who are sick or injured. However, Border Patrol (a part of CBP) does not have a mechanism to verify that staff have done so across field locations.

We recommended that Border Patrol implement such a mechanism.

Washington, DC: U.S. Government Accountability Office. 2022. 50p.

By Bob Libal with contributions from Setareh Ghandehari and Silky Shah

  Every day, thousands of people are cruelly deprived of their liberty in a vast system of mass immigration detention in the United States. For years, detained people and advocates have organized to close troubled immigration detention centers and exposed the horrors of a detention system rife with extreme negligence, abuse, and even death. Numerous studies document that detention is also wholly unnecessary.1 Despite overwhelming evidence that immigrants successfully navigate their immigration cases in community, the immigration detention system — now with over 230 facilities in the United States — has seen exponential growth across the last three presidential administrations. In just the last four years, the number of people detained by Immigration and Customs Enforcement (ICE) grew dramatically to an average daily population of more than 50,000 people in Fiscal Year (FY) 2019, by far the most in the agency’s history. This unprecendented expansion of detention was propelled not by changing migration trends, but by a resurgence of nativist and xenophobic rhetoric translated into harsher policies towards both arriving immigrants and long-term non-citizen residents. Detention expansion continued under the Trump administration despite draconian enforcement policies such as the Migrant Protection Protocols (MPP), the expansion of the border wall, and Title 42, meant to keep people from    arriving at the border or seeking asylum once they did. Due to a variety of factors, including detention numbers trending down, an ongoing global pandemic, and a shifting political landscape, the Biden administration has an opportunity to begin the process of phasing out immigration detention entirely. This report addresses one stated barrier to detention center  closures — the economic impacts of detention centers on host communities

Washington, DC: Detention Watch Network 2021. 29p.

By Neil Donnelly

To examine changes in the clear up rate for New South Wales (NSW) crimes by considering offence-specific trends in the NSW Police Force 90-day legal action rate over the 15 years from 2009 to 2023. METHOD Data were extracted from the NSW Police Force’s Computerised Operational Policing System (COPS) for incidents reported between 2009 and 2023. This includes the total number of incidents reported, the number of persons of interest legally proceeded against, and the percentage of incidents where police commenced legal action against at least one person within 90 days of reporting. Trends were examined separately across 11 major offence categories using the percentage point change and the total percentage change. The Kendall’s rank order correlation test to determine statistical significance. RESULTS Over the 15 years from 2009 to 2023, there were significant increases in the 90-day legal action rate for nine of the 11 offences examined. The largest percentage point changes were observed for robbery (up 19.9 percentage points (p.p.)), malicious damage to property (up 14.2 p.p.), break and enter non-dwelling (up 12.7 p.p.), break and enter dwelling (up 8.5 p.p.), motor vehicle theft (up 7.9 p.p.) and domestic violence (DV) related assault (up 6.5 p.p.). Smaller, though significant, increases were also found for steal from motor vehicle (up 3.0 p.p.), non-DV related assault (up 2.5 p.p.) and sexual touching, sexual act and other sexual offences (up 2.0 p.p.). By contrast there was a significant decline in the 90-day legal action rate for reported incidents of sexual assault over the 15-year period examined (down -3.4 p.p.). For most property offences, the upward trend in legal action rates has not caused a significant growth in criminal court workload. This is because the fall in crime rates over the last 15 years has been even greater than the rise in legal actions. This is not true for DV assault and sexual touching, sexual acts and other sexual offences. The combination of a rise in incident counts and the improvements in legal action rates has resulted in more people being brought before the criminal courts for these offences. The overall number of sexual assault incidents recorded by police over the 15-year period also increased. This increase offset the decline in the legal action rate in recent years, leading to a higher number of offenders being proceeded against to court for this offence.

Parramatta, NSW: NSW Bureau of Crime Statistics and Research (BOCSAR), 2025. 11p.

