Making appropriate parenting arrangements in family violence cases: applying the literature to identify promising practices

2.0 Literature review on impact of family violence

2.1 Impact of Family Violence

Family violence is considered to be any form of physical, sexual, emotional, or psychological abuse that occurs in the context of family relationships. The term family violence encompasses child abuse and neglect, spousal violence (intimate partner violence), and elder abuse. Throughout this document the term family violence is intended to be inclusive of all forms of abuse in the family and the term spousal violence signifies abuse within the context of an intimate adult relationship. In the divorce literature, high–conflict couples are identified as those that require extensive and lengthy court involvement to resolve disputes post–separation. Family violence issues are present in a majority (but not all) of high conflict separations (Jaffe, Austin, & Poisson, 1995; Johnston, 1994). This distinction is important because not all conflict can be deemed violence, but conversely, violence should not be euphemized as conflict.

Family violence continues to negatively impact the healthy development of children and families across the country. In Canada, 27% of reported violent crime victims are victims of family violence, and similar rates have been documented in the US. In both countries the number of female victims outnumber the male victims by at least 300% in the context of intimate violence (Statistics Canada, 2004a; Bureau of Justice Statistics, 2000). These rates are comparable to those found in Europe, although the reports of estimated prevalence of family violence vary due to differences in definition, data sources and sampling (Hagemann–White, 2001; Kury, Obergfell–Fuchs, & Woessner, 2004). For example, the British Crime Survey estimates 26% of women and 17% of men are physically assaulted and/or threatened with violence by an intimate partner (Byron & Mirlees–Black, 1999). Similarly, estimates from the Australian Women's Safety Survey, which strictly focused on the prevalence of physical and sexual violence experienced by women and the nature of this violence, reported that 8% of women have experienced at least one incidence of violence, perpetrated by an intimate partner. These cross–national estimates capture the reported (actual or threatened) violent incidents from crime victim surveys.

There continues to be debate within the research literature, and among practitioners and other members of the violence prevention community, about using official crime statistics versus random surveys as tools for determining the incidence and prevalence of family violence. (Johnson & Bunge, 2001; Tjanden & Theonnes, 2000). There is general agreement that family violence is an underreported crime. There continues to be a lack of information nationally and cross–nationally regarding the likely number of unreported incidents, as well as the extent and trajectory of family violence. However, Canada has been a forerunner in collecting these data through methods other than crime surveys. Statistics Canada has completed several comprehensive telephone surveys on the topic of family violence (Statistics Canada 2001; 2004a, 2005). While these surveys suggest that rates of victimization of intimate partners is similar to other cross–national samples, there is particularly rich additional information that is captured in these surveys including trends, context, sentencing implications, family violence against children and youth, violence against older adults, and homicide risk. At one level, rates of victimization for females and males look very similar (7% of women vs. 6% of men reported being victims of an act of spousal violence in the previous five years); however, the additional contextual information identified important gender patterns in severity, impact and lethality of violence. Notably, these findings revealed that:

  • Female victims of spousal violence were twice as likely to suffer ten or more incidents of violence in comparison to male victims (Statistics Canada, 2005).
  • Female victims of spousal violence were significantly more likely than male victims to suffer injuries, require medical attention, lose time from work, live in fear, and worry about the safety of their children (Statistics Canada, 2005).
  • Data from the Homicide Survey (Dauvergne, 2003) indicate that between 1993 and 2002, women were four times more likely to be killed by their spouse (8 female homicide victims per million couples compared to 2 male homicide victims per million couples).
  • Cases of spousal homicide–suicide involve female spouses as the target in 97% of these cases (Statistics Canada, 2005).

The most recent survey completed looked at violence after separation and the association with child contact. Twenty–seven percent of estranged spouses with children under 18 years of age reported physical or sexual assault in the previous five years. More than twice as many abused spouses in comparison to non–abused spouses reported that their ex–spouse had no contact with the children (14% vs. 6%, Statistics Canada, 2005). Family violence has an impact on children in both direct child abuse and the indirect impact of exposure to spousal violence. This impact on children has garnered heightened awareness as scholars and those in the broader spousal violence network continue to call for better answers about how to accurately measure the incidence, impact and prevalence of family violence, its impact on family dynamics, and how to create meaningful interventions (Mears & Visher, 2005). While there has been considerable progress in the identification of cases and coordination of community responses to family violence, there is still much to achieve. In particular, the complexity of family violence and its impact on all facets of family functioning and child development is the source of ongoing efforts to improve intervention and prevention. There is a growing awareness of the need for longitudinal research on the impact of family violence on children. Challenging but important issues to study include research into what happens to children after parental separation, and what are the effects of different post–separation parenting arrangements on children who have experienced family violence.

