CHILD WELFARE RE S E AR C H RE P O R T Evaluating California's Efforts to Address the Commercial Sexual Exploitation of Children Ivy Hammond, Wendy Wiegmann, Joseph Magruder, and Daniel Webster UNIVERSITY OF CALIFORNIA, BERKELEY, CALIFORNIA CHILD WELFARE INDICATORS PROJECT Bridgette Lery, Sarah Benatar, Jaclyn Chambers, Laura Packard Tucker, Katrina Brewsaugh, Annelise Loveless, and Jonah Norwitt URBAN INSTITUTE July 2023 AB O U T T HE U R BA N I NS T I T U TE The Urban Institute is a nonprofit research organization that provides data and evidence to help advance upward mobility and equity. We are a trusted source for changemakers who seek to strengthen decisionmaking, create inclusive economic growth, and improve the well-being of families and communities. For more than 50 years, Urban has delivered facts that inspire solutions-and this remains our charge today. AB O U T T HE C AL IF O R NIA C H IL D WE LF A RE I NDI C A T O RS P R O JE C T The California Child Welfare Indicators Project (CCWIP) is a collaborative venture between the University of California at Berkeley (UCB) and the California Department of Social Services (CDSS). The project is housed in the School of Social Welfare and provides agency staff, policymakers, researchers, and the public with access to critical outcome information on California's child welfare system. Further information can be found at https://ccwip.berkeley.edu/. Copyright © July 2023. Urban Institute. Permission is granted for reproduction of this file, with attribution to the Urban Institute. Cover image by Tim Meko. Contents Acknowledgments iv Executive Summary v Background 1 CSEC Program Description 3 Evaluation Overview 5 Methods 9 Data Collection and Sources 9 Analysis 16 Findings 19 Implementation Study 19 Outcome Study 45 Discussion 65 Limitations 65 Key Findings and Recommendations 67 Conclusion 76 Notes 79 References 82 About the Authors 84 Statement of Independence 86 Acknowledgments This report was funded by the California Department of Social Services (CDSS). We are grateful to them and to all our funders, who make it possible for Urban to advance its mission. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders. Funders do not determine research findings or the insights and recommendations of Urban experts. Further information on the Urban Institute's funding principles is available at urban.org/fundingprinciples. We are grateful to our partners in the Child Trafficking Response Team and the Research, Automation, and Data Division, who provided rich content knowledge and support for this research. We thank the public child welfare agency staff and other county staff who participated in the survey and interviews for this study. We also thank the lived experience experts who gave their time and generously shared their experiences and insights in focus groups. This evaluation was informed by subject matter experts, many of whom experienced CSE during childhood. We would like to recognize the integral contributions of Alia Dewees, Cari Herthel and Tkeyah Dupclay, who collaborated with us to implement the evaluation, interpret findings, and present our work to stakeholders. We also want to acknowledge the Child Welfare Council's Commercially Sexually Exploited Children (CSEC) Action Team and the Preventing and Addressing Child Trafficking (PACT) program for their critical guidance and feedback. We thank Mike Pergamit, Bill Dawson, and Stephanie Cuccaro-Alamin for their careful review, and Shannon Gedo, Emily Ross, John Prindle, and Sofia Guo for their important contributions. iv ACKNOWLEDGMENTS Executive Summary The commercial sexual exploitation (CSE) of children and young people is a human rights concern and a public health challenge. CSE refers broadly to any activity or crime that involves the sexual abuse and exploitation of a child for monetary or nonmonetary benefit ranging from providing a phone or jewelry in exchange for a suggestive picture to coercive sex in exchange for money or a place to live. While information about CSE prevalence among young people in the United States is limited, research suggests that CSE during childhood can have serious consequences for its survivors, including exposure to violence and other traumatic events, mental health disorders, reproductive health complications, and internalized coping behaviors. Many states have decriminalized young people's involvement in the commercial sex industry. CSE has been formally established as a type of child maltreatment, with child welfare agencies given the responsibility to investigate and intervene on behalf of young people experiencing CSE. Federal legislative changes over the past decade have prompted states, including California, to develop policies aimed at addressing commercial sexual exploitation of children and young people. In 2014, California's Senate Bill (S. B.) 855 amended the state's welfare codes to create the state's Opt-In Commercially Sexually Exploited Child (CSEC) Program, which gives participating county child welfare agencies guidance and funding to prevent and intervene on behalf of children who are experiencing or at risk of CSE. Between July 1, 2015, and June 30, 2021, California's child protection system received roughly 70,000 maltreatment reports alleging CSEC. About a quarter of these reports were substantiated, meaning there was enough evidence to conclude that CSE of a child likely occurred. The California Department of Social Services (CDSS) contracted with the California Child Welfare Indicators Project at the University of California, Berkeley (CCWIP) and the Urban Institute to conduct an implementation evaluation led by Urban and an outcome evaluation led by CCWIP and CDSS's Research, Automation, and Data Division. This executive summary reports key findings, promising practices, and recommendations from our evaluation of the state's CSEC program. Key Findings from the Implementation Study The implementation study focused on opportunities for continuous quality improvement and cross- system collaboration. We gathered data from annual county program plans and a CSEC program administrator survey. We also conducted key informant interviews with agency and provider staff and focus groups with adults who experienced CSE as minors in a subset of 12 counties. Key findings from EXECUTIVE SUMMARY v the study highlight the importance of effective collaboration, training, staffing, data management, and system coordination to better serve and support young people at risk of or experiencing CSE.  Implementing S. B. 855 has fostered strong interagency collaboration and communication. Interagency collaboration has improved following S. B. 855, and those we spoke with reported positive relationships among agencies engaged in the county's CSEC response. In most counties, agencies other than child welfare services also screen for young people at risk of or experiencing CSE. Key informants cited progress toward working closely across agencies that had previously been siloed, and many counties were successfully incorporating continuous communication. However, collaboration is uneven. Approximately one-third of counties indicated that collaborating agencies are not in full agreement on the success of their efforts to work together, and the same share expressed dissatisfaction with interagency communication.  A shortage of placements exacerbates CSE service challenges. The shortage of placements appropriate for young people experiencing or at risk of CSE came up repeatedly. Many foster caregivers are unwilling or unready to take young people experiencing CSE. California's Continuum of Care Reform (CCR) and the federal Family First Prevention Services Act1 reduced the availability of congregate care beds in favor of family-based placements. Even though CCR invested in placement alternatives, the demand for appropriate placement settings far exceeds what is available.  Staff shortages are ubiquitous. Turnover reduces trust among children, families, and county agencies; fragments training efforts; and erodes institutional knowledge about CSE.  Standardized training on CSEC is widespread, but not enough training is available for parents, relative caregivers, and foster parents. Training should also be updated to align with evolving CSE recruitment trends. Standardized CSEC 101 and 102 training, which is funded by S. B. 855, is widespread within child welfare agencies and generally well regarded. Sometimes training is offered to the broader community and often in schools. Training for parents, relative caregivers, and foster parents is less common. Many people we spoke with said this is a major gap, noting that caregivers need to know the signs of CSE and what to do if they are concerned.  The Commercial Sexual Exploitation-Identification Tool (CSE-IT) is regularly used, although its implementation is not universal. Because S. B. 855 does not mandate a specific screening tool, some counties are using other or additional assessments to screen for CSE, typically either in child welfare or probation agencies. We heard from some county staff that the tool is not always used as effectively as it could be-specifically, screening may happen but findings may vi EXECUTIVE SUMMARY not be shared with the staff responsible for identifying young people at risk of CSE or deploying resources for them. In addition, there are concerns that the CSE-IT is overly sensitive and that the definition of "at risk" is too broad.  S. B. 855 gave child welfare agencies responsibility for caring for this population, but many feel they have inadequate tools to be successful and sometimes feel undermined by other agency priorities. Child welfare staff said they are not given enough support from partner agencies or the necessary tools to fulfill their mandate to effectively help young people experiencing CSE. Child welfare staff bear the primary responsibility for the safety and care of these children but expressed concern that their mandates sometimes conflict with other stakeholders. For instance, some counties are dual-status counties (i.e., a youth can be under the jurisdiction of both the child welfare agency and probation) and some are not (i.e., the county has to pick one agency to have primary responsibility for the youth). This can facilitate collaboration, but also sometimes produce disagreements and territorialism. The lack of a shared agenda can undermine interagency collaboration.  Multidisciplinary Teaming (MDT) approaches are inconsistent, but the consequences of uneven implementation appear to be minimal. Counties vary in how they understand and use an MDT process to support young people experiencing CSE. In some cases, MDTs and child and family team meetings are conflated. In other cases, MDTs and steering committee meetings are interchangeable. In both situations, staff feel that the teaming required to coordinate resources and collaborate on decisions is largely taking place effectively. The act of bringing collaborators together regularly appears to make the difference for collaboration, not the name, structure, or specific implementation of the MDT approach.  Counties would benefit from greater intercounty service coordination. The collaboration required under S. B. 855 is narrow, referring only to specific agencies within a county. But some small (and a few large) counties collaborate across county lines where services are not available in county or where there is significant geographic mobility among young people. Some counties noted that key information was often lost when a child welfare case was transferred between counties. State support and encouragement for broader collaboration across more in-county systems and across county lines can better serve young people who may have complex needs from any number of systems and may be placed out of county.  Improved timely and accurate data collection, cross-county data sharing, and standardized measures to track the CSEC population would enhance counties' responses. Many counties reported that data tracking contributed to interagency collaboration, served as helpful baseline EXECUTIVE SUMMARY vii information for agencies to begin discussing a case, and assisted with faster referrals and specialized placements. There has been progress and success in collecting and sharing data among agencies, but there is still room for improvement in accurately and comprehensively identifying and serving young people who are at risk of or experiencing CSE.  It is challenging to serve young people experiencing CSE who are not formally involved with the child welfare system. Many counties did not have a process for serving young people who do not have an open child welfare case, nor a clear understanding of roles and responsibilities for which agency has oversight for these young people. Many key informants felt that this was a major gap in a county's ability to effectively respond to CSE. Additionally, young people and families in voluntary family maintenance cases do not have access to the same funding streams for services as court ordered in-home or placement cases (e.g., Medi-Cal). Key Findings from the Outcome Study In the outcome study, we examined child welfare system involvement for young people after S. B. 855's implementation. We analyzed information recorded in the statewide administrative database to describe the experiences of 38,168 young people who met California's definition of CSEC or were identified as being at heightened risk of experiencing CSE. We studied the identification of CSEC, documentation practices, revictimization, case involvement, and foster care placement trajectories.  The majority of CSE reports are screened in for investigation, but a minority of those investigated are substantiated. Nearly two-thirds of the 70,334 CSE reports made between July 1, 2015, and June 30, 2022, in opted-in counties were screened in for investigation. Among those, one in five were deemed inconclusive and nearly one in four reports were substantiated. The high percentages of law enforcement and school personnel represented among CSE maltreatment reporters suggests that partner agencies have benefited from education about CSE.  Young people with documented concerns of CSE are primarily female, English-speaking, and people of color. Almost two-thirds of S. B. 855–eligible young people were identified as female. Slightly less than half were identified as Hispanic/Latinx, one-quarter were identified as White, one in six were identified as Black, almost one in twenty were identified as Asian or Pacific Islander, and less than 1 percent were identified as Native American/Alaskan Native. More than 80 percent of eligible young people spoke English as their primary language. The median age of children at the time of initial CSE concern was 12 years. viii EXECUTIVE SUMMARY  Repeat confirmation of CSE victimization does not appear to be a common practice among opted-in counties. Among the young people with CSE reports, one in five had a subsequent CSE report that was screened in for investigation within a year of their initial CSE report, and one in twenty children had a substantiated CSE report within 12 months of their initial investigation.  A minority of young people were in child welfare cases or placements when CSE concerns were identified. Nearly one in seven young people were in an open child welfare case when CSE concerns were first documented, a little more than one in ten were in an out-of-home care placement, and 2 percent were absent from placement but in an open foster care episode at the time of CSE identification. Among young people with confirmed CSE, 4 percent had some foster care placement history but were not in care when CSE concerns were documented, more than one in nine were in a placement, and nearly 3 percent were absent from placement.  Congregate care was the most common predominant placement setting, followed by Foster Family Agency–supervised homes and kinship care. The median number of unique homes or facilities young people experienced within 12 months of an initial CSE concern was two, and one in three young people spent time in a single placement home or facility. Those predominantly placed in group care had more placement moves than those placed primarily with either kin, a county foster home, or in a Foster Family Agency home.  Cases and placement openings were more common among young people with confirmed CSE victimization. Among all young people with CSE concerns, about 15 percent had a case opened and one in ten entered foster care within 12 months of the concern. In contrast, nearly half of those with confirmed experiences of CSE had a case opened, and one in three entered foster care. One in four young people with confirmed CSE had a documented placement absence at some point during the 12 months that followed an initial CSE concern.  Cases and placements were opened more quickly among young people with confirmed CSE victimization. The median time to foster care placement was 29 days among those with confirmed CSE and 38 days across all young people with CSE concerns. Promising Practices for S. B. 855 Implementation County staff and other stakeholders identified several strategies and approaches that have helped them implement S. B. 855 and better serve this population. EXECUTIVE SUMMARY ix  Assigning CSE cases to specific frontline workers rather than distributing them throughout the workforce was described as a helpful practice, even in counties where it is not practical to have workers who only carry CSE cases. Children experiencing or at risk of CSE benefit from workers who have developed specialized knowledge about their needs.  24/7 dual responses from child welfare and CSE advocates (voluntary nonprofit) when going out for investigations may improve assessment of CSE and engagement with young people.  On-staff clinicians and staff dedicated to recovering missing young people may benefit county efforts. One county has a unique staffing structure. Within their CSE team, they have (1) a CSEC clinician who is a licensed therapist and serves as a mental health resource for young people experiencing CSE and (2) a missing persons liaison who carries all missing-youth cases and does active searches for those young people.  Partnering with outside organizations can be effective in connecting at-risk young people who are not child welfare involved. At least two counties refer young people who are not child welfare involved to partner organizations equipped to make referrals outside of child welfare, hold an MDT, and connect young people to services.  Weighting CSE cases more heavily when calculating caseloads may protect against burnout by acknowledging that such cases are more work intensive. For example, a CSE case is counted as two when calculating caseloads in one county.  Some large counties have a trauma-informed court for CSE cases, which provides a safe and knowledgeable setting to consider child welfare and associated issues for this population. Overall, S. B. 855 has provided resources that have improved collaboration, supported innovation, and in some cases boosted services available to young people at risk of or experiencing CSE. Several contextual factors beyond scope of the legislation, however, have posed challenges to implementing S. B. 855-most notably, staffing, placement, and service shortages. The ability to provide placement homes that are willing, trained, and supported stands out as one of the greatest capacity constraints that arose in our research. Further investments could include workforce support to improve retention and more services designed to specifically and effectively engage young people in this special population, including more family-based placements where caregivers (including relatives) are well trained and prepared to welcome young people. Funding specialized service slots in areas such as mental health and substance use treatment through the collaborating agencies is another way to respond to the increasing policy pressure to reduce the child welfare system's involvement with children and families while upholding S. B. 855's goal to increase collaboration across child-serving systems. x EXECUTIVE SUMMARY Background The commercial sexual exploitation (CSE) of children and young people is a human rights concern and a public health challenge. CSE refers broadly to any activity or crime that involves the sexual abuse and exploitation of a child for monetary or nonmonetary benefit ranging from providing a phone or jewelry in exchange for a suggestive picture to coercive sex in exchange for money or a place to live.2 Research suggests that the negative consequences of CSE during childhood can include exposure to violence and other traumatic events (e.g., rape, physical assault, gang violence); mental health disorders (e.g., PTSD); reproductive health complications (e.g., sexually transmitted infections, fertility concerns); internalized coping behaviors (e.g., substance use, self-injury, suicide); and nontraumatic health complications resulting from a lack of preventative care (Le et al. 2018). Research documenting the prevalence or cumulative incidence of CSE among young people in the United States is limited because of a combination of practical, ethical, and methodological challenges (Franchino-Olsen et al. 2022). Annually, about 1,500 incidents of CSE were reported to law enforcement authorities as juvenile crimes in the US during the early 2000s (Mitchell, Finkelhor, and Wolak 2010). Since then, the federal government and many states have decriminalized young people's involvement in the commercial sex industry, meaning they cannot be arrested or prosecuted for illegal activity committed as part of their exploitation. (California officially decriminalized juvenile prostitution in 2016.) Because of these changes, CSE was formally established as a type of child maltreatment, while child welfare agencies were given responsibility for investigating and intervening on behalf of young people experiencing CSE. Nevertheless, CSE differs from other forms of child maltreatment in two ways that have the potential to impede effective child welfare responses. First, fewer CSE reports identify parents as perpetrators compared with other forms of maltreatment (Havlicek et al. 2016; Sprang and Cole 2018). Second, financial motives play a unique and central role in CSE, both for third-party exploiters and young people (Cook, Le, and García 2022; Finigan-Carr et al. 2019; Wijkman and Kleemans 2019). Prevalence data on child welfare–identified CSE remain sparse in part because statewide data systems have only recently been modified to capture CSE information in response to legislation. Recent research suggests that allegations of child sex trafficking represent less than 1 percent of all allegations made to child protection systems (Gibbs et al. 2018; Latzman et al. 2019). Nevertheless, a majority of children who experience CSE have a history of child maltreatment and many are involved in the child welfare system (CWS) (De Vries and Goggin 2020). BACKGROUND 1 The federal Preventing Sex Trafficking and Strengthening Families Act of 20143 amended titles IV- B, IV-E, and section 1114A of the Social Security Act4 to identify, document, and determine appropriate services for any child or youth under the care or supervision of the CWS who is at risk of being or confirmed to be sexually exploited, including young people who receive in-home services, are absent from placement, or are nonminor dependents. The Justice for Victims of Trafficking Act of 20155 modified the definition of child abuse and neglect such that any child who is identified by a state as a victim of sex trafficking or other forms of trafficking must be considered a victim of "child abuse and neglect" and "sexual abuse." Annual counts of children in foster care who are identified as victims of sex trafficking, either before or while they were in foster care, were recently added to the National Adoption and Foster Care Analysis and Reporting System (AFCARS).6 These federal legislative changes implemented over the past decade have prompted states, including California, to develop policies aimed at CSE of children and young people, with a particular focus on those under the jurisdiction of the CWS. One California initiative was the Preventing and Addressing Child Trafficking (PACT) pilot project. The pilot began in 2014 and involved a collaborative of 10 county child welfare agencies, which shared a goal of serving and improving outcomes for children at risk of or experiencing human trafficking. In June 2014, the California legislature enacted Senate Bill (S. B.) 855,7 which amended the state's Welfare and Institutions Codes in three ways to better address CSE of children:  defined a commercially sexually exploited child (in line with federal guidance) as any child who is sexually trafficked or who receives food or shelter in exchange for, or who is paid to perform, sexual acts and whose parent or guardian failed or was unable to protect the child from CSE  clarified the CWS's ability to intervene on behalf of children experiencing CSE when a parent is unable to protect their child  created the Opt-In CSEC Program, which provides participating county child welfare agencies with guidance and funding to prevent and intervene on behalf of children who are experiencing or at risk of experiencing CSE8 California's CWS is administered by its 58 counties, with state oversight. Since the inception of S. B. 855, more than 75 percent of counties in California that account for more than 97 percent of the state's child population have sought funding through the Opt-In CSEC Program. Counties can use CSEC program funding to develop and implement interagency protocols, procedures, and specialized services to identify, document, address, and prevent the CSE of children. Nearly a decade later, with 47 of California's 58 counties having opted in to the program, California is well positioned to evaluate this 2 BACKGROUND policy, particularly its implementation across the state, but also the extent to which the legislation may be influencing desired outcomes for young people. A full list of the 47 counties that opted in to the program can be found in appendix A, provided in a separate technical appendix.1 The California Department of Social Services (CDSS) contracted with the California Child Welfare Indicators Project (CCWIP) at the University of California, Berkeley, and the Urban Institute, to conduct an evaluation, which included an implementation study led by Urban and an outcome study led by CCWIP and CDSS's Research, Automation, and Data Division.9 CSEC Program Description Counties that elect to participate in California's CSEC program receive an annual allocation for the costs of implementing prevention and intervention services, as well as training related to serving children who are experiencing or at risk of CSE. To be eligible for such funding, counties must implement the following CSEC program components required by S. B. 855: 1. an interagency protocol developed by, at a minimum, representatives from child welfare, probation, mental health, public health, and juvenile courts-the statute was amended to mandate the Sheriff's Department and County Office of Education to be included in Interagency Protocols in 2017 via A. B. 122710 2. use of a multidisciplinary team (MDT) approach to case planning with participation from, at a minimum, representatives from child welfare, probation, mental health, public health, and substance abuse Shortly after S. B. 855 was passed, S. B. 79411 required all county child welfare and probation departments to implement policies and procedures related to CSE. S. B. 794 mandated identification of children who are experiencing or at risk of CSE, documentation of these children in the child welfare state administrative data system, Child Welfare Services Case Management System (CWS/CMS), determination of appropriate services, and receipt of relevant training. The core components of S. B. *See technical appendix here: Ivy Hammond, Wendy Wiegmann, Joseph Magruder, Daniel Webster, Bridgette Lery, Sarah Benatar, Jaclyn Chambers, Laura Packard Tucker, Katrina Brewsaugh, Annelise Loveless, and Jonah Norwitt, Evaluating California's Efforts to Address the Commercial Sexual Exploitation of Children: Appendices Volume I (Washington, DC: Urban Institute, 2023), https://www.urban.org/research/publication/evaluating-californias- efforts-address-commercial-sexual-exploitation-children. BACKGROUND 3 794 made a child welfare response to CSEC a requirement for all counties, regardless of whether they opt in to the CSEC program under S. B. 855. To support implementation of S. B. 855 and S. B. 794, counties received several key points of guidance from the California Child Welfare Council's CSEC Action Team12 on how to implement a Memorandum of Understanding (MOU) between agencies in service of developing an interagency protocol. First, the CSEC Action Team recommended that the interagency protocol should define the mutually agreed upon responsibilities of each agency. Further, it recommended that policies and procedures established in the interagency protocol should codify cross-system collaboration at the system and individual-case levels, incorporate mechanisms to identify and assess CSEC at key decision points, and address the unique physical and emotional safety considerations of children experiencing CSE. Finally, the CSEC Action Team also recommended that counties form a CSEC steering committee to provide ongoing oversight and support to county agencies and partners to effectively collaborate and fulfill their requirements to identify and serve this population. As part of S. B. 855, counties formed multidisciplinary teams (MDTs) composed of staff qualified to provide a broad range of services to respond to the needs of young people experiencing CSE in a coordinated manner and build on children's strengths. The goal of a multidisciplinary approach to case management, service planning, and provision is to sufficiently address a child's needs from initial identification through ongoing stabilization. The approach may include but is not limited to holding cross-system case-planning meetings. MDTs are expected to facilitate cross-system collaboration by sharing information between agencies that could be used for prevention, identification, and treatment purposes. Further, MDTs formalize the practice of social workers leveraging existing connections to the community during their case management processes. Research shows that young people who experience CSE have service requirements markedly different from those of the children usually served by the CWS. They include, but are not limited to, behavioral, social, legal, educational, housing, and employment service needs (Hickle and Roe-Sepowitz 2018; Le et al. 2018; Rafferty 2018). Collaboration between child welfare and other systems is necessary to adequately meet young people's myriad service needs. Little is known about young people's cross- system service receipt following experiences of CSE, and available research focuses primarily on child welfare and probation system involvement (Bath et al. 2020). Nine years after S. B. 855 was passed, counties have a breadth of experience that can inform the state's understanding of how opted-in counties implemented CSEC programming, what challenges have emerged, and where counties identified promising practices for identifying and supporting children at risk of or experiencing CSE. To that end, this 4 BACKGROUND evaluation was guided by a theory of change developed by Urban to anticipate how components of S. B. 855 would result in change for victimized young people and those at risk of CSE (figure 1). FIGURE 1 S. B. 855 Theory of Change Problem: Young people who are commercially sexually exploited are not well accounted for and do not receive the services they need. Hypothesis: It's because child welfare and adjacent systems do not systematically identify, assess, and coordinate services for young people who are commercially sexually exploited. Intervention: implement S. B. 855, which defines CSEC, clarifies the CWS's ability to intervene, and provides funding and guidance for how to do so collaboratively across systems. Outcomes: (1) more young people being identified as having CSE victimization experiences or at risk of CSE; (2) more young people being assessed for their needs across child welfare and adjacent systems; (3) young people receiving appropriate services from those systems; (4) and safety, permanency, and well-being for young people. Source: Adapted from "EDGE: Evidence-Driven Growth and Excellence," Chapin Hall at the University of Chicago, The Center for State Child Welfare Data, accessed June 22, 2023, https://fcda.chapinhall.org/wp-content/uploads/2018/08/EDGE.pdf. Evaluation Overview The evaluation approach for California's CSEC program consists of two main components: an implementation study and an outcome study. Table 1 provides each component's research questions. BACKGROUND 5 TABLE 1 S. B. 855 Evaluation Research Questions Evaluation component Research questions Implementation study 1. What are the components of S. B. 855 in terms of services, staffing, and organizational structure? 