Chairman Brownback, Senator Lautenberg, and Members of the Subcommittee:
My name is Gordon Berlin. I am the executive vice president of MDRC, a unique nonpartisan social policy research and demonstration organization dedicated to learning what works to improve the well-being of disadvantaged families. We strive to achieve this mission by conducting real world field tests of new policy and program ideas using the most rigorous methods possible to assess their effectiveness.
I am honored to be invited to address your committee about what we know and do not know about the effects of marriage and divorce on families and children and about what policies and programs might work to promote and strengthen healthy marriages, especially among the poor. My goal is to briefly summarize the evidence in three areas: (1) what we know about the effects of marriage, divorce, and single parenthood on children; (2) what we know about the effectiveness of policies and programs that seek to stem persistently high rates of divorce and out-of-wedlock childbearing; and (3) what we know about the likely effects of these policies on low-income families and children. The central focus of my remarks will be to explicate the role that marital education, family counseling, and related services might play in promoting and strengthening healthy marriages and to discuss what we know about the potential of strategies that seek to ameliorate the key stressors (for example, job loss, lack of income, domestic violence, and childbearing) that make it difficult to form marriages in the first place or act as a catalyst that eventually breaks up existing marriages.
To summarize my conclusions:
- First, children who grow up in an intact, two-parent family with both biological parents present do better on a wide range of outcomes than children who grow up in a single-parent family. Single parenthood is not the only, nor even the most important, cause of the higher rates of school dropout, teenage pregnancy, juvenile delinquency, or other negative outcomes we see; but it does contribute independently to these problems. Neither does single parenthood guarantee that children will not succeed; many, if not most, children who grow up in a single-parent household do succeed.
- Second, an emerging body of evidence suggests that marital education, family counseling, and related services can improve middle-class couples' communication and problem-solving skills, resulting initially in greater marital satisfaction and, in some cases, reduced divorce, although these effects appear to fade over time.
- Third, we do not know whether these same marital education services would be effective in reducing marital stress and eventual divorce among low-income populations or in promoting marriage among the unmarried. Low-income populations confront a wide range of stressors that middle-class families do not. The evidence is limited, and mixed, on whether strategies designed to overcome these stressors, for example, by providing job search assistance or by supplementing low earnings, rather than relying solely on teaching marital communication and problem-solving skills would also increase the likelihood that low-income couples would marry or that married couples would stay together.
- Fourth, to find out whether and what types of policies and programs might successfully strengthen marriage as an institution among low-income populations as well as among a wide variety of ethnically and culturally diverse populations, our national focus should be on the design, implementation, and rigorous evaluation of these initiatives.
Marriage, Divorce, and Single Parenthood
Encouraging and supporting healthy marriages is a cornerstone of the Bush Administration's proposed policies for addressing the poverty-related woes of single-parent households and, importantly, for improving the well-being of low-income children. The rationale is reasonably straightforward: About a third of all children born in the United States each year are born out of wedlock. Similarly, about half of all first marriages end in divorce, and when children are involved, many of the resulting single-parent households are poor. For example, less than 10 percent of married couples with children are poor as compared with about 35 to 40 percent of single-mother families. The combination of an alarmingly high proportion of all new births occurring out of wedlock and discouragingly high divorce rates among families with children ensures that the majority of America's children will spend a significant amount of their childhood in single-parent households. Moreover, research shows that even after one controls for a range of family background differences, children who grow up living in an intact household with both biological parents present seem to do better, on average, on a wide range of social indicators than do children who grow up in a single-parent household (McLanahan and Sandefur, 1994). For example, they are less likely to drop out of school, become a teen parent, be arrested, and be unemployed. While single parenthood is not the main nor the sole cause of children's increased likelihood of engaging in one of these detrimental behaviors, it is one contributing factor. Put another way, equalizing income and opportunity do improve the life outcomes of children growing up in single-parent households, but children raised in two-parent families still have an advantage.
If the failure of parents to marry and persistently high rates of divorce are behind the high percentage of children who grow up in a single-parent family, can and should policy attempt to reverse these trends? Since Daniel Patrick Moynihan first lamented what he identified as the decline of the black family in his 1965 report, The Negro Family: The Case for National Action, marriage has been a controversial subject for social policy and scholarship. The initial reaction to Moynihan was harsh; scholars argued vehemently that family structure and, thus, father absence was not a determinant of child well-being. But then in the 1980s, psychologists (Wallerstein and Kelly, 1980; Hetherington, 1982) began producing evidence that divorce among middle-class families was harmful to children. Renewed interest among sociologists and demographers (Furstenberg and Cherlin, 1994) in the link between poverty and single parenthood soon emerged, and as noted above, that work increasingly began building toward the conclusion that family structure did matter (McLanahan and Sandefur, 1994). Of course, the debate was not just about family structure and income differences; it was also about race and gender. When Moynihan wrote in 1965, 24 percent of all births among African-Americans occurred outside of marriage. Today, the black out-of-wedlock birthrate is almost 70 percent, and the white rate has reached nearly 24 percent. If single parenthood is a problem, that problem cuts across race and ethnicity.
But the story has nuance. Yes, growing up with two parents is better for children, but only when both mother and father are the biological or “intact” (as opposed to remarried) parents. In fact, there is some evidence that second marriages can actually be harmful to adolescents. Moreover, marriage can help children only if the marriage is a healthy one. While the definition of a “healthy marriage” is itself subject to debate, it is typically characterized as high in positive interaction, satisfaction, and stability and low in conflict. Unhealthy marriages characterized by substantial parental conflict pose a clear risk for child well-being, both because of the direct negative effects that result when children witness conflict between parents, and because of conflict's indirect effects on parenting skills. Marital hostility is associated with increased aggression and disruptive behaviors on the part of children which, in turn, seem to lead to peer rejection, academic failure, and other antisocial behaviors (Cummings and Davies, 1994; Webster-Stratton, 2003).
While our collective hand-wringing about the number of American births that occur out-of-wedlock is justified, what is often missed is that the birthrate among unmarried women accounts for only part of the story. In fact, birthrates among unmarried teens and African-Americans have been falling — by a fourth among unmarried African-American women since 1960, for example (Offner, 2001).
How, then, does one explain the fact that more and more of the nation's children are being born out of wedlock? Because the nonmarital birth ratio is a function of (1) the out-of-wedlock birthrate (births per 1,000 unmarried women), (2) the marriage rate, and (3) the birthrate among married women (births per 1,000 married women) - the share of all children born out of wedlock has risen over the last thirty years, in large measure, because women were increasingly delaying marriage, creating an ever larger pool of unmarried women of childbearing age, and because married women were having fewer children. Indeed, families acted to maintain their standard of living in the face of stagnant and falling wages, earnings, and incomes during the 1970s and 1980s by having fewer children and sending both parents into the workforce, a strategy that undoubtedly has increased the stress on low-income two-parent families (Levy, 1988), and that contributed to the rise in out-of-wedlock births as a proportion of all births.
Concern about these trends in out-of-wedlock births and divorce, coupled with the gnawing reality that child poverty is inextricably bound up with family structure, has encouraged conservatives and some liberals to focus on marriage as a solution. Proponents of this approach argued that many social policies — welfare and tax policy, for example — were actually anti-marriage, even if research only weakly demonstrated that the disincentives to marry embedded in these policies actually affected behavior. Moreover, they maintained that social policy should not be neutral — it should encourage and support healthy marriages — and they stressed the link between child poverty and single parenthood and the positive child effects associated with two-parent families.
The focus on marriage was met with skepticism by others. Critics argued that marriage was not an appropriate province for government intervention and that income and opportunity structures were much more important factors than family structure. They questioned why the focus was on low-income families when the normative changes underlying the growth in single-parent households permeated throughout society, as witnessed by the prevalence of divorce across all economic classes.
“Fragile Families” Are Pro-Marriage
More recent evidence from the Fragile Families and Child Well-Being Study tipped the balance for many in favor of the pro-marriage arguments. Designed by two prominent academics, Sara McLanahan and Irv Garfinkel, the study is a longitudinal survey of 5,000 low-income married and nonmarried parents conducted in 75 hospitals in twenty cities at the time of their child's birth. Among mothers who were not married when their child was born, 83 percent reported that they were romantically involved with the father, and half of the parents were living together. Nearly all of the romantically involved couples expressed interest in developing long-term stable relationships, and there was universal interest in marriage, with most indicating that there was at least a fifty-fifty chance that they would marry in the future. Looking at employment history and other factors, researchers estimated that about a third of the couples had high potential to marry; another third had some problems, like lack of a job, that could be remedied; while the final third were not good candidates due to a history of violence, incarceration, and the like (McLanahan, Garfinkel, and Mincy, 2001).
There was certainly reason to be cautious about presuming a link between what people said and what they might actually do, and longer follow-up data did indeed throw some cold water on initial optimism. However, when the Fragile Families data were thrown into the mix with the trend data and with the data that suggested that family structure was a determinant of poverty, the reaction was catalytic. The notion was reinforced that more marriage and less child poverty would result if public policies could just be brought in line with the expressed interests of low-income couples.
Marital Education Can Work
But what, if anything, could government actually do to promote marriage among low-income families? For some policy analysts, the discovery of marriage education programs seemed to provide the missing link. To the surprise of many, not only did these programs exist, but there was a body of evidence, including more than a dozen randomized trials, indicating that marriage education programs could be effective. Marriage education refers to services that help couples who are married or planning to marry to strengthen their communication and problem-solving skills and thus their relationships. Models range from those that adopt a skills-based instructional approach to those that use a therapeutic “hands on” approach that addresses the specific marital problems facing individual couples.