While there have been numerous studies related to all forms of childhood victimization and its short term and long term effects on social, emotional, physical and psychological development, this research tends to be compartmentalized. That is, "a relatively small proportion of studies concerned with childhood violence has assessed participants for exposure to multiple forms of violence, multiple incidents of the same type of violence, or exposure to potentially stressful or traumatic events other than violence" (Saunders, 2003, p. 359). This lack of research speaks to the complexity of family violence and how the effects of violence can vary greatly based on an array of variables. Cunningham & Baker's (2004) recent exhaustive review of family violence and child maltreatment literature proposed a model that examines the variables hypothetically associated with the impact of family violence (See Figure 1). This illustration captures the complexity and the substantial number of variables that must be taken into consideration when examining the impact of family violence.

2.2 Child abuse[2]

The literature related to child abuse is dominated by empirical studies that examine the characteristics, behavioural and emotional impact (immediate and long–term implications), developmental considerations, and societal consequences of child maltreatment. The majority of studies have shown that maltreated children, when compared to children who have not experienced maltreatment, are more likely to display major behaviour problems and emotional difficulties (Egeland, Yates, Appleyard, & van Dulmen, 2002; Jungmeen & Cicchetti, 2003; Maughan & Cicchetti, 2001; Hildyard & Wolfe, 2002), demonstrate more discipline problems at school (Kendall–Tackett & Eckenrode, 1996), are more aggressive towards their peers or more socially withdrawn (Shields & Cicchetti, 2001), have fewer social skills (Levendosky, Okun, & Parker, 1995) and are more likely to be rejected by their peers (Shields & Cicchetti, 2001). The serious long–term effects of child maltreatment have also been noted, including adverse mental health, physical health impairments and societal consequences (National Clearinghouse on Child Abuse and Neglect, 2004a; Higgins & McCabe, 2003; Johnson et al., 2002). While there have been a number of studies that document the characteristics and behavioural impact of child maltreatment, there has not yet been enough research conducted on the relationship between the characteristics of maltreatment and the development of behavioural or emotional problems over time (Ethier, Lemelin, & Lacharit, 2004).

It is generally acknowledged in the literature that child abuse and its fundamental causes can be traced to various systems including the family, the community and larger society (Belsky, 1993; National Clearinghouse on Child Abuse and Neglect, 2004b). It is the family system and its impact on child maltreatment that is most important and pertinent to this discussion. In Canada, it is estimated that biological parents are responsible for the majority of child maltreatment, with approximately 90% of all instances of child abuse being committed by at least one biological parent (Trocme, MacLaurin, Fallon, Dciuk, Billingsley, Tourigny, et al., 2001)[3]. In addition, extensive research indicates perpetrators of spousal violence are significantly more likely than non–perpetrators to physically abuse their children (review in Bancroft & Silverman, 2002).

However, separation of the maternal caregiver from her abusive partner significantly reduces the risk for child maltreatment when spousal violence is reported (Cox, Kotch, & Everson, 2003). In these cases, the identification of these families as high risk may facilitate appropriate intervention and safety planning for the caregivers and their children.

Figure 1 - Variables Hypothetically Associated with Impact of Family Violence

Figure 1 Variables Hypothetically Associated with Impact of Family Violence

[ Description ]

Source: Cunningham, A. & Baker, L. (2004). What about Me! Seeking to Understand the Child's View of Violence in the Family. Available at