2. Are counties implementing S. B. 855 as expected? 3. Do counties have the capacity to meet the needs of CSEC? 4. What were the barriers and facilitators to implementing S. B. 855? 5. How did the implementation of S. B. 855 vary across counties and why? 6. To what extent did agencies within counties collaborate to implement S. B. 855? 7. What are the key successes of S. B. 855? How do those vary across counties? 8. What are promising practices for implementing local CSEC response programs? 9. Did the COVID-19 pandemic disrupt or facilitate CSEC services or protocols? And, if so, how? Outcome study 1. Among young people who met the state's definition of CSEC or were identified as at risk of CSE, what were their experiences in the CWS as they relate to the following? a. CSE risk and confirmed victimization b. revictimization c. maltreatment in foster care d. case service receipt e. placement setting f. placement stability 2. Among young people who met the state's definition of CSEC or were identified as at risk of CSE, are there common or otherwise meaningful trajectories of CWS involvement? 3. In what ways did associations between S. B. 855 implementation and youth service delivery differ across regions and/or counties? Source: Authors' evaluation plan. Implementation Study We designed the implementation evaluation of California's CSEC program to describe the extent to which S. B. 855's core components are being carried out across California, examine the variation in how the policy is being applied across participating counties, and identify promising practices and major challenges. To this end, we gathered data from annual county program plans, a CSEC program administrator survey, interviews with agency and provider staff, and focus groups with adults who experienced CSE in childhood. CONTINUOUS QUALITY IMPROVEMENT APPROACH The implementation study used a continuous quality improvement (CQI) approach to frame research questions and analyses. Introducing a new policy and its associated procedures requires investments in three major areas: the processes to support the program, the quality with which the program services are 6 BACKGROUND being delivered, and the capacity of the providers to deliver process and quality with fidelity. These areas give specificity to the "Do" phase in what is commonly referred to as the "Plan-Do-Study-Act" (PDSA) cycle of CQI (Wulczyn et al. 2014). At a high level, the processes related to service delivery are triggered when children who are at risk of or experiencing CSE are identified and their needs are assessed through interagency response protocols and MDTs. This requires contact with the identified child to establish their need for help, referrals to high-quality services (e.g., providers are victim centered, trauma informed, and trained in human trafficking), and ready access to those services (i.e., capacity). We collected data in all three areas (process, quality, and capacity) to document whether processes were being followed, services were being delivered as intended, and whether the service providing entities were adequately staffed to meet participants' needs. CROSS-SYSTEM COLLABORATION FRAMEWORK Efforts to identify, assess, and coordinate services for young people experiencing CSE require child- serving systems to collaborate toward unified goals. To understand that coordination and partnership, we looked to two frameworks that concretely define what collaboration looks like in complex social service systems. Collective impact is a deliberative and disciplined approach to collaboration (Kania and Kramer 2011).13 It recognizes that major social problems are too large and multifaceted to be solved by a single organization. Rather, complex social problems, such as CSEC, require long-term coordinated efforts across multiple organizations working toward a common goal. Collective impact has five components, outlined in table 2. Coordinated community response refers to collaboration among the critical systems, organizations, and people in a community (Burt and Spellman 2007). Like collective impact, coordinated community response recognizes that solving complex social problems requires active, long-term planning and involvement of public systems. The approach has five components (table 2), which parallel those of collective impact's components, with an additional focus on access to services. We used this combined set of components to document the extent of collaboration among organizations within implementing counties. BACKGROUND 7 TABLE 2 Cross-System Collaboration (CSC) Frameworks Framework component Definition Collective impact 1. Common agenda All participants have a shared vision for change, including a common understanding of the problem and a joint approach to solving it through agreed-upon actions. 2. Shared measurement Collecting data and measuring results consistently across all participants ensures efforts remain aligned and participants hold each other accountable. 3. Mutually reinforcing Participant activities must be differentiated while still being coordinated activities through a mutually reinforcing plan of action. 4. Continuous Consistent and open communication is needed across the many players to communication build trust, assure mutual objectives, and create common motivation. 5. Backbone support Creating and managing collective impact requires a separate organization(s) with staff and a specific set of skills to serve as the backbone for the entire initiative and coordinate participating organizations and agencies. Coordinated community response 1. Involvement of all All of the systems in a community essential to [preventing and ending CSEC] essential systems must be involved. 2. Mechanism for assigning Integrating services through one or more of several mechanisms. The result of clients to services this streamlined service delivery at the client level should be improved client outcomes as well as more efficient and effective use of resources. 3. Functioning feedback In many communities, this is a monthly (or more frequent) meeting of those mechanism most actively involved in developing appropriate interventions or smoothing bureaucratic pathways. 4. Mechanism for ongoing An ongoing mechanism for thinking about what comes next; asking what and future planning needs to be done, how best to accomplish it, and, finally, what needs to change for the goals to be accomplished. 5. Paid project manager or Someone is being paid to serve as coordinator to organize and staff the coordinator interagency working groups and committees necessary to accomplish community-wide goals. Sources: Burt and Spellman (2007); Fay Hanleybrown, John Kania, and Mark Kramer, "Channeling Change: Making Collective Impact Work," Stanford Social Innovation Review, January 26, 2012, https://ssir.org/articles/entry/channeling_change_making_collective_impact_work. Outcome Study The outcome study examines CWS involvement and outcomes for young people who met the state's definition of CSEC before and after S. B. 855 implementation. This analysis used information recorded in the statewide administrative database to describe CWS events that were experienced by young people who met California's definition of CSEC or were identified as being at heightened risk of experiencing CSE. Among the measures examined were the identification of CSEC, along with documentation practices, revictimization, case involvement, and foster care placement trajectories. 8 BACKGROUND Methods Data Collection and Sources Implementation Study We relied on three data collection approaches to understand what counties have implemented in response to S. B. 855, the challenges associated with those efforts, and promising practices for serving young people who are at risk of or experiencing CSE. Initially, we conducted an environmental scan of program documents (e.g., annual county plans); we then designed and fielded a survey of county CSEC program administrators to systematically gather information that was not available in county plans. Finally, we conducted county case studies to gather nuanced information of how these programs have been functioning from a diverse set of perspectives. The environmental scan involved an in-depth analysis of more than 200 annual county program plans and other relevant administrative documents. Informed by findings from the environmental scan, we surveyed CSEC program administrators to fill knowledge gaps. Based on both the environmental scan and survey results, we selected a subset of 12 counties for in-depth case study through (1) stakeholder interviews and (2) focus groups with adults who experienced CSE during childhood in California between 2015 and 2022 (hereafter referred to as focus groups). Table 3 below connects the implementation study data collection methods to the study's research questions. TABLE 3 Research Questions and Data Collection Methods Method Environmental Key informant Research question scan Staff survey interviews Focus groups Components of S. B. 855 Services X X X X Staff X X X Organizational structure X X X Barriers and facilitators X X X Variation across counties X X X X Interagency collaboration X X X Key successes X X Best practices X X Impact of COVID-19 X X X Source: Authors' evaluation plan. Notes: Focus group participants were ages 18 and older and had experienced CSE as a minor. We did not include young people currently in the CSEC program. METHODS 9 ENVIRONMENTAL SCAN The environmental scan captured information about the intended and realized enactment of the core components of S. B. 855, identified key implementation milestones, and noted barriers and facilitators to success. We explored intracounty collaboration and cross-county collaboration, where applicable. We analyzed a total of 203 annual county plans and other documents that counties submitted to CDSS because each county opted in to S. B. 855 between 2015 and 2020. We coded these documents for evidence of prior (i.e., before S. B. 855), planned, and active screening; data collection; service; training; collaboration; and staffing activities done in service of supporting young people who experienced CSE victimization. Specific staff and partner organizations were identified, with an emphasis placed on highlighting areas where counties have incorporated lived expertise from people who experienced CSE in childhood and/or advocate voice in their implementation efforts. We also looked for evidence of collaboration as defined by the cross-system collaboration (CSC) framework, such as the breadth of partner organizations identified and descriptions of CSE service referral processes. We looked for both barriers and facilitators to implementation that we later explored more deeply in the staff survey and county case studies. STAFF SURVEY Informed by the environmental scan, we developed a survey to gather information that was missing, unclear, or incomplete in the county plans. We developed an online survey targeted to the primary CSEC contact in each county's child welfare agency. The survey inquired about process, quality, and capacity changes and cross-system collaboration components that were operating in the participating counties. Questions covered the CSEC program core components (interagency protocol, MDT processes, CSE screening, and services), CSE data collection and use, and case management structure for CSE cases and CSE-related training. Guided by the CSC framework, the survey included items that addressed the CSC components. For example, items covered the presence and functioning of a steering committee, involvement of stakeholders beyond those required by S. B. 855 (e.g., education, foster care placement providers), and perceptions of communication among stakeholders. The full survey instrument can be found in appendix B. Before full deployment of the survey, we pilot tested it with six administrators from three counties and three state-level administrators from the Child Trafficking Response Team. Five pilot respondents completed the survey or otherwise provided feedback. We administered the full survey from May to June 2022 and received responses from 46 of the 47 (98 percent) participating counties. 10 METHODS COUNTY CASE STUDIES  We carried out county case studies to contextualize the findings from the environmental scan and staff survey. The environmental scan and staff survey relied on information provided by a small number of child welfare staff in each county. While we looked for evidence of the CSC components as part of the environmental scan and staff survey, these data sources provided a limited view of collaboration. The county case studies allowed us to better understand S. B. 855's implementation and the depth of cross-system collaboration by hearing from various stakeholders, particularly people who experienced CSE. Between November 2022 and February 2023, we visited a total of 12 counties (nine in-person and three virtually). We visited the following counties: » Alameda » Riverside » Shasta » Los Angeles » Sacramento » Trinity » Madera » San Diego » Tulare » Monterey » San Francisco » Ventura Site selection. The combination of information gleaned from the environmental scan and staff survey informed how we selected 12 counties for in-depth case study. Our aim was to select a mix of counties with variation in contexts that could influence program implementation and outcomes for young people who were at risk of or had experienced CSE. We used the following factors to select case study counties:  Geography. We selected a minimum of two counties from each of four geographic regions as defined by the PACT pilot project: Northern California, Central Valley, Bay Area, and Southern California. Nonurban areas typically have fewer services, both in terms of type and capacity, than urban areas-something we observed in our review of county plans. We purposively selected two counties classified as nonmetro by the Economic Research Service of the US Department of Agriculture to better explore this phenomenon.  CSE population size. While CSE can occur in any community, some areas have higher rates of documented occurrence than others. We used child welfare administrative data from California's Child Welfare Services Case Management System (CWS/CMS) to create a preliminary count of CSE cases in each county. For the purpose of site visit selection, a CSE case was defined as a child age 11 or older with an entry on the CSEC Data Grid in CWS/CMS between July 1, 2010, and June 30, 2021.14 The Data Grid captures young people with both confirmed CSE and those at risk of CSE. Within each region, at least one selected county had CSE case counts that were in the top third of all counties. We also calculated counties' rates of METHODS 11 CSE referrals as a share of all referrals for children ages 11 or older and ensured that the selected counties reflected a mix of those with high, medium, and low CSE referral rates.  Implementation score. Using survey responses, we calculated county-level implementation scores for process, quality, capacity, and collaboration. This analysis is detailed in the next section. Implementation scores were classified as low, medium, or high. Within each region, we selected at least one county with high implementation scores and at least one with low implementation scores. We also used implementation scores in conjunction with urbanicity to select counties that reflected a mix of "typical" and "outlier" expected performance. Large counties with low implementation scores and small counties with high implementation scores were both considered "outliers" while the reverse of both instances was considered "typical."  CSEC funding. Participating counties receive different amounts of CSEC program funding from CDSS, with more populous counties generally receiving more funding. We created a per capita measure to standardize CSEC funding levels using their amount of program funding in FY 2020–21 and the total number of young people ages 11 to 17 in care on January 1, 2021. We then classified counties into terciles of per capita funding levels: low, medium, or high. Because funding can influence implementation, we selected some counties because they reflected one of four combinations of funding levels and implementation scores: high on both, low on both, low funding but high implementation, or high funding but low implementation.  CSEC pilot county. Counties that volunteered to be part of the CSEC program pilot may be qualitatively different than nonpilot counties. In addition to simply having a longer implementation period, pilot counties may have more leadership or community support or may have a sizeable number of young people experiencing or at risk of CSE. Ten counties participating in the CSEC program in 2022 were also part of the pilot. Half of the case studies were in pilot counties.  Other considerations. CDSS is interested in understanding how counties engage tribes in CSEC program implementation. We considered stated tribal engagement documented in the county plans and/or survey responses when making county selections. We also considered the presence of a steering committee. While steering committees are not a required component of S. B. 855, they were a mechanism suggested by the state CSEC Action Team to promote interagency communication, and they are a marker for cross-system collaboration.  Key informant interviews. In each of the 12 case study counties, we requested interviews with staff from child welfare, juvenile probation, service provider agencies, and other partners (e.g., education, district attorney). We specified the key informant types we thought would have important perspectives to share as well as the child welfare CSEC coordinator's 12 METHODS recommendations about who was best positioned to report on the development and execution of the county's CSEC program. Not all people we contacted agreed to be interviewed. Two researchers (one senior lead interviewer and one research assistant) conducted each interview. In all, we interviewed 62 child welfare staff, 12 juvenile probation staff, 29 provider staff, and 29 staff from other partner agencies (appendix I).  Focus groups. We led focus groups with adults who had experienced CSE as minors during the implementation period in 2 of the 12 counties. Throughout the evaluation period, CSEC Advisory Board members (a group of experts with direct experience funded by the CDSS) emphasized the importance of hearing directly from those who were impacted by the policy and practice changes associated with S. B. 855. The research team worked with service providers and child welfare staff to invite young adults who had experienced CSE since the inception of S. B. 855 in 2014 to discuss their experiences with receiving support from county staff and service providers. Participants were offered a meal and a $75 gift card as a thank you for their participation. We developed a moderator's guide that was reviewed by advisory board members in advance of any groups being conducted to ensure the approach was trauma informed and sensitive to this population's needs. In all, 12 people with direct experiences of CSE participated in focus groups. Outcome Study The outcome study used statewide administrative data to analyze a set of system performance metrics, including both event- and child-level indicators, to achieve three aims. First, we sought to describe child welfare agencies' efforts to identify, intervene, and provide placement for young people who experienced or were at risk of experiencing CSE. Second, we documented changes in system performance over the course of S. B. 855 implementation between state fiscal years 2015 and 2021. And finally, we tested for associations between these system performance indicators and county-level context. STUDY POPULATION We examined CWS involvement of all children (from birth to age 17) who met the legislature's definition of a commercially sexually exploited child as described in S. B. 855 and were identified between July 1, 2015, and June 30, 2021. A child was defined as being sexually exploited or at risk of sexual exploitation if they had a CSE concern documented in the state's Child Welfare Services Case Management System (CWS/CMS) in any of the following ways: 1. had a special project code attached to a case, client or placement notebook that identified CSE victimization, risk, or specialized service receipt in a county that opted in to S. B. 855 METHODS 13 OR 2. had an "at-risk" or "victim" entry on a specialized CSEC screen on a child client's description page (i.e., "CSEC Data Grid") by a county that opted in to S. B. 855; OR 3. had a Child Protective Services (CPS) maltreatment report alleging CSE that was not concluded as "entered in error" in a county that opted in to S. B. 855. The CWS/CMS contains three forms of CSE documentation because the first form, special project codes, could be implemented almost immediately, but the implementation of the other two forms took several years while CWS/CMS architecture was being revised. A detailed description of S. B. 855 defined CSE victimization and risk along with screenshots of the CSEC Grid and Allegation page can be found in appendix K. DATA SOURCE The outcome study relied on administrative records from the CWS/CMS. Through a longstanding partnership between CDSS and the University of California, CCWIP receives quarterly extracts of administrative data from CWS/CMS. All personally identifiable information is removed from records on a nonnetworked computer, and the deidentified dataset is stored on a secured remote server. The ongoing storage and analysis of CWS/CMS data by CCWIP receives annual review by the state's institutional review board. MEASURES Broadly, event- and child-level measures of system performance fall into five system performance categories, which include identification of CSEC; documentation practices; front-end responses to CSEC; and back-end responses to CSEC including case involvement and foster care placement. Below we summarize each measure within these categories and its corresponding methodology. Additional information can be found in appendix L. Documentation practices. Frequency counts of CPS reports represent the standard practice for measuring child maltreatment and recurrence, regardless of the type of abuse or neglect. Although exploitation already existed as an allegation type, the CSEC Data Grid was created in response to S. B. 855 as a way to document and distinguish CSE risk and victimization within the case records of child clients receiving CWS interventions. Use patterns for both fields are described separately and in combination. Detailed findings on CSE documentation practices are provided separately in appendix J. 14 METHODS CSEC identification. CWS identification of CSE victimization was measured at both the event level (via a maltreatment report) and child level using the CWS/CMS indicators noted in the "Study Population" section above. Documentation practices stratified by child demographics and county characteristics are in appendix L. We used two event-level indicators to summarize CSE maltreatment reporting and responses. First, counts of all CPS reports documenting concerns of CSE (hereafter referred to as CSE reports) were used to describe the share of CSE reports that were screened in by the CWS for investigation and were substantiated by the CWS. Second, CSE reports were stratified by reporter type, which describes the professional role of the person who reported suspected CSE and/or their personal relationship to the reported child. One goal of S. B. 855's approach was to improve the identification of CSE through MDTs. To document the extent to which identification of CSE concerns involved cross-system collaboration, reporter types were grouped to distinguish each type of MDT-eligible agency identified in the legislation. At the child level, 38,168 children had CSE concerns documented one (or more) of three ways in CWS/CMS between July 1, 2015, and December 31, 2021. Nearly all (80.1 percent) had at least one child maltreatment report alleging CSE that was screened in for investigation, 25.7 percent had one or more entries on the CSEC Data Grid (indicating CSE risk or victimization), and 4.3 percent had a special project code but no CSE report or Data Grid entry documented within the observation period. Front-end responses to CSEC. Front-end response measures describe the percentages of children with initial CSE reports that were (1) investigated and (2) substantiated. They also describe the percentages of children with initial CSE reports that had (1) subsequent CSE report investigations and (2) subsequent CSE report substantiated within 12 months of the initial report. Case involvement. Mutually exclusive profiles of case involvement following CSE concerns documented the percentages of children where (1) there was already an open CWS case; (2) a case was opened within one month of CSE concern; (3) a case was opened within 2 to 12 months of CSE concern; or (4) no case was opened within the 12 months following the child's initial CSE concern. Case status at 12 months post-CSE was also measured. To gauge the timeliness of formal child welfare intervention on behalf of young people experiencing CSE, we measured the median time between an initial CSE concern and case opening and stratified findings by level of CSE certainty and implementation characters. Placement trajectories. We examined placement trajectories for S. B. 855–eligible children through a series of indicators. First, placement involvement was summarized according to whether the child (1) had no placement episodes (before, during, or after CSE concern); (2) had only closed placement episodes at the time of CSE concern; (3) was in an open placement episode and in placement at the time METHODS 15 CSE concern was identified; (4) was in an open placement episode but not in a placement home or facility at the time CSE concern was identified; and (5) had a placement episode opened within 12 months of CSE concern. Those with multiple placement episodes during and after CSE concerns were described. In addition, we looked at predominant placement types for S. B. 855–eligible young people who experienced foster care placement and the number of unique placements across given placement types. The median time between an initial CSE concern and case opening was described and stratified by level of CSE certainty and implementation characters. Finally, the median percentage of placement days that young people were absent from care was calculated by dividing the total number of days each child was absent from placement by the total number of days they were in placement episodes. Analysis Qualitative Analysis ENVIRONMENTAL SCAN For the environmental scan, we coded and analyzed the county CSEC plans using Dedoose, a qualitative analysis software package. The study's research questions guided the construction of the coding structure, including themes related to services, screening, data collection, training, staffing, and collaboration (including codes reflecting the components of the cross-system collaboration framework). Individual team members coded the county plans for each fiscal year after being trained on the coding scheme. We regularly checked team members' coding to ensure reliability and resolve any disputes about interpretation. Themes were examined across counties, including looking at how implementation varied among county subgroups (e.g., rural versus urban counties). One caveat when interpreting observations from the county plans was that if a county plan did not mention a particular program component, facilitator, or barrier, this was not necessarily evidence that it did not exist or happen in that county. All we can say for certain is that it was not mentioned in that county's plan. Because of this limitation, we chose to rarely cite exact counts of counties when summarizing findings from the county plans. COUNTY CASE STUDIES Following each case study, researchers cleaned all notes and did a close read for emerging topics per a thematic analysis framework for qualitative data analysis. Identified themes were shared across the team for consistency and any disagreement was resolved. Researchers wrote site-specific summaries and shared them with the rest of the team to review (see appendix H). Following the protocol structure, the summaries describe the county's CSEC program context, organizational structure, and staffing. We 16 METHODS capture screening procedures used by each county and their training practices (including perceived adequacy of training). We describe collaboration between involved agencies-who is involved, what is working well, and what has been challenging. And finally, we highlight both ongoing barriers to counties' implementation of their CSEC program and promising practices that could be shared with other counties. At the conclusion of our analyses, summaries were shared with county staff to ensure the accuracy of our interpretations. Quantitative Analysis STAFF SURVEY We placed survey items into scales aligned with the CSC framework described above. The Collaboration scale contains 18 items focused on the involvement of collaborating agencies in the county's CSEC program. Collaborators included agencies required by S. B. 855 as well as partnerships with other organizations to address CSE that are not legislatively required (e.g., tribal agencies, placement providers). Collaboration could include participating in MDTs, providing CSEC training or services, or participating in the steering committee. The Collaboration scale also assesses the presence and functioning of a CSE steering committee, involvement of agencies in MDTs or steering committees, and perceived quality of collaboration. Four additional scales were also created to align with the CQI framework: Minimum Process, Bonus Process, Quality, and Capacity.  