On this page:

Violence Against Women: A literature review

« Previous | Contents | Next »

Listen

Violence against Women: A literature review commissioned by the National Group to Address Violence Against Women

4. Responding to violence against women

The starting point for this report is the assumption that all forms of violence against women are linked. However, the research literature focuses mainly on discrete aspects of violence against women, with different aspects attracting attention within different fields.

This section of the report reviews research which assesses the effectiveness of interventions to address violence against women, identifies some of the gaps in the literature, and attempts to assess how far what has been learned about one aspect of violence against women can be transferred to another.

4.1 Responding to rape and sexual assault

4.1.1 The criminal justice system response

The reporting and subsequent investigation and prosecution of rape and sexual assault are the focus of much of the available research literature on the subject, particularly in Britain. As already indicated, the establishment of specialist police units in Scotland to deal with sexual assault dates back to the mid 1980s, and followed highly publicised research which critiqued existing police practice (Chambers and Millar, 1983). The trend towards specialism in this area continued throughout the 1990s, and is now standard across the Scottish police forces. Although the model varies slightly from one force area to another, key components include dedicated interview suites, specialist officers, and a 'victim-centred' approach. In some areas there is also a dedicated forensic suite.

Practitioners acknowledge that there have been significant improvements in the police response to rape and sexual assault complainers over the last 20 years (Christianson and Greenan, 2001), and this is supported by research. A study of 23 women in Sussex who had reported to the police between 1991 and 1993 found that 57% of them felt mostly positive about the response of police officers, while 43% were mostly negative about the response of police officers (Temkin, 1997). None of the women, including those who felt negative about the service overall, felt that they were disbelieved, or that the police were 'heavy-handed ' in their approach. In addition, the majority of the women (19 out of 23) valued the manner and attitude of the police officers who dealt with them. For the women who were 'mostly negative' about their experience with the police, poor follow up, difficulty accessing information, disbelieving attitudes and insensitive handling were the main features of their complaints about the service.

Temkin concludes that "the believing, sympathetic, non-judgmental attitude of the police, the unpressured pace and supportive manner in which their statements were taken, the access which they had to police officers and to information thereafter and the help and backing they received...during the course of the investigation and afterwards" were the main reasons for women feeling positive about the experience of reporting (Temkin, 2001: 524). Follow-up, she maintains, continues to be a problem, particularly in the area of information on the progress of the case.

These findings are similar to those from a survey of 48 women who reported to police in New Zealand between 1990 and 1994, in which 40% of the women expressed some degree of satisfaction with the police response, and 38% were dissatisfied (Jordan, 2001). The author acknowledges some of the difficulties inherent in measuring 'satisfaction' with a process which by its nature is bound to be distressing. She notes:

"Because rape is such an intense and sensitive area, when the police act with professional caring and demonstrate their respect for the victim, this is noticeable and greatly appreciated. When such qualities are lacking, however, their absence is also very noticeable." (Jordan, 2001: 696).

She goes on to explore the balance to be struck between the need (of women) for the process to be manageable, and the need (of police officers) to focus on the end result of that process:

"...at the very time that a raped woman is seeking to be believed and validated, the police will be intent on obtaining proof and verification that she is telling the truth. Her need for validation may clash with the police search for verification, and the techniques used by the police in their quest for evidence may threaten and undermine her sense of confidence and safety in them. While she struggles to regain a sense of autonomy following the rape, the police feel they as professionals must retain control of the proceedings." (Jordan, 2001: 701).

Jordan asserts that the achievement of a sense of control over the proceedings need not be achieved by one party at the expense of the other. Citing Temkin, she notes the value women place on belief, respect for the complainer, and a non-judgmental approach by the police (Jordan, 2001).

Recorded crime statistics for Scotland show a steady increase in the reporting of rape (Scottish Executive, 2003). This picture is similar in England and Wales (Harris and Grace, 1999; HMIC/HMCPS, 2002) and across Europe (Regan and Kelly, 2003). However, in none of these jurisdictions has the increase in reported rapes been matched by an increase in prosecutions or convictions. In fact, the conviction rate for rape has fallen during the period in which the reporting levels have risen (Harris and Grace, 1999; Regan and Kelly, 2003).

