What Have We Learned From
30 Years of Research on Families
in Divorce Conflict?
By Elizabeth M. Ellis, Ph.D.,
Atlanta, Georgia, emellis@mindspring.com
Originally published in The Trowbridge Foundation
Report, Vol II, Issue III, Summer 2001
The following is a summary of highlights of the text, Divorce Wars,
Interventions with Families in Conflict, (APA Books, 2000). The text
is a review of research in a number of key areas and is intended as
a reference work and manual for psychological experts in the courtroom
and the attorney who wishes to be knowledgeable regarding the scientific
literature in this field. The book also contains case studies, guidelines
for clinicians, charts and data tables, as well as a complete bibliography.
(The last two chapters, on ethical dilemmas and future trends in the
area of divorce and family law, have not been reviewed here).
This summary is organized in a list of pertinent questions and brief
responses to these questions. For the full text please order a copy
from APA Books, 1-800-374-2721, or view the entire series of books on
Law and Psychology online at 222.apa.org/books/
What are the Effects of Divorce on Children?
(1) In terms of the effects of divorce on children, as a group, we
can conclude that divorce roughly doubles the rates of emotional and
behavioral adjustment problems in children. Though the effect sizes
are small, they are very consistent (see: Gregory, 1965; Hetherington,
1989; Hetherington, Cox, & Cox, 1985; Hetherington, Stanley-Hagan,
& Anderson, 1989; Wallerstein, 1985; Wallerstein, 1991; Wallerstein
& Blakeslee, 1989; Wallerstein & Kelly, 1980b; Zill, Morrison,
& Coiro, 1993; Zimiles & Lee, 1991).
(2) The effects of divorce on boys appear to be more immediate and
dramatic, especially in mother-headed households. These are increases
in aggressive, disruptive, acting out behaviors. Boys in single-mother
households are considered to be "developmentally vulnerable"
and at risk for high levels of acting out behavior (Gregory, 1965; Kalter,
1977, 1987; Hetherington, 1989; Hetherington, Cox, & Cox, 1985;
Warshak, 1992; Zaslow, 1988, 1989; Zimiles & Lee, 1991).
(3) The effects of divorce on girls (provided mother doesn’t
re-marry), are minimal, until adolescence. Girls of divorce show no
increased risk of behavior problems, as compared with girls from intact
families, until adolescence. Then, they show increased rates of running
away, skipping school, sexual promiscuity, and acting out. These girls
are more likely to drop out of school and become pregnant outside of
marriage. This has been dubbed "the sleeper effect". ( Allison
& Furstenberg, 1989; Hetherington, 1989; Hetherington & Deur,
1971; Hetherington, 1991; Kalter, 1977; Kalter, et al., 1984; Hetherington,
1972, 1973).
(4) The adult children of divorce experience lower levels of reported
happiness, higher levels of psychological problems, lower levels of
marital happiness, and a higher rate of divorce in their own marriages
. This is known as the intergenerational transmission of divorce (Amato,
1996; Kulka & Weingarten, 1979; Glenn & Kramer, 1985; Glenn
& Shelton, 1983).
(5) Women who come from divorced homes are more likely to have their
own marriages end in divorce than men from divorced homes. When both
partners come from divorced homes, the chances that their own marriage
will end in divorce is triple that of a couple who both grow up in intact
homes (Amato, 1996).
(6) What seems to be transmitted to the children of divorce is a tendency
toward lower rates of education, early marriage, living together before
marriage, and a group of behaviors which can be described as: lower
commitment to marriage, infidelity, problems with anger management,
feelings of insecurity, neediness, demandingness, denial and blame,
contempt, and poor conflict resolution skills (Amato, 1996; Hetherington,
Bridges, Insabella, 1998).
At this time, we are not sure if what is being transmitted is genetic-biochemical,
or behavior patterns that are learned, or some interaction of both.
See O'Connor, et al. (2000), for a complete discussion of this topic.
What is the Mechanism Through Which Divorce Exerts a Negative Effect
on Children?
(7) The effects of divorce on children are not explained through parental
loss or economic hardship. Children who lose a parent through a lengthy
illness or even death fare better than the children of divorce, and
economic status post-divorce is not a good predictor of children’s
adjustment. Mother’s remarriage, which is associated with a higher
standard of living, is not associated with improved outcomes for children
(Amato & Keith, 1991; Rutter, 1971; Rutter, et al., 1976; Shaw &
Emery, 1987).
