From Taxi to Takeoff: Planning and Implementing Early Diversion in Los Angeles

By: Chidinma Ume, Darcy Hauslik

Courts Diversion Jail Populations June 15, 2023

The last several years have ushered in a seismic shift to Los Angeles County’s criminal justice landscape. Home to the world’s largest jail system, LA County achieved an unprecedented 25 percent decline in its jail population–the largest in the nation during the height of the COVID-19 pandemic. While the overall number of people in jail decreased, the percentage of people of color and people with mental health needs behind bars in LA increased. This changing composition mirrored a national trend and illustrated a key lesson: without a parallel effort to promote racial equity and provide safe, community-based care for people who need it, reducing jail populations may actually worsen disparities.

To address this, LA County—a grantee in the John D. and Catherine T. MacArthur Foundation’s Safety and Justice Challenge—announced a new vision of “Care First, Jails Last” and deepened its investment in community-based support for vulnerable populations and alternatives to incarceration. LA County also launched the Justice, Care, and Opportunities Department to consolidate most of these efforts under one roof.

As members of the Center for Justice Innovation (The Center) team, we are technical assistance partners in the Safety and Justice Challenge and worked with LA County to expand their alternatives to incarceration with particular emphasis on equity. Inspired by the learnings from this work—and grounded in our long history of launching and operating diversion programs—our West Coast Initiatives team authored a new report. The report offers concrete insights to inform the development of equitable diversion programming around the country. For practitioners seeking to create diversion programs, this document offers advice on designing early alternatives to incarceration, leveraging data to identify and connect with target populations, and working towards racially equitable outcomes.

We collaborated with LA County to launch two initiatives—the Rapid Diversion Program, which is court-based, and the Prefiling Diversion Program, which is based in law enforcement stations. Both programs aim to reduce the use of jail by connecting participants to safe, appropriate, and community-based care.

While successful diversion programs can safely reduce the use of incarceration, special care must be taken to ensure that these programs are carried out in an equitable way. We hope the insights of the report can provide guidance for developing diversion initiatives that bridge the gap between legal systems and communities while caring for vulnerable populations.

The first section of the report offers tips for developing the essential components of an early diversion program. Recommendations include:

  • Create infrastructure to divert people at the early stages of a case.
  • When determining eligibility criteria, prioritize the client profile over charges. When creating diversion programs, justice partners typically determine charge types to include and exclude in programs. Instead, we suggest program partners determine the profile of people they want to serve—for example, mental health, people with three or more arrests/system contacts, young people, race and ethnic groups that are disproportionately represented, etc.—and let that guide program development.
  • Make charge-based exclusions (sex offenses) presumptively instead of categorically ineligible. This means that instead of having categorical exclusions based on what people may be charged with, presume people are eligible for programs and, when someone is facing a presumptively ineligible charge, allow for discussion on a case-by-case basis.
  • Even within the same municipality, recognize that each diversion site may operate differently and have a distinct culture.
  • Seek out cross-sector collaborations and expertise in the program planning phase.

The opportunity to provide community-based care to people with unmet social service needs can happen at stages that far precede a criminal conviction—any time before a criminal case is adjudicated, and indeed, even before criminal charges are filed. For example, LA’s Prefiling and Rapid Diversion Programs utilize police stations and courts as potential off-ramps from the traditional legal system path.

Both programs pursue a common objective: to expand early interventions for people with unmet needs rather than continued detention or release without any supportive resources. To accomplish this, LA located behavioral health care professionals in the jails and courthouse. For Prefiling Diversion, this meant physically converting unused breathalyzer rooms and offices into spaces for care by placing service navigators in the station. Service navigators find programs and help people enroll in them and understand how to get connected to the program (i.e. when and where to go for intake, whom to call for questions). The Rapid Diversion Program embedded pairs of service navigators and clinicians in courthouses.

Four roles can improve the diversion infrastructure:

  • Mental Health Clinician—screens candidates for behavioral health conditions and appropriate level of services for the behavioral health program someone will need.
  • Service Navigator—identifies healthcare and social service needs, finds local programs and providers, and connects participants to these organizations and services.
  • Case Manager—supports participants one-on-one. Often the main point of contact for program participants, case managers provide referrals for continuing needs, such as education, employment resources, benefits, and housing, and help participants stay engaged in the program.
  • Driver—takes participants to their agreed upon destinations, oftentimes directly from the police station or courthouse to appointments, referred services, and future court dates. This is especially important for jurisdictions where transportation equity is a challenge. Although the driver’s primary role is to transport program participants, the driver frequently interacts with program participants and serves as an additional level of support.