The Minimum Process scale focuses on the core components explicitly required by S. B. 855 and S. B. 794. We include S. B. 794 components as all counties must address these requirements to meet current expectations for CSEC programs. For instance, a screening and assessment process is needed to identify young people at risk of or experiencing CSE (per S. B. 794) so their needs can be addressed through an MDT (per S. B. 855).  The Bonus Process scale asked about additional promising practices recommended by the CSEC Action Team, such as including advocates with lived expertise related to CSE or specialized training in working with clients post-CSE (hereafter referred to as advocates) as part of their initial or ongoing MDT approach and providing CSE training to placement providers. The Bonus Process scale also includes processes that are likely to enhance program implementation, such as collecting data to track MDTs and service delivery or having CSE-specific services. METHODS 17  The Quality scale contains items to gauge how well program activities and processes are being completed, as survey respondents perceive them. This includes consistent input from required parties of MDTs, having multiple events that could trigger a CSE screening, and perceptions about whether screening effectively identifies young people at risk of or experiencing CSE.  The Capacity scale contains items related to staffing, waitlists for services, and supply of family-based placements for CSEC program young people. To create scale scores, we assigned points to each survey item and summed. The full scoring rubric can be found in appendix E. For most items, an affirmative response (e.g., agree, yes) was one point. Items that asked participants to rate on a scale from 0 to 100 were given a point if the response was greater than 50 percent (or less than 50 percent for reverse-worded items). Some items allowed respondents to check all that applied from a list (e.g., types of CSE-specific services available), and respondents were given half points for each affirmative. For scoring purposes, "don't know" and missing responses were counted as zero. For all scales, a ranking of low, medium, or high reflects whether a county fell in the first, second, or third tercile, respectively, for scores on that scale. The scale ranking reflects how a county scored on a scale relative to others rather than an absolute score. Appendix F displays the absolute scores for each tercile. These rankings were used to aid in selecting counties for case studies, and we also examined variation in administrative outcome data based on these rankings. In addition to creating these scales, we ran basic descriptive tables for key items from the survey to better understand indicators of collaboration, staffing and training, screening, service provision, and data collection and usage. We then examined variation in key survey items and implementation rankings based on several subgroups, including urbanicity, population size, region, CSE count, CSE rate, and whether the county was an S. B. 855 pilot county or not, as described in the "Methods" section. CWS/CMS DATA ANALYSIS We conducted descriptive analyses of child-level data from the CWS/CMS. Several existing measures were used to answer the research questions on CWS child outcomes. Additional measures were developed using allegation, case, and placement data to further describe child outcomes following a documented instance of confirmed or suspected CSE victimization (i.e., a CSE concern). We documented the count of unique placements a child entered during the year following CSE. We also described periods of absence from placement among children in open foster care episodes and examined the status of their system involvement 12 months after a CSE concern was first identified.15 18 METHODS Findings Implementation Study Together, information from the county plans, survey, and case studies formed a picture of how counties were responding to CSE. Below we report on findings in the major areas of implementation that reflect S. B. 855's directives: collaboration, staffing and training, screening, services, and data. Collaboration S. B. 855 aims to bring together key county agencies to coordinate a unified response for people experiencing or at risk of CSE. A unified response that meets young people's needs requires collaboration. Implementation encompasses what collaborative activities counties engage in, how they go about those activities, and how well those efforts are going. The primary collaborative activities under S. B. 855 included developing interagency protocols, holding case conferencing meetings with MDTs or using other approaches to collaborate across disciplines, and convening CSEC steering committees. We calculated an overall collaboration score from 18 questions in the survey to create a summary indicator of collaboration. In general, larger counties scored higher on collaboration. Large counties also tend to be urban and use the CSEC Data Grid more frequently, suggesting that places with greater resources can dedicate some of those resources to collaborative efforts, and places with a larger share of young people who are identified as experiencing or at risk of exploitation may be the most motivated to collaborate. Counties generally reported that agencies they defined as part of the CSEC collaboration work well together. About 84 percent of survey respondents agreed that collaborating agencies were aligned in their CSE efforts, and 89 percent of respondents said that agencies were able to work together to support young people experiencing or at risk of CSE, suggesting that most counties have demonstrated the CSC component of mutually reinforcing activities (table 2). This sentiment was echoed in case study interviews. Key informants heralded the progress toward working closely across agencies that had previously been siloed, and interviews suggested that many counties were successfully incorporating the CSC component of continuous communication (table 2). One CSEC coordinator from an urban county said, Coming together on a quarterly basis for nine years, knowing who the partners are, it makes a difference when you feel like you know that person, when a problem comes up just talk with them, and we're just trying to serve the youth. FINDINGS 19 When asked specific questions about the quality of the collaboration, survey respondents were somewhat more equivocal. Despite almost all respondents reporting positively on county collaboration overall, some reported areas where collaboration could improve. Approximately one-third of counties indicated that collaborating agencies were not on the same page, suggesting that not all counties have met the CSC component of a common agenda (table 2). The same share expressed some dissatisfaction with communication between child welfare agency staff and other agencies, again suggesting that a substantial share of counties were experiencing challenges with the CSC component of continuous communication (figure 2; table 2). However, three-quarters of counties agreed that child welfare leadership prioritized responding to CSE, providing support for the CSC component of backbone support (table 2). FIGURE 2 County Collaboration Quality Counties' survey responses to collaboration quality questions Agree Neither agree nor disagree Disagree Missing Agencies in my county are able to work together to support 89% 11% young people at risk of or experiencing CSE. My agency and other collaborating agencies are aligned in our efforts to support and serve children at risk of or 80% 13% experiencing CSE. Child welfare agency leadership prioritizes responding to 76% 20% CSE in my county. I am satisfied with how my agency communicates with 65% 33% collaborating agencies about CSE. I feel as though my agency and other collaborating agencies 61% 31% are on the "same page" in our county's response to CSE. Share of responding counties Source: Survey of CSEC program administrators in California counties that had opted in to the CSEC program. Note: N = 46 counties that responded to the survey. Case study interviews elaborated on these relationships. In general, we heard that agencies were interested in working together, but it was difficult when priorities did not align. Staff often spoke about challenges working with juvenile probation. With efforts to decriminalize sex crimes, probation's orientation shifted away from detention (even if other crimes may be suspected); this frustrated some 20 FINDINGS child welfare staff who are eager to hold onto children long enough to connect them to services. One service provider said, Everyone is willing to collaborate, but when a case comes up, it's like agencies are protective/territorial. INTERAGENCY PROTOCOL S. B. 855 requires opted-in counties to have an interagency protocol to facilitate collaboration and coordination in serving young people experiencing CSE. The protocol typically defines the organizational structure of each county's CSEC response. Shortly after opting in to S. B. 855, county agencies came together and designed their own protocol under the legislation's requirements. Once the protocol was established, however, few counties specifically mentioned in their county plans changes to the protocol or how it shaped their organizational structure in the following years. The original S. B. 855 legislation required the interagency protocol to include (at a minimum) child welfare, probation, mental health, public health, and juvenile courts. The statute was amended to mandate the sheriff department and county office of education to be included in interagency protocols in 2017 through A. B. 1227; however, the survey did not specifically ask whether these two agencies were part of the interagency protocol. Among the 46 counties that responded to the survey, 39 reported that they have a signed interagency protocol (85 percent). Counties that have a signed interagency protocol with participation from all agencies required by the original S. B. 855 legislation (N = 27) and those who do not (N = 19) varied significantly by region and total population size, with counties in Southern California and the Bay Area more likely to report having all required agencies involved in their interagency protocol (100 percent and 70 percent, respectively) compared with counties in Northern California and the Central Valley (44 percent and 36 percent, respectively). The likelihood of having all required agencies involved in the protocol also increased with population size. Small counties were the least likely to have a signed protocol with all required agencies (31 percent), moderately sized counties were more likely (60 percent), and large counties were the most likely (87 percent). Counties with and without signed agreements did not differ on urbanicity, CSE population size, CSE rate, or whether a county was an S. B. 855 pilot county. Appendix D displays all subgroup variation analyses. Many counties used interagency memoranda of understanding (MOUs) to operationalize their interagency protocols. These MOUs could define the steering committee, clarify identification processes, lay out reporting requirements, explain assessment processes, guide proper documentation, demonstrate the county's multidisciplinary response, and describe information-sharing agreements.16 Counties usually FINDINGS 21 began work on their MOUs concurrently with their CSEC response protocol. However, many counties reported problems in establishing their MOUs; when one agency delayed their approval, it slowed the entire collaboration system for CSEC cases. By 2020-four-to-five years into implementation-all but one county had established MOUs, though many were initially delayed in doing so. Appendix C, table C.1 shows agency participation in each county's interagency protocol. A total of 27 counties reported that they had a signed interagency protocol with participation from all required agencies. Juvenile probation was included in the interagency protocol for all counties who responded to the survey question. In county plans and key informant interviews, the majority of counties reported that probation was a key collaborator in CSEC screening and service provision, though child welfare staff generally had ultimate responsibility for this population, and probation's role in serving young people experiencing or at risk of CSE was sometimes limited by factors such as staffing capacity and whether the county was a dual-status county or not.17 County mental health departments were included in interagency protocols for nearly all counties who responded to the survey question, yet case study findings indicated that their role was often more peripheral and limited by both procedural and capacity constraints. For example, a few counties noted that there were not enough therapists to provide behavioral health services for young people. And staff from one county said that although young people may be referred to their behavioral health department for services, staff in the department are not trained on working specifically with young people experiencing or at risk of CSE. Substance abuse services agencies were represented in the interagency protocol for about three-quarters of counties. For almost all counties, substance abuse services were part of the behavioral health agency. However, many of the case study counties reported not having adequate substance use services available for young people, even if these services were represented in their protocol. About two-thirds of counties reported participation from public health. Among the case studies, only two counties identified public health officials as key staff to be interviewed. Juvenile courts were included in 63 percent of interagency protocols according to the survey. County staff told us that juvenile court involvement varied widely. In some counties, juvenile courts were instrumental to the CSE response, and some counties had designated courts for young people experiencing CSE. In other counties, juvenile courts were not identified as a key partner. The variation in juvenile court involvement seemed to be related to the size of the identified population of young people experiencing CSE and the quality of collaboration between agencies. 22 FINDINGS Interagency protocols frequently included partner agencies beyond those required by S. B. 855. Other agencies often included were educational entities (likely prompted by A. B. 1227 in 2017, which required adding education departments to CSEC protocols), district attorneys, and other community- based organizations. We asked counties which agencies were involved in their CSE response (other than their interagency protocol partners). Education partners (80 percent of counties) and the court system (83 percent of counties) were frequent partners. However, case study findings suggest that the survey may have overrepresented education's role. About half of counties said that child welfare placement agencies were part of their CSE response-but that placement homes for young people experiencing CSE were in extremely short supply. Representatives from federally recognized tribes were the least common partner, with only one-third of respondents indicating that tribal organizations were part of CSE collaboration (see appendix C, table C.2). MULTIDISCIPLINARY TEAMING S. B. 855 requires counties to use an MDT approach to case management and service planning. The legislation did not specify the mechanics of how this MDT approach should be carried out, and counties were given latitude to develop their own MDT approach. Most counties implemented MDT meetings as their primary method for facilitating an MDT approach to case planning and service provision, although some counties also incorporated other approaches. A few counties are still determining and refining their MDT processes as they learn more about what an MDT approach is meant to entail. Because counties overwhelmingly utilized meetings as an MDT approach, our findings focus primarily on MDT meetings. However, we also describe other MDT approaches that counties have implemented. MDTs are intended to develop treatment plans that meet children's cross-system needs, and thereby function as the CSC component of mechanism for assigning clients to services (table 2). They may also be called upon to troubleshoot during a case or plan for successful case closure. The legislation requires MDT participation from child welfare, probation, mental health, public health, and substance abuse. MDT participation from other service providers, education agencies, law enforcement, and advocates is not required but may be included. In the survey, we asked counties what type of MDTs they use, how often they engage in MDTs, and which entities are involved. Among the 46 counties that responded to the survey, almost all counties had ever used an MDT approach in their work with children experiencing or at risk of CSE (96 percent).18 The share of young people whose care utilized an MDT approach, as reported on the survey, FINDINGS 23 did not vary significantly based on counties' region, urbanicity, overall population size, CSE population size, CSE rate, or whether a county was an S. B. 855 pilot county. Counties reported using a variety of methods for engaging in MDTs. Among the 44 counties that reported ever using an MDT process in case management or service provision, 84 percent conducted formal CSEC MDT meetings, 70 percent combined CSEC MDTs with an existing meeting process, such as case consultations or child and family team meetings, and 30 percent engaged in MDTs through a nonmeeting process, such as email or phone calls. Most counties used more than one of these methods (84 percent). Key informant interviews confirmed that counties approach MDTs in a variety of ways. Some had integrated MDTs in their child and family teams, and some had incorporated MDTs in their specialized CSE juvenile courts. Others used MDT meetings for more general policy and practice discussions. Further, the frequency with which an agency participated in counties' MDT processes varied widely (see appendix C, table C.3). Juvenile probation and mental health agencies participated in about three-quarters of MDTs while juvenile courts and substance abuse agencies participated in about one-quarter and one-third of MDTs, respectively. advocates often participated in MDTs, even though the legislation does not require it. Other CSEC collaborating agencies also participated in MDTs (see appendix C, table C.4). Education agencies, in particular, were named as regular participants in MDTs for many counties in survey responses, though they were rarely mentioned in case study interviews. Table C.3 displays how often each type of other collaborating partner attended MDTs. Some counties also incorporated an MDT approach throughout the entire process of working with young people, not only within specific MDT meetings. For example, a few counties conducted joint response child welfare investigations where child welfare staff and nonprofit agencies would go out together for initial investigations-exemplifying an MDT approach outside of a standalone meeting. In counties that held MDT meetings, there was a great deal of variation in their process and frequency. Some larger counties held MDT meetings weekly for frequent case review of newly identified young people, while other counties only met monthly or as needed because they had a lower volume of identified young people. All but one county said that they review individual cases at their meetings, and 26 counties said that they invite the involved young people to attend. Some counties held MDT meetings in addition to other case-planning meetings and integrated an MDT approach in existing meetings (e.g., child and family teams, or CFTs). Because so many counties indicated that their MDT approach was meeting driven, we asked counties on the survey how many young people identified as at risk of or experiencing CSE had an initial MDT meeting and how many had an MDT meeting during an 24 FINDINGS ongoing case. On average, counties reported that 60 percent of young people received an initial MDT meeting, but there was wide variation (ranging from 3 percent to 100 percent). Results were similar for young people also receiving ongoing MDTs. STEERING COMMITTEE AND INTERAGENCY COLLABORATION While not an explicit requirement of the S. B. 855 legislation, the CSEC Action Team recommends forming a steering committee as a mechanism to coordinate CSEC efforts across systems. The county plans indicated that steering committees were essential to developing the interagency protocol and continued to be effective for promoting adherence to the protocol. Many counties relied on their CSEC steering committees to monitor their interagency protocol and propose adjustments as necessary. In the survey, we asked counties whether they have a steering committee, how often it meets, and how well it functions. Most counties had a steering committee (37 of 46 counties that responded to the survey). Among the 37 counties with a steering committee, 81 percent met quarterly or more often. In their plans, counties reported using their steering committee to discuss system-level problems, CSEC- specific trends, and provide oversight to the county's CSEC coordination. Counties with a steering committee that met regularly did not differ significantly from counties without a steering committee that met regularly in terms of region, urbanicity, overall population size, CSE population size, CSE rate, or whether a county was an S. B. 855 pilot county. Steering committees usually included most or all county agencies involved in CSEC response-such as child welfare, juvenile probation, behavioral health, public health, sheriff's department, or law enforcement-along with community partners such as service providers, placement agencies, advocates, legal services, and more. However, the specific membership and attendance varied among counties. Agencies represented on steering committees were often very similar to the MDT partners. Appendix C, table C.5 displays steering committee attendance rates by agency. In the county plans, smaller, more rural counties were more likely to note that some county agencies-for example, law enforcement partners-were less likely to engage with the steering committee consistently over time, even if all agencies enthusiastically participated at its conception. Most steering committees seem to have a unified purpose, in line with the CSC framework component of a common agenda (table 2). More than two-thirds of the 37 counties with a steering committee believed that their committee had a clearly defined goal (68 percent). However, many counties reported challenges with member composition or attendance. Only about half of counties with a steering committee agreed that all essential agencies were involved (54 percent), suggesting challenges with the CSC component of involvement of all essential systems (table 2). Additionally, the differences between FINDINGS 25 steering committee meetings and MDTs were blurry in some counties. As the program has evolved, the purpose and utility of the steering committee changed in many counties. During case study interviews, many shared that steering committees were essential for establishing CSEC protocols and building interagency relationships, but now that processes were in place, the need for these meetings has waned. Several key informants suggested that nearly 10 years into implementation, this may be a good time to revisit the mission of the steering committee. These interviews indicated that while many counties had the CSC component of a mechanism for future and ongoing planning (table 2), this future planning process could use reinvigoration in some counties. In both county plans and case studies, counties highlighted that frequent staff turnover across agencies has had a detrimental effect on the ability to collaborate across systems. For example, staffing changes among steering committee members delayed one county's ability to develop and finalize their interagency protocol. Key informants said that collaboration relied on relationship building, and turnover can disrupt collaborative progress. Disruption was mitigated where more than one person facilitated the cross-agency work, or when case collaboration was part of a job description. Related to staffing capacity, cross-agency capacity to attend MDTs was a notable challenge. Twenty- four counties reported that it was hard to schedule MDTs, which delayed services. We learned that while some young people regularly receive an MDT, in other counties MDTs are very infrequent. This varies tremendously across counties. Further, key informants repeatedly remarked that they did not know of a way to systematically serve young people who are experiencing CSE without involving the CWS. Many interviewees felt that this was a key gap in the county's capacity to effectively respond to CSE. However, a few counties developed a CSEC MDT process for non–child welfare involved young people. Similarly, several counties expressed feeling limited in what they can do for young people who are not removed from the home. This highlighted a natural tension between national, state, and local pressures to reduce the use of foster care and the CWS's funding and authority boundaries. When day-to-day case coordination is delayed, overall collaboration stalls and young people do