(8) The effects of divorce on children occur largely through the effects
of inter-parental conflict on children, both before and after the divorce
(Amato & Keith, 1991; Amato & Rezac, 1994; Amato, Loomis, &
Booth, 1995; Buchanan, Maccoby, & Dornbusch, 1991; Camara &
Resnick, 1988, 1989; Emery, 1982; Kline, Johnston, & Tschann, 1991;
Long, Slater, Forehand, & Fauber, 1988; Rutter, 1971).
This is one of the most consistent findings and the above reference
list is abbreviated. Amato and Keith (1991) reviewed 15 studies. Emery
(1982) reviewed over 20 studies, most of which are not listed above.
The "high conflict" factor predicts a poor parent-child relationship
(#11), predicts the outcome of joint custody (#28), and predicts whether
time with the non-custodial parent is beneficial or distressing (#38).
(9) Children and teens from divorced homes where there is no conflict,
fare better than children and teens from intact homes where there is
chronic conflict. On the surface, one would expect children to benefit
from divorce based on this finding. However, in reality, the children
of divorce seem to endure years of conflict prior to their parents’
divorce; the conflict often continues unabated after the divorce, or
even intensifies; and many parents go on to marry again and have highly
conflicted remarriages. (Rutter, 1971; Long, Slater, Forehand, &
Fauber, 1988; Camara & Resnick, 1988, 1989; Kline, Johnston, &
Tschann, 1991; McCord & McCord, 1959).
(10) The type of conflict matters. Children do not appear to be affected
much by conflict which is characterized by emotional distance and tension.
They are more negatively affected by conflict that is open, attacking
(both verbally and physically), and where the children are exposed to
the conflict and caught up in it (Camara & Resnick, 1988, 1989;
Emery, 1982; Kline, et al., 1991; Buchanan, Maccoby, & Dornbusch,
1991).
(11) The mechanism through which inter-parental conflict exerts its
negative influence on the children of divorce is through a deterioration
of the parent-child relationship. Second and third divorces deteriorate
the parent-child relationship even further (Amato & Booth, 1996;
Emery, 1982; Hess & Camara, 1979; Peterson & Zill, 1986; Kline,
et al., 1991; Tschann, Johnston, Kline, & Wallerstein, 1989).
When children live in intact families where conflict is low, 62% report
having a good relationship with both parents. Where conflict is high
and persistent, only 20% do so Where children live with a mother who
has divorced a second time, only 8% report having a good relationship
with both parents. Fully 46% report a poor relationship with both parents
(Peterson & Zill, 1986).
(12) The emotional stability, warmth, and consistency of the primary
custodial parent is a strong factor in predicting children’s adjustment
to divorce. Children who reside with a parent who is emotionally unstable,
depressed, and angry experience poorer outcomes (Emery, 1982; Hess &
Camara, 1979; McCombs & Forehand, 1989; Thomas & Forehand, 1993;
Tschann, Johnston, Kline, & Wallerstein, 1989; Wallerstein &
Blakeslee, 1989).
(13) The effects of chronic conflict on children also exert their
effect through inducing in the child a feeling of chronic stress, insecurity,
and agitation (Davies & Cummings, 1994); shame, self-blame, and
guilt (Grych & Fincham, 1993); a chronic sense of helplessness (Davies
& Cummings, 1994); fears for their own physical safety (Davies &
Cummings, 1994); a sense of rejection, neglect, unresponsiveness, and
lack of interest in the child’s well being (Davies & Cummings,
1994; Fauber, et al., 1990; Fantuzzo, et al., 1991; Johnston, 1992)
What are the Statistics on Parental Divorce and Remarriage?
(14) Women are by far more unhappy with the current state of marriage.
They initiate 75% of divorces (Wallerstein and Kelly, 1980b), 76% (Koel
et al., 1988), or 80% (Furstenberg & Cherlin, 1991), depending on
the study.
(15) Second marriages have a higher rate of failure. Furstenberg (1987)
found the rate of divorce of second marriages to be 56%. Glick (1984)
found the divorce rate of second marriages to be 57%. Baydar (1988)
found that 42% of second marriages ended within 5 years.