Building these roles into any diversion program—and locating these professionals at the booking station or courthouse where possible—can help ensure that people with specialized knowledge connect participants to resources in a coordinated way. Interactions with program staff are also supportive touchpoints, which is made possible by hiring staff who understand the needs of participants and want to help. This includes people with lived experiences (including families impacted by the criminal justice system), previous program graduates, and people with clinical backgrounds. Remaining intentional about including and staffing each role, especially case managers and drivers who interact with participants frequently—can make even the mandatory components of diversion programs motivational.

The second section of the report includes recommendations for using data to promote equitable practices for diversion. The report stresses the need for prioritizing data collection as a critical tool in ensuring equity. Data analysis can help to identify underlying needs and shape the design process prior to program launch. After launch, a consistent flow of data among partners is necessary to sustain the program and gives planners the ability to adjust the program as needed.

Recommendations include:

  • Use relevant and detailed data at the planning stage to ensure equity and effectiveness of programming.
  • Review program performance data on an ongoing basis to ensure the right people are being served.
  • Make data planning a team effort.
  • Clarify roles and responsibilities around data management.

When launching a program meant to achieve equity, program planners should learn how to meet the needs of the desired population. Instead of looking at the types of charges to divert, examine who is being charged and any trends that can inform the demographic to serve (e.g. people who are unhoused but arrested for quality of life offenses, or people arrested on a drug offense but who are excluded based on prior drug convictions). Having a better understanding of common issues—not just charges—that people face can radically shape programming.

In addition to using data on the front end of design, the report recommends reviewing program performance data on an ongoing basis to ensure program efficacy and equity. Recurring reviews should include program staff who can speak to participants’ growth in the programs, especially to help the program collaborators understand ways to improve and sustain the program so it continues to have its intended impact.

Many partners in Los Angeles County make the Prefiling and Rapid Diversion Programs possible, including the LA County’s Justice Care, and Opportunities Department,the Offices of the Los Angeles County Public Defender and Alternate Defender, the Los Angeles City Attorney’s Office, and Los Angeles County District Attorney’s Office, the Department of Mental Health, Project 180, and Exodus Recovery, Inc.


Diversion Incarceration Trends Probation Sanctions January 12, 2023

Probation Violations as Drivers of Jail Incarceration in St. Louis County, Missouri

Beth M. Huebner, Lee Ann Slocum, Andrea Giuffre, Kimberly Kras, and Bobby Boxerman

Many have argued that we are in the era of mass probation, as more people are under probation supervision than under any other correctional sanction. Although there have been declines in the national probation population over the past decade, one in 84 adult US residents is currently on probation. Nationwide, local jail populations have also grown—from 184,000 in 1980 to 741,900 in 2019. The increased use of probation inflates the population at risk of subsequent confinement in jail or prison. Individuals who violate their probation, in some states, are detained in jail and await a hearing. Despite the growth in probation revocations and the increased use of jail stays as a response to technical violations, however, there is little evidence to suggest that short-term stays of incarceration reduce recidivism.

Adding to the growing rate of probation is the problem of racial disparity in incarceration. People of color are disproportionately represented among the probation population. In 2018, Black people represented 30% of the US probation population, twice their proportion in the national population. Further, almost half of all young Black men (24 to 32 years old) with no high school degree reported having been on probation at some point. Black individuals, particularly young men, are also more likely than White individuals to struggle on probation and to be given multiple conditions of supervision. Although there is evidence that Black individuals are more likely to have their probation revoked, less is known about how revocation to jail influences trajectories and outcomes for this group.

Jail stays also have deleterious effects in the short and long term. For example, Harding and colleagues found that short terms of jail incarceration resulting from technical violations suppressed the earnings of individuals by about 13% in the nine months after release from custody. The churn of multiple jail stays, even if short in length, also causes strain and instability among families, leaving them feeling hopeless under the constant eye of supervision. Yet, the unique needs of jail populations overall, and those of individuals who violate probation terms, are rarely considered in correctional reforms.

Additional Downloads


Diversion Featured Jurisdictions Plea Bargains November 29, 2022

An Exploration of Prosecutorial Discretion in Plea Bargaining in Philadelphia

Andreea Matei, Lily Robin, Kelly Roberts Freeman, and Leigh Courtney

As we have come to reckon with our nation's overreliance on carceral punishment and the mass incarceration of people of color, particularly Black people, experts are turning to a key system point that is the primary method for resolving most criminal cases: plea bargaining. Plea bargaining involves negotiation between a prosecutor and, often, a defense provider on behalf of their client. Prosecutors hold a lot of discretion over how to proceed regarding plea bargains, including whether to offer a plea agreement, when to do so, and what they wish to offer. Despite the wide use of plea bargaining, little is known about the practice, largely because it happens outside of public view and little is documented by the key actors involved—prosecutors.