not receive timely services. County staff shared that the inability to share case information-both across counties and across agencies in the same county-was a significant hinderance to collaboration. Under S. B. 855, counties are required to establish MOUs among the agencies in their interagency protocols. MOUs grant agencies permission to share information but they do not facilitate the process, which can be complex. MOUs also do not facilitate data sharing between counties. While S. B. 855 fostered collaboration between agencies and helped counties create shared processes for case planning, the site visit interviews revealed that staff from different agencies sometimes disagree on 26 FINDINGS what approaches that they think best serve young people. This was often raised in the context of social workers who perceived ultimate responsibility for the safety and protection of young people being exploited but felt they did not have adequate tools at their disposal. In turn, they expressed that the efforts of other agencies would undermine their attempts to help young people on their caseload. One social worker said, For more recent cases it's collaboration with probation and [the] child's counsel not quite understanding the abilities and limitations that the [child welfare] department has. If they don't agree how we want to work a case, they trump it and it's been a whole mess. At least four recent cases we've had, they come in with "that's what we want to do." Then the youth sits and watches the whole thing. While the survey and county plans overall suggested that the CSC component of mutually reinforcing activities (table 2) was present in most counties, the site visits suggested this component may be challenging for some counties when there was frequent mention of tension or misunderstanding about each agency's role and responsibilities. For example, probation and child welfare sometimes disagreed about which agency should hold primary responsibility for the youth, a common problem for dual-status counties. Similarly, child welfare staff and voluntary advocacy services often had to navigate their very different mandates and approaches. Child welfare staff expressed that some partners expect child welfare to do certain things that are not within their power or policies to do. One social worker said, For example, they got an order for a youth to be put in a certain home, but the home isn't certified…That happens often where they come up with ideas for the department to carry out, and we're not able to or we're roadblocked. Likewise, other partners indicated that they believe child welfare does not always understand the limits facing probation or law enforcement agencies. These misunderstandings or misalignments often impede collaboration. As one probation staff member said, I think that's the challenge collaborating with CPS. They are the lead agency and sometimes they ask probation [to do] what we really can't do….Sometimes behavior is not criminal, and they think it warrants them going into custody. Whether a county is a dual-status county or not also seemed to impact collaboration. In counties that were not dual status (i.e., a youth cannot be under the jurisdiction of both the child welfare agency and probation simultaneously), the child welfare agency and probation often disagreed about which agency should be lead on the case. This tension was not raised in the dual-status counties that were part of the case studies. Some common factors facilitated collaboration, including building from existing coalitions and creating formal partnerships. In both the county plans and case studies, we found that some larger counties began FINDINGS 27 their steering committee by building from existing coalitions related to CSEC. For example, both San Francisco and San Diego counties maintained their own human-trafficking coalitions for years before S. B. 855; they noted that this framework was useful in forming the CSEC steering committee and broader CSEC coordination. Many counties mentioned that formal partnerships with multiple organizations was essential to effectively integrating CSEC services across systems. Multiple partners with overlapping services allowed them to move past service delays at any one provider. While most counties reported difficulties in coordinating services out of county, some reported that they worked with providers in neighboring counties to expand the services available to their young people. On rare occasions, counties set up formal agreements for cross-county services. Staffing and Training STAFFING Child welfare agencies took the lead on coordinating CSEC response and communicating across partners, indicating that the CSC component of backbone support (table 2) is present. Many counties hired a dedicated CSEC coordinator housed in child welfare to oversee the day-to-day CSEC response, in line with the CSC component of a paid project coordinator or manager (table 2). In the county plans, counties that mentioned a dedicated CSEC coordinator included the majority of urban counties (N = 9), some suburban counties (N = 6), and one rural county. This person worked with other county agencies and community partners to coordinate CSEC activities and monitor the progress of individual cases. While the coordinator role varied from county to county, they often convened steering committee meetings, coordinated MDTs, managed contracts with county-specific service providers, arranged and/or led trainings, provided technical assistance to workers, coordinated CSEC awareness activities, and managed data collection. Even where there was no dedicated CSEC coordinator, child welfare nearly universally took leadership over county CSEC activities. The survey was typically filled out by county CSEC coordinators, or the person most responsible for the county's CSEC response in counties that did not have a dedicated full-time coordinator. Survey respondents reported spending about one-third of their workweek coordinating their county's CSEC program, though responses varied widely (M = 32 percent, SD = 34). In seven counties, survey respondents reported that CSEC coordination comprised at least 75 percent of their workweek, while in 15 counties it accounted for 10 percent or less. Respondents who spent more of their workweek on CSEC program coordination were more likely to be funded, either partially or fully, by S. B. 855 opt-in 28 FINDINGS funding. Respondents spent more of their workweek on CSEC coordination in counties with larger CSE populations. On average, counties with a low CSE population spent 8 percent of their workweek on CSEC coordination, counties with a medium-sized CSE population spent 16 percent, and counties with a high CSE population spent 65 percent. As described above, CSEC coordinators were more common in urban areas, and some suburban counties employed CSEC coordinators, but others expressed the need for one. For example, in one county, a staff member reported that they were able to dedicate 20 percent of a full-time equivalent to CSEC coordination but expressed that it should really be a full-time job. About half of counties reported on the survey that they have specific units or dedicated social workers that respond to CSE. In four of these counties, these units or workers were solely dedicated to supporting young people experiencing or at risk of CSE. Rural counties were less likely to employ CSE- specific staff and were more likely to rely on general county child welfare staff to handle CSEC program activities. Sometimes, these additional responsibilities were added to social workers' existing duties. Some counties, especially larger urban counties, employed multiple CSEC-specific staff who were a CSEC team. For example, Los Angeles County employed two CSEC court liaisons, two CSEC administrative staff (for administration, events, reports, trainings, and more), and two hotline staff who split their time between CSEC-specific responses and general emergency responses. These larger counties said that specialized teams helped coordinate agency activities and support young people without overburdening any particular agency. For counties without dedicated CSEC staff, county child welfare program managers often led CSEC activities in addition to their other duties. Apart from dedicated CSEC staff, child welfare staff led or were actively engaged in CSEC collaboration in essentially every county. Frontline child welfare workers were the most common point of contact for young people experiencing or at risk of CSE. Frontline workers often had the most knowledge of young people's needs and were best equipped to connect them with services. Case- carrying child welfare staff led service provision and MDTs, but they also sometimes contributed to broader CSEC program activities such as steering committees. Child welfare staff also generally had the most knowledge of CSEC procedures, including screenings such as the Commercial Sexual Exploitation- Identification Tool (CSE-IT) developed by the WestCoast Children's Clinic.19 After child welfare, juvenile probation was the agency most involved with CSEC programs. Probation staff commonly conducted CSE screenings and made service referrals for young people under their supervision. Health care staff were also key in many counties, although there were many counties in which health care (medical and behavioral) staff were minimally involved. Counties that directly employed FINDINGS 29 health care staff such as therapists or public health nurses (usually urban counties) reported in county plans that these staff were key in the initial services provided to young people. Notably, however, only one of the case study counties spoke about health care staff playing a role in initial services. Law enforcement staff and nonprofit service provider staff were also key in many counties' CSEC responses. Staffing capacity is a challenge statewide. On the survey, counties reported they have enough staff for less than half of the necessary time on average to serve young people who are at risk of or experiencing CSE. Staff shortages and turnover were also cited as a challenge in nearly all case study counties. One rural county declined to participate as a case study because of a severe staffing shortage. Reduced frontline staffing in both child welfare and provider agencies pushed caseloads higher, making it harder to be responsive to collaborative partners. Counties described how challenges with staff and service capacity increased the time a youth has to wait for services. Participants we spoke with said that staff turnover reduced trust between young people and families and the county agency, making them reticent to accept or continue with services offered. Staff turnover also undermines institutional knowledge about CSE. To reduce staff stress and burnout and ultimately reduce turnover, some counties accounted for the greater time required for CSE cases by weighting them more heavily in caseload assignments. As one child welfare supervisor said, It makes it hard for the families because they don't know who to talk to because there's that inconsistency with staffing. When the family's already been investigated and has to tell the same story over again to another person, I think people lose hope on the provider. It can be frustrating for the workers trying to refer but not nearly as frustrating for the parent who is trying to access the service. We fall short there. Ensuring that the right staff are hired into CSE-specific roles was also cited as critical to retention. Participants we spoke with emphasized it was important for people to care about and have a passion for this work to be effective and have longevity in their positions. Many counties recounted in the case studies and their county plans the value of engaged social workers for cases where young people were at risk of or experiencing CSE. When social workers were able to form connections with young people, they were more successful in convincing them to participate in services. Larger counties with resources to form a CSEC response team were the most equipped to identify and quickly serve young people experiencing or at risk of CSE. Staff said that having a unit specifically designated to respond to CSE, but not exclusively responding to CSE, was helpful. This staffing model concentrated CSE expertise in one unit, but it reduced burnout because staff were not solely responding to CSE cases. One county described a unique staffing structure: within their CSE team, they have (1) a CSEC clinician who is a licensed therapist and serves as a mental health resource for young 30 FINDINGS people and (2) a missing persons liaison who carries cases for all young people who are absent from placement without permission and actively searches for them. TRAINING S. B. 855 requires counties to offer CSE training to county children's services workers and foster caregivers. Most counties reported on the survey that they meet these requirements, with 40 counties (87 percent) reporting that they obliged frontline workers and supervisors to complete at least one CSE training and 39 counties (85 percent) reporting that they offered CSE training to foster caregivers. While not mandated by the legislation, 29 counties (63 percent) required child welfare managers to complete at least one CSE training, and 42 counties (91 percent) offered voluntary CSE training to child welfare staff during 2021. All counties offer child welfare staff a wide range of trainings. The county plans indicated that some trainings were offered across all opted-in counties, while others were specific to only a few counties. The most common trainings were standardized CSEC 101, CSEC 102, and CSE-IT training by a vast margin. CSEC 101 and CSEC 102 are funded by S. B. 855 and are also open to the public in addition to being a common choice for child welfare staff. Trainings on harm reduction, CSEC intervention, and survivor perspectives were also offered in the majority of counties, though these trainings were less standardized. In the county plans, fifteen counties mentioned specific trainings for child welfare staff: youth empowerment; supporting LGBTQ+ young people; "runaway" youth training; the intersection of domestic violence, sexual assault, and human trafficking; CSEC in tribal communities; and behavioral skills training. Counties sometimes used S. B. 855 funds to provide training and education on CSEC topics to a variety of stakeholders beyond child welfare staff. In county plans, both rural and urban counties noted that some agencies (such as law enforcement) had less CSEC knowledge and training or differed in the level of CSEC awareness and buy-in to integrated service delivery. This was important because staff said that this lack of training impeded collaboration. The vast majority of counties also reported offering a range of CSEC trainings to non–child welfare professionals and community partners in survey responses, county plans, and/or through site visit interviews. The recipients of these trainings were most often staff in probation, law enforcement, or community partners, though other groups such as court system staff were sometimes mentioned. Most often, trainings for other professionals or community partners were the same CSEC 101, CSEC 102, and CSE-IT trainings offered to child welfare staff. Almost all counties provided some type of training to young people, though the specifics varied widely by county. The most common training topics were social media safety, raising awareness of CSE, and how to identify or avoid a potentially harmful situation, healthy relationships, and general human- FINDINGS 31 trafficking awareness. Trainings for children, young people, and families in the general community usually took place in a middle school, a high school, or at a community event. Many counties contracted with organizations to run licensed events such as "Word on the Street," "Ending the Game," or "Traps of a Trafficker" to educate young people on strategies for protecting themselves against CSE. These trainings were also occasionally held for young people experiencing or at risk of CSE. A few counties mentioned specific trainings for young people experiencing or at risk of CSE that focused on issues such as behavioral wellness, empowerment, or harm reduction. Many case study counties used a prevention training curriculum entitled the "Cool Aunt" series. Young people in foster care often received these standard trainings, but trainings specifically tailored for young people in care were not often mentioned in county plans or case studies. Trainings for caregivers, and foster care caregivers specifically, were common but occurred most often in urban counties. County plans mentioned trainings for foster care caregivers in 41 percent of rural counties, 50 percent of suburban counties, and 86 percent of urban counties. These trainings addressed CSEC generally and/or focused on harm reduction. Trainings for other caregivers (who may come from the community or be a parent involved in child welfare) occurred in 24 percent of rural counties, 64 percent of suburban counties, and 71 percent of urban counties. These trainings could focus on harm reduction, social media safety, general CSEC awareness, trauma-informed skills, or support for LGBTQ+ young people. They were most often offered as a part of larger community education efforts on CSE, as a pairing to a training offered to young people, or specifically to parents or caregivers of young people at risk of or experiencing CSE. Trainings were often delivered by community organizations. In some circumstances, these organizations prepared county staff to conduct trainings locally on an ongoing basis (i.e., train the trainer). For example, San Diego County placed many different preventative trainings under the umbrella of a new San Diego Trafficking Prevention Collective, creating a centralized location for human-trafficking education. Los Angeles County developed an annual CSEC empowerment conference for young people, and a few other counties mentioned sending young people to this conference. Some counties expressed the desire for some updated and additional training. Some interviewees said that trainings should be updated to account for the most current CSE trends. For example, key informants in several case study counties expressed a need for training on cyber recruitment, such as which websites are most commonly used and how to prevent and address this type of recruitment. Another case study county said that they would like training about tattoos and other visual indicators that may indicate CSE. 32 FINDINGS Screening S. B. 794 established a requirement for public child welfare and juvenile probation agencies to identify young people experiencing or at risk of CSE, and S. B. 855 further reiterated counties' responsibility in identifying young people for whom to intervene and provide services. We asked counties what situations prompt a CSE screening, what screening tool they use, and their perceptions of how effective their screening processes are at identifying young people experiencing or at risk of CSE. County plans also described how counties teach staff about screening. SCREENING CIRCUMSTANCES Table 4 shows situations that prompted a county to screen for CSE. All counties that responded to the survey question said that they screened when a child returns after running away from placement. CSE screening also often occurs during maltreatment investigations and hotline screenings, after interactions with law enforcement, and upon placement in juvenile detention. Fewer counties reported that key CWS events-opening a family maintenance case (39 percent) or placement of a child in out-of- home care (48 percent), or when there is a placement move (24 percent)-triggered screening. About a quarter of counties indicated other situations that would lead a social worker to suspect CSE, including pregnancy, frequent sexually transmitted infections, connections to those known to be involved in CSE, having unexplained money or gifts, or talking to strangers online. TABLE 4 Situations That Prompt CSE Screening Does not Prompts prompt Missing/don't Situation screening screening Know Upon a child's return after absence from placement 89% 0% 11% During a maltreatment investigation 83% 13% 4% During hotline screening 80% 11% 9% After interactions with law enforcement 70% 11% 20% Upon placement in detention or juvenile hall 65% 7% 28% During involvement with a probation agency 52% 11% 37% Upon placement in out-of-home care 48% 28% 24% Upon starting family maintenance services 39% 41% 20% Whenever the child's placement changes 24% 43% 33% Other event or situation 22% 17% 61% Source: Survey of CSEC program administrators in California counties that had opted in to the CSEC program. Note: Row percentages are the share of the 46 counties that responded to the survey. FINDINGS 33 Figure 3 displays the share of counties that reported that each agency screened for CSE. These agency categories were not mutually exclusive, and a county could have reported multiple agencies that screened for CSE. We saw that child welfare and probation screened for CSE in nearly all counties, as required by S. B. 855. Counties reported less frequently that agencies such as law enforcement or community partners screened for CSE. But many counties (48 percent) responded that at least one other organization, besides child welfare or probation, screened for CSE. FIGURE 3 CSE Screening Rates by Agency Share of surveyed counties where the agency screens for CSE 96% 76% 28% 24% 13% 9% Child welfare Juvenile probation Law enforcement Juvenile courts Local homeless Other youth shelter or drop-in center Source: Survey of CSEC program administrators in California counties that had opted in to the CSEC program. Notes: N = 46 counties that responded to the survey. The organizations in the "Other" category were most commonly community partners, such as child advocates or behavioral health service providers. SCREENING TOOLS The Commercial Sexual Exploitation-Identification Tool (CSE-IT) developed by WestCoast Children's Clinic is the only evidence-based screening tool available and was the most commonly used screening tool, with about two-thirds of counties using it (67 percent).20 Five counties (11 percent) reported that they used the Child and Adolescent Needs and Strengths Assessment (CANS), and four counties (9 percent) reported that they used Structured Decision Making (SDM) in addition to the CSE-IT to identify CSE. Eight counties (17 percent) reported using another CSE screening tool that is not the CSE-IT, CANS, or SDM. Most of the other screening tools reported on the survey were instruments developed by counties themselves. Four counties listed multiple tools. A few counties described in their county plans that different agencies used different assessments (e.g., the child welfare agency using CSE-IT and another using their own tool). In some counties, screening was happening with a lot of frequency, but 34 FINDINGS that did not necessarily translate to increased efficacy and sometimes felt more like "checking a box." As one social worker said, We know we aren't getting one from every worker for every kid every six months. When CSEC screening tools were used in addition to other standard practice assessments like SDM and/or CANS, staff reported that the additional screening tool is burdensome. A few urban counties reported such burden, as they already used a formal assessment before CSE-IT became popular. One key informant, a child welfare program planner, said that staff felt disincentivized to complete the CSE-IT: The CSE-IT truthfully was not being done really well…We use SDM and CANS so we're already capturing it. So, the CSE-IT wasn't being used as regularly because it gave you results you knew anyway. In their county plans, about two-thirds of counties explicitly mentioned that they regularly trained staff on screening. Some counties held collective trainings for all agencies that were intended to use the CSE-IT, while others expected each individual agency to train their own staff. Most of the time the staff targeted for trainings were social workers within the child welfare agencies themselves. Juvenile probation staff were also frequently offered trainings, though some counties noted that these staff were not very comfortable with the tools after being trained. Some of the counties that used the CSE- IT contracted with the WestCoast Children's Clinic to provide both staff trainings and train-the- trainer trainings. Others held informal trainings by staff knowledgeable in CSEC, who personally trained new hires. Fourteen counties explicitly mentioned in their most recent county plans that their staff improved their use of screening tools in later years of implementation. Some counties mentioned that their screening tools helped them better understand the prevalence of CSEC. A significant minority of counties were also able to better leverage screeners in their collaboration in later years by integrating them in collaborative activities such as MDTs. One child welfare supervisor said, Before our county implemented the CSEC tool and before that assessment, we didn't know much about CSEC in general. We would just categorize it as sexual abuse or neglect…Now when I think about the kids from the past, I recognize that it was trafficking, I just didn't know at the time. Despite improvements over time, counties expressed some concerns about the efficacy of their screening. Some counties reported that supervisors do not have the time or expertise to assess fidelity to the tool and ensure it is used on every eligible youth. On a sliding scale where zero was not at all effective and 100 was very effective, counties on average rated their screening processes for identifying young people who have experienced CSE as 65. Counties reported slightly higher success identifying young people at risk of CSE, with an average rating of 71. Some key informants said they are FINDINGS 35 worried that they are not identifying some young people. One person mentioned that their county's age cutoff of 12 and older might miss younger people at risk. The age cutoff for screening varied among counties, with many counties maintaining a cutoff of 10 years old. Counties' perceptions of their screening effectiveness did not vary significantly based on their region, urbanicity, overall population size, CSE population size, CSE rate, or whether a county was an S. B. 855 pilot county. Additionally, counties' screening effectiveness was not related to the total number of situations that prompt CSE screening in the county or whether they use the CSE-IT as their primary screening tool. Counties identified multiple barriers that obstructed their CSEC screening. The most common barrier was staff turnover. High turnover increases training demand because new frontline staff need to be trained on the tool. In some counties, newer frontline staff were not quickly trained on the tool and were unable to properly fill out the county's screener. Turnover among staff that conducted screening trainings left fewer staff with the knowledge to train others on CSE-IT and other tools. Some counties also said that screening results for ongoing cases were not updated because of staff turnover. Data sharing was another frequently mentioned challenge related to screening. In some counties, different agencies used different screening tools, which sometimes led to data duplication or inconsistent data collection depending on the questions in each tool. Counties also mentioned that they identified young people coming from a neighboring county but could not see previous screening results because they did not have data-sharing agreements with the origin county. Many counties said that some partner agencies did not screen for CSE, possibly contributing to underidentification. Multiple key informants said that law enforcement did not fill out the screening, though they were the first contact for many young people experiencing or at risk of CSE. While law enforcement contacted the child welfare agency as part of the CSEC protocols, some wished that law enforcement would complete a screener when they reported the incident. Multiple key informants hoped for formal or informal CSEC screening to occur directly in schools in the future. They felt that schools were in the best position to identify at-risk young people. A key informant in an urban county, a CSEC coordinator, said, It's education, it needs to be in schools, for parents, looking at signs…By the time we get them we see there could have been services and education prior to committing a crime…There just has to be more education because by the time they get to it, it's like they were probably victimized a year ago. 36 FINDINGS Finally, some key informants mentioned that staff are resistant to the formal screening data collection measures, feeling that mandatory formal screenings were an inefficient use of time. One child welfare supervisor described, More and more workers are resentful about it…They would rather use their assessment skills to determine if it's needed. They want to do it if they feel it's warranted and not just on every child over age 10. At the identification stage, many counties said that universal screening (especially CSE-IT) helped quickly identify young people who were experiencing or at risk of CSE and begin the process of delivering CSEC services. Many counties spoke highly of CSE-IT's accuracy in identifying