ACKNOWLEDGEMENTS

EXECUTIVE SUMMARY

PART A: CUSTODY, VISITATION, AND CHILD WELL-BEING A SYNTHESIS OF THE LITERATURE
Introduction
Children's Development
The Role of the Father
Ways in which Family Structure May Affect Children's Lives
Loss of a Parent

Weaknesses and Gaps in Literature
Policy Ramifications
Bibliography

PART B. ANNOTATED BIBLIOGRAPHY: SELECTED CHILD SUPPORT ARTICLES

PART C: SELECTED BIBLIOGRAPHY ON CUSTODY VISITATION, AND CHILD WELL-BEING
Consequences for Children
Divorce
Family Structure/Single Parent Families
Remarriage/Step-Families
Parental Conflict/Family Processes
Parental Functioning
General and in Divorced Families
Poor Families
Parental Involvement After Marital Disruption
Custody Arrangements
Child Support
Mediation
Developmental Needs of Children
Fatherhood/Fathers
Fathers (general)
Young, Unwed, or Poor Fathers
Theories
Marital Disruption
Families and Society
Determinants of Marital Disruption
Other Related Research

• Twenty-one percent or 1.3 million custodial parents with formal written child support agreements report that they have a joint custody arrangement. Of these, over 1 million (80%) have a joint legal only arrangement. The remaining 262 thousand report that they have a joint legal and physical custody arrangement.



• A large minority of custodial parents (45%) have never had a child support agreement of any type. About 14 percent of custodial parents without a written award report that the nonresident parent provided child support or non-cash assistance in the previous year.



• Court-ordered agreements are the most common type of written agreement. Seventy-three percent of custodial parents with written agreement report that they had court-ordered agreements. Twenty-three percent report that the agreements were voluntary ones ratified by the court. Three percent report that they had some other type of written agreement, such as one not ratified by the court.



• Nearly two-thirds (64.6%) or almost 4 million custodial parents with written agreements report that their agreements contain visitation provisions.



• According to the reports of custodial mothers with written agreements, nearly one4hird (32%) of non-resident fathers have not spent time with their children in the previous 12 months. However, nearly one-quarter (24%) of non-resident fathers see their children at least once a week.



• Non-resident mothers are more likely to visit their children and to see them more often than non-resident fathers. Sixteen percent of non-resident mothers had not visited their children in the past year compared to 32 percent of non-resident fathers. Thirty-five percent of nonresident mothers saw their children once a week or more compared to 24 percent of nonresident fathers.



• Custodial parents with written child support agreements who were owed child support, received about 65 percent of what they were due. Parents with voluntary agreements received 73 percent of what they were due, while those with court-ordered agreements received 62 percent of what they were due. Parents living in the same city or county as the non-resident parent received 70 percent of what was due compared to 58 parent if the non-resident parent lived in a different state.

• Encouraging parents to establish child support agreements through a process of bargaining and mutual agreement, whenever feasible, rather than through litigation and court mandate, and providing services, if needed, to assist in the process. Multivariate models indicated that even after controlling for background characteristics, fathers who had a voluntary written agreement ratified by the court maintained more contact with their children, were more likely to pay some child support, complied more fully with the child support orders, and paid greater amounts of child support than fathers who had court-ordered child support agreements.
  
  
• Encouraging couples to specify visitation provisions in their agreements. Nonresidential fathers had significantly more contact with their children when the child support agreement had an explicit provision specifying the frequency and schedule of visitation than when such a provision was not present in the agreement. The existence of a visitation provision, however, had no direct effect on the payment of child support.
  
  
• Encouraging and facilitating contact between non-resident fathers and their children, when feasible. The SIPP analyses suggest that continued contact with the non-resident father has a beneficial influence on older teens and young adults. Moreover contact had a positive association with both the payment of child support and with compliance with child support orders. Although this relationship diminished with the addition of past child support behavior to the model, the positive sign remained suggesting that contact is not hindering the payment of child support and is actually exerting pressure towards the payment of child support.
  
  
• Promoting joint custody arrangements. The results of the analyses were broadly supportive of arguments for joint custody, though the influence of joint custody differed somewhat depending upon whether it was joint legal or joint legal and physical. Nonresidential fathers with joint legal and physical custody were more likely to have paid some child support in the current year (significant at .10 level) and to have complied more fully with their support obligations than fathers in other arrangements. There was no difference in the probability of paying support or in the degree of compliance with the child support orders between fathers with joint legal only custody arrangements and other fathers. However, fathers with joint legal only arrangements paid larger amounts of support than other fathers, even after controlling for the demographic and socioeconomic characteristics of the custodial parent and other possible mediating factors.

• All persons who are involved in working with single parent families and with divorcing families need to be made more aware of the developmental needs of children, the potential difficulties that they will face from family disruption or turmoil, and steps that could ease those difficulties.
  
  
• Given that conflict if inappropriately handled can be harmful to children and can alienate the two adults so that they cannot cooperate over matters concerning the child's well-being, it is important to provide services to reduce conflict or to express it in a healthier manner.
  
  
• Several studies showed that all the parties in a divorce experience stress. The stress can adversely affect the custodial parent's ability to function effectively; it can serve to inhibit the non-custodial parent from remaining involved in the child's life; and it may affect how the child adjusts to family disruption and turmoil. More research and training on how to reduce stress is needed.
  
  
• Given that most children desire the continuing presence of a father in their lives and that fathers may disengage from their parental responsibilities in part because they feel no sense of control over the new arrangements, steps should be taken to enable fathers when it is at all feasible to have a more active post-divorce role.
  
  
• There is a scarcity of research that has examined couples who never establish awards and couples who never married each other. Such couples are particularly vulnerable economically and the children face risks as well.

WESTAT
An Employee-Owned Company

MEMORANDUM


TO:Anne Benson and Linda MellgrenAugust 20, 1996

FROM:Christine Nord

SUBJECT:Final Report

---

Enclosed are ten bound copies of the final report, a camera ready copy of Volumes I and II, an unbound copy of the Supplemental Tables, and a disk containing copies of Volumes I and II and the Supplemental Tables.

It was quite satisfying to me to actually hold the bound copies in my hand. I now have them on my shelf and intend to use them (especially the Supplemental Tables) as a handy reference. I hope that they will be useful to you.

I enjoyed working with you and look forward to doing so again in the future.

---

PART A: CUSTODY, VISITATION, AND CHILD WELL-BEING
A SYNTHESIS OF THE LITERATURE

Christine Winquist Nord and Nicholas Zill
Westat, Inc.