To better understand prosecutorial discretion in plea bargaining, the Urban Institute was funded by the MacArthur Foundation through the Safety and Justice Challenge (SJC) Research Consortium, which is managed by the CUNY Institute for State and Local Governance (ISLG), to conduct a study on plea bargaining policies, practices, and outcomes. The Philadelphia District Attorney's Office (DAO) agreed to partner with Urban to shed light on the inner workings of plea negotiations and how they are viewed by different parties involved in the process, including attorneys and people who accept pleas. The DAO's partnership provided a rare opportunity to learn more about prosecutorial decisionmaking in plea bargaining in a single office and how this could inform policy and practice more broadly. This unparalleled look into prosecutorial decisionmaking owes to the forthrightness of the assistant district attorneys (ADAs) we interviewed and surveyed. The DAO's cooperation made it possible for Urban's research team to read policies on plea offers, analyze a deidentified sample of the office's case files, and hear from the ADAs to learn more about their decisionmaking during plea negotiations. Notably, this report is an exploration of discretion in plea bargaining in one office, not an impact evaluation of policies.

In this report, we discuss findings from our exploratory single-site study, in which we used qualitative and quantitative data to answer research questions of interest. Our activities included a policy review; analysis of administrative data from 2018 to 2021; interviews with 11 Philadelphia ADAs, 9 defense providers, and 5 people who accepted pleas; a case file review of 115 cases; and a survey of 65 ADAs. Because prosecutorial discretion in plea bargaining is not well documented in data, the best way to learn about discretion is by speaking with prosecutors; thus, this report focuses primarily on our qualitative findings. We organized our findings by three main topics: policies and goals of plea bargaining, trends in plea offers and outcomes, and decisionmaking and perceptions of key actors. We end the report with a discussion of policy implications.

Additional Downloads


Diversion Plea Bargains Racial Disparities November 29, 2022

Exploring Plea Negotiation Processes and Outcomes in Milwaukee and St. Louis County

Don Stemen, Beth M. Huebner, Marisa Omori, Elizabeth Webster, Alessandra Early, and Luis Torres

Although guilty pleas are the modal method for criminal case resolution in the US, relatively little attention has been paid to the plea negotiation process. Research suggests that prosecutors drive plea decision-making; however, the decision process is largely hidden and informal. Consequently, little is known about the role that prosecutors and other criminal justice actors play across the process, and even less is known about how these mechanisms have changed over time, particularly during the COVID-19 pandemic. Unpacking these plea negotiation decisions are especially key to understanding racial and ethnic disparities in criminal case processing.

Funded as part of the John D. and Catherine T. MacArthur Foundation's Safety and Justice Challenge Research Consortium, the current study considers guilty plea negotiation processes and outcomes in Milwaukee County, Wisconsin, and St. Louis County, Missouri. Both offices are currently lead by reform-oriented attorneys, are are medium-sized offices serving urban and suburban jurisdictions. Over the long tenure of elected District Attorney John Chisholm in Milwaukee, the office has implemented innovative prosecution models such as community prosecution units and diversion programs. In St Louis, recently elected District Attorney Wesley Bell is the first Black person to hold the office, and he ran on a platform of ensuring equity in the system and reducing mass incarceration. The goal of the study is to explore how prosecutors and other court actors approach and make decisions surrounding the plea negotiation process, in addition to, investigating the factors that affect plea outcomes. The data used in this report include narratives from interviews with and surveys of local stakeholders including prosecutors, public defenders, judges, private attorneys, and system-involved persons. The report also centers on administrative data collected through agencies' case management systems for criminal cases filed in Milwaukee and St. Louis Counties through 2020.

Additional Downloads

Why Law Enforcement Should Be Doing More Deflection as A Primary Response

By: Shannon Magnuson, Amy Dezember

Diversion Frequent Utilizers Interagency Collaboration August 17, 2022

New research by Justice System Partners supported by the MacArthur Foundation’s Safety and Justice Challenge (SJC) shows the positive impacts of police-led deflection strategies on jail reduction efforts. Overall, “deflection first, arrest rare” as a primary policy for eligible offenses helps reduce criminal legal system involvement and improve equity by connecting individuals to the services they need. It no longer makes access to treatment conditional or contingent on arrest.

Download the report here. 

Deflection is different from diversion. Diversion programs make use of pending criminal charges as the mechanism to elicit treatment initiation and compliance from people living with severe mental health disorders and substance use disorders. Although diversion programs do not always include a formal booking to jail, the person does technically enter the legal system’s front door. In contrast, deflection programs entail no criminal legal system involvement beyond the interaction with the police officer in the field. There is no mechanism to coerce treatment initiation or compliance beyond an individual’s own wishes to enter a program. If an individual ultimately decides not to participate in the program they are referred to, there are no legal consequences. Police-led deflection programs also provide police agencies an opportunity to return to the streets and answer calls from 911 more quickly because transporting individuals to community services can take substantially less time than booking an individual in jail. Combined, police-led deflection can make police agencies more efficient while eliminating the collateral consequences of the legal system on individuals.

