An Overview of the Risks and Protectors for Children of Separation and Divorce

3. The Association Between Components of Divorce and Children's Adjustment

In section 2, we saw that while most children cope successfully with the stress of parental divorce, children from divorced families are at an increased risk of developing adjustment problems.  In this section, we review four components of divorce that have been suggested to account for the increased risk of maladjustment in children of divorce.  They are:

3.1  Absence of Non-resident Parent

Parents are important resources for children in terms of providing them with attention, assistance, love, support and supervision.  However, following parental divorce, children typically have decreased contact with their non-resident parent, usually the father.  One large-scale study in the United States found that in the previous five years, 23 percent of fathers had no contact with their children, while an additional 20 percent had not seen their children during the preceding year (Furstenberg and Nord, 1985).  Estimates for Canadian children are only a little lower (Marcil-Gratton and Le Bourdais, 1999).  A second American study found that approximately one third of fathers saw their children only once or not at all during the previous year, that approximately four out of ten fathers saw their children a few times a year to a few times a month, and that a quarter saw their children once a week or more (Seltzer, 1991).  Marcil-Gratton and Bourdais (1999) found that the type of union the parents had prior to their separation influenced the frequency of father-child contact after the divorce.  Common-law union prior to separation was associated with less non-resident parental contact after separation than married unions (common law 21 percent versus 11 percent married unions).  The age of children at the time of separation has also been found to be an important factor in father-child contact.  Contact has been found to be greater when children were older at the time of separation (Le Bourdais, Juby and Marcil-Gratton, 2001).

Does this decreased contact with the non-resident parent account for the maladjustment experienced by some children of divorce?  Empirical studies have provided inconsistent findings that generally do not support a positive answer to this question.  Some researchers have found no relationship between visitation frequency and child outcomes (McLanahan and Sandefur, 1994), while others have found that frequent contact with the non-resident parent is associated with better adjustment, but only when interparental conflict is low (Kelly, 2000; Wallerstein and Kelly, 1980).

Amato and Keith (1991a) argued that if decreased contact with the non-resident parent is associated with child maladjustment, then children of divorce should fare no worse than children who have experienced the death of a parent.  To assess this argument, researchers combined the results of 23 studies that included data on children who had experienced the death of a parent and children from divorced and intact families.  The results of the meta-analysis indicated that the children whose parent had died were significantly lower in academic achievement, and had more conduct problems and difficulties in psychological adjustment and self-esteem than children in intact families.  However, children whose parent died were significantly higher in academic achievement and displayed fewer conduct problems than children of divorced parents.  As well, by collapsing all child outcomes into a single category, children whose parent had died had a lower overall adjustment score than children whose parents were still together, but functioned at a higher level than children whose parents had divorced.

As noted above, the research literature suggests a weak association between paternal visitation frequency and child well-being.  Although researchers have used visitation frequency as an indicator of the general quality of a father-child relationship, relational theorists suggest that this association is problematic.  Thus, the frequency of paternal contact is a necessary but not sufficient condition for a close relationship to develop between a father and child.  Following from this theoretical assertion, Amato and Gilbreth (1999) recently conducted a meta-analysis to examine the association between four dimensions of father involvement, including visitation frequency, payment of child support, feelings of closeness, and authoritative parenting and child adjustment.  In accordance with previous research, the results of the meta-analysis indicated that, in general, children do not appear to benefit from frequent visitation with their fathers.  Rather, what a father does with his child when he has contact with him or her is much more important than how often he sees the child.  For example, it was found that child support, feelings of closeness, and authoritative parenting were all significantly associated with positive child outcomes, with authoritative parenting being the most consistent predictor of child outcomes.

In conclusion, the literature suggests a weak association between the absence of the non-resident parent or visitation frequency and a child's well-being, and emphasizes that additional mechanisms must be operating in divorced families that have an affect on child well-being.

3.2  Troubled Parent-Child Relationships

A large number of studies have consistently documented that, on average, parents and children from divorced families have less positive relationships than parents and children from intact families (Amato and Keith, 1991a; Hetherington, 1989, 1993).  For example, data suggest that divorce is related to significant declines in the quality of parenting, including poor communication, inconsistent warmth and affection, inconsistent discipline, and lower levels of monitoring (Peterson and Zill, 1986; Simons et al., 1999).  In addition, as previously noted, prospective studies suggest that difficulties in the parent-child relationship of divorced families were evident prior to the parental divorce (Amato and Booth, 1996).  This suggests that the stress of marital difficulties indirectly influences parent-child relationships both before and after the parental divorce.

In a national study of 1,147 American children from divorced and intact families, children were interviewed at two points in time about their relationship with their parents (Zill et al., 1993).  Children were interviewed when they were between the ages of 12 to 16 years (Time 1) and 18 to 22 years (Time 2).  Of the 1,147 children, 240 of them experienced the divorce of their parents at the mean age of 6 years.  "Poor" parent-child relationships were defined as the child's positive response to zero or only one of the following:  1) feel close to the parent; 2) satisfied with the amount of affection received from the parent; 3) desire to be the kind of person the parent is; and 4) doing things with the parent that the child enjoys.  At Time 2, the fourth item was replaced with a statement to determine how well the youth could share ideas or talk with the parent.

The results of the study indicated that 32 percent of children from divorced families reported they had a poor relationship with their father at Time 1 and 65 percent reported they had a poor relationship with their father at Time 2.  In contrast, 14 percent of children from intact families reported they had a poor relationship with their father at Time 1 versus 29 percent at Time 2 (Zill et al., 1993).  With respect to the mother-child relationship, in both divorced and intact families, only 8 percent of youth reported a poor relationship at Time 1.  However, at Time 2, there was a significant increase in the number of children from divorced families who reported a poor relationship with their mother.  Specifically, 25 percent of children from divorced families reported they had a poor relationship with their mother, versus 18 percent of children from intact families (Zill et al., 1993).  Taken together, these findings suggest that, on average, the father-child relationship in divorced families is significantly worse than that between mothers and their children.  In addition, the fact that an increasing number of youth reported poor relationships with their parents as time progressed indicates that divorce may exacerbate normative difficulties in parent-child relationships (Emery, 1999).

A number of studies have linked troubled parent-child relationships with increased internalizing and externalizing of problems among children from divorced families (Lengua et al., 2000).  In addition, an intervention program aimed at improving parenting skills has shown that increasing consistent discipline and the positive aspects of the mother-child relationship leads to improvements in children's adjustment following divorce (Lengua et al., 2000; Martinez and Forgatch, 2001).  Thus, empirical data suggest that an additional stressor for children in divorcing families is the increased risk of troubled parent-child relationships and the decreased quality of parenting.

3.3  Economic Disadvantage

According to U.S. statistics, 52 percent of children living with a single mother lived below the poverty line in 1994, as compared to 24 percent of children living with a single father, and 11 percent of children living in two-parent households (Emery, 1999).  Since divorced parents fare better than parents who were never married, 38 percent of children with divorced mothers lived below the poverty line in 1994 as compared to 15 percent of children living with divorced fathers (Emery, 1999).  In addition, longitudinal data show that the living standards of women and children fall by approximately 10 percent in the first year after divorce (Emery, 1999).  These statistics highlight that family status, often associated with parental divorce, is a powerful predictor of child poverty.

Decreased family income following divorce may negatively affect children in a number of indirect ways.  For example, children may have to move to a less expensive neighbourhood, which could result in losing contact with friends and supports in the old neighbourhood, moving to a less desirable school, and being exposed to more negative peer groups.  In addition, there may be less money for resources such as books, computers and tutors, which could adversely affect academic achievement.  Custodial parents may also have to find employment or work longer hours to support the family.  This in turn could result in a significant decline in the quality of parenting as parents are burdened with financial concerns and consumed with the added responsibilities of single parenting.  Any of the above noted stressors can affect a child's well-being, and all of these changes can result from decreased economic stability following divorce.