Introduction

The family situations in which American children are cared for and raised have changed dramatically in the last 30 years. The number of children who live with only one parent, with a birth parent and a stepparent, or with neither of their parents has increased substantially. In 1981, 33 percent of all children under 18 in the United States did not live in a traditional two-parent family, i.e., with both of their biological parents. By 1992, that figure had risen to 43 percent of children. Estimates are that at least half of all children today will spend some time in a single-parent family before they reach age 18 (Bumpass and Raley, 1993; Furstenberg and Cherlin, 1991; Furstenberg et al., 1983).
Currently, more than 40 percent of all first marriages end in divorce (Norton & Miller, 1992). The U.S. divorce rate doubled between the late 1960s and the late 1970s. The rate has stabilized and even declined slightly since then, but remains at very high levels (National Center for Health Statistics, 1994). Each year, more than 1.5 million children -- nearly 2.5 percent of all U.S. children -- undergo the painful experience of having their parents separate or become divorced (Bianchi & McArthur, 1991; Zill & Nord, 1994). Following separation or divorce, a sizable proportion of children lose regular contact with the non-custodial parent (Seltzer, 1991a; Furstenberg et al., 1983; Furstenberg and Nord, 1985; Seltzer and Bianchi, 1988). In the National Survey of Families and Households, nearly a third of the children had seen their non-resident parent only once or not at all in the preceding year, and only one in four averaged weekly contact or better (Seltzer, 1991a).
In addition to divorce, the number of children born to unmarried mothers has grown in recent years. In 1993, the nationwide count of such births amounted to 1.2 million, or 31 percent of all births. This was nearly triple the percentage of births that occurred outside of marriage in 1970. Among births to white women, the percentage that occurred outside marriage quadrupled between 1970 and 1993, going from 5.5. percent to 24 percent. Among births to black women, the unmarried percentage rose from 38 percent in 1970 to 69 percent in 1993. There are signs that unmarried birth rates are starting to level off; birth rates to unmarried women have remained stable for the past three years at 45.3 births per 1,000 unmarried women (Ventura et al., 1995). If the proportion of unmarried women in the population does not increase, the number of births to unmarried mothers should soon level off, as well. Many children born to an unmarried mother have little regular contact with their fathers (Seltzer, 1991a; Furstenberg et al., 1983). In the National Survey of Families and Households, nearly half the children born to an unmarried mother had seen their father only once or not at all in the last year (Seltzer, 1991a).
These trends in marriage and divorce have had negative consequences for the overall wellbeing of young people who grow up in single-parent families, stepfamilies, or other non4raditional family types. Over the last decade and a half, social scientists have generated a considerable body of solid research on family disruption and children's well-being. This rich literature has been reviewed and synthesized by a number of researchers (e.g., Hetherington, 1979, 1981; Longfellow, 1979; Emery 1982, 1988; Chase-Lansdale and Hetherington, 1990; Furstenberg and Cherlin, 1991; Wallerstein, 1991, Amato and Keith 1991a, 199b, Amato, 1993, 1994; Kelly, 1993; McLanahan and Sandelur, 1994). There is widespread agreement that, on average, children are better off financially, psychologically, and emotionally when they are raised by two parents. Children who grow up in single-parent families are much more likely than children living in two-parent families to develop an array of problems, including behavior problems in school, emotional problems, and academic difficulties (Zill and Schoenborn, 1990). Children raised in step-families are not immune to these problems. They, too, show an elevated risk of maladjustment and school failure (Zill, 1988).
It is also widely agreed, however, that the effects of divorce should not be overstated. Although they are real, they are not inevitable nor are they necessarily long-lasting. The majority of children whose families are disrupted by divorce show no adverse signs several years later, though they may go through a crisis period in the two-year period following the divorce (Morrison and Cherlin, 1995; Hetherington, 1979). During the crisis period the children may exhibit behavioral problems and difficulties in social and psychological functioning. Generally, however, these difficulties diminish and disappear for most children within two or three years of the divorce. At the same time, it is important not to understate the effects of divorce: researchers have found effects of marital disruption 12 to 22 years later in such outcomes as poor relationships with parents, increased levels of problem behavior, increased likelihood of dropping out of school and receiving psychological help, lower likelihood of attending college, an increased likelihood among young men of being idle (i.e., not enrolled in school or employed), and an increased risk of teenage childbearing among young women (McLanahan and Sandefur, 1994; Zill et al., 1993). Although the overall likelihood of these outcomes was relatively small, for some outcomes, such as dropping out of school, the risk was doubled, which could have considerable impact on the particular individuals affected. As Zill and his colleagues also note, the fact that divorce is associated with a poorer relationship with one's parents in young adulthood is a cause for societal concern. These young people will be more susceptible to negative influences around them than they would be if they had strong relationships with their parents. It should be pointed out that most of this research pertains to children of divorce. Relatively little research has been conducted that examines the psychological and social development of children born to women who were not married at the time of the birth. Although there is a rich literature about the economic insecurity experienced by single mothers and their children and its effects on their development and well-being, there is a scarcity of research that has tried to separate out the effects of single parenthood from the effects of poverty.
In spite of the extensive research conducted thus far, much remains to be learned, especially about why so many non-custodial parents become disengaged from their children's lives or take little or no responsibility and the effects of different custody arrangements on the continued involvement of the non-residential parent and on the well-being of children. More research is also needed on the consequences for children if their parents never marry. Because many good syntheses of the research on divorce and children's well-being already exist, this paper focuses on the connections between custody arrangements, child support payments, parental involvement, and children's well-being in both the divorced and never-married populations. The literature on these topics is less extensive and the results more mixed (Kelly, 1993).
To understand how parental involvement, custody arrangements, and child support interact to affect child well-being, it would be useful to have a theoretical model that suggested the specific pathways of influence. Unfortunately, very little theoretical work has been done in this area (Kurdek, 1993). Researchers, however, have offered a variety of explanations for why divorce may affect children's lives. At least some of the reasons given may also apply to children whose parents never married. Among the explanations are loss of a parent (decrease in quantity and quality of contact with the noncustodial parent or feelings of loss, abandonment, and worthlessness stemming from the decreased interaction with the parent), poor adjustment and decrease in parenting skills of the parents (particularly the custodial parent), parental conflict, economic hardship, stressful life changes, and the loss of other resources such as social supports (Amato, 1993, 1994; McLanahan and Sandefur, 1994; Kelly, 1993). Child custody and child support influences on child well-being are likely to transmit any effects that they have through one or more of the above forces.
Also important in understanding the potential effects of custody arrangements and child support payments on parental involvement and on children's well-being is an understanding of child development. The research on divorce indicates that effects may differ depending upon the age and sex of the child, with preschool children and boys more likely to exhibit short-term effects and older children and boys more likely to exhibit long-term effects. More consistent results have been found by age of the child than by sex of the child (Amato, 1994; Allison and Furstenberg, 1989; Zaslow, 1988, 1989; Zill et al, 1993). Both age and sex are proxies for the developmental pathways and needs of the children. It may well be the case that the optimal custody arrangements for a child change with the developmental stage of the child.
We adopt a child development perspective in examining the research results. Although fathers are increasingly being awarded custody of their minor children in divorce cases, it is still far more common for children to live with their mothers (U.S. Bureau of the Census, 1995). In instances where the parents never married, children are much more likely to reside with their mothers or perhaps with grandparents. Because the majority of non-custodial parents are fathers, it is important to understand the role that fathers play in children's lives and how that role changes as children grow older. We begin the discussion with an overview of children's development and of the role of the father. We then discuss the different reasons that have been offered to explain why family disruption may affect children's lives. To highlight that some family problems are not unique to disrupted families, we discuss how children's well-being can be affected by these forces even in intact families. As part of this discussion, we discuss what the research reveals about the relationship between child custody arrangements, child support payments, these forces, and children's well-being. We conclude this paper with a discussion of the policy implications of the synthesis.

