The empirical literature on the longer-term adjustment of children of divorce is reviewed from the perspective of (a) the stressors and elevated risks that divorce presents for children and (b) protective factors associated with better adjustment. The resiliency demonstrated by the majority of children is discussed, as are controversies regarding the adjustment of adult children of divorce. A third dimension of children's responses to divorce, that of lingering painful memories, is distinguished from pathology in order to add a useful complement to risk and resilience perspectives. The potential benefits of using an increasingly differentiated body of divorce research to shape the content of interventions, such as divorce education, by designing programs that focus on known risk factors for children and that assist parents to institute more protective behaviors that may enhance children's longer-term adjustment is discussed.
Key Words: adjustment, children, divorce, resiliency, risk.
Parental divorce has been viewed for 40 years as the cause of a range of serious and enduring behavioral and emotional problems in children and adolescents. Divorced families have been widely portrayed by the media, mental health professionals, and conservative political voices as seriously flawed structures and environments, whereas, historically, married families were assumed to be wholesome and nurturing environments for children (Popenoe, Elshtain, & Blankenhorn, 1996; Whitehead, 1998). Although, on average, children fare better in a happy two-parent family than in a divorced family, two essential caveats that distinguish our position from the stereotypical view are underscored. First, unfortunately, many two-parent families do not offer a happy environment for parents or for children (e.g., Cummings & Davies, 1994; Amato, Loomis, & Booth, 1995). Second, although there are differences in the average psychological well-being of children from happy married families and divorced families, it also is true that the majority of children from divorced families are emotionally well-adjusted (Amato, 1994, 2001; Hetherington, 1999).
A continuing stream of sophisticated social science and developmental research has contributed a more complex understanding of factors associated with children's positive outcomes and psychological problems in the context of both marriage and divorce. As a result, most social scientists relinquished a simplistic view of the impact of divorce more than a decade ago. Research demonstrating that children's behavioral symptoms and academic problems could be identified, in some instances, for a number of years before their parents' divorces was particularly important in facilitating this conceptual shift (Block, Block, & Gjerde, 1986; Cherlin et al., 1991). However, compelling stories of negative outcomes for children of divorce continued to be reported by the media in the past decade, stimulated in part by a 10-year longitudinal study of divorced families that emphasized the enduring psychological damage for children of divorce (Wallerstein & Blakeslee, 1989). More recently, two longitudinal studies that report quite different long-term outcomes for children and young adults (Hetherington & Kelly, 2002; Wallerstein, Lewis, & Blakeslee, 2000) have interested the media in taking a more discriminating look at divorce research, although the preference in the media for drama and simple dichotomous answers remains evident (e.g., Time Magazine, September 25, 2000).
We believe that social science researchers need to look more closely at the varied evidence on children and divorce within and across disciplines and across methodological approaches. Among the basic empirical issues of concern are (a) the confounding of correlation with cause such that any psychological problems found among children from divorced families often are portrayed as "consequences" of divorce, whereas both logic and empirical evidence demonstrate otherwise; (b) the overgeneralization of results from relatively small, unrepresentative, often highly select samples, most notably clinical or troubled samples as in the widely discussed work of Wallerstein; (c) the too ready acceptance of the null hypothesis of no differences in the face of limited and sometimes superficial assessment, particularly in large, often representative samples; and (d) the failure to distinguish between normative outcomes and individual differences in drawing implications for practice and policy, for example, by noting that the majority of children from divorced families are not "at risk" and that family processes after divorce are strong predictors of risk versus resilience. …