Exploring Child Maltreatment Under and Overreporting Drivers

Background and Purpose: Child abuse and neglect (CAN) is a significant public health and social problem in the US. In 2016, approximately 676,000 children and youth were victims of child maltreatment. CAN is correlated with poor child and youth development, including increased risk of mental health disorders, suicidality, physical health problems, cognitive difficulties, drug use, and risky sexual behavior. However, across 45 states that report on referral rates, states on average “screened in” 58% of referrals, and “screened out” the remaining 42%, raising questions about whether CAN may be overreported in some cases. In fact, recent research shows that CAN may be both over and underreported, but few studies have explored this phenomenon. This paper addresses this gap by exploring community informants’ opinions about the drivers leading to CAN over and underreporting.

Methods: Snowball sampling was used to identify 30 professional community key informants. Participants ranged from 27 to 69 years old (M = 47) and 74% were female. The majority of the participants (44%) were Caucasian, 29% were African American, 24% were Latinx, and 3% were Asian. Informants identified their professional role in the community, ranging from business owner to educational administrator, with most (n=16) being a child welfare worker. Informants had an average of 18 years of experience in the community about which they were interviewed. Semi-structured qualitative interviews were conducted with community informants in Southern California. Most informants were interviewed about one specific census tract, but three were interviewed about two census tracts. Up to three interviews were conducted per census tract, resulting in 32 interviews representing 19 census tracts. Informants were not directly asked about the drivers motivating under or overreporting practices, but often described these incidents as they considered broader questions related to CAN. Data analysis was, therefore, based upon Charmaz’s grounded theory approach. Interviews were transcribed and coded thematically using NVivo.

Results: Key informants’ perceptions of CAN reporting practices offered five themes that emerged from the qualitative data, including 1) Conflict in neighborhood settings 2) Professional discretion 3) Poor CAN knowledge 4) Fear and mistrust and 5) Community norms.  

Conclusions and Implications: Whereas professional discretion and poor CAN knowledge were perceived to affect both under and overreporting practices, fear and mistrust and community norms appeared to only affect underreporting practices and conflict in neighborhood settings seemed to only affect overreporting practices. Furthermore, tracts with unusually low and high rates of CAN referrals demonstrated unique patterns. Respondents perceived that underreporting practices driven by fear and mistrust and established community norms were more likely to occur in tracts with unusually low rates of CAN referrals. Conversely, respondents perceived that overreporting practices driven by conflict in neighborhood settings and professional discretion were more likely to occur in tracts with unusually high rates of CAN referrals.

Results show that neighborhood-based family support initiatives should avoid a one size fits all approach to child abuse prevention. Strategically considering drivers can help inform targeted community interventions, such as service announcements defining CAN or implicit bias training for community informants.