Children's Development

As children grow older they change in many ways: physically, socially, cognitively, and emotionally (Hetherington and Parke, 1993). Developmentalists often identify four broad stages of development: infancy (from birth to age 2), early childhood (ages 3 to 5), middle childhood (ages 6 to 11) and adolescence (ages 12 to 17). Although the ages in parentheses should not be taken as fixed transition points, children's physical capacities, cognitive abilities, moral reasoning abilities, and social and emotional development undergo distinct transformations within each of these periods. There may be critical or sensitive periods in which children are more susceptible to certain stresses. For example, research on adoption suggests that children who are adopted prior to age two develop normally and exhibit few problems regardless of the conditions they lived in prior to age two, whereas children adopted after age two exhibit more behavior problems and other difficulties (Clarke and Clarke, 1976; Zill, 1995 ). It has been suggested that preschoolers (ages 3 to 5) may be especially vulnerable to the stresses of familial divorce because of their cognitive and social immaturity (Hetherington, 1981).
According to the ecological systems theory of child development, children's development is the result of a complex interplay between the child's own natural endowment and characteristics and a variety of social systems and environmental influences that move from immediate influences such as the child's home to more distal influences such as the child's community (Bronfenbrenner, 1986). Those factors closest to the child will have the most direct and powerful influence on the child, especially when the child is young. The mother, in particular, because she is usually the one who has the most intimate and direct contact with the child, is a central figure in understanding children's development and has been the focus of much study. In general, internal family processes, the psychological and physical health of the parents, and the relationship of the parents with each other and with the young child are more important than influences from outside the household, though factors external to the household will exert some influence through the effects they have on the household and on the dynamics among persons within the household. As children grow older and establish contacts outside the household, influences external to the household will gain in importance.
This theory recognizes that children's development is molded in part by characteristics of the child such as the child's sex, age, temperament, and psychological adjustment. Fussy infants can contribute to parent irritability which can lead to poorer parenting, setting up a cycle that is not conducive to a healthy parent-child relationship or to healthy child development. Research on resiliency suggests that individual characteristics including age, sex, genetic factors, temperament, and other problem-solving skills are also important determinants of how an individual responds to stressful events (Rutter, 1979, 1983, 1987; Garmezy, 1985).
At a minimum, children need to be protected from disease and other hazards in their environnment, and to receive proper nutrition and rest. Beyond these basics, research on attachment indicates that children also need to have a continuous, warm, and loving relationship with at least one parent and preferably with two. Attachment theory has most frequently been studied in infants and referred to the child's attachment to the mother. Recently, however, the theory has been expanded to include children of older ages and to look at the child's relationship with the father, as well (Hetherington and Parke, 1993). Research in a variety of fields indicates that parental involvement increases children's well-being on an array of measures from academic achievement to being less prone to participate in delinquent activities. Research also suggests that children benefit, cognitively at least, when they have two highly involved parents instead of just one (Lamb, 1986). One reason for this may be that mothers and fathers tend to have very different styles of interacting with their children. Children may be stimulated by the diversity of the parenting styles (Lamb, 1986). It may also be that each parent can respond better to particular aspects of the child's personality than the other.
Children's development can be enhanced or impeded by others in the child's life and by the child's general environment. Researchers studying resiliency in children have found that having close social networks, close personal relationships, and close relationships with larger groups and institutions help protect children from the ill effects of stressful life events (Rutter, 1983). Thus, having a close relationship with others may compensate to some extent for a poor relationship with one or both parents.
Ecological systems theory, however, explicitly acknowledges that children are influenced not only by direct interactions with others, but also by the interactions that the others have with each other. The child benefits, for example, if the two parents have a harmonious relationship or if the parents have a strong connection to the child's school and teachers (Bronfenbrenner, 1986). Research supports this view: children thrive when their parents love one another as well as the child. One reason for this is that the climate of the household is more harmonious and the two parents can provide emotional and physical support to each other (Hetherington and Parke, 1993). Coleman's concept of social capital complements this perspective (Coleman, 1990). Social capital is embedded in the relations that exist among persons. The more connections that there are within a family and between members of the family and persons and institutions in a community, the greater the social capital, and the greater the ability to transfer human, financial, and other types of resources to individuals within the family.
Children also seem to benefit from continuity in their environment, predictable schedules, and consistent, firm (but not harsh) discipline (Miller, 1970; Crockenberg and Litman, 1990; Patterson, Reid, and Dishion, 1989). Structuring home routines, for example, is associated with higher achievement and reading scores in school (Miller, 1970). Reasonable and consistent discipline is associated with children who are more cooperative and who are more likely to internalize the standards of their parents and are less likely to exhibit behavior problems (Crockenberg and Litman, 199O; Patterson, Reid, and Dishion, 1989). All of these activities are more difficult in single parent households. Psychological research also suggests that children's sex-role development is enhanced by e presence of the same sex parent. Of particular concern are preschool boys who lose their fathers at a when they are beginning the process of same sex identification.
Children's development is in many ways a cumulative process. Experiences that a child has earlier can continue to influence the child later. Rutter in his research on resiliency found that chronic stresses, such as chronic poverty, may itself be a factor in how a child responds to an acute stressful event, such as the divorce of one's parents. A single stress may have no undue consequences, but two or more stresses may greatly increase the chances of adverse outcomes (Rutter, 1979).
These perspectives on child development carry implications for research on the consequences of marital disruption for children's lives. Factors that this research indicate should be important are the age of the child at the time of disruption, other characteristics of the child including his or her experiences prior to the disruption, the psychological well-being of the primary caretaker, the dynamics between the two parents and between each parent and the child, and the extent to which continuity in the child's day-to-day routines and surroundings are maintained following the disruption.
First, child development theory would lead one to expect young children whose lives are more narrowly focused within the household to be more affected than older children. Infants, however, may escape some of the negative effects. On the other hand, to the extent that younger children are more dependent on a single primary caregiver, they may feel the loss of the second parent less than older children, particularly if that single parent can continue to function well. Although many studies on divorce have detected differences by age of child at the time of disruption, systematic studies that attempt to disentangle age at disruption from duration since disruption are still needed. Moreover, relatively large age groups have been used so that the potentially distinctive effects on infants, toddlers, school-aged children, and adolescents, and young adults cannot be readily assessed (Kurdek, 1993). The way in which other characteristics of the child affect their response to divorce has not been adequately addressed in the research. Second, the psychological well-being of the primary caretaker can directly influence children's lives. Research should examine more carefully factors that increase the psychological health of the caretaking parent and those that cause it to deteriorate. Third, to the extent that routines and familiar surroundings can be maintained, children should fare better. None of the studies which we examined documented the extent to which children's routines are disrupted and whether such disruptions do influence the children's response to divorce. John Guidubaldi and his colleagues in one study did find that children in disrupted families who had more consistent home routines were better adjusted two years later compared to similar children with less consistent routines (Guidubaldi et al., 1986). Fourth, attempts should be made to minimize the number and types of stresses that each child experiences due to the marital disruption. Again, however, to date we do not have a good sense of whether consequences for children vary by the number and types of stresses present.
In determining appropriate custody arrangements, parents and the courts need to be more sensitive to children's developmental needs. In a family with several children, it is quite possible to have children in different developmental stages. Regardless, the parties involved in determining custody should try to consider the different needs of each of the children. It appears rare, however, for judges who preside over family court cases to receive specific training that would allow them to be sensitive to the developmental needs of the children in making their decisions (Kelly, 1994).