the need for CSEC services (when staff are trained to use the form). Multiple key informants appreciated CSE-IT's observational approach so that staff did not have to unduly question young people. One CSEC coordinator said, One of our favorite things [about the CSE-IT] is that it's observational-they don't have to ask questions. It's based on what they know about the youth. That makes it easier than having to ask uncomfortable questions to the youth, but also hard to do the tool if you haven't been working with the youth for that long. Once staff filled out a screener, the results served as a helpful baseline for initial collaboration. For example, many counties share the CSE-IT results among partner organizations to prepare everyone for an MDT meeting. Services One goal of S. B. 855 is coordinated, targeted services that meet the unique needs of young people experiencing or at risk of CSE. Below we describe our findings about service availability in opted-in counties, including the types of services offered and their ability to meet young people's needs. We found that, overall, the primary way specific service needs were identified was through a child welfare caseworker or juvenile probation officer (63 percent) or during a CSEC MDT or a CFT (20 percent). But while counties had common pathways to identify service needs, they varied widely in the type, availability, and uptake of services. Most counties offered services for independent living skills, employment, and substance abuse treatment; although in most counties these services were offered to all young people, and rarely were they designed specifically to serve those experiencing CSE (figure 4). The service most likely to be available as a CSEC-specialized service was youth advocates or mentors followed by mental health and medical care. FINDINGS 37 FIGURE 4 Service Availability For which service types do counties have CSEC-specialized or traditional services available for young people at risk of or experiencing CSE? Traditional CSE specialized Both Not available Missing Independent living skills 85% 9% Employment 85% 4% 4% 4% Substance abuse treatment 80% 4% 7% 7% Medical 70% 22% 7% Mental health 61% 28% 9% Youth advocates/mentors 43% 41% 7% 4% Other 9% 9% 57% 24% Share of responding counties Source: Survey of CSEC program administrators in California counties that had opted in to the CSEC program. Notes: N = 46 counties that responded to the survey. Figure summarizes answers to the survey question, "For each service type, please indicate if it is available as a CSEC-specialized service, traditional service, or not available." A large majority of counties offered mental and physical health services (70 percent as a traditional service, 22 percent as a CSE specialized service, 9 percent as both; see figure 4 above). These services were sometimes provided by the child welfare, public health, or mental health agency, but the most common strategy, by far, was for county staff to refer young people to an outside health care provider. Counties did not detail the specific services offered-or not offered-by health care providers. Twenty-nine counties mentioned in the county plans that they directly provided health services to young people, including therapy and forensic exams through child welfare, public health, mental health, or a general social services agency. In the county plans, two counties mentioned a county-employed public health nurse who provided services to young people. However, most counties did not reference their agencies' capacity for direct health services. On the survey, most counties also reported offering substance use treatment services. When these services were mentioned in the county plans, county staff reported referring young people out to substance use service providers. In the county plans, nine counties mentioned the availability of wraparound services. Most mentioned 38 FINDINGS general wraparound services without describing them further. Wraparound providers differed across counties and the services could be situated in child welfare, juvenile probation, or a community agency. County plans mentioned a wide variety of housing support services for young people. Many counties offered some form of temporary housing, usually through community partners. For both short- and long-term housing, one-third of the counties-mostly but not all rural-used CSEC funds to financially support young people's housing, even offering to pay rent deposits or buy furniture. Independent living programs, operated by county child welfare agencies, provided housing assistance along with a variety of supports and trainings to young people. Some counties mentioned placing young people experiencing or at risk of CSE with foster families. However, many counties lamented not having enough (or any) placement homes willing to take this population. Counties reported that they had enough family-based placements only 11 percent of the time on average. We discuss this in more detail below, where we address the shortage of family-based placements. Many counties offered other services such as employment and education, with some counties directly providing CSE young people with learning or work opportunities. For example, 93 percent of counties offered employment services, either as a traditional or CSE specialized service. Another common type of service was normalizing activities to help young people recover from CSE. For example, some counties offered tattoo removal that might be connected to their experiences of exploitation as well as, music lessons, art supplies, or dental work for young people who were trafficked. Los Angeles County maintained an individualized incidental restoration fund to help young people with any of these activities or more. A few counties mentioned vouchers for clothing or transportation (such as a bus voucher). Counties used a variety of harm reduction strategies woven in as a service strategy in broader service offerings. About a third of counties said in their county plans that adults with direct experiences of CSE were part of the CSEC program services, either as advocates, mentors to young people, or programmatic advisors, but only one of those counties was rural. Counties expressed the importance of an advocate or mentor, though often offering that it was impossible to provide every youth at risk of or experiencing CSE with a mentor. Advocates (with and without direct experiences of CSE) were sometimes employed by a county agency, but more often they were employed by a community partner. Counties noted a dearth of service slots, family-based placements, and CSE-specialized providers (table 5). Counties reported that they had enough CSE-specialized providers only 29 percent of the time. Young people were often referred to other counties for specialized services because they were not available locally (54 percent of the time). FINDINGS 39 TABLE 5 CSEC Service Capacity How often did counties report enough service capacity for young people at risk of or experiencing CSE? Please rate how often each statement below is true.* Mean There are waitlists for CSE-specialized services. 21% We are able to match young people who are at risk of or experiencing CSE to the best service to meet their needs, rather than referring them to the first available program slot. 51% There are enough CSE-specialized providers to meet the needs of all children in our county. 29% We have to send young people who are at risk of or experiencing CSE out of county to receive specialized services. 54% Source: Survey of CSEC program administrators in California counties that had opted in to the CSEC program. Notes: Mean response of the 46 counties that responded to the survey. *Scale went from 0 to never to 100 to always. Counties noted in their county plans that many necessary services had waitlists. The most common services with low capacity were family-based placements, followed by health care and advocate services, but counties often faced shortages of other services as well. A few counties said they felt that they did not have enough funding to provide all necessary services for young people at risk of or experiencing CSE. Low resources for CSEC and human services in general in small counties were barriers to service coordination and integration, even when agencies were motivated to collaborate. Social workers in rural counties said it is especially hard to coordinate services because their roles were stretched. The shortage of family-based placements for young people experiencing or at risk of CSE came up repeatedly. Key informants shared that placement homes were frequently unwilling to take young people experiencing CSE, and this constrained staff members' ability to respond to young people's needs. Staff spent significant time searching for appropriate placements. On average, counties reported that they have enough family-based placements only 11 percent of the time. This concern was commonly reiterated in the county plans, and placement availability was noted as a challenge in almost all counties we spoke with. The issue is inextricably tied to other services, which in theory would be readily available and help stabilize young people, making family-based placements more realistic. One county council leader we talked to described in frustration the lack of suitable placements for this population: There is a lack of a mental health care-lack of substance abuse treatment. Juvenile hall should not be a frontline for mental health and substance use. We're all aware of that, and we try to work collaboratively but one of the biggest frustrations has been the continuing care format. Everyone wants kids in a foster home, [but] there are highly dysregulated and reactive behaviors; the ones coming into juvenile hall-the girls who are there who are CSEC-they have strangled someone, it has been a very violent assault…they're still traumatized but it's also a huge public safety risk…everyone waits for an adequate placement, and we don't have enough of them. We don't have foster homes for youth who are AWOL and drug using, and they can't go out of state. 40 FINDINGS Involving experts with direct experience of CSE in services was a goal for many counties, but implementation has been challenging. Some counties wanted to involve survivors in services, but they either did not have enough willing survivors to engage every CSE youth or they faced high turnover among the survivors they do hire. These counties, usually smaller and more rural, lamented having no way to offer survivor support to young people. Among the survivor consultants we interviewed for the site visits, some noted challenges such as a lack of clarity about their role. Many counties described support programs that they were unable to implement because the county lacked available service providers. Some counties mentioned challenges with getting young people access to necessary health care services. For example, staff from one county said that some of their young people had to wait months for only one health appointment at a health care provider. And many counties said that they could not hire staff or find a proper community partner to provide substance abuse services to young people at risk of or experiencing CSE. Counties also reported multiple barriers in providing proper mental health services for CSE young people. The most common problem was a lack of treatment slots, leading to long delays for consistent treatment. Secondly, many counties could not find mental health care providers with CSEC knowledge or trauma-informed practice. Sometimes counties knew of services, but they were simply too far away to be practical for young people to attend. Many counties reported that young people at risk of or experiencing CSE sometimes ran away from placement either before or during services, or simply refused recommended services. Counties viewed these scenarios as being central challenges to service provision with this population. In terms of practices that facilitated services, we heard during the case studies that having dedicated CSEC staff or advocates facilitated the uptake of CSEC services. A few counties have a 24/7 dual response from child welfare and a CSE advocate (from a service provider partner agency) when going out for investigations where CSE is a concern. One nonprofit CSE advocate said, I like how they brought us into the mix. We're not mandatory-that is a really special element to our specific protocol. That they brought in some external folks that are not law enforcement-that's unique. We're attractive to youth because of that. We're not wearing a badge, not part of the investigation piece. I'm not sure how many other counties do that, but it is well received from youth. Low-barrier drop-in centers were also noted as very helpful in several counties. These drop-in centers provided a range of services for young people ranging from basic one-time needs such as food and clothing to more complex ongoing services such as comprehensive case management. Staff spoke about FINDINGS 41 how the drop-in centers were often successful at engaging young people, particularly when they had not engaged with other agencies. One service provider shared, Youth realize we're not mandated reporters, which means they can be honest. We won't make them do anything they're not ready for, like reporting [their exploiter] to law enforcement. We're able to pick up a lot of things that child welfare misses because the youth may know they're going to be removed…We try to make it more relaxed for the youth, not like an office. It's more like a house and we're trauma-informed so everyone's comfortable. Service access, availability, and provision varied geographically. In general, rural counties reported having fewer available services and fewer specialized services based on the county plans and case studies. Rural counties reported difficulties developing the services that other counties had, from immediate CSEC response teams to supportive services. Some small- to mid-sized counties explicitly stated that lack of variety and limited resources were their primary barriers to fully supporting young people. While placements are a problem everywhere, it was even more difficult for rural counties to find appropriate placements for young people experiencing CSE because they have very few foster homes to begin with. Some rural counties also mentioned that they did not have the workforce to properly engage young people, especially through survivor mentors. The survey analyses supported these findings. Urban counties were significantly more likely to report being able to match young people to services that best meet their needs. Urban and large counties were more likely to report having enough CSE-specific service providers and less likely to report having to send young people out of county to receive services. Counties with high CSE counts were also more likely to report having enough CSE-specific service providers compared with counties with medium or low CSE counts. Urban counties generally had a wider swath of services and more comprehensive versions of services. In the immediate response phase, urban counties reported more comprehensive response teams. For example, counties such as San Diego or Santa Barbara had comprehensive CSEC Response Teams that could include social workers, law enforcement, health services, and more, depending on the individual case. Similarly, both urban and rural counties mentioned supplementary services to assist the diverse needs of CSE young people; but urban counties often had institutionalized systems for providing such services. One example was Los Angeles County's special Individualized Incidental Restoration Funds, which maximized normalizing opportunities for CSE young people, from removing tattoos to getting dental work to pursuing an education. Urban counties were also more likely to have service providers employed by the county, such as a county therapist or public health nurse. Rural counties almost always referred health concerns to an outside provider. 42 FINDINGS Wraparound service availability depended on the county context: counties happily referred young people to an existing resource center for wraparound services if such a center existed in their county. Some counties, such as Santa Cruz, referred Native American young people to specialized tribal service providers for wraparound services. Data Accurate data collection is an important part of a coordinated CSE response. We asked counties how they collected and shared data among agencies, as well as their perceptions of data quality. The answers suggested that while there has been progress toward the CSC component of shared measurement (table 2), this is an area for improvement for many counties. DATA COLLECTION Of the 46 counties that responded to the survey, 39 (85 percent) reported that they entered data using CSEC special project codes21 in CWS/CMS. In county plans, counties described collecting a variety of data in CWS/CMS, including information from community partners, health history, and school involvement. About two-thirds of counties reported sharing data among county partner organizations. Thirty child welfare agencies (65 percent) shared aggregate CSEC data with other collaborating agencies, and 28 counties (61 percent) reported that collaborating agencies shared their aggregate CSEC data with the child welfare agency. UNDERSTANDING CSE DATA In county plans, many counties said that they standardized their CSEC screening and related data entry over time as staff across agencies became familiar with the processes. However, progress was often uneven. For example, many child welfare staff quickly began using the CSEC special project codes in CWS/CMS while staff in adjacent agencies did not use the codes reliably because they were not experienced with CWS/CMS. Several counties mentioned that they provided staff in these partner agencies with a documentation guide to introduce CWS/CMS, with visual examples of pertinent actions like the special project codes. The county plans indicated that this guide was especially useful in learning and accurately using CWS/CMS. We asked counties to share their perceptions of how accurate and helpful their data are for identifying and serving this population (table 6). On a scale where 0 was not at all accurate and 100 was very accurate, counties reported that their CSEC program data were moderately accurate (mean = 58). FINDINGS 43 Counties also reported that they had the data they needed 43 percent of the time to effectively serve young people who are experiencing or at risk of CSE. TABLE 6 CSEC Data Efficacy Question Mean How accurate is the CSEC program data collected by your agency?* 58 Does your county have the data needed to effectively serve young people who are at risk of or experiencing CSE?** 43 How often does your agency and collaborating agencies jointly review data on CSEC program outcomes?*** 37 Source: Survey of CSEC program administrators in California counties that had opted in to the CSEC program. Notes: Mean response of the 46 counties that responded to the survey. *0 is not at all accurate, 100 is very accurate; **0 is none of the data needed, 100 is all of the data needed; ***0 is never, 100 is very often. Thirty counties mentioned in their annual plans that their staff needed more technical assistance and cross-system training in screening, documenting, and tracking cases in data systems. Staff turnover across county agencies compounded the problem because it was difficult to continuously train staff. Many counties also said that problems sharing case information across agencies delayed case reviews. Efforts to set up data-sharing agreements were often protracted. In some counties that completed data- sharing agreements, specific agencies used data systems that did not sync with CWS/CMS, or they did not have any formal data system in place whatsoever. When county agencies did not use the same data systems, they would often create independent cases across the different systems for the same young people. This made it harder for staff to collaborate and to enumerate the total number of unduplicated CSEC cases countywide. Some counties reported that they hold case consultation meetings with their MDT members but that these meetings are inconsistently scheduled, which reduced opportunities for data sharing. When they did occur, staff noted that it was difficult to track in CWS/CMS. Finally, many counties described difficulties receiving data for young people with cases open in another county. Only six counties mentioned having a data-sharing agreement with any neighboring counties. Key informants and county plans indicated that CSE data in cases were often not updated on time, were inaccurately updated, or were not formally entered into CWS/CMS. County plans described staff confusion on promising practices for managing and entering data in CWS/CMS across multiple agencies. Counties partially attributed the confusion to high turnover, reporting that it was difficult for counties to keep all staff fully trained on data procedures. County plans also described staff being unable to find all relevant data for cases even if case data were up to date. Because there was no CSEC tab or filter in CWS/CMS (only the CSEC special project codes), staff could be confused about how to best find CSEC information. 44 FINDINGS Data tracking contributed to interagency collaboration on cases involving young people experiencing or at risk of CSE, such as in MDTs. When multiple county agencies used the same data system-usually CWS/CMS-they were generally able to access information and collaborate regularly. In MDTs, information from CWS/CMS or CSE-IT served as a helpful baseline for agencies to begin discussing a case. Multiple counties found success in using CWS/CMS data to identify specific areas of concern for young people who are at risk before they are in a crisis. They said that the data helped them make faster referrals and specialized placements. Data also helped some counties identify the relationship between service use and outcomes for young people. For instance, one urban county used case-level data to connect a youth's service use with their progression through stages of change, with focus areas ranging from homelessness to mental health. Most counties reported successes with their data systems-especially CSE-IT and CWS/CMS-in identifying CSEC and at-risk young people. However, suburban and urban counties were more likely to report that they actively used their CWS/CMS data to understand trends or improve their CSEC processes. For instance, one urban county analyzed their data to investigate the most significant risk factors for CSE in the county. Outcome Study S. B. 855 defines CSEC and establishes responsibilities and goals for the CWS to lead efforts to address the sexual exploitation of children, while offering funding and materials to support coordination of care across agencies. This two-pronged approach specified in the legislation was expected to generate increases in the following system performance areas:  identification of minors as at risk or with confirmed CSE victimization  assessment of service needs within child welfare and related systems  delivery of appropriate services by those systems  better outcomes for young people While the implementation study provided insight into the first three areas, the outcome study is focused on the first (identification) and the last (outcomes) to the extent that these expected changes were evident in the statewide administrative data collected by counties. Information available from CWS/CMS offers a partial picture of child welfare outcomes as a result of S. B. 855 implementation efforts, but because service delivery data are not often inputted in CWS/CMS, direct connections FINDINGS 45 cannot be made between service delivery and better outcomes. In the next section, we report findings that document child welfare agencies' efforts to identify and intervene on behalf of young people who were identified as at risk of CSE or met the state's definition of CSEC victimization. In addition, we describe significant associations between identification, system performance outcomes, and county context with the aim of identifying the specific settings where CSE concerns may be more prevalent. Detailed tests of associations between identification, system outcomes, and county-level factors such as whether a county was an S. B. 855 pilot, urbanicity, region, collaboration score, and whether the county implemented the minimum implementation requirements, are included in appendix L. CSEC Identification Participating child welfare agencies are required under S. B. 855 to track whether victimization occurred either (1) before a child entered foster care; (2) during an in-home case but not during foster care; (3) during foster care; or (4) during a foster care placement episode, but not in a placement home or facility. FIGURE 5 Trajectories of CSE Victimization and CWS Involvement Source: Authors' depiction of possible trajectories. The timing of CSE identification relative to CWS involvement has implications for what the system-including cross-system partner agencies-can do to address it. Figure 5 illustrates the trajectories of CSE victimization and child welfare involvement. Children whose service pattern matches trajectory one (Victimization  Foster Care) are those for whom foster care may have been a response to the CSE concern. Trajectories two through four suggest situations where more services may be needed to prevent victimization during child welfare involvement. By July 1, 2015, CDSS required all counties to document CSE victimization at the event level via CPS maltreatment reports and at the child level as victimization entries on a child's client description 46 FINDINGS page.22 We therefore used CWS/CMS data collected between July 1, 2015, and June 30, 2022, to examine CSE events documented in CPS maltreatment reports and on the client notebook, as well as additional documentation methods (i.e., special project codes) used among a subset of opted-in counties. CSE MALTREATMENT REPORTS To understand the degree to which counties implemented CSE documentation protocols and whether CSE reporting changed over the implementation period, we examined CSE reporting, investigations, and substantiation at both the event and child levels, along with the reporting parties that generated CSE reports. Between FY 2015–16 and FY 2021–22 (i.e., July 1, 2015, and June 30, 2022), a total of 70,334 CSE reports were documented by CPS (figure 6). A majority of CSE reports identified a single child; however, some children were the subject of multiple reports. Nearly two-thirds of those reports (64 percent) were screened in for investigation. Among those, more than one in five were labeled inconclusive (22 percent) and one in four reports were substantiated (25 percent). We found that pilot and urban counties were more likely to investigate reports of CSE maltreatment but were not much more likely to substantiate those allegations. These findings may indicate a greater capacity among pilot or urban counties for child maltreatment investigations but may also reflect that urban counties were also more likely to report CSEC via child maltreatment reports than through a note on a child's client description page. Regionally, the most notable finding is that despite being an urban locale, counties in the Bay Area were the least likely to both investigate and substantiate allegations of CSE. On average, counties in Southern California had higher percentages of initial investigations as compared with Bay Area, Central Valley, and Northern Californian counties. Finally, we found that counties that scored high in collaboration and met the minimum S. B. 855 implementation requirements were more likely to investigate and substantiate reports of CSE. Next, we considered reports, investigations, and substantiations that occurred throughout the implementation period to determine if the CSEC program had an impact on CSE documentation and management. Slightly more than half (53 percent) of the total 70,334 CSE reports were made during the first half of S. B. 855 implementation. Further, while the share of screened-in reports was slightly higher during early implementation (65 percent versus 62 percent), the share of substantiations remained consistent across implementation periods (24 percent versus 25 percent). FINDINGS 47 FIGURE 6 CSE Maltreatment Reports and Child Protection Responses Count of CSE Reports in Opted-In Counties between FY 2015–16 and FY 2021–22 70,334 44,672 10,933 Total CSE maltreatment reports Investigated CSE maltreatment reports Substantiated CSE maltreatment (63.5%) reports (15.5%) Source: Administrative records from California's child welfare state administrative data system, Child Welfare Services Case Management System (CWS/CMS). Note: The counts of investigated and substantiated CSE maltreatment reports are subsets of the total CSE maltreatment reports. Across counties that opted in to S. B. 855 funding, CSE maltreatment reporting peaked during FY 2017–18 before declining in FY 2021–22 (figure 7). This trend is consistent with overall child maltreatment reporting, which peaked between 2017 and 2018 and declined slightly over time until a steeper decline during the COVID-19 pandemic. Comparing early implementation (2015 through 2018) with late implementation (2019 through 2022), we observed the largest changes in CSE report responses occurred in rural counties, where the percentage of investigated and substantiated CSE reports each decreased by 11 percent. Significant changes were also observed among counties located in Northern California, where the percentage of investigated and substantiated CSE reports decreased by 11 percent and 8 percent, respectively. Moderate decreases in the share of investigated and substantiated CSE reports were also observed among Bay Area counties, while Southern California counties experienced a 4 percent increase in the percentage of substantiated CSE reports from early to late implementation. Pilot counties, counties that scored high on collaboration, and those implementing the minimum requirements of S. B. 855 all experienced a decrease in the percentage of investigated CSE reports but an increase in the percentage of substantiated reports. We have been unable to determine why these changes occurred, but they may point to the impact of S. B. 855 on CSE screening and information gathering. 