Psychoanalytic Psychology
2004, Vol. 21, No. 3, 353–370

Copyright 2004 by the Educational Publishing Foundation
0736-9735/04/$12.00 DOI: 10.1037/0736-9735.21.3.353

THE UNEXPECTED LEGACY OF DIVORCE
Report of a 25-Year Study
Judith S. Wallerstein, PhD
Judith Wallerstein Center for the Family in Transition and University of
California, Berkeley

Julia M. Lewis, PhD
San Francisco State University

This follow-up study of 131 children, who were 3–18 years old when their parents divorced in the early 1970s, marks the culmination of 25 years of research. The use of extensive clinical interviews allowed for exploration in great depth of their thoughts, feelings, and behaviors as they negotiated childhood, adolescence, young adulthood, and adulthood. At the 25-year follow-up, a comparison group of their peers from the same community was added. Described in rich clinical detail, the findings highlight the unexpected gulf between growing up in intact versus divorced families, and the difficulties children of divorce encounter in achieving love, sexual intimacy, and commitment to marriage and parenthood. These findings have significant implications for new clinical and educational interventions.
The study we report here begins with the first no-fault divorce legislation in the nation and tracks a group of 131 California children whose parents divorced in the early 1970s. They were seen at regular intervals over the 25-year span that followed. When we first met our young participants, they were between ages 3 and 18; by the mid- to late 1990s, when our study ended, they were 28–43 years old. They were the vanguard of an army of adults raised in divorced families who made up one quarter of the American population between the ages of 18 and 44, as reported in 1991 in the National Survey of Families and
Households (personal communication, Norval Glenn, November 1991).
Whereas it is well known that in the closing three decades of the last century the incidence of divorce hovered at nearly half of all first marriages, it is less known that half
Judith S. Wallerstein, PhD, Judith Wallerstein Center for the Family in Transition, Corte Madera,
California, and School of Social Welfare, University of California, Berkeley; Julia M. Lewis, PhD,
Department of Psychology, San Francisco State University.
Correspondence concerning this article should be addressed to Judith S. Wallerstein, PhD, 290
Beach Road, Belvedere, CA 94920. E-mail: judywall@mindspring.com

353

354

WALLERSTEIN AND LEWIS

of the one million children whose parents make up the annual divorce rate are age 6 or under at the breakup (personal communication, Norval Glenn, November 1991). Like our subjects, these children will spend the bulk of their growing-up years in postdivorce families, often within a range of new relationships of one or both parents that include cohabitations and remarriages, and they will experience new losses due to their parents’ broken love affairs or second, and even third, divorces. This is the first and only such report that tells the story of growing up in the postdivorce family through the eyes of children. The divorced family is a new kind of family and not a truncated version of the familiar intact family that has been studied within and across many disciplines. Relationships with stepparents, visiting parents, stepsiblings, and lifestyles that include joint custody have no counterpart in the intact family. Moreover, as we report, when the marital bond is severed, parent–child relationships are likely to change radically in ways that are not predictable from their course during the marriage. Both childhood and parenthood are challenged and often heavily burdened within the divorced family, at the same time that many adults are set free from unhappy and sometimes tragic situations. If we recall what Erikson taught us about the close connection between childhood and society, then we are, as a society, in the midst of profound changes in our relationships with each other and in relationships between the generations. The impact of these far-reaching changes on the society as a whole, as well as on the many individuals whose lives have been profoundly affected, has been hardly addressed or even appreciated.

The Study
The aim of our study was to illuminate the social and psychological experiences of children and parents at the marital breakup and during the postdivorce years. We were especially interested in the impact of the divorce experience on the child’s developmental course, self-concept, and feelings, and on critical passages of growing up to adulthood.
Early on, we were concerned with the bewildered and frightened responses of children to their parents’ breakup. Several years later, when most of the initial crisis-engendered responses had faded, we focused on the many issues associated with adolescence in this group. We were especially concerned with the vulnerability of young adolescent girls in divorced families, the anger and widespread acting out of both the girls and the boys, and their profound need for both committed parents at this often turbulent developmental stage. In the final years of the study, as the youngsters entered adulthood and moved into their 20s and 30s, we were interested in their relationships and overall adjustment within adult society, including their self-concept and values. Our findings at each point have been reported in many publications (Wallerstein & Blakeslee, 1989; Wallerstein & Kelly, 1980;
Wallerstein, Lewis, & Blakeslee, 2002) and are well known to professional and lay audiences both here and abroad. Here we review our earlier findings briefly and then turn to some of our major findings at the 25-year mark, when the participants in the study were well into adulthood. We also construct a beginning agenda for clinical and educational interventions designed to alleviate the widespread anxiety and suffering that we found at that later time.

Our Sample
The study began in 1971 in a northern California county. Divorcing families with minor children were referred by their attorneys with the encouragement of the local family court

THE LEGACY OF DIVORCE

355

judge, who supported our study. Those families selected reflected the willingness of both parents to participate. The children were required to have achieved appropriate academic and developmental norms and to have never been referred to counseling for emotional problems. Thus, the children were, at the outset, a psychologically sturdy group, who had been able to maintain their developmental course on target despite having lived in troubled homes. The children ranged in age from 3 to 18. Fifty-three percent were age 8 or younger.
Fifty-two percent were female. Eighty percent of the families were Caucasian. Parents were well educated: 72% of fathers and 38% of mothers were college graduates. Twentyfive percent of the fathers had graduate degrees. Their socioeconomic status was predominantly middle class. The average age of the parents was in the mid-30s. They had been married an average of 11 years, ranging from 4 to 23 years. It was the first marriage for over 90%. Our purpose in putting together this relatively homogeneous, White, middle-class sample was to look at divorce under the best of circumstances and to omit as far as possible issues of poverty and ethnic discrimination.
Parents and children were recontacted at 18 months postseparation and again at the
5-year and 10-year marks. For the 25-year follow-up, only the children, now adults, were located and interviewed. At the 18-month follow-up, 56 of the 60 families, including 108 children, participated; at the 10-year mark, 52 families and 113 children participated. At the 25-year follow-up, 48 (80%) of the original 60 families were located. In two of them the located fathers were unwilling to put us in contact with their children, and in one the located mother did not know the whereabouts of her child. Three adult children were deceased (one had died of AIDS, one died in a one-car accident that may have been a suicide, and one succumbed to an immune disorder). A fourth was incarcerated for armed robbery. The 25-year sample consisted of 93 adult children (73% of the original sample) from 45 families (75% of the original 60). Their ages when last seen ranged between 28 and 43, with a mean age of 33 years. The high retention rate over 25 years was due to the trust engendered in the interviews, which the subjects commented on appreciatively.
Comparison of families that did not participate in the 25-year follow-up and those that did showed no differences in parent socioeconomic or educational level or in offspring age or educational level, as last reported at the 10-year follow-up.