The Role of the Father

To understand how the role of the father changes in divorced families or in families in which the father was never a member of the child's household, it is important to understand the role of father in intact families first. Only then can changes in his role be truly assessed.
Fathers are an important source of financial support for their families, but they also provide psychological and emotional support to the mothers, and can provide physical support to them as well by sharing in the housework and other duties associated with running a household. They also serve as models for their children and affect their children's lives directly through their interaction with them whether it be through play, helping with homework, or acting as their confidants. Fathers also, like mothers, serve as interpreters of community and family norms, teaching children which behaviors are acceptable and which are not. They also serve as a disciplinary backup to the mother. In some circumstances, particularly for older children, the physical strength of the father may make a difference when it comes to enforcing rules. In addition, fathers may be able to understand and respond to certain aspects of the child better than the mother in that the child is a genetic combination of both parents. The same, of course, is also true of mothers. Fathers also directly influence their children through the quality of the relationship that they have together. Children's well-being as measured by sex-role adjustment, achievement, and psychosocial adjustment is enhanced when they have a close, warm, and loving relationship with their fathers (Lamb, 1986). As noted earlier, children with such relationships with both parents benefit even more (Lamb, 1986).
Lamb identified three components of parental involvement (applicable to mothers as well as fathers): direct interaction in which parent is actively engaged in activity with child; accessible in which parent is not directly involved with child, but is in the same room or nearby and is, therefore, readily accessible; and responsibility in which parent ensures that the needs of child are met such as scheduling pediatric visits, making sure meals are ready, and that child has clothes to wear. Mothers overwhelmingly take responsibility for their children's well-being (Lamb, 1986). This is true even if the mother works full-time. The reason for this is not clear. Some suggest that mothers strongly desire to assume this responsibility role and to some extent actively discourage fathers from trying to assume the role (Lamb, 1986). On the other hand, a substantial proportion of fathers do not appear to strongly desire the role and are generally willing not to assume it (Lamb, 1986).
The nature of the interaction that fathers and mothers have with their children is quite different. Fathers spend proportionately more of their time playing with their children, while mothers spend a greater proportion of their total time with their children in caretaking activities (Lamb, 1986). Because mothers spend a greater amount of time overall with their children, they may actually spend more time playing with them than the fathers do, yet caretaking still is what best characterizes the overall time, while play characterizes the fathers overall time with his children. The type of play that fathers engage in is also different than that of the mothers, being much more likely to be rough and tumble (Hetherington and Parke, 1993).
The extent of fathers' involvement with their children changes as the children grow older and also varies by whether the child is a boy or a girl. Regardless of the child's age, fathers are more likely to be involved with their sons than with their daughters (Lamb, 1986; Marsiglio, 1991). Fathers (and mothers) spend less time with their children as the children grow older, in part because children themselves desire to spend more time with friends and peers. However, in spite of spending less time as children grow older, the importance of fathers to children's development increases as children grow older, especially for sons (Thompson, 1986).

Ways in Which Family Structure May Affect Children's Lives

Loss of a Parent

As the previous discussion on the role of the father revealed, fathers play an important role in intact households. Their absence in a household potentially has multiple effects. Their contribution to the family income is reduced, their role as caretaker and role model to their children is sharply reduced, their role as psychological, emotional, and physical supporter to the mother is reduced (though in divorcing families this role may have diminished well before the divorce), their role as parent is reduced, and the quality and nature of their relationship with their children may be changed. In this section, we discuss what research reveals about their changed role, the quality and nature of relationship that they have with their children, and what effects their absence has on children's lives. We then go on to discuss what factors are associated with a father remaining involved, what factors explain why fathers are absent from their children's lives, whether other adults can substitute for the non-residential father, and the interconnections between paternal involvement, child support, and custody arrangements. In this discussion we are assuming that the fathers have not been abusive or have a history of substance abuse or psychological difficulties that could be harmful to their children or former spouse. When such conditions are present, paternal loss may well improve the quality of life for children and for the former spouse. Such cases, however, are not typical.

Role as Parents

Studies have shown that the amount of contact non-residential fathers have with their children diminishes over time (Furstenberg et al, 1983; Furstenberg and Nord, 1985; Seltzer and Bianchi, 1988). 1981 data indicate that approximately half of all children with a father living elsewhere see that father less than once a month or had not seen him at all in the past year (Seltzer and Bianchi, 1988). There is some evidence, however, that the amount of contact, though it still is low, may have increased in recent years (Kelly, 1993; Furstenberg and Harris, 1992). Some have also noted that the rate at which contact decreases may be smaller, the older the child is at disruption (Furstenberg and Harris, 1992). Such a pattern would make sense in that fathers of older children have had a longer period of time to develop strong relationships and to have made substantial emotional and economic investments in their children and thus would be more reluctant to sever the relationship (or to have it severed by others). It should also be noted that some studies have shown that noncustodial parents report more visits with their children than do custodial parents (Seltzer and Brandreth, 1994; Braver et al., 1991). Even so, the proportion of children with little or no contact with their non-residential parent is quite large.
Most studies examining father contact have focussed on divorced fathers. A few studies, however, have also examined parental contact in families formed by adolescent mothers (Furstenberg and Harris, 1993; Mott, 1990). These studies document movement of fathers into and out of the child's household. Furstenberg and Harris found that just under half the fathers of children born to adolescent mothers had lived with the child at some point while the child was 18 or younger, but only 9 percent of these fathers had lived with the child until the child turned 18. Most of the nonresidential fathers did not maintain regular contact during the child's entire childhood (Furstenberg and Harris, 1993). Others have found that fathers who had not married a child's mother were less likely to pay child support, visit the child, or be involved in decision-making concerning the child compared to divorced fathers (Seltzer, 1991; Peterson and Nord, 1990).
Even when contact is regular, the role that non-residential fathers play in their children's lives is often quite limited (Furstenberg and Nord, 1985). This statement applies to divorced and never married fathers. Non-residential fathers may engage in social activities with the child such as occasionally taking them out to dinner, on a trip, or playing with them, but they rarely help them with school work or carry out some project together (Furstenberg and Nord, 1985). Using Lamb's terminology, fathers may interact with or be available to their children, but rarely take (or have the opportunity to take) responsibility for their children. As noted earlier, this is true of fathers in intact families as well. Fathers are also less likely to impose rules on their children (Furstenberg and Nord, 1985). Some have argued that there are no clear social rules about how non-resident fathers should behave and therefore fathers are uncertain about their responsibilities after divorce or if they never married the child's mother (Seltzer, 1991).

