Enhancing Safety: When Domestic Violence Cases are in Multiple Legal Systems
(Criminal, family, child protection)
A Family Law, Domestic Violence Perspective

Part 6: Risk Assessment, Continuing Domestic Violence

6.1 Introduction

After determining the type of domestic violence, the next step is assessment of risk. This requires, in addition to ascertaining the type of domestic violence, detailed scrutiny of patterns of behaviour. When deciding what information can, should, and must be shared across court systems, one of the most important factors is level of risk, including potential for lethal outcome.

In the following discussion, the terms 'potential for lethal outcome' and 'potential for lethality' are used interchangeably. A number of domestic violence researchers use the term 'danger' to describe the same phenomenon.

While many of the indicators of potential for lethal outcome are the same as those for risk of continuing domestic violence, others are unique to lethality. In other words, domestic violence perpetrators who kill have at least some characteristics that distinguish them (as a group) from perpetrators of repeat but non-lethal violence. Therefore, in this manual, we consider risk and the potential for lethality separately.

Violence and abuse in the home harm not only the person targeted, but also any children who reside there. In a small but significant number of cases, perpetrators kill former intimate partners, or the children, and then themselves by committing suicide.Footnote 65 Safety measures should be specifically designed to address the case-specific particulars of the risk of continuing violence and of the potential for lethal outcome.

6.2 Information Sources

Domestic violence advocates, transition house workers, victim services professionals, as well as domestic violence academic and professional experts in the community can offer invaluable knowledge and or experiential insight and assistance pertinent to risk.

In addition to consulting such experts, lawyers representing clients in domestic violence cases will also want to take the time and initiative to become knowledgeable about the particulars of police, Crown, and victim policies and services; transition house and longer term safe housing; domestic violence intervention and parenting programs for perpetrators; supervised child access programs; alcohol and drug and mental health treatment programs; services relating to culture and disability; as well as victim, child, and family domestic violence support services in the community.

Finally, one should keep in mind that persons targeted by domestic violence are often themselves the best source of information relating to risk and potential for lethal outcome.

6.3 Risk and Lethality: Similarities and Differences

To reiterate: it is important to distinguish facts and characteristics associated with the onset of domestic violence, from facts that indicate the likelihood domestic violence will continue, from facts associated with the potential for lethal outcome. While some facts - such as the pattern of abusive and violent conduct - are relevant to more than one category, others relate to only one.

For example, researchFootnote 66 has revealed that depression and suicidal thoughts are associated with a potential for a lethal outcome but not necessarily with the likelihood of repetitive domestic violence; witnessing domestic violence as a child is associated with the likelihood a person will engage in at least one incident of domestic violence as an adult, but has not been identified as a good predictor of whether or not a particular person will continue to engage in domestic violence. Access to guns is empirically linked to potential for lethal outcome but has not been considered an accurate predictor that a person will continue to engage in repetitive non-lethal domestic violence.Footnote 67 Therefore, Part 6 discusses continuing risk, while Part 7 discusses potential for lethal outcome.

6.4 Indicators of Risk of Continuing Violence

A systematic review of the research reveals that the following facts are associated repeatedly with continuing domestic violence:

Family lawyers should consider as well indicators of risk outlined in part 6.4.1 below.

6.4.1 Risk Factors Identified in Some Studies, not in Others

Facts that have been associated with enhanced risk of continuing domestic violence in many studies but not in others are outlined below. One should consider these facts when considering risk and the need to implement safety measures, particularly if the facts identified in part 6.4 are present.

When a collection of facts associated with risk appear in evidence, one should weigh carefully whether or not information with respect to risk should be reported to police, victim services and/or to child protection authorities.

For additional information on risk: See Praxis International Blueprint for Safety http://www.praxisinternational.org/praxis_blue_print_for_safety.aspx, particularly "Practitioners' Guide to Risk and Danger in Domestic Violence Cases" at page 14.

6.5 Sharing Information Relating to Risk

While the best option for sharing information pertinent to risk is with client consent, in its absence, service providers and lawyers will want to weigh carefully client concerns relating to safety, privacy and liability associated with revealing information, on the one hand, and child and adult safety concerns associated with failure to divulge, on the other. Most Professional Codes of Professional Conduct, including those for lawyers authorize the revelation of confidential information in the face of imminent risk of harm to identifiable persons. See, for example, Chapter IV, Rule 2 'Public Safety Exception' p. 17 of Canadian Bar Association (2009) Code of Professional Conduct. See also Rule 3.3, and 3.3.3 of the Federation of Law Societies of Canada (2012) Model Code of Professional Conduct, approved December 2012, which states that a lawyer may disclose confidential information, limited to what is required, on "reasonable grounds that there is an imminent risk of death or serious bodily harm." The Code advises consideration of likelihood, imminence, absence of alternative means to prevent injury, and the circumstances under which the information was acquired. Consider also the factors to be taken into account, when assessing whether public safety outweighs solicitor-client privilege, identified in Smith v. Jones, [1999] 1 S.C.R. 455; 169 D.L.R. (4th) 385.