By Irshad Altheimer, Hayder Alhafedhi, Rashid Muhammad

his report provides a descriptive analysis of Rochester Police Department (RPD) traffic stop data. We analyzed 41,201 traffic stops performed by RPD between 2015 and 2023. The stop data consisted of 83 variables, but the analysis of many of those variables was hampered by high numbers of missing values. Increased scrutiny of law enforcement practices has led to calls for the collection and analysis of police stop data. Such data analysis is essential for fostering accountability and transparency within the police departments. Police officers in the United States conduct at least 50,000 traffic stops every day, making these stops a central part of modern policing and the most common way in which the public interacts with law enforcement (Policing Project, 2019). Beyond the frequency, traffic stops often lead to significant financial hardship for individuals through fines and fees, and evidence suggests that both traffic and pedestrian stops disproportionately impact people of color (Policing Project, 2019). Despite how common these stops are, there is still much we do not understand about traffic stop practices and the full effects of traffic stops. Analyzing this data enables the identification of patterns and trends within policing practices, including disparities in traffic stops based on factors such as race, gender, or geographic location (Policing Project, 2019). This helps to assess the effectiveness of current policing strategies and ultimately develop informed policy reforms aimed at promoting equitable policing practices. Transparency is crucial in law enforcement because it builds trust between police departments and the communities they serve. When the actions, decisions, and data of law enforcement agencies are open to public scrutiny, it demonstrates a commitment to accountability and ethical practices. This openness allows citizens to understand the reasoning behind police actions, reducing misunderstandings and promoting cooperation. Strong relationships based on mutual trust between police and the communities they serve are vital for effective policing and public safety; community members are more likely to cooperate with law enforcement and provide crucial information when they trust that police actions are fair and reflect community values (Community Relations Service, n.d.). Transparency is a key element in fostering this trust, as timely and open communication about critical incidents helps reassure the public that information is not being deliberately withheld (Community Relations Service, n.d.). This report is divided into the following sections. First, we provide an overview of the data. Second, we provide a descriptive analysis of the data that could be reliably analyzed. Third, we provide recommendations for future data collection.

Working Paper 2025-01 - 01

Rochester, NY: Center for Public Safety Initiatives Rochester Institute of Technology, 2025. 15p.

By Iain Britton

The Commonwealth of Massachusetts has a proud tradition of volunteers in law enforcement stretching back to colonial times (Greenberg, 2015). Until very recently many police departments, large and small, had volunteer auxiliary police. Many of which had existed essentially in their current forms since the 1940s and 1950s (Spigel, 2017). This reflects the abundant picture of tens of thousands of volunteer auxiliary and reserve police in hundreds of law enforcement agencies across most states in the USA (Dobrin and Wolf, 2016). Sadly, recent changes have seen the auxiliary model in Massachusetts almost entirely administratively swept away. What was a large, and in many ways thriving, and highly impactful, volunteer auxiliary policing model in Massachusetts, has all but gone. Changes in ‘POST’,The Massachusetts Peace Officer Standards and Training (POST) Commission1 certification standards at state level across Massachusetts, had the effect of wiping out all but a tiny number of auxiliary units. This rapid extinction event removed volunteer auxiliary police almost overnight from their communities across Massachusetts, communities that many of those individual auxiliaries had served in with dedication for decades. This represents the largest single volunteer police extinction event in the USA for decades. It serves as a worrying harbinger for volunteer policing more widely, especially in other states across the USA but also worldwide. Heralding the very real and urgent strategic threat to continuity of volunteer police programmes. Something that is resonant in the UK, as we see for the twelfth successive year an annual reduction in the number of volunteer special constables.2 Across the USA and internationally, some volunteer police models are thriving, with volunteer police impacting in a wide range of ways (Wolf and Borland Jones, 2018). Some are growing or innovating – the Netherlands, Estonia and Hungary are strong national examples, with significant scale, and interesting innovation of roles. However, many more programmes are struggling to survive (Britton, 2024). Many volunteer police programmes around the world are struggling to maintain sustainable numbers, with the combined headwinds of generational changes in reduced patterns of voluntarism, unfavourable attitudes towards police volunteers, volunteer police models with too narrow appeal and accessibility, and increasingly hostile public perceptions of policing particularly in respect of conduct, race, and misogyny. Despite strong evidence of return on investment (Britton et al., 2023) many forces are also struggling to effectively adapt and redefine the role, capability, professional identities and operating models for the volunteer police officer in a fast-changing, and ever more complex, litigious, dangerous and contested policing environment (Britton and Callender, 2018). Retention of longer-service volunteers is proving in many contexts to be a particular challenge (Britton, 2023).