48 FINDINGS FIGURE 7 CSE Maltreatment Reports over Time during S. B. 855 Implementation Count of CSE maltreatment reports by state fiscal year 12,000 10,000 8,000 6,000 4,000 2,000 0 FY 2015–16 FY 2016–17 FY 2017–18 FY 2018–19 FY 2019–20 FY 2020–21 FY 2021–22 Source: Administrative records from California's child welfare state administrative data system, Child Welfare Services Case Management System (CWS/CMS). Note: This figure represents the longitudinal trend of the total CSE maltreatment reports made between FY 2015–16 and FY 2021–22 (n = 70,334). REPORTER TYPE Figure 8 presents the roles of people who made CSE maltreatment reports to better understand the drivers of CWS interventions around CSE and points of potential support and prevention. Specifically, we examined CSE reports to document the percentage of concerns that were alleged by professionals that may contribute to S. B. 855 MDTs. Across the implementation period, we found that about one in six CSE reports (18 percent) were initiated by law enforcement personnel, a primary collaborator around CSEC, and key member of the multidisciplinary process. Fifteen percent of reports were made by counselors or therapists, and 15 percent were reported by educational professionals. Slightly less than 7 percent were reported by both medical and CWS personnel, respectively, which is surprising given the extent to which available research identifies these two types of professionals as key screeners for CSE. These shares differ from general child maltreatment reporting mainly for educational professionals and CWS personnel. Among all maltreatment categories, education professionals generally report up to 20 to 25 percent of children, a higher percentage than in this analysis, and child welfare staff generally report between 3 and 5 percent of children, a lower percentage than in this study. Between early and late implementation, there were no significant changes in reporter type among opted-in counties or counties grouped by implementation characteristics. Fourteen percent of reports were made by other professionals, and 15 percent of reports were made by anonymous reporters. While not discernable given these data, it would be interesting to learn more about these last FINDINGS 49 two categories. For example, are they composed of advocates and other professionals who support young people impacted by CSE? FIGURE 8 CSE Maltreatment Reports by Reporter Type Share of CSE maltreatment reports originating from each reporter type Law enforcement 18% Not identified 15% Counselor/therapist 15% Education 15% professional Other professional 14% Medical professional 7% CWS personnel 7% Family/friend 6% Other 4% nonprofessional Neighbor/bystander 1% Source: Administrative records from California's child welfare state administrative data system, Child Welfare Services Case Management System (CWS/CMS). Note: Share of total CSE maltreatment reports made between FY 2015–16 and FY 2021–22 (n = 70,334). In addition to substantiating maltreatment allegations, child welfare investigators conclude each CSE maltreatment report with a child disposition, which can shed some light on the safety threats and follow-up resources a child may have access to. Overwhelmingly, a majority of CSE maltreatment reports were closed with no investigation or because the situation in which CSE may have occurred was assessed and determined to be stable enough not to offer a formal case intervention (figure 9). This finding underscores the importance of community-based advocacy and supportive services. 50 FINDINGS FIGURE 9 CSE Maltreatment Reports by Report Disposition over Time Count of CSE maltreatment report by report disposition and state fiscal year No investigation Child already in open case No safety threat identified Situation stabilized Incomplete investigation/other Case opened 4,000 3,500 3,000 2,500 2,000 1,500 1,000 500 0 FY 2015–16 FY 2016–17 FY 2017–18 FY 2018–19 FY 2019–20 FY 2020–21 FY 2021–22 Source: Administrative records from California's child welfare state administrative data system, Child Welfare Services Case Management System (CWS/CMS). Note: This figure represents the longitudinal trend of the total CSE maltreatment reports made between FY 2015–16 and FY 2021–22 (n = 70,334) by report disposition. S. B. 855–ELIGIBLE COMMERCIALLY SEXUALLY EXPLOITED CHILDREN In the next section, we describe the demographic characteristics of children who were found to be at risk of or who had experienced CSE victimization. These data might be used to tailor interventions and messaging around CSE to the particular young people who may benefit from prevention efforts. Across all three methods for documenting CSE concerns (inclusive of CSE risk as well as suspected and confirmed victimization), a total of 38,168 children were eligible for S. B. 855–funded intervention or services based on CSE concerns dated between July 1, 2015, and December 31, 2021 (figure 10).23 FINDINGS 51 FIGURE 10 Annual Counts of Children with Confirmed and Unconfirmed Concerns24 CSE Victimization Confirmed Unconfirmed CSE Concerns 2,258 1,709 1,417 5,522 1,259 5,267 4,839 1,151 1,021 4,059 3,851 3,653 453 1,709 FY 2015–16 FY 2016–17 FY 2017–18 FY 2018–19 FY 2019–20 FY 2020–21 FY 2021–22 Source: Administrative records from California's child welfare state administrative data system, Child Welfare Services Case Management System (CWS/CMS). Note: Annual counts of the 38,168 children who were eligible for S. B. 855–funded intervention or services based on CSE concerns dated between July 1, 2015, and December 31, 2021. As seen in table 7, two-thirds of the S. B. 855–eligible population identified25 as female, while the remaining third had male listed as their assigned sex (less than 1 percent were intersex or had no assigned sex documented). These shares are significantly different from the demographic description of children generally reported for child maltreatment in California, which are typically divided 50–50 between young people identified as male and female. Information on sexual orientation and gender identity was only available for children involved in the CWS after 2018 and was therefore excluded from the current analysis. Slightly less than one-half of the S. B. 855–eligible child population was identified in CWS/CMS as Hispanic/Latinx (N = 19,062), while one-quarter of eligible young people were reported as White (24 percent), 16 percent of young people were identified as Black, nearly 5 percent were reported to be Asian or Pacific Islander, and less than 1 percent were documented as Native American/Alaskan Native. Six percent of S. B. 855–eligible young people were missing racial/ethnic identifiers. Compared with all children reported for maltreatment in California, these shares may indicate an overrepresentation for both White and Black young people (compared with 21 percent and 12 percent among all children reported for maltreatment, respectively) although these differences could be accounted for by the fact that among all children reported for maltreatment, approximately 15 percent were missing racial/ethnic identifiers compared with 6 percent in this analysis. A considerable majority (84 percent) 52 FINDINGS of young people spoke English as their primary language; however, 12 percent were identified as primarily Spanish speaking, and 1 percent spoke another language. The median age of children at the time of initial CSE concern was 12 years, lending some credibility to guidance provided by screening protocols that children be assessed for CSE regularly after age 10. However, some of these children may have also been the nonexploited siblings or children of young people with CSE concerns. TABLE 7 Children with CSE Concerns by Child Characteristics and Level of Certainty Child Characteristics Total Level of Certainty Not indicated CSE indicated N Col % N Col % N Col % Total 38,168 28,900 75.7% 9,268 24.3% Child race/ethnicity Hispanic/Latinx 18,510 48.5% 13,958 48.3% 4,552 49.1% White 9,222 24.2% 7,127 24.7% 2,095 22.6% Black 6,103 16.0% 4,227 14.6% 1,876 20.2% Asian/PI 1,754 4.6% 1,354 4.7% 400 4.3% NA/AN 299 0.8% 216 0.7% 83 0.9% Not listed 2,280 6.0% 2,018 7.0% 262 2.8% Primary language English 31,976 83.8% 23,968 82.9% 8,008 86.4% Spanish 4,496 11.8% 3,469 12.0% 1,027 11.1% Other 374 1.0% 311 1.1% 63 0.7% Not documented 1,322 3.5% 1,152 4.0% 170 1.8% Assigned sex Female 24,805 65.0% 18,446 63.8% 6,359 68.6% Male 13,292 34.8% 10,388 35.9% 2,904 31.3% Other/unknown * * * * * * Age at initial CSE concern Fewer than 5 years 7,556 19.8% 5,044 17.5% 2,512 27.1% 5–10 years 8,213 21.5% 6,602 22.8% 1,611 17.4% 11–13 years 6,433 16.9% 5,203 18.0% 1,230 13.3% 14–15 years 7,493 19.6% 5,674 19.6% 1,819 19.6% 16–17 years 8,473 22.2% 6,377 22.1% 2,096 22.6% Source: Administrative records from California's child welfare state administrative data system, Child Welfare Services Case Management System (CWS/CMS). Notes: Characteristics and level of certainty for the 38,168 children who were eligible for S. B. 855–funded intervention or services based on CSE concerns dated between July 1, 2015, and December 31, 2021. Asian/PI = Asian or Pacific Islander, NA/AN = Native American/Alaskan Native. Among children with a higher level of certainty-that is, CSE was investigated and confirmed not only suspected (i.e., indicated)-more than two-thirds (69 percent) of the S. B. 855–eligible population were identified as female, while 31 percent had male listed as their assigned sex. Slightly less than one- FINDINGS 53 half of the S. B. 855–eligible child population with a higher level of certainty was reported to be Hispanic/Latinx (N = 4,641), while 23 percent of such young people were reported to be White (N = 2,137), 20 percent of young people with confirmed exploitation were reported to be Black (N = 1,910), 4 percent were reported to be Asian or Pacific Islander (N = 407), and less than 1 percent were reported to be Native American/Alaskan Native (N = 85). Only 3 percent of young people with a higher level of certainty were missing racial/ethnic identifiers. Compared with all children with substantiated allegations of maltreatment in California, these shares indicate an underrepresentation of Latinx children (55 percent among all children with substantiated allegations) and an overrepresentation of both White and Black young people (22 percent and 14 percent among all children with substantiated maltreatment, respectively). Similar to the total population of S. B. 855–eligible young people, a considerable majority (86 percent) of children spoke English as their primary language, while one in nine (11 percent) primarily spoke Spanish and nearly 1 percent spoke another language. The median age of young people with a higher level of certainty at the time of initial CSE concern was also 12 years. Front-End CSEC Responses: Investigations and Substantiations The primary responsibility of the CWS is to assess child maltreatment and intervene when necessary to protect children and prevent re-abuse. With the passage of S. B. 855, the California legislature tacitly endorsed the CWS as the most appropriate agency to assess, engage, and serve young people impacted by CSE. In this section, we report on the initial responses of CWSs to CSE, including the percentage of reported children whose allegations were investigated and substantiated and whether such children experienced subsequent investigations and substantiations of abuse. Examining those identified through the maltreatment referral process only-that is, excluding children reported via the CSEC Data Grid or special project codes because there is no standardized process for investigating or confirming CSE without a maltreatment report-a total of 37,383 children26 had one or more maltreatment reports that fell within the parameters of this analysis (see "Study Population" section). For children with multiple CSE reports, the most certain report disposition was analyzed. At the child level, 68 percent (N = 25,491) of all those with CSE reports had their initial report screened in for investigation, and one in seven (16 percent, N = 5,782) had their initial report substantiated. These shares are similar to investigation rates for all children reported for maltreatment in California but are slightly higher than the share of substantiated allegations (12 percent) among all children reported in the state. Available evidence on the experience and effects of CSEC indicates that it is a process of prolonged or recurring abuse rather than isolated events, as in cases of physical abuse and severe neglect (Bruhns et al. 2018). We assessed the extent to which CSE revictimization was documented in CWS/CMS in two 54 FINDINGS ways. First, we measured subsequent CSE concerns within 12 months of an initial maltreatment report. Among the 37,383 young people with CSE maltreatment reports, one in five (23 percent, N = 8,539) had a subsequent CSE report that was screened in for investigation within a year of their initial CSE involvement, and one in twenty (5 percent, N = 1,832) had a substantiated CSE report within 12 months (appendix L). Compared with all children reported for maltreatment in California, the rate at which S. B. 855 young people experienced a second substantiated CSE report was lower than those reported for any type of maltreatment (8 percent). Among young people with CSE documented on their individual client description page, a notable majority (82 percent) of those with victimization documented had only one "victim" entry, and only 5 percent had more than two entries. That said, it is not clear that counties received clear guidance on whether they should indicate multiple victimizations on the client description page or use different methods for indicating repeat maltreatment. Further, the data grid on the child's description page was used inconsistently across counties. Based on these findings, repeat confirmation of CSE victimization does not appear to be a common practice among opted-in counties, although this finding should not be construed to mean that young people do not experience repeat maltreatment. Our findings did not reveal significant differences in revictimization across county contexts. Across county contexts, young people were far more likely to be identified as at risk of CSE than to have CSE victimization confirmed (appendix L). In line with the implementation study findings, this could be because some child welfare staff have a broad understanding of CSE risk, even going so far as to assume that any youth in a congregate care placement is at risk. The definition of victimization could also vary across counties, with some counties requiring or preferring direct child disclosure while others may rely on various forms of evidence. Large, urban counties and counties piloting the CSEC program had slightly higher percentages of victimization, which could indicate that young people in those regions were more likely to be victimized or that those counties were better at or more resourced to investigate and confirm harm than smaller, rural, or nonpilot counties. The following analysis describes the characteristics of children who had a known or suspected experience of CSE (i.e., a CSE concern) documented in their child welfare records. Children were identified through the maltreatment referral process or by caseworkers who documented CSE concerns in their records using the CSEC Data Grid and/or special project codes. Some children were identified on multiple sources. A detailed description of how CSE concerns were defined and measured is available in appendix K. FINDINGS 55 Back-End CSEC Responses: Case Involvement In addition to investigating and substantiating reports of child maltreatment, CWSs also have a duty to assess and serve young people to promote their ongoing safety, permanency, and well-being. While not required, these public agencies often open cases for children and families to formalize their assessments and support. As noted in the implementation study, child welfare staff in some counties have reported that it can be difficult to serve young people impacted by CSE when workers do not or cannot open a case and place a child in foster care. In the following section, we will describe the degree to which young people impacted by CSE27 experienced case openings before and after CSE concerns were identified. For these analyses, the study population includes 38,168 S. B. 855–eligible young people who could be tracked using their administrative data for at least 12 months following their earliest CSE concern (figure 11). FIGURE 11 Case Involvement within 12 Months of CSE Concern Share of children with CSE concerns who experienced each level of case involvement within 12 months of the CSE concern In a case at time of CSE concern Case opened immediately at time of CSE concern Case opened within 2–12 months of CSE concern No case opened within 12 months of CSE concern 14% 4% 11% 71% Source: Administrative records from California's child welfare state administrative data system, Child Welfare Services Case Management System (CWS/CMS). Notes: We document here the case involvement within 12 months of a CSE concern. These percentages represent the share of the 38,168 children who were eligible for S. B. 855–funded intervention or services based on CSE concerns dated between July 1, 2015, and December 31, 2021. CSE VICTIMIZATION TIMING Among the 38,168 S. B. 855–eligible young people (i.e., those with suspected and confirmed CSE identified via any of three methods) who could be tracked for 12 months, 14 percent (N = 5,419) were in a case at the time CSE concerns were first documented and 15 percent (N = 5,748) experienced a case opening within 12 months of the initial CSE concern, whether that was immediately (N = 1,640) or by the end of the year (N = 4,108). While it is true that 71 percent (N = 27,001) of these young people had no case opened within 12 months of the CSE concern, it should be noted that this population of S. B. 56 FINDINGS 855–eligible young people included both those with suspected (yet unconfirmed) CSE and young people who were identified as being "at risk of CSE only." Given the trauma and harm associated with CSE, and the potential for CWSs to provide protection and support, it may be important to know whether case openings happen more often, with greater expediency for young people with confirmed rather than suspected CSE victimization. Our findings indicate that compared with children with unconfirmed victimization, a greater percentage of cases were open, and the time between a CSE concern and case opening was shorter for those with confirmed victimization. Specifically, among the 9,268 young people with confirmed CSE victimization who could be tracked for one year following their initial CSE concern (figure 12), 17 percent (N = 1,554) were in a case at the time CSE concerns were first documented, and 46 percent (N = 4,225) experienced a case opening within 12 months of the initial CSE concern, either immediately after CSE identification (N = 1,327) or by the end of the year (N = 2,898). Among these children, 38 percent (N = 3,489) experienced no case opening within 12 months of CSE concern, and we cannot know with the data available to us, whether they were provided with services or support via other community providers (appendix L). FIGURE 12 Case Involvement within 12 Months of a CSE Concern (Children with Confirmed CSE only) Share of children with a confirmed CSE concern who experienced each level of case involvement within 12 months of the CSE concern In a case at time of CSE concern Case opened immediately at time of CSE concern Case opened within 2–12 months of CSE concern No case opened within 12 months of CSE concern 17% 14% 31% 38% Source: Administrative records from California's child welfare state administrative data system, Child Welfare Services Case Management System (CWS/CMS). Notes: We document here the case involvement within 12 months of a confirmed CSE concern. These percentages represent the share of the 9,268 children who were eligible for S. B. 855–funded intervention or services based on a confirmed CSE concern dated between July 1, 2015, and December 31, 2021. Approximately 70 percent of new cases were opened on behalf of young people who had confirmed CSE documented in CWS/CMS, and a majority (86.5 percent) were opened for children with no prior foster care placements. The median time to case opening is 18 days. A regression analysis of median time to case opening showed that time to be shorter among young people (1) with confirmed CSE FINDINGS 57 documented in their child welfare records; (2) served in urban counties; and (3) from counties that reported having implemented the minimum requirements of S. B. 855. In addition, the median time to case opening was negatively associated with collaboration scores. Young people from high collaboration scoring counties had their cases opened more quickly than those in counties with medium collaboration, and the median time to case opening was longest for young people in counties with low collaboration, with high versus low collaboration scores. OUT-OF-HOME PLACEMENT Foster care is one of the primary interventions the CWS offers to meet children's basic needs, protect them from harm, and provide care that parents are unable to provide on their own. In a 2017 report to the state legislature, limited availability of suitable placements for young people experiencing CSE was identified as a high priority. As such, out-of-home care is an important component of the collaborative case-planning process. To better understand the foster care involvement of S. B. 855–eligible young people, we examined the degree to which they experienced out-of-home placement and whether they were in a placement when CSE concerns were first documented. Because multiple placements can be traumatizing to children and are presumed to be more common among young people with CSE concerns, we also present data on the number of unique placements experienced by S. B. 855–eligible young people. According to our findings (table 8), of the entire group of young people meeting S. B. 855 eligibility criteria (N = 38,168), about two-thirds (N = 25,765) did not have any known placement history before or during CSE concerns. About one in thirteen (N = 2,911) had some placement history but were not in care when CSE concerns were first documented. More than one in ten (N = 4,035) were in a placement at the time of CSE identification, and an additional 2 percent (N = 580) were absent from placement but in an open episode. An additional one in ten (N = 4,070) entered foster care within 12 months of the first CSE concern. A small percentage (N = 807) of young people had multiple placement episodes during and after CSE concerns were documented. Among the 9,268 children with confirmed CSE victimization (table 8), a smaller percentage (44 percent, N = 4,029) did not have any known placement history before or during CSE concerns than among the full S. B. 855–eligible group. Four percent (N = 396) had some placement history but were not in care when CSE concerns were first documented. More than one in nine (N = 1,113) were in a placement at the time of CSE identification, and nearly 3 percent (N = 241) were absent from placement but in an open episode. An additional one in three (N= 3,162) entered foster care within 12 months of the first CSE concern. A small percentage (N= 327) of young people had multiple placement 58 FINDINGS episodes during and after CSE concerns were documented. Compared with the full S. B. 855–eligible group, there were greater percentages of children in placement during CSE concerns, absent from placement during CSE concerns, and who had a placement opened within 12 months of CSE concerns among the group of children with confirmed CSE. TABLE 8 Foster Care Activity for Young People Eligible for S. B. 855 Placement episodes for young people eligible for S. B. 855 in the year after a CSE concern Placement episodes before, during, and Young people with after CSE concern S. B. 855–eligible young people confirmed CSE N Col % N Col % All S. B. 855–eligible young people 38,168 100% 9,268 100% No placement episodes identified 25,765 67.5% 4,029 43.5% Closed placement episode only 2,911 7.6% 396 4.3% Open placement episode during CSE 4,035 10.6% 1,113 12.0% concern Absent from open placement episode 580 1.3% 241 2.6% during CSE concern Pl. episode opened (w/in 12 mo.) after CSE 4,070 10.7% 3,162 34.1% concern Multiple placement episodes during and 807 2.1% 327 3.5% after CSE concern Source: Administrative records from California's child welfare state administrative data system, Child Welfare Services Case Management System (CWS/CMS). Notes: Count and share of the 38,168 children who were eligible for S. B. 855–funded intervention or services based on CSE concerns dated between July 1, 2015, and December 31, 2021. Placement activity was observed for 12 months after the CSE concern. As with case openings, slightly more than 70 percent of new placement episodes were opened on behalf of young people who had confirmed CSE documented in CWS/CMS, and a majority (84.3 percent) were opened for children with no prior foster care placements. The median time to placement opening is 38 days. The time between an initial CSE concern and placement episode opening was shorter among young people with confirmed CSE documented in their child welfare records and among those served in urban counties. PLACEMENT TYPE The following section describes placement exposures for the 8,541 children who were in care at the time of the initial CSE concern or entered care within 12 months of the concern. Specifically, we examined predominant placement type (i.e., the setting a child resided in for at least 50 percent of a placement episode) among all children in placement within 12 months of the initial CSE concern. Given the high-level service needs of this population, not surprisingly, the largest number of young people had FINDINGS 59 group care as their predominant placement, followed by foster family agencies (FFAs) and kinship placements (table 9). Interestingly, predominant placement with kin was almost double the share of those whose primary placement type was a county foster home. This perhaps counterintuitive finding reveals the notable number of relatives caring for children with CSE concerns and points out an important potential area to bolster with service supports. TABLE 9 Predominant Placement for Young People Eligible for S. B. 855 Predominant placement settings for young people eligible for S. B. 855 in the year after a CSE concern Predominant placement type N Col % Total 8,541 100% Group care 2,273 27% Foster family agency 1,829 21% Kin placement 1,629 19% County foster home 890 10% Absent from care 712 8% Mixed placement 545 6% Juvenile hall/camp 159 2% Guardian 141 2% Non–foster care 130 2% Shelter 100 1% Supervised independent living 73 1% Other 60 1% Source: Administrative records from California's child welfare state administrative data system, Child Welfare Services Case Management System (CWS/CMS). Notes: Count and share of the 8,541 children who were in care at the time of the initial CSE concern or entered care within 12 months of the concern. Placement activity was observed for 12 months after the CSE concern. Predominant placement type is defined as the setting a child resided in for at least 50 percent of a placement episode. PLACEMENT STABILITY Poor placement stability has been identified in prior research as associated with CSE victimization, and data from the implementation study indicate challenges in finding placements that accept young people who may be experiencing CSE. We examined the number of unique foster homes or placement facilities (hereafter referred to as placements) that S. B. 855–eligible young people resided in, meaning that if a child returned to a prior placement, the home or facility would only be counted once (figure 13). Nearly one-third (N = 2,902) spent time in a single placement, one-quarter (N = 2,346) in two placements, approximately one in six (N = 1,507) in three placements, nearly one in ten (N = 869) in four placements, and one in ten spent time in five or more unique placements. Overall, the median number of unique placements these children spent time in during the year following a CSE concern was two. 60 FINDINGS FIGURE 13 Unique Number of Placements among Young People Eligible for S. B. 855 One placement 34% Two placements 28% Three placements 18% Four placements 10% Five+ placements 