Seeing Poverty: The Consequences of Economic Disadvantage for Increasing Risk on a Child Welfare Decision-Making Tool

Introduction: The Structured Decision Making (SDM) model is the most common set of decision making tools used in child welfare systems with the actuarial-based Risk Assessment (RA) being its most prominent component. The SDM was developed to estimate the prognosis for the family at different decision points in a child welfare case. The RA, for example, prognosticates future maltreatment risk and is designed to be used during the investigation process. Since the RA’s development over 25 years ago, new research has emerged on the role of financial hardship in causing child maltreatment, particularly neglect. Concurrently, the significant disproportionality between rates of child-welfare involvement for Black, White and Latinx families has become a major focus of child welfare systems. Some contend this disproportionality is a reflection of current and historic societal disparities, particularly around poverty and structurally racist macro policies that differentially impact communities and families of color. Others attribute the disproportionality to implicit bias and explicit racist treatment within child welfare systems. Identifying the actual cause of the disproportionality is critical to arriving at an effective solution. In this study, we examine the question: How are those structural factors accounted for in the SDM RA?

Method: In this conceptual paper, we utilize primary and secondary source documents on the SDM RA to understand its construction and choices made related to sensitivity and specificity in its approach. We then analyze the variables in one state’s SDM RA in order to identify how the SDM RA accounts for poverty and race in its assessment of families. Drawing on new research related to poverty as a cause of maltreatment, we theorize about the implications for how relationships between structural factors such as race and poverty are captured and defined for assessing future maltreatment risk.

Results: Variables in the SDM RA are limited to factors at the individual and family level and ignore meso-, exo- and macro-system factors. Additionally, the role of poverty or financial hardship is not explicitly included. This focus on individual-level factors misses a significant number of known factors associated with risk of maltreatment. Additionally, choices were made to favor false positives over false negatives, resulting in an inflated prognosis of future maltreatment for many families.

Conclusion: Despite the longstanding commitment to the ecological model in social work the family characteristics included in the RA are all at the individual-level and financial hardship is left out. The focus on non-financial characteristics likely results in a spurious relationship between the risk of future maltreatment and other demographic factors, such as number of children, that are correlated with poverty. The RA’s focus on individual factors also ignores the nesting of effects of structural racism and the significantly higher prevalence of poverty among families of color, resulting in an inaccurate identification of primary causes of maltreatment. Without including these critical and causal variables the risk of future maltreatment, may at best, be inaccurately assessed and at worst misdirects the focus of treatment away from causal factors, preventing effective solutions.