Some of the cutting-edge work now underway provides a flavor of the approaches being developed. Dr. Phil Cowan and Dr. Carolyn Cowan, both professors of psychology at the University of California, Berkeley, have been involved in the development and rigorous testing of family instruction models for more than twenty years. Dr. Benjamin Karney, a psychologist at the University of Florida, has been conducting a longitudinal study of newly married couples. Dr. Richard Heyman, a psychologist at the State University of New York at Stony Brook, has 15 years’ experience conducting prevention and treatment research on couple and family interaction. Dr. John Gottman, who leads the Relationship Research Institute where he focuses on marriage, family, and child development, has developed and carefully evaluated some of the most innovative new approaches to marital education and group instruction. Dr. Pamela Jordan developed the Becoming Parents Program, a couple-focused educational research program being tested in a large randomized trial. Dr. Howard J. Markman and Dr. Scott Stanley, both of the University of Denver, developed and refined the Preparation and Relationship Enhancement Program (PREP).
Among the skills-training programs, PREP is the most widely used with couples who are about to marry. It teaches skills such as active listening and self-regulation of emotions for conflict management and positive communication. PREP also includes substantial content on topics such as commitment, forgiveness, and expectations clarification. PREP appears to have a significant effect on marital satisfaction initially, but the effect appears to fade over time (Gottman, 1979), and there is some indication that it improves communication among high-risk couples but not low-risk couples (Halford, Sanders, and Behrens, 2001). Therapeutic interventions are more open-ended and involve group discussions, usually guided by trained professionals to help partners identify and work through the marriage problems they are facing. The most carefully evaluated of the structured group discussion models targeted couples around the time of their child's birth, an event that triggers substantial and sustained decline in marital satisfaction. Couples meet in a group with a trained therapist over a six-month period that begins before the child is born and continues for another three months after the birth. Initially, marital satisfaction soared and divorce rates plummeted relative to a similar group of families that did not participate in the program. But the divorce effects waned by the five-year follow-up point, even while marital satisfaction remained high for those couples who stayed together (Schultz and Cowan, 2001). More recent work by Cowan and Cowan and by John Gottman appears to produce more promising results.
Both the Cowans’ model of education via structured group discussions and a marital-education and skills-development model pioneered by John Gottman led to positive effects on children. The Cowans found positive effects in the school performance of children whose parents participated in their couples instruction and group discussion program. Gottman describes improved cooperative interaction between the parents and their infant child and sustained increased involvement by fathers.
While the results from the marriage education programs are encouraging, they are not definitive. Most of the studies are small, several have serious flaws, and only a few have long-term follow-up data (and those that do seem to show decay in effectiveness over time). Moreover, only a handful of the studies collected information on child well-being. Most importantly, all of the programs studied served mostly white, middle-class families, not the low-income and diverse populations that would be included in a wider government initiative.
Context and Low-income Families
Not surprisingly, low-income couples have fewer resources to cope with life's vagaries. They are more likely to experience job loss, have an unexpected health or family crisis, be evicted from or burned out of their home, be the victim of a violent crime, and so forth. As a result, they face greater difficulty than middle-class individuals in forming and sustaining marriages. With the exception of African-Americans, low-income couples are not less likely to marry; but they are more likely to divorce when they do marry. Yet evidence from the Fragile Families survey of 5,000 low-income couples who have just given birth to a child and ethnographic interviews conducted with low-income women in Philadelphia by Kathy Edin of Northwestern University provide convincing evidence that low-income people share the same normative commitment to marriage that middle-class families demonstrate. As Kathy Edin told the Senate Finance Committee last week, “[T]he poor already believe in marriage, profoundly so. The poor want to marry, but they insist on marrying well. This…is the only way to avoid an almost certain divorce.”
If poor families share the same commitment to marriage as better-off couples, what is it about their low-income status that inhibits the formation of stable marriages? One possible explanation is the mismatch between a large number of stressful events they face and few resources with which to respond to those stressors. The imbalance places greater demands on the individuals in a dyad, leaving less time together and less time to dedicate to relationship building than might be the case for a middle-class couple. In addition, the problems low-income couples have to manage — problems such as substance abuse, job loss, eviction, chronic infidelity, a child with a chronic condition like asthma or developmental delays, and criminal activities — may be more severe than those confronted by better-off couples. (Edin, 2004; Karney, Story, and Bradbury, 2003; Heyman, 2000).
Because the problems low-income couples confront are likely to be more acute and chronic than those faced by middle-class couples, it is an open question whether the problem-solving and communication skills taught by marital education programs will be as effective among low-income couples as they appear to have been for middle-class couples (where the evidence base is still evolving). Clearly, the skill sets taught in those programs and the strategies applied by therapists and counselors to solve the problems couples present will need to be adapted. Moreover, it is possible that these kinds of stressors overwhelm the abilities of individuals to use the skills they are taught. It is difficult to be understanding of a partner's failings when the rent is due and there is not enough money to pay it.
Such concerns have elicited two kinds of responses: first, efforts to adapt marital education programs to better meet the needs of low-income families; and second, proposals to combine marital education with strategies that would directly tackle the poverty-related stressors on family life — for example, with help in finding a job, income supplements to make up for low wages, child care assistance, and medical coverage.
Adapting Marital Education to the Needs of Low-Income Families
Underpinning the interest in public support for marital education programs is a conviction that low-income individuals do not have good information about the benefits of marriage. In part, this dearth results from their experience of having grown up in single-parent households where they were simply not exposed to role models that might inform their own relationships. In part, it is a consequence of their lack of access to the same kinds of supports and information, counseling, and therapy that are often available to middle-class couples contemplating marriage or divorce. Buoyed by the success of the model marriage education programs with middle-class families, and following the lead of former Oklahoma Governor Frank Keating, who was determined to end his state's embarrassing status as the nation's divorce capital, practitioners of marital education programs have begun applying and adapting these models to the needs of low-income couples. The objective is to equip low-income couples with relationship skills to improve couple interaction by reducing negative exchanges (anger, criticism, contempt, and blaming) and strengthening positive behaviors (expressions of support, humor, empathy, and affection). The logic is obvious: When couples enjoy positive interaction and are successful in handling conflict, their confidence and commitment would be reinforced, thereby fostering satisfaction and stability. But the designers of these programs recognize that they must adapt marital education as middle-class families know it to better meet the different needs of low-income households. This might involve changes in the types of agencies that deliver services, the training leaders would get, the content and examples used in the training, the duration and intensity of services, and the balance between strengthening internal communication and the forging of links to community programs that can provide support related to the contexts in which poor families live.
Does Reducing Financial Stress Promote Marital Stability?
While there is a strong relationship between poverty and marital breakup, would programs that ameliorate poverty by providing supports to the working poor actually improve marital relationships? There have been few tests of this question; the most relevant recent reform that has been carefully evaluated for two-parent families is the Minnesota Family Investment Program (MFIP). Implemented in 1994, MFIP used the welfare system to make work pay by supplementing the earnings of recipients who took jobs until their income reached 140 percent of the poverty line, and it required nonworkers to participate in a range of employment, training, and support services. For two-parent families, MFIP also eliminated the arcane work-history requirements and the “100-hour rule,” a policy that limited the number of hours a primary earner could work and still receive welfare but which had the perverse, unintended effect of encouraging couples to divorce so they could remain eligible for welfare.
MDRC's evaluation of MFIP examined program effects on employment, income, marriage, and other family outcomes up to three years after entry. Because MFIP treated two-parent family recipients (who were receiving welfare at the onset of the study) and new applicants differently, outcomes for these groups were examined separately. We found that two-parent recipient families in MFIP were as likely as those in a comparable group of welfare recipients who were not eligible for MFIP to have at least one parent work; but the MFIP sample was less likely to have both parents work, leading to an overall reduction in their combined earnings of approximately $500 per quarter. Yet because the program supplemented the earnings of participating families, the two-parent recipient families who participated in MFIP still had slightly higher family incomes (up $190 per quarter more, on average, when taking into account their decreased likelihood of separating or divorcing — and, thus, retaining access to both partners' earnings). In contrast, MFIP had fewer effects on parental employment, earnings, and income for welfare applicants, a finding that is not entirely surprising given their short welfare spells.
One of the striking findings of the three-year evaluation was that, among the 290 two-parent recipient families who were part of a follow-up survey sample, families in the MFIP group were 19.1 percentage points more likely than families in the group who received traditional welfare payments under the Aid for Families with Dependent Children (AFDC) program to report being married and living with their spouse. Most of this increase in marital stability was a result of fewer reported separations in MFIP families as compared to AFDC families, although some of it was a result of small reductions in divorce. Because there is some question about how families on welfare might report their marital status, MDRC also obtained and analyzed data from publicly available divorce records. We did this for some 188 two-parent recipient families who were married at study entry. (The other 100 or so families in the original survey sample were cohabiting, and we did not look for marriage records for them). The data confirmed that these couples were 7 percentage points less likely than their AFDC counterparts to divorce. This gave us confidence that MFIP did indeed reduce marital instability. (Again, divorce records would not tell us about the separations we found in the survey, so the effect should be smaller than the 19 percentage point effect we found there).