Comparison Group for the 25-Year Follow-Up
For the 25-year follow-up, we recruited a comparison group of 44 adults in the same age range who had grown up in the same neighborhoods and graduated from the same elementary and high schools. The accounts of these adults who grew up in intact families—some of which were idyllic, some chaotic and conflict ridden, with the majority in between—enabled us to expand our observations into the ways in which growing up in a postdivorce family both resembles and differs from growing up in a “good enough” (or even unhappy) intact family: how living with parents who remain married influences attitudes, choices, and behaviors in courtship and marriage; how living in abusive, chaotic families affects development when the parents remain married.
We use the term comparison rather than control for the group recruited from intact marriages, because it accurately portrays a sample matched for relevant variables of age, socioeconomic status of parents, growing up in the same neighborhood, and graduating from the same elementary and high schools. A personal approach brought in the majority of our participants. The organizers of alumni reunions at the high schools our children had attended were helpful in providing us with lists of classmates and guiding us to those

356

WALLERSTEIN AND LEWIS

whose parents had stayed married. We approached potential subjects with an honest description of our study and asked for their participation. Our requests evoked a lot of interest despite these individuals’ busy lives. Subjects were not paid.
The comparison sample consisted of 44 adults from 27 families. Like our original sample of children at the 25-year mark, they ranged in age from 28 to 43 years, with a mean age of 34. Over 90% were Caucasian, and 64% were female. Parent educational level, the most reliable indicator we could obtain of family socioeconomic status while the children were growing up, was similar to that of the divorced parents: 74% of comparison fathers and 72% of divorced fathers and 41% of comparison mothers and 38% of divorced mothers had completed college. Current income levels of the comparison parents and the divorced fathers were very similar, with 30% of comparison parents versus 33% of divorced fathers being very comfortable and 44% versus 47% being above adequate. A small number (11%) of comparison parents were in below adequate financial circumstances. No fathers in either group were living at poverty level. As expected, divorced mothers showed a different financial range, with 10% at below adequate or poverty level and only 5% in the very comfortable category at the 25-year mark.

Initial Study
The diagnostic process comprised 15 hr of clinical interviews per family. Both parents and all of the children were interviewed separately except in families where the adolescent children insisted on being seen together. The younger children, ages 3–8, were seen in a well-stocked playroom with two fully furnished dollhouses, puppets, drawing materials, toy guns and soldiers, and other miniature paraphernalia. The two dollhouses made it easy for children to play out their feelings about the family; as the opening move of their sessions, almost all of the young children pushed the two houses together and placed the mother and father dolls in the same bed. Children were seen at least twice and often three times, to establish a relationship in which the child could comfortably show her feelings, her fears, her hopes, and her wishes. The assessment interviews explored fantasy material from the play and artistic creation of each child about the nature of changes in the family, what or who the child thought had caused the breakup, and what the child expected would occur as a result.
A history of the courtship, the marriage, the causes for the divorce, and the legal and interpersonal interactions during the breakup was elicited in detail from each parent. They were also asked about violence or infidelity, about their plans for remarriage, and about relationships of the children with other family members and friends. They were asked for details of their child’s behavior in the immediate past and present, as well as their overall impression of the child, and their past and present relationship with the child, including who was responsible for daily care and discipline, what each of the children had been told about the breakup or might have been exposed to in connection with it, and what the child’s reaction had been. Their plans for the future care of the child, including custody and visitation, were elicited as well.
The interviews were conducted by experienced clinicians, all of whom had had special training in working with children. The interviews were recorded in detail, and considerable attention was paid to capturing the exact statements and body language of child and parent. At the initial study and the 18-month follow-up, each child’s teacher was interviewed by a clinician who knew that the child came from a divorced family but was entirely

THE LEGACY OF DIVORCE

357

unfamiliar with the family’s history. Interviewing the teachers in person yielded more detailed and subtle information about the child’s behavior and personality than what could have been recorded on child behavior checklists or questionnaires. Thus, neither the boy who sat for many months refusing to unbutton his heavy jacket in the warm classroom nor the little girl who wove an elaborate fantasy about a new baby brother who did not exist, but about whom she reported to the teacher daily, exhibited the kind of behavior that would have been reported on a checklist. But the boy’s fearful expectation that he would be required to leave at any moment and the little girl’s use of fantasy to fill the void in her family turned out to be themes that remained prominent in their personalities as they grew up. At each of the follow-up contacts, interviews were conducted in an open-ended fashion whereby participants were encouraged to explain in detail their feelings, perceptions, attitudes, behaviors, decisions, opinions, hopes, and anxieties about targeted areas of their lives. Interviews were taped with the subjects’ permission, which was generally granted.
Questions were broad, that is, designed to allow the subject maximum freedom to bring in memories, opinions, and incidents that might have seemed unrelated. Typical questions were “Tell me about love,” “Tell me about sex,” “Talk to me about your relationship with your folks.” These interviews with familiar clinicians, who had participated from the outset, yielded material that was rich in emotional content and in expression. Parents were interviewed for at least an hour and sometimes much longer at each follow-up. Children were interviewed or, depending on their age, seen in structured play sessions and faceto-face conversations.
Participants in the comparison group were similarly interviewed at the 25-year mark.
Interviews with these adult children from intact homes averaged over 2 hr in length, with many lasting well beyond 3 hr in one or more interviews. The interview format used for the comparison group mirrored that used for the divorce group except that parental divorce-related questions were replaced by questions about the parental marriage.
At the 25-year follow-up, a questionnaire was added and administered to both divorce and comparison participants. It asked demographic questions, such as living situation, education, occupation, income, marital status, divorce, and children. It also contained self-report rating scales regarding degree and quality of contact with family members, including children (and stepfamilies in the case of the divorce sample), quality of adult relationships, parenting style, occupational satisfaction, life events, and mood states of self-esteem, happiness, and depression, as well as several open-ended questions about opinions and feelings regarding either the impact of the parental divorce or, for the comparison group, the continuing parental marriage. The questionnaire served as a complementary method of data gathering that we could then compare with the interview responses. Detailed case summaries, containing both the responses of the participant and the clinical impressions of the interviewer, were prepared after each contact. These reports provided the basis for clinical case conferences to develop a family narrative as well as an individual narrative about each child that included those factors in the family, in the milieu, and in the child that placed her at risk, and the protective factors that were available in and outside the family and within the child. The narrative also formulated the child’s central psychological response to the breakup based on the play and fantasy material and the child’s behavior at home and in school. Each transcription (plus the questionnaire responses at 25 years) was coded according to a series of rating scales and categorical items. This coding at 25 years contained information from three methods of data gathering: responses during the interview, self-report responses from the question-

