Post-Permanency Discontinuity: A Moderated Mediation Model
Methods: In 2006, 346 Illinois caregivers who finalized an adoption or guardianship between 1998 and 2002 completed surveys about child behavior problems, the adequacy of financial assistance, and thoughts about maintaining the permanency relationship. Responses were linked to administrative data that tracked continuity of care through 2012. A mediation analysis examined whether caregivers’ thoughts of ending the permanency relationship mediate the impact of child behavior problems (X) on post-permanence continuity (Y). Also, a moderated mediation analysis explored whether the observed primordial and bureaucratic factors moderated the mediation relationship.
Results: Thoughts expressed at survey time about ending the permanency relationship mediated the effect of child behavior problems on post-permanency discontinuity rates. This indirect effect was more pronounced among distantly related kin, lone caregivers, and caregivers who felt the subsidy was inadequate to cover their expenses. The odds of placement continuity are significantly lower when a child has a high problem behavior score (c = -0.061; 95%CI [-0.104, -0.170]). However, when caregiver thoughts of ending the permanency relations was added as a mediator, the results showed the direct effect of child behavioral problem on placement continuity was no longer statistically distinguishable from zero (c’ = -0.040; 95%CI [-0.087, 0.007]). We found a significant positive effect of child behavioral problems on thoughts of ending the placement (0.022; 95%CI [0.015, 0.028]). The results showed a significant indirect negative effect of child behavior problems on permanency continuity that was mediated through the caregivers’ thoughts of ending the relationship (ab = -0.015; 95%CI [-0.029, -0.003]). The main effect of subsidy adequacy also tested significant at conventional levels, but unexpectedly it was in the negative direction (-1.093; 95%CI [-1.907,-0.278). Neither the interaction terms nor the main effects for licensure or guardianship tested significant, and none of the bureaucratic agency factors moderated the influence of child behavioral problems on caregivers’ thoughts of ending the placement when each was added separately to the statistical model. The effect of child behavioral problems on caregiver’s thoughts was contingent both on distant kin relationship (0.081; 95%CI [0.055, 0.107]) and married caregiver (-0.022; 95%CI [-0.34, -0.009]).
Conclusion: Our findings suggest that post-permanency services should target a narrow segment of caregivers who express weakened permanency commitments that arise from the challenges of parenting a child with multiple behavioral problems.