The Shift (cont.)

Among family scholars, what I called elsewhere “the shift” - put simply, from optimism to pessimism regarding the effects of family break-up on children - is becoming ever more noticeable. The latest important scholar publicly to reorient his perspective on this matter is the sociologist Andrew J. Cherlin of Johns Hopkins University, and therein hangs a modest but revealing tale.

In 1991, Cherlin and six colleagues published an influential article in the journal Science in which they sought to discredit the view, widely believed by both experts and parents in earlier generations, that children’s problems after divorce “stem mainly from the difficult adjustment children must make after their parents separate.” In particular, Cherlin and his colleagues took aim at Judith S. Wallerstein, the clinical psychologist who had argued that “the central hazard which divorce poses to the psychological health and development of children and adolescents is in the diminished or disrupted parenting which so often follows in the wake of the rupture and which can become consolidated within the post-divorce family.”

In contrast to this view, the main conclusion of the Science article is that “a substantial portion of what is usually considered the effect of divorce on children is visible before the parents separate.” Consequently, “those concerned with the effects of divorce on children should consider reorienting their thinking. At least as much attention needs to be paid to the processes that occur in troubled, intact families as to the trauma that children suffer after their parents separate.” 

The Science article presents data from two studies of children of divorce, one from Great Britain, the other from the United States. Though margin of error considerations make the numerical estimates less than completely reliable, both British and U.S. data suggest that pre-divorce family problems (what Cherlin would later call “pre-disruptive effects”) account for about half of the increased behavioral and school achievement problems experienced by boys from divorced families. For British girls from divorced homes, about one-quarter of the deterioration in school achievement (though none of the worsening behavior) was attributed by the researchers to family problems that predated the divorce. Among U.S. girls from divorced homes, Cherlin and his colleagues find that behavior actually improves after divorce, while school achievement remains about the same.

Widely discussed in both the academic and popular press, the Science article soon became, as it remains today, probably the single most important academic citation for those who believe that bad marriages are “at least,” as Cherlin implies, as harmful as divorce. Cherlin and his colleagues had apparently established an intellectual foundation for one of the most frequently repeated ideas of the family debate of the 1990s: Better for unhappy parents to divorce rather than expose their children to ongoing marital conflict and distress.

Here is how the family scholars Philip and Carolyn Cowan of the University of California at Berkeley put it in 1997 in the New York Times: “Children are at risk when their parents fight a lot - and it is this conflict, not divorce, that is so harmful to children.” Here is Arlene Skolnick touting the Science article recently in The American Prospect: “The results showed that, compared to those who remained in intact families, children whose parents had divorced in the interim did have more problems, but they had shown those problems at age 7, before the parents divorced.”

Leave aside the fact that Skolnick misreports the findings of the Science article. Cherlin and his colleagues never said, as Skolnick implies they did, that all of the problems experienced by these children of divorce had “shown” themselves prior to the divorce. Nor has Cherlin ever put forward, as the Cowans do, the simple-minded slogan that “conflict, not divorce” is what harms children. Everybody, or at least almost everybody, agrees that both things matter. The important questions for scholars are ones of relative weight and, especially, causation. And on these matters the public influence of Cherlin and his colleagues has been widespread and unmistakable: We should “reorient” our thinking about the harmful effects of divorce. Marital problems predating divorce may be at least as important as divorce itself and therefore deserve at least as much of our attention.

Throughout most of the 1990s, Cherlin remained a steadfast supporter of the proposition that, while divorce may be a problem, people who worry too much about divorce are at least as big a problem. For example, he continued his criticism of Judith Wallerstein, especially disputing the significance of what Wallerstein called divorce’s “sleeper effects”: longer-term emotional problems affecting the children of divorce that emerge only during late adolescence and early adulthood.

In 1993, Cherlin strongly attacked Richard T. Gill, an economist who had criticized the Science article, primarily on methodological grounds, in The Public Interest. Like many observers of the family debate, Gill understood Cherlin’s findings as essentially supportive of the idea, increasingly embraced by parents as well as experts, that “divorce or separation itself is not the real problem,” since “divorce is an acceptable solution to problems that might have even more deleterious effects on children if the marriage continued.” Yet for Gill, the Science article is methodologically “flawed” in several respects, thus rendering its argument “seriously incomplete.”

For example, what about that strange finding that, in the U.S., but apparently not in Britain, parental divorce improves the behavior of girls? Since such a finding is flatly inconsistent with the weight of existing evidence, doesn’t it cast technical doubt on the study as a whole? Try as he might, Cherlin could not, in his reply to Gill, come up with an explanation for this anomaly that both explained the girls’ behavior and plausibly defended his study’s methodology.

