Child Custody and Visitation Decisions in Domestic Violence – Saunders

By Daniel Saunders

Daniel G. Saunders, Ph.D.
University of Michigan
School of Social Work

“…This document describes some of the legal and cultural trends surrounding custody and visitation decisions and the social science evidence supporting a need to consider domestic violence in these decisions….”


“Social science evidence can help to establish which parent is most likely to harm children. The most convincing evidence for the potential of men who batter their partners also to batter their children comes from a nationally representative survey (Straus, 1983). Half the men who battered their wives also abused their children. Abuse was defined as violence more severe than a slap or a spanking. Battered women were half as likely as men to abuse their children. Several non-representative surveys show similar results (reviewed in Saunders, 1994). When battered women are not in a violent relationship, there is some evidence that they are much less likely to direct anger toward their children (Walker, 1984).
Emotional abuse of children by men who batter is even more likely because nearly all of these men’s children are exposed to domestic violence (Pagelow, 1990). This exposure often constitutes a severe form of child abuse since the problems associated with witnessing abuse are now clearly documented (e.g., Edleson, 1997). There are short and long-term emotional and behavioral consequences for both boys and girls. Parents may not realize that their children can be affected even if they do not see the violence. For example, the children may be hiding in their bedrooms listening to repeated threats, blows, and breaking objects. Obviously, they may be afraid their mother will be injured or killed, but they may also have divided loyalties between their parents, guilt about not being able to intervene, and anger at their mothers for not leaving (Saunders, 1994). If mothers cannot find safety, their fears and depression may keep them from being as nurturing and supportive to their children as they normally would be….”

“…Mothers may also be blamed for harming their children in cases where evaluators and practitioners do not understand the dynamics of abuse (Edleson, 1997). Their cases are sometimes labelled as “”failure to protect”” since they are supposedly able to protect their children from the physical and emotional abuse of their partners (Enos, 1996). Battered women may even face criminal charges (Sierra, 1997). However, battered women’s actions often come from their desire to care for their children. They may not attempt to leave because of financial needs, because they believe that the children need a father, or because they fear losing the children to their abuser. They often leave the relationship when they see the impact of violence on their children, only to return when threatened with even greater violence or out of economic necessity. Innovative programs, like Project Protect in Massachusetts, were developed to address these concerns. They use specially trained staff and multidisciplinary teams to integrate interventions for child abuse and domestic violence (Davidson, 1995). On a policy level, states generally allow evidence to show that the non-abusive spouse feared retaliation from her partner and thus could not try to stop or prevent abuse to the child. However, only a few states explicitly authorize this type of evidence….”

“….Factors Related to Risk to the Children
In a given custody case, a number of factors related to or incorrectly attributed to child abuse and exposure to domestic violence may be present. Several factors–parental separation, childhood victimization of the parents, the parents’ psychological characteristics, and abuser interventions– are discussed next.
Parental Separation. Parental separation or divorce does not prevent abuse to children or their mothers. On the contrary, physical abuse, harassment, and stalking of women continue at fairly high rates after separation and divorce. In one study, a fourth of the women reported threats against their lives during visitation (Leighton, 1989). Separation is a time of increased risk of homicide for battered women (Wilson & Daly, 1994) and these homicides sometimes occur during custody hearings or visitation exchanges of children. In rare cases, men kill children in retaliation for their female partners leaving them.
Children are also likely to be exposed to renewed violence if their fathers become involved with other women. Over half of men who batter go on to abuse a second woman (Wofford, Elliot, & Menard, 1994). Judges who consider the remarriage of a man to be a sign of stability and maturity should instead consider it as a possible sign that the children will once again be emotionally harmed.
Parents’ Childhood Victimization. Evaluators may look to childhood risk factors of each parent to assess their child abuse potential. The link between being abused in childhood and becoming a child abuser is not as strong as was once thought, with about 30% of child abuse victims becoming abusers (Kaufman & Zigler, 1987). Some evidence suggests that the link is stronger in men than in women (Miller & Challas, 1981).
Parents’ Psychological Characteristics. The parents’ personality traits and psychological disorders are generally poor predictors of child abuse (Wolfe, 1985). Neither parent is likely to have chronic mental disorders of genetic origin (e.g., schizophrenia, or bipolar disorder). Personality disorders are much more likely to appear on the psychological tests of both parents. Great care must be taken, however, when interpreting parents’ behaviors and psychological tests. Men who batter often have the types of personality disorders that keep childhood traumas, anxiety, and other problems hidden (Holtzworth-Munroe & Stuart, 1994).
To the extent that psychological disorders continue to be used to describe battered women, they can be placed at a serious disadvantage. Compared with the chronic problems of their partners, battered women’s psychological problems are much more likely to decrease as she becomes safer. Many battered women may seem very unstable, nervous, and angry (Crites & Coker, 1988). Other battered women may speak with a flat affect and appear indifferent to the violence they describe (Meier, 1993). These women probably suffer from the numbing symptoms of traumatic stress. The psychological test scores of some battered women may indicate severe personality disorders and mental illness. However, their behaviors and test scores must be interpreted in the context of the traumas they have faced or continue to face (Rosewater, 1987). The tactics used by their abusers parallel those used against prisoners of war and include threats of violence, forced isolation, degradation, and attempts to distort reality and increase psychological dependence. Severe depression and traumatic stress symptoms are the likely results. When women fear losing custody of children to an abusive partner, the stress can be overwhelming.
Interventions for the Abuser. Successful completion of treatment does not at all mean that the risks of child and woman abuse are eliminated. Although the evaluation of programs for men who batter is still in its infancy (Saunders, 1996), it is clear that a substantial proportion of women (35%, averaged across a number of studies) report that physical abuse by their partners occurs within 6-12 months after treatment. Psychological abuse is even more prevalent. Only two studies of programs for men who batter investigated the reduction of actual or potential violence toward the children (Myers, 1984; Stacey & Shupe, 1984). Both of these studies showed promising results, yet did not specifically focus on parenting issues. Only one description could be found of a special parent training program for men who batter (Mathews, 1995)…”

