Association for Public Policy Analysis & Management

That both first births and higher order births represent an increased risk for child maltreatment are often-cited findings in the child maltreatment literature. However this work has frequently relied on retrospective or cross-sectional samples of maltreated children. This paper reexamines these two well-established findings in the context of a prospective longitudinal birth cohort study.

A range of empirical and theoretical research has found that first births are at increased risk for child maltreatment (Sedlak, 2010). Theoretical models have described the ways in which the introduction of a child into the family system can result in a range of adverse parenting practices (Belsky, 1980; Cowan & Cowan, 1995). Empirical research has also demonstrated that first births, particularly to young, single, low-income mothers, increases the risk for maltreatment (Lee & George, 1999; Olds et al., 1995).

An extant body of research has also found associations between the number of children in a family and child maltreatment. Nair and colleagues (2003) describe the ways in which having multiple children increases stress and depletes resources resulting in an increased likelihood of child maltreatment. Related work has found that four or more children in a household represents a possible tipping point for child maltreatment (Fuller, 2005; Wu et al., 2004).

The Fragile Families and Child Wellbeing Study (FFCWS) is well-suited for examining associations between first birth, parity, birth order, and the risk for child maltreatment. The FFCWS is a longitudinal birth cohort study of approximately 5,000 families, including an oversample of non-marital births. Drawing on a prospective longitudinal birth cohort study is useful because it avoids many of the limitations of prior work.

I examine whether first born children are more likely to experience maternal spanking, physical aggression, and psychological aggression (collectively a risk for child maltreatment) than higher order children. Next, I examine the duration of the risk (when children are 1, 3, 5, and 9 years old) and whether it is mediated by indicators of maternal economic hardship or mental health. Last, I investigate whether this association is moderated by marital and poverty status, as prior research has indicated. Drawing on individual fixed effects models, I take up the question of whether the number of children in a family is associated with an increased risk for child maltreatment, and I examine whether the association is mediated or moderated by the same set of indicators listed above.

Preliminary results indicate that first birth is associated with an increase in the risk for maternal child maltreatment when children are 1 and 3 years old. Strikingly, I find no evidence of mediation nor does the association appear to be moderated by marital or poverty status. In contrast to prior work, I find that the number of children in a household is associated with a decrease in the risk for child maltreatment, and that each additional child represents a further decrease in the odds. Furthermore, I find no evidence that these associations are mediated or moderated by indicators of economic hardship, mental health, or marital status.