Quality and Nature of their Relationships with their Children

Children seem to desire a continuing relationship with their fathers. Studies consistently find that loss of regular contact is the main complaint about divorce voiced by children (Kelly, 1993). Children also seem to apply a different standard in defining their relationship to a non-resident parent, particularly their non-resident mothers. Perhaps because they expect less, they are more positive about the little that they receive (Furstenberg and Nord, 1985). There is some evidence that the perceived emotional bond that children feel for their parents is more predictive of well-being than actual contact (Amato, 1994b). For these reasons, even if studies show no positive benefit of paternal participation on children's well-being, the children's expressed wishes to see their fathers should not be taken lightly.

Consequences for Children's Lives

Although in theory the loss of the continuing presence of the father should be detrimental to children's development, research results are decidedly mixed in this regard (King, 1994; Kelly, 1993; Furstenberg and Cherlin, 1991). Several large-scale survey studies have found no association between the amount of contact a non-custodial father has with his children and an assortment of measures of child well-being (King, 1994; Furstenberg, Morgan, and Allison, 1987). Other studies have found continued contact to be related to improved psychological scores, fewer behavioral problems, and better peer relationships (Peterson and Zill, 1986; Wallerstein and Kelly, 1980). Some studies have found differences by sex and by the mother's attitude toward the father's contact with the child (Kelly, 1993).
What might explain the inconsistent and generally weak or non-existent links between paternal participation and children's well-being. Furstenberg and Harris (1992) posit three possibilities: contact is too low even among those with regular contact for there to be any measurable effect; (2) the positive benefits of higher contact may be offset by more conflict with the mother or other changes in family dynamics that adversely affect the child; and (3) fathers are not so important to children's lives as theory would lead one to expect, as long as there is a good relationship with the mother or other adult. Another possibility is that contact, itself, is a mixed blessing. Although contact allows the relationship of the father and his children to continue, it also forces the children (and the father) to endure repeated separations. Studies have shown that the artificial nature of visitation is a major source of stress for fathers. Of the literature we reviewed, none examined the stress that it places on children. Over ten years ago, this same gap in the literature was noted (Clingempeel and Reppucci, 1982). It seems conceivable that children who feel closest to their father would also be the ones who feel the most pain at having to repeatedly leave him. To truly understand the role of parental involvement in mediating children's adjustment to divorce, studies need to include measures of parental involvement, information on the quality of the father-child relationship both before and after divorce, the extent of conflict with the custodial parent, and measures of stress felt by fathers and children in the same model. The relative importance of each of these factors on children's adjustment could also be ascertained by including them in one model.

Why Some Fathers are Absent from Their Children's Lives

The reasons why fathers lose contact with their children over time and "drop out" of their lives are not well understood (Furstenberg and Harris, 1992). Factors that research suggests play a role include geographic mobility, remarriage of either parent, inability to establish a workable childrearing arrangement with the former spouse, lack of access due to actions of the former spouse, psychological pain at not being able to see their children in the same manner as before which causes fathers to remove themselves entirely to reduce the pain, and inadequate financial resources (Furstenberg and Harris, 1992; Furstenberg and Cherlin, 1991; Umberson and Williams, 1993; Seltzer and Bianchi, 1988).
In a focus group study of young black men who fathered children born to adolescent girls, the men gave several reasons for not remaining involved in the children's lives. These were that the children were not theirs, the mothers had another partner who had replaced them, their support wasn't going for the children, they didn't have the money, and the mother prevented them from remaining involved (Furstenberg, 1992). These young black fathers gave very similar responses to those given in more representative samples of fathers.

Predictors of Remaining Involved

The quality of the relationship between fathers and their ex-spouses is one of the strongest predictors of remaining involved: the better the relationship the more likely the fathers are to remain involved (Chase-Lansdale and Hetherington, 1990; King, 1994). This appears to be true of ever-married couples, as well as low income, black, never-married couples. One reason for this association is that mothers under such circumstances would have fewer reasons to limit the fathers' access. Another factor is how well the fathers, themselves, are able to adjust to divorce. Fathers who are better able to cope with the stresses and changes are more likely to remain involved (Chase-Lansdale and Hetherington, 1990).
In a study of 289 single, teen-mother families on AFDC in Wisconsin, the father's work experience was the strongest predictor of the father remaining involved in the child's life. Fathers who had worked in the last year were significantly more likely to help out in a variety of childrearing tasks including dressing and playing with the child, to talk with the mother about the child, and to have a good quality relationship with the child as described by the mother (Danziger and Radin, 1990). The authors speculate that the association may be due to the fact that both mothers and fathers may hold the belief that a sign of a good father is one that can provide for his family. Thus, fathers who are employed may have a greater sense of pride in their ability to be a father and thus perform other expected duties, as well. Second, the mother and her extended family may be more willing to let an employed father remain involved in the child's life (Danziger and Radin, 1990).
Although it appears logical to assume that fathers who were more highly involved with their children prior to separation would remain more highly involved than other fathers after separation, the evidence for this supposition is weak at best (Chase-Lansdale and Hetherington, 1990). Again, however, it is precisely these fathers who are most likely to feel the pain of the artificial nature of visitation. Policy makers may want to consider ways in which they could reduce the psychological stress that non-resident fathers feel. Steps that try to involve the fathers in truly meaningful ways in their children's lives and that award them a sense of control may ease their stress and reduce the tendency for them to remove themselves.
In a study of two-parent families, Marsiglio found that the best predictors of father involvement in minor children's lives were characteristics of the children, themselves. Spouse/partner attitudes, occupational prestige, and work hours were not important. However, fathers were more likely to do things with their children as their children got older, if they had all boys, and if the children were their biological children (Marsiglio, 1991). Although it is clear that custodial mothers can act as "gatekeepers" in permitting or preventing non-residential fathers access to their children (Price et al., 1994), the role of the children themselves in the amount of contact, particularly as they get older, needs more examination.

Can Other Adults Substitute for Biological Fathers?

From a policy perspective, it is useful to know whether other adults can substitute for biological fathers. In a study of children four years of age or younger born to young mothers, Mott (1990) documented considerable movement of fathers into and out of the children's households and also the presence of other adults who could serve as father-figures in the children's lives, particularly in black female-headed families. Thus, at least in the early years, the potential for children to form relationships with fathers or father-substitutes is greater than survey results on family living arrangements would lead one to believe. However, as noted above, other studies have found that over time, most of these young fathers, like their older ever-married counterparts, lose contact with their children (Furstenberg and Harris, 1992).
Research on stepfamilies has shown that stepfathers do contribute income to their families, but in many other respects children in stepfamilies do not fare better than children who remain in single-parent families. There are age and sex differences in children's reactions to their mother's remarriage. Yet the stepfather appears to compete for the mother's time with her children, brings yet one more adjustment that the child must make, and may not have the same level of commitment to the child that a biological father would have (McLanahan and Sandefur, 1994).
Elder and his colleagues (1995) in a study of inner-city families did not find that the presence of other adults in black single-parent families buffered the parents or children. Hawkins and Eggebeen (1991) using a more representative sample of young children also found no positive benefit of having other adults present in the household. In fact, among white children the presence of a grandfather and mother was associated with greater psychosocial problems compared to children living in intact families (Hawkins and Eggebeen, 1991).
Although grandparents are biologically linked to the child and, therefore, presumably would exhibit greater commitment to the child, studies of three-generation black families have also shown that children in these families are similar to children in step-families and to children in single-parent families (McLanahan and Sandefur, 1994). They suggest that there may be differences in parenting styles, diffusion of parental responsibility, and a weakening of the quality of parenting in both the mother and grandparent. Furstenberg and Cherlin, however, found that the psychological well-being of children raised by a mother and grandmother was the same as for children raised in two parent families (Furstenberg and Cherlin, 1986).