London: The Police Foundation, 2025. 7p.

By Gabrielle T. Isaza, Ryan T. Motz, Hannah D. McManus, Nicholas Corsaro, and Amanda M. Shoulberg

The report “Examining Police Reforms in New Jersey: Impacts on Officer Attitudes and Self-Reported Behavior” is the first of a series of publications presenting the findings from the statewide evaluation of police use of force reform in New Jersey.

It describes the research team’s examination of the impact of the mandated training—including the Police Executive Research Forum’s Integrating Communications, Assessment, and Tactics (ICAT) de-escalation training and Georgetown University’s Active Bystandership for Law Enforcement (ABLE) peer intervention training—on officers’ perceptions, attitudes, and self-

reported behaviors.

The findings represent offics’ responses to training surveys immediately before, after, and one to two years following their training participation. The high response rates to the surveys—ranging from 12,623 to 17,036 responses at pre- and post-training—offer insights representative of law enforcement officers across New Jersey.

Arlington, VA: National Policing Institute, 2025. 217p.

By Andy Higgins

The Police Foundation’s Blueprints project explores the design choices made by English and Welsh police forces in the delivery of local policing – specifically in relation to the operating models they adopt to provide incident response, neighbourhood policing, local investigation and public protection. It takes as its starting point the considerable diversification and frequent change in local policing models that has occurred within and between police forces over the last one to two decades.

Project aims

Describe and (as far as possible) codify, the variety of local police operating models being practised across England and Wales, explain why they have developed and explore the rationales behind them.

Investigate what (if anything) can be concluded about the strengths and weaknesses of different approaches, based on quantitative performance

analysis (where possible) and local experiences and learning.

Look, in-depth, at the way policing models enable and constrain police practitioners working in four functional areas: incident response, neighbourhood policing, local investigation and public protection.

Explore what can be learned from the way local policing is organised in other countries.

Assess the suitability of different design options against future policing challenges, drawing on the analysis presented in the Strategic Review of Policing in England and Wales, and focusing on strategic enablers such as preventative partnership and public cooperation.

The Police Foundation, 2024. 21p.

By Sarah Bosman

Fencing constitutes knowingly and willingly dealing in stolen goods, such as smartphones, vehicles or jewellery. It includes a large variety of activities, not only buying, trading and selling stolen goods, but also safely transporting and storing these goods. It is often referred to as a ‘victimless crime’. Typically, the selling and purchasing of stolen goods is a voluntary exchange in which both the seller and customer are helped instead of harmed. Moreover, many of the customers do not even realise they are buying stolen goods. For this reason, no one will report it, meaning that little is known about the prevalence of this phenomenon. Fencing inherently cannot exist without a different prior offence (e.g. burglary or theft) occurring. According to criminal law, this constitutes two separate crimes, as first there is a theft, followed by selling or trading these stolen goods. Yet, from the offender’s perspective, these actions fall under the same objective, which is obtaining money. On one hand, the fences make an easy profit, and on the other, the thieves can acquire money to buy something else. In order to prevent fencing, it is necessary to implement a holistic approach that includes all the relevant partners and targets every aspect of the phenomenon. For this reason, four separate focus areas have been identified that together make up the general phenomenon (see figure 1). These focus areas are: (1) preventing (valuable) goods from being stolen, (2) preventing fences and their customers from selling as well as buying stolen goods, (3) tackling stolen goods markets in general, and (4) the creation of local and (inter)national partnerships. Many of the initiatives contained in this paper are already being implemented in practice and have great potential. The aim of these initiatives is to make it more challenging to deal in stolen goods and consequently also to discourage theft to begin with.

Brussels: European Crime Prevention Network (EUCPN) 2022. 24p.