In an attempt to identify some of the reasons for this, the Home Office requested a joint inspection by HM Inspectorate of Constabulary and HM Crown Prosecution Service Inspectorate into the investigation and prosecution of rape cases. Their report was published in April 2002. The terms of reference were:

"...to carry out an analysis of investigations, decision-making and prosecutions of allegations of rape, from initial report through to case disposal." (HMCPSI/HMIC, 2002: 2).

The review covered all offences of rape against women, men and children. In relation to the investigation stage, the key findings echo some of the research findings already discussed, included the need for consistent training of police officers and forensic examiners, and improvements in the physical environments in which interviews and examinations take place. In addition, the review identifies partnership working with other agencies (e.g. through dedicated sexual assault referral centres) as key to improving the response to victims. Improved and standardised recording systems, and a review of timescales needed for submission of files to the Crown Prosecution Service are identified as the main administrative improvements required.

Although the report does not consider the role of forensic examiners in detail, it does note the limitations on choice posed by the lack of female forensic examiners, and the implications of forensic examiners learning 'on the job' rather than through accredited training programmes. The report considers that quality of forensic evidence is crucial to effective prosecution of rape and sexual assault, increasing the likelihood that prosecution will happen, and that a conviction will result (HMCPSI/HMIC, 2002). In addition, it is suggested, any measures which reduce the trauma of the investigative process for individual women are to be welcomed, improving not only the likelihood of achieving a conviction, but also the woman's recovery rate following a sexual assault. This view is supported elsewhere in the literature (Campbell et al, 2001; Kelly and Regan, 2003).

In a study commissioned by Rape Crisis Network Europe (RCNE), Kelly and Regan reviewed recent literature on the conduct and outcomes of forensic examinations. They identify some of the key elements of good practice in relation to forensic examinations, emphasising "the rights and dignity of the victim" (Kelly and Regan, 2003: 6). These include "speedy response; avoiding the triage system in hospital A&E departments; a private dedicated space; a well equipped examination room; trained and skilled practitioners; female examiners; a streamlined victim-centred information gathering process; time to move at the speed the victim/survivor is comfortable with; protocols and evidence kits which are applied flexibly, according to the facts of the case; space to discuss the process, debrief and undertake crisis intervention, and provision of, or links to, medical follow up and advocacy/support services." (Kelly and Regan, 2003: 12)

These conditions, they argue, are crucial, both to the quality of the evidence gathered, and to the comfort and health of the complainers (Kelly and Regan, 2003: 12). In their subsequent review of five different approaches to forensic examination, they note that the use of trained doctors is one of the more common models. They identify a number of difficulties inherent in this approach, including problems with recruiting women doctors, the need for participating doctors to take on a generic forensic role, thus perhaps limiting their knowledge about sexual assault, and problems with limited availability of doctors at certain times. They also suggest that there may be limited co-ordination and integration across the services, and an absence of advocacy and support.

Some of these issues are addressed by the use of forensic nurses, who have a more extensive role in providing healthcare advice, advocacy and support to complainers, in addition to evidence gathering and providing forensic reports for the courts. This model is widely used in North America and has several advantages, including a higher degree of specialism, cost effectiveness, and a more holistic approach to health intervention following sexual assault (Kelly and Regan, 2003; Ledray, 1999).

A holistic approach is also found in the provision of Sexual Assault Centres, which in Canada are designed to "attend to the medical, emotional, social and medico-legal needs of clients in a prompt, professional, and compassionate manner and to provide leadership in the prevention of sexual assault" (Du Mont and Parnis, 2002, cited in Kelly and Regan, 2003: 15). These are usually hospital based, often attached to accident and emergency facilities, with a dedicated examination room and possibly interviewing facilities. There are several examples of similar centres in England, including The Haven, a referral centre based in a sexual health setting which provides forensic examination and sexual health follow up in southeast London (Kerr et al, 2003), and the St Mary's Centre in Manchester, which was the first such centre in England. As yet, there are no such facilities in Scotland, although discussions are ongoing in Glasgow about how such a service might be developed (Dutton and Cavanagh, 2003).

Kelly and Regan conclude by identifying the key components required in order to begin developing minimum standards:

"Privacy through the development of dedicated rooms, or a centre;

Philosophical principles underpinning practice that emphasises respect, dignity, rights and choice;

Enhancing forensic practice through capacity building - both the number of trained examiners (often through involving nurses) and their skills base;

Access to female examiners;

Ensuring that even if people have to wait for a medical practitioner, that a staff member is available to greet them, take them to the more private rooms, and explain their rights and what may happen next;

Linking provision of immediate medical care, forensic examinations, crisis and short term counselling, follow up medical care and advocacy;

Combining service provision with training, awareness raising and system advocacy;

Leadership within the service, and some form of community accountability;

Ensuring access is as wide as possible, and that outreach efforts are targeted at under-served populations."

