A Process Evaluation of San Francisco’s Law Enforcement Assisted Diversion Program

• Author: Erica Jovanna Magaña,Aili Malm,Dina Perrone
• Published date: 01 March 2022
• Date: 01 March 2022
• DOI: http://doi.org/10.1177/08874034211033328
• Subject Matter: Articles

https://doi.org/10.1177/08874034211033328

Criminal Justice Policy Review

2022, Vol. 33(2) 148 –176

© The Author(s) 2021

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/08874034211033328

journals.sagepub.com/home/cjp

Article

A Process Evaluation of

San Francisco’s Law

Enforcement Assisted

Diversion Program

Erica Jovanna Magaña1, Dina Perrone2,

and Aili Malm2

Abstract

In 2016, San Francisco (SF) implemented the Law Enforcement Assisted Diversion

(LEAD) program, a harm reduction–based pre-booking diversion system for people

who violate drug laws and/or are engaged in sex work. LEAD is set apart from existing

diversion programs, as it uses police as point of entry. Prior LEAD studies indicate

some success in reducing recidivism and improving life outcomes. However, less is

known about program implementation, including barriers and facilitators. Relying on

policy documents, interviews, and focus groups, this study describes the LEAD SF’s

development, operations, adaptations, and challenges. It also identifies the unique

context of LEAD SF that led to implementation barriers and facilitators. Results

show that SF experienced success in collaboration, relationship building, and client

connections to services but experienced challenges in securing and maintaining police

officer buy-in and keeping clear and open lines of communication regarding LEAD

goals, objectives, policies, and procedures. This led to the termination of LEAD SF

in 2020.

Keywords

drug offenders, process evaluation, realistic evaluation, street-level bureaucrats,

pre-arrest diversion

1Washington State University, Pullman, USA

2California State University, Long Beach, USA

Corresponding Author:

Dina Perrone, California State University, Long Beach, 1250 Bellflower Blvd., ET 246, Long Beach,

CA 90840, USA.

Email: dina.perrone@csulb.edu

1033328CJPXXX10.1177/08874034211033328Criminal Justice Policy ReviewMagaña et al.

research-article2021

Magaña et al. 149

Introduction

Historically, effective policing was considered a craft, in which experience equated

with success. More recently, however, a new framework—Evidence-Based Policing

(EBP)—emerged that is shaping policing and shifting the understanding of effective

policing. Under an EBP approach, “police officers and staff create, review and use the

best available evidence to inform and challenge policies, practices and decisions”

(Ratcliffe, 2018, p. 185). Strong empirical evidence helps departments persuade politi-

cians to fund programs that work and avoid spending money on activities that are

shown ineffective. But how do we assess the strength of evidence?

The Maryland Scale of Scientific Evidence or SMS is one tool used to evaluate the

methodological robustness of outcome, impact, and cost–benefit evaluations (National

Institute of Justice [NIJ], 1998). The SMS ranks evaluations on a scale from 1 to 5,

with higher scores equaling stronger evidence support. Using the SMS, the NIJ denotes

effective programs to be those with at least two Level 3 or higher-rated studies

showing effectiveness (Justice Research and Statistics Association [JRSA], 2014;

e.g., CrimeSolutions.gov).

Police departments tend to rely, then, on quantitative methods to assess if a program

is evidence based. This has significant value but is limited. Most have weak external

validity (JRSA, 2014; NIJ, 1998; Pawson & Tilley, 1997), which makes generalizing

to other contexts difficult and poses replication challenges (Campbell et al., 2019;

JRSA, 2014; Pawson & Tilley, 1997). Outcome and impact evaluations also do not

explain how or why a program worked or failed (Mears, 2010; Miller & Miller, 2015;

Pawson & Tilley, 1997).

To overcome the shortcomings of experimental research and to best understand

why programs are effective or ineffective, Pawson and Tilley (1997) put forth realistic

evaluation. Realistic evaluation moves away from experimental approaches and

instead seeks to understand the underlying mechanism and context that produced the

results. The realistic evaluation approach is more than a description of what happened

but rather how it happened. Simply stated, it seeks to identify how a program works

(or does not) and under what conditions.

One underutilized evaluation method that can help understand mechanism–context

relationships is the process evaluation (Mears, 2010; Miller & Miller, 2015). Process

evaluations provide detailed information about a program’s underlying theory, model

design, goals, objectives, operations, service delivery, quality of services, and imple-

mentation barriers and facilitators (Krisberg, 1980; Mears, 2010). They also contextu-

alize impact and outcome findings by describing how and why an intervention

experienced certain results, which is helpful for replication purposes. This article pro-

vides a process evaluation to understand program implementation for the now widely

implemented Law Enforcement Assisted Diversion (LEAD) in San Francisco (SF),

California.

LEAD is one of the first U.S. pre-booking diversion programs that redirects indi-

viduals from criminal justice system involvement into community-based social,

health, and behavioral health services (LEAD National Support Bureau, n.d.-b). To