More fundamentally, Gill questions the failure of Cherlin and his colleagues even to consider what Gill calls the “indirect effects” of divorce on children and society. Drawing upon ideas developed by Norval Glenn of the University of Texas and others, Gill hypothesizes that, in a high divorce society, not only are troubled marriages more likely to end in divorce, but more marriages are likely to become troubled and unhappy. As the ideal of marital permanence gets weaker, and as the observed reality of marital permanence becomes rarer, people tend increasingly to hedge their bets with regard to marriage. Commitment becomes less complete; options stay more open. Consequently, marital satisfaction declines. Thus, a high divorce rate becomes one generator of precisely those “pre-divorce” problems that - let’s see if we’ve got this straight - Cherlin then cites as a reason not to worry too much about effects of divorce.

In his reply to Gill, Cherlin first dismisses this entire theory as either wrong or unproven, then warms up to his real mission: denouncing Gill as a “conservative.” For even though Gill “may not consider himself a conservative,” Cherlin still finds him to be a member of that “largely male” group of “conservative observers” whose “entire program” is based on treating women unfairly, including “impoverishing women into staying married” and pressuring them to “quit their jobs and return home.” Moreover, these observers are driven to do these things by something called “nostalgia,” which fosters a world view based on “returning to the 1950s.”

Never mind that none of these charges have anything to do with the Science article in particular, or with the effects of divorce on children in general. Never mind that Gill nowhere advocates what Cherlin calls the “conservative program,” or anywhere says that we ought to pressure mothers to stay home, turn back the clock to an earlier era, etc. To learn what Gill actually believes, readers may wish to consult his 1997 book, Posterity Lost. The point for Cherlin at the time, it seems, was simply to change the subject by making broad political accusations.

Cut now to 1998. In the American Sociological Review, Cherlin and two colleagues, relying as they had in 1991 on data from Britain’s National Child Development Study, seek to document the long-term effects of divorce on the mental health of children of divorce as they mature into adulthood. The news is not good. Yes, a “previous study” (Cherlin et. al., 1991) had concluded that some of the emotional problems experienced by the pre-adolescent children of divorce (children between the ages of 7 and 11) could be attributed to family problems that predated the divorce. But: “The present study suggests, however, these earlier findings should be modified.”

For it turns out that, as these children of divorce enter into adulthood, the mental health gap separating them from the children of intact marriage widens significantly, with children of divorce suffering an increasingly disproportionate share of emotional problems. What could be causing this growing divergence?  Cherlin says that he and his colleagues are not sure. Maybe something that the researchers can’t put their fingers on, some “unmeasured factors,” are now revealing themselves. 

But it may also be true, Cherlin reports, that this “continuing effect” that is harming the mental health of the adult children of divorce is “a result of the divorce.” If so, “it would suggest that this childhood event can set in motion a chain of circumstances that affects individuals’ lives even after they have left home, married, and entered the labor force. The absence of a strong post-disruption effect at age 11 [the conclusion of the “previous study”] suggests that the long-term effect may emerge only in adolescence or young adulthood. Parental divorce could trigger events such as early childbearing or curtailed education that, in turn, affect adult outcomes.”

Let’s see, now. The act of parental divorce might cause multiple problems for children, some of which are long-term, and some of which surface only as the children get older. In short, for children, the main problem with divorce might be. . .divorce. Sounds very much like reoriented thinking, based on new and accumulating evidence, from one of the nation’s most well-known family scholars. The shift continues.

Sources: Andrew J. Cherlin, Frank F. Furstenberg, Jr., P. Lindsay Chase-Lansdale, Kathleen E. Kiernan, Philip K. Robins, Donna Ruane Morrison, and Julien O. Teitler, “Longitudinal Studies of Effects of Divorce on Children in Great Britain and the United States,” Science 252 (June 1991): 1386-1389; Cherlin, “Nostalgia as Family Policy,” The Public Interest (Winter 1993): 1-8; and Cherlin, Chase-Lansdale, and Christine McRae, “Effects of Parental Divorce on Mental Health Through the Life Course,” American Sociological Review 63 (April 1998): 239-249. Philip and Carolyn Cowan, Letter to the Editor, New York Times, November 5, 1997. Arlene Skolnick, “Family Values: The Sequel,” The American Prospect (May-June 1997): 86-94. Richard T. Gill, “For the sake of the children,” The Public Interest (Summer 1992): 81-96; “Family Breakdown as Family Policy,” The Public Interest (Winter 1993): 8-15; and Posterity Lost: Progress, Ideology, and the Decline of the American Family (Lanham, MD: Rowman & Littlefield, 1997).

First published Winter 1999.