11% Share of children who are eligible for S. B. 855 Source: Administrative records from California's child welfare state administrative data system, Child Welfare Services Case Management System (CWS/CMS). Notes: The number of placements experienced by the 8,541 children who were in care at the time of the initial CSE concern or entered care within 12 months of the concern. Next, we compared the median number of unique placements per child based on predominant placement type (figure 14). As indicated above, young people with highly unstable placements over a yearly period appear to be the minority of S. B. 855–eligible young people. Consistent with findings on other subpopulations of child welfare–involved children, those primarily placed in kinship care (N = 1,629) had fewer unique placements than peers who were predominantly placed in FFAs (N = 1,829), foster home (N = 890), and group care (N = 2,273) settings. Those predominantly in FFA placements had fewer unique placements than others in foster home and group care settings, and young people who spent most of their time in foster home settings had fewer unique placements than those predominantly residing in group care. FINDINGS 61 FIGURE 14 Unique Number of Placements among Young People Eligible for S. B. 855 by Predominant Placement Type One placement Two placements Three placements Four placements Five+ placements 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% Kin Foster Foster Group Non–child Runaway SILP Juvenile Mixed home family welfare hall/camp agency setting Source: Administrative records from California's child welfare state administrative data system, Child Welfare Services Case Management System (CWS/CMS). Note: The number of placements experienced by the 8,541 children who were in care at the time of the initial CSE concern or entered care within 12 months of the concern. ABSENCES FROM PLACEMENT Absences from placement (i.e., "runaway" episodes) have also been identified in prior research as associated with CSE victimization. To examine this phenomenon, we examined absences from placement for S. B. 855–eligible young people (table 10). We found that slightly more than one-quarter of children with placement episodes open during the year after an initial CSE concern (N = 2,306) had at least one documented absence from placement during the 12-month observation window. Half of those children had only one documented absence, but absences were not brief; half of children were absent for 44 days. This result was inconsistent with anecdotal evidence and child welfare workers' reports. Taken together, our findings may indicate that shorter absences from placement were not documented within CWS/CMS and highlight the need for data tools to document brief absences in a manner that would facilitate tracking over time. 62 FINDINGS TABLE 8 Children with One or More Absences from Placement Total Children absent from placement % Median number of days absent All S. B. 855– eligible children 8,541 2,306 27.0% 44.3 Children with confirmed CSE 4,519 1,083 23.9% 44.0 Source: Administrative records from California's child welfare state administrative data system, Child Welfare Services Case Management System (CWS/CMS). Notes: Population is limited to children with open out-of-home placement episodes. Because of missing data, values may not add to total. Summary Over the course of S. B. 855 implementation, roughly 70,000 maltreatment reports alleging CSEC were submitted to California's child protection system throughout opted-in counties, about a quarter of which were substantiated. A majority (60 percent) were generated in Southern California, and one in ten were routed to the Central Valley. CSE maltreatment reports were largely made by law enforcement personnel (18 percent), counselors or therapists (15 percent), educational professionals (15 percent), unspecified professionals (14 percent), and unidentified reporters (15 percent), while only 7 percent of CSE maltreatment reports were generated by child welfare personnel. Slightly fewer than 40,000 minors were identified as having possible CSE victimization experiences, either via a maltreatment report or because it was documented in a child's case record, and nearly a quarter were confirmed as having been victimized. Multiple reports of victimization were uncommon, as most children had a single CSE allegation or victimization entry documented. This should not be interpreted as evidence that CSE revictimization did not occur among this population; rather, it reveals that CSE is not commonly documented as distinct events or experiences. This has important implications for contextualizing CSE victimization within broader understandings of maltreatment recurrence. Across the full population of young people with identified CSE concerns, 8 percent had prior placement history but were not in care when CSE concerns were first documented, 11 percent were in a placement at the time of CSE identification, and an additional 1 percent were absent from placement but in an open episode. Among young people who received child welfare interventions during the year following an initial CSE concern, cases and placement episodes were opened more quickly among those with confirmed CSE documented in their child welfare records and in urban counties. FINDINGS 63 Of those in out-of-home care, a majority were primarily placed in congregate care settings; however, almost 19 percent spent a majority of the year following a CSE concern placed with relative caregivers. Both case and placement involvement were more common among young people with confirmed victimization. Our system performance study captured the magnitude of CSE concerns and affected clients among counties that opted in to S. B. 855's CSEC program. However, available administrative records did not enable us to examine the provision of CSE-specific services. This lack of information is important to note because nearly three-quarters of all young people with CSE concerns and more than one-third of those with confirmed experiences of CSE victimization were not in child welfare cases at the time or during the year following a CSE. Future, more comprehensive examinations of CSE will therefore need to access and analyze the identification of service needs, provision, intensity, and duration of services for this population. 64 FINDINGS Discussion Limitations We collected data from available sources, including annual county plans, a survey of the CSEC coordinator or other point of contact in each county, interviews with program staff, focus groups with adults who experienced CSE as children, and data entered into the state child welfare administrative data system. Each data source provided information that we triangulated with other data wherever possible. Notably, none of the sources alone or together provide complete information. The largest caveat to this study is that the absence of information is not evidence of the absence of CSE or counties' efforts to address it. Further, each data source has its own limitations. Annual county CSEC program plans are completed by a single person from each county in response to a state mandate. These documents may not fully reflect the entirety of a county's plans or activities. Similarly, the county surveys were typically completed by a single respondent in each county. Although we asked that it be completed by a person with primary responsibility for coordinating the county's CSEC program, answers may reflect that person's singular perspective. The survey also represents implementation at a single point in time. We asked about prior activity, but the accuracy of the responses could be low if the respondent did not have or was unable to gather institutional knowledge about the past, especially if they were new to the coordinator role. We conducted 12 county case studies to gain a deeper understanding of implementation and collaboration. We selected counties of differing sizes, locations, populations, and levels of CSEC program implementation to try to represent the major sources of program variation across California. However, 12 counties cannot fully represent the experiences of all 47 implementing counties. And while we were able to interview 132 stakeholders from child welfare and collaborating agencies, we made multiple attempts to engage organizations that would connect us to adults who experienced CSE as children, but in the end we were only able to gain the perspectives of 12 adults with CSE experience through focus groups. The outcome study findings informed by administrative records were limited by the data available, which have known limitations with quality and completeness. CSE documentation became mandatory shortly before the study period and training for the child welfare workforce on screening, investigating, and documenting CSE victimization were under development during S. B. 855 implementation. As a DISCUSSION 65 result, some of the documented risk and victimization described in the study may reflect false positives or negatives. Summary statistics included in this report should not be referenced as prevalence estimates of CSE. Rather, they reflect county child welfare agencies' awareness of CSE and performance at screening and documenting CSEC. Further, diagnostic criteria established by the legislation to define "at risk" regarding CSE were broad and, in some cases, subjective. In addition, we could not detect any brief absences from care when an out-of-home placement remained open on the child's record. Despite this, information provided about absences that trigger placement closures are relevant to placement policy and expensing. Arguably the most important limitation within this study was the lack of administrative records on specialized CSE-related service receipt, which could be used to determine how and when young people were being offered those interventions. Respondents were asked about the extent to which their county offers specialized CSE interventions or referrals, and administrative records of these interventions would have allowed us to corroborate survey responses. Without data on specialized service receipt, we are unable to determine whether or not delivery of those interventions differed among those with and without formal child welfare involvement. This limitation also prohibited us from identifying promising upstream prevention efforts. In addition, outcome study findings comparing early and late implementation may have been impacted by the COVID-19 pandemic during which child welfare agencies across the state saw a decrease in the number and rate of children referred to, entering, and remaining in care. At the child level, several limitations to the outcome study emerge. First, the data may include duplicate child records if, for example, a child reentered the system after being adopted; however, most children are only documented once within CWS/CMS. Second, the racial, ethnic, gender, and other identity indicators included in the study were reported as documented in CWS/CMS by the CWS workforce. Policy requires that workers update these fields to reflect the self-reported responses of young people and their families, but the options are limited to the categories available in CWS/CMS. Additionally, we could not include sexual orientation and gender identities in the analysis because the CWS only began to collect this information during the latter half of S. B. 855 implementation.28 Finally, CSE policies differ by state, so our findings may not be generalizable beyond California. However, our approach sought to capture the wide variation in geographic and social contexts in which CSEC occurs. As an early adopter of diversion and treatment responses to CSE, California's experience can hopefully inform efforts in other states and territories. 66 DISCUSSION Key Findings and Recommendations The evaluation approach for California's CSEC program focused on the components of the program that can we directly tie to S. B. 855 implementation and documentation of CSE. This is critical for understanding how the legislation impacted case coordination and service provision, but other contextual factors also need to be considered. Below we summarize key findings and recommendations and consider how other policies, local operating context, and capacity constraints interact with efforts to implement S. B. 855. Key Findings from the Implementation Study The implementation study assessed the extent to which S. B. 855's core components were being implemented across California, examined the variation in how the policy was being applied across participating counties, and identified promising practices and major challenges. Our key findings highlight the importance of effective collaboration, training, staffing, data management, and system coordination to better serve and support young people at risk of or experiencing CSE.  Implementing S. B. 855 has fostered strong interagency collaboration and communication. Interagency collaboration has improved following S. B. 855, and those we spoke with reported positive relationships among agencies engaged in the county's CSEC response. In most counties, agencies other than child welfare services also screen for young people at risk of or experiencing CSE. Key informants cited progress toward working closely across agencies that had previously been siloed, and many counties were successfully incorporating continuous communication. However, collaboration is uneven. Approximately one-third of counties indicated that collaborating agencies are not in full agreement on the success of their efforts to work together, and the same share expressed dissatisfaction with interagency communication. Collaboration and case planning cannot be effective if the county does not have the capacity and resources to adequately serve young people. Some examples of how this has played out include the following: » Counties commonly reported that certain agencies may officially be part of the interagency protocol and participate in collaborative activities but have limited service slots or frequent turnover, making it hard for them to provide consistent, effective services that are in high demand for this population-such as substance abuse and mental health treatment. » Steering committees have helped agencies plan and launch county CSEC programs, but the goals and utility of these groups have shifted over time. Some counties have evolved to DISCUSSION 67 continually focus on the next stage of addressing CSEC in their county, while others have grown stagnant and are currently lacking direction. » MOUs codify intent but do not make data sharing happen. They often do not solve the technical or legal challenges of sharing data across systems. They also represent agreement among leadership, not necessarily those on the ground who are meant to share information for case collaboration-reflecting both inadequate communication as well as misunderstandings about information-sharing barriers.  A shortage of placements exacerbates CSE service challenges. The shortage of placements appropriate for young people experiencing or at risk of CSE came up repeatedly. Many foster caregivers are unwilling or unready to take young people experiencing CSE, and this constrains staff members' ability to respond to young people's needs and provide them a safe and stable living situation. Staff spend significant time searching for appropriate placements that are accepting of these young people and will not put them at elevated risk. Children experiencing CSE are a small fraction of California's CWS, but the problem tends to affect teens, a population already experiencing a dearth of services and placements, further squeezing options for all teens in the system. A related policy issue that has complicated responses to children experiencing or at risk of CSE pertains to changes in the use of congregate care under California's CCR and FFPSA. Even though CCR invested in placement alternatives, the demand for appropriate placement settings far exceeds what is available.  Staff shortages are ubiquitous and negatively impact how well counties can serve and protect young people at risk of or experiencing CSE. Staffing capacity was a challenge in almost all counties, with shortages and turnover being common problems. Staff turnover reduces trust between children, families, and county agencies; fragments ongoing training efforts; and chips away at institutional knowledge about CSE. Moreover, counties reported that frequent staff turnover complicated training efforts making it difficult to keep the ever-revolving staff supporting this population trained and ready to effectively serve young people impacted by CSE.  Standardized training on CSEC is widespread, but not enough training is available for parents, relative caregivers, and foster parents. Training should also be updated to align with current, ever-evolving CSE recruitment trends. Standardized CSEC 101 and 102 training is widespread within child welfare agencies and generally well regarded. Training is also typically offered to, or required of, probation and other partner agencies. Sometimes training is offered to the broader community and often in schools. Training for parents, relative caregivers, and foster parents is less common, and many people we spoke with said this is a major gap, noting that 68 DISCUSSION caregivers need to know the signs of CSE and how to proceed if they are concerned. For example, we heard from several counties that cyber recruitment is increasing.  The Commercial Sexual Exploitation-Identification Tool (CSE-IT) is regularly implemented but not universally. Because S. B. 855 does not mandate a specific screening tool, some counties use other or additional assessments to screen for CSE, typically either in child welfare or probation agencies. We heard from some county staff that the tool is not always used as effectively as it could be-specifically, screening may happen but findings may not be shared with the staff that are responsible for identifying young people at risk of CSE or deploying resources for them. In addition, there are concerns that the CSE-IT is overly sensitive and that the definition of "at risk" is too broad-identifying essentially everyone in foster care as at risk without considering the specifics of their situation. On the other hand, some social workers suggested that anyone in congregate care should be considered at risk.  S. B. 855 gave child welfare agencies responsibility for caring for this population, but many feel they have inadequate tools to be successful and sometimes feel undermined by other agency priorities. Child welfare staff said they are not given enough support from partner agencies or the necessary tools to fulfill their mandate to effectively help young people experiencing CSE. Child welfare staff bear the primary responsibility for the safety and care of these children but expressed concern that their mandates sometimes conflict with other stakeholders. For instance, some counties are dual-status counties (i.e., a youth can be under the jurisdiction of both the child welfare agency and probation) and some are not (i.e., the county has to pick one agency to be the main one responsible for the youth). This can facilitate collaboration, but also sometimes disagreements and territorialism. The lack of a shared agenda can undermine interagency collaboration.  Multidisciplinary Teaming (MDT) approaches are inconsistent, but the consequences of uneven implementation appear to be minimal. Counties vary in how they understand and use an MDT process to support young people experiencing CSE. In some cases, MDTs and child and family team meetings are conflated. In other cases, MDTs and steering committee meetings are interchangeable. In both situations, staff feel that the teaming required to coordinate resources and collaborate on decisions is largely taking place effectively. The act of bringing collaborators together regularly appears to make the difference for collaboration, not the name, structure, or specific implementation of the MDT approach.  Counties would benefit from greater intercounty service coordination. The collaboration required under S. B. 855 is narrow, referring only to specific agencies within a county. But some DISCUSSION 69 small (and a few large) counties collaborate across county lines where services are not available in county or where there is significant geographic mobility among young people. Some counties noted that key information was often lost when a child welfare case was transferred between counties. State support and encouragement for broader collaboration across more in-county systems and across county lines can better serve this population of young people who may have complex needs from any number of systems and may be placed out of county.  Improved timely and accurate data collection, cross-county data sharing, and standardized measures to track the CSEC population would enhance counties' responses. Many counties reported that data tracking contributed to interagency collaboration, served as helpful baseline information for agencies to begin discussing a case, and assisted with faster referrals and specialized placements. There has been progress and success in collecting and sharing data among agencies, but there is still room for improvement in accurately and comprehensively identifying and serving young people who are at risk of or experiencing CSE. » We observed that uneven knowledge about inputting and accessing CSEC data is a barrier to collaboration. When staff do not know how to access or update case information in CWS/CMS, partner agencies cannot stay abreast of case decisions or a youth's service use. Staff turnover across county agencies, staff confusion on best practices for managing and entering data, and problems sharing case information across agencies were identified as challenges.  It is challenging to serve young people experiencing CSE who are not formally involved with the child welfare system. Child welfare serves as the backbone support agency for system-involved young people, but many counties did not have a clear process for serving them without an open child welfare case, nor a clear understanding of roles and responsibilities for which agency has oversight for these young people. Many key informants felt that this was a major gap in a county's ability to effectively respond to CSE. S. B. 855 gives the child welfare agency jurisdiction to address this population, but funding rules limit their ability to support young people without court involvement. Young people and families in voluntary family maintenance cases do not have access to the same funding streams for services as court ordered in-home or placement cases (e.g., Medi-Cal). Thus, agencies feel unable to provide specialized services without going to court, which lowers families' trust and engagement. Most counties only discuss child welfare–involved young people in their MDT process, which is the main venue for service planning for young people experiencing or at risk of CSE. A few counties have developed a CSEC MDT process for non–child welfare involved young people 70 DISCUSSION (e.g., Tulare, San Francisco). However, several counties expressed feeling limited in what they can do for young people who are not removed from home. Other policies such as the federal Family First Prevention Services Act (FFPSA)29 pressure child welfare to keep more children and families out of their purview. This highlights a tension among national, state, and local policy pressures. Efforts are being made to reduce the use of foster care and the CWS's funding and authority boundaries to avoid removing children whenever possible. But funding to provide services to the population of young people at risk of or experiencing CSE is tied to the CWS, particularly the federal reimbursement that agencies can claim when placing children and young people in out-of-home care under Title IV-E.  Collaboration and implementation vary by urbanicity. While some themes are common across counties regardless of size or location, differences by urbanicity play out in a few ways: » Counties that scored high on collaboration tended to be larger and urban with higher CSE counts, suggesting that places with greater resources can dedicate some of those resources to collaborative efforts and that places with higher identified levels of the CSE problem may be the most motivated to collaborate. » Smaller, more rural counties were more likely to note that some county agencies were less likely to engage with the steering committee consistently over time, even if all agencies participated at its conception. » Most urban and some suburban counties have a dedicated CSEC coordinator, who works with other county agencies and community partners to coordinate CSEC activities and monitor individual cases. In rural counties, child welfare is likely to handle CSEC program activities without any dedicated CSEC staff. » More specialized trainings (e.g., those beyond CSEC 101 or CSEC 102) were less common in rural counties. » In general, rural counties have fewer specialized services and fewer services overall. Compared with other counties, rural counties reported having more difficulty developing services, finding appropriate placements for young people experiencing CSE, and employing the workforce needed to successfully serve young people. Key Findings from the Outcome Study In the outcome study, we examined child welfare system involvement for young people after S. B. 855's implementation. We analyzed information recorded in the statewide administrative database to DISCUSSION 71 describe the child welfare system experiences of 38,168 young people who met California's definition of CSEC or were identified as being at heightened risk of experiencing CSE. We studied the identification of CSEC, documentation practices, revictimization, case involvement, and foster care placement trajectories.  The majority of CSE reports are screened in for investigation, but a minority of those investigated are substantiated. Nearly two-thirds of the 70,334 CSE reports made between July 1, 2015, and June 30, 2022, in opted-in counties were screened in for investigation. Among those, one in five were deemed inconclusive and nearly one in four reports were substantiated. The high percentages of law enforcement and school personnel represented among CSE maltreatment reporters suggests that partner agencies have benefited from education about CSE. » Slightly more than half of the CSE reports made over the full implementation period were made during early implementation. The share of screened-in reports was slightly higher during early implementation, but the share of substantiations remained consistent across implementation periods. » One in six CSE reports were initiated by law enforcement personnel; 15 percent were made by counselors or therapists and educational professionals, and 7 percent were reported by medical personnel and CWS personnel.  Young people with documented concerns of CSE are primarily female, English-speaking, and people of color. Across all three methods for documenting CSE concerns (inclusive of CSE risk as well as suspected and confirmed victimization), a total of 38,168 children were determined eligible for S. B. 855 intervention based on CSE concerns dated between July 1st, 2015, and December 31, 2022. » Almost two-thirds of the S. B. 855–eligible population was identified as female, while the remaining third had male listed as their assigned sex. » Slightly less than one-half of the eligible population was Hispanic/Latinx, while one-quarter of eligible young people were White, one in six young people were Black, almost one in twenty were Asian or Pacific Islander, and less than 1 percent were Native American/Alaskan Native. » More than 80 percent of eligible young people spoke English as their primary language. » The median age of children at the time of initial CSE concern was 12 years, regardless of whether CSE victimization was confirmed. 72 DISCUSSION  CSE reports and CSEC Data Grid entries were not being used concurrently. Highly populated counties had comparatively lower CSEC Data Grid usage but greater CPS reporting of CSE. Southern California counties used a greater percentage of CPS reports to document concerns of CSE than all other regional groups and had the lowest percentage of young people with CSEC Data Grid entries.  Repeat confirmation of CSE victimization does not appear to be a common practice among opted-in counties. Among the young people with CSE reports, one in five had a subsequent CSE report that was screened in for investigation within a year of their initial CSE report, and one in twenty children had a substantiated CSE report within 12 months of their initial investigation. » Victimization was documented more frequently among females than males. » Across racial/ethnic subpopulations, the share of children with confirmed victimization ranged from a low of 22 percent (for Asian and Pacific Islander young people) to 30 percent (among Black young people). » Large child population size was positively associated with victimization identification.  