358

WALLERSTEIN AND LEWIS

naire, and clinical impression of the interviewer and the clinical conference. Interviewers did not code subjects whom they had interviewed. Before the 10-year follow-up, interrater reliability in coding was established using the consensus method, where coding discrepancies were discussed and resolved, and was retested on another record. Where consensus was achieved, items were retained; where consensus could not be reached, items were dropped. At the 10- and 25-year follow-ups, a more formal method for reliability was followed, where coding items requiring clinical judgment were first discussed extensively between raters and operational descriptors anchoring rating points were developed. Using selected transcripts from 10% of the sample, we computed interrater reliability using
Kendall’s tau b statistic for ordinal-level items and the kappa statistic for nominal-type coding items. Items below acceptable statistical levels of agreement were reviewed and operational scale points further refined. Where acceptable agreement could not be achieved after this second round, items were dropped. Outcome was assessed using a global measure and three specific measures (Lewis & Wallerstein, 1987).

Brief Literature Review
Although research on children in divorced families has grown from almost none when this study started to a full library, longitudinal studies that follow the course of divorce in children to adulthood are few. We briefly note findings from studies by Amato, Cherlin, and Hetherington. All are based on large-scale sample populations in the United States or the United Kingdom. They use standardized rating scales and symptom checklists, administered either over the phone or in person to parents, teachers, and, as the child grew up, the adult child, for their findings about psychological and social outcomes. Despite differences in particulars, these and other long-term studies using these methods largely agree that there were significant but relatively small differences between adults from divorced families and those who grew up in intact homes. Thus in a full-scale review of research over the 1990s, Amato and his colleagues reported that adults and children scored lower than their counterparts in married-couple families on a variety of indicators of well-being. They found that adults raised in divorced families suffered from a deficit in social skills and had special problems in handling conflicts within their own marriages
(Amato, 2000; Amato & DeBoer, 2001). Cherlin and his colleagues drew their divorce population from a long-term, nationwide public health study in the United Kingdom, begun in 1958 (Cherlin, Chase-Lansdale, & McRae, 1988; Cherlin, Kiernan, & ChaseLansdale, 1995). Although Cherlin found initially that the children’s difficulties were evident prior to the divorce, his most recent work (Cherlin et al., 1995) showed that the subjects from divorced families were experiencing serious psychological difficulties after they reached adulthood, which had not been foreseen. Hetherington studied 144 families, half of them intact and half divorced, with a target child who was 4 years old at the divorce
(Hetherington & Kelly, 2002). This number was subsequently supplemented to obtain a total of 450 families with 900 youths distributed across intact, divorced, and remarried families at the 25-year assessment. Hetherington relied on a combination of psychological tests, video observations, questionnaires, highly structured interviews, and symptom checklists. Many of her findings are in full accord with those reported in this article, particularly regarding the high anxiety young adults from divorced families experience in relationships with the opposite sex and in parenthood. Measuring against a list of psychiatric symptoms, she found that 20%–25% of the children were troubled adults as compared with 10% among those raised in intact families. She noted, “Now, that two-fold

THE LEGACY OF DIVORCE

359

increase is not to be taken lightly. It’s larger than the association between smoking and cancer” (Hetherington, 2002, p. D6).
Although each study differs somewhat in design and methodology, all share the use of group aggregated data based on questionnaires, highly structured interviews, and symptom checklists, subjected then to statistical analysis. Our research stands alone in its qualitative study of each child within his or her family over several decades.

Findings
Early Changes
We present here only a few of the major findings from our work. To summarize some early findings: Growing up was harder for most of the children during the postdivorce years. The lives of parents and children changed radically almost overnight, as parents struggled to reestablish economic, social, and parental functioning, while trying to rebuild the tattered social network of their lives. Children of every age struggled with bewildering, demanding adjustments in their contact with both parents. Often they faced relocation to a new neighborhood and a new school, along with consequent disruptions and losses in their friendships and activities. At home, they confronted seriously diminished parenting just at the time when they needed their parents’ help to make sense of what was happening and to support their efforts to adjust to the major changes within and outside the family.
Typically, the parents themselves became the source of the child’s worry. “Who is taking care of my dad?” was a frequent question. The consequences of the family’s disruption were especially serious for the younger children. “I need a new mommy,” an anxious
5-year-old insisted. To her young mind, her loving, devoted mother had disappeared and been replaced by a tense, cranky, unavailable stranger. The major changes in both parent– child relationships, along with the high anxiety of the children, almost all of whom were taken by surprise by the breakup, have been well documented in our earlier work (Wallerstein & Blakeslee, 1989; Wallerstein & Kelly, 1980) and by others (Hetherington & Kelly,
2002).
Out of their experience of the parental breakup, children of all ages reached a conclusion that terrified them: Personal relationships are unreliable, and even the closest family relationships cannot be expected to hold firm. As we discovered later, this was an enduring theme that rose to new prominence as the youngsters reached adulthood. Their conviction that relationships are unlikely to endure was reinforced by their experiences over the postdivorce years and were entirely unaffected by the amount of time they spent with each parent. Two thirds of the children experienced the multiple marriages and divorces, plus the unrecorded broken love affairs and temporary cohabitations, of one or both parents. Less than 10% of the children had parents who established stable, lasting second marriages in which the children felt fully welcome and included. The frequent discrepancy in the postdivorce adjustment of their parents was also a source of deep distress to the children well into adulthood.

