CHILDREN OF BATTERED WOMEN
BY TANIA POCOCK* AND FIONA CRAM
Children of battered women are worried, frightened, confused and vulnerable: they are worried about the safety of their mothers, themselves, and their siblings; they are frightened, not just by what they have seen and heard, but by what is yet to come; they are confused about what is happening and all the conflicting emotions they have both around the abuse and the abuser; and they are vulnerable; vulnerable because of their dependence upon and relationship with their mothers, vulnerable because of the lack of stability, security, and protection afforded by their environment, and vulnerable because of the lack of power they have over the onset or outcome of the violence. You do not need to have an intimate knowledge of abuse, or to have grown up in a violent home in order have some understanding or sense of how traumatic and disruptive witnessing the abuse of your mother and living with the myriad forms of fallout, could be, for any child. Common sense alone tells us that the impact is likely to be profound.
II. THE RISKS AND CONSEQUENCES OF LIVING WITH WOMEN ABUSE
The risks posed to children who live and learn within the context of domestic violence are substantial. Indeed, living in a domestically violent home may constitute one of the most serious, frequently encountered, and potentially ongoing threats to the health, welfare and safety of many children and young adults. It is of real interest and concern then, that this ‘risk’, and these children, legally and socially speaking, are hardly visible. This is clearly evident in the distinct lack of guidelines, services, or legislation directed towards addressing the special circumstances and needs of children who witness and live with the abuse of their mothers.
However, perhaps the most poignant indicator of the risk and invisibility of children of battered women lies in the sheer number of children who are believed to live in homes where their mothers are abused. Although we cannot determine with any degree accuracy (there are no epidemiological studies either in New Zealand or overseas to document in a precise fashion the incidence or prevalence of children in this country who witness wife abuse, we do know that in 1991, women who sought help from Refuges indicated that 90% of their children had witnessed violence and that 50% of these children had themselves experienced abuse. Based on data recorded as part of the Hamilton Abuse Intervention Pilot Project (HAIPP) we also know that children were present during 87% of the incidents in which their parent was victimised, and that the children themselves were either the accidental or direct targets of violence in almost one in five of all recorded incidents. These statistics highlight with unmistakable clarity, New Zealand children’s intimate and inescapable connection with family violence. Indeed, on the basis of these statistics and conservative estimates of the number of women who seek refuge in this country, it appears safe to conclude that when talk about the risk and the impact associated with wife abuse, we are talking about a disturbingly significant number of children in this country.
In this article the literature on the impact of domestic violence on children is reviewed and the implications of this knowledge for children in a legal sense is explored. Risk factors associated with witnessing domestic violence are identified. These include the trauma associated with witnessing violence, the social and behavioural problems evident in many of these children, their vulnerability to physical and psychological abuse within the context of domestic violence, and their ongoing susceptibility to the threat and actuality of violence throughout the dissolution process. Where possible results of research conducted with children of battered women in Aotearoa/New Zealand are included. The implications of our knowledge of these children’s experiences and the rationale behind the legal prioritisation of the safety of children of battered women in custody and access decisionmaking is then briefly discussed.
III. THE TRAUMA OF WITNESSING THE ABUSE
Although research has yet to verify on an empirical level the extent to which child witnesses’ initial responses to the abuse of their mothers are consistent with the broader concept of traumatic stress, clinical accounts and the findings of related research clearly indicate the traumatising potential of living with woman abuse for the children of battered women. That many children are initially shocked, unsettled and potentially ‘overwhelmed’ by what they have seen and heard is most evident. Children in Refuges have been noted by staff to experience sleep disturbance, to be disinterested in food, and to show obvious signs of psychosomatic distress (eg. headaches, stomach aches, asthma, and ulcers). In addition, children in Refuges have consistently been described as distraught, confused, concerned, fearful, untrusting of others, unsettled, anxious, insecure, and emotionally needy. That children of battered women may experience loss of safety and control during a battering episode is also relatively self-evident.
However, perhaps the most crucial consideration in our efforts to understand the nature of this trauma, is the realisation, that for many children, witnessing the abuse of their mothers is not an isolated incident. Many children, over an indefinite period of time, witness both the physical and psychological victimisation of a caregiver. This understanding raises serious concerns regarding the potential cumulative impact of these experiences. Beyond what children see and hear, the frequency, severity and inescapability of the violence in combination with a consideration and understanding of the child’s dependence upon and relationship with both the victim and the perpetrator, are all factors capable of compounding the stress and trauma experienced by these children. During the battering phase and for an indefinite period of time afterwards, many of these children’s reactions will no doubt be consistent with the broader concept of traumatic stress. Their world at this point in time is tense and traumatic, fragile and uncertain, and their reactions reflect this
IV. SOCIAL FUNCTIONING AND ADJUSTMENT
From a psychological perspective it follows that children who are experiencing trauma of this magnitude, would concomitantly experience problems in other important social and developmental spheres. Research conducted by Jaffe and colleagues indicated that child witnesses were below their peers in many areas associated with social competence. Some light was cast on the possible origin of these reported differences by de Lange. In an observational study, de Lange described Refuge children as socially isolated and estimated that as many as forty percent of these children were unable to join in the activities of, or relate to the interests of their age mates, and struggled to interact appropriately with either peers or adults. From a different, but still relevant perspective, Rosenberg and Rossman drew attention to the ‘isolationist’ and transitional nature of life for many child witnesses of domestic violence, reporting that children of battered women tended to move more often than their peers, were more likely to attend multiple schools, and were the most likely to have poor school attendance records. Clearly, the isolationist and disruptive context conferred by the presence of violence in the family can interfere with the development of social competence in these children. But what about their social networks? Are children of battered women isolated from their peers and/or other important sources of support?
In our first small Aotearoa/New Zealand based study, we investigated the degree to which children perceived themselves to be isolated from three broad but significant sources of support (their peers, family, and teachers). Our interest in children of battered women’s subjective appraisals of how socially supportive their immediate social network is, stemmed from two sources: recent findings which indicate that children’s perceptions of social support relate directly to their level of psychological wellbeing and adjustment, and concerns regarding the impact abuse has recently been found to have on children of battered women’s social network structure and involvement. It seems that these children’s opportunities to learn, engage in, and access socially supportive responses, may be particularly limited given the social isolation that characterises abusive families in general.
