In this investigation of the relationship between marital disruption and children's health, child health status is modelled as a function of three sets of determinants: (1) the level of parental investment in the child (provision of food, shelter, supervision, and medical care); (2) the direct effects of the child's environment on his/her health (exposure to stress, infection or physical hazards); and (3) the child's inherent propensity for poor health (frailty). Data from the Child Health Supplement to the (U.S.) 1981 National Health Interview Survey show a limited association of illness with family disruption among children under twelve, an association that changed over time in ways that vary with maternal education. The importance of the level of family health investments (which may change with available resources) and of direct depletions to health, and the possible selection of frail children into divorce, are explored as explanations for the divorce-illness link.
Children who were preschoolers (aged 1-5) in the 1966-1970 period, whose mothers had not attended college, and who had experienced a parental separation, had rates of illness significantly higher than the rates for similar children who were in intact marriages in the same period. This health deficit among preschoolers of disrupted families shrank but was still present in the 1971-1975 period, but was not seen in the most recent (1976-1980) period.
The health deficit associated with disruption displayed the opposite trend over time for children of college-educated women. In the most recent (1976-1980) period, disrupted children had significantly more illness than children in intact families, whereas in earlier years disrupted children had only elevated risks of illness (an insignificant difference).
These findings suggest that the expansion between 1965 and 1975 of income assistance and health programs for single-parent, low-income families (specifically the introduction of Medicaid), significantly improved the health of children of divorced families. It appears that where divorce detrimentally affects children's health, it does so through the loss in income and parental time that follow divorce, which leads to less investment in children's health. Neither the stress of divorce nor selectivity into divorce account for the association of disruption with childhood illness.