Examining Potential Gender Bias in the Child Welfare System: A Comparison of Father and Mother Engagement

Background: The childwelfare literature suggests that fathers are routinely not engaged to the same extent as mothers in the child welfare system. Past research finds that many fathers are not contacted by their children’s child welfare workers, are not offered services, and are not considered as potential placements for their children and suggests that child welfare worker bias against fathers underlies this phenomenon. However, the literature has yet to compare engagement of fathers to that of mothers and thus, has not yet established that there exists a gender bias against fathers. The present study seeks to address this gap by directly comparing fathers’ levels of engagement to those of mothers.

Methods: This study utilized child welfare case records for 359 fathers and 359 mothers whose children entered out-of-home care for the first time between October 2013 and September 2015 in one urban county.

Information relevant to father engagement was extracted via a systematic content analysis of two case documents. The Emergency Response Transfer Summary contains a summary of maltreatment allegations investigative findings. The Disposition Report containsa comprehensive family assessment and case plan. Engagement was operationalized as a 10-stage gradient that begins with workers attempting to identity fathers/mothers and concludes with workers considering fathers/mothers as potential placements for their children.

Multilevel linear regressions with parent type (mother vs. father) as the primary predictor were fit to determine the degree to which father engagement is the same or different from mother engagement while accounting for parent and family covariates. The covariates included resident parent status (resident vs. non-resident), perpetrator status (maltreatment perpetrator vs. non-perpetrator), prior child welfare involvement, and criminal justice involvement as well as the number of children in the family and the children’s ages.

Results: In the unadjusted model, parent type was significantly associated with level of engagement such that on average, mothers achieved two additional stages on the engagement gradient (p<0.001). However, after introducing covariates, type of parent was no longer a significant predictor; instead, resident parent status, perpetrator status, and current criminal justice involvement were significant predictors. 

Among non-resident parents, being the perpetrator of maltreatment was significantly associated with, on average, a 1-stage decrease in level of engagement (p<0.01). However, among resident parents, being the perpetrator of the maltreatment did not impact level of engagement. Being on probation or parole was, on average, associated with 0.54-stage decrease in level of engagement (p< 0.01); being incarcerated was associated with a 1.28-stage decrease in level of engagement (p< 0.001).

 Conclusions and Implications: There is a significant difference in level of engagement between fathers and mothers on average, such that mothers are engaged to a greater extent than fathers. However, this difference is attributable to parent characteristics, specifically resident parent status and criminal justice involvement and not to gender bias. Thus, a lack of father engagement is of concern but for incarcerated and non-resident fathers and efforts to address a lack of father engagement should focus on these subgroups.