Loneliness and Fear of Abandonment Recalled
Specific events, including the conflicts surrounding the separation and divorce of their parents, had faded but were not forgotten by these children of divorce 25 years later. Early conflict from before the breakup did not dominate their memories, unless the conflict continued to plague the family over several years. In the main, children recalled their own

360

WALLERSTEIN AND LEWIS

feelings of shock and unhappiness at the time of the separation and its aftermath. Almost all remembered feelings of loneliness, bewilderment, and anger at the parents. Many cried as they recounted their history and their childhood fears that they would be forgotten by their preoccupied parents. These feelings were especially powerful, decades later, among those who had been 6 years old or younger at that time. For them, the loss of the parents’ availability was most distressing, because they had so little capacity to comfort themselves. “I remember feeling so alone. I would go for days with no one to talk to or play with.” “I remember being angry at everyone.”

Memories of Loss and Violence
These themes were paired with vividly remembered scenes by those who had experienced a parent’s abandonment or witnessed violence between their parents. Those few who were abandoned recalled in minutest detail the last time and place they saw the lost parent. “I remember the sun striking the patterns on the living room carpet in the late afternoon. It was the last time that I saw my dad. I was 4 years old,” said one 30-year-old woman. In nearly 25% of the children, memories of violent scenes were vivid and detailed. Their fear and sense of helplessness at that time had been fully retained in their adult consciousness.
In half of these families, the violence began or increased during the breakup. A few of the adult children, who had been 4 or 5 at the breakup, had no conscious memories of violent events but reported recurring dreams in which violent episodes were reenacted. One
30-year-old suffered with severe nightmares that occurred twice weekly and recapitulated a particularly violent scene in which her father burst into the home with a gun and attempted to shoot her mother but was arrested in time. When told of the dream, her mother explained that it had happened just that way, when the girl was 4. The daughter answered, “I don’t remember it.” One 34-year-old man described how, at age 5, he would bang his head repeatedly against the wall when his father hit his mother in the adjoining room. Violence was sometimes an overture to sex for the parents, which the children also remembered overhearing. Although the violence stopped after the divorce was final, the children’s posttraumatic symptoms endured. None received treatment prior to adulthood, when some sought therapy on their own.

Memory Fragments
Over half of our subjects reported memory fragments that captured key moments of the breakup or the years that followed. These images intruded into their adult relationships at crisis points. One woman in her 30s told us that her strongest memory of her parents’ divorce, when she was 11 years old, was of her father crying as he walked slowly down the flower-bordered path away from the family home, after her mom threw him out because of his adultery. This memory flashed before her eyes whenever she contemplated leaving her alcoholic boyfriend. By her account, her boyfriend’s tears brought back the image of her weeping father and prevented her from leaving. Such fragments, which so frequently loomed large in their adult relationships, reflected the suffering of the parent that the child had perceived and internalized. The rawness of the parental suffering following the breakup left an indelible emotional mark. One woman said, “I could never do to another human being what my mother did to my father.” We found that more than half of our subjects carried similar memory fragments, which became powerful intruders into their adult relationships. The child’s lasting internalizing of the parent’s suffering or joy has largely been unrecognized.

THE LEGACY OF DIVORCE

361

Childhood
Hardly any of our subjects described a happy childhood; in fact, a number of children told us that “the day they divorced was the day my childhood ended.” Older children took on a lot of responsibility in the household, taking care of younger children and often taking care of needy parents as well. They were proud of their helpful role and developed compassion and a sense of moral responsibility at an early age. For those who did too much over too many years, the price was high. They lost out on their childhood and adolescent pleasures and important aspects of their social development. A discovery at the
25-year mark was how frequently they installed the familiar caregiving role into their own adult relationships and how often they sought out needy, troubled partners whom they nurtured to their own emotional detriment.
As compared with their peers in intact families, children and teenagers reported less play; far less participation in extracurricular activities, such as sports or music; and less involvement in enrichment programs, such as after school classes or summer programs.
This was due to a combination of less money to pay for such activities, less availability of parents to routinely transport the child and attend lessons and events, more frequent neighborhood and school dislocations, constant interruptions in team sports and other activities because of visiting and custody schedules, and less interaction between the divorced parent and other neighborhood parents.
The wistfulness of the young people and their continued longing after those lost opportunities for play came out decades later, when, as adults, they spoke about their hopes for their own children. “I don’t want any child of mine to have my childhood.” “I want to take care of her, I don’t want her to take care of me.” By the 25-year mark, the majority had decided not to have children. Often they gave as their reason the deprivations in their early experiences. “Nothing in my childhood prepared me for parenting.”

Adolescence
As adolescents, most of our children of divorce experienced less protection than their peers in intact families. They took greater responsibilities for themselves. They had fewer rules, and those were often poorly enforced. In many homes there was no curfew. When the youngster was “grounded” or otherwise punished in one parent’s home for misbehavior, it was often not enforced in the other’s. Overall, there was more acting out among the adolescents in the divorced homes than those in the intact homes. This finding coincides with reports from large-scale studies (McLanahan & Sandefur, 1994). One in five of the girls had her first sexual experience before age 14. Over half were sexually active with multiple partners during high school. The girls told us they often frequented bars where pickups were common. They were driven by the wish to be held by a man and to be sought after. Sex itself was not their goal but rather the price they paid for the attention. Several said that they would have sex only if they were drunk or high on drugs. Precautions were rarely taken. Although we neglected to tally the incidence, several of the high school girls reported having multiple abortions.
By contrast, the great majority of girls from intact families postponed sex until late high school or early college, with most reporting having sex only within the context of an established relationship. All had curfews during their high school years. Most of these families had stricter rules and provided greater supervision for girls than for boys. Their parents did not expect virginity by high school graduation, but they did expect that sex would occur within an ongoing, respectful relationship. Although the number of adults from the divorced and comparison groups who reported using drugs and alcohol during

362

WALLERSTEIN AND LEWIS

high school and college was similar (50% vs. 40%), we found, in accord with other studies
(e.g., Resnick, 1997), that students in divorced families reported using earlier and using more heavily and over a longer period.