Accordingly, we found significant differences between the Representative Population Subsample and Refuge children’s perceptions of peer support, family support, and the quality and availability of support overall. Although children’s perceptions cannot be interpreted as a ‘true’ measure of the responsiveness and availability of children’s social network, there is some evidence to suggest that children’s perceptions may be the strongest predictor of children’s adjustment. This is not very surprising given that, in this instance, when we assessed children’s perceptions, we were in effect tapping children’s ideas and perceptions about their own isolation, their own ‘acceptability’ within their social network, and their own expectations or learning experiences regarding the availability and dependability of others.
All the evidence to date indicates that children of battered women are isolated from their social network, in both a situational and perceived sense. However we need to be aware that when we talk about lower levels of ‘social competence’ among children of battered women, we are in part, talking about the constraints of their environment; and the potential tyranny and control so evident in abusive environments, has to interfere with, or prevent, children from partaking in such rudimentary social activities as going to someone else’s home, or having someone over to play. As well we are talking about how these responsibilities or transient existence could hinder their involvement in school, social, and extracurricular activities. Perceptually speaking, we cannot determine on the basis of these findings, the extent to which children of battered women’s poor perceptions reveal real limitations in their social networks, or the degree to which those real limitations are influenced by their relational history or ability to both engage in and facilitate socially supportive behaviours. However, these children’s perceptions do reflect a lack of confidence in the availability and/or approachability of others; a lack of confidence in the ability of their social network to met their needs; and in part, a sense of themselves as unworthy or undeserving when it comes to receiving help from others.
Whatever the underlying factors, Refuge children’s perceived isolation from their peer group raises serious concerns, particularly in light of recent findings which indicate that peer relationship difficulties in childhood are predictive of serious adjustment difficulties later in life. Children whose behaviour is difficult and whose home life is precarious risk isolating themselves from their peers, especially if their efforts to cope with the overwhelming nature of their situation ‘spills over’ into their interactions with or perceptions of others. Given the current weight of support lent to the notion of intergenerational transmission of violence, this risk, and the mechanisms by which children’s experiences of abuse in the context of the family, influence their perceptions of, and approach or access to relationships with others, is of great interest and concern.
V. CHILDREN OF BATTERED WOMEN AND THE INTERGENERATIONAL TRANSMISSION OF VIOLENCE
Children of battered women may be at risk of developing emotional and behavioural problems because of the learning opportunities associated with family violence. Specifically what is of greatest concern in terms of socialisation, is that witnessing violence in the context of the family will teach children that violence is an extremely powerful and effective means of control, and that women and children are both legitimate and deserving victims. So, are child witnesses in danger of learning the same lessons and repeating the same patterns? Children are vulnerable yes, largely because parents are often the most salient and influential models in their young lives. This assertion is supported by retrospective accounts by battered women and their partners which clearly indicate that exposure to family violence predisposes children in general, although male children in particular, to be abusive in their own intimate relationships.
The danger, and hence the risk associated with children’s learning opportunities and the cycle of violence hypothesis, is however complicated by several factors. While a large body of research has clearly documented the role of modelling in both the acquisition and modification of aggressive behaviour, its unidirectional approach (ie. parents influence their children’s behaviour because children model this behaviour on that of their parents) precludes a more sophisticated and comprehensive overview of the developmental progression of aggressiveness. Patterson’s social interactional model of aggression has highlighted these limitations by recognising the reciprocal influences of parent and child in the development and expression of aggressive behaviour. The child is an integral part of this system who learns not only to respond in like, but eventually to elicit the kind of aversive interchanges they have witnessed and experienced. In line with the arguments of Patterson, it is argued here that child witnesses have the opportunity to learn more than just the execution and mimicry of behaviours; they have the opportunity to learn rules and regulations regarding the acceptability and appropriateness of behaviour.
Unfortunately we still lack a great deal of information regarding the conditions under which children will adopt observed behavioural and interactional patterns as their own. Retrospective evidence suggests that the personal experience of abuse may be an important or salient factor in determining whether or not individuals are likely to become embroiled in the cycle of violence. But apart from this distinction, there is not at the present time, enough social information available to be able to determine with any degree of accuracy which children, or more specifically what child characteristics are the most susceptible or vulnerable regarding exposure to violent models. Some recent studies suggest that the acquisition of antisocial and generally aggressive behavioural patterns may require more subtle cognitive and/or effective changes in addition to the vicarious observational learning presumed to underlie the modelling hypothesis.
It is argued here, as it has recently been argued elsewhere that children’s social-cognitive appraisals are likely to play an extremely important, although as yet, largely unexplored part. Based upon their own social-cognitive appraisals, and the meaning they attach to their experiences, children no doubt begin to form expectations about the roles of themselves and others in relationships, about the appropriateness and acceptability of a whole range of behaviours, and about the nature of responsibility. On a behavioural level, children are more likely to adopt and perceive aggressive and abusive behaviour as appropriate when such behaviour finds support in cultural norms and is accompanied by rationalisations, or under conditions which fail to highlight the irregularity or excessiveness of the behaviour.
In light of this it seemed conceivable to us, that the well documented familial experiences acting on children’s behavioural development, could also influence their social-cognitive development and social information processing patterns. It also seemed conceivable that children’s social-cognitive functioning, specifically their perceptions or interpretations of social situations could be affected, as a function of their family experiences. Accordingly, we tentatively hypothesised that family experiences could influence the quality of children’s interpersonal relations and behavioural adjustment via the influence they exert on children’s interpersonal problem solving skills. Refuge children were indeed found to be significantly less resourceful and behaviourally flexible in their approach to interpersonal problem solving. Refuge males distinguished themselves from children in the representative population subsample by their overly aggressive and often inappropriate responses. Female Refuge children distinguished themselves by the passive and aggressive extremes of their interpersonal problem solving responses.