Custody Arrangements, Child Support, and Parental Involvement

The most common custody arrangement is for the mother to obtain sole legal and physical custody. However, the proportion of joint custody arrangements in which both mother and father retain legal control over their children is growing. In 1991, nearly 17 percent of custodial parents reported that they had a joint custody arrangement with the non-residential parent (U.S. Bureau of the Census, 1995). There are two types of joint custody arrangements: joint legal custody in which both parents have legal control over decisions that affect the child, but the child resides with one parent, generally the mother and joint physical arrangements in which both parents have legal control over the child and the child actually spends time (typically similar amounts of time) in both households (Arditti, 1992). Given that these types of arrangements are still not widespread, little research has examined the consequences for children of joint-legal, joint-physical, and sole custody arrangements (Furstenberg and Cherlin, 1991).
Part of the impetus behind the movement to encourage joint custody arrangements was the presumption that such arrangements would keep fathers more involved in their children's lives and would increase their financial contributions to the child's upbringing (Furstenberg and Cherlin, 1991). These presumptions were based on research studies that had observed a higher level of contact and more regular payment of child support among fathers who had joint custody arrangements (Peterson and Nord, 1990; Furstenberg and Cherlin, 1991). Results from more recent studies are mixed. Some researchers have found that joint custody arrangements are associated with greater payment of child support, greater paternal involvement, and with greater paternal satisfaction with the custody arrangements (Arditti and Keith, 1993; Pearson and Thoennes, 1988). Others, however, after controlling for family background, find the type of custody arrangement does not influence levels of payment, father involvement, or the relationship between children and parents (Veum, 1993; Donnelly and Fineklhor, 1992; Seltzer, 1990).
There is an association between the amount of contact a father has with his children and his likelihood of paying child support. Fathers with more contact tend to pay more child support (Arditti and Keith, 1993; Furstenberg and Cherlin, 1991; Seltzer, et al., 1989; Furstenberg et al., 1983). The reason for this association, however, is not necessarily that increased contact leads to more payment. Veum (1993) using the National Longitudinal Survey of Youth and simultaneous equations found that there was no causal relationship between visitation and child support payment. It is possible that a third variable, commitment to the child, may explain both the level of visitation and the level of payment (Veum, 1993; Furstenberg and Cherlin, 1991). Another possible variable that could affect both level of child support payment and visitation is the quality of the relationship between the residential and non-residential parents (Veum, 1993).
Some have suggested that one reason parents do not pay child support is that they cannot monitor or control how the money is spent (McLanahan and Sandefur, 1994). Other studies support this view in that fathers who feel a sense of control over their children after divorce are more likely to visit and to pay child support (Price et al., 1994). Similarly, others have found that the feeling of loss of control is a major source of stress among non-residential fathers, particularly when it comes to knowing how the child support monies that they pay are spent (Umberson and Williams, 1993). Compounding fathers' inability to monitor or control the funds they pay is the fact that at least some of the money goes towards supporting their former spouse or partner and others in the household (McLanahan and Sandefur, 1994; Umberson and Williams, 1993). Becker and others have suggested that when fathers leave a household, they lose commitment to their children over time and as they establish new relationships (McLanahan and Sandefur, 1994; Becker, 1981; Furstenberg and Nord, 1985). Fathers who had high levels of involvement prior to the divorce are more likely to make the specified child support payments (Peters et al., 1993).
The relationship between custody arrangement, paternal involvement, and child support needs more study using larger, more representative data and incorporating potentially confounding variables such as the quality of the relationship between the two parents and whether the arrangement was voluntarily agreed to or was mandated by the courts. Moreover, few measures exist for potentially important concepts, such as paternal commitment. It might be worthwhile to spend time developing such measures that could be included in future studies. It would then be possible to more directly study factors that predict paternal commitment and the effect of paternal commitment on child support payments and on paternal involvement.

Adjustment of Custodial Parent and Parenting Skills

Divorces are usually very stressful for all the people involved, though the person who initiated the divorce may also feel a sense of relief. To varying degrees, though, all members of a divorcing family are vulnerable to feelings of depression, loneliness, regret, anger, and helplessness (Maccoby and Mnookin, 1992; Menaghan and Lieberman, 1986). The adults can become so preoccupied with their own emotional and financial difficulties that they may not be able to meet the emotional and psychological needs of their children. The custodial parent may find it very difficult to gather the energy to monitor his or her children and to enforce consistent expectations and demands on them. Research has shown that children in single parent families receive less adult supervision and attention than children living with two parents (Seltzer, 1991). One of the strongest predictors of children's adjustment following divorce is the overall well-being of the custodial parent (Kelly, 1993).
It is not only the custodial parent's adjustment to divorce that is important for children's well-being, however. There is good evidence in the psychological literature that having a mother with mental health problems, whether it be depression, anxiety, or other types, is bad for children (McLoyd and Wilson, 1991). Even if fathers in these families do not assume a large role in taking care of the children, they can offset some of the negative effects on children by providing adult companionship to the mother and by other types of support. There may be a higher rate of depression and anxiety among single parents because there is no adult companionship or support readily at hand.
Single parents also face a balancing act between meeting their own needs and meeting those of their children. In married couple families, both parents need to balance their needs of love and affection and time to be together with demands that their children make upon their time. In most cases, both parents also have strong ties to their children and work out satisfactory arrangements. In single parent families, however, the parent may be dating or actively looking for a suitor. The suitor has fewer incentives to be concerned with the needs of the children than a parent would and may be less understanding of demands that children make. Thus, there may be a conflict between being in role of dating or looking for a suitor and being a parent, particularly as children grow older. Few studies have examined how these other demands on a single mother affect her parenting behavior or her relationship with her children.

Custody Arrangements, Child Support, and Adjustment of the Custodial Parent

There has been virtually no research that examines the interrelationship between custody arrangements, child support, and the adjustment of the custodial parent following divorce. One study found that parents who are able to emotionally disengage from each other following a divorce show better psychological adjustment two years later compared to those who still are preoccupied with the divorce or have residual anger towards the ex-spouse (Tschann et al., 1989). Custody and child support issues are often emotionally charged issues that may well make it difficult for either parent to emotionally disengage from the other. As noted above, research has yielded mixed findings about the effects of custody arrangements and child support on children's well-being. A fruitful area to examine is the extent to which custody arrangements and regularity of child support payment affect the custodial parents well-being and, in turn, her children.