2.3 Exposure to spousal violence

There has been much research and policy focus on children exposed to spousal violence. The term "exposure" covers such a wide range of circumstances which include hearing a violent event, visually witnessing the event, intervening, being used as a part of a violent event (e.g., being used as a shield against abusive actions), and experiencing the aftermath of a violent event (Edleson, 1999c). The negative effects of childhood exposure to spousal violence have been presented in numerous studies and meta–analyses (Edleson, 1999a; Kitzmann, Gaylord, Holt, & Kenny, 2003; Wolfe, Crooks, Lee, McIntyre–Smith, & Jaffe, 2003). Most notably, research indicates that children exposed to spousal violence are more likely than other children to be aggressive and have behavioural problems (Graham–Bermann, 1998), have different physiological presentations (Saltzman, Holden, & Holahan, 2005), exhibit higher rates of Post–Traumatic Stress Disorder[4] symptomatology (Kilpatrick, Litt, & Williams, 1997), are likely to try to intervene on behalf of the victimized parent (Peled, 1998), and may also develop a 'traumatic bond' (a longing for kindness, leading to confusion between love and abuse) with the perpetrator (Bancroft & Silverman, 2002). In some cases, children express preference to live with the abusive parent, who is perceived as more powerful.

We are only beginning to understand the broader picture as it relates to children's exposure to spousal violence. Research related to the effects of being exposed to spousal violence has evolved over the past decade, but has largely relied on the reporting of victims or other adults (teachers, service providers, etc.) to identify the problematic effects using standardized measures (Ornduff & Monohan, 1999; Morrel, Dubowitz, Kerr, & Black, 2003). A recent review of available studies estimated that less than 20% (of 220 empirical studies) directly asked children for information (Cunningham & Baker, 2004). Recently, researchers have begun turning their attention to capturing children's voices and their experience of being exposed to violence. This research has shown that children usually are aware of the spousal violence that occurs in their family and also freely disclose incidents of their own abuse (Cunningham & Baker, 2004; Ornduff & Monahan, 1999; Holden, 2003). These first–person accounts from children that describe multiple forms of violence in the home converge with other research that indicates child maltreatment occurs most frequently in families where there is also spousal violence present (Edleson, 1999b; Hartley, 2002).

2.4 Overlap among child abuse, spousal violence and sibling abuse

Spousal violence and child abuse often occur in the same family and until recently very few interventions were targeted at addressing this duality in families (Straus & Gelles, 1990; Schechter & Edleson, 1999). The majority of studies reveal that in families where there is spousal violence or child maltreatment present, in 30% to 60% of the cases both forms of abuse exist (Edleson, 1999b; Appel & Holden, 1998). The impact on children in these families varies based on the degree and frequency of violence, how much is witnessed and how much is directly experienced, as well as risk and protective factors (Edleson, 2004). Risk factors such as young caregiver age, low education, and low income, and lack of the social support network compound the risk for child abuse associated with spousal violence (Cox et al., 2003). Emerging Canadian interventions, such as the Caring Dads program, recognize this overlap by providing intervention for fathers who have maltreated their partners and children. This program addresses both spousal violence and child abuse (Scott & Crooks, 2004; Crooks, Scott, Francis, Kelly, & Reid, in press)[5].

The presence of spousal violence also increases the likelihood of the presence of violence and abuse between siblings (Hoffman & Edwards, 2004). There are few studies that document the incidence and prevalence of sibling abuse, with some researchers suggesting that there are no systematic studies that address the incidence and prevalence of sibling abuse and its impact on future adult functioning (Graham–Bermann, Cutler, Litzenberger, & Schwartz, 1994). One of the most reliable studies, conducted well over a decade ago, reported that sibling abuse is the most common form of violence in the family, with 8 out 10 children reporting physical violence against a sibling (Gelles & Straus, 1988). In addition, parents may view the violence between siblings as mutual and therefore never really consider the possible perpetrator and victim roles that exist in sibling violence (Graham–Bermann et al., 1994).

While some degree of intersibling aggression is normal, more severe sibling abuse is a cause of concern, especially in families where there are other family violence issues. Recently, Wiehe's (1997) study on severe sibling abuse described a cascading effect, with the oldest sibling targeting the second child, and this sibling attacking the next youngest child. In this same study, victims of severe sibling abuse reported that their self–esteem and their ability to trust others were negatively impacted, resulting in future problems such as depression, substance abuse, and poor intimate relationships. For siblings who have unresolved abusive relationships throughout childhood, their opportunity to develop a mutually supportive and healthy adult intimate relationships may be compromised (Brody, 1998).