Higher Education
Differences between the divorced and comparison groups were etched in high relief by the end of the high school years. Almost all of the comparison group remembered talking with both of their parents about plans for college and career. No child of divorce in our study was invited by both parents, either separately or together, to discuss college plans. Only
30% of the children of divorce received full or consistent partial support from their parents throughout college and graduate school, as compared with 90% of the comparison group.
One fairly common pattern was for support to be available during the freshman and sophomore years and to diminish suddenly or disappear in the last two years, creating panic and leading students to drop out.
In California, legal child support ends at age 18, and in the great majority of states there is no parental obligation to provide any financial support afterward. In a few states child support extends to age 21 (or to age 23, as in Massachusetts) for a university student.
Family-law attorneys are generally reluctant to include support for higher education in divorce agreements, holding that if the youngster visits the father regularly and their relationship is friendly, he will be motivated to provide college support. Such was not the experience of these young people.
Patterns of support differed markedly between mothers and fathers. Two thirds of the young people had fathers who were successful professionals, most of them physicians, attorneys, or businessmen. Those mothers who were professionals were teachers, nurses, or social workers. Despite the wide economic disparity between the mothers and fathers, only one third of the fathers provided consistent, partial support for their children’s college expenses, although the fathers were visited regularly and relationships were friendly. Two thirds of those in college received regular partial support from their mothers only.
Major differences between the two groups began at college entry. In high schools where 92% of seniors go on to attend college, only 80% of the divorced group entered college at all. The differences widened as the divorced group combined full-time work with their school studies or alternated semesters of full-time work with college attendance.
Because of their financial difficulties, half of the young people ended up with less formal education than that achieved by their parents. Only 57% of the divorce group achieved their bachelor’s degree as compared with 90% in the comparison group. Another consequence of their endangered education was that few were able to pursue a career in science or any demanding major that would preclude their holding several jobs while attending school. Unhappily, they settled for fields of study that were not their first choice, at lower ranked institutions than their parents had attended. It was at this time that one young person, echoing the emotions of many others, commented bitterly, “I paid for my parents’ divorce.” The Workplace
We note briefly that those who were successful in achieving a good education were successful in the workplace. They combined their formal education with their early independence and their lifelong experience of taking responsibility for themselves and others. They spoke with confidence of their well-honed skills in getting along with

THE LEGACY OF DIVORCE

363

difficult people, because they had learned to negotiate their way between their estranged parents. “I had to rely on my own judgment,” or “I had to be my own parent,” many said with pride. By and large, they behaved responsibly in their commitment to economic self-support and that of their children, and often in their efforts to provide help to needy parents. Almost all were working full time at a wide range of jobs. Those who lacked education worked at less desirable, low-paying jobs that were below their intellectual abilities, but they were nevertheless self-supporting. A few made their living as call girls or peddling drugs. The discrepancy between the responsible behavior of most of these young people in the workplace and their troubled personal relationships has made it difficult to use any one measure to capture their overall adjustment.

Adult Relationships
The central finding of this study is that parental divorce impacts detrimentally the capacity to love and be loved within a lasting, committed relationship. At young adulthood, when love, sexual intimacy, commitment, and marriage take center stage, children of divorce are haunted by the ghosts of their parents’ divorce and are frightened that the same fate awaits them. These fears, which reach a crescendo at young adulthood, impede their developmental progress into full adulthood. Many eventually overcome their fears, but the struggle to do so is painful and can consume a decade or more of their lives. In addition to overcoming their fear of failure, they have a great deal to learn about the give and take of living with another person, about how to deal with differences, and about how to resolve conflicts. This is knowledge that children acquire from growing up with both parents in reasonably harmonious, intact families. As our study ended, 60% of the women and 40% of the men had been able to establish reasonably gratifying and enduring relationships that included a satisfying sexual relationship. Close to 40% had opted for parenthood. The remainder said they were not interested in having children. A good number enjoyed successful careers but suffered from severe loneliness. Because most of these people were still in their 30s, we may yet see changes in their attitudes toward relationships and parenthood.
One third of the men and women were openly pessimistic about marriage and divorce and sought to avoid both. “If you don’t marry, then you don’t divorce,” was their mantra.
Only a few were outright cynical. The majority were eager, even desperate, for a lasting relationship, and fearful that they would never achieve it. They did not want the lives their parents had. Their message was clear: “My parents’ divorce is still incomprehensible to me. They met in college. They fell in love. They were compatible in their tastes and values. So, what is to keep the same fate from happening to me?” Over and over, they told us, “I’d love to get married, but I’m sure that I’d jinx it.” Or, “Any relationship I’m in will dissolve.” Even being in love and living for years in a reasonably harmonious relationship did not mute their fear of commitment. One man told us, “We have been living together for four years. She brought love and laughter into my life. I can’t imagine being with anyone else.
But every time she brings up marriage, I feel this great sadness welling up inside of me.
That was exactly the way I felt when I was 7 years old, when my folks split up.” Even those in apparently stable, satisfactory marriages and relationships had a sense of unease, a strong foreboding that their happiness might be short lived, that they somehow did not deserve to be in a happy, long-lasting union. They felt particularly helpless when dealing with conflict. Any argument or even difference of opinion in a close relationship represented the dreaded slippery slope leading to breakup and loss. Their first response was