By Michael Skidmore 

In 2022-23 there were an estimated 3.5 million fraud offences in England and Wales, with members of the public now more likely to fall victim to fraud than any other type of crime (ONS, 2023a). And accordingly, the police are seeing an overwhelming rise in reported fraud, with levels of recorded crime exceeding one million offences, reflecting not only fraud against the public but also the considerable impact on businesses (ONS, 2023a). These patterns are largely the consequence of living in an increasingly digitised society in which the opportunities to perpetrate fraud have proliferated. All this crime is reduced into one single offence category of ‘fraud’, which covers both a large volume and a wide variety of offenders, offending, victims, harm, and vulnerability. This paper focuses specifically on ‘online fraud’, forming part of a wider programme of work looking at fraud that is enabled by the internet and digital technology. The paper reviews the literature with the aim of unpacking the nature and particular characteristics of online fraud. It also examines how data and knowledge about fraud inform and direct the strategic and operational responses of the police and other organisations. The complexities of producing a ‘true’ picture of fraud are explored, including a discussion of the meaning and significance of fraud when it is ‘online’. It highlights the gaps and challenges in our current understanding of online fraud that will be addressed in our ongoing research programme  

Paper 10. 

London: The Police Foundation April 2024  16p.

By Melissa M. Labriola, Jill Portnoy Donaghy, Tiffany Keyes, Sarah Junghee Kang

oncerns about the physical health, mental health, and safety of first responders and law enforcement officers have been increasing for some time. The goal of this research is to synthesize evidence from the growing literature on mental health and wellness programs studied with law enforcement and first responder populations to help the U.S. Department of Homeland Security (DHS) identify and strengthen programs and policies and to conduct an evaluability assessment (EA) to provide direction for future research.

This report presents findings from multiple research tasks, including a review of domestic and international literature on first responder wellness programs and interviews with key stakeholders in DHS about existing DHS wellness programs, wellness program implementation, and subsequent challenges. Authors conducted an EA of programs identified as potentially ready for evaluation in the stakeholder interviews. The authors of this report synthesized the findings from these tasks to develop a research agenda for future DHS wellness research efforts.

Key Findings

• According to the literature within the scope of our parameters, the most studied wellness programs for law enforcement and first responders were group prevention skills and knowledge training, psychotherapy, physical fitness, and mindfulness training.

• According to our interviews, certain program types, such as suicide prevention training, physical fitness programs, mindfulness training, and mandatory postvention efforts, were generally viewed as effective.

• Overall, more research is needed with larger sample sizes, rigorous designs, and outcomes other than knowledge change.

• Some interviewees recommended adopting robust, evidence-informed, non-DHS programs and adapting them for DHS's population, with the related concern of the uniqueness of the agency and the applicability of programs designed for different populations.

• Stigma in seeking and receiving services is a critical barrier but, according to interviewees, might be improving.

• Organizational barriers to participation include (1) a lack of resources to fund programs and/or on-site mental health clinicians, (2) employees feeling that they do not have sufficient time to seek out support (especially employees who have very demanding roles), (3) employee concerns about confidentiality, (4) insufficient leadership support, and (5) a lack of awareness about the resources available to employees.

• More research is needed with larger sample sizes, rigorous designs, and outcomes other than knowledge change, such as mental health and wellness outcomes.

Santa Monica, CA: RAND, 2025, 107p.

By The Organization for Security and Co-operation in Europe

This paper is a summary of the first of a series of roundtable discussions that aim to identify opportunities for law enforcement to harness new technologies to support its work, help formulate policy recommendations and explore potential OSCE capacity-building support in this area. The first event of this series was dedicated to the topic of harnessing new technologies to enhance crime analysis, and focused on opportunities and challenges in deploying artificial intelligence (AI) for analysis and the potential impact of these technologies on human rights.

\

Prague: OSCE 2025. 12p.

By Brian A. Jackson, Pauline Moore, Jennifer T. Leschitz, Benjamin Boudreaux, Jo Caulkins, Shoshana R. Shelton

Violence by K–12 students is disturbingly common. Ensuring that schools have effective ways to identify and prevent such incidents is becoming increasingly important. Various concerning behaviors or disturbing communications, including direct threats, can precede acts of violence. Although removing every student exhibiting such behaviors might seem prudent, doing so can be counterproductive, limiting the effectiveness of safety efforts. With effective systems for behavioral threat assessment and management (BTAM), schools can assess and respond to concerning behavior to protect the community and respond to the student whose behavior caused concern.