A minority of young people were in child welfare cases or placements at the time CSE concerns were identified. One in ten young people with CSE concerns were in care at the time of CSE identification, and approximately 10 percent had a placement episode opened within 12 months of an initial CSE concern. » Cases and placement episodes were opened more quickly among young people with confirmed CSE documented in their child welfare records and those who were served in urban counties. » Of those with placement episodes open during the 12 months after an initial CSE concern, one in three young people were placed in only one unique foster home or facility. » Approximately one in four young people had a documented absence from placement. A larger share of those with confirmed CSE (as opposed to the full S. B. 855–eligible population) had documented absences from care.  Congregate care was the most common predominant placement setting, followed by FFA- supervised homes and kinship care. The median number of unique homes or facilities young people experienced within 12 months of an initial CSE concern was two, and one in three young people spent time in a single placement home or facility. Those predominantly placed in group care had more placement moves than those placed primarily with either kin, a county foster home, or in a Foster Family Agency home. DISCUSSION 73  Cases and placement openings were more common among young people with confirmed CSE victimization. Among all young people with CSE concerns, about 15 percent had a case opened and one in ten entered foster care within 12 months of the concern. In contrast, nearly half of those with confirmed experiences of CSE had a case opened, and one in three entered foster care. One in four young people with confirmed CSE had a documented placement absence at some point during the 12 months that followed an initial CSE concern.  Cases were opened more quickly for young people with confirmed CSE victimization. » The time between an initial CSE concern and a case opening was also shorter among young people in urban counties, in counties that reported having implemented the minimum requirements of S. B. 855, and in counties with high versus low collaboration scores. » Among children who were in an open case when CSE was identified, cases in counties in the Bay Area were open for less time leading up to the CSE concern compared with cases in the Central Valley and Southern California. » For children who only received a case after CSE was identified, counties in Southern California had a shorter median time between the identified concern and a case opening than the Bay Area and Northern California counties. Promising Practices County staff and other stakeholders identified several strategies and approaches that have helped them implement S. B. 855 and better serve this population.  Assigning CSE cases to specific frontline workers rather than distributing them throughout the workforce was described as a helpful practice in counties that do this (e.g., Monterey, San Francisco, Tulare) and was a common wish among counties that do not. Children experiencing or at risk of CSE benefit from workers who have developed specialized knowledge about their needs. Staff recommend this staffing model even in counties where CSE is relatively uncommon, and it is not practical to have workers who only carry CSE cases.  24/7 dual response from child welfare and CSE advocates (voluntary nonprofit) when going out for investigations (e.g., Monterey, Los Angeles, San Francisco) may improve assessment of CSE and engagement with young people.  On-staff clinicians and staff dedicated to recovering missing young people may benefit county efforts. Tulare has a unique staffing structure. Within their CSE team, they have (1) a 74 DISCUSSION CSEC clinician who is a licensed therapist and serves as a mental health resource for young people experiencing CSE and (2) a missing persons liaison who carries all AWOL cases and does active searches for those young people.  Partnering with outside organizations can be effective in connecting at-risk young people who are not child welfare involved. Tulare and San Francisco have established processes for referring young people who are not CWS involved to partner organizations that are equipped to make referrals outside of CWS, hold an MDT, and connect them to services.  Weighting CSE cases more heavily when calculating caseloads acknowledges that they are more intensive and may protect against burnout. For example, a CSE case is counted as two when calculating caseloads (e.g., Sacramento).  Some large counties (e.g., Los Angeles, Sacramento, San Diego) have a trauma-informed court specifically designated to hear CSE cases, which has provided a safe and knowledgeable setting for consideration of child welfare and associated issues for this population. DISCUSSION 75 Conclusion Findings from our evaluation highlight the role that S. B. 855 has had on county efforts to identify and serve young people experiencing CSE. Overall, S. B. 855 has provided resources that have improved collaboration, supported innovation, and in some cases boosted services available to young people at risk of or experiencing CSE. To a large extent, we found that counties have forged successful and productive collaborations, have made deliberate infrastructure and practice changes, and have worked to identify and document CSE. Several contextual factors beyond the scope of the legislation, however, have posed challenges to implementing S. B. 855. For example, the legislation was not designed to address features of the commercial sex industry that drive demand for CSE and provide economic incentives for young people to participate. Other such factors-most notably staffing, placement, and service shortages-could be addressed with additional resources. The ability to provide placement homes willing, trained, and supported to care for this population stands out as one of the greatest capacity constraints that arose in our research. Placement shortages for young people involved in the CWS have been a perennial concern in California, but S. B. 855 came just before a time when federal and state policies started to put concerted pressure on the supply of homes. A host of programs in California leading up to CCR in 2017, followed by the FFPSA in 2018 sought to address the overreliance on congregate care by limiting its use and reducing the supply toward that end. The transition has been bumpy both in California and nationally (Hughes 2023) and has had specific implications for young people who are experiencing or at risk of CSE. Ongoing consideration of how to manage placement of young people at risk of and experiencing CSE would be valuable and might be coupled with efforts to ensure adequate supply of appropriate care options. Interestingly, the outcome study found that most S. B. 855–eligible young people are not placed out of home. This could be an artifact of inconsistent data practices related to documenting CSE. However, it highlights the need for services not tied to formal child welfare placement. Because of the absence of data on CSE–specialized services, we were unable to examine whether or not being under formal child welfare jurisdiction was associated with service provision. Our analysis also found that many S. B. 855– eligible young people admitted to out-of-home care are placed with kin caregivers, an observation that has not received much attention to date. Child welfare and collaborating systems should consider the unique needs of kin caregivers in their CSEC screening, training, and service planning. In particular, training could equip parents and relative care providers with the knowledge and skills to promote healthy social, emotional, and sexual development during childhood and adolescence. 76 CONCLUSION Finally, our system performance indicators reveal that, on average, a majority of young people with documented CSE risk, suspected victimization, or confirmed CSE experience few changes in their out- of-home care setting and infrequent absences from case within 12 months of a CSE concern. Although the data may not reflect short absences that do not result in the closure of the placement home or facility, our findings reveal considerable variation in placement trajectories among young people following confirmed or unconfirmed CSE. Future research on formal child welfare involvement in response to CSE should examine the timing of absences from care in relation to CSE identification and account for the competing risks of case and placement openings. Systems have struggled to increase intensive mental health services-particularly in family-based settings-as well as continuing care when young people are placed out of county. The laudable goals of S. B. 855 to better identify and serve children at risk of for or experiencing CSE attached those goals to the CWS, with the expectation that cross-system collaborations would form to ensure a "no wrong door" service continuum. In addition, limitations regarding mental health agencies serving young people who are placed in another county have added complication to this effort. Innovation and investment regarding cross-county collaboration may help ease some of the capacity and process constraints that affect ongoing mental health care service provision for this population. Inconsistencies in the use of the CSE Data Grid, exploitation allegations, and special project codes between counties document a need for additional clarification regarding CSE documentation. Unless a maltreatment report is substantiated, repeat CSE victimization cannot be monitored using the standard methods used to monitor other types of child abuse and neglect. Further, the existing data fields that capture information related to CSE provide no information about whether the youth was offered services relevant to their experiences of CSE and exposure to the commercial sex industry. Our observations about who is documented as at risk of or experiencing CSE and their outcomes are also somewhat attributable to capacity. Large counties may have been more resourced to take advantage of the legislation's funding and more motivated to participate in S. B. 855 because they simply have more children coming to their doors than smaller counties. Ongoing investments in understanding methods and approaches to documentation and supporting county efforts to that end may improve measurement of who is at risk of or experiencing CSE, when, and whether services are provided. Efforts to improve policy and practice responses to CSE should incorporate the input of people who experienced CSE in the past. Further investments should include workforce support to improve retention and more services designed to specifically and effectively engage young people in this special population, including more family-based placements where caregivers (including relatives) are well CONCLUSION 77 trained and prepared to welcome young people experiencing CSE because they are confident in the services available to them. Funding specialized in areas such as mental health and substance use treatment through the collaborating agencies is another way to respond to the increasing policy pressure to reduce the CWS's involvement with children and families while upholding S. B. 855's goal to increase collaboration across child-serving systems. 78 CONCLUSION Notes 1 Family First Prevention Services Act of 2017, H. R. 253, 115th Cong., 1st Sess. (Jan. 4, 2017), https://www.congress.gov/bill/115th-congress/house-bill/253/text. 2 "Sexual exploitation of children," Office of Juvenile Justice and Delinquency Prevention, accessed November 10, 2021, https://ojjdp.ojp.gov/programs/sexual-exploitation-children. 3 Preventing Sex Trafficking and Strengthening Families Act of 2014, Pub. L. No. 113-183, 118 Stat. 1919 (2014), https://www.congress.gov/bill/113th-congress/house-bill/4980/text. 4 "Compilation of Titles IV-B, IV-E and Related Sections of the Social Security Act," US Department of Health and Human Services, Administration for Children and Families, Children's Bureau, updated July 7, 2021, https://www.acf.hhs.gov/cb/law-regulation/compilation-related-sections-security-act; section 1114A, National Advisory Committee on the Sex Trafficking of Children and Youth in the United States, of the Social Security Act of 1935, https://www.ssa.gov/OP_Home/ssact/title11/1114A.htm. 5 Justice for Victims of Trafficking Act of 2015, Pub. L. No. 114-22, 129 Stat. 227 (2015), https://www.congress.gov/bill/114th-congress/senate-bill/178/text. 6 Adoption and Foster Care Analysis and Reporting System, Final Rule, 85 Fed. Reg. 92 (May 12, 2020), https://www.govinfo.gov/content/pkg/FR-2020-05-12/pdf/2020-09817.pdf. 7 S. B. 855, chapter 29, statutes of 2014, http://www.leginfo.ca.gov/pub/13-14/bill/sen/sb_0851- 0900/sb_855_bill_20140620_chaptered.pdf. 8 State of California Health and Human Services Agency, Department of Social Services, "Commercially Sexually Exploited Children (CSEC) Updated Model Interagency Protocol Framework," notice to All County Child Welfare Directors, All County Boards of Supervisors, All Chief Probation Officers, and All Title IV-E Agreement Tribes, November 23, 2016, https://www.cdss.ca.gov/lettersnotices/entres/getinfo/acin/2016/i-83_16.pdf. 9 This evaluation focused on minors served by the county child welfare agencies, although S. B. 855 also allows for the support of nonminor dependents. 10 A. B. 1227, Human Trafficking Prevention Education and Training Act (Nov. 7, 2017), https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=201720180AB1227&search_keywords=Hu man+Trafficking+Prevention+Education+and. 11 S. B. 794, chapter 425, statutes of 2015, https://leginfo.legislature.ca.gov/faces/billPdf.xhtml?bill_id=201520160SB794&version=20150SB79493CHP. 12 The California Child Welfare Council was established by the Child Welfare Leadership and Accountability Act of 2006 and serves as an advisory body responsible for improving the collaboration and processes of the multiple agencies and courts that serve the children in the CWS. The council is cochaired by the secretary of the California Health and Human Services Agency and the designee of the chief justice of the California Supreme Court, and membership includes state departments, county departments, nonprofit service providers, advocates, parents, and young people formerly in foster care. The council is charged with monitoring and reporting on the extent to which the agencies and courts are responsive to the needs of children in their joint care. 13 Fay Hanleybrown, John Kania, and Mark Kramer, "Channeling Change: Making Collective Impact Work," Stanford Social Innovation Review, January 26, 2012, https://ssir.org/articles/entry/channeling_change_making_collective_impact_work. NOTES 79 14 The preliminary measure of CSE cases used to select counties for site visits included only a portion of all S. B. 855–eligible young people studied in the outcome analysis because it exclusively relied on S. B. 855–dictacted CSEC Data Grid use. 15 For all categorical variables, we conducted chi-squared tests of independence to determine whether the shares observed among system outcome measures differed by county contextual factors (S. B. 855 pilot, urbanicity, region within California, level of collaboration, and whether the minimum requirements of S. B. 855 were met). These tests tell us whether system outcomes are associated with any of the various county contexts or seem to be independent. For system outcome measures that included three or more categories (e.g., region), post hoc enumeration was conducted to identify the value(s) associated with the observed differences in frequency. This was done using the Bonferroni correction for a chi-squared analysis. Mean and median values of continuous system performance measures were examined and displayed significant skewness in the distribution, indicating differences between the mean and the median. Quantile regression was then used to compare the medians of system performance outcomes by county context measures and obtain p-values and 95 percent confidence intervals. 16 "Memorandum of Understanding Template for State and Federal Commercially Sexually Exploited Children (CESC) Requirements," Child Welfare Council CSEC Action Team, 2015, https://www.courts.ca.gov/documents/BTB_23_4F_1.pdf. 17 Dual-status counties are those with existing protocols to address children who are simultaneously involved with child welfare and juvenile probation. 18 One county indicated that they had never conducted an MDT because they have never identified a youth experiencing or at risk of CSE, and one county did not respond to the survey question. 19 "commercial sexual exploitation – identification tool (cse-it)," accessed June 22, 2023, WestCoast Children's Clinic https://www.westcoastcc.org/cse-it/. 20 The CSE-IT instrument is a validated, open domain tool for use in service delivery systems that serve children and young people; however, a three-hour user training is required before use and fees are associated with its ongoing use and data access. The copyright is held by WestCoast Children's Clinic. Other standard assessment tools used in California, such as CANS and SDM, do not charge individual counties for training or technical support. 21 One CSE special project code created by CDSS (that identifies CSE maltreatment reports) remains active. Additional CSE-related special project codes were created by county agencies and remain active. 22 Documentation protocols relevant to reporting requirements established by S. B. 855 were disseminated on May 28, 2015, in All County Letter (ACL) No. 15-49 and directed counties to identify CSEC in maltreatment reports and establish the conditions in which CSEC victimization and risk should be documented. Amended guidelines were shared with counties on May 23, 2016, as part of ACL No. 16-49. This update described changes to CWS/CMS, including the addition of the CSEC Data Grid, and further clarified the conditions in which CSEC victimization and risk should be documented. 23 To observe system involvement within the 12 months after a CSE concern was identified, we excluded 3 percent of the study population (N = 1,193) whose CSE concerns were documented within the final year of our period under review. As a result, this table reflects censoring of data for 2022. 24 Confirmed CSE indicates that a CSE maltreatment report was substantiated or that a "victim" entry was listed on the child's client page. Unconfirmed CSE concerns include documented CSE risk and CSE maltreatment reports that were screened in for investigation but were not substantiated. 25 Although the child welfare workforce in California has been encouraged to have meaningful conversations with young people about their gender and racial/ethnic identity, research literature commonly documents that such data are often based on the perceptions of child welfare professionals and may be misclassified (Jawetz et al. 2021). 80 NOTES 26 The analysis of children's initial CSE maltreatment reports included 37,383 children, some of whom only ever received an evaluated-out CSE maltreatment report during the course of S. B. 855 implementation. Those children were not included in child-level analyses of CSE confirmation, case trajectories, and placement experiences because the CWS did not document any known or suspected CSE victimization on their behalf. 27 In this case, "young people impacted by CSE" means their CPS reports allege CSE and/or "at risk" or "victim" indicators in the CSEC Data Grid and/or a CSEC special project code (see appendix K) attached to their case. 28 Per CDSS ACL 19-20 (March 13, 2019): State of California Health and Human Services Agency, Department of Social Services, "Documentation of Sexual Orientation and Gender Identity (SOGIE) in the Child Welfare Services/Case Management System (CWS/CMS)," memorandum for all county welfare directors, chief probation officers, independent living program managers, independent living program coordinators, foster care managers, Title IV-E agreement tribes, and transitional housing coordinators, March 13, 2019, https://www.cdss.ca.gov/Portals/9/ACL/2019/19-20.pdf. 29 Family First Prevention Services Act of 2017, H. R. 253, 115th Cong., 1st Sess. (Jan. 4, 2017), https://www.congress.gov/bill/115th-congress/house-bill/253/text. NOTES 81 References Bath, Eraka P., Sarah M. Godoy, Taylor C. Morris, Ivy Hammond, Sangeeta Mondal, Saron Goitom, David Farabee, and Elizabeth S. Barnert. 2020. "A Specialty Court for US Youth Impacted by Commercial Sexual Exploitation." Child abuse & neglect 100:104041. https://doi.org/10.1016/j.chiabu.2019.104041. Bruhns, Maya E., Alicia del Prado, Jana Slezakova, Alexander J. Lapinski, Toni Li, and Becky Pizer. 2018. "Survivors' Perspectives on Recovery from Commercial Sexual Exploitation Beginning in Childhood." The Counseling Psychologist 46 (4): 413–55. https://doi.org/10.1177/0011000018777544. Burt, Martha R., and Brooke E. Spellman. 2007. "Changing Homeless and Mainstream Service Systems: Essential Approaches to Ending Homelessness." In 2007 National Symposium on Homelessness Research, 2-1–2-46. Cook, Mekeila C., PhuongThao D. Le, and Jennifer J. García. 2022. Addressing Racism in the Domestic Minor Sex Trafficking of Black Girls: The Role of Public Health Critical Race Praxis. Public Health Reports 137 (1-suppl.): 10S–16S. https://doi.org/10.1177/00333549211058735. De Vries, Ieke, and Kelly E. Goggin. 2020. "The Impact of Childhood Abuse on the Commercial Sexual Exploitation of Youth: A Systematic Review and Meta-Analysis." Trauma, Violence, & Abuse 21 (5): 886–903. https://doi.org/10.1177/1524838018801332. Finigan-Carr, Nadine M., Melissa H. Johnson, Michael D. Pullmann, C. Joy Stewart, and Anne E. Fromknecht. 2019. "A Traumagenic Social Ecological Framework for Understanding and Intervening with Sex Trafficked Children and Youth." Child and adolescent social work journal 36 (1): 49–63. Franchino-Olsen, Hannahbeth, Brittney R. Chesworth, Colleen Boyle, Cynthia F. Rizo, Sandra L. Martin, Brooke Jordan, Rebecca J. Macy, and Lily Stevens. 2022. "The Prevalence of Sex Trafficking of Children and Adolescents in the United States: A Scoping Review." Trauma, Violence, & Abuse 23 (1): 182–95. https://doi.org/10.1177/1524838020933873. Gibbs, Deborah A., Alana M. Henninger, Stephen J. Tueller, and Marianne N. Kluckman. 2018. "Human Trafficking and the Child Welfare Population in Florida." Children and Youth Services Review 88:1–10. https://doi.org/10.1016/j.childyouth.2018.02.045. Havlicek, Judy, Shannon Huston, Seth Boughton, and Saijun Zhang. 2016. "Human Trafficking of Children in Illinois: Prevalence and Characteristics." Children and Youth Services Review 69:127–35. Hickle, Kristine, and Dominique Roe-Sepowitz. 2018. "Adversity and Intervention Needs among Girls in Residential Care with Experiences of Commercial Sexual Exploitation." Children and Youth Services Review 93:17–23. https://doi.org/10.1016/j.childyouth.2018.06.043. Hughes, Sean. 2023. "Why Foster Children Are Sleeping in Office and What We Can Do about It." Washington, DC: American Enterprise Institute. Jawetz, Ali, Lisa Mishraky-Javier, and Rachel Hand. 2021. "Our Identities, Ourselves: An Anti-Racist Review on Collecting Accurate Data on Race and Ethnicity." Washington, DC: Center for the Study of Social Policy. Kania, John, and Mark Kramer. 2011. "Collective Impact." Stanford Social Innovation Review 9 (1): 36–41. Latzman, Natasha E., Deborah A. Gibbs, Rose Feinberg, Marianne N. Kluckman, and Sue Aboul-Hosn. 2019. "Human Trafficking Victimization Among Youth Who Run Away from Foster Care." Children and Youth Services Review 98:113–24. https://doi.org/10.1016/j.childyouth.2018.12.022. Le, PhuongThao D., Nessa Ryan, Yael Rosenstock, and Emily Goldmann. 2018. "Health Issues Associated with Commercial Sexual Exploitation and Sex Trafficking of Children in the United States: A Systematic Review." Behavioral Medicine 44 (3): 219–33. https://doi.org/10.1080/08964289.2018.1432554. 82 REFERENCES Mitchell, Kimberly J., David Finkelhor, and Janis Wolak. 2010. "Conceptualizing Juvenile Prostitution as Child Maltreatment: Findings from the National Juvenile Prostitution Study." Child Maltreatment 15 (1): 18–36. https://doi.org/10.1177/1077559509349443. Rafferty, Yvonne. 2018. "Mental Health Services as a Vital Component of Psychosocial Recovery for Victims of Child Trafficking for Commercial Sexual Exploitation." American Journal of Orthopsychiatry 88 (3): 249–60. https://doi.org/10.1037/ort0000268. Sprang, Ginny, and Jennifer Cole. 2018. "Familial Sex Trafficking of Minors: Trafficking Conditions, Clinical Presentation, and System Involvement." Journal of Family Violence 33:185–95. Wijkman, Miriam, and Edward Kleemans. 2019. "Female Offenders of Human Trafficking and Sexual Exploitation." Crime, Law and Social Change 72:53–72. Wulczyn, Fred, Lily Alpert, Britany Orlebeke, and Jennifer Haight. 2014. "Principles, Language, and Shared Meaning: Toward a Common Understanding of CQI in Child Welfare." Chicago: Chapin Hall at the University of Chicago; Center for State Child Welfare Data. REFERENCES 83 About the Authors Ivy Hammond, MSW, PhD, is a research associate for the California Child Welfare Indicators Project. Her research focuses on child protection and maltreatment prevention efforts, with an emphasis on CSEC. Before pursuing a doctorate, she worked for LA County Department of Children and Family Services as an emergency response worker. Joseph Magruder is a research specialist for the California Child Welfare Indicators Project. He is a social work researcher who specializes is analyzing longitudinal administrative data to understand children's experience in the CWS. He is especially interested in the long-term effects of child welfare decisions on children's lives. Joe joined the Child Welfare Indicators Project after working as a child welfare program specialist for the California Department of Social Services for several decades. Wendy Wiegmann, MSW, PhD, is the project director for the California Child Welfare Indicators Project. She has at least 20 years of experience analyzing linked and unlinked data to examine child welfare practice. Wendy draws on her experience as a former foster youth, relative caregiver, and child welfare worker to inform her research. Daniel Webster, MSW, PhD, is principal investigator of the California Child Welfare Indicators Project. He has more than two decades of experience analyzing longitudinal administrative data and providing technical assistance to promote system improvement for public child welfare agencies in California and in jurisdictions across the country. Bridgette Lery, MSW, PhD, is a senior fellow at the Urban Institute. She has spent the past two decades working within or on behalf of child welfare agencies to understand the implementation and impact of their programs and policies and to make use of their administrative data for better decisionmaking. Sarah Benatar is a principal research associate in the Health Policy Center at the Urban Institute. Her research is focused on how public policy and program design affect program enrollment, access to care, use of services, and the health and well-being of underserved or inequitably cared for communities. Jaclyn Chambers is a research associate in the Center on Labor, Human Services, and Population at the Urban Institute. Her research focuses on child welfare, with emphasis on program evaluation and 84 ABOUT THE AUTHORS administrative data analysis. She holds an MSW from New York University and a PhD in social welfare from the University of California, Berkeley. Laura Packard Tucker is a senior research associate in the Center on Labor, Human Services, and Population. Her research focuses on issues related to child welfare, with an emphasis on fiscal analysis, program evaluations, and transition-age young adults. Packard Tucker holds an MS in financial analysis from Portland State University. Katrina Brewsaugh, MSW, PhD, is a senior research associate in the Center on Labor, Human Services, and Population. Her research uses both quantitative and qualitative methods to investigate the impact and implementation of child welfare programs and policies. Annelise Loveless is a research analyst in the Center on Labor, Human Services, and Population. Her research focuses on evaluating the impact and implementation of child welfare programs. She holds a BS in psychology and minor in social welfare from Seattle University. Jonah Norwitt is a research assistant in the Center on Labor, Human Services, and Population. He works on a variety of child care, early education, and child welfare projects. Before joining Urban, Norwitt worked as a research assistant at the University of Chicago Crown Family School of Social Work, Policy, and Practice, where his research focused on investigating racial disparities in welfare programs. He holds a BA in sociology and political science from the University of Chicago. ABOUT THE AUTHORS 85 STATEMENT OF INDEPENDENCE The Urban Institute strives to meet the highest standards of integrity and quality in its research and analyses and in the evidence-based policy recommendations offered by its researchers and experts. We believe that operating consistent with the values of independence, rigor, and transparency is essential to maintaining those standards. As an organization, the Urban Institute does not take positions on issues, but it does empower and support its experts in sharing their own evidence-based views and policy recommendations that have been shaped by scholarship. Funders do not determine our research findings or the insights and recommendations of our experts. Urban scholars and experts are expected to be objective and follow the evidence wherever it may lead. 500 L'Enfant Plaza SW Washington, DC 20024 www.urban.org