These findings have two important implications. First, make-work-pay strategies might reduce financial stress and increase the likelihood that two-parent families stay together. Second, given the small number of people followed in the MFIP survey sample, MFIP's marriage effects on all two-parent families should be investigated and the results should be replicated in other locations before the findings are used to make policy.
As a first step in that process, MDRC went back to the state of Minnesota to obtain divorce and marriage records for the full sample of 2,200 two-parent MFIP families (including both recipients and applicants) for a follow-up period of more than six years. This fuller record would give us the opportunity to understand whether the positive effects on divorce (but not the much larger effects on separation) we found for the 290 two-parent families in the survey sample applied to the larger group of two-parent MFIP families. In addition, we wanted to learn about MFIP's possible effect on subgroups of two-parent families that we could not previously examine.
Six years later, the full-sample story on divorce is decidedly mixed. Overall, for the full sample of two-parent families, there is no discernable pattern of effects on divorce over time. When we look at the two-parent recipient families only, those eligible for the MFIP program appear to be less likely to get divorced, but the finding is not statistically significant until the last year of follow-up, leaving open the possibility that the pattern we see could still be due to chance. Moreover, the pattern among applicants is also uncertain — barely statistically significant in one year, but favoring more rather than less divorce. The different direction in the findings for the recipient and applicant groups explains the absence of an overall effect on divorce. And in both cases, the effects we did see were small — about a 3 to 4 percentage point difference in divorce between the MFIP group and the AFDC group. Finally, recall that public marriage and divorce records can capture only a family's legally documented marital status. They cannot distinguish informal statuses like separations, the form of marital dissolution that drove the dramatic 36-month recipient findings mentioned above. We are currently planning further analyses to better understand MFIP's effects on divorce for these and other subgroups. We have no reliable way of exploring the separation findings.
MFIP's initial results were tantalizing in large part because MFIP was not specifically targeted to affect marriage, divorce, or separations, and yet it appeared to produce large effects on the likelihood that some two-parent families would stay together, suggesting that strategies that tackle the vagaries of poverty could promote marital stability by reducing some of the economic stress on poor families. But the full-sample findings cast some doubt on that promise (with regard to divorce but not separations), reinforcing the need to replicate programs like MFIP for two-parent families in different settings before reaching conclusions about the contribution such strategies might make toward strengthening marriage. The findings particularly leave open the question of the possible range of effects that programs could achieve if policies providing marital education were combined with policies designed to affect employment and income.
What We Don't Know
While the evidence base on marital education is extensive, there is much left to learn. For example:
- Will participation in marital education programs by low-income couples lead to an increase in marriage and in marital harmony and, in turn, have lasting effects on couples' satisfaction, on parenting skills and practices, and on children?
- Will the skills taught in marital education programs be a match for the poverty-related stresses experienced by low-income families, or are additional supports such as employment and income also needed to reduce divorce and increase the number of healthy marriages?
- Will marriage education programs be effective regardless of race, ethnic identity, and cultural norms, and how should these programs be adapted to better meet different groups' divergent needs?
- Who will participate in marital education programs? Will they attract predominantly couples who already have a deep commitment to each other or couples whose problems are acute? Will a broad cross-section of low-income couples participate or only a narrow slice of the population?
- Will these programs facilitate the dissolution of unhealthy marriages as proponents contend, or will they prolong marriages that might be better off dissolving or not forming in the first place?
- Can a relatively short education course — say, 10 to 20 hours spread over a few months — have a long-lasting effect on marital and couple discord, or are more long-term strategies and even one-on-one back-up couple-counseling services necessary? What is the right duration and intensity of an initiative? Can courses be short term and intense, or must they be longer and more sustained to yield longer-lasting effects? What is the right content? What are the implications for affordability and scale?
An Opportunity to Learn
On substantive, policy, and financial grounds, there are good arguments to be made for public involvement in the marriage field. If marital education programs could be mounted at scale, if participation rates among those eligible were high, and if the programs were effective in encouraging and sustaining healthy two-parent families, the effects on children could be important. The key word is if!
The strong correlation between growing up in a two-parent family and improved child outcomes does not ensure that intervening to encourage more marriage and less divorce will have the intended results. Indeed, social policymaking based on correlation has an uncanny way of ending with unintended consequences. The only reliable way to understand whether marital education and other supports designed to strengthen marriage produces such results is to conduct a social experiment with the right mix of quantitative and qualitative methods to answer the “what difference,” “how,” and “why” questions.
The Administration of Children and Families within the U.S. Department of Health and Human Services has launched two new projects to do just that. Managed by Mathematica Policy Research, the Building Strong Families evaluation is targeted to low-income unwed couples beginning around the time of their child's birth. The Supporting Healthy Marriage initiative, which is being overseen by MDRC, is aimed at low-income married couples. Both projects will involve large-scale, multisite, rigorous random assignment tests of marriage-skills programs for low-income couples. The goal is to measure the effectiveness of programs that provide instruction and support to improve relationship skills. Some programs might also include services to help low-income couples address barriers to healthy marriages, such as poor parenting skills or problems with employment, health, or substance abuse. Programs operated under these demonstration umbrellas will screen for domestic violence and help participants gain access to appropriate services. Done well, the results from these path-breaking projects should inform the marriage field, and they should add value to our existing understanding of the potential and the pitfalls of government intervention in this critically important arena.
Cummings, E. M., and P. Davies. 1994. Children and Marital Conflict. New York: Guilford.
Edin, K. 2004. Testimony Before the United States Senate Committee on Finance Subcommittee on Social Security and Family Policy. The Benefits of Healthy Marriage Hearing, May 5.
Edin, K., and M. Kefalas. 2004. Promises I Can Keep: Why Poor Women Put Motherhood Before Marriage. Berkeley: University of California Press.
Gennetian, L. and V. Knox. 2004. Getting and Staying Married: The Effects of a Minnesota Welfare Reform Program on Marital Stability. New York: MDRC.
Gottman, J. M. 1979. Marital Interaction: Experimental Investigations. Oxford, England: Elsevier.
Furstenberg, F. and A Cherlin. 1994. Divided Families: What Happens to Children when Parents Part. Cambridge, MA: Harvard University Press.
Halford, W. K., M. R. Sanders, and B. C. Behrens. 2001. “Can Skills Training Prevent Relationship Problems in At-Risk Couples? Four-Year Effects of a Behavioral Relationship Education Program.” Journal of Family Psychology 15, 4: 750-768.
Hetherington, E. M., M. Cox, and R. Cox. 1982. “Effects of Divorce on Parents and Children.” In M. Lamb (ed.), Nontraditional Families. Hillsdale, NJ: Erlbaum.
Heyman, R. E. 2001. “Observation of Couple Conflicts: Clinical Assessment Applications, Stubborn Truths, and Shaky Foundations.” Psychological Assessment 13: 5-35.
Karney, B. R., L. Story, and T. Bradbury. 2003. “Marriages in Context: Interactions Between Chronic and Acute Stress Among Newlyweds.” Presentation at the International Meeting on the Developmental Course of Couples Coping with Stress, October 12-14, 2002, Boston College, Chestnut Hill, MA.
Levy, F. 1988. Dollars and Dreams: The Changing American Income Distribution. New York: Norton.
McLanahan, S., I. Garfinkel., and R. B. Mincy. 2001. “Fragile Families, Welfare Reform, and Marriage.” Policy Brief No. 10. Washington DC: Brookings Institution.
McLanahan, S., and G. D. Sandefur. 1994. Growing Up with a Single Parent: What Hurts? What Helps? Cambridge, MA: Harvard University Press.
Moynihan, D. P. 1965. The Negro Family: The Case for National Action. Washington, DC: U.S. Department of Labor, Office of Policy Planning and Research.
Offner, P. 2001. “Reducing Non-Marital Births.” Policy Brief No. 5. Washington, DC: Welfare Reform and Beyond.
Schultz, M., and C. P. Cowan. 2001. Promoting Healthy Beginnings During the Transition to Parenthood. Minneapolis: Society for Research in Child Development.
Wallerstein, J., and J. Kelly. 1980. Surviving the Breakup: How Children and Parents Cope with Divorce. New York: Basic Books.
Webster-Stratton, C. 2003. The Incredible Years. Toronto: Umbrella Press.
The Impact of Family Structure and Family change on Child Outcomes: A Personal Reading of the Research Literature
Ministry of Social Development
The paper provides a brief overview of the research literature on the impacts of family structure and family change on child outcomes, with a particular focus on parental separation. It takes as a starting point the existence of pervasive associations between family change and child outcomes and addresses a range of issues that are examined in the research literature. Do family changes primarily have short-term impacts on children, or do they also have more enduring impacts? How does remarriage affect child outcomes? What impact do frequent changes of family structure have on child outcomes? What are the mechanisms that link family structure and family change to child outcomes? How much of the impact is attributable to income changes consequent on parental separation? How much is attributable to the absence of a parent figure? How much is attributable to poorer mental health of lone parents following a parental separation? How much is attributable to the conflict between parents which often accompanies a parental separation? And how much of the association between family change and child outcomes is due to non-causal mechanisms, such as selection effects? The paper will sketch out answers to these questions, as far as these can be determined from the published results of research.
Over the past two decades or so, a significant literature has developed on the impact of family structure and family change on child wellbeing. This literature documents an accumulating body of evidence that children raised in different family contexts display differential patterns of outcomes across a wide range of developmental domains. In particular, children raised in lone-parent families have been found, on average, to do less well across a range of measures of wellbeing than their peers in two-parent families, while parental separation has been found to be associated with an array of adverse outcomes for children. Behind these patterns of associations between family contexts and child outcomes, however, lies a complex web of overlapping and interacting influences, which means that interpreting these results is far from straightforward. It is the aim of this paper to throw some light on the reasons why child outcomes are contingent on family contexts.
