Predictors of Reunification Among Child Welfare System-Involved Children of Incarcerated Parents

Background: The objective of the study is to examine factors related to reunification among children in foster care for whom at least one parent is incarcerated. Main effects including reason for removal and race/ethnicity were examined for impact on reunification. Selected interactions between reason for removal and race/ethnicity were also examined. Methods: The sample was comprised of children represented in the AFCARS files from 2004-2006, filtered to exclude children who did not have parental incarceration listed as one reason for removal and children with discharge reasons other than reunification, adoption, living with relatives, guardianship, and emancipation. The final sample was comprised of 47,470 children. Control variables included child age, gender, any diagnosis, and number of times removed. Predictor variables included race/ethnicity and reason for removal. Hierarchical linear modeling was used to take into account the higher-order organizational structure represented by each state as a unique policy entity. Results: Results indicated significant intra-class correlation, suggesting that a significant portion of the variance is at the state level. Approximately 9% of the total variance was at the state level for this outcome. Thus, the structure of the model was retained. Child's diagnosis was predictive of whether reunification would occur. Children without diagnoses were 1.82 times more likely to be reunified. Moreover, children with a disability or behavior problem listed as a reason for removal were less likely to be reunified. Additional reasons for removal were also predictive of reunification. Race/ethnicity was predictive of reunification in that Hispanic children were 1.22 times more likely to be reunified than African American children in this population. Several interactions between reason for removal and race/ethnicity were found. African American children for whom parental drug abuse was an identified reason for removal were less likely to be reunified than White children. Specifically, White children were 1.51 times more likely to be reunified than African American children when parental drug abuse was a listed reason for removal. Implications: For the population of children in foster care for whom parental incarceration was listed as a reason for removal, both child characteristics and reasons for removal were predictive of reunification. Identifying who, among this population is at heightened risk of failure to achieve reunification is necessary in order to ensure that these children are the focus of greater policy and programmatic attention. Children of incarcerated parents may be at greater risk for emotional, behavioral, or developmental problems. The finding that children in this population who have clinical diagnoses, disabilities, and behavior problems are less likely to achieve reunification suggests that timely screening, referral, and service delivery for this population may have the potential to positively impact permanency trajectories. Moreover, the finding that children of different races experience different permanency outcomes when parental drug use is a reason for removal suggests a disparate process may be in place at some level of the decision-making hierarchy. Additional exploration of the reasons for the disparity in outcome is warranted in order to ensure equitable treatment for all children in foster care.