To do so, schools need the tools to respond. Tools may include restrictive measures or law enforcement involvement in the most serious cases, but other options can be more effective. Those additional options include different types of mental health intervention, counseling, and other supports. Teams with extensive tools available to them can better customize interventions, increasing the chance of positive outcomes for all involved.

In this report, the authors draw on published literature and extensive interviews with education and public safety practitioners to build an inventory of the many intervention options that are valuable for schools in the management phase of BTAM. In addition, drawing on varied approaches from the fields of counseling, school discipline and behavioral management, and other professions that must match appropriate services to the needs of youth in their care, the report discusses decision support tools to help management teams implement this critical part of efforts in preventing targeted violence and keeping school communities safe.

Key Findings

Various Intervention Options Are Available for K–12 BTAM Efforts

Support-focused interventions can address the underlying causes of problematic student behavior and also lead a student toward a more favorable, positive path into the future.

By using supportive counseling and other interventions, BTAM is widening the options available for school leaders and staff to address problematic behavior that has the potential to develop into violence.

To be effective, school BTAM teams need a broad set of tools, including options appropriately matched (1) to the specifics of a student's problematic behaviors, (2) to the unique school community and environment, and (3) to the needs and circumstances of the student or students involved.

Insights from Education, Public Safety, and Other Fields Can Be Combined to Support Matching Effective Interventions to Student Needs

Decision support tools and resource-matching guidance can help ensure that school-based teams are collecting the information required to taking a holistic approach to address a student's varied needs and help promote appropriate consistency to ensure that disparities in how BTAM teams respond do not substantiate concerns that BTAM processes are unfair or inequitable.

Using structured systems to capture data when a BTAM team (1) interviews students involved in an incident, (2) collects school or law enforcement data, or (3) contacts others for information about a student of concern provides a more straightforward starting point for selecting among multiple intervention options.

Recommendations

To better inform intervention planning, intervention tools should be designed so that they prioritize collecting data on factors that can be changed because pieces of information in BTAM that may be a useful part of assessing the danger posed by an individual may be useless for intervention planning.

The inventory of intervention options developed in this study could provide a starting point for schools to make conscious decisions as they (1) review the options available to their teams and (2) identify any options they do not have access to but that could become valuable near-term priorities to strengthen their school safety efforts.

Progress monitoring data of BTAM efforts can help better support students while also helping schools become more responsive to external oversight of their BTAM programs and allay concerns about the fairness and equity of outcomes across different student populations.

Including positive mileposts into threat management planning not only could help lay out a path to full completion of all intervention activities but could also help define goals more specifically for an at-risk student, motivating even more beneficial outcomes.

Santa Monica, CA: RAND, 2025.

By National Academies of Sciences, Engineering, and Medicine

Mental, emotional, and behavioral (MEB) disorders—including mental illness and substance use disorders—are at the heart of several ongoing national crises and affect every U.S. population group, community, and neighborhood. Existing infrastructure responds to these crises predominantly with treatment and recovery, or addressing MEB disorders once they already exist, rather than working to prevent them. Available prevention services are insufficiently funded, fragmented, and better developed for substance use prevention than mental health promotion and for children and youth than for other age groups. Improved prevention services could help people thrive, avert the harms that accompany MEB disorders, and reduce the burden on an already overtaxed system. In response to a request from the Centers for Disease Control and Prevention (CDC), the National Institutes of Health (NIH), and the Substance Abuse and Mental Health Services Administration (SAMHSA), the National Academies of Sciences, Engineering, and Medicine (the National Academies) convened a committee of experts to develop a blueprint, including actionable steps for building and sustaining an infrastructure, for delivering prevention interventions for behavioral health disorders. The committee’s report, Blueprint for a National Prevention Infrastructure for Mental, Emotional, and Behavioral Disorders, presents its conclusions and recommendations. The committee asserts that the existing MEB disorder prevention infrastructure—partially present in other systems like education, health care, and human services—provides a foundation to build on and that creating another system would be inefficient. Instead, the report’s conclusions and recommendations focus on strengthening, coordinating, and funding existing structures to close gaps, prepare workers, and maximize available data to deliver needed interventions .