Beyond abuse by a sibling, children can be affected by witnessing a parent abuse a sibling, regardless of whether they themselves are targeted for abuse. Although few studies have been done in this area, it seems likely that witnessing a sibling being abused by a parent figure threatens the emotional security a child experiences (Cummings & Davies, 1996; Davies, Harold, Goeke–Morey, Cummings, Shelton, & Tasi, 2002). That is, the child may have a secure relationship with the parent, but the experience of seeing a sibling victimized by that parent may profoundly shape a child's view of the world and relationships. In this case, the child may be physically safe, but may suffer from anxiety related to the possibility that he or she might be a future victim. Furthermore, the observer child may feel guilty about being safe, or conversely, come to see the victimized child as deserving of the abuse, to make sense of the violence.

2.5 Variability of the Impact on Children

The serious implications for children who are maltreated or exposed to spousal abuse have been well documented. There are a number of studies which indicate that not all children who directly and indirectly experience family violence later develop severe emotional and behavioural problems (National Clearinghouse for Child Abuse and Neglect Information, 2004b). Cunningham & Baker (2004) caution against making assumptions that (1) all children are negatively affected by spousal violence, (2) all children are affected in the same way and (3) that spousal violence should be the sole focus of interventions. Outcomes of individual cases vary widely and are affected by a combination of factors, including the child's age and developmental status when the abuse or neglect occurred, the type of abuse (physical abuse, neglect, sexual abuse, etc.), frequency, duration, and severity of abuse, and the relationship between the victim and the abuser (Chalk, Gibbons, & Scarupa, 2002). These varying outcomes can be seen in families where children have similar risk factors and exposure experiences, but have very different short–term and long–term outcomes.

Researchers have begun to explore why some children experience long–term consequences of abuse and neglect while others emerge relatively unharmed under similar circumstances. The ability to cope effectively following a negative experience is sometimes referred to as "resilience." A number of protective factors may contribute to an abused or neglected child's resilience. These include individual characteristics, such as optimism, self–esteem, intelligence, creativity, humour, and independence; parent or family factors such as extended family support, highly educated parents, household rules and boundaries, and a caring adult in the child's life; and social factors such as community well–being, including neighborhood stability and access to health care (National Clearinghouse for Child Abuse and Neglect Information, 2004b).

2.6 Typologies of Family Violence

In the same way that there is variability among outcomes for children, there is also great variability among the patterns and contexts of violence between adults in a relationship. A thoughtful analysis of the impact of family violence must consider typologies of violence and the various contexts in which spousal violence can occur. A number of helpful typologies have been developed. The different types of spousal violence have different expectations of future dangerousness and require different social and legal interventions.

Johnston and Campbell (1993) were among the first to offer a model for understanding different patterns of spousal violence within high–conflict divorcing families, operating under the assumption that spousal violence arises from multiple sources and follows different patterns in different families. Recognizing that theories from the literature related to family violence are numerous (psychodynamic, biological, family systems, sociopolitical, etc.), these researchers created linkages between these theories to create five categories of spousal violence (with special consideration given to paranoid and psychotic forms of violence). These five types include:

1. Ongoing / Episodic Male Battering

This type of violence most closely resembles the traditional understanding about batterers as it relates to the cycle of violence theory. Men's perpetration of violence is attributed to "their low tolerance for frustration, their problems with impulse control, and their angry, possessive, or jealous reactions to any perceived threat to their potency, masculinity and 'proprietary male rights'." (p. 193). These men generally are a threat to women, and over time their propensity to use violence increases with the threat of separation and long after separation.

2. Female Initiated Violence

Women's use of violence (not in the context of self–defense) is seen as a reaction to their own stress and tension. While women demonstrate physical, emotional and verbal abuse within relationships, these acts do not affect the power differential between partners (in relation to perceived or actual power and control dynamics between partners).

3. Male Controlling Interactive Violence

This type of violence most closely resembles what has come to be known as "mutual violence". This type of violence arises out of a mutual disagreement or verbal altercation and escalates into a physical struggle. It should be noted that the term "mutual violence" is not without controversy, as most advocates and others working in the anti–violence field acknowledge that context and power dynamics are not often recognized in the understanding of this type of violence. Indeed, the name of this category has been identified as problematic due to the seeming paradox of "interactive" and "male controlling" (see Bancroft, 1998 for critique).

4. Separation / Divorce Trauma

This category refers to acts of violence which only occur about the time of separation, but were not present in the relationship prior to separation. Often, after an escalation of outrage, anger and abandonment, physical violence is typically perpetrated by the partner who is being 'left'. The violence does not develop into an ongoing pattern of violence, but stops following a few isolated incidents at the height of the separation.