According to the economic disadvantage perspective, children experience maladjustment following the separation of their parents because of a decrease in their standard of living.  If this view is valid, then few differences should be observed between children from divorced and intact families once income is statistically controlled.  To assess this hypothesis, one would compare children from divorced families with children from intact families on measures of well-being, both before and after controlling for family income.  One such study offered support for this hypothesis by finding that differences in children, due to parents' marital status, were reduced by half once income was accounted for (McLanahan and Sandefur, 1994).  A second study found that when children from divorced and intact families were compared, and when income was not controlled for, children from divorced families scored significantly lower than children in intact families on 27 of 34 outcomes.  However, when income was statistically accounted for, the number of significant differences between the groups of children decreased from 27 to 13 outcomes (Guidubaldi, Perry and Nastasi, 1987).  A recent study (Clarke-Stewart et al., 2000) found that differences between children in one- and two-parent families, on assessments of cognitive and social abilities, problem behaviour and attachment security, were reduced when family income (and the mother's education, which is usually associated with income) was accounted for.  Moreover, research has consistently documented a significant association between the father's payment of child support and positive child outcomes following parental divorce (Amato and Gilbreth, 1999).

In conclusion, then, there is ample evidence that divorce is usually accompanied by a decrease in a family's economic standing (Amato and Keith, 1991a; Emery, 1999), and that this financial stress accounts for some of the maladjustment observed in children of divorced parents.  However, this same evidence indicates that children in divorced families continue to score below children in intact families on various indices of well-being even after family income is accounted for.  Thus, although economic disadvantage accounts for some of the variance, it is not the sole explanation for the impact of divorce on children.

3.4  Parental Conflict

The two perspectives that have been discussed up to this point both involve "post-divorce factors." That is, following a divorce, children may experience poorer relationships with their parents, they may have less contact with their non-resident parent, and they may experience a decrease in their standard of living.  Empirical evidence suggests that these three post-divorce stressors can have a negative effect on children's well-being.  However, longitudinal studies indicate that a proportion of adjustment problems observed in children from divorced families can be accounted for by "pre-divorce" factors such as interparental conflict (Cherlin et al., 1991).

Researchers believe that interparental conflict is one of the most important factors associated with maladjustment in children from divorced families (Long et al., 1988).  The interparental conflict perspective takes two distinct forms.  The first states that children from divorced families experience maladjustment not because of divorce per se, but because of the interparental conflict they are exposed to prior to the divorce.  The second posits that child well-being is inversely related to interparental conflict following divorce.  Does empirical support exist for one, or both, of these hypotheses?

3.4.1 Pre-divorce Conflict

In one study, Amato, Loomis and Booth (1995) estimated that 30 percent of divorces involve severe marital conflict (i.e. frequent disagreements, serious quarrels, verbal and/or physical abuse) prior to the divorce.  If the hypothesis that interparental conflict prior to divorce, rather than divorce itself, is responsible for child maladjustment, then children from divorced families should be similar in their adjustment to children from intact, high conflict families.  In other words, interparental conflict is a risk factor for child maladjustment regardless of parental marital status.  Many studies have found support for this hypothesis.  For instance, Camara and Resnick (1988) found that marital status (divorced versus intact) was significantly associated with five outcomes of child well-being until measures of conflict were entered into the regression equations.  Following the inclusion of conflict measures, marital status was no longer significantly associated with any outcome of child well-being.  Block, Block and Gjerde (1986), in one of the earliest studies suggesting that prior marital conflict might explain the negative effects of divorce on children, found that children whose parents subsequently divorced showed disturbance in their behaviour prior to the divorce.

In an attempt to validate the pre-divorce conflict hypothesis, Amato and Keith (1991a) conducted a meta-analysis of eight studies that allowed for the comparison of children from low conflict intact families, high conflict intact families, and divorced families.  The results of the meta-analysis revealed that in comparison to children in intact low conflict families, children in intact high conflict families scored significantly lower on measures of well-being.  The effect sizes between these two groups of children were substantial for conduct (mean effect size = -0.60), psychological adjustment (-0.68) and self-concept (-0.59).  Moreover, children in intact high conflict families scored significantly lower even in comparison to children from divorced families with respect to psychological adjustment (-.31) and self-esteem (-.35).  When all outcomes were combined to calculate a single effect size, children in high conflict intact families scored 0.32 of a standard deviation below children in intact low conflict families and 0.12 of a standard deviation below children from divorced families.  These findings are consistent with the results of a recent study that found similar rates of delinquency in boys from intact and divorced high conflict families (Juby and Farrington, 2001), which suggests that parental conflict is a risk factor for child maladjustment in both divorced and intact families.