Government officials and service providers (and lawyers) will want to consider provisions relating to freedom of information, privacy and public safety in pertinent privacy and personal information protection legislation.Footnote 74

Note that many Access to Information and Privacy statutes authorize the release of information without consent in specified circumstances in order to protect health or safety of an individual. See, for example section 42 (h) of Freedom of Information and Protection of Privacy Act R.S.O. 1990, Chapter F.31. Note also, however, the qualification in a number of these statutes such as "compelling" and the duty to give notice.

Best Practice from a domestic violence safety perspective: British Columbia's Freedom of Information and Protection of Privacy Act [R.S.B.C. 1996] Chapter 165. This statute authorizes the release of personal information relating to risk of domestic violence. Section 33.1 subsection (m.1) states expressly that personal information may be released for "the purpose of reducing the risk that an individual will be a victim of domestic violence, if domestic violence is likely to occur". Note, however, that while the provision authorizes release of information by service provides and other officials, lawyers are additionally bound by solicitor-client professional privilege and codes of professional conduct.

In connection with limitations on capacity to disclose information obtained during discovery processes, see 8.7.

6.6 Culture, Age and Social Status can Increase Risk

The following situational and cultural factors are commonly associated with increased risk: being Aboriginal/First Nations, Métis, Inuit; being young (18-25); having a physical or mental disability; being a member of a disadvantaged cultural group; being isolated from sources of help as a result of religious belief, culture, or as a result of rural location; being poor; being in a same sex-relationship; living common law or in an unmarried intimate relationship; being pregnant; experiencing mental health or substance abuse problems or being involved (or formerly involved) with a violent intimate partner who has such problems. In addition, rates of domestic violence are known to increase in times of emergency, social upheaval and stress. Such circumstances can increase risk either directly or indirectly by limiting access to support services. In these circumstances, one should pay special attention to safety measures and attend to any obstacles that limit access to support services specific to the cultural or social context.

6.7 Targeted Party Fear

Reminder: victim fear has been documented empirically as one of the most accurate predictors of future domestic violence. When victims are frightened, lawyers and service providers should take note and check for the presence of other risk indicators - matching safety provisions to level of fear.

6.8 Concluding Comments on Indicators of Risk

If a collection of indicators suggests heightened risk of physical domestic violence, one should undertake protective measures, considering carefully the need to exchange information about risk with service providers and professionals who can provide protection and support in other court sectors.

6.9 Information Exchange Protocols

Given the confusion surrounding the circumstances in which information on risk can be disclosed in the absence of consent, consider engaging the community in the development of information-sharing protocols across legal systems. The purpose should be to identify the circumstances in which service providers, victim services, police, child protection authorities, court coordinators, lawyers, and other professionals may share information in connection with high risk and particularly changing risk (such as when an alleged perpetrator has dropped out of mandated programmes, has breached a no contact order, or has dropped out of therapy). Clear information exchange rules can reduce delays, help to promote safety, save professional time, and avoid cumbersome applications to courts.

Note the importance of ensuring protection of confidential information from victims that might adversely affect safety, and the duty not to share more confidential information than is necessary to prevent harm.

A number of jurisdictions, for example Alberta, Nova Scotia, and parts of Ontario and British Columbia, have implemented cross sector committees to advise on and respond to the need for collaboration and the swift information exchanges in high risk domestic violence cases.Footnote 75

6.10 Risk Assessment Tools

6.10.1 Introduction

The list of indicators of risk identified above is drawn from ten years' scrutiny and analysis of risk assessment research conducted in Canada, the United States, Australia, England and New Zealand.Footnote 76 The indicators listed emerge consistently across studies and jurisdictions. They are presented here for the use of lawyers and other professionals to enable a matching of services and safety measures to level of risk.

Nonetheless the indicators listed are not weighted for predictability, meaning that while they are useful in ensuring that appropriate safety measures are put in place matched to level of risk, the lists do not constitute a predictive risk-assessment tool.