Do Children Benefit from their Parent's Remarriage?
(16) In general, no. If one reviews the large volume of studies on
this question, one finds that in about half the studies, children do
a little better. In about half the studies, children do a little worse.
Although the children might benefit a little from an increased standard
of living, these gains are offset by increases in inter-parental conflict,
and conflictual relationships with stepparents and stepsiblings. To
look at somewhat predictable findings, you need to look at factors such
as the child’s age, and gender, and who it is that is remarrying.
(17) Broadly speaking, remarriage appears to benefit young boys who
live in single mother households and whose fathers are absent or uninvolved.
These boys are eager for a male role model. Acting out behavior is reduced
(Hetherington, Cox, & Cox, 1985; Hetherington, 1989; Kalter, 1977;
Zaslow, 1988, 1989; Zill, Morrison, & Coiro, 1991; Zimiles &
Lee, 1991).
(18) Broadly speaking, remarriage fares rather badly when one is looking
at adolescent girls in single mother custody homes. They universally
reject stepfathers, despite the best of efforts on the part of mother
and stepdad (Hetherington, Cox, & Cox, 1985; Hetherington, 1989;
Kalter, 1977; Zaslow, 1988, 1989; Zill, Morrison, & Coiro, 1991;
Zimiles & Lee, 1991).
How Successful has Mediation Been?
(19) Of those who attend and complete mediation, rates of reaching
out of court agreements are fairly high. Emery's (1987) review found
rates of 55% in a California study and 64% in a Connecticut study. Emery
& Wyer (1987) report rates of 75%; Saposnek (1984) also reported
rates of 75%. Pearson & Thoennes (1984) reported that 50% settled
through mediation, and another 30% settled just before going to court.
(20) Regarding relitigation rates, Pearson and Thoennes (1982), Emery
& Wyer (1987) and Scott & Emery (1987) all report lower rates
of relitigation in the year following divorce in the group who mediated
and settled out of court.
(21) Mediation appears to have no effect on children's post-divorce adjustment,
nor on the parents' adjustment, nor on the ability of the parents to work
together over the long-term (Pearson, Thoennes, & Hodges, 1984; Emery,
1988; Dillon & Emery, 1996; Kelly, 1996; Saposnek, 1998).
Does Gender Matter When Considering Primary Residential Custody?
In general, yes.
(22) Boys as a group are happier and show lower rates of delinquency
and school drop-out in father-custody homes. (Camara & Resnick,
1988; Clarke-Stewart & Hayward, 1996; Gregory, 1965; Peterson &
Zill, 1986; Santrock & Warshak, 1979; Warshak, 1996; Warshak &
Santrock, 1983; Zimiles & Lee, 1991).
(23) Girls, as a group, are happier and show lower rates of delinquency
and school drop-out in mother-custody homes (Camara & Resnick, 1988;
Clarke-Stewart & Hayward, 1996; Gregory, 1965; Peterson & Zill,
1986; Santrock & Warshak, 1979; Warshak, 1996; Warshak & Santrock,
1983; Zimiles & Lee, 1991).
These results are consistent and robust. There are no studies which
find the reverse--that children function better with the opposite-sex
parent. However, some have suggested that this may be an artifact of
demographic differences between mothers and fathers who are primary
custodial parents, e.g., fathers who pursue and are awarded custody
are generally more educated, more affluent, have more professional occupations,
and have been more involved with their children.
Is Split Custody (Splitting the Children) a Bad Idea?
In general, no.
(24) Split custody is rarely recommended by courts and by custody
evaluators (Maccoby & Mnookin, 1992; Simons, Grossman, & Wiener,
1990; Greif, 1990; Hauser, 1995). However, there is no established research
that has determined that it is associated with poor adjustment to divorce.
Many writers in the field have noted anecdotally that many families
drift toward split custody over time following the divorce as the children
mature and their needs change, or as the family composition of the two
households change. Therefore, it is a custodial arrangement that is
frequently naturally selected by families, and those families report
good outcomes (Greif, 1990, Hawthorne, 2000).
Is Joint Custody Better for Children?
(25) Parents who share child rearing in joint custody arrangements
(the type of "joint" custody is frequently unspecified) are
generally happy with it and satisfied with it. They report that children
see their father more often. Mothers report feeling less overwhelmed
(Pearson & Thoennes, 1990; Rothberg, 1983; Luepnitz, 1986).
