Single Fathers and the Cycle of CPS Involvement: A Longitudinal Multilevel Approach

Background and Purpose: Child maltreatment research has consistently acknowledged the role of family structure in shaping the recurrence of Child Protective Services (CPS) re-reports. While much attention has been given to single mother and married households, less is known about the unique experiences of single father households. Single fathers represent a growing demographic among primary caregivers, yet they remain underexamined in the context of CPS involvement. In the last 20 years, the growth of single fathers with a CPS investigation has almost doubled from 2.6% in 2004 to almost 5% in 2022 (DHHS, 2023), yet information about whether the needs of these families differ from other family types is virtually nonexistent. Understanding the distinct risk factors and household dynamics associated with single fathers is essential for developing accurate risk assessments and prevention strategies. This study compares National Child Abuse and Neglect Data System (NCANDS) records from 2004 to 2023 asking: (1) what family and child-level factors predict re-report to Child Protective Services (CPS) over time, and how does this relationship differ by family structure? and (2) to what extent does single father household status moderate the relationship between prior substantiated investigations and the likelihood of re-report to CPS within 18 months?

Methods: Data were from NCANDS 2004–2023 Child Files. Using a multilevel model, the dataset included unique child-level cases nested within annual reporting periods (Level 1 = Time; Level 2 = Family and child characteristics). Level 2 predictors included family structure (married caregivers, single mother, single father), allegation type (neglect, physical abuse, sexual abuse, emotional abuse, or multiple allegations), child risk indicators (e.g., prior CPS report, behavioral issues, disabilities), and prior substantiated investigations. Control variables included child demographics (age, gender, and race/ethnicity). The primary outcome was re-report to CPS within 18 months of an initial report.

Results: MLM results show that overall re-report risk decreased modestly over time (β = -0.012, p < .01). However, single father households consistently demonstrated higher baseline risk of re-report compared to married caregivers (γ = 0.38, p < .001) and single mother households (γ = 0.21, p < .05), with no significant time interaction, indicating that disparities remained stable across two decades. Among single father households, multiple allegations (γ = 0.42, p < .001) and prior substantiated reports (γ = 0.55, p < .001) were the strongest predictors of increased re-report risk. Physical abuse allegations had stronger associations with future involvement (γ = 0.28) as did homes with children with disabilities (γ = 0.31, p < .01). Mediation analysis indicated that prior substantiated reports partially mediated the relationship between single father status and re-report, accounting for 32% of the effect.

Conclusion and Implications: These findings emphasize the enduring risk associated with single father households in CPS involvement trajectories and highlight the cumulative impact of prior system contact. Results suggest a critical need for tailored, long-term interventions that consider substantiation history and family structure. Policy efforts should ensure longitudinal monitoring systems disaggregate data by caregiver type to identify family-specific risk patterns.