364

WALLERSTEIN AND LEWIS

panic, followed by the wish to flee. Almost universally, they expressed what they called
“the fear of the second shoe dropping.” They were convinced that if they went to bed happy, the source of their happiness, whether a great marriage or a beloved child, would be gone by morning. What is so devastating about this foreboding, they explained, is that the happier one feels, the greater the threat of sudden loss becomes.
They complained that they were unprepared for marriage. “I’ve never seen a man and a woman on the same beam.” “Sometimes I feel that I have been raised on a desert island.
Combining love with sexual intimacy is a strange idea to me.” Although, for many, their parents’ divorce may have been wise and necessary, the children did not reach this conclusion until well into adulthood. Most regarded their parents’ divorce as representing the failure to achieve one of the most important of life’s tasks. They complained about their parents’ ineptness. “My mom never taught me about men. She didn’t know anything.” Asked whether they would go to their father for advice, less than 10% said that they would consult him about a personal problem. “I learned from my dad how not to parent,” said one man, who was then expecting his first child. “Is that enough?” we asked.
“That’s all I have. It will have to do,” was the grim reply.
Their anxieties often led them to search for love in strange places, to make impulsive destructive choices, to hang on for years to exploitative partners, to hide their search for love behind promiscuity, to accept whoever volunteered to move in, or to avoid intimate contact altogether, referring to themselves as “children of divorce,” as if this were a fixed identity that defined them forevermore. Men and women reacted differently to their fears, although both groups shared the same desires for love and the same wariness of being hurt.
Men were more likely to withdraw from involvement. A significant number of the young men avoided relationships altogether. As the study ended, 42% of the men had never married or cohabited for longer than 6 months, compared with 6% in the comparison group. Half of the single men in the divorce group led sad, isolated lives. One young man went so far as to discipline himself to go without dinner so that he could avoid the misery of eating alone. Another group of men were inordinately hurt by the failure of a first love affair and withdrew from the dating scene for years thereafter. Many were astonishingly passive in their relationships with women and altogether clueless in responding to the woman’s wishes or complaints when they lived together.
By contrast, none of the women were consistently alone. The dominant pattern was that they jumped headlong, counterphobically, into relationships, often with men they hardly knew or with men in need of rescue. Unlike the men, all of the women from both the divorced and comparison groups had been in relationships, either brief or longer lasting affairs. A subgroup of over 20 women from the divorced group sought out multiple lovers. Many of these women, by their own admission, felt compelled to attract, conquer, reject, and quickly move on. As one woman said, “I wanted to be powerful, like a man.”
Another woman admitted to having one-night stands with over 50 men during her years at graduate school. Their sexual encounters seemed driven by anger at men, which even their close relationships with their fathers did not seem to mute. Ten women told us that when they were with a man they did not care for, they enjoyed the sex, but that when they liked or loved the man, they froze. Some attractive, very young women accepted the first marriage offer they received, whatever the man’s attributes. When asked why they had married, they replied, “I was afraid no one else would ever ask me.” In one such instance, the 23-year-old woman turned to a man she hardly knew, on their second date, and said,
“Marry me. It’s my birthday.”
By contrast, members of the comparison group, even those raised in disappointing marriages, were hopeful that sooner or later they would meet the right person and enter

THE LEGACY OF DIVORCE

365

into a satisfying, committed relationship, usually involving marriage. Considering the high incidence of divorce in our culture, we expected more doubts, but only a small minority admitted to worry. “I never doubted I’d marry and have a family” was a typical comment.
They expected ups and downs in their relationships, but they did not expect to fail, if they chose carefully. The issue of choice of partner, which was so baffling to the children of divorce, was where the comparison group told us they put their greatest efforts. Their confidence that things would eventually work out well enabled most to survive heartbreak and to delay marriage until they felt ready. Often they drew on their family of origin for images of what they wanted. “I didn’t want a volatile lady like my mom.” Many men and women mentioned that they wanted someone who would be a good parent to their future children. Asked how she chose her husband, one woman laughingly answered, “Besides his being devastatingly good looking, you mean? I wanted someone who wasn’t too serious, who would treat me well, who would be a good father, and was someone I’d like to wake up with 50 years later.” This way of thinking, which came easily to many of those raised in intact homes, was omitted from the voices of the women in the divorce sample.
The wide differences in the incidence of marriage and divorce between the children of divorce and the children of intact families are in line with national data (personal communication, Norval Glenn, November 1997, based on figures from the Statistical Abstract of the U.S. Bureau of the Census, 1997, for adults ages 18–44). By the end of our study,
60% of the divorced group had married, as compared with 80% of the comparison group and 84% of the national sample. Fifty percent of those married before age 25. Of those,
57% divorced. In the comparison group, only 11% married before age 25, and of those,
25% divorced. The overall divorce rate by the end of the study was 40% for the children of divorce as compared with 35% nationwide for persons in that age group, but only 9% for our comparison group. The outcomes of later marriage, for both the divorce group and the comparison group, are unpredictable. In looking at these delayed marriages, the majority seem to have improved. Several went out of their way to marry people from intact families. “He has no baggage,” one woman declared triumphantly in describing her spouse. “No one has ever been divorced in his family.”
Of the one third of the adults who had gone into therapy because of their worries about adult relationships, one half found the experience “very helpful.” The group that delayed marriage until they were in their 30s had experienced sexual liaisons and failed love affairs during their 20s but appeared to have learned to choose more carefully and with greater self-confidence by the time they reached their 30s. Several were restrained from continuing their promiscuous or provocative behavior by acquiring stable spouses, who were generous and loving and aware of the problems. They also had been able finally to loosen their ties to their parents and to gain the emotional and geographical distance needed to free themselves from the sense that the failure of their own relationships was inevitable. A Brief Look at Parent–Child Relationships
Parenting suffered from greater inconsistency in the postdivorce families compared with the intact families. We found that the importance of the stable marital bond and family structure in protecting parent–child relationships and stabilizing the commitment of each parent to the child had been underestimated. Parenting in the absence of these protective factors was subject to fluctuation and distraction, especially when the mother or father became involved in a passionate love affair or entered a second or third marriage, or