This, along with findings which indicate that aggressiveness and shyness/withdrawal are significant antecedents of peer rejection, does not bode well for refuge children’s sociometric standing among peers and suggests that children’s perceptions of and approaches to interpersonal problem solving situations may be important determinants of children’s peer status. If this is indeed the case we have reason to be concerned, especially given recent findings indicating that disturbed peer relations are predictive of later maladjustment. However, even though refuge males were found to be more aggressive with their peers than refuge females, the implications for the girls are potentially more serious given that the standards for the acceptability of aggressive behaviour within the peer group are harsher for girls than for boys. Although children’s reactions to wife abuse are likely to vary over time, the fact that violence has already seeped into their relations with peers and has been incorporated as part of their problem solving and behaviour regulating repertoire suggests that there is a distinct possibility that these children have a significant potential to perpetrate the cycle of violence.
VI. BEHAVIOURAL FUNCTIONING AND ADJUSTMENT
The small but steadily growing number of empirical studies undertaken in this area reveal that children who are exposed to such emotionally and anxiety promoting events as wife abuse, are more likely than children who are not, to develop short-term, and possibly long-term adjustment difficulties. On the basis of these undertakings researchers have been able to infer that these children are vulnerable to a host of behavioural and adjustment difficulties ranging from poor concentration and erratic school attendance, to such emotional, behavioural and self-concept disturbances as withdrawal, anxiety, and a proness to depression. However, in light of the violent, aggressive and antisocial nature of domestic violence, it is of little surprise that the bulk of research on these children has sought to identify and quantify the aggressive and disruptive elements of child witnesses’ behaviour in relation to their peers.Accordingly, the presence of externalising (aggressive, disruptive, and delinquent) behaviours are widely cited in the literature. Significantly, the vast majority of overseas research has established that the externalising behaviour problems of children of battered women frequently exceed clinical cut-off margins, and borders, in the most severe cases, on delinquency.
These results are clearly indicative of the serious nature of many of these children’s adjustment difficulties. Of real concern are recent indications that children of battered women in Aotearoa/New Zealand, are experiencing problems of a comparable magnitude. In our 1994 study, Refuge children evidenced behavioural difficulties to a far greater degree than Representative Population Subsample children overall, and on each of the assessment tools subscales (ie. hyperactive, anxious, and aggressive). More than three quarters of the Refuge children (88.2%) were reported as having behavioural problems severe enough to fall within the clinical range. In stark contrast, just over one quarter (28.2%) of Representative Population Subsample children were reported by their parents as evidencing behavioural problems of a clinical nature. Notably, the number of children in the Representative Population Subsample identified by their parents as behaviourally problematic (ie. within the clinical range of behavioural problems) is quite comparable to prevalence rates reported by McGee, Silva, and Williams, who sought to investigate the prevalence, nature, and stability of behaviour problems (as measured by the Rutter) in a large sample of seven year old New Zealand children. This compatability between our findings and theirs strongly suggests that we have found meaningful differences between Refuge children and the representative population subsample in terms of clinically defined behavioural problems.
In many respects, the finding that children of battered women are significantly more behaviourally troubled (particularly with respect to the aggressive and anxious subscales) than a representative sample of their peers is not surprising considering the violent, aggressive, and distressing nature of domestic violence. It seems reasonable to assume that the high level of behavioural problems reported among the children of battered women who participated in this study, relates directly back to their family context, and the nature of their stressful circumstances. There are a vast array of broad yet interrelated factors capable, either singly or in combination, of significantly contributing to the severity of behavioural problems identified here. Children of battered women are vulnerable. Their hyperactive and anxious behaviours have the potential to further exacerbate their vulnerability, or at the very least place unwanted strain upon their efforts to develop, grow, and adjust. Indeed, the degree of anxious behaviour evident in this child population suggests that they are troubled emotionally.
However, the seriousness of many of these children’s difficulties is at the same time, complicated both by the questionable cultural appropriateness and relevance of the behavioural measure employed for Aotearoa/New Zealand, and by our understanding of the relationship that exists between maternal wellbeing and child adjustment . It is clear that our use of parental report data to measure child behaviour may be potentially misleading to the extent that the parental reports in this study contained a component of variance which was directly related to parental health. Importantly however, when comparable groups were compared (ie. non-clinical and clinical samples) on the basis of the parents’ responses to the General Health Questionnaire, Refuge children still distinguished themselves from their peers on the aggressive subscale. Moreover, for many of the children in this study, exposure to violent and abusive behaviour was not restricted to the mere observation of their mothers’ victimisation. Over fifty percent of Refuge children were identified by their mothers as having personally experienced abuse. Not surprisingly, these children were identified as having the most serious behavioural problems.
VII. CHILDREN’S EXPERIENCES OF ABUSE
Not only would a discussion of the behavioural problems evident in child witnesses not be complete without drawing attention to the differential impact of experienced abuse, it would be misleading. There is a very real reason to believe that many child witnesses suffer repeated exposure to violence, both as direct and indirect victims. That children of battered women may be ‘doubly’ exposed to violence, has serious implications for their adjustment given that child abuse in and of itself is an extremely traumatic and usually detrimental experience with its own established socio-emotional and behavioural sequelae. To date, personally experienced abuse is the greatest risk factor to have been identified for children of battered women, largely because of what these experiences are likely to convey to the child, as well as the sheer number of child witnesses who are thought to experience repeated exposure to violence, both as direct and indirect victims.
However, although children’s personal experiences of abuse have emerged as potentially more traumatic, developmentally hazardous, and generally speaking, more detrimental for children, (at least one that has more serious implications with regard to children’s behavioural functioning), the extent of behavioural problems evident in children who have witnessed but supposedly not experienced abuse, suggests that there are important shared similarities and consequences inherent in living in an abusive environment, regardless of who is the direct victim. These observed and documented similarities raise several important questions: to what extent are recent attempts in the literature to arbitrarily categorise children as “abused” or “not abused” necessarily partial and over simplistic? To what degree do these similarities reflect similarities in the circumstances that give rise, or contribute to, abuse in the family, regardless of whether the victim is the mother or child? However, perhaps the most pertinent questions are those centring around the as yet unexplored role of emotional abuse in child witnesses’ adjustment. Can witnessing the abuse of a caregiver be defined as an abusive experience, ie. does witnessing violence involve “mental or emotional injury”?