The term 'risk assessment tool' refers to actuarial and other professional assessment tools designed to help police and other professionals assess and predict the risk that domestic violence will occur in the future. Two of the most researched domestic violence risk assessment tools in use in Canada today are the Spousal Assault Risk Assessment (SARA) and Ontario Domestic Assault Risk Assessment (ODARA). Both have some degree of research verification although controversies relating to predictive validity remain.Footnote 77

In connection with potential for lethal outcome, see Part 7 below.

6.10.2 Strengths and Limitations

All domestic violence risk prediction tools have limitations. Research indicates that such tools, in 20 to 30 % of cases fail to predict continuing physical domestic violence. Furthermore, in about 20 % of cases, the tools identify as high risk perpetrators who fail to engage in further domestic violence.Footnote 78 Nonetheless, research is also indicating that risk assessment tools can accurately predict between 66% and 77% of continuing violent domestic violence, depending on the tool used, and that the use of tools is an improvement over professional judgement alone.

Consequently, information from risk assessment tools can be particularly helpful in a criminal police context IF considered along with careful scrutiny and analysis of detailed evidence of domestic violence and the individual circumstances of each case.Footnote 79

In the family law and child protection context, however, the usefulness of these tools is rather limited since they focus on the risk of continuing physical violent action (for example assault) to the exclusion of other forms of domestic violence that can be equally damaging to families and to children (such as continuing harassment, economic and psychological abuse, child abuse, and forms of manipulation, coercion and negative parenting).

Nonetheless, given that risk assessment can help lawyers anticipate (and respond to) the level of risk of continuing physical violence, family lawyers representing survivors of domestic violence may wish, particularly when the risk of continuing violence is thought to be moderate or high, to seek risk assessment conclusions from the police. This information can be taken into account during settlement discussions in order to ensure that adequate attention is paid to preventing risk of future continuing physical violence. It can also be presented, along with other factors applicable to risk, to family or child protection courts.

Similarly, family law lawyers representing perpetrators of domestic violence may wish to seek risk assessment information from police to present to the family court if it is thought that the assessment indicates low risk.

Anticipate the potential for delay and difficulty in gaining access to information, however. Refer to Part 8 below for potential options when the police are unable to consent to release of information.

6.11 Changing Circumstances

One must keep in mind that risk is situational and changes with circumstances. It can increase (for example when the perpetrator is no longer employed, during times of stress and emergency, or when the targeted party seeks to relocate). It can also decrease (for example following intervention programs, following acceptance of separation, following treatment for mental health problems, or when the targeted party is well-protected by community services). Consequently, in addition to considering information, if any, from risk assessment tools, one should continuously monitor facts associated with risk (outlined above), taking into account the support networks available to the parties. Safety plans should be revised accordingly.

6.12 Risk Assessment in Family and Child Protection Cases, Admissibility and Use

Generally, in family law cases, the relevance and probative value of safety and risk assessment tool evidence is likely to outweigh potential prejudice because of the way the evidence is used. Concerns about potential prejudice from false positives, while important, will be of less concern in family than in criminal cases where a false positive could affect personal liberty and freedom. In family law cases, the evidence is considered and used for a different purpose; it is used preventively to identify the need for protective measures, to identify service needs, and to respond to the best interests of children.

Consider as well that domestic violence risk assessment evidence should not be considered conclusive, particularly if other evidence indicates the need for safety precautions. See, for example: Roach v. Kelly, 2003 CanLII 1991 (ON S.C.) (CanLII).

Finally, one should keep in mind that domestic violence risk assessment tools tend to measure the continuing risk of merely some forms of physical domestic violence. Thus they can be useful as a reminder that safety measures are needed but should not be used to discount the need for protective measures, particularly the need for protection from other forms of domestic violence in a family law context. This issue is discussed more thoroughly in part 6.14.

6.13 Caution: Meaning of a "Low Risk of Domestic Violence" Assessment

What does it mean when a witness testifies that a violator has scored 'low risk' on a recognized domestic violence risk assessment tool (such as SARA or ODARA)? It means only that other people who have committed violent acts of domestic violence, who have similar attributes and who have faced similar circumstances, have tended not to engage in repetitive domestic violence. It also means that most people who engage in repetitive domestic violence have different attributes and circumstances. It is thus not an absolute finding. And, since risk is situational, it can change rapidly as circumstances change. In addition, a low score does not rule out the possibility that a particular domestic violator has an unusual set of circumstances that are not measured by risk assessment tools. Risk and safety assessment should be periodic, not a one-time occurrence.

Moreover, in a family law context, since the risk assessment tools tend to focus on violent physical acts, as opposed to other forms of domestic violence, a low risk assessment may not offer much reassurance with respect to other forms of domestic violence, including forms of coercion associated with child abuse and poor parenting.