The paper provides a brief overview of the research literature in this field. For reasons of space, the paper focuses rather narrowly on the impact of parental separation on child outcomes, although it also briefly examines the impact of remarriage and multiple family transitions on child wellbeing. Within this constrained purview, however, the paper examines a range of issues that are canvassed in the research literature. It takes as a starting point the existence of pervasive associations between family change and child outcomes and considers a range of questions that follow from this: Do family changes such as parental separation primarily have short-term impacts on children, or do they also have more enduring impacts? How does remarriage affect child outcomes? What impacts do frequent changes of family structure have on child outcomes? What are the mechanisms that link family structure and family change to child outcomes? Are there causal connections between family change and child outcomes or are there other reasons for these associations? The paper also examines an exemplar intervention that has been shown to ameliorate the adverse impacts of family change on children’s wellbeing.
The literature on these questions is large, complex and growing so fast that it is no longer possible even to keep abreast of new papers produced each year, let alone master everything that has been published in the past two decades. This poses a challenge for a brief survey of the literature such as this. It needs to be said that this paper is not based on a systematic review of the literature in this field. Although I have tried to read widely and without bias, the portion of the literature I have been able to read is necessarily selective – and the portion I can reference in this paper is much more constrained – while the very act of selection has, no doubt, been shaped by my own views and interests. The paper should thus be regarded as no more than a personal reading of the literature.
Parental Separation and Child Outcomes
Parental separation has been reported in the literature as being associated with a wide range of adverse effects on children’s wellbeing, both as a short-term consequence of the transition and in the form of more enduring effects that persist into adulthood. Effects reported include adverse impacts on cognitive capacity (Fergusson, Lynskey and Horwood 1994), schooling (Evans et al. 2001), physical health (Dawson 1991), mental and emotional health (Chase-Lansdale et al. 1995), social conduct and behaviour (Morrison and Coiro 1999), peer relations (Demo and Acock 1988), criminal offending (Hanson 1999), cigarette smoking (Ermisch and Francesconi 2001), substance use (Fergusson, Horwood and Lynskey 1994), early departure from home (Mitchell et al. 1989), early-onset sexual behaviour (Ellis et al. 2003) and teenage pregnancy (Woodward et al. 2001).
A further range of impacts in early adulthood and beyond include higher rates of early childbearing (McLanahan and Bumpass 1994), early marriage (Keith and Finlay 1988), marital dissolution (Amato and DeBoer 2001), lone parenthood (McLanahan and Booth 1989), low occupational status (Biblarz and Gottainer 2000), economic hardship (McLanahan and Booth 1989), poor-quality relationships with parents (Aquilino 1994), unhappiness (Biblarz and Gottainer 2000), discontentment with life (Furstenberg and Teitler 1994), mistrust in others (Ross and Mirowsky 1999), and reduced longevity (Tucker et al. 1997).
On the face of it, this seems like a long and forlorn listing, which suggests that parental separation bears down heavily on children and blights their lives to a significant degree across all domains of functioning. Yet the picture is not as bleak as this litany of problems might suggest. In most cases the size of the reported effects is small; a minority of children are negatively affected, generally only in the presence of other exacerbating factors; and in many cases the existence of a causal connection is contested and other competing explanations for these associations have been put forward. In other words, it is important to be cautious in interpreting the meaning of these patterns of association.
Many scholars who have identified associations between family structure and family change and child outcomes have drawn attention to the relatively small size of the effects. Joshi et al. (1999) describe the effect sizes they measured as “modest”, while Burns et al. (1997) refer to effects that were “very weak”. Allison and Furstenberg (1989) report that the proportion of variation in outcome measures that could be attributed to marital dissolution was generally small, never amounting to more than 3%.
The modest nature of the associations between separation and children’s outcomes means that knowing that a child comes from a separated family, and knowing nothing else about the child, has little predictive power in terms of the child’s wellbeing. There is a wide diversity of outcomes among both groups of children from divorced and intact families, and the adjustment of children following divorce depends on a wide range of other factors.
Demo and Acock (1996) note that “the differences in adolescent well-being within family types are greater than the differences across family types, suggesting that family processes are more important than family composition”. Indeed, O’Connor et al. (2001) showed that differences in adjustment between children within the same family are as great as, and even slightly greater than, differences between children in different families. Demo and Acock (1996) note further that measures of family relations explained the largest proportion of variance in adolescent wellbeing.
The majority of children whose parents have divorced function within normal or average limits in the years after divorce (Kelly 1993). As a group, they can not be characterised as “disturbed”. Furthermore, there is a considerable range of functioning within both groups of children from divorced and intact families. Among children whose parents have divorced are many who are functioning quite well, while among children from intact families are many with major adjustment problems. In short, there is no one-to-one relationship between divorce and psychological adjustment problems in children.
In fact, not only do some children do well despite the divorce of their parents, but some children actually benefit from the divorce. Demo and Acock 1988 note that adolescents living in single-parent families can “acquire certain strengths, notably a sense of responsibility, as a consequence of altered family routines”. It is likely, however, that such benefits will accrue only where the altered routines are structured and predictable. Changes that involve the emergence of more chaotic patterns of family life are unlikely to be beneficial for children, even if some strive to furnish a sense of order where their parents fail to do so. Butler et al. (2002) note that the children in their study demonstrated “an active role helping their parents cope with divorce, even in circumstances where parents did not seem able to contain their more negative emotions and impulses”.
Children also benefit where a parental separation provides release from an aversive family situation; for example, where the parental relationship is highly conflicted and the children are drawn into the conflict (Booth and Amato 2001, Jekielek 1998) or where the child’s relationship with a parent figure is of poor quality (Videon 2002). Videon (2002) notes that:
The prophylactic effects of parental separation are amplified as adolescents’ satisfaction with the parent–adolescent relationship decreases. When adolescents are residentially separated from an unsatisfying same-sex parent relationship … their level of delinquent behaviour is lower than adolescents who continue to reside with a same-sex parent with whom they have a poor relationship.
A further circumstance where children may benefit from a parental separation is where a parent exhibits antisocial behaviour. Jaffee et al. (2003) found that the less time fathers lived with their children, the more conduct problems the children had, but only if the fathers exhibit low levels of antisocial behaviour. In contrast, when fathers exhibit high levels of antisocial behaviour, the longer they lived with their children the more conduct problems the children exhibited. In such cases, children are likely to be receiving a double whammy of genetic and environmental factors that heighten the risk of conduct problems.
Nevertheless, despite all these caveats and qualifications, it remains true that children whose parents separate do less well, on average, across a range of measures of wellbeing. A pressing question that follows from this is why these associations arise. Before examining this question, I will consider briefly whether remarriage changes the outlook for children who have experienced a parental separation, what impact multiple family transitions have on child wellbeing and whether the effects of parental separation are primarily short-term or whether it also has more persistent and enduring consequences for children’s wellbeing.
Remarriage and Child Outcomes
Remarriage does not generally improve outcomes for children, despite the potential gains from both improved economic circumstances and the presence of an additional adult to help with parenting tasks. Indeed, some studies have shown children to be worse off after a parent’s remarriage. Elliott and Richards (1991) found that having a stepfather1 had a deleterious effect on children’s behaviour scores. Fergusson et al. (1986) found that, among children who had experienced a parental separation, those whose parents reconciled or whose mother remarried exhibited more behavioural difficulties than children who remained in a single-parent family. Baydar (1988) found that, although divorce was not negatively related to mothers’ reports of children’s behavioural and emotional problems, remarriage was.
It appears, then, that there is something about the complexity of family life in stepfamilies that hinders them from benefiting from the additional resources that are available when a lone mother remarries. Relationships within stepfamilies are complex and need time and goodwill on all sides to work well. Unlike the relationship between mother and stepfather, that between stepfather and stepchild is not a relationship of choice, which means that goodwill may sometimes be in short supply, at least in the early stages of establishing a stepfamily. Children are often suspicious of their mothers’ new partners and slow to open up to the benefits the new relationship might confer on them, while stepfathers are often uncertain about how to respond to the children of their new partner (Amato 1987). Typically, this uncertainty results in lower levels of involvement: as Fine et al. (1993) note, stepfathers appear to actively refrain from becoming involved with their stepchildren, engaging in both fewer positive and fewer negative behaviours. Perhaps as a result, cohesion remains lower among stepfamilies than among intact families (Pryor & Rodgers 2001). Even so, improvements in stepfamily functioning are evident over time (Amato 1987), which suggests that many families manage to master the challenges they face.
Multiple Family Transitions
Several studies have found that multiple family transitions are especially damaging for children. Dunn et al. (1998) reported that the number of transitions impacted both on children’s adjustment problems and on levels of prosocial behaviour. Kurdek et al. (1994) found that, although the effects of the number of parenting transitions were significant, these accounted for a relatively small percentage of the variation of adjustment, ranging between 5% and 8% across three separate samples.
Aquilino (1996) reported that the experience of multiple transitions and multiple family types, among a sample of children not born into an intact biological family, was associated with lower educational attainment and greatly increased the likelihood that children would try to establish an independent household and enter the labour force at an early age.