If conflict is harmful for children, does child adjustment improve when highly conflictual marriages are dissolved?  In one study, after controlling for parent's age, sex, race, education and children's sex and age, children whose highly conflictual parents separated did better as adults than those that remained married (Amato et al., 1995).  A more recent study, based on the National Longitudinal Survey of Youth data set, also examined the question of whether children benefit from the dissolution of high conflict marriages (Morrison and Coiro, 1999).  One hundred and thirty-seven children, with a mean age of six years, from disrupted families were studied.  It was revealed that 10 percent of the disrupted marriages in the sample were highly conflictual.  The results indicated that children showed more behaviour problems in highly conflictual marriages, but regardless of the level of conflict prior to divorce they showed an increase in behaviour problems following parental separation.  The greatest increase in behavioural problems, however, was shown by children whose parents remained married despite frequent conflict.  They also found that children who were in high-conflict homes before the separation did not show a drop in behavioural disturbance after the separation.  One possible interpretation of this finding was that the benefits of divorce were not yet evident, since the duration between when the children were studied and the divorce was not long.

Taken together, the results of such studies indicate that children from intact high conflict families exhibit lower levels of well-being than children from divorced families.  Therefore, interparental conflict is a strong predictor, and risk factor, of child maladjustment regardless of the family type.

3.4.2 Post-divorce Conflict

As previously noted, a second hypothesis stemming from the interparental conflict perspective is that children's well-being is inversely associated with the level of post-divorce conflict that persists between parents.  Such conflicts are typically about child custody, access and financial support.  Several studies have reported data relevant to this hypothesis.

Johnston, Kline and Tschann (1989) found that less conflict and greater parental cooperation following divorce predicted better child adjustment than continued or increased conflict and poor parental cooperation.  Guidubaldi et al. (1987) found that a decrease in parental conflict following divorce predicted better adjustment in boys.  Long et al. (1988) also examined how changes in interparental conflict following divorce are related to adolescent adjustment.  The sample included adolescents from intact families, adolescents who reported high interparental conflict both before and after divorce (the continued high conflict group) and adolescents who reported high interparental conflict before but not after the divorce (the reduced conflict group).  Because most of the adolescents in the study were males, it is not clear whether the study's results can be generalized to girls.  In any case, the results indicated that adolescents in the reduced conflict group did not differ significantly from adolescents in the intact family group on academic achievement or anxiety and withdrawal.  However, adolescents in the continued high conflict group had significantly lower academic achievement and significantly higher anxiety/withdrawal scores than both adolescents in the reduced conflict group or intact family group.

In an early study, Johnston, Gonzales and Campbell (1987) examined the effects of child custody disputes on children's well-being.  The sample consisted of 56 children between the ages of four and twelve years whose well-being was assessed at the time of the custody dispute and two and a half years later.  Half of the sample's parents had been involved in repeated custody disputes that spanned a number of years and thus would be classified as highly conflictual divorces.  The results of the study indicated that children subject to custody and access disputes were likely to exhibit symptoms of depression and withdrawal, and to voice somatic complaints.  This finding has been replicated by additional related studies which found that severe marital conflict that focusses on the child is more predictive of child behaviour problems than conflict that is not centred on the child (Grych and Finchman, 1993).

In a longitudinal study, Johnston et al. (1989) investigated the emotional, behavioural and social adjustment of 100 children whose parents were involved in disputes over custody and visitation for four and a half years (on average) after separation and two and a half years after legal dispute.  Thus, such families would be considered highly conflictual.  The study's findings did not indicate that children were better adjusted in either joint or single custody.  However, regardless of the parents' socioeconomic status, income level, ethnicity, or the number of children in the family, it was found that children who had more frequent access and made more transitions between parents were more likely to be clinically disturbed.  What explains this association?  Parents who were more actively hostile at the time of the custody dispute were more conflictual two to three years later.  Also, more verbal and physical aggression was present between parents when children had more frequent access to both parents.  Thus, the continued conflict between parents, coupled with frequent child access, increased the potential for children to be exposed to parental conflict.  This increased exposure to parental conflict in turn explains the higher incidence of emotional and behavioural problems among the children.  Also, it was found that children of severe divorce disputes appeared to be more symptomatic in response to making transitions between homes, even when their parents were no longer fighting.  This suggests that severe interparental conflict, particularly if it is focussed on the child, can have long-term effects on children's well-being even after the parental conflict ceases.