366

WALLERSTEIN AND LEWIS

divorced again, or when a parent became bored or annoyed with the foibles of their adolescent children.
In keeping with the attention that has been focused in recent years on father–child relationships (Thompson, 1994), we report that at the 25-year mark, one third of the children had seen their fathers weekly during their childhood and with some regularity during their adulthood. Others who saw their fathers frequently as children had little contact as adults. This drop-off of contact at adulthood, especially among the boys, was striking in comparison with fathers and sons in the comparison group, many of whom grew closer after the father retired. Of those who maintained frequent contact with their father after the divorce, several lived in his home for a year or more during their childhood or adolescence. A number worked in their father’s business and saw him daily over many years. Yet these young people did not show any less anxiety about love and commitment than those who saw their father less frequently or hardly at all. Actually, the father’s influence often was evident in the child’s career choice but not in his or her understanding or trust in relationships. Thus, one young woman who became a dedicated environmental attorney attributed her career choice to her father’s ardent support of conservation activities, which she had shared with him as a child and as an adolescent. At age 30, she described being involved in multiple relationships that lasted an average of 6 months and of experiencing sexual frigidity if she liked the man. She said, “It’s not sex that scares me.
It’s getting close.”
Children observed their parents carefully and formed strong moral judgments about each. They watched how parents treated each other and tried hard to read the true feelings beneath the surface behavior. Their interest in their parents’ interaction continued throughout their growing-up years. Parents’ continued love or anger toward each other, and their envy of ex-spouses’ new husbands, lovers, or younger wives, were discussed endlessly among siblings. If one parent had rebuilt his or her life successfully and the other was unhappy, the discrepancy distressed the children. As adults, they pitied the lonely parent and blamed the fortunate parent for lack of compassion or for having caused the other’s suffering. As children, their feelings toward their parents seemed more passionate than those of their peers in intact families, who took their parents more or less for granted, unless a parent became ill or distressed.
As adults, many of our participants still felt angry that they had not been informed about the cause of the divorce. Those whose reluctant visits to a parent were rigidly enforced by a court order during their adolescence remained especially angry. They rejected contact with that parent after they reached age 18 and often did not resume contact during the decades that followed. Stepparent–stepchild relationships varied greatly. Some stepparents were liked, admired, and respected; others were resented or ignored. Unless they were active parents when the child was very young, they rarely evoked the intense love or passionate concern that the children felt for their biological parents. Discussion
The call to liberalize divorce in the early 1970s promised happier and better marriages.
Ironically, findings from this study show that although divorce sets many adults free, and many second marriages are happier, these benefits do not extend to their children. Divorce begets fewer marriages, poorer marriages, and more divorces. This should not encourage us to retreat from regarding divorce as an adult right. However, it does call attention to

THE LEGACY OF DIVORCE

367

enduring problems in the lives of the children involved. Where did we go wrong, and what can we do?
The findings from this study call for a shift in our dominant paradigm of understanding the impact of divorce on children and in the interventions that have been developed to mitigate its effects. The widely accepted premise has been that divorce represents an acute crisis from which resilient children recover, typically within a 2-year period, and then resume their normal developmental progress, if three conditions obtain: (a) the parents are able to settle their differences without fighting; (b) the financial arrangements are fair; and
(c) the child has continued contact with both parents over the years that follow. Implicit in this model is the notion that after the turmoil of the divorce, the parent–child relationships return to the status quo ante; parenting resumes much as it was before the split, and the child continues to do well, or even better, minus the marital conflict of the predivorce years. A parallel paradigm places loss at the center of the divorce: The hazard to the child is primarily the loss of one parent, usually the father. In this view, it is held that the child will be protected against long-term problems if continued contact with both parents is ensured. The first model has led to a range of interventions centering on reducing conflict between the parents, including mediation, collaborative divorce, programs provided under the aegis of the courts to educate parents in ways to eschew conflict and litigation, and a range of other educational programs to help high-conflict parents bring their anger under control. The second model has found its expression in joint custody, in legal efforts to block the mother’s move away from the community where the father resides, and in encouraging fathers to value their continuing role of active participation in their child’s upbringing after the breakup.
However, most of the children in this study were in ongoing contact with their fathers throughout their childhood. One third visited weekly or more frequently. None of the parents engaged in conflict through the courts over visitation or custody. When parents got along and both maintained caring relationships with their children of the first marriage, undiminished by their postdivorce relationships, and when both parents were doing reasonably well in their personal lives, the childhood and adolescence of the children were protected. However, even a protected childhood did not shield the children, at late adolescence and young adulthood, from the fear that their love relationships would fail.
This 25-year study points to divorce not as an acute stress from which the child recovers but as a life-transforming experience for the child. The divorced family is not simply an intact family from which the troubled marital bond has been removed. There are many stresses in the postdivorce family, and a great many daunting adjustments are required of the children. Hence, though the divorce was designed to relieve stress and may well have done so for the adult, for the child the stresses of the divorced family may be more burdensome, and he may feel correctly that he has lost more than he has gained. This is especially the case if, like most children in this study, the child was relatively content before the breakup and had no expectation of the upheaval ahead. Our findings suggest that whereas children in intact homes often seek continuity with their parents, those from divorced homes seek discontinuity. They fear identification with their parents. Those in our study who were close to stable grandparents felt reassured and comforted by the models that the grandparents provided, but only a minority had extended family members who remained, in the words of the children, “faithful” to them. Contrary to the loss model, remaining in frequent contact with both parents did not alleviate their suffering in adulthood, especially if the condition of the parents was discrepant and one parent remained lonely and unhappy.

368

WALLERSTEIN AND LEWIS

It appears that when the child of divorce arrives on the stage of adulthood, the setting is lacking in good images of how an adult man and woman can live together in a stable relationship—and this becomes the central impediment that blocks the child’s developmental journey. The need for a good internal image of the parents, as a couple, is important to every child during his growing-up years. The significance of this internal template increases in adolescence. Sad memories from the past and observations from the present build to a dramatic crescendo as young people from divorced homes confront the issues of love, sex, and lasting commitment, and as they address the practical workaday problems of choosing a life partner, of forming a realistic image of what they are looking for, of distinguishing love from dependency, and of creating an intimate relationship that holds. How is the inner template of the child of divorce different from that of the young adult in the intact family, especially if the child has access to both parents and the parents refrain from fighting? As every “child of divorce” in our sample told us, no matter how often they see their parents through the years, the image of them together as a couple is forever lost; and a father in one home and a mother in another does not represent a marriage. Joint custody does not teach children how to create adult intimacy and mutual affection, how to resolve marital conflicts, or how to deal, as a couple, with a family crisis. As they grew up, these children lacked this central reassuring image. By strong contrast, the children from intact families told many stories about their home life and how their parents met and married. They had spent their growing-up years observing their parents’ interactions and learning about marriage, and they were well aware of the expectable ups and downs. For the children of divorce, the parents’ interactions—including the courtship, the marriage, and the divorce—collapsed into a black hole, as if the parents as a unified couple had vanished from the world and from the child’s inner life.