(Kelly and Regan, 2003: 17)

Finally, they note the emerging debate about how far forensic evidence actually influences the outcome of sexual assault trials, citing Canadian research which demonstrates that only documented injury appears to have a predictive influence on convictions (Du Mont and Myhr, 2002; cited in Kelly and Regan, 2003). Given the trauma for women undergoing forensic examination, further research would seem to be indicated in this area.

In another report for Rape Crisis Network Europe, Regan and Kelly address the issue of attrition in reported rape cases, raising serious concerns about the extent to which convictions for rape have fallen across Europe, as shown in their pan-European study (Regan and Kelly, 2003). They maintain that this downward trend in conviction rates is contrary to what might be expected, given the role of the women's movement in raising awareness and challenging rape stereotypes, the development of rape crisis lines and other women's counselling projects, the development of training and practice guidelines, increased media awareness and legal reforms. However, they argue that it is symptomatic of a situation in which rape has received little attention compared with domestic violence and trafficking. The study indicates that countries with adversarial legal systems have the highest attrition rate - England and Wales, Scotland and Ireland all have conviction rates below 10%. At 6%, the conviction rate for rape in Scotland is second only to that in Ireland.

The Justice ministries for the countries involved in the study offered a range of technical and procedural 'barriers to successful prosecution', including limited or absent evidence, under-reporting or delayed reporting, lack of support services, delays in court proceedings and 'limited incentives for prosecutors'. The authors contend that, despite a wide range of legal and procedural reforms which have been enacted across Europe since 1980, there is still an absence of good practice in enabling rape complainers to give their best evidence or in supporting and protecting "their dignity and integrity" during the trial process (Regan and Kelly, 2003: 17).

Overall, they argue, rape is very much 'a forgotten issue' on political and social policy agendas, attracting neither the debate nor the resources which have gone into highlighting domestic violence as a social policy priority. They make a number of recommendations for change, including the suggestion that research should be undertaken to explore the points of attrition in rape cases and identify possible reasons for the increase in attrition.

A small scale pilot study which addresses these issues has already been undertaken in Scotland. The study retrospectively tracked the progress of 191 complaints involving sexual offences through the criminal justice system, by examining crime reports, interviewing police officers, examining fiscal files and interviewing precognition officers and procurators fiscal. Two police forces were involved, one urban and one rural, and seven fiscal offices. Of a total of 47 cases which began as complaints of rape, 17 did not progress beyond the police, a further 15 did not progress beyond the fiscal, and of the 15 which went to court, eight resulted in a conviction (Jamieson, 2001).

Although this seems an improvement on the 22% conviction rate reported in an earlier study (Brown, Burman, Jamieson, 1992), Jamieson notes that more than half of the cases which proceeded to court involved child complainers. A further breakdown of the figures shows that of the nine cases involving child complainers, five resulted in a conviction, compared with only two out of the 14 cases involving an adult complainer. Although Jamieson comments on the range of reasons given in police crime reports for not proceeding, including withdrawal of the complaint, false allegation, and no known suspect, she does not identify any one area of police procedure as particularly problematic. In relation to cases marked 'no proceedings' by the procurator fiscal, she suggests that the basis for deciding there is 'insufficient evidence' might bear further exploration. Acknowledging that the fiscals interviewed all maintained that decisions should be made on the basis of sufficiency of evidence, and not on the credibility of the complainer, she nonetheless notes:

"In the case files we examined, we formed the impression that judgements about credibility were most often recorded in cases in which there is equivocation about the sufficiency of the evidence." (Jamieson, 2001: 80).

The HMCPSI/HMIC report also notes concern about the role of the complainer's credibility in cases where there is limited evidence, and in particular "...found that the prosecutor's approach too often tended to be one of only considering any weaknesses, rather than also playing a more proactive role in seeking more information and trying to build or develop the case." (HMCPSI/HMIC, 2002: 9). Amongst a raft of measures outlined in the subsequent Action Plan it is noted that revised guidance on rape has already been made available to prosecutors, and that a revised training package for sexual offences will be commissioned. In addition, it is noted that the CPS agrees with the recommendation that rape cases should be handled by specialist prosecutors, and that consideration is already being given to how to implement this.