(26) Reports on whether parents litigate less often with joint custody
or have less conflict are less clear. Some studies report lower rates
of re-litigation (Luepnitz, 1986; Ilfield, Ilfield, & Alexander,
1982); some report higher rates of re-litigation (Albiston, Maccoby,
& Mnookin 1990; some studies found no differences (Koel, et al.,
1988; Pearson & Thoennes, 1990).
(27) Children's emotional adjustment is not associated with custodial
arrangement (Steinman, 1981; Luepnitz, 1986; Kline, et al., 1989; Pearson
& Theonnes, 1990; Johnston, 1995).
This is one of the most robust findings in the literature. Johnston
(1995) reviewed 14 studies and concluded that there were few, if any,
differences in the adjustment of children by virtue of the type of custodial
arrangement in which they lived.
Who Settles and Who Goes to Court Over Custody?
(28) Studies show that the vast majority of divorcing families do
manage, through attorney bargaining, and through mediation, to settle
out of court. Estimates vary from "less than ten per cent"
(Kornhauser, 1979); 10% (Maccoby & Mnookin (1992), to 10% to 17%
(Rohman, Sales, & Lou, 1987).
(29) Parents' race, age, education, and income are not predictive
of who goes to court. Parents typically go to court over one or two
very young children (Maccoby & Mnookin, 1992; Dixon, 1991).
(30) Since most mothers ask for and receive primary custody, the decision
to litigate over custody is usually made by the father, when the father
expresses disagreement with this arrangement (Dixon, 1991; Maccoby &
Mnookin, 1991).
How Good are the Psychological Tests Which are Used in Custody Evaluations?
(31) Though psychological tests are widely used in custody evaluations,
their use in this context is, for the most part, highly spurious and
scientifically unsupportable (Brodzinsky, 1993; Grisso, 1986; Hawthorne,
2000; Heinze & Grisso, 1996; Melton, Petrila, Poythress, & Slobogin,
1997; Weithorn & Grisso, 1987).
(32) There are only a few tests which have demonstrated validity and
usefulness in custody evaluations. These are: the Child Abuse Potential
Inventory (Milner, 1990); Parent Child Relationship Inventory (Gerard,
1994); the Parenting Stress Index (Abidin, 1990); and the Children's
Reports of Parental Behavior (Schaefer, 1965).
What Factors are Considered most Important in Making Recommendations
for Custody?
(33) In making recommendations for custody, both judges and mental
health professionals place high value on the parent's stability and
the presence of alcoholism. However, mental health professionals tend
to place higher value on the children's bond with the parent and the
children's wishes; judges tend to place greater emphasis on issues such
as whether the parent is cohabiting, whether the parent is gay, whether
the parent has had an affair, etc. In other words, judges may come at
these issues with a more conservative bent and a focus on what society
may consider to be "moral character" issues (Lowery, 1981;
Felner, et al., 1985; Sorenson & Goldman, 1989; Reidy, Silver, &
Carlson, 1989; Keilin & Bloom, 1986; Ackerman & Ackerman, 1997).
How does Participation in a Custody Evaluation Impact Children?
(34) Studies of the impact on children who have been the subject of
custody disputes are surprising. They consistently indicate that participating
in the evaluation does not appear to be associated with poor outcomes
but rather with higher self esteem and a greater sense of control (Ash
& Guyer, 1986; Fulton, 1979; Wolman & Taylor, 1991).
What Can We Say with Some Scientific Certainty as to What Visitation
Arrangements are Best for Children of Different Ages?
(35) Not much at all. There are no controlled studies which indicate
that any visitation pattern is best for a child of a certain age. All
we have to go on is clinical experience, and extrapolations and educated
guesses which have been derived from the child development literature
(studies of babies in daycare; studies of infant attachment; studies
of separation anxiety, etc.)
How Common is the Denying or Blocking of Visitations?
(36) Quite common. Non-custodial fathers report higher figures than
custodial mothers, naturally. Fulton (1979) reported that 53% of fathers
had had their visitations blocked; Pearson & Thoennes (1988) said
20% of fathers reported problems with blocked visitations; Braver et
al. (1991) said 33% of non-custodial fathers reported loss of visitations;
Arditti (1992) reported a figure of 50%.