Implications for Interventions
This study, along with others, has spawned educational and clinical programs throughout the country that address parents and children at the time of the breakup. There are no studies as yet of the long-term effectiveness of these or other interventions.
There are several policy issues that emerge from this study. They include: (a) equalizing access to higher education by extending child support nationwide beyond age 18 for youngsters in college, in families where the youngster would expectably have received substantial financial support had the parents remained together; (b) greater recognition by courts, mediators, and parents of the importance of considering the interests and concerns of adolescents in setting custody schedules; and (c) treatment at the time of the breakup for children and parents in families where the children have witnessed parental violence, in order to prevent posttraumatic symptoms from consolidating. We believe that these measures would ease the suffering and reduce the lasting anger of many children toward their parents.
The major challenges of engendering hope, creating good images of man–woman relationships, and teaching young people to choose appropriate partners and create a relationship that will hold are staggering in their complexity and go far beyond any interventions yet attempted. What follows are some initial suggestions based largely on reports from clinicians and reports of treatment from the subjects of this study.
There are indications from university counseling services that many adults who grew up in divorced families seek out therapy, especially during their first two college years

THE LEGACY OF DIVORCE

369

(personal communications, including consultations at Stanford University, University of
California at Berkeley, University of Michigan, and San Francisco State University).
Counseling centers have successfully initiated groups and individual therapy for these students, who come with urgent pleas for help with their failed relationships or grave concerns about their parents—including those parents who waited to divorce until the youngest child went to college. This population would provide a splendid opportunity for a range of pilot projects.
Findings from this study have provided a detailed agenda for groups in several locales, including groups run by private practitioners. One such program, run by Dr. S. Demby in
New York City, entitled “Leftover Business From My Folks’ Divorce,” has drawn a lively response (personal communication, 2001). Also, our experience at the Judith Wallerstein
Center for the Family in Transition in Corte Madera, California, showed similar strong interest among young professionals who were suffering from failures in relationships, sexual inhibitions, and difficulty in separating emotionally from their parents. Our experience has been that daughters feel especially guilty about enjoying a happier relationship with a man than their mothers were able to achieve.
Organized groups or courses in high school and, especially, in college might prove effective in eliciting attitudes of doubt and cynicism as well as stereotypes about men’s and women’s behavior in close relationships. The challenge would be to find counselors, therapists, or teachers who could hold the students’ interest, raise provocative questions rather than preaching at these young people, and deal candidly with issues of trust, love, and sex, while conveying honesty, integrity, and hope.
One third of the subjects in this study sought individual therapy in adulthood. It is encouraging that those who benefited were able to terminate exploitative relationships quickly and went on to find appropriate partners. Clinicians reported that these people were excellent candidates for expressive therapy because of their youth, their pain, and their high motivation to work hard to change their lives. The problems they presented are in keeping with this study. Therapists need training in understanding and developing appropriate strategies geared specifically to the special challenges these young people bring. They make very quick contact with the therapist, but as they begin to value the therapy, their fears of being abandoned emerge powerfully in the transference, and their impulse is to flee before the therapist leaves them. If the therapist addresses these fears early in the treatment, it will enable the therapy to continue. These young adults are also in danger of feeling overwhelmed by sorrows and angers that lie close to the surface, as if their parents’ divorce happened only yesterday. The therapist can help by acknowledging how long and how bravely these individuals have kept their suffering to themselves, perhaps in order to protect a needy parent, but that it is now safe to close the door on the past. This, then, defines their task.
Finally, a major theme in family life education might be to help parents discuss the reasons for the divorce with their children, as they become older adolescents. Silence or vague explanations offered by most parents only contribute to the young person’s sense that divorce strikes suddenly, without warning. The family-life educator could also help the parents review with the child the mistakes that were made by both parties. Most important, the parents should assure the adolescent of their hope that their youngster will succeed in creating lasting relationships of his or her own. Such explicit assurances might alleviate the “fear of success” that haunts so many children of divorce. The goal should be to help the young person view divorce not as inevitable but as a result of avoidable human error (Wallerstein & Blakeslee, 2003).
It remains to be seen how much these and other, yet to be developed interventions can

370

WALLERSTEIN AND LEWIS

reduce anxieties and change attitudes that are continually reinforced by the surrounding culture. We are in new territory as clinicians and educators, and as members of a society in flux.

References
Amato, P. R. (2000). The consequences of divorce for adults and children. Journal of Marriage and the Family, 62, 1269–1287.
Amato, P. R., & DeBoer, D. D. (2001). The transmission of marital stability across generations:
Relationship skills or commitment to marriage? Journal of Marriage and Family, 63,
1038–1051.
Cherlin, A. J., Chase-Lansdale, P. L., & McRae, C. (1988). Effects of parental divorce on mental health throughout the life course. American Sociological Review, 63, 239–249.
Cherlin, A. J., Kiernan, K. E., & Chase-Lansdale, P. L. (1995). Parental divorce in childhood and demographic outcomes in young adulthood. Demography, 32, 299–318.
Hetherington, E. M. (2002, March 26). [Interview.] New York Times, p. D6.
Hetherington, E. M., & Kelly, J. (2002). For better or for worse: Divorce reconsidered. New York:
Norton.
Lewis, J. L., & Wallerstein, J. S. (1987). Methodological issues in longitudinal research on divorced families. In J. P. Vincent (Ed.), Family intervention, assessment and theory (Vol. 4, pp.
121–142). Greenwich, CT: JAI Press.
McLanahan, S., & Sandefur, G. (1994). Growing up with a single parent: What hurts, what helps.
Cambridge, MA: Harvard University Press.
Resnick, M. D. (1997). Protecting adolescents from harm: Findings from the National Longitudinal
Study on Adolescent Health. JAMA, 278, 823–832.
Thompson, R. A. (1994). The role of the father after divorce. In Richard E. Behrman (Ed.), The future of children (Vol. 4, pp. 210–235). Menlo Park, CA: The David and Lucille Packard
Foundation.
Wallerstein, J. S., & Blakeslee, S. (1989). Second chances: Men, women and children a decade after divorce. Boston: Houghton Mifflin.
Wallerstein, J. S., & Blakeslee, S. (2003). What about the kids? Raising your children before, during, and after divorce. New York: Hyperion.
Wallerstein, J. S., & Kelly, J. B. (1980). Surviving the breakup: How parents and children cope with divorce. New York: Basic Books.
Wallerstein, J. S., Lewis, J. M., & Blakeslee, S. (2002). The unexpected legacy of divorce. A 25-year landmark study. New York: Hyperion.