One possible explanation is that having multiple transitions presents children with a succession of caregivers … and this experience may weaken children’s attachment to any particular caregiver, making early autonomy seem more attractive. Similarly, having a variety of caregiving arrangements and multiple separations from caregivers may weaken both parents’ and children’s sense of mutual obligation and thus reduce the exchange of support across generations.
The evidence on this, however, is not entirely consistent. A range of other studies failed to turn up any evidence that multiple transitions are more damaging to children’s wellbeing (Booth and Amato 2001, Carlson and Corcoran 2001, Teachman 2002). It may be that the impact of multiple transitions depends to some extent on the circumstances associated with transitions. Where transitions are well managed and conducted with goodwill, they may do little damage, while transitions that are chaotic, unpredictable and infused with rancour and disputation may have malign effects on children’s wellbeing.
Short-Term and Long-Term Impacts
Many of the reported effects of parental separation on child wellbeing are based on observations that are taken in the short term. However, other studies have examined effects over longer-term durations, some into adulthood. While there is evidence that many of the difficulties that children encounter as a result of parental separation decline as time passes, there is also evidence that some effects are persistent and enduring.
Chase-Lansdale and Hetherington (1990) found that during the first two years after a divorce both children and adults experienced pragmatic, physical and emotional problems as well as declines in family functioning. By two years after the divorce the majority of families had made significant adjustments, although among children there were variations by age and gender. While girls seemed to recover fully during the primary school years, boys in mother-custody homes exhibited behaviour problems for as long as six years.
However, Chase-Lansdale et al. (1995) found that parental divorce had negative consequences for the mental health of some offspring that persisted into adulthood. Parental divorce was associated with a moderate increase in the average score on a measure of mental health (indicating deterioration) and a 39% increase in the risk of psychopathology. Despite this significant effect, it is important to note that only a minority of people were at such risk: 82% of women and 94% of men whose parents divorced were predicted to fall below the clinical cut-off for psychopathology.
Amato and Keith (1991), in a meta-analysis of studies that examined long-term consequences of parental divorce, reported adverse impacts on a range of domains of adult wellbeing, including psychological adjustment, use of mental health services, behaviour and conduct, educational attainment, material quality of life and divorce.
The last effect implies that the risk of a failed marriage is transmitted intergenerationally, a finding that is supported by other studies (Mueller and Pope 1977, Amato and DeBoer 2001, Teachman 2002). Amato and DeBoer (2001) found that parental divorce approximately doubled the odds that children’s own marriages would end in divorce. These increased odds appear to be the end result of a longer chain of effects. Children whose parents separated have been found to be more likely to engage in early-onset sexual activity, to leave home at an early age, to enter into an intimate partnership at an earlier age and to become parents at an early age. Early entry into marriage is known to heighten the risk of separation and divorce. In addition, Mueller and Pope (1977) hypothesised that these effects arise in part because youthful marriages involve less socially and emotionally mature individuals, are subject to greater economic hardship and receive less social support, both normatively from wider society and from family and kin.
Even though the majority of children of divorced families are functioning within normal ranges or better on a variety of objective measures of adjustment, Kelly (2003) notes that divorce can create lingering feelings of sadness, longing, worry and regret. Perhaps as many as half of young adults recall distress and painful memories and experiences caused by their parents’ behaviours and post-divorce custody arrangements. Even if many children do not experience mental health disorders according to a clinical diagnosis, there is no doubt that for most it causes pain and sadness in their lives.
Wallerstein and Corbin (1989) draw attention to the period of late adolescence as a time when delayed responses to an earlier parental divorce emerge in young women, giving rise to anxieties in the domain of their relationships with young men. They also point to adolescence as a period when young women are more sensitive to the relationship between their parents:
It is the relationship between the parents, after all, that forms the template for heterosexual relationships and provides the young woman with a basis for her own hopes and expectations … Thus, it may not suffice for divorced parents to refrain from angry fighting. It may be equally important to their daughters for parents to treat each other fairly and with continued kindness.
Mechanisms That Link Parental Separation to Child Outcomes
A range of mechanisms has been postulated to explain the link between parental separation and adverse child outcomes. Five mechanisms will be considered in the following discussion:
- income changes consequent on parental separation
- paternal absence
- poor maternal mental health following a separation
- interparental conflict
- compromised parenting practices and child-parent relations.
Each of these mechanisms implies a causal connection between associations between parental separation and adverse child outcomes. A range of alternative explanations for the associations that do not involve causal connections has also been proposed. These non-causal explanations are examined in the following section.
Income Changes Consequent to Parental Separation
The economic circumstances of families decline after divorce, especially among mother-headed families. Amato (1993) outlined a range of ways in which the economic position of a family might exert effects on child wellbeing:
Financial hardship may negatively affect children’s nutrition and health; it reduces parental investment in books, educational toys, computers, private lessons; it constrains choice of residential location, which means that the family may have to live in a neighbourhood where school programmes are poorly financed, services are inadequate and crime rates are high; children are more likely in such neighbourhoods to associate with delinquent peers.
As well as having a direct impact on child outcomes, economic factors are also likely to have impacts through indirect pathways. The stress associated with economic hardship can have negative impacts on parental mental health, which in turn can have consequences for children’s wellbeing.
A number of studies have found that when controls for income are applied, the effects of parental separation decline significantly (Carlson and Corcoran 2001) or even vanish entirely (e.g. Blum et al. 1988), which implies that post-separation economic circumstances account for much of the deficit in wellbeing among children in separated families. However, other studies show that the post-separation economic situation of families is not fully responsible for adverse outcomes among children and, moreover, that this has varying impacts on different outcomes. Wu (1996) found that the impact of a change in family structure on the probability of a premarital birth was largely unaffected when controls for income measures were applied, and noted that this suggested that family instability and income have largely independent effects on the probability that a young woman would bear her first child outside marriage.
Hetherington et al. (1998) also found only modest support for the economic deprivation hypothesis. They cite a number of studies that found that even when income is controlled, children in divorced families exhibit more problems than do children in non-divorced families. They also note that although the income in stepfamilies is only slightly lower than that in non-divorced families, children in these families show a similar level of problem behaviour to that in divorced mother-custody families. They conclude that the effects of income do not seem to be primary and are largely indirect.
Overall, it might be concluded that declines in economic circumstances following separation may explain part, but by no means all, of the poorer outcomes among children who have experienced a parental separation.
Following a parental separation, most children live in the primary custody of one parent, although joint custody arrangements have become increasingly common over recent years. In most cases, the custodial parent is the mother, which means that a significant aspect of the experience of post-separation family life, for most children, is the absence of their father. Although other custody arrangements are increasingly common, the research in this area has still tended to focus on “father absence”.
There are a range of a priori reasons to hypothesise that the absence of the father from the home might have a negative impact on children’s wellbeing. As Amato (1993) notes, the absence of one parent means a deficit in terms of parental time available to do the work of parenting (and all the other work in the household, which further restricts the available time for parenting). Children will also lack exposure both to an adult male role model and to the skills and processes involved in a committed adult relationship, including such things as communication, negotiation, compromise and expression of intimacy (although it must be said that many couples in intact relationships model such things imperfectly at least part of the time). In addition, children are likely to suffer where the absence of their father from the home means that they have lost effective contact with him.
Despite these hypothetical grounds for expecting a “father absence” effect, research studies have generally failed to find evidence to show that this plays a strong role in explaining the differential outcomes experienced by children from divorced and intact families. Two pieces of evidence, in particular, weigh against it.
First, children whose parents separated do worse than children who have experienced a parental bereavement. Biblarz and Gottainer (2000) found that, compared with children of widowed mothers, children of divorced mothers had significantly lower levels of education, occupational status and happiness in adulthood. They found no evidence that divorced mothers were less competent parents than widowed mothers and speculated that the contrasting positions in the social structure of different types of single-mother families may account for observed differences in child outcomes. In particular, they note that widows occupied an advantaged position in the social structure, in terms of employment, financial position and occupational status, compared with divorced mothers. This suggests that the absence of the father, if it has an effect, has a much weaker effect than that of these economic factors.
Secondly, as has already been noted, remarriage does not generally improve the wellbeing of children, despite the gain of another adult to help with the task of parenting. As a number of studies have noted, outcomes for children in remarried families are generally little different from those of children in sole-parent families. It is important to note also that remarriage generally results in an improvement in economic circumstances. As noted above, there appears to be something associated with stepfamilies – perhaps the complexities of the new pattern of relationships that need to be established and worked at before the family can settle down into new comfortable ways of living together – that weighs against both the economic gain and the gain of an additional adult figure. Once again, this suggests that the absence of the father, by itself, does not play a strong role in explaining the differences between children from divorced and intact families.
There are various reasons why the impact of the father’s absence might be less than expected. Other adults may be filling the gap by providing adult role models and support to lone parents, and many fathers continue to make significant contributions to their children’s wellbeing after separation. It is not just the father’s presence in the home that is important; it is his presence in the child’s life.
Maternal Mental Health
Maternal mental health is another mechanism through which parental separation exerts effects on children’s wellbeing. The pathways that connect separation, maternal mental health and child wellbeing are somewhat complex and are likely to operate via the route of impairments to parenting. The process of separation can take a toll on the mental health of separating parents, which can in turn impair the quality of parenting.
Block et al. (1988) note that divorced mothers describe themselves in terms indicating low self-esteem and that a failed or failing marriage affects mothers more strongly than fathers. Hetherington et al. (1982, cited in Amato 1993) showed not only that separation took a toll on the mental health of custodial mothers, in the form of higher rates of anxiety, depression, anger and self-doubt, but also that this in turn impacted on their parenting, exemplified by less affectionate, less communicative, more punitive and more inconsistent disciplinary interactions with their children. Such sub-optimal parenting behaviours, in turn, have adverse consequences for children’s wellbeing.
