Variation in decision-making on out-of-home placements in child and family welfare services

Background: Variation across local administrative units (municipalities) is critical in child welfare, as inconsistencies in out-of-home placement decisions can create inequalities in children's access to support. Given the impact of placement decisions on a child's well-being, and long-term outcomes, ensuring fairness and consistency is essential in promoting equal opportunities for all children.

Objective and setting: This study explores municipal variation in out-of-home placement decisions following referrals in Danish child and family welfare using 52.649 out-of-sample predicted risk scores on 27.341 unique children across 94 municipalities in 2016 and 2017. While existing research primarily focuses on overall placement rates or individual placement probabilities, we extend the literature by analyzing municipal differences in referral handling beyond raw rate variations.

Methods: Using a novel approach that integrates Predictive Risk Models (PRMs) with linear regressions to control for individual predicted risk scores, we isolate municipal decision-making variation from broader socioeconomic influences. This provides a clearer understanding of how referral outcomes differ across municipalities.

Results: First, our findings reveal that significant inter-municipal differences persist after we adjust raw placement rates for variation attributable to family and child characteristics, and calendar time variation. Second, variations in socioeconomic structures influencing child risk outcomes do not fully account for these differences. Thus, it is discussed whether the remaining variation may stem from local decision-making practices, shaped by local organizational dynamics and inherent uncertainties (such as lack of information and control over future events) in child welfare assessments.

Conclusions: By emphasizing the potential role of local decision-making practices, this study advances research on decision-making and highlights the need for further investigation into how local organizational factors (such as agency policies, resource availability, work load, organizational culture and support) shape referral-placement outcomes. Implications for policy and practice are discussed.