One of the five federal performance measures for state child support agencies is the rate of paternity establishment for nonmarital births. Fathers who voluntarily establish paternity in the hospital are more likely to be involved and supportive of their children, both of which are associated with positive child outcomes. Further, paternity establishment is a necessary precursor to opening a child support case; therefore a higher rate of in-hospital paternity establishment expedites the establishment of subsequent child support orders.

Approximately 70% of unmarried fathers establish paternity in the hospital. The paternity establishment rate, however, is highly contingent on the father’s presence at the birth. More than three-quarters of fathers are present at the birth, and of these, 90% establish paternity. Roughly 22% of fathers, however, are not present at the hospital when the opportunity to establish paternity is offered. Though absent fathers overwhelmingly fail to establish paternity voluntarily, prior research has largely neglected the father’s presence at birth as a relevant variable in paternity establishment outcomes. We propose that the failure to establish paternity involves two distinct circumstances, often conflated in past research—fathers who are absent from the birth, and fathers who are present but decline to establish paternity. A better understanding of in-hospital paternity establishment requires examining each of these in isolation. In addition, we argue that the precursors to not establishing paternity are evident long before parents’ arrival at the hospital.

Using new data collected through the Paternity Establishment Study (PES), a longitudinal birth cohort study of approximately 800 Texas mothers who gave birth outside of marriage in January 2013, we examine parents’ in-hospital paternity establishment decision conditioning on fathers’ presence at the birth. First, we provide a descriptive portrait of parents (n=754) in each of three cases: father present, paternity established; father present, not established; and father absent, not established. Second, we model the in-hospital paternity establishment decision as a multinomial logit with these three cases as the dependent variable given a vector of socioeconomic controls, the parents’ relationship status at birth, and a broad set of prenatal factors intended to capture (1) relationship commitment prior to pregnancy, (2) prenatal relationship quality, and (3) prenatal father involvement. Finally, we consider the role of operational and informational barriers to in-hospital paternity establishment.

Results suggest that the factors predicting non-establishment differ between fathers who are present and absent from the hospital. Fathers who are absent from the hospital have had limited contact with the mother, and provide her little support during the pregnancy. A fathers’ absence from the 20-week ultrasound emerges as an especially strong predictor of his absence from the birth, suggesting this standard prenatal checkup may be an optimal time to provide unaccompanied mothers with information on paternity establishment and child support. For fathers who are at the hospital, only father doubting the child’s paternity significantly predicts the failure to establish paternity. Offering paternity testing to these fathers may facilitate accurate paternity establishment. Overall, results suggest the in-hospital paternity establishment rate may be near a maximum level.