On the other hand, where custodial mothers are psychologically able to provide a loving, effective parent–child relationship, children will be buffered from the stress divorce engenders and will tend to prosper developmentally (Kalter et al. 1989). However, when economic deprivation, interparental hostility and the burdens of single parenting take their toll on the mental health of custodial mothers, children will tend to fare less well.
The connection between marital separation and marital conflict is complex. Clearly the two factors are interrelated, in that at the time of a marital dissolution the separating partners are likely to be at odds and many are involved in serious conflict. Hanson (1999) reported that about half of all couples who divorced exhibited high levels of conflict beforehand, compared with about one-quarter of families who remained continuously married.
However, the connection between marital separation and marital conflict is not at all straightforward, since some partners manage to separate on relatively amicable terms, while many marriages survive for long periods despite the presence of ongoing conflict. Hanson (1999) found that approximately 75% of high-conflict couples chose not to divorce, indicating that, for the vast majority of children exposed to high levels of parental conflict, divorce is not an avenue through which their exposure to conflict is reduced, at least in the short term.
To understand the relationship between marital conflict and separation, it is important to distinguish between conflict that precedes the separation and conflict that follows the separation. Many families experience conflict both before and after separation, so it is not possible to draw a clear demarcation in this way. Nevertheless, in some cases a prolonged period of conflict is terminated when parents separate, while in other cases the separation itself provokes a round of conflict which may persist for years afterward.
The evidence about the impact of separation and pre-separation conflict is somewhat complex. First, both marital conflict and separation have been found to be independently associated with child outcomes. Peterson and Zill (1986) found that marital conflict in intact homes, especially if persistent, was as harmful as separation. Indeed, they found that scores on measures of overcontrolled and undercontrolled behaviour of children living amid persistent conflict were even higher than for those living with one biological parent.
However, many studies have also reported the presence of an interaction between separation and conflict, so that in high-conflict families children benefit when their parents divorce, while in low-conflict families children do worse when their parents divorce (Amato et al. 1995). Other studies show similar results, although with a twist. Hanson (1999) found that children exposed to low levels of parental conflict appeared to suffer disadvantages when their parents separated, although he also found that children exposed to high levels of parental conflict were neither better nor worse off, on average, when their parents divorced. Morrison and Coiro (1999) found that while their results did not indicate a benefit for children exiting high-conflict marriages (problem behaviours among children increased after separation regardless of the level of conflict that predated the separation), nevertheless the greatest increase in behaviour problems was observed among children whose parents remained married, despite very frequent quarrels.
All of these results indicate a complex relationship between marital conflict, separation and child outcomes. Taken together, they suggest that children in high-conflict families are likely to be better off, while children in low-conflict families are likely to be worse off, if their parents separate. Booth and Amato (2001) note that:
while escape from a high-conflict marriage benefits children because it removes them from an aversive, stressful home environment, in contrast a divorce that is not preceded by a prolonged period of overt discord may represent an unexpected, unwelcome, and uncontrollable event, an event that children are likely to experience as stressful.
On the other hand, the evidence about post-separation conflict is much more straightforward. The more conflict there is, and the more this involves the children, the more damaging it is to children’s wellbeing.
Bream and Buchanan (2003) found that, among a sample of children whose parents could not agree on arrangements for them, high proportions had significant adjustment problems: about half of both boys and girls immediately after the proceedings, and around two-thirds of boys and one-third of girls a year later. Children aged under seven were particularly vulnerable to such difficulties. Where half of such children were distressed at the time of the first interview, this had risen to 80% at the second. This equates to four times the rate that would be expected in the general population. Indeed, the rate of difficulties among these children was similar to that among a sample of children who were subject to care proceedings.
Johnston et al. (1985) report that children who are the subject of lengthy post-separation disputes between their parents have been identified as the most at-risk among the divorcing population. For this group the major benefit of the divorce – the cessation of parental hostilities – does not accrue. Johnston (1994) notes that children of high-conflict divorces scored as significantly more disturbed, and were two to four times more likely to have the kinds of adjustment problems typically seen in children being treated for emotional and behavioural disturbance, when compared with national norms.
Conflict takes different forms and some types of conflict are especially damaging for children. Hetherington (2003) found that parental conflict that is about the child or directly involves the child, conflict that is physically violent, threatening or abusive, and conflict in which the child feels caught in the middle between two warring parents have the most adverse consequences for children.
Even from this small selection of studies it seems clear that post-separation conflict between parents carries the risk of significant levels of adverse impacts on children.
Parenting and Parent–Child Relationships
Various studies have shown that separation and divorce lead to disruptions in parenting practices. Simons et al. (1999) found that the quality of the mother’s parenting mediated much of the association between divorce and child adjustment. In addition, the level of the father’s involvement in parenting explained part of the association between divorce and the externalising problems of boys. Compared with fathers in intact families, non-custodial fathers were less likely to provide their children with help in solving problems, to discuss standards of conduct or to enforce discipline. This reduced involvement in parenting was associated with an increased probability that a boy would display conduct problems. This suggests that a divorced father who remain actively involved as a parent may significantly reduce his son’s chances of conduct problems. Indeed, Simons et al. (1999) note that boys with divorced parents, where both parents exhibit competent parenting behaviours, are at no greater risk of involvement in delinquent behaviour than boys living in an intact family.
McLanahan and Bumpass (1994) investigated several hypotheses for the adverse childbearing and marital outcomes of children of divorced parents and concluded that parental role models and parental supervision were the major factors in determining the future family-formation behaviour of offspring. As they note, “it seems obvious that single parents would have more difficulty maintaining authority and control over daughter’s dating, which, in turn, is directly related to early family-formation behaviour”.
Holdnack (1993) notes that parental divorce interrupts the emotional closeness between parents and children, leading to negative impacts on children’s self-esteem. An unresolved issue is whether poor-quality family relationships arise as an effect of the divorce or whether these may have pre-dated (and perhaps given rise to) the divorce. This raises the possibility that the results reflect selection into divorce rather than demonstrating the effects of divorce. Sun (2001) found that, indeed, families on the verge of breakup are characterised by less intimate parent–parent and parent–child relationships, as well as less parental commitment to children’s education and fewer economic resources. Prior to the marital disruption, families that broke down showed consistent signs of dysfunction on every indicator of family environment examined. Sun (2001) concludes that this suggests that “a dysfunctional family environment serves as an important mechanism by which marital disruption process affects children”. However, it is also possible that this reflects selection effects. I turn now to an examination of such effects.
The discussion so far has assumed that the associations between parental separation and child outcomes are brought about through causal connections that link the former to the latter. However, it is also possible that the associations arise through non-causal mechanisms; in particular, through selection effects. The discussion now turns to an examination of such effects.
Several studies have demonstrated that many of the presumed effects of parental separation on children are evident many years in advance of the actual separation. Block et al. (1986) found the behaviour of boys as early as 11 years prior to parental separation to be characterised by undercontrol of impulse, aggression, and excessive energy. Elliott and Richards (1991) report that children whose parents divorced when they were between seven and 16 years old had worse scores on a range of measures of wellbeing than children whose parents remained married, not only at age 16 (after the separation) but also at age seven.
A question that arises is whether these results reflect the fact that the process of parental separation can take place over a long period (while some families break down quickly, often in spectacular ways with much heat, in other families the process is a longer and slower burn), or whether they result from selection effects (that is to say, some parents bring into a marriage a set of characteristics that are likely both to raise the possibility that the marriage will break down and to heighten the risk of adverse outcomes for their children). There are a number of characteristics that might perform such a role, such as poor mental health, antisocial behaviour and substance dependencies. Parents with such personal difficulties are likely to have greater difficulties both in maintaining stable and enduring intimate relationships and in providing their children with a family environment that is likely to promote their wellbeing. Part of the patterns of association between parental separation and child outcomes might therefore simply reflect the fact that some adults are not well equipped either to perform well as a marriage partner or as a parent. Furstenberg and Teitler (1994) note that:
Families that eventually divorce may be different in a variety of ways from those that do not long before marital disruption occurs. They may be more likely to exhibit poor parenting practices, high levels of marital conflict, or suffer from persistent economic stress ... exposure to these conditions may compromise children’s economic, social and psychological wellbeing in later life whether or not a separation takes place.
Sun (2001) found that, compared with parents that remain continuously married, parents who later divorce are more likely to have personal, sexual, psychological or financial problems throughout their marriage, and these problems continue to affect children negatively. Given the persistence of these problems, a separation may actually reduce the stress associated with such problems, resulting in relatively little further damage to child wellbeing.
Emery et al. (1999) found that while children from never-married and divorced families had higher rates of externalising behaviour problems, much of this could be explained by their mothers’ histories of delinquent behaviour in adolescence. In fact, delinquent behaviour reported when future mothers were single, childless adolescents prospectively predicted behaviour problems among their offspring 14 years later.
Thus, it appears that the contribution of divorce and its aftermath to children’s problems in later life is not nearly as great as might be inferred from findings that do not take adequate account of family conditions prior to the separation. Parental separation does not occur randomly, and the causes that underlie it may also be part of the explanation for the apparent impacts on children.