Current definitions describe “feeling threatened and living in fear” as an integral part of psychological abuse, wherein the meaning of the word ‘threatened’ includes “those threats made directly to the child, and those that are made indirectly, for example, where one parent verbally, psychologically, and/or physically abuses or threatens to abuse another in the child’s presence”. On the basis of this definition, children of battered women are clearly highly vulnerable to psychological abuse, and the fear, powerlessness, frustration, anger, low sense of worth, guilt, shame, embarrassment and grief associated with this type of maltreatment. Of course we still have some way to go in both defining and explicating the nature and the consequences of witnessing and living with abuse for children. In order to attain a more complete, coherent, and complex picture of both child witnesses and the nature of their adjustment difficulties, we as researchers will need to be sensitive to and start to concentrate our efforts upon exploring and elucidating the more subtle and individual influences or impacts that living with woman abuse has on children’s developing view of themselves and their world. Exploring the degree to which children’s reported experiences of woman abuse are consistent with current definitions of emotional abuse will clearly contribute to both our understanding of these children’s predicament and our ability to intervene in a sensitive and appropriate way. But we also need to explore and understand the link or nexus between domestic violence and child maltreatment. As yet, neither our understanding of the of the dynamics of violent relationships, nor the predicament of battered women and their children, has had a significant impact on how child abuse within the context of these relationships is identified, understood or managed.
VIII. THE LINK BETWEEN WOMEN BATTERING AND CHILD ABUSE
Traditionally, woman abuse and child abuse have been examined and addressed as distinctly separate issues in the field of family violence. However recent research indicates that we need to seriously question the validity of this prevailing assumption. A clear picture of the overlap has yet to emerge, but the link between woman abuse and child abuse certainly represents a significant and important area for research, policy, and intervention. So, what do we know about the link? We now know that children of battered women may be ‘doubly’ exposed to violence, either accidentally or intentionally. Using data from the Hamilton Abuse Intervention Project, Robertson and Busch reported that fifteen percent of children present during violent episodes may have become the unintended victims of violence either through attempts to intervene or their proximity to their mother at the time of the assault.
But what about the instances of intentionally directed abuse? Unfortunately, the fact these children’s mothers’ are usually the primary targets of abuse, does not mean that these children are less likely to be victimised: quite the contrary. At present we know that battering is the most common context for child abuse, with children of battered women being fifteen times more likely to be abused than children whose mothers are not battered. We also know that men who abuse their partners are more likely to abuse their children. From a feminist perspective it is not surprising that where there is woman abuse there is also likely to be child abuse, that the severity of the abuse inflicted upon the mother is predictive of the severity of abuse inflicted upon the child, or that the way in which children are victimised strongly resembles that experienced by their mothers. These trends indicate that like their mothers, many of these children may also be abused by their fathers or by the significant male in their mothers’ lives. Statistically speaking, child witnesses are in fact three to six times more likely to be abused by their father or father surrogate than any other adult in the family.
However, while children of battered women are approximately three to six times more likely to be abused by their father or father surrogate, there is also evidence to indicate that some of these children are being abused by their mothers. Straus reported that approximately twenty-five percent of battered women recounted abuse towards their children, double the child abuse rate reported by mothers who were not battered. It is worth noting, however, that the number of battered women directly implicated in the abuse of their children arguably remains surprisingly small in light of their own experiences of abuse and their often disproportionate responsibility for child care.
Even though comparatively only a small percentage of battered women are directly implicated in the abuse of their children, in a social and often professional sense, battered women are held responsible for and treated as complicit in, the abuse of their children. This is clearly evident in the disproportionate number of children who have been removed from women who are battered and in a recent American report released by the National Council of Juvenile and Family Court Judges wherein it was declared that a double standard is most evident in child protection, ‘failure to protect’ and custody cases. It is important to note that this report also went on to specifically state that this double standard is more often than not, exacerbated when the mother is also clearly a victim of abuse. Within the report numerous examples of case law are cited to exemplify how mothers are often prosecuted without full recognition of the issues involved in the abuse of women.
Unfortunately our lack of recognition and our lack of understanding of the issues involved in the co-existence of wife abuse and child abuse has some potentially serious implications: the possibility that battered women will not be identified when their children come to the attention of Child Services, and as a result, may subsequently become inappropriately implicated in the abuse of their children; or that in neglecting to differentiate between abuse and periodic ‘unavailability’ we will mistakenly assume that battered women’s and their children’s interests are diametrically opposed, thereby negating the influential and crucial part many battered women have to play in their children’s adjustment. Clearly, the abuse of a parent has serious implications for children’s relationships with both their parents, and for the security, consistency and stability afforded by these children’s environments.
IX. PARENT-CHILD RELATIONSHIP AND MATERNAL STRESS IN THE CONTEXT OF DOMESTIC VIOLENCE
Children of battered women may also be at an increased risk for developing emotional and behavioural problems because of the disruption in familial and parental functioning associated with family conflict and violence. In the first instance, the development of a strong and secure attachment or bond between parent and child could be undermined in an abusive context if the abuser perceives his partner’s relationship with a child as threatening, or if the child’s mother is abused during her pregnancy, and/or after the birth of her child. A rather rapidly growing body of research indicates that the ability of parents to provide children with the kind of consistent structure and involvement they need and demand is greatly undermined in the context of abusive and discordant relationships, ie. in this context parents have been found to communicate less clearly with their children, to decrease their use of approval statements whilst increasing their use of disapproval statements, to attend primarily to the negative behaviour of their children, and generally, to engage in the kind of behaviour that is often likely to facilitate and perpetuate negative parent-child interchanges.
