Nostalgia as family policy - emotional and economic effects of divorce on children

IN "For the Sake of the Children" [Summer 1992], Richard Gill criticizes a study several collaborators and I published on the effects of divorce on children.([dagger]) Gill first charges that our methodology and our interpretation are flawed, then launches into a broader discussion of contemporary marriage and divorce. In an earlier Public Interest article, "Day Care or Parental Care?" [Fall 1991], Gill expressed similar concerns about the direction of family policy and the well-being of American children.

Gill's concern about children is well-founded, and we share it. It was never our intent to conduct a study that would absolve divorcing parents of guilt. In our current research, in fact, we have found some harmful effects of divorce on young children who were assessed shortly after their parents' marriages broke up. Nevertheless, we think that Gill's procedural criticisms are groundless and, more important, that the underlying themes of both his articles are unhelpful as a guide to public policy concerning children and families. In this reply, I will address Gill's specific criticisms and then respond more generally to his two articles.

For our study of divorce, we analyzed statistics from two surveys, one British and one American, that had followed national samples of children and their families for several years. We identified about 12,000 seven-year-old British children and 800 seven- to eleven-year-old American children whose parents were married at the beginning of the survey period, and we tracked these children for the next four to five years. Not surprisingly, children whose parents separated or divorced displayed more behavior problems and performed more poorly in school than children whose parents remained married. But when we looked back to the beginnings of the surveys, we found that the children--particularly boys--whose parents were then married but would later divorce were already displaying more behavior problems and performing more poorly in school than the children whose parents would remain married. We concluded:

Overall, the evidence suggests that much of the effect of divorce on

children can be predicted by conditions that existed well before the

separation occurred.... [T]he British and U.S. longitudinal studies

suggest that those concerned with the effects of divorce on children

should consider reorienting their thinking. At least as much atten-

tion needs to be paid to the processes that occur in troubled, intact

families as to the trauma that children suffer after their parents

separate.

ALTHOUGH Gill agrees with our call for more attention to troubled, intact families, he disputes our conclusion that much of the harm experienced by children of divorce is due to conditions that exist before separation occurs. He notes that the British study, by far the larger of the two, followed children from 1965 to 1969, when the divorce rate was much lower than it is today. Only 11 percent of British children during that period experienced a divorce before they were sixteen, compared to about 40 percent of U.S. children today. The social climate in Great Britain, Gill argues, must have been more favorable for keeping marriages together and making them work. Gill then goes on to make the crucial assumption--without any evidence--that a high proportion of those British marriages ending in divorce must have been "very troubled." He speculates that these marriages were characterized by "physical, sexual, or psychological abuses, whether directed toward child or spouse, that not only justify but clearly mandate separation and divorce." No wonder, he concludes, that the British children were often troubled before the separation--only deeply troubled families got divorced.

But were the British marriages that ended in divorce much more troubled than American marriages that end in divorce today? Although we cannot be sure, we doubt it for two reasons. First, we deleted from our British study all families in which separation occurred before the child's seventh birthday. We did this because the interviews conducted with the parents and teachers of seven-year-olds in intact families provided us with the first detailed family information that could serve as a pre-separation benchmark. A fortuitous result is that couples whose marriages lasted fewer than seven years were excluded. In fact, we excluded more divorced couples than we retained. We think it very likely that seriously troubled, abusive marriages tend to break up more quickly than marriages with more minor sources of conflict. If so, then the families we retained are probably more like contemporary families than Gill believes.

Second, each interview with the parent of a seven-year-old was conducted by a Local Authority Health Visitor, a trained nurse employed by the municipality, who normally visited each mother before her child was born and returned afterward, often several times if the family was experiencing problems. The Health Visitor's questionnaire included a checklist that noted any family difficulties or use of social-welfare services. Relatively few problems were noted for the families that later divorced. Health Visitors checked "domestic tension" for 17 percent of later-divorced families, "financial difficulties" for 16 percent, contact with the child-welfare department for 8 percent, mental illness or neurosis for 7 percent, and alcoholism for 1 percent. To be sure, the Health Visitors may have been unaware of many problems. Still, the reports do not suggest that these families were in great conflict.