Researchers in the 1990s have employed a variety of theories and conceptual perspectives to explain how divorce affects adults and children; these include feminist theory (Carbonne, 1994), attachment theory (Hazan & Shaver, 1992), attribution theory (Grych & Fincham, 1992), symbolic interactionism (Orbuch, 1992), systems theory (Emery, 1994), the social capital perspective (Teachman, Paasch, & Carver, 1996), and the life-course perspective (Amato & Booth, 1997). The largest number of studies, however, begin with the assumption that marital disruption is a stressful life transition to which adults and children must adjust. Many researchers link their work to established stress perspectives, such as family stress and coping theory (Hill, 1949; McCubbin & Patterson, 1983; Plunkett, Sanchez, Henry, & Robinson, 1997), general stress theory (Pearlin, Menaghan, Lieberman, & Mullan, 1981; Thoits, 1995), and the risk and resiliency perspective (Cowan, Cowan, & Schulz, 1996; Hetherington, 1999; Rutter, 1987). Because stress frameworks dominate the literature on divorce, I give them particular attention here. And because these frameworks have much in common, I combine their various elements into a general divorce-stress-adjustment perspective. This conceptual model integrates the assumptions found in many discrete pieces of research, helps to summarize and organize specific research findings from the 1990s, and provides a guide for future research on divorce. This perspective also is useful because it can be applied to children as well as adults.
The divorce-stress-adjustment perspective
The divorce-stress-adjustment perspective, outlined in the Figure, views marital dissolution not as a discrete event but as a process that begins while the couple lives together and ends long after the legal divorce is concluded. The uncoupling process typically sets into motion numerous events that people experience as stressful. These stressors, in turn, increase the risk of negative emotional, behavioral, and health outcomes for adults and children. The severity and duration of these negative outcomes varies from person to person, depending on the presence of a variety of moderating or protective factors. Successful adjustment occurs to the extent that individuals experience few divorce-related symptoms, are able to function well in new family, work, or school roles, and have developed an identity and lifestyle that is no longer tied to the former marriage (Kitson, 1992; Kitson & Morgan, 1990).
Thinking of divorce as a process leads to several useful insights. Uncoupling begins with feelings of estrangement—feelings that typically emerge after a period of growing dissatisfaction with the relationship (Kayser, 1993). Because virtually all people enter marriage with the expectation (or the hope) that it will be a mutually supportive, rewarding, lifelong relationship, estrangement from one's spouse is typically a painful experience. Estranged spouses might spend considerable time attempting to renegotiate the relationship, seeking advice from others, or simply avoiding (denying) the problem. Consequently, the first negative effects of divorce on adults can occur years prior to final separation and legal dissolution. In addition, overt conflict between parents during this period might lead to behavior problems in children—problems that can be viewed as early effects of marital dissolution (Davies & Cummings, 1994).
Furthermore, it is often the case that one spouse wants the marriage to end more than the other spouse does (Emery, 1994). When this happens, the spouse who is considering divorce might mourn the end of the marriage even though it is still legally and physically intact. Indeed, when the marriage is legally terminated, the initiating spouse often experiences a great deal of relief. The spouse who wanted the marriage to continue, in contrast, might not mourn the end of the marriage until the legal divorce is completed. Spouses, therefore, often experience the greatest degree of emotional distress at different points in the divorce process (Emery, 1994). The same principle applies to children. For example, an older child might experience stress prior to the divorce, during the period when the parents' marriage is unraveling. For this older child, the physical separation of constantly warring parents might come as a relief. For a younger child in the same family, however, the departure of one parent from the household might be a bewildering event that generates considerable anxiety. In other words, members of divorcing families can experience different trajectories of stress and adjustment.
Legal divorce does not necessarily bring an end to the stress associated with an unhappy marriage, even for the partner who initiates the divorce. Instead, during the time in which the marriage is ending, and in the immediate postdivorce period, new events and processes (mediators) emerge that have the potential to affect people's emotions, behavior, and health. For adults, mediators include: having sole responsibility for the care of children (among custodial parents); losing contact with one's children (among noncustodial parents); continuing conflict with the ex-spouse over child support, visitation, or custody; loss of emotional support due to declining contact with in-laws, married friends, and neighbors; downward economic mobility (especially for mothers); and other disruptive life events, such as moving from the family home into less expensive accommodation in a poorer neighborhood. With regard to children, divorce can result in less effective parenting from the custodial parent, a decrease in involvement with the noncustodial parent, exposure to continuing interparental discord, a decline in economic resources, and other disruptive life events such as moving, changing schools, and additional parental marriages and divorces. These mediating factors represent the mechanisms through which divorce affects people's functioning and well-being. (For discussions of mediators, see Amato, 1993; Kitson, 1992; McLanahan & Booth, 1989; McLanahan & Sandefur, 1994; Rodgers & Pryor, 1998; Simons and Associates, 1996.)
It is important to recognize that mediators can be viewed as outcomes in their own right. For example, a particular study might focus on the impact of divorce on single mothers' standard of living. But a declining standard of living, in turn, can have consequences for single mothers' sense of financial security, children's nutrition, and older adolescents' opportunities to attend college. Mediators, therefore, represent short- or medium-term outcomes of divorce that can have additional long-term consequences for adults' and children's well-being.