There have been some significant changes in the legislative response to rape and sexual assault over the past three years in Scotland. A recent Lord Advocate's reference on the definition of rape clarifies Scots law and makes it clear that rape is based on an absence of consent, and does not require the use or threat of force. The introduction of the Sexual Offences (Procedure and Evidence) (Scotland) Act 2002 prohibits the accused in sexual offence trials from conducting his own defence, and tightens the restrictions regarding use of sexual history/character evidence. The Solicitor General has recently announced a review of the prosecution of rape and sexual offences in Scotland. All of these initiatives might be expected to improve the treatment of rape complainers and hopefully the attrition rate. Regan and Kelly (2003) recommend that governments evaluate recent and new legal and procedural reforms. A precedent has been set in Scotland with the evaluation of the Protection from Abuse (Scotland) Act 2001, and it would seem that this exercise would bear repeating in relation to the recent sexual offences legislation.

A search of the literature produced little from Canada on rape or sexual assault. As noted earlier in this report, Canada has no specific offence of rape, having made the shift to a broader 'sexual assault' spectrum of offences in the early 1980s. In a recent overview of the Canadian experience, Hague et al note that this creates some difficulty in trying to compare reporting and prosecution of rape in Canada with experiences in the UK, as the figures available from Justice Canada provide the totals for all sexual assaults, across all three of the levels of sexual assault defined in law, and with no distinction made between offences against children and offences against adults (Hague, Kelly and Mullender, 2001).

The legislative reform in Canada also included the removal of the requirement for corroborative evidence in sexual assault cases. In practice, however, it would seem that prosecutors are still reluctant to proceed with cases which do not have some form of corroboration (Du Mont and Myhr, 2000, cited in Krug et al, 2003: 170).

Hague et al note with some surprise a decline in reported sexual assaults over the five years before their report, and with some disappointment the absence of the kind of detailed data that is available on 'family violence' from Statistics Canada (Hague, Kelly and Mullender, 2001). It would seem that in Canada, as in Europe, rape and sexual assault are 'forgotten issues', despite the best efforts of women's advocacy services.

4.1.2 Supporting survivors of rape and sexual assault

Rape crisis centres have provided support services for rape and sexual assault survivors in Scotland since the 1970s. In common with similar services in other parts of the world, centres were usually based on a feminist political perspective, with support provided 'by women, for women'. The support aimed to be woman-centred, non-judgmental and non-directive. It was free, and it was confidential. The stated goal was to help the woman regain control over her life, and support was delivered in ways designed to enhance this:

"Most women have their initial contact with us through the telephone. Women assaulted by men have had their sense of control over their own lives and bodies destroyed. In using the telephone a woman has the power over her contact with us - by hanging up when she wants, by ringing back if she wants, by making arrangements to meet face-to-face and keeping or breaking them - she chooses the extent of her involvement." (Edinburgh Rape Crisis Centre, 1981: 6).

The mechanics of service delivery varied from one area to another, often dependent on funding. Where possible, centres provided medical and legal advocacy for women, and accompaniment through the criminal justice process. Until the late 1980s, most centres were run entirely by volunteers.

Believing women and validating their reactions to sexual assault were core to the process. Challenging the social norms of the time, rape crisis centres gave a clear message to individual women and to the rest of society - women were not responsible for rape, men were:

"When a woman has been raped she often encounters disbelief and blame from all corners - family, friends, police, doctors. She suffers from the fear that somehow she contributed to the attack. Our acceptance without judgement of whatever the woman wants to tell us can help begin the process of banishing this guilt. We do not doubt or question what she says." (Edinburgh Rape Crisis Centre,1981:6).

Belief and unconditional acceptance continue to be central to the ethos of rape crisis. In a recent handbook on rape and sexual assault, "believe the woman" is still at the top of the list (Rape Crisis Centre, Glasgow, 2003).

The basic philosophy has changed little over the last 20 years. The services, however, have changed and adapted, partly in order to meet the requirements of funding and regulatory bodies. However, change has also been prompted by the demands of women using the services. Services which were established around a 'crisis line' as the primary source of support have shifted emphasis, as women making contact with centres have developed more of an expectation of receiving 'counselling', reflecting a wider public acceptance of counselling than was the case in the early days of the movement (Christianson and Greenan, 2001). Rape crisis centres today are more likely to offer face-to-face support routinely, using a formal appointment system (Rape Crisis Scotland, 2003). Centres also develop their services in response to the needs of particular women, for example women asylum seekers and women working in prostitution (Rape Crisis Scotland, 2004).

Although some centres carry out in-house evaluations of their service, usually based on feedback questionnaires from service users, there has been little formal or independent evaluation of the rape crisis response to survivors of rape and sexual assault anywhere. One U.S. study which sought the views of rape survivors on the responsiveness of services following an assault found that 75% of women rated their contact with rape crisis centres positively (Campbell et al, 2001). Elsewhere, the same group of authors note that evaluation of rape crisis services in the U.S. has come hand-in-hand with increased dependence on State funding (Riger et al, 2002), and suggest that this link to funding requirements is instrumental in creating tensions between practitioners and programme evaluators. They go on to advocate an evaluation approach which is more holistic, with a focus on best practice, as well as best value; on accountability as well as accounting (Riger et al, 2002).

Consistent with this view, Rape Crisis Network Europe (RCNE) has developed a set of 'best practice guidelines' for NGO's working with women who have experienced sexual violence. Their report, based on information drawn from their membership, outlines a good practice model based on a political understanding of sexual violence as an abuse of power (RCNE, 2003), and the need, therefore, for a politicised response to it. This underpins all aspects of the work to be undertaken. Client-centred, accessible services, working with each woman to identify what she needs and then helping her to find appropriate support for her situation, are seen as core elements. The involvement of women who are survivors of sexual violence in the delivery of the service is also seen as important, demonstrating to women who are still in crisis that they can survive and be strong again. Evaluation of services is seen as crucial, through information gathered from service users, but also through regular networking and contact with other centres to share elements of good practice (RCNE, 2003).

Linked with the delivery of direct services to women is the perceived need to work for change in the societal values and attitudes which allow sexual violence against women to be perpetuated. The report identifies campaigning for legislative and social policy reform, public education work, research and training as essential components of a strategy to improve service responses to women who have experienced sexual violence (RCNE, 2003).

Some of the elements of the 'best practice' approach identified in the RCNE guidelines are shared by many other services involved in supporting survivors of rape and sexual assault. Outwith the voluntary sector, mental health services are the main source of support, therapy and counselling for survivors. Campbell (2001) reviewed the available literature, primarily from the U.S., on the effectiveness of different interventions used by mental health practitioners. She notes that most of the research has been conducted with white women (and by implication middle class women), because in the U.S., this is the group which accesses mental health services. The findings, therefore, may not reflect the experiences of minority ethnic women or working class women. Citing Wyatt (1992) she notes that African American women are perhaps more likely to use informal support systems, i.e. family and friends, but goes on to suggest that more work is needed to provide a fuller picture of why minoritised groups are less likely to access mental health services (Campbell, 2001).

Campbell's review acknowledges that, in general, there is a need for more research which evaluates interventions. The existing body of research focuses primarily on cognitive behavioural therapy (CBT) and feminist therapy. CBT approaches are primarily used to help women deal with the immediate aftermath of an assault, particularly in response to the high levels of fear or anxiety they may experience. Feminist therapy is utilised more as a response to the longer term difficulties of women who have been raped, particularly in relation to guilt and self-blame. Although noting the value of CBT techniques in reducing anxiety and fear, she questions whether there is any evidence that they are as helpful in reducing self-blame and guilt. Feminist therapy, she suggests, may be a more effective intervention in relation to guilt and self-blame, since it encourages the woman to see her experience in the context of societal inequalities, not as her individual problem. In practice, she acknowledges, mental health workers are likely to use a combination of these approaches, since they will be seeking to address both short and longer term difficulties (Campbell, 2001).

In an effort to identify what kind of intervention was helpful to women, Campbell et al surveyed 102 women survivors of rape in Chicago, examining the impact of a range of service responses, or lack of response, on their psychological and physical health outcomes (Campbell et al, 2001). The services considered included the legal system, medical/forensic services, mental health services, rape crisis centres and religious groups 14. Over half of the sample found contact with the legal system 'hurtful' rather than 'helpful', and although 47% of those who sought medical attention found it 'healing', almost a third found it 'hurtful'. In contrast, the majority (over 70%) of survivors experienced interventions by mental health services, rape crisis centres or religious communities as 'healing'.

The study found that ethnic minority women and women who had experienced 'acquaintance rape' had a particularly poor response. Overall, the researchers concluded that "a key focus on violence against women research and interventions must be the prevention of secondary victimisation." (Campbell et al, 2001: 1253). They recommend three approaches to achieve this - an increased involvement of rape crisis services; increased training for all service providers, and the development of coordinated multi-agency responses.

4.1.3 Developing a multiagency response to rape and sexual assault

A recent report on the subject of multi-agency responses to sexual violence suggests that the literature in this area is largely descriptive rather than evaluative in focus (Dutton and Cavanagh, 2003). The report, commissioned by the Glasgow Violence Against Women Partnership, provides a comprehensive review of the literature on multi-agency models in Britain and the United States. The authors examine the development of Sexual Assault Response Teams (S.A.R.T.s) in the United States, noting that the model may operate slightly differently in different parts of the country. Generally there will be a degree of consistency in the membership of the teams, which may involve a Sexual Assault Nurse Examiner (S.A.N.E.), a representative from the police or sheriff's office, a detective, a prosecutor, a rape crisis advocate and staff from the emergency department. Following examination by the S.A.N.E., a nurse who has undergone specialist training in gathering forensic evidence, the complainer is interviewed in the presence of, usually, the S.A.N.E., the investigating police officer, and a rape crisis advocate. From there on, the S.A.R.T. will maintain contact throughout the investigation and prosecution process (Dutton and Cavanagh, 2003).

The original S.A.R.T. model, developed in California, is built on the assumption of an immediate police involvement. Subsequent adaptations have sometimes moved away from this approach, arguing that women may feel under pressure to report to the police. In addition, if they believe that seeking help with injuries or other health concerns may involve them in having to report to the police, some women will avoid accessing healthcare services, and thus miss out on screening for sexually transmitted infections and pregnancy (Ledray, 1999). However, in all S.A.R.T. derivatives, the principle of providing a coordinated response is unchanged, simply the range of services involved (Dutton and Cavanagh, 2003).

Dutton and Cavanagh raise a justifiable concern about the lack of evidence to support the view that S.A.R.T. and S.A.N.E. initiatives are 'best practice' in responding to survivors of rape and sexual assault. Advantages of S.A.N.E. schemes are said to include - better collaboration with law enforcement, higher reporting rates, shorter examination time, better forensic evidence collection, more complete documentation and improved prosecution rates (Ledray, cited in Dutton and Cavanagh, 2003). However, no substantial empirical study has yet been carried out.

Although rape crisis centres in the U.S. are now routinely involved in S.A.R.T.s, there have been significant periods of tension and debate during the development of these multiagency responses. The campaigning and political lobbying work of rape crisis centres has undoubtedly played a massive role in increasing awareness of sexual violence and challenging the myths and stereotypes which abound about women and rape. In the process, rape crisis centres have raised women's expectations that they should be able to access services if they are raped, leading to a steady increase in the numbers of women approaching services. The pragmatic response to this is to seek the most efficient way to deliver services. The involvement of statutory service providers in developing this response has inevitably meant the loss of much of the political focus of the early work, as the emphasis has shifted from 'stopping violence to managing rape' (Matthews, 1994).

Similar tensions have followed the development of Sexual Assault Referral Centres (S.A.R.C.s) in the U.K. There are currently four S.A.R.C.s in England, with plans for more. Based on a medico-legal response to rape, none of the centres involve rape crisis or other women's groups. Counselling is provided by professional counsellors, and the services are 'gender neutral', i.e. providing services to all adults.

In summary, Dutton and Cavanagh identify some common themes in the development of multi-agency responses to women survivors of sexual violence. They note that services are "primarily instigated by statutory service providers; primarily based on medico-legal service provision; gender-neutral in the provision of services" (Dutton and Cavanagh, 2003: 75). Their over-riding conclusion is that there is "a significant lack of research examining the effectiveness of these initiatives" (Dutton and Cavanagh, 2003). In the light of this, a forthcoming evaluation of the U.K. S.A.R.C.s is to be welcomed. 15

4.2 Responding to women working in prostitution

"Some of the questions about prostitution cannot be resolved by research, since they are fundamental questions about the kind of society one wishes to see, how one understands gender equality, and what it means to sell sex."