Washington, DC: The National Academies Press. 2025. 384p

By Emma,Sutherland Disley, Alex,Sussex, Jon,Pollard, Jack,Saunders, Catherine L.,Morley, Katherine I.,Gkousis, Evangelos,Hulme, Shann

This report presents an outcome evaluation of the National Model for Liaison and Diversion (L&D) in England, which aims to identify and support vulnerable individuals within the criminal justice system (CJS) by linking them with appropriate health and social care services. The evaluation utilizes a novel, large-scale linked dataset, combining healthcare and criminal justice records, to assess the impact of the National Model on health service utilization, re-offending rates, diversion from the CJS, and the timeliness of court processes. Findings suggest L&D services engage individuals with multiple vulnerabilities, intervene at crisis points, and may increase diversion from custodial sentences. However, the evaluation found no overall reduction in re-offending following L&D referral.

Santa Monica, CA: Cambridge, UK: RAND, 2021. 232p.

By Kelly Lyn Mitchell, Julia Laskorunsky, Ebony Ruhland, and Tammy Dean

Community supervision, commonly known as probation or parole, involves people serving part of their sentence under supervision while living in the community. Supervision conditions are requirements that individuals must comply with during this period, such as engaging in a treatment program, maintaining employment, or regularly checking in with their probation or parole officer. However, the current condition-setting approach often focuses on setting restrictions on behavior without providing meaningful guidance for behavioral change. This policy brief proposes aligning supervision conditions with the Risk-Needs-Responsivity (RNR) framework to improve outcomes for individuals on supervision and the community. 

Policy Brief, Minneapolis, MN: Robina Institute of Criminal Law and Criminal Justice. 2023. 7p.

By Megan Welsh Carroll , Shawn Teresa Flanigan , and Nicolas Gutierrez III

This paper examines racialized encounters with the police from the per-spectives of people experiencing homelessness in San Diego, California in2020. By some estimates, homelessness doubled in San Diego during the first year of the COVID-19 pandemic. We conducted a survey of (n ¼ 244)and interviews with (n ¼ 57) homeless San Diegans during initial shelter-in-place orders, oversampling for Black respondents, whose voices are often under-represented despite high rates of homelessness nationally. Ourrespondents reported high rates of police contact, frequent lack of respect;overt racism, sexism, and homophobia; and a failure to offer basic services during these encounters. Centering our Black respondents’ experiences of criminalization and racism in what Clair calls “criminalized subjectivity,” we develop a conceptual framework that brings together critical theoretical perspectives on the role of race in the governance of poverty and crime.When people experiencing extreme poverty face apathy, disrespect, and discrimination from police—as our data show—the result is a reluctance to seek services and to engage with outreach when offered. This reinforces stereotypes of unhoused people as not “wanting” help or “choosing” to be homeless. We reflect on these findings and our framework for envisioning a system of public safety that supports and cares for—rather thanpunishes—the most vulnerable members of our society