“…As stated earlier, the model state statute of the National Council of Juvenile and Family Court Judges clearly states that there should be a presumption that it is detrimental to the child to be placed in sole or joint custody with a perpetrator of family violence (NCJFCJ, 1994). The model statute emphasizes that the safety and well-being of the child and the parent who is the victim must be primary. The perpetrator’s history of causing fear as well as physical harm should be considered. A parent’s absence or relocation in an attempt to escape violence by the other parent should not be used as a factor to determine custody. Courts sometimes label battered women as “”impulsive”” or “”uncooperative”” if they leave suddenly to find safety in another city or state. The model statute specifies that it is in the best interest of the child to reside with the non-violent parent and that this parent should be able to choose the location of the residence, even if it is in another state. The noncustodial parent may also be denied access to the child’s medical and educational records if such information could be used to locate the custodial parent.
Visitation guidelines should be based on the following general principles: a) contact between child and parent should be structured in a way that limits the child’s exposure to parental conflict; b) transitions should be infrequent in cases of ongoing conflict and the reasonable fear of violence; and c) substantial amounts of time with both parents may not be advisable (Johnston, 1992). Ideally, a court order should detail the conditions of supervised visitation, including the role of the supervisor (NCJFCJ, 1995). Unsupervised visitation should be allowed only after the abuser completes a specialized program for men who batter and does not threaten or become violent for a substantial period of time. Practitioners need to be aware of the strong likelihood that men who batter will become violent in a new relationship and that they often use nonviolent tactics that can harm the children. Rather than rely on official records of recidivism, the best way to establish that the perpetrator is nonviolent is to interview current and past partners.
Visitation should be suspended if there are repeated violations of the terms of visitation, the child is severely distressed in response to visitation, or there are clear indications that the violent parent has threatened to harm or flee with the child. Even with unsupervised visitation, it is best to have telephone contact between parents only at scheduled times, to maintain restraining orders to keep the offender away from the victim, and to transfer the child in a neutral, safe place with the help of a third party (Johnston, 1992). Hart (1990) describes a number of safety planning strategies that can be taught to children in these situations.
The model statute (NCJFCJ, 1994) states that visitation should only be awarded to the perpetrator if adequate safety provisions for the child and adult victim can be made. Orders of visitation can specify, among other things: the exchange of the child in a protected setting, supervised visitation by a person or agency, completion by the perpetrator of “”a program of intervention for perpetrators””, and no overnight visitation. If the court allows a family or household member to supervise the visitation, the court can set the conditions to be followed during visitation. For example, an order might specify that the batterer not use alcohol prior to or during a visit and that the child be allowed to call the mother at any time….”