UC Berkeley

Child maltreatment is an issue of pressing concern in the United States. As researchers and providers work to stem the tide of the opioid epidemic, increased funding has been allocated to address the associated rise in maltreatment and foster care entry. This places pressure on child welfare providers to implement more effective strategies in maltreatment cases involving parental substance abuse and mental health problems. Implementing effective intervention and prevention strategies for households affected by mental health problems or substance abuse requires that child welfare providers understand why these households are at increased risk of penetrating the system, and what factors might mitigate this risk. Fortunately, the popularization of actuarial child welfare decision-making frameworks has resulted in a growing body of administrative data that is available to researchers interested in understanding how workers make determinations about households at each juncture in the system. In California, critical decision-making junctures include whether to investigate a report, whether a child can remain safely in the home during an investigation, and whether to open a case, among others. For households affected by parental mental health problems or substance abuse, analysis of child welfare decision-making data has the potential to expose why they are at increased risk of more serious involvement, as well as identify factors that may guard against future involvement. Such knowledge may inform early intervention and prevention strategies for these parents and their children. To that end, this dissertation examines several critical junctures in the child welfare decision-making process using administrative data collected on 4,070 households referred for the first time for maltreatment allegations in San Francisco County. Chapter 1 applies mediation analysis to an inventory of safety threats in order to account for why workers are more likely to remove children from the home during an investigation if a parent has mental health problems or substance abuse. Chapter 2 uses moderation analysis to examine which protective factors are present when workers decide that children of parents with mental health problems or substance abuse can remain in the home despite the presence of safety threats. Chapter 3 examines associations between safety decisions and likelihood of several key child welfare outcomes, including allegation substantiation, in- and out-of-home case openings, and maltreatment re-referral. Taken together, dissertation results provide a more in-depth perspective on how and why parents with mental health problems or substance abuse and their children become more seriously involved in the child welfare system, and what might prevent that involvement in the future. When parental mental health problems and substance abuse are present, workers tend to document evidence of prior maltreatment, unmet immediate needs, and prenatal exposure to substances, but not physical abuse; the combination of these safety threats puts children at increased risk of being determined unsafe in the home. This increased risk appears to be mitigated by greater numbers of protective factors, although specific factors such as child and caregiver capacity, history of problem-solving, and willingness to take protective action appear to have individual protective effects. When parental mental health problems or substance abuse are present, workers are more likely to change their safety decisions during an investigation, and are more likely to substantiate allegations and open cases, both in- and out-of-home, even among households whose children have been determined safe. However, parental mental health problems or substance abuse do not increase risk of maltreatment re-referral, an indication that child welfare intervention may reduce recurrence.