Not surprisingly, changes in parenting practices and parents’ relationships with their children constitute the most significant indirect mechanisms to have been identified yet. Unfortunately at present we know very little about children of battered women’s relationships with their abuser-fathers, but given findings which indicate that in domestically violent homes, the way in which children are victimised strongly resembles that of their mothers, the possibility that these fathers are often punitive, power assertive, and perhaps violent in their relationships with their children is very real. We do however know more about battered women’s relationships with their children, and the significant part women’s experiences of abuse, and maternal stress play in determining both parent and child outcomes.
Indeed three studies have examined the significance of maternal health in relation to behavioural problems evident in children of battered women. Importantly, in all three studies, a strong relationship was found between mothers’ general level of health functioning and children’s level of adjustment. But the strongest relationship was found by Holden and Ritchie. They investigated the relation, or more precisely, the degree of interdependence between extreme marital discord, parenting, and child development and were able to account for an impressive 33% of the variance with maternal stress and parental irritability emerging as the two most significant predictors of child behaviour problems.
However, given that the data gathered in all three studies was based largely on mothers’ reports, it is possible that battered women were more likely to report feeling stressed in their role as parents than comparison mothers, and that their assessments of their children’s behaviour were subsequently unduly influenced by the stress and strain they themselves were experiencing at the time. Moreover, on the whole, these studies, due to their correlational nature, tell us very little about the quality of battered women’s relationships with their children, the difficulties and complexities of parenting in an abusive relationship, or the specific aspects of parenting that are of particular relevance or concern to battered women and their children. It is easy to see how a women’s ability to parent in a consistent, involved, and nurturant manner could be disrupted and undermined by her own responses and reactions to the abuse she is experiencing. So although maternal stress and wellbeing have consistently been identified as “the most important factors in a child’s psychological development and wellbeing over the years”, we know little about how, or the extent to which, women’s experience of intimate abuse make ‘mothering’ a more acutely stressful and potentially hazardous experience.
Research the authors are in the process of undertaking represents an initial effort to try and explore, elucidate, and make visible, some of the added difficulties and complexities of parenting for women who have, or who are, continuing to experience and endure physical and emotional abuse. Preliminary findings indicate several areas of consensus: that women experience violence as a process of victimisation, that they are acutely aware of how their ‘process’ impacts directly on their ‘availability’ to their children, and the regret, guilt, and resentment this evokes; that their children are focal, both in the kinds of decisions they make around the abusive relationship, and in terms of the concerns they have regarding their children’s responses and reactions to their abuse; that loyalty, discipline, responsibility and blame are consistently identified as difficult issues and sources of tension between battered women and their children; that the satisfaction these women glean from ‘mothering’ stems in part from the importance and centrality of their children in their lives, and in part from their ability over time to reconstitute a sense of themselves as worthy, capable, and deserving of the mother role; that women carry with them into their relationships with their children an overwhelming sense or need to compensate for their children’s experiences; and the ways in which women feel their efforts are undermined by the harsh realities and custody battles many they face when they decide to make a life for themselves and their children, independent of their partners. Although little attention has been directed towards what happens after the immediate crisis, the costs associated with leaving an abusive relationship are often not just high, but ongoing. As women and children seek to re-establish their lives in the community, at home, at work, and at school, many stressors, although slightly different, are still apparent. But perhaps the gravest concern for battered women and their children are those that centre around their safety during the process of separation.
X. WOMEN AND CHILDREN’S ONGOING SUSCEPTIBILITY TO THE THREAT AND ACTUALITY OF SEPARATION VIOLENCE
Despite persistent cultural beliefs that women and children will be safe once they leave an abusive relationship, evidence suggests that women and children are extremely vulnerable at this time. Although, in research terms, data on the incidence of separation violence in New Zealand is not currently available, overseas studies indicate that violence directed towards women and their children, is in fact likely to escalate once women have sought physically and/or legally to end the relationship. An investigation by Langen and Innes demonstrated that as many as seventy-five percent of visits to hospital emergency rooms by battered women took place after separation, while Stark and Flitcraft found that on the whole seventy-five percent of the cases where Police were called upon to intervene involved the victimisation of a women by ex-partner or spouse.
However the seriousness of the ongoing threat posed to women after separation is perhaps best conveyed by the homicide statistics. One study revealed that half of the female homicide cases in that year were committed by their ex-partner or spouse, after separation or divorce. Another in Philadelphia found that a quarter of the women murdered by their abuser were killed after separation. Several women have also been killed whilst attempting to ‘leave’. Sadly, a recent tragedy in New Zealand brought home the risks associated with leaving a violent relationship for women and children in this country. In 1994, three children were murdered by a batterer/father as he exercised his custody rights: rights which their mother had strongly contested in what was a long, losing and in hindsight, tragic custody battle.
In light of the seriousness of separation violence and the obvious danger that surrounds the dissolution process for women and children, it is imperative that we understand how contesting custody and gaining access privileges, can provide batterers the opportunity to continue to exercise, either in a directly abusive or more subtly intimidating way, ‘control’ over ‘their’ family. Several overseas studies document this potential. In Canada, Leighton found that as many as a quarter of women involved in custody proceedings received explicit threats against their lives during custody visitations. Similarly Shepard reported that sixty percent of the women in her study were subject to ongoing psychological abuse in the form of threats; these threats however, usually involved, or centred around the children.
Children it would seem, not only provide batterers with the opportunity for ongoing contact with their mothers’ recent evidence indicates that the children themselves may become both the source, and in some instances the target, of the batterers ongoing exploitation. Walker and Edwall found that abusive men could successfully continue to harass their ex-partners and spouses by returning children late, demanding to see the children at unscheduled times, making unscheduled visits, and removing children from the home without legal authority to do so. Perhaps the most extreme example of such custodial interference is that relating to the abduction of children. It is worth noting that over half of all reported child abductions in the United States occurred within the context of woman abuse.