5. Psychotic / Paranoid

The fifth category addresses violence that is associated with psychotic or paranoid reactions due to mental illness or "drug–induced dementia." Psychiatric treatment is recommended as the preferred intervention. However, Bancroft's critique (1998) notes that a person who batters and also has a mental health problem may have two important issues requiring multiple intervention strategies. Furthermore, treating the mental health problem alone may not eliminate spousal violence. Bancroft further argues that a perpetrator of spousal violence who has co–existing mental health problems may require an approach similar to the one needed for the substance abusing batterer; that is, both problems need to be specifically addressed in intervention.

Frederick and Tilley (2001) contend that "in order to intervene effectively, it is important to understand the (1) intent of the offender, (2) the meaning of the act to the victim and (3) the effect of the violence on the victim." (p. 1). They describe 5 contexts that must be considered when gathering historical information about spousal violence in a family. Thus, any act of physical aggression must be evaluated in the larger context of these factors. These include:

1. Generally violent (a "fighter")

Some people are violent regardless of the context. These are people who use violence in situations inside and outside of the family to resolve conflict or to satisfy aggressive impulses.

2. Battering

Battering consists of not only acts of violence and abuse, but is a component of a larger system of intimidation, control and isolation that purposefully puts the victim at a power disadvantage, severely compromising the victim's independence, self–esteem and safety. While some batterers are also "fighters," many are violent only in a familial setting.

3. Isolated act (not a "batterer")

The use of violence is highly uncharacteristic and not used in the relationship to exert power or control. The violent incident may occur in a highly stressful situation and the perpetrator normally recognizes the behaviour as inappropriate.

4. Mental incapacitation

Mental illness, substance use and dependency, and medications contribute to use of violence. For perpetrators who have some mental health impairment, their use of violence in a relationship may be illegal, but may reflect their mental health issues.

5. Responsive to battering (self–defensive)

Self–defensive violence is always in response to a partner's violence or threat of violence. The use of violence by this person is not part of an attempt to gain control of the relationship, but rather is a response to attempt to protect oneself or gain control in a particular, violent situation.

Depending on the combination of type of violence and the context, each situation can call for a different systems (criminal justice, civil justice including family law and child protection aspects, health care, etc.) response. Also, perpetrators of violence can fit into more than one context (i.e. they can be a batterer and be generally violent).

Another researcher who has argued for delineation of different patterns of spousal violence is Michael Johnson (Johnson, 1995; Johnson & Ferraro, 2000). His early work identified the important distinction between patriarchal terrorism and common couple violence. More recently, LaViolette has extended this framework to develop a continuum of aggression and abuse. This continuum conceptualizes spousal violence ranging from common couple aggression to terrorism/stalking (LaViolette, 2005). LaViolette has hypothesized a number of dimensions upon which the five (Johnson) types differ, including the contextual factors identified by Frederick and Tilley (2001). Figure 2 depicts this continuum and the characteristics of each type of aggression / abuse. Understanding the differences among these types of violence provides an important foundation for assessing the appropriateness of a particular post–separation parenting arrangement.

Examination of the various patterns of family violence also highlights gender differences that need to be discussed. A gendered analysis of family violence is a controversial topic that tends to divide both practitioners and researchers. There is no doubt that male perpetrated violence against women is most often reported to police, results in more serious physical injury, is associated with fear and concern about children's well–being, and accounts for the majority of domestic homicides (Statistics Canada, 1999; Tjaden & Thoennes, 2000; Ontario Domestic Violence Death Review Committee, 2004; Washington State Fatality Review Committee, 2004). At the same time, not all female perpetrated violence is in self–defense, and it is generally accepted that males are more hesitant to report victimization experiences to authorities.

Furthermore, although male domestic homicide victims constitute a minority of intimate partner homicide victims, these cases of victimization of male intimate partners present the same challenges for early identification and prevention. their victimization can have the same profound impact on children and extended family members. most recognized experts in the field would agree that one death is one too many and there is a paucity of research on violent relationships in which the female partner is the primary perpetrator. a similar gap exists for understanding same–sex intimate partner violence; this violence is underreported due to the need to disclose both intimate violence and sexual orientation to authorities who may be perceived to be homophobic.

Figure 2 - Continuum of Aggression and Abuse (La Violette, 2005)

Date modified: