Parental divorce and outcomes for children: evidence and interpretation

Using longitudinal data from the UK National Child Development Study (NCDS) 1958 Birth Cohort), we examine the association, net of pre-disruption background factors, between the experience of family disruption between ages 7 and 16 and later experience of a range of outcomes usually considered adverse: early school leaving, leaving home early, teenage first partnership, early entry into parenthood, and extra-marital fertility. Logistic regression analyses indicate that some of these outcomes are more frequent among the children of disrupted families are not distinctive: except for leaving school at the minimum age, characteristic of a majority of all family types in this cohort, all of the outcomes examined here are experienced by a minority of children in all family groups. Furthermore, the associations observed may not be interpreted as indicating that family disruption causes a greater risk of the outcomes studies. The NCDs, though large and rich in detail compared with many other data sources, is nevertheless incomplete in crucial respects, and so many key alternative hypotheses cannot be tested. As a result, a selective origin to the findings cannot be ruled out. Causal inference needs to proceed with particular care in this area.