Many children then are indirectly, but unmistakably, drawn into the conflict. Unfortunately a woman’s efforts to free herself and her children from violence, may be meet with a stronger likelihood of the children becoming more focal in the conflict than ever, as their ex-partners or spouses, consciously or not, attempt to undermine, manipulate and control her actions through her concern for the children. The nature of the impact of these experiences on children has yet to be explored, but the extent of their involvement, the stress, and the lack of security associated with the presence of such an ongoing threat, would undoubtedly jeopardise, or place unwanted strain upon their development and adjustment. These studies clearly highlight the danger the surrounds separation and access. In so doing, they raise some serious concerns: concerns regarding the comprehensiveness of legal protection accorded to women and children throughout the dissolution process; concerns regarding the appropriateness of legal decisions which may directly or indirectly compromise this precarious safety; and concerns regarding the potential deployment of custody as a battering tool.
XI. THE LEGISLATIVE RESPONSE AND THE RATIONALE FOR SUPERVISED ACCESS
The provisions of the 1995 Guardianship Amendment Act are designed to prioritise the safety of the child by mitigating against the risk of further violence by the batterer-father. The implications of our knowledge of these children’s experiences and the rationale behind the legal prioritisation of the safety of children of battered women in custody and access decisions will only be discussed briefly as detailed discussions of this issue are available elsewhere. The following discussion centres around contentions we believe establish a firm basis or rationale for supervised access in the context of domestic violence.
Clearly the collective wisdom that both parents are fundamentally equal in the investment they have in promoting their child’s wellbeing, and in the decision making power they have with regard to children in a co-operative joint custody arrangement is rendered meaningless when one parent uses violence to ensure the compliance of the other, and in so doing, jeopardises the physical and emotional wellbeing of their children in the process. Joint custody, and/or ‘friendly parent’ presumptions are not just untenable, they effectively undercut the safety and autonomy battered women seek, both for themselves and their children, when they leave an abusive relationship.
Effectively, the Guardianship Amendment Act provisions offer safeguards for women and children from ongoing violence and harassment. We now know that perpetrators of domestic violence contest physical custody of children more often than non-abusive men, and that their requests for custody are often an attempt to continue to exercise, either in a directly abusive, or more subtly intimidating way, control over their partners. As a society and as professionals, we need to ensure that in the context of domestic violence, the safety of children of battered women is not compromised by a shift in focus to the rights of battering men as parents. The guise of the batterer as a reasonable man/father is no longer tenable, and there are real grounds both legally and psychologically speaking to address requests on the part of battered women to limit and monitor the role of the abusive party in her and her children’s lives. Circumstances specific to the abuse of women demand a carefully planned, considered, and cautious approach. The higher the number of precautions, the less the risk that children will have to witness the continuation, and often escalation of violence.
Our third contention concerns the centrality and significance of women’s wellbeing in their children’s development and adjustment. A rapidly growing body of evidence now exists which indicates that the ‘best interests’ of children of battered women are uniquely tied to the wellbeing of their mothers. Jaffe and Austin reported that children whose mothers had been protected by legal sanctions from ongoing violence evidenced a considerable degree of recovery. Conversely Shepard found that the psychological abuse of mothers during access visits was significantly correlated with children’s adjustment problems. Clearly, legal efforts to protect and support the ‘psychological parenthood’ of battered women are paramount. Instances of severe, clearly defined maternal maltreatment withstanding, what is in the ‘best interests’ of battered women will almost certainly be in the ‘best interests’ of their children.
Of course last but not least is the contention that the rights and wellbeing of children of battered women have been undermined and threatened in such a way that safeguards and protective measures are necessary. These children, as we have seen, are highly vulnerable to both physical and psychological abuse. This knowledge, coupled with evidence which indicates that the risk of child abuse increases with the advent of divorce and that children at this time are likely to become more focal in the conflict and hence more vulnerable to coercion and manipulation indicates that these children’s physical and psychological safety is extremely precarious in the access context.
XII. PRIORITISING CHILDREN’S SAFETY: BENEFITS AND REQUIREMENTS
Prioritising children’s safety through supervised access will go a long way towards facilitating the physical safety of children. In this respect the benefits to children are immediately apparent. However children may also benefit in other ways from the legal parameters that regulate their relationships with their fathers. Access to fathers may in some circumstances help children adjust more quickly to their situation, especially when one considers how in the advent of violence, children’s separation from both their fathers’ and their homes’ is often abrupt and dramatic. Children may also experience a degree of security and control they have never known in their relationships with their fathers, and there is a good chance that the time they now spend with their fathers will better directed towards meeting their relational needs. There is of course also the significance of the message: that their fathers’ behaviour is negatively sanctioned by the legal system. In this way children’s own feelings regarding their responsibility for, or their role in, the violence will be directly challenged.
For the new legislation to really work in the ‘best interests’ of children will require more than a legal commitment to upholding these interests. It will require recognition of both the practical and psychological needs that will as a matter of course, arise from children’s new found situation. From a practical perspective, the dangers inherent in child custody and access within the context of domestic violence will require ongoing court supervision of access arrangements, as well as the implementation and co-ordination of services and safe locations for access exchanges. From a psychological perspective, parenting education and support groups for parents and children will be needed. Supervised access provisions will raise several issues for children. Children will need ongoing assistance in learning how to come to terms with the past, and adjust to their current situation. Essentially supervised access represents a difficult struggle to ensure children’s safety, while at the same time, addressing in as full a manner as possible, children’s needs, wishes and concerns.
XIII. A CAUTIONARY NOTE
While prioritising children’s safety through supervised access will go a long way towards facilitating the ‘physical’ safety of children, questions remain regarding the extent to which legal stipulations have adequately addressed the ‘psychological’ safety of children. Unfortunately, verbal abuse, intimidation, harassment, and manipulation, all of which constitute psychological abuse, are not sufficient to invoke the safety presumption of Subsection 16B. Given these constraints, supervised access arguably does not, in and of itself, guarantee or ensure children’s safety. Indeed the diverse array of abusive behaviours battering men employ in the course of their relations with their partners could very well extend to and continue to play a part in battering men’s relations with their children. Clearly, due consideration has in this instance, not been given to the dynamics of family violence, the interconnected nature of various abusive behaviours, their functions, or their consequences. The potential for children’s psychological safety to be undermined within the access context will need to be closely monitored if children are to be extended the kind of comprehensive protection the recent amendments to the Guardianship Act (1968) sought to ensure.