Parental Conflict

Numerous studies have revealed that children do not fare well when there is substantial conflict or hostility in the family (Hetherington and Parke, 1993; Peterson and Zill, 1986; Rutter, 1979). Studies have also shown that children whose parents continue to have high conflict after divorce also do not fare as well as children whose parents have a better relationship (Kelly, 1993). However, it appears that the influence of conflict is indirect, through its effect on parenting behavior (Kline et al., 1991; Tschann et al, 1989). Moreover, the extent of adverse consequences due to conflict appears to be mediated by the strategies that parents use to resolve their conflicts and by whether children feel "caught in the middle of" parental conflict (Kelly, 1993; Camara and Resnick, 1988).

Custody Arrangements, Child Support, and Parental Conflict

None of the research we reviewed directly examined the link between custody arrangements, child support payments, and parental conflict. There is evidence that when parents can cooperate, joint custody arrangements are beneficial for children, but when there is a high level of dispute between the parents, such arrangements can be bad for children (Kelly, 1993). We do not know, however, the extent to which custody arrangements or irregularity in child support payment creates or exacerbates existing conflict.

Economic Hardship

Single parent families are more likely to be poor than two-parent families, especially if the lone parent is the mother. The poverty rate for single-parent, female-headed families in 1992 (46 percent) was nearly six times higher than the poverty rate for married-couple families with children (8 percent). The poverty rate for single-parent, male-headed families (22 percent) was nearly three times higher than the married-couple rate, but only half as high as the female-headed, single-parent rate (U.S. Bureau of the Census, 1993). Such families also have lower median income levels. In 1993, the median income level for a U.S. child in a two-parent family was about $43,600 whereas the median income for a child in a mother-only family was about $12,100 (Zill and Nord, 1994). There is no doubt that family disruption and single-parenthood are associated with lower income levels and that family income is one of the determinants of a child's life chances and future success (McLanahan and Sandefur, 1994; Bianchi and McArthur, 1991).
There are three primary reasons why low income can adversely affect children's lives. First, if resources are sufficiently low, children do not receive the nutrition and medical care that they need for healthy development. Many of our government programs, including AFDC and WIC, are to ensure that children are protected from these serious consequences of low income.
Second, economic pressures can adversely affect maternal emotional well-being and maternal parenting practices. Elder and his colleagues in a study of 429 inner-city Philadelphia families found that low income and unstable work/income patterns increased the risk of emotional distress which adversely affected parental beliefs about their own efficacy and their parenting strategies (Elder et al, 1995). Among blacks, the researchers found that economic pressures affected parental efficacy both directly and indirectly through depressed feelings. Among whites, economic pressures affected parental efficacy indirectly through depressed feelings (Elder et al., 1995). The authors speculate that the direct effects observed among blacks are due to the fact that blacks tended to have fewer economic resources and therefore losses in economic resources could have a very acute affect on them and their children. Among blacks, having a strong marriage protected them so that economic pressures did not affect their emotional well-being or sense of parental efficacy. However, among single-parents the presence of other adults in the household did not provide the same emotional protection (Elder et al. , 1995). There were an insufficient number of single-white families to repeat the analyses among whites.
Third, low income can adversely affect children's lives through the neighborhoods in which they reside. Low income families are often forced to live in neighborhoods that are beset by problems. Elder and his colleagues in the study noted above, found that whites described their neighborhoods more favorably than did blacks and had access to more services (Elder et al., 1995).
A distinction needs to be drawn, however, between children who have experienced chronic poverty and those who have experienced an income loss (McLoyd, 1990). Children who have always lived in poverty are more likely to live in the most problematic neighborhoods and to have lived in them for a longer period of time. Moreover children who have experienced an economic loss have some reason to expect that their new situation is not permanent. As Vonnie McLoyd notes, "Poverty among black children is marked by its persistence and geographic concentration, whereas it is primarily a transitory, geographically diffuse phenomenon among white children" (McLoyd, 1990, p. 335). In a similar vein, children of divorce are more likely than children of never married parents to have experienced economic loss, whereas children of never married parents are more likely than children of divorce to have experienced chronic income deprivation or outright poverty.

Child Support, Economic Hardship, and Children's Well-Being

Child support offsets, to some extent, the loss of the father's income. However, many fathers pay no child support or less than was agreed upon. Several studies have demonstrated that divorced women are more likely than never married women to have child support awards and, therefore, are more likely to receive at least some child support (U.S. Bureau of the Census, 1995; Peterson and Nord, 1990). Although studies may overstate the amount of underpayment because some couples may have informally renegotiated their agreements, the reduction in income is substantial (Peters et al., 1995; U.S. Bureau of the Census, 1995).
Child support can certainly reduce the economic hardship in a family. The connection between payment of child support and non-financial aspects of children's well-being is less clear. Several researchers have shown a link between receipt of child support and educational attainment and academic achievement (Knox and Bane, 1994; Graham et al., 1994; Baydar and Brooks-Gunn, 1994). Payment of child support also appears to be associated with a lower level of school behavior problems (McLanahan et al., 1994). However receipt of child support seems to be less strongly related to other aspects of children's well-being (Marsiglio, 1993). Reasons for this may be that the amounts paid are insufficient to reduce the economic pressures on the custodial parent and associated psychological stress experienced. The typical amount paid, which is usually lower than what is owed, is below the actual costs of raising a child. One study found that state child support guidelines generally fall within the range of estimates of raising children (Lewin/ICF, 1990), which suggests that the economic pressures would be reduced if the non-resident parent paid the full amount owed. There may also be accompanying conflict over the child support payments so that potential benefits for children's well-being are masked by the adverse effects of parental conflict.

Stressful Life Changes and the Loss of Social Supports and Other Resources

When divorces occur, many other changes also tend to occur. At a minimum, one parent must find a new place to live. Often, because of financial difficulties, both parents end up moving. Thus, at a time when the members of the family are at an increased need for emotional support, they are uprooted from familiar surroundings and from friends and family. Thus, the risk for social isolation is greatly increased. In time, many children must cope with their parents dating behavior and, for a substantial proportion, the remarriage of one or both parents. It is, therefore, not uncommon for children of divorce to be faced with the challenges of adjusting to their parents' boy friends or girl friends, a step-parent and step-siblings, or to the birth of a half-sibling. Children whose parents never married face similar challenges. How children cope when facing these major life changes had not been well-studied, although there is ample evidence that life in step-families is not equivalent to life with both of one's own parents. As noted earlier, children in step-families also experience a higher incidence of behavioral problerns and emotional difficulties. No study that we reviewed adequately examined the extent to which children of divorce are exposed to multiple stressful life changes and the relationship of number of stressors or type of stressors to their adjustment. Amato (1993) suggests that this is a fruitful direction for research.

Custody Arrangements, Child Support, and Stressful Life Changes

Theoretically, joint custody arrangements and adequate child support levels should reduce the number of stressful changes that children must experience. Under ideal circumstances, joint custody arrangements would keep the father actively involved in the child's life and the child support would ensure that acute changes in income did not force the child to drastically change his or her lifestyle. Research, however, has not examined the extent to which children's lives change by the type of custody arrangement or due to differences in regularity of child support payments. Joint physical custody arrangements force the child to adapt to two homes and potentially to two different sets of standards and rules. We have no sense, however, of how often this is the case or what consequences it has for children of different ages.

