The child welfare system at state and local levels faces a persistent challenge: ensuring every child in foster care finds a safe and appropriate placement that meets his or her unique needs. Despite a variety of options (kinship care, foster care, congregate care), finding the best fit for children's specific needs is challenging because of the volatility demand for these placements. This often results in inappropriate temporary placements (offices, hospitals) or increases in preventable placement disruptions. This research takes a system-wide perspective, analyzing placement data to describe bed utilization and predict future needs with greater precision. Prior research suggests increasing bed capacity can inadvertently inflate demand, particularly for congregate care. We move beyond simply adding beds and delve into the demand-supply dynamics within the system.
Methods:
Building on system science, we bring several innovative solutions to child welfare agencies to proactively monitor placement demand and then strategically allocate. First, we bypass the complexities of measuring true private agency bed capacity by focusing on daily bed utilization, creating a real-time snapshot of providers’ actual operating capacity. Second, we leverage rich longitudinal administrative data to study 10 years of daily child placements by care type, contract type, provider, geography, and child demographics, creating a detailed picture of system utilization over time. This allows us to see trends and identify gaps in specific placement and contract types needed for children. Lastly, by analyzing historical patterns, we predict future bed requirements at the specific care type level while also considering the anticipated needs of children entering the system. This empowers child welfare leaders to ensure the right resources are available for the right children at the right time.
Results:
In our study jurisdictions, we find a steady rise in foster care and kinship placement utilization, while congregate care utilization varies by level. Using a historical maximum quarterly average utilization as a set capacity, we found that annual care days for foster care beds are over utilized more than half of the time, and most severely for higher-level foster care beds. Using regression models, we forecast future care days (and corresponding bed needs) for each placement type and level in the coming years.
Conclusions and Implications:
Social policy in the US favors using as little foster care as possible along with an understanding that some foster care is inevitable. Because the supply of beds serves as a source of its own demand, our research emphasizes the importance of taking an evidence-driven, system-level approach to studying foster care capacity. By pinpointing the observed utilization and predicting future needs, the evidence now available to child welfare leaders helps them: (1) reduce inappropriate placements, maximize the likelihood of suitable placements, and minimize disruptions to ensure children are placed in safe and appropriate environments that address their individual needs and (2) allocate resources more effectively, allowing for increased spending on keeping children with families.