Circumstances specific to the abuse of women demand a carefully planned, considered and cautious approach. Such an approach requires more from the social service and legal communities than a knowledge of the threat posed to women and children both during and after the dissolution process and an understanding that mediation and joint custody are inappropriate in the given circumstances. What it requires is that this knowledge and these considerations are reflected, and evident in every decision that is made pertaining to custody and the conditions of access, that the safeguards required to achieve this are firmly in place, and that these checks extend beyond the decision making process to ensure that the custodial arrangement continues to work ‘in the best interests’ of the children involved. Providing the protection accorded to battered women and their children is comprehensive enough, these legal stipulations should go some way towards protecting children from being used as pawns in on-going abuse scenarios by batterers whilst ensuring that women and children are safe from the threat of further violence. The success of this new legislation will be conditional in part, on the ability of the legal and social service communities to continue to review and evaluate the ‘legal process’ in relation to the health and safety needs of women and their children, the development and implementation of specialised services and procedures (eg. screening protocols, demonstration projects, supervised access centres), and on the strength of our political and social commitment to protecting and upholding the ‘best interests’ of children.
[*] MA(Hons) (Auckland), Research Officer, Department of Psychology, University of Auckland.
** PhD (Otago), Lecturer in Social Psychology, University of Auckland
 National Collective of Independent Refuges Inc. Submission to the Family Court(Unpublished Paper, 1991).
 Maxwell, Children and Family Violence. The Unnoticed Victims, Report for the Office of the Commissioner for Children (1994).
 Arroyo and Eth, “Assessment following violence witnessing trauma” in Peled, E (ed)Ending the Cycle of Violence: Community Responses to Children of Battered Women(1995). See also Eth, S and Pynoos, R, Post Traumatic Stress Disorders in Children(1985).
 Hilberman and Munson, “Sixty battered women”  2 Victimology 460.
 Hughes, “Psychological and Behavioral Correlates of Family Violence in Child Witnesses and Victims” (1988) 58 American Journal of Orthopsychiatry 77.
 Jaffe, Wolfe, Wilson and Zak, “Similarities in Behavioural and Social Maladjustment Among Child Victims and Witnesses to Family Violence” (1986) 56 American Journal of Orthopsychiatry 142.
 “Spousal Violence and Children’s Social Competence” (1986) 56 American Journal of Orthopsychiatry 255.
 “The Child Witness to Marital Violence” in Ammerman, R and Herson, M (eds),Treatments of Family Violence: A Sourcebook. (1989).
 Pocock, T and Cram, F Children of Battered Women: The Forgotten Victims of Domestic Violence (Unpublished Masters Thesis, University of Auckland, 1994).
 Dubow and Ulman, “Assessing Social Support in Elementary School Children: The Survey of Children’s Social Support” (1989) 18 Journal of Clinical Child Psychology 52.
Rosenberg and Rossman, supra n. 8.
 Sarason, Pierce, Shearin, Sarason, and Waltz, “Perceived Social Support and Working Models of Self and Actual Others (1991) 60 Journal of Personality and Social Psychology 273.
 Parker and Asher, “Peer Relations and Later Personal Adjustment: Are Low Accepted Children at Risk?” (1987) 102 Psychological Bulletin 357.
 Widom, “Does Violence Beget Violence? A Critical Examination of the Literature” (1989) 106 Psychological Bulletin 3.
 Forstrom-Cohen and Rosenbaum, “The Effects of Parental Marital Violence on Young Adults: an Adult Exploratory Investigation” (1985) 47 Journal of Marriage and the Family 467.
 Bandura, A, Aggression: A Social Learning Analysis (1973).
 Coercive Family Process (1982).
 Kalmuss, “The Intergenerational Transmission of Marital Aggression” (1984) 46 Journal of Marriage and the Family 11..
 Jaffe, P, Wolfe, D, and Wilson, S, Children of Battered Women (1990).
 Christopolous, Cohn, Shaw, Joyce, Sullivan-Hansen, Kraft, and Emery “Children of Abused Women: Adjustment at Time of Residence” (1987) 49 Journal of Marriage and the Family 611. See also Hertzberger, “Social Cognition and the Transmission of Abuse” in Finkelhor, D (ed), The dark side of families: Current Family Violence Research (1983).
 Hertzberger, ibid., 22.
 Smetana, Kelly and Twentyman “Abused, Neglected and Non-maltreated Children’s Perceptions of Moral and Social Transgressions” (1984) 55 Child Development 277.
Pocock and Cram, supra n. 9, at 10.
 Renshaw and Asher, “Children’s Goals and Strategies for Social Interaction” (1983) 29 Merrill-Palmer Quarterly 353.
 Parker and Asher, supra n. 13.
Cairns, R, Cairns, B, Neckerman, Gest, and Gairiepy, “Social Networks and Aggressive Behavior: Peer Support or Peer Rejection?” (1988) 24 Developmental Psychology 815.
 Goodman and Rosenberg, “The Child Witness to Family Violence: Clinical and Legal Considerations” in Sonkin, D (ed), Domestic Violence on Trial: Psychological and Legal Dimensions of Family Violence (1987).
 Davis and Carlson, “Observation of Spouse Abuse: What Happens to the Children?” (1987) 2 Journal of Interpersonal Violence 278.
 Fantuzzo and Lindquist, “The Effects of Observing Conjugal Violence on Children: A Review and Analysis of Research Methodology” (1989) 4 Journal of Family Violence 77.
 Christopolous, et. al., supra n. 20; Mathias, Mertin, and Murray, “The Psychological Functioning of Children from Backgrounds of Domestic Violence” (1995) 30 Australian Psychologist 124; and also Sternberg, Lamb, Greenbaum and Cicchetti, “Effects of Domestic Violence on Children’s Behaviour Problems and Depression” (1993) 29 Developmental Psychology 44.
 Pocock and Cram, supra n. 9, at 10.
 “Behaviour Problems in a Population of Seven-Year-Old Children: Prevalence, Stability and Type of Disorder: A Research Report” (1984) 11 Australian & New Zealand Journal of Family Therapy 65.