Genetic Transmission Mechanisms
One means by which selection effects might arise is via genetic transmission of characteristics and behaviours between parents and children. Studies of the impact of parental separation on children in adopted and biological families provide a window on this issue, since parents and children in biological families share both genes and environment, while parents and children in adoptive families share their environment but not their genes.
O’Connor et al. (2000) report some suggestive findings from such a study. They found that, while biological and adopted children who had experienced a parental divorce displayed similarly elevated rates of behavioural problems and substance use compared with their peers in intact families, a different pattern was found for academic and social competence outcomes. While children from biological families also had lower levels of academic achievement and social competence than their peers in intact families, there were no differences between adopted children in divorced and intact families. They note that “the findings for psychopathology are consistent with an environmentally mediated explanation for the association between parents’ divorce and children’s adjustment [while] the findings for achievement and social adjustment are consistent with a genetically mediated explanation”. These results show that if genetic mechanisms are involved they have differential effects in different spheres of development.
In a subsequent paper, O’Connor et al. (2003) reported that the association between genetic risk (as reflected in a measure of negative reactivity in the biological parents) and child adjustment among a sample of adopted children was moderated by parental separation. While genetic risk was uncorrelated with the adjustment of adopted children in intact families, among children who had experienced a parental divorce there were substantial and significant associations between genetic risk and poor adjustment. This result indicates a complex interplay between genetic and environmental factors: genetic risk only poses a problem for children’s wellbeing in the presence of an environmental stressor such as parental divorce.
It appears, then, that genetic factors do play a role in the association between parental separation and child outcomes, although their impact varies across different outcome domains and interacts with environmental triggers.
An Exemplar Intervention: The New Beginnings Program
A large question raised by the above results for policy makers and their advisers is what, if anything, might be done to alleviate the distress that parental separation causes in children’s lives. While consideration of this question is beyond the scope of the present paper, it will be useful to sketch out the promising results that have been achieved through one particular intervention, which indicates that there is indeed scope for effective action.
The exemplar intervention I have chosen to highlight is the New Beginnings Program in the United States, an intervention for custodial mothers following a separation, which was subject to a true experimental trial (Wolchik et al. 2002). The programme involved randomised assignment to one of two treatment conditions (a mother-only programme, involving 11 group sessions with other custodial mothers, plus two structured individual sessions and a dual component mother-plus-child programme, which also included 11 group sessions for the children) or a control condition. Participants who were assigned to the control condition were issued with books on adjustment to divorce. The sample was randomly drawn from divorce court records. Children in the study were followed up six years after the intervention.
Wolchik et al. (2002) reported that both treatments were found to yield significant benefits for the children who participated in them, compared with children in the control group. In particular, they exhibited reduced rates of mental disorders, reduced levels of externalising problems, reduced rates of substance abuse and reduced numbers of sexual partners. It appears, then, that it is possible to design interventions that afford children with significant protection from the adverse consequences of their parents’ divorce, at the expense of a dozen or so sessions of group treatment for mothers.
A number of conclusions can be drawn from this brief survey of the literature on parental separation and child outcomes. First, there is an abundance of evidence that children who experience a parental separation are, on average, worse off than their peers in intact families, on a number of measures of wellbeing. However, the scale of the differences in wellbeing between the two groups of children is not large and most children are not adversely affected. Parental separation then bears down most heavily on a minority of children, generally in the presence of other exacerbating factors.
Underlying these effects are multiple mechanisms: income declines following separation, declines in the mental health of custodial mothers, interparental conflict and compromised parenting. These mechanisms do not operate independently, but are related in complex ways. For example, income declines following separation place mother-headed households at risk of material and economic deprivation, which can take a toll on mothers’ mental health. This in turn can lead to compromised parenting behaviours. All of these factors can impact adversely on child wellbeing.
Part of the effects also arise from non-causal mechanisms: that is to say, not all of the adverse child outcomes following separation can be laid at the door of the separation itself. Many of the difficulties have deeper roots that date from many years prior to the separation and are due to the fact that some parents bring into a marriage characteristics and behaviours – such as poor mental health, antisocial behaviour or substance addictions – that are likely both to jeopardise the success of the marriage and heighten the risk of poor child outcomes. Furthermore, some of the associations between separation and child outcomes are due to genetic inheritance.
One factor that plays a more complex role is interparental conflict. Conflict between parents plays a dual role, both as part of the explanation for the link between parental separation and child outcomes and as an independent influence on child outcomes. It is clear, nevertheless, that post-separation conflict which is bitter and ongoing and which places the children at the centre of disputation has highly malign effects on child wellbeing.
Yet this is a factor which is surely amenable to treatment. If separating couples can be helped to reduce levels of conflict following a separation, or at least to understand the importance of conducting their affairs out of the way of the children and in ways that do not implicate them, then this is likely to have significant benefits for the wellbeing of the children. As Moxnes (2003) notes, “extensive parental cooperation is ... the most important means by which to reduce the negative effects of divorce for children.”
The evidence from the evaluation of the New Beginnings Program shows that it is possible to design programmes aimed at ameliorating the negative fallout from a parental separation that yield real benefits for children, in terms of their mental health, behaviour and general wellbeing. This suggests it would be useful to conduct further investigations to identify promising approaches that afford children protection from a parental separation that could be considered for trial in the New Zealand context.
Allison, Paul D. and Frank F. Furstenberg, Jr. (1989) “How marital dissolution affects children: Variations by age and sex” Developmental Psychology, 25(4):540-549.
Amato, Paul R. (1993) “Children’s adjustment to divorce: Theories, hypotheses, and empirical support” Journal of Marriage and the Family, 55:23-38.
Amato, Paul R. (1987) “Family processes in one-parent, stepparent, and intact families: The child’s point of view” ā, 49(2):327-337.
Amato, Paul R. and Danelle D. DeBoer (2001) “The transmission of marital instability across generations: Relationship skills or commitment to marriage?” Journal of Marriage and Family, 63:1038-1051.
Amato, Paul R. and Bruce Keith (1991) “Parental divorce and adult well-being: A meta-analysis.” Journal of Marriage and the Family, 53(1):43-58.
Amato, Paul R., Laura Spencer Loomis and Alan Booth (1995) “Parental divorce, marital conflict, and offspring well-being during early adulthood” Social Forces, 73(3):895-915.
Aquilino, William S. (1994) “Impact of childhood family disruption on young adults’ relationships with parents” Journal of Marriage and the Family, 56(2):295-313.
Aquilino, William S. (1996) “The life course of children born to unmarried mothers: Childhood living arrangements and young adult outcomes” Journal of Marriage and the Family, 58:293-310.
Baydar, Nazli (1988) “Effects of parental separation and reentry into union on the emotional well-being of children” Journal of Marriage and the Family, 50(4):967-981.
Biblarz, Timothy J. and Greg Gottainer (2000) “Family structure and children’s success: A comparison of widowed and divorced single-mother families” Journal of Marriage and the Family, 62(2):533-548.
Block, Jack, Jeanne H. Block and Per F. Gjerde (1988) “Parental functioning and the home environment in families of divorce: Prospective and concurrent analyses” Journal of the American Academy of Child and Adolescent Psychiatry, 27(2):207-213.
Block, Jeanne H., Jack Block and Per F. Gjerde (1986) “The personality of children prior to divorce: A prospective study” Child Development, 57:827-840.
Blum, Heather Monroe, Michael H. Boyle and David R. Offord (1988) “Single-parent families: Child psychiatric disorders and school performance” Journal of the American Academy of Child and Adolescent Psychiatry. 27(2):214-219.
Booth, Alan and Paul R. Amato (2001) “Parental predivorce relations and offspring postdivorce well-being” Journal of Marriage and Family, 63:197-212.
Bream, Victoria and Ann Buchanan (2003) “Distress among children whose separated or divorced parents cannot agree arrangements for them” British Journal of Social Work, 33:227-238.
Burns, Ailsa, Rosemary Dunlop and Alan Taylor (1997) “Children and parental divorce: The meaning of small effects” Australian Psychologist, 32(2):136-138.
Butler, Ian, Lesley Scanlan, Margaret Robinson, Gillian Douglas and Mervyn Murch (2002) “Children’s involvement in their parents’ divorce: Implications for practice” Children and Society, 16:89-102.
Carlson, Marica J. and Mary E. Corcoran (2001) “Family structure and children’s behavioral and cognitive outcomes” Journal of Marriage and Family, 63:779-792.
Chase-Lansdale, P. Lindsay, Andrew J. Cherlin and Kathleen E. Kiernan (1995) “The long-term effects of parental divorce on the mental health of young adults: A developmental perspective” Child Development, 66:1614-1634.
Chase-Lansdale, P. Lindsay and Mavis E. Hetherington (1990) “The impact of divorce on life-span development: Short and long-term effects” in P. Baltes, D. Featherman and R. Lerner (eds.) Life Span Development and Behavior, Lawrence Erlbaum, Hillsdale, New Jersey.
Dawson, Deborah A. (1991) “Family structure and children’s health and well-being: Data from the 1988 National Health Interview Survey on Child Health” Journal of Marriage and the Family, 53:573-584.
Demo, David H. and Alan C. Acock (1988) “The impact of divorce on children” Journal of Marriage and the Family, 50(3):619-648.Demo, David H. and Alan C. Acock (1996) “Family structure, family process, and adolescent well-being” Journal of Research on Adolescence, 6(4):457-488.