PUBLIC INTEGRITY2023, VOL. 25, NO. 3, 285–300

By Will Engelhardt and Daniel S. Lawrence

Before the Juvenile Justice Reinvestment Act was implemented in December 2019, North Carolina was the last state that still automatically charged 16- and 17-year-olds as adults in its criminal legal system. In March 2014, led by then–chief district court judge Marcia Morey, a group of stakeholders from Durham County, North Carolina, started the Misdemeanor Diversion Program (MDP) to prevent 16- and 17-year-olds from entering the criminal legal system. The first of its kind in North Carolina, the program provides services including life skills courses, restorative justice efforts, and behavioral health treatment over a 90-day period and has expanded to include adults of all ages. It has also been replicated in certain counties throughout the state. The MDP enables law enforcement officers in Durham County to redirect people accused of committing their first misdemeanor crime(s) to community-based services in lieu of charge, citation or arrest. The purpose is to diminish unnecessary arrests and jail time and the collateral consequences of being charged with and convicted of a crime. A central feature of this program is that it occurs prearrest and precharge, meaning someone law enforcement officers believe may have committed a crime will not be arrested or charged and will not formally enter the justice criminal legal system in any way. In 2020 and 2021, with support from the John D. and Catherine T. MacArthur Foundation’s Safety and Justice Challenge Research Consortium, the Urban Institute conducted in-depth process and impact evaluations of the MDP, the findings of which we summarize in this report. By conducting both types of evaluations, the research team was able to better understand the processes and context that led to observed impacts. In addition, this is the first time a third-party research organization has evaluated the program’s impact, and such an evaluation is critical to demonstrating the program’s usefulness. Key takeaways from the process evaluation (A Process Evaluation of the Misdemeanor Diversion Program in Durham County, North Carolina) and the impact evaluation (An Impact Evaluation of the Misdemeanor Diversion Program in Durham County, North Carolina) are detailed in box 1.

Washington, DC: Urban Institute, 2021. 24p.

By Daniel S. Lawrence, Will Engelhardt, Storm Ervin, Rudy Perez

Before the implementation of the Juvenile Justice Reinvestment Act in December 2019, North Carolina was the last state that still automatically charged 16-to-17-year-olds as adults in its justice system. In March 2014, a group of stakeholders from Durham County—led by then–chief district court judge Marcia Morey—started the Misdemeanor Diversion Program (MDP) to prevent 16-to-17-year-olds from entering the justice system. The program has since expanded to include adults up to 26 years old. The first program of its kind in North Carolina, the MDP gives law enforcement officers in Durham County the discretion to redirect people accused of committing their first misdemeanor offense(s) to community-based services (such as life skills courses, restorative justice efforts, and behavioral health treatment) in lieu of citation or arrest. The purpose was to diminish unnecessary arrests and time in jail and the collateral consequences of being charged with and potentially convicted of a crime. What is particularly unique about this program is that it occurs prearrest and precharge, meaning someone law enforcement officers believe may have committed a crime will not be arrested or charged and will not formally enter the justice system in any way. This impact evaluation, the first conducted for the MDP, found that from March 2014 to February 2020, law enforcement officers in Durham County referred fewer than one-quarter of all people eligible for diversion to the MDP, though when they did, the program had positive impacts. In 2020 and 2021, with support from the John D. and Catherine T. MacArthur Foundation’s Safety and Justice Challenge Research Consortium, the Urban Institute conducted an in-depth impact evaluation of the MDP, the findings of which are detailed in this report. This impact evaluation was one component of Urban’s research on the MDP; Urban also conducted a detailed process evaluation that was described in a July 2021 report, A Process Evaluation of the Misdemeanor Diversion Program in Durham County, North Carolina (Engelhardt et al. 2021). Key Takeaways The data examined in this report cover January 2012 to February 2020 and were collected from North Carolina’s Administrative Office of the Courts, the MDP, the Durham Police Department (DPD), and the Durham County Sheriff’s Office. Box 1 provides five key findings the research team derived from these data. In this report, we assess the following: ◼ MDP enrollment ◼ MDP completion rates ◼ the MDP’s impact on new arrests, convictions, and jail admissions for program participants ◼ the MDP’s impact on disparities by race and ethnicity, sex, and age ◼ the MDP’s impact on system-level arrests, convictions, and jail admissions Analyses were separated into two population groups—people ages 16 to 17 and people ages 18 to 21—because each group was eligible for the MDP during different periods. These groups were statistically matched to comparison groups through propensity score matching for the analyses that examined new arrests, convictions, and jail admissions. The comparison groups were well balanced with the MDP participant groups (see appendix D) and were pulled from pools of people who were concurrently eligible for the program but did not participate. Five Key Findings ◼ Approximately 77 percent of people eligible for the MDP were not referred to the program while it was operational from March 2014 to February 2020. ◼ Of those who did participate in the program, there was a very high completion rate of 95 percent. ◼ MDP participants had significantly lower rates of rearrests, convictions, or jail admissions than comparison groups within six months, one year, and two years. ◼ Participation in the MDP significantly reduced disparities in new arrests within two years and in new convictions and jail admissions within six months between 16-to-17-year-old Black people and non-Black people, making the differences in the levels of new arrests between these groups much more equivalent than between Black and non-Black people who did not participate in the MDP. ◼ The MDP did not have a larger impact on countywide rates of arrests, convictions, or jail admissions for either of the two age groups we analyzed

Washington, DC: Urban Institute, 2021. 83p.