 Hughes, “Psychological and behavioral correlates of family violence in child witnesses and victims” supra n. 5.
 Cicchetti, “How Research on Child Maltreatment Has Informed the Study of Child Development: Perspectives from Developmental Psychopathology” in Cicchetti, D and Carlson, V (eds), Child Maltreatment (1989).
 Pocock and Cram, supra n. 9, 9-38.
 Ibid., 20.
 McDowell, H Emotional Child Abuse and Resiliency: An Aotearoa/New Zealand Study (Unpublished Doctoral Thesis, University of Auckland, 1995).
 Ibid., 43.
 Pocock, T and Cram, F, Mother-Child Relations in Wife Abusive Homes, Study in progress (1996).
 Hart, B, Children of Domestic Violence: Risks and Remedies (Unpublished Paper presented to the Fourth International Conference on Family Violence in Durham, South Carolina, 1995).
 Cappell and Heiner, “The Intergenerational Transmission of Family Aggression” (1990) 5 Journal of Family Violence 132.
 Bowker, Arbitell and McFerron, “On the Relationship between Wife Beating and Child Abuse” in Yllo, K and Bograd, M (eds), Feminist Perspectives on Wife Abuse(1988). See also Stark and Flitcraft, “Woman-Battering, Child Abuse and Social Heredity: What is the Relationship in Marital Violence?” in Johnson, N (ed), Sociological Review Monograph: No. 31 (1988), 147.
 Edleson, J and Schecter, S, Model Initiatives Linking Domestic Violence and Child Welfare (Briefing paper presented at the conference on Domestic Violence and Child Welfare: Integrating Policy and Practice for Families in Racine, Wisconsin, June 8-10, 1994).
 Hughes, supra at n.5.
 Hamilton Abuse Intervention Pilot Project (HAIPP): A Two Year Review , Report No. 5 (1994).
 Roy, M Children in the Crossfire: Violence in the Home: How does it Affect our Children? (1988).
 Stark and Flitcraft, supra n. 43.
 Stacey, W and Shupe, A,The Family Secret: Domestic Violence in America (1983).
 Stark and Flitcraft, supra n. 43.
 Stauss, “Ordinary Violence, Child Abuse and Wife Beating: What do They have in Common?” in Finkelhor, D (ed), supra n 20.
 Stark and Flitcraft, supra n. 43.
 Ibid. See also Pocock and Cram, supra n. 40.
 Stark and Flitcraft, supra n. 43.
 Davis, “Failure to Protect and Its Impact on Battered Mothers” (1995) 1 Courts and Communities: Confronting violence in the family 6-7.
 Bullock and McFarlane, “A Programme to Prevent Battering of Pregnant Women” (1988) 11 Response to the Victimisation of Women and Children 18.
 Emery, “Interparental Conflict and the Children of Discord and Divorce” (1982) 92 Psychological Bulletin 310.
 Jouriles, Barling and O’Leary, “Predicting Child Behaviour Problems in Maritally Violent Families” (1987) 15 Journal of Abnormal Child Psychology 165.
 Zussman, “Situational Determinants of Parental Behaviour: Effects of Competing Cognitive Activity” (1980) 51 Child Development 792.
 Jouriles, Murphy and O’Leary, “Interspousal Aggression, Marital Discord and Child Problems (1989) 57 Journal of Consulting and Clinical Psychology 453.
Stacey and Shupe, supra n. 49.
 Wolfe, Jaffe, Wilson and Zak, “Children of Battered Women: The Relation between Child Behaviour, Family Violence, and Maternal Stress” (1985) 53 Journal of Consulting and Clinical Psychology 657.
 “Linking Extreme Marital Discord, Child Rearing, and Child Behaviour Problems: Evidence from battered women” (1991) 62 Child Development 311.
 Wallerstein, J, and Blakeslee, S, Second Chances (1989).
 Pocock and Cram, supra n. 40.
Bowker et. al., supra n. 43.
 Preventing Domestic Violence against Women, Bureau of Justice Statistics Reports (1986).
 Supra n. 43.
 Bernard, Vera, H, Vera, M, and Newman, “Till Death Do Us Part: A Study of Spouse Murder” (1982) 10 Bulletin of the American Academy of Psychiatry & the Law 271.
 Caazenave, N and Zahn, M Women, Murder and Male Domination: Police Reports of Domestic Homicide in Chicago and Philadelphia (Unpublished Paper presented at the American Society of Criminology Conference, Atlanta, 1986).
 Leighton, B, Spousal Abuse in Metropolitan Toronto: Research Report on the Response of the Criminal Justice System, Report No. 1989-02 (1989).
 “Child Visiting and Domestic Abuse” (1992) 71 Child Welfare League of America 357.
 “Domestic Violence and Determination of Visitation and Custody in Divorce” in Sonkin, D (ed), supra n. 28.
 Grief, G and Hegar, R, When Parents Kidnap (1992).
Bowker, et. al., supra n. 43.
 Jaffe, Wolfe, and Wilson, supra n. 19.
 Geffener and Pagelow, “Meditation and Child Custody Issues in Abusive Relationships” (1990) 8 Behavioral Sciences and the Law 115.
 Shepard, supra n. 72.
 Geffner and Pagelow, supra n. 77.
 Wallerstein and Blakeslee, supra n. 64.
 Jaffe, P and Austin, G, The Impact of Witnessing Violence on Children in Custody and Visitation Disputes (Unpublished Paper presented at the 4th International Conference on Family Violence, New Hampshire, 1995).
 Supra n.72.
 Klosinski, “Psychological Maltreatment in the Context of Separation and Divorce” (1993) 17 Child Abuse and Neglect 557.
 McDowell, supra n. 38. See also Stark and Flitcraft, supra n. 43.
 Finkelhor, supra n. 20.
 Walker and Edwall, supra n. 73.
 Busch, Safeguarding the Welfare of Children, (1995) 4 Butterworths Mental Health & the Law Bulletin 46.
Hart, supra n. 41.