Moderators introduce variability into the manner in which divorce and mediating factors are linked to personal outcomes. Protective factors act like shock absorbers and weaken the links between divorce-related events and people's experience of stress, and hence the extent to which divorce is followed by negative emotional, behavioral, or health outcomes (Rutter, 1987). Resources that lessen the negative impact of divorce might reside within the individual (self-efficacy, coping skills, social skills), in interpersonal relationships (social support), and in structural roles and settings (employment, community services, supportive government policies). For example, although divorce often brings about an initial decline in emotional support, people vary in their ability to reconstruct social networks following divorce, including how quickly they are able to form new, supportive intimate relationships. Another moderator refers to the manner in which people regard divorce, with some individuals viewing it as a personal tragedy (typically the partner who is left behind) and others viewing it as an opportunity for personal growth or as an escape from an aversive or dysfunctional marriage (typically the partner who initiates the divorce). Finally, a number of demographic characteristics, such as gender, age, race, ethnicity, and culture can moderate the effects of divorce. As a result of the particular configuration of moderating factors, some individuals are resilient and others are vulnerable following divorce, resulting in a diversity of outcomes. (For discussions of these and other moderators see Bloom, Asher, & White, 1978; Booth & Amato, 1991; Pearlin et al., 1981; Wheaton, 1990).
Imbedded within the divorce-stress-adjustment perspective are two contrary models. The first, a crisis model, assumes that divorce represents a disturbance to which most individuals adjust over time. According to the crisis model, factors such as personal resources and definitions determine the speed with which adjustment occurs. But given a sufficient amount of time, the great majority of individuals return to their predivorce level of functioning. The second model, a chronic strain model, assumes that being divorced involves persistent strains, such as economic hardship, loneliness, and, for single parents, sole parenting responsibilities. Because these problems do not go away, declines in well-being associated with divorce might continue more or less indefinitely. According to the chronic strain model, factors such as personal resources and definitions determine the level of distress that individuals experience, but divorced individuals do not, in general, return to the same level of well-being they experienced early in the marriage.
Some researchers have argued that stress perspectives tend to focus exclusively on the negative aspects of divorce and ignore positive outcomes for adults (Ahrons, 1994; Wheaton, 1990) and children (Barber & Eccles, 1992; Gately & Schwebel, 1991). For example, women (as well as their children) might feel that they are substantially better off when a relationship with an abusive husband ends. The notion that divorce can be beneficial, however, is not inconsistent with the divorce-stress-adjustment perspective. Many stress theorists, such as Thoits (1995) and Wheaton (1990), have argued that potentially stressful events, such as divorce, can have positive long-term consequences when people resolve their problems successfully. Indeed, the divorce-stress-adjustment perspective explicitly focuses on the contingencies that lead to negative, positive, or mixed outcomes for individuals. Nevertheless, the divorce-stress-adjustment perspective assumes that for most people, the ending of a marriage is a stressful experience, even if much of the stress occurs prior to the legal divorce, is temporary, or is accompanied by some positive outcomes.
The selection perspective
The main alternative to the divorce-stress-adjustment perspective is based on the notion that poorly adjusted people are selected out of marriage. According to the selection perspective, certain individuals possess problematic personal and social characteristics that not only predispose them to divorce, but also lead them to score low on indicators of well-being after the marriage ends. Consequently, the adjustment problems frequently observed among the divorced might be present early in the marriage or might predate the marriage. Some evidence is consistent with the assumption that people bring traits to marriage that increase the risk of divorce, including antisocial personality traits, depression, and a general history of psychological problems (Capaldi & Patterson, 1991; Davies, Avison, & McAlpine, 1997; Hope, Power, & Rodgers, 1999; Kitson, 1992; Kurdek, 1990). Whereas the divorce-stress-adjustment perspective assumes that marital disruption causes adjustment problems, the selection perspective assumes that adjustment problems cause marital disruption. Selection also can occur if the best adjusted divorced individuals are especially likely to remarry. If this is true, then the mean level of functioning in the divorced (and not remarried) population should decline over time.
The selection perspective, as applied to children, assumes that at least some child problems observed following divorce are present during the marriage—an assumption consistent with several longitudinal studies (Amato & Booth, 1996; Cherlin et al., 1991; Elliot & Richards, 1991; Hetherington, 1999). Many researchers assume that these problems are caused by parents' marital discord or by inept parenting on the part of distressed or antisocial parents. Of course, to the extent that dysfunctional family patterns are reflections of the unraveling of the marriage, then these early effects on children can be viewed as part of the divorce process. But the selection perspective goes one step further and argues that inherent characteristics of parents, such as antisocial personality traits, are direct causes of dysfunctional family patterns and divorce, as well as child problems. The discovery that concordance (similarity between siblings) for divorce among adults is higher among monozygotic than dizygotic twins suggests that genes might predispose some people to behaviors that increase the risk of divorce (McGue & Lykken, 1992; Jockin, McGue, & Lykken, 1996). Consequently, some children from divorced families might exhibit problems because they have inherited genetic traits from their (presumably troubled) parents. According to this perspective, to the extent that parents' personalities and genetically transmitted predispositions are causes of divorce as well as child problems, the apparent effects of divorce on children are spurious.