6.14 Safety for Children: Current Risk Assessments Tool Limitations

Given that domestic violence risk assessment tools relate primarily to physical violence between adult intimate partners or former intimate partners and are not designed to assess risk to children, they should not be used in a family law context to assess the safety of children, or to justify denial, reduction or delayed access to assistance or as a replacement for detailed consideration of factual evidence.  For example, in Roach v. Kelly, 2003 CanLII 1991 (ON S.C.) (CanLII) cited earlier, despite expert assertions of a low risk of domestic violence but moderate risk of violence, the trial judge relied on evidence of past conduct and denied reinstatement of supervised access on the basis that the mother's terror would have a negative impact on her parenting and the family should not have to live in constant fear.  (The father had used supervised access to question the child about the mother and child's whereabouts.)

Reminder: the operative factor for children is the level and effect of stress, including the level of continuing conflict between the parents; continuing trauma-related harm from past domestic and family violence in the home; the effects of contact on the child and on the care giving parent; the presence or absence of parenting practices that mirror the coercive elements of the domestic violence; and the presence or absence of child abuse. The psychological elements and effects of coercive domestic violence have as much, perhaps more, long-term impact on children. Nonetheless when risk of physical DV is high for a parent or a parent is in danger, children are in danger too. Thus, while risk assessments should not be used to assess children, when risk assessments indicate high risk and particularly the potential for lethal outcome for a parent, children require safety and security measures. For further discussion, see Part 7.

6.15 When Should a Domestic Violence Expert be Called In?

The Supreme Court of Canada recognizes, in R. v. Lavallée, [1990] 1 S.C.R. 852, [1990] 4 W.W.R. 1, (1990), 55 C.C.C. (3d) 97, (1990), 76 C.R. (3d) 329, (1990), 67 Man. R. (2d) 1, (1990), 67 Man. R. (2e) 1, 1990 CanLII 95 (S.C.C.) (CanLII), that domestic violence is a complex phenomenon lying outside the experience and understanding of most Canadian judges and lawyers. Misconceptions in the domestic violence field are common. Consequently, expert evidence will often be helpful particularly when issues associated with potential risk and safety are unclear; when additional information is needed to assess whether or not safety measures are needed; when each partner has been violent and is making allegations of domestic violence against the other; when it is necessary to understand the psychological impact of domestic violence on a child or on a targeted parent; when it would be helpful to the court to understand connections between perpetration of domestic violence and perpetration of child abuse; when evidence is insufficient to enable the court to conclude whether or not the targeted person (adult or child) requires protection from future domestic violence, child abuse, or destructive and manipulative parenting.

Basic domestic violence training seldom qualifies an evaluator an expert. Many 'experts' who conduct parent-child evaluations (assessments) for courts lack specialized knowledge of domestic violence. Assessing the needs or interests of children in a domestic violence context requires considerable knowledge not only of the complexities of domestic violence but also of the developmental and social needs of children. While Canada lacks national standards defining a 'domestic violence expert', there are a number of questions one can ask in order to help to identify an expert:

In the absence of domestic violence expertise, assessments of parents and children in a domestic violence context can be misleading. One should try to ensure that evaluators who assess parents and children in domestic violence cases are recognized domestic violence experts or, if that is not possible, that the evaluators consult with a domestic violence expert.

One of the problems at the intersection of criminal, child protection, and family law proceedings is that experts, who regularly conduct assessments in legal proceedings, possess different types and levels of expertise. For example, professionals who evaluate child best interests in a child protection context or in a family law context may have appreciable expertise in connection with child development but may have limited understanding of how domestic violence affects adult parenting or children. Police in the criminal sector may have received specialized training in the assessment of risk of continuing domestic violence in a criminal law context but may lack understanding of pertinent child development and child safety issues.Footnote 80 Moreover, few assessors have an adequate understanding of how an evaluation in one sector can affect the family in another legal context.

In order to maximize resources, consider arranging, in consultation with a domestic violence expert, joint consultations among experts across legal sectors to coordinate and consolidate pertinent information and to reach agreement on how the information will be used in the various legal systems.

6.16 When a Domestic Violence Expert May Not be Needed

Expert assistance and evidence can be time-consuming and expensive. Many families experiencing domestic violence have limited financial resources and thus limited ability to hire experts. Expert information may not be needed if the evidence is clear and the parties agree that the domestic violence was clearly minor and isolated. Expert information may also not be necessary when the level of risk is clear and safety measures have been carefully considered and put in place to protect the targeted parent and children.

When specialized domestic violence expertise has not been considered or is not available, the best course of action is to err on the side of caution and safety.