Weaknesses and Gaps in Literature

Much of the existing research has been based on middle-class samples of divorcing couples and on clinical samples or samples of convenience. This is a major weakness because we do not know the extent to which current findings are generalizable to the broader U.S. population and to non-middle class and minority populations. There also is only limited information on children whose parents never married. Much attention has been paid to their poor economic prospects, but less attention has been devoted to other aspects of their lives. From a data perspective, more longitudinal, large-scale studies are needed that can test some of the findings that have been based on small-scale, clinical, or other non-representative samples. These studies should include measures about family processes and changes in potentially mediating factors over time such as remarriage, entry into adolescence, changes in family relationships. They would also be able to sort out age at disruption and duration effects. Prospective studies would be ideal. A new National Survey of Children or National Longitudinal Survey of Youth that followed a representative sample of young people over a five to ten year period could provide a wealth of information about the consequences of non-marriage and divorce on children's lives. Such a survey should interview both parents (regardless of whether they live in the household) and at least two children in the household. Moreover, attempts should be made to locate and interview at later points in time any parent that leaves the household.
Aside from weaknesses in our knowledge base due to limitations of the samples that have been used to study these questions, there are also several major gaps in our knowledge. A major gap in the literature is the lack of theoretical work that links what psychologists have learned about the development of children and those factors that foster healthy development with what family sociologists and economists have learned in studies of marital disruption and divorce. To truly understand the consequences of different custody arrangements, visitation, and child support on children's lives, these perspectives need to be brought together into one model.
There is also a need for studies that can determine the relative influence of different variables that predict to parental involvement and to child well-being. To accomplish this, the models need to incorporate measures of the various forces that can affect parental involvement or children's adjustment to a single-parent family, such as those discussed earlier: the loss of a parent and the psychological pain that may accompany it; poor adjustment of the custodial parent and a deterioration in his or her parenting skills; parental conflict; economic hardship; and other stressful life changes. Until measures from each of these domains is included in a single model, it will remain difficult to understand children's adjustment to family disruption.
Another major gap in the literature is research on how children in non-middle-class families react to marital disruption or the lack of marital formation, in the first place.
Researchers have attempted to examine the association between child support or child custody and children's well-being, but they have not attempted to theorize why there should be an association or disentangle the pathways by which either should affect children's well-being. As noted above, there is virtually no research that examines the connections between custody arrangement, child support and maternal adjustment or parental conflict. More research needs to be done in this area because it is quite likely that effects on children of different custody arrangements or child support payment patterns will be mediated in large part by how the mother adapts and by the extent of parental conflict and the manner in which it is expressed.
There is a need for more research on non-custodial fathers -- the stresses they face, how they cope, their emotional adjustment, how they feel about changes in their parenting role, and factors that alleviate stresses. It is commonly believed that men, because they are generally not the primary caretaker, are less reliable respondents than mothers with respect to providing factual information about their children. However, fathers would be the best source of information about their own adjustment and reactions to divorce and about having children through a non-marital relationship. Such information could be obtained through the survey described above or through a series of selected focus group or small-scale studies with fathers who represent distinct groups of interest such as low-income blacks in the inner city (Furstenberg and others are currently pursuing such studies), nonmiddle-class whites, non-middle-class minority fathers, and middle-class black and other minority men.
There is need for more research that devotes attention to the developmental needs of children. Such research would address questions such as the following: do optimal custody arrangements vary by the age of the child? Does the influence of any given force change with the developmental stage of the child? For example, is parental absence or parental conflict more damaging when children are toddlers, when they are school-aged, or when they are adolescents? Similarly, does the influence of the various forces vary with the child's temperament or other characteristics? For example, do children who are outgoing and those who are shy react differently to parental conflict or to adjustment problems in the custodial parent?
From a policy perspective, an important question is, does the influence of custody arrangements and child support payments on children's well-being differ according to the particular configuration of forces at work in any given situation? For example, if conflict between the parents is the primary reason for difficulties for children in one family, then a joint custody arrangement may exacerbate the conflict, which could lead to poorer outcomes for the children. However, if maternal adjustment factors are the prime difficulty, then joint custody arrangements, where the father shares the child-rearing responsibilities and buffers his children, may benefit the children. Similarly, in a family experiencing financial stress, child support payments may be very important, but in a family not experiencing such difficulties, child support payments may not matter as much. One reason for the inconsistent findings in the literature may be that the effects have been masked by not controlling for the type of forces involved.

Policy Ramifications

The above review has several implications for policy makers.

All persons who are involved in working with single parent families and with divorcing families need to be made more aware of the developmental needs of children, the potential difficulties that they will face from family disruption or turmoil, and steps that could ease those difficulties.

• Educate fathers, mothers, and service providers of the importance of fathers in the lives of children.
  
  
• Educate parents as to the consequences of divorce for children;
  
  
• Ensure that judges who preside over family courts receive training in the developmental needs of children.
  
  
• Reduce, to extent possible, changes in the children's lives, at least in the immediate post-divorce/separation period -- children may be better able to cope with changes in routines later. If possible, allow children to remain in same house or neighborhood with the non-residential parent living nearby in the first year or two;

Given that conflict if inappropriately handled can be harmful to children and can alienate the two adults so that they cannot cooperate over matters concerning the child's well-being, it is important to provide services to reduce conflict or to express it in a healthier manner.

• Encourage mediation and conflict reduction during the divorce process;
  
  
• Inform parents about the potentially harmful effects of conflict for their children;
  
  
• Teach parents about the different ways of handling conflict and how some forms of handling conflict are particularly detrimental to children.
  
  

Several studies showed that all the parties in a divorce experience stress. The stress can adversely affect the custodial parent's ability to function effectively; it can serve to inhibit the non-custodial parent from remaining involved in the child's life; and it may affect how the child adjusts to family disruption and turmoil.

• Support research which studies ways to reduce stress of custodial parent, the non-custodial parent, and the children during the divorce period;
  
  
• Support research which can provide more information about how to reduce stress of the parents and children in never married families;
  
  
• Ensure that service providers and social workers are attuned to the potentially high levels of stress that single parents feel and that they seek ways to alleviate those stresses.

Given that most children desire the continuing presence of a father in their lives and that fathers may disengage from their parental responsibilities in part because they feel no sense of control over the new arrangements, steps should be taken to enable fathers to have a more active post-divorce role. There will, of course, be cases where this will be impossible because of the inability of the two parents to cooperate, because the father has no interest in remaining involved, or because of a history of past abuse. But, when it is at all feasible, policy should encourage paternal involvement.

• Allow paternal role to continue, to the extent possible;
  
  
• Find ways to allow fathers to have a meaningful role in their children's lives where they can shoulder some of the responsibility of raising the child;
  
  
• Find ways to enable non-custodial parents to have some control over child's life.

There is a scarcity of research that has examined couples who never establish awards and couples who never married each other. Such couples are particularly vulnerable economically and the children also face risks as well.

• Support research which looks at low income fathers and their children.
  
  
• Support experimental interventions aimed at increasing the establishment of awards among couples who have not been married and among low income couples.

---