Police-led deflection allows officers to use discretion to replace arrest with outreach to community-based service providers for select offenses. It transforms police contact into opportunities to broker community resources, especially for individuals with severe mental health disorders and substance use disorders. It is key that we understand how deflection programs work in practice if we are going to improve and expand these programs, reduce jail populations, and improve equity and access to care by connecting people to the help they need.

The goal of the research was to understand how deflection works in Pima County, AZ, and in Charleston County, SC—two SJC sites. We sought to understand how police make decisions about who to deflect and how deflection to a local crisis center impacts people’s subsequent experience. In 2011, Pima County built the Crisis Response Center (CRC) with county bond funds. It is part of the Banner-University of Arizona Medical Center South Campus. The CRC is a short-term inpatient unit with a maximum length of stay of five days and has a “no wrong door” policy – which means they accept nearly everyone, except individuals who require hospitalization, from any law enforcement agency in the county. Similarly, Charleston County runs the Tri-County Crisis Stabilization Center (TCSC) which is a ten-bed voluntary adult crisis center embedded within the Charleston Drug and Alcohol Center where individuals can stay up to 14 days. Both counties’ crisis centers provide immediate treatment options and psychiatric care for individuals.

In Pima County, when people receive at least two voluntary deflections to the local crisis center, they are more likely to continue agreeing to subsequent deflections and, each time they return to the crisis center, they stay longer. In Charleston County, two-thirds of individuals deflected to the local crisis center had a previous case with the county’s mental health department, showing that both police and service providers are often interacting with the same people. These findings suggest that police-led deflection can help connect or re-connect individuals with treatment while reducing the number of individuals who are brought to jail—effectively creating a parallel treatment open door. Importantly, these repeat access points to treatment reflect research that suggests people need multiple opportunities to access treatment services before agreeing and engaging with the program. In this way, police-led deflection, particularly for individuals who have disproportionate police contact, can transform pathways to jail into pathways to the help they need for individuals historically excluded from access.

A parallel treatment open door does not mean failure to initiate or complete treatment. The treatment open door acknowledges the complexity and nuance of treatment and reflects the research about the need for multiple opportunities to remain engaged. When police policy allows multiple deflections of the same individuals, as in Pima and Charleston Counties, it means every interaction with individuals is another opportunity to engage them in treatment while eliminating the collateral consequences of the legal system and jail for these vulnerable populations.

In both counties, if the offense is eligible, the policy allows police to offer deflection to an individual regardless of how many times they offered deflection to the individual in the past. This means that while the policy itself reflects the research on multiple opportunities to engage in treatment, police in practice have the ultimate decision-making authority to deflect or arrest. As a result, police hold a lot of decision-making power for triaging people out of the legal system revolving door, and into a treatment system revolving door. As we continue to unpack how officers make decisions about who to deflect and under what conditions, it is important to consider the intersection of race, gender, and disability and how that impacts officer decision-making about who is “worthy” of deflection.

Officers in Pima County reported people’s willingness to start treatment as the most critical factor when deciding to deflect to a community-based resource. When people did not want to initiate treatment, officers tended to rely on arrest, even when they knew that jail would not help the individual. But the ability to deflect for certain eligible offenses means that the police have determined that no arrest is an acceptable response. This tension demands a critical examination: If people do not wish to go into treatment, are there other ways officers can diffuse and handle the situation without relying on arrest?

Officers in Charleston County cited victims’ wishes—including those of business owners—as the most critical factor when deciding whether or not to begin the process of deflection. For example, when victims remain at the scene and want police to arrest the individual, even when the offense is deflection-eligible and the officer recognizes jail is not helpful, police expressed feeling inclined to defer to the victim and make the arrest, anyway. This means victims are, in part, driving who is offered deflection and potentially contributing to disparate deflections. As such, we must critically examine the role of victims in the deflection initiation process and consider how victims’ own perceptions of justice and implicit bias can temper the positive impact of police-led deflection programs.

“Deflection first, arrest rare” as both policy and principle connects vulnerable individuals, who are historically excluded from the services they need, with easy access to treatment. It also lessens opportunities for implicit bias, determinations of “worthiness,” and non-clinical judgments about readiness for change to impact the decision to deflect. When agencies distance themselves from jail and deflect as the primary response, and do so for everyone, they no longer make access to treatment conditional or contingent.