By Brittany C. Cunningham, Vincent Bauer, Kira Cincotta, Jessica Dockstader, Benjamin Carleton, Bridgette Bryson, Daniel S. Lawrence

Officer safety is of critical importance in an era of increased risk for law enforcement officers (hereinafter “officers”). Officers respond to some of the most unpredictable, traumatic, and violent encounters of any profession. Although much of an officer’s workday entails repetitive interactions, some calls for service or self-initiated contacts by officers may escalate into dangerous encounters. For officers to adequately mitigate the risks they may encounter while responding to calls for service, they must be well informed regarding the types of risks they face, the situations that may pose greater risk, and the strategies that will mitigate these risks.

Although previous empirical work on officer safety has yielded many important insights, to our knowledge, no prior work has applied machine learning models to produce risk assessments to promote officer safety. This project explored the potential for machine learning to identify high-risk incidents to officers using only the information available to dispatchers. A risk assessment model that could successfully flag high-risk incidents at dispatch would be immensely useful to law enforcement agencies, making it possible for officers to be better informed about potential risk factors before arriving on scene. Such a model would also be useful to agencies as they decide how to allocate scarce resources, such as deciding which calls should receive single- or dual-officer vehicles, where to send alternative response teams, and whether to deploy specialized units.

Readers should be aware that the model reflects the data upon which it is built. Biases in reporting and collecting officer injuries, as well as in how officers respond to calls for service, will be mirrored in the model’s risk assessments. While we have gone to great lengths to build the model using objective factors, these biases could sometimes lead the model to identify a situation as high risk when in fact that situation reflects low risk to officers. Concerns about the potential for bias in machine learning are important to evaluate, and these techniques offer opportunities for objective empirical examination of divisive topics to minimize the bias that is already present in the real world.

Calls for service and Law Enforcement Officers Killed and Assaulted (LEOKA) data were merged from each of the four agencies, revealing the following findings:

Overall, the machine learning model performed well, correctly identifying officer injuries about half of the time. Given the rarity of officer injuries within the four agencies, being able to identify half of such rare situations is notable.

The model was also able to identify the factors that were the most important in predicting risk to officer safety and the types of incidents that posed the highest risk to officer safety. The results demonstrate that such a model can identify officer injuries from data on call characteristics; thus, whether such a model could be built into the dispatch process should be explored so that officers would be informed about potential risk factors before arriving at the location of a call.

The model highlighted factors and calls for service types that posed greater risks to officer safety.

The results of the machine learning model, along with the results from the officer interviews and surveys, also highlighted an often-overlooked aspect of police operations that is critically important to officer safety: dispatch.

Beyond producing statistical models, this project also collaborated with participating agencies to explore officer perspectives on safety and identify promising practices and recommendations to reduce risks to officers.

This project provides several practical benefits for improving officer safety. These benefits include the following:

Quantifying concepts that until now have been only informally or qualitatively understood (e.g., the relative risks of different calls for service types).

Comparing officer perceptions about injury risk to the quantitative data and identifying where gaps in understanding exist.

Highlighting the important relationship between dispatch and patrol, as well as the implications that this relationship has for officer safety.

Helping agencies assess the efficacy of their trainings and policies that directly affect officer safety.

Providing guidance on the information agencies collect and make available to dispatchers.

Supporting agencies to improve the amount and quality of risk and injury data agencies collect and use.

We hope that by providing agencies with a foundational knowledge of risks to officer safety, agencies will have a basis for modifying policy, training, and operations, leading to the implementation of strategies, processes, and procedures to keep officers and the communities they serve safe.

Arlington, VA: CNA Corporation, 2024. 52p.