Dunn, Judy, Kirby Deater-Deckard, Kevin Pickering and Thomas G. O’Connor (1998) “Children’s adjustment and prosocial behaviour in step-, single-parent, and non-stepfamily settings: Findings from a community study” Journal of Child Psychology and Psychiatry, 39(8):1083-1095.
Elliott, Jane B. and Martin P.M. Richards (1991) “Children and divorce: Educational performance and behaviour before and after parental separation” International Journal of Law and the Family, 5:258-276.
Ellis, B.J., J.E. Bates, K.A. Dodge, D.M. Fergusson, L.J. Horwood, G.S. Pettit and L.J. Woodward (2003) “Does father absence place daughters at special risk for early sexual activity and teenage pregnancy?” Child Development, 74(3):801-821.
Emery, Robert E., M. Waldron, Katherine M. Kitzman and J. Aaron (1999) “Delinquent behavior, future divorce or nonmarital childbearing, and externalizing behavior among offspring: A 14-year prospective study” Journal of Family Psychology, 13(4):568-79.
Ermisch, John F. and Marco Francesconi (2001) “Family structure and children’s achievements” Journal of Population Economics, 14:249-270.
Evans, M.D.R., Jonathan Kelley and Richard A. Wanner (2001) “Educational attainment of the children of divorce: Australia, 1940-90” Journal of Sociology, 37(3):275-297.
Fergusson, D.M., M.E. Dimond and L.J. Horwood (1986) “Childhood family placement history and behaviour problems in 6-year-old children” Journal of Child Psychiatry, 27(2):213-226.
Fergusson, David M., L. John Horwood and Michael T. Lynskey (1994) “Parental separation, adolescent psychopathology, and problem behaviors” Journal of the American Academy of Child and Adolescent Psychiatry, 33(8):1122-1131.
Fergusson, David M., Michael T. Lynskey and L. John Horwood (1994) “The effects of parental separation, the timing of separation and gender on children’s performance on cognitive tests” Journal of Child Psychology and Psychiatry, 35(6):1077-1092.
Fine, Mark A., Voydanoff, Patricia & Donnelly, Brenda W. (1993) “Relations between parental control and warmth and child well-being in stepfamilies” Journal of Family Psychology, 7(2): 222-232.
Furstenberg, Frank Jr. and Julien O. Teitler (1994) “Reconsidering the effects of marital disruption: What happens to children of divorce in early adulthood?” Journal of Family Issues, 15(2):173-190.
Hanson, Thomas L. (1999) “Does parental conflict explain why divorce is negatively associated with child welfare?” Social Forces, 77(4):1283-1315.
Hetherington, E. Mavis (2003) “Social support and the adjustment of children in divorced and remarried families” Childhood, 10(2):217-236.
Hetherington, E. Mavis, Margaret Bridges and Glendessa M. Insabella (1998) “What matters? What does not? Five perspectives on the association between marital transitions and children’s adjustment” American Psychologist, 53(2):167-184.
Hetherington, E. M., M. Cox and R. Cox (1982) “Effects of divorce on parents and children” in M. Lamb (ed.) Nontraditional Families, Erlbaum, Hillsdale, New Jersey.
Holdnack, James A. (1993) “The long-term effects of parental divorce on family relationships and the effects on adult children’s self-concept” in Craig A. Everett (ed.) Divorce and the Next Generation: Effects on Young Adults’ Patterns of Intimacy and Expectations for Marriage, Haworth Press, Binghamton, New York.
Jaffee, Sara R., Terrie E.Moffitt, Avshalom Caspi and Alan Taylor (2003) “Life with (or without) father: The benefits of living with two biological parents depend on the father’s antisocial behavior” Child Development, 74(1):109-126.
Jekielek, Susan M. (1998) “Parental conflict, marital disruption and children’s emotional well-being” Social Forces, 76(3):905-936.
Johnston, Janet R. (1994) “High-conflict divorce” The Future of Children, 4(1):165-182.
Johnston, Janet R., Linda E.G. Campbell and Sharon S. Mayes (1985) “Latency children in post-separation and divorce disputes” Journal of the American Academy of Child Psychiatry, 24(5):563-574.
Joshi, Heather, Elizabeth C. Cooksey, Richard D. Wiggins, Andrew McCullough, Georgia Verropoulou and Lynda Clarke (1999) “Diverse family living situations and child development: A multi-level analysis comparing longitudinal evidence from Britain and the United States” International Journal of Law, Policy and the Family, 13:292-314.
Kalter, Neil, Amy Kloner, Shelley Schreier and Katherine Okla (1989) “Predictors of Children’s Postdivorce Adjustment” American Journal of Orthopsychiatry, 59(4):605-618.
Keith, V.M. and B. Finlay (1988) “The impact of parental divorce on children’s educational attainment, marital timing, and likelihood of divorce” Journal of Marriage and the Family, 50:797-809.
Kelly, Joan B. (1993) “Current research on children’s postdivorce adjustment: No simple answers” Family and Conciliation Courts Review, 31(1):29-49.
Kelly, Joan B. (2003) “Changing perspectives on children’s adjustment following divorce: A view from the United States” Childhood, 10(2):237-254.
Kurdek, Lawrence A., Mark A. Fine and Ronald J. Sinclair (1994) “The relation between parenting transitions and adjustment in young adolescents: A multisample investigation” Journal of Early Adolescence, 14(4):412-432.
McLanahan, Sara S. and Karen Booth (1989) “Mother-only families: Problems, prospects, and politics” Journal of Marriage and the Family, 51(3):557-580.
McLanahan, Sara S. and Larry Bumpass (1994) “Intergenerational consequences of family disruption” American Journal of Sociology, 1:130-152.
Mitchell, Barbara A., Andrew V. Wister and Thomas K. Burch (1989) “The family environment and leaving the parental home” Journal of Marriage and the Family, 51(3):605-613.
Morrison, Donna Ruane and Mary J. Coiro (1999) “Parental conflict and marital disruption: Do children benefit when high-conflict marriages are dissolved?” Journal of Marriage and the Family, 61:626-637.
Moxnes, Kari (2003) “Risk factors in divorce: Perceptions by the children involved” Childhood, 10(2):131-146.
Mueller, Charles W. and Hallowell Pope (1977) “Marital instability: A study of
its transmission between generations” Journal of Marriage and the Family,
39(1):83-93.O’Connor, Thomas G., Avshalom Caspi, John C. DeFries and Robert Plomin (2000) “Are associations between parental divorce and children’s adjustment genetically mediated? An adoption study” Developmental Psychology, 36(4):429-437.
O’Connor, Thomas G., Avshalom Caspi, John C. DeFries and Robert Plomin (2003) “Genotype–environment interaction in children’s adjustment to parental separation” Journal of Child Psychology and Psychiatry, 44(6):849-856.
O’Connor, Thomas G., Judy Dunn, Jennifer M. Jenkins, Kevin Pickering and Jon Rasbash (2001) “Family settings and children’s adjustment: Differential adjustment within and across families” British Journal of Psychiatry, 179:110-115.
Peterson, James L. and Nicholas Zill (1986) “Marital disruption, parent–child relationships, and behavior problems in children” Journal of Marriage and the Family, 48:295-307.
Pryor, Jan and Rodgers, Bryan (2001) Children in Changing Families: Life After Parental Separation, Blackwell Publishers, Oxford.
Ross, Catherine E. and John Mirowsky (1999) “Parental divorce, life-course disruption, and adult depression” Journal of Marriage and the Family, 61:1034-1045.
Simons, Ronald L., Kuei-Hsiu Lin, Leslie C. Gordon, Rand D. Conger and Frederick O. Lorenz (1999) “Explaining the higher incidence of adjustment problems among children of divorce compared with those in two-parent families” Journal of Marriage and the Family, 61:1020-1033.
Sun, Yongmin (2001) “Family environment and adolescents’ well-being before and after parents’ marital disruption: A longitudinal analysis” Journal of Marriage and Family, 63:693-713.
Teachman, Jay D. (2002) “Childhood living arrangements and the intergenerational transmission of divorce” Journal of Marriage and Family, 64:717-729.
Tucker, Joan S., Howard S. Friedman, Joseph E. Schwartz, Michael H. Criqui, Carol Tomlinson-Keasey, Deborah L. Wingard and Leslie R. Martin (1997) “Parental divorce: Effects on individual behavior and longevity” Journal of Personality and Social Psychology, 73(2):381-391.
Videon, Tami M. (2002) “The effects of parent-adolescent relationships and parental separation on adolescent well-being” Journal of Marriage and Family, 64:489-503.
Wallerstein, Judith S. and Shauna B. Corbin (1989) “Daughters of divorce: Report from a ten-year follow-up” American Journal of Orthopsychiatry, 59(4):593-604.
Wolchik, Sharlene A., Irwin N. Sandler, Roger E. Millsap, Brett A. Plummer, Shannon M. Greene, Edward R. Anderson, Spring R. Dawson-McClure, Kathleen Hipke and Rachel A. Haine (2002) “Six-year follow-up of preventive interventions for children of divorce” Journal of the American Medical Association, 288(15):1874-1881.
Woodward, Lianne, David M. Fergusson and L. John Horwood (2001) “Risk factors and life processes associated with teenage pregnancy: Results of a prospective study from birth to 20 years” Journal of Marriage and Family, 63:1170-1184.
Wu, Lawrence L. (1996) “Effects of family instability, income, and income instability on the risk of a premarital birth” American Sociological Review, 61:386-406.
1 Much of the research literature on stepfamilies focuses on stepfather families, because, following a separation, most children live with their mother