Custodial mothers have admitted denying visitations in these ratios:
40% (Fulton, 1979); 20% (Wallerstein & Kelly, 1980); 40% Kressel
(1985); 23% (Braver et al.).
Don't Most Children Benefit from as Much Time as Possible with the
Non-Custodial Parent?
(37) Children do not necessarily benefit from more time with the non-custodial
parent. If there are high levels of inter-parental conflict, children
may exhibit more adjustment problems with increased contact with the
non-custodial parent. (Amato & Rezac, 1994; Ash & Guyer, 1986;
Baydar, 1988; Bray & Berger, 1990; Furstenberg, Morgan, & Allison,
1987; Healey, Malley, & Stewart, 1990; Hess & Camara, 1979;
Hetherington & Parke, 1979; Hodges, Wechsler, & Ballantine,
1979; Johnston, Kline, & Tschann, 1989; Kalter, Kloner, Schreier,
& Okla, 1989; King, 1994; Kline, Johnston, & Tschann, 1991;
Kurdek & Berg, 983; A. M. Thomas & Forehand, 1993; Wallerstein
& Corbin, 1989; Zill, 1988).
This is one of the most robust findings in the research literature.
I have compiled 17 studies at this time in which no association is found
between frequency and length of visitations with non-custodial dad and
children's emotional/behavioral adjustment. It is troubling, but remarkably
consistent.
(38) Children seem to benefit from increased time with the non-custodial
parent when certain conditions are met: low levels of inter-parental
conflict and a warm, consistent relationship with the non-custodial
parent. They benefit from authoritative parenting with the non-resident
parent (i.e., advice and help with projects, supervision of homework;
discipline)... In other words, how often fathers see their children
is less important than what they actually do with them. (Amato &
Gilbreath, 1999; Clarke-Stewart & Hayward, 1996; Hetherington, Cox,
& Cox, 1982).
What can we Say About How Children Cope with Visitation Disputes and
Inter-Parental Conflict?
(39) There are some broad, but consistent patterns in children's coping
styles. Young children (3 to 8), respond with intense physical and emotional
distress. Early elementary children responded with somatic symptoms,
avoidance, and active intervention. Late elementary and early middle
school age children tend to decrease feelings of distress as they enter
into alignments with one parent (Cummings, et al., 1984; Johnston &
Campbell, 1988; Johnston, Campbell, & Mayes, 1985; Johnston, Gonzalez,
& Campbell, 1987; Johnston, et al., 1989)
What are Factors Within the Child Which Predict Good and Bad Outcomes
from Divorce and Post-Divorce Conflict?
(40) The child’s temperament is a strong factor in predicting
poor outcomes to divorce and various visitation arrangements. Children
with disruptive, acting-out temperaments experience poorer outcomes
than those with easy temperaments. (Block, Block, & Gjerde, 1986;
Hetherington, 1991; Kasen, et al., 1996; Grych & Fincham, 1997).
(41) Boys with difficult temperaments fare very poorly with divorce,
and these boys can be identified as young as age 3 (Amato & Booth,
1996; Block, Block, & Gjerde, 1986; Cherlin, et al., 1991; Jouriles,
et al., 1991). This seems to be due to a complex interaction between
the genetic transmission of certain temperament factors, poor parenting,
high levels of marital conflict, and even the destructive effect of
the boys’ behavior on what was already a poor marriage.
(42) Cognitive style is also a factor. Children who can "reframe"
the divorce in a positive way do better than those who don't (Radovanovic,
1993). Similarly, those children who appraise the divorce with a sense
of lack of control, a sense of loss, feelings of helplessness, and self-critical
feelings, fare much more poorly following divorce and/or post-divorce
conflict (Lengua & Sandler, 1996; Sandler, Tein & West, 1994).
Is Parental Alienation Syndrome (P.A.S.) a Valid Phenomena? Is it
the Same Thing as Brainwashing? Is it Simply an Attack on Mothers?
(43) Parental alienation syndrome is a predictable response in children
exposed to high levels of inter-parental conflict. It serves to reduce
high levels of stress, fear, and cognitive dissonance through a firm
alliance with one parent and the exclusion of contact with the other
parent (Gardner, 1982; Garrity & Baris, 1994; Johnston, 1993; Johnston
& Campbell, 1988; Johnston & Roseby, 1997; Lampel, 1996; Racusin,
Copans, & Mills, 1994; Rand, 1997).