3.4.3 Aspects of Marital Conflict Related to Child Well-being

Conflict is a frequent feature of close relationships such as between spouses/partners (Eisenberg, 1992; Vuchinich, 1987).  If managed well, conflicts can actually serve several positive functions, such as promoting open communication, enabling people to feel that their goals can be met in relationships, and clarifying roles and boundaries (Vuchinich, 1987).  However, poorly managed conflicts can result in several negative outcomes, such as the erosion of relational bonds, coercive interactions, and negative social and emotional outcomes for children.

Conflict is a multi-dimensional construct that can take numerous forms.  Recent research has focussed on the distinction between the presence of interparental conflict that is covert and that which is overt or anger-based.  Covert parental conflict is defined as a passive-aggressive type of conflict that includes unspoken tension, resentment, being upset (Buehler et al., 1998; Jenkins and Smith, 1991), influencing one's child to side with them against the other parent, using one's child to get information about the other parent, and asking one's child to carry messages to the other parent (Buehler et al., 1998).  Overt interparental conflict is defined as aggressive hostile conflict marked by angry and/or violent behaviours such as belligerence, screaming, insulting, threatening, contempt, ridiculing and slapping (Buehler et al., 1998).

To date, many studies have looked at the relationship between child adjustment and types of interparental conflict in both intact and divorced families.  Buehler and Trotter (1990) found that the frequency of parental conflict was unrelated to youth problem behaviour when anger-based conflict was controlled for.  Katz and Gottman (1993) found that a "mutually hostile" (overt) pattern of marital conflict predicted an externalizing behaviour pattern in children three years after initial assessment and that a "withdrawn" (covert) pattern of marital conflict predicted an internalizing behaviour pattern only.  Jenkins and Smith (1991) found that in their sample of mothers, fathers and children, anger-based parental conflict was strongly and consistently associated with children's externalizing difficulties after controlling for two other marital dimensions (covert conflict and parental arguments over child-rearing practices).  In addition, they found that children exposed to high frequencies of anger-based parental conflict were significantly more symptomatic than children exposed to lower frequencies of it.

A meta-analysis conducted by Buehler et al. (1997) found that the mean effect size between frequency of parental conflict (or presence of parental conflict) and total youth problem behaviour is small (.19).  However, the same meta-analysis revealed that the mean effect size between anger-based parental conflict and youth problem behaviour was .35, greater than that found for the presence of parental conflict.  Moreover, in a recent cross-sectional study, Buehler et al. (1998) found that regardless of marital status, the effects of anger-based parental conflict were more strongly associated with youth problem behaviour when parental conflict was frequent.  Recently, Jenkins (2000) found that anger-based marital conflicts were significant predictors of children's angry behaviours.  More specifically, mothers' and fathers' verbal and physical angry expressions were found to be associated with the presence of anger, but not sadness, in children as rated by peers, teachers and mothers.  Thus, exposure to anger-based marital conflicts increased the likelihood that children would themselves exhibit high levels of aggressive behaviours in various interpersonal relationships (e.g. child-peer, child-teacher, child-mother) (Jenkins, 2000).

Compared to other divorce factors associated with child maladjustment (absence of non-resident parent, troubled parent-child relationships and economic disadvantage), the relationship between parental conflict and child maladjustment is consistent and strong (Buehler et al., 1997; Davies and Cummings, 1994; Lengua et al., 2000).  As well, although the mere presence of parental conflict is not necessarily harmful to children's well-being, anger-based parental conflict (compared to less hostile or angry conflict) places children at greater risk for internalizing and externalizing difficulties regardless of the status of the parents' marriage.