Standardizing Discretion in Child Protective Service Work: Does Structured Decision-Making Make a Difference?
Since the mid-1960s, researchers have noted that there is little shared understanding among child welfare workers about what constitutes a case of neglect that necessitates state intervention and what criteria should be used to initiate legal proceedings against a family. To address issues of subjectivity and variability in decision-making, policymakers have introduced the Structured Decision Making Model (SDM) in over 30 states to guide action in child welfare cases. The use of SDM is intended to create objective measures that yield identical results for similar situations.
While SDM has been lauded for replacing subjective processes of clinical discretion with a system that increases the consistency and validity of decision-making, there has been little research on how, and to what extent, these tools are used in practice. This study explores how SDM is used in routine practice in two states to better understand if it increases consistency in child welfare decision-making. Further, this research examines whether SDM controls for subjectivity or encodes subjectivity in new ways.
Methods:
Data from this study draws from participant observation and 66 semi-structured interviews with CPS workers in Michigan and Connecticut. Interviews were audio-taped and transcribed with analysis conducted in accordance with the principles of grounded theory. Using Nvivo 10, interviews were coded thematically with specific attention paid to how CPS workers incorporated SDM into their everyday work routines, negotiated issues of authority and discretion posed by the SDM model, and ultimately decided which cases warranted legal or social service intervention for families.
Findings:
Challenging claims that SDM tools are neutral objects, the findings suggest that the use of SDM is both socially situated and relationally contingent on the characteristics of the child welfare worker and the larger role that SDM plays in the organizational environment. For example, in environments where SDM tools were perceived to drive decision-making, CPS workers approached these tools in ways that reflected their social position. Workers with over 10 years of experience and white male workers reported that they “manipulated the form” to ensure that the SDM tool’s final score reflected a decision they had already made, while African-American and female workers reported that they spent hours filling out the tools to make sure numerical scores were tightly correlated with the evidence. In environments where workers perceived themselves to have more discretion, differences based on social categories disappeared and workers reported that they approached the tools similarly. These workers also reported that the SDM tools recommendations did not strongly factor into their decision-making, calling into question whether the tools met their stated objectives.
Implications:
Results raise questions about the ability of SDM tools to eliminate subjective assessments in child welfare cases and highlight the role that organizational context has in the implementation of interventions. Child Protective Service organizations must carefully consider how cultural, relational, demographic, and organizational factors may impact the adoption and implementation of standardized decision-making tools.