(44) P.A.S also has many features in common with folie a deux--a more
dependent and submissive individual takes on the world-view of a more
dominant individual due to dependency, enmeshment, concern for the emotional
distress of the dominant person, and poor capacity to reality test.
The disorder is most often improved by separation of the parties. The
most common parent-child duo in cases of folie a deux are mother-daughter
pairs (Deutsch, 1938; Gralnick, 1942; Dewhurst & Todd, 1956; Soni
& Rockley, 1974; Sacks, 1988; Munro, 1986; Brooks, 1987; Dippel,
Kempel, & Berger, 1991, Mentjoux, van Houten, & Koolman, 1993).
(45) P.A.S. is most often a problem of mother-daughter pairs (Dunne
& Hedrick, 1994); mother-son alignments (Wallerstein & Kelly,
1980b); or mother-daughter and mother-son pairs equally (Racusin, Copans,
& Mills, 1994).
(46) Clinicians and researchers differ as to suitable treatments for
P.A.S. Some lean toward swift intervention through the courts in extreme
cases and separation of the enmeshed mother-child pair (Gardner, 1992;
Cartwright, 1993; Lund, 1995; Kopetski, 1991). Others strongly oppose
separation of the mother-child pair (Garrity & Baris, 1994; Johnston
& Roseby, 1997). All agree with Dunne & Hedrick (1994) that
traditional methods of treatment are ineffective in severe, intractable
cases.
Why do a Minority of Parents Engage in Bitter, Protracted Legal Disputes
with Each Other Over the Children? What Makes them Different from Parents
who have Amicable Settlements?
(47) Parents engaged in high levels of inter-parental conflict, before,
during, and after the divorce, appear to exhibit a cluster of personality
traits which are similar to the personality disorders (Johnston &
Campbell, 1988; Garrity & Baris, 1994; Wallerstein & Kelly,
1980b).
(48) Research studies have established that parents involved in protracted
post-divorce disputes have some personality/cognitive traits in common.
They have poor impulse control, are needy and demanding, have poor problem
solving skills, and deal with highly emotional situations through the
use of denial, blame, and defensiveness. They have difficulty taking
another person's perspective and have poor interpersonal skills. They
are more self-serving and less child-centered. (Ehrenberg, Hunter, &
Elterman, 1996; Hoppe & Kenney, 1994)
How Common are False Allegations of Sexual Abuse in Custody Disputes?
(49) In intractable cases, where experts are called in to conduct
evaluations, rates of sexual abuse allegations being declared false
or unfounded are 55% (Benedek & Schetky, 1985); 36% (Green, 1986);
50% (Thoennes & Tjaden, 1990, 160 cases); 75% (Wakefield & Underwager,
1991, 500 cases); 77% (Dwyer, 1986).
What is the Most Common Profile of a Mother Who Makes a False Accusation
of Sexual Abuse in a Custody Dispute?
(50) Various writers agree that this individual most commonly meets
criteria for a diagnosis of personality disorder--hysterical, borderline,
passive aggressive, or paranoid personality. For such an individual,
believing in the accusation fits into a cognitive schema in which the
accuser is a victim of a malicious husband, the child is an extension
of the accuser, and thus the child is a victim also (Benedek & Schetky,
1985; Blush & Ross, 1987; Bresee, Stearns, Bess, & Packet, 1986;
Elterman & Ehrenberg, 1991; Green, 1986; Klajner-Diamond, et al.,
1987; Mikkelsen, Gutheil, & Emens, 1992; Wakefield & Underwager,
1991).
________________________________________________________________________
Dr. Elizabeth M. Ellis is a clinical psychologist in private practice
in a suburb of Atlanta. She is the author of Raising a Responsible Child
(Birchlane, 1995), and over 20 scholarly papers in the fields of post
traumatic stress disorder, psychotherapy, and child and family forensic
evaluations. Her second book, Divorce Wars, was published by APA Books
in May of 2000. You may reach her at: 3346 Gwinnett Plantation Way,
B-1, Duluth, Georgia. 770-476-1967.)