(Bindel and Kelly, 2004: 1)

Although there is little dispute that women working in prostitution are at significantly higher risk of being physically or sexually assaulted than women in the general population (Farley and Barkan, 1998; Farley and Kelly, 2000), there are different degrees of understanding or acceptance of prostitution as violence against women. For some there is a distinction to be made between 'forced prostitution', including trafficking, and women 'choosing' prostitution as an occupation (Butcher, 2003). Others assert that the harm caused to women by prostitution should define it as a form of violence against women ( Farley and Kelly, 2000; Miller and Jayasundara, 2001). The nature of policy and practice responses to women in prostitution is determined largely by which of these positions is adopted.

4.2.1 The criminal justice system response

Policy and legislative frameworks play a significant role in determining the focus of responses to women working in prostitution. In Scotland, the focus of the criminal justice system has been largely reactive, working from a 'crime management' perspective which defines women in prostitution as offenders, more likely to be arrested than the men who buy sex from them. The imposition of fines adds to the debt many women are working to repay, they may then be jailed for non-payment of fines, and the debt problems are then exacerbated. In the meantime, women who are assaulted while working in prostitution may be reluctant to report assaults to the police because they can then be arrested on outstanding warrants. Thus women's experiences as victims of male violence remain largely invisible (Women's Support Project, 2002).

Looking for ways to change this situation, the Routes Out Partnership recently commissioned a review of legal responses to prostitution in four countries. The review assesses current and past approaches to prostitution in Victoria (Australia), Ireland, the Netherlands and Sweden; identifies the rationale for the changes made in each country's position and assesses the impact of the changes on women involved in prostitution and the men who use them. The countries studied utilise one of three regimes - Victoria and the Netherlands have both moved to legalisation involving state sanctioned brothels, Ireland has adopted a regulatory approach, and Sweden has moved to a position of criminalising the buying of sex and decriminalising the selling of it (Bindel and Kelly, 2004).

In the state of Victoria, Australia, and in the Netherlands, legalisation of prostitution is primarily focussed on licensed brothels; some aspects of street prostitution are still illegal. The rationale for adopting this approach can be seen, in part, as being about removing the 'nuisance' element of street prostitution by providing a state sanctioned indoor environment. The provision of a 'clean' and safe environment is also seen as an advantage, improving the sexual health of prostitutes (and the men who use them) and breaking the links between prostitution and organised crime (Bindel and Kelly, 2004).

The authors comment that this does not appear to be the case in practice. In both Victoria and the Netherlands, there is still evidence of strong links between organised crime and prostitution, along with a significant increase in the number of licenced and unlicenced brothels. Police and local authorities have inadequate resources to enable effective monitoring of which brothels are operating without licences, and in the Netherlands, it is noted that the expansion of 'legal' prostitution has been matched by a similar expansion in 'illegal' prostitution. Women working in prostitution are still stigmatised, and many continue to work illegally because they do not want to be officially recorded as prostitutes (Bindel and Kelly, 2004).

'Regulation' of prostitution involves a mix of approaches including responding to the 'nuisance' aspect of prostitution and/or adopting an unofficial position of tolerance. The goal is 'management' by maintaining public order, rather than 'prevention'. Implementation of a regulatory approach in Ireland since 1993 has not resulted in any discernable weakening of the links between prostitution and organised crime. It has, however, had a negative impact on women, as the powers given to the police to deal with public soliciting have primarily been used against women, rather than customers and pimps (Bindel and Kelly, 2004).

Sweden passed legislation in 1999 which made it a criminal offence to buy any form of sexual services, in line with an overall policy on gender equality which determined that ending prostitution rather than managing it must be the goal. At the same time, the provision of sexual services by women was decriminalised, and their status shifted from 'offender' to 'victim'. Linked with the legislation is a commitment to increasing the resources of police and prosecutors to deal with it, along with the development of drug rehabilitation programmes and exit strategies for women. Street prostitution in Stockholm has subsequently reduced by two thirds. This is seen by the Swedish government as a long term project, with the goal not only of reducing the numbers involved in prostitution but also of changing public attitudes. In the meantime, the decriminalisation of women working in prostitution, and the consequent freedom from the cycle of arrest/fine/jail, has removed at least one barrier to women's ability to leave (Bindel and Kelly, 2004).

The researchers, while noting that "Virtually no evaluation of overall approaches has been undertaken...", go on to suggest that the evidence that does exist "does not commend a legalisation approach." (Bindel and Kelly, 2004: 31). They conclude that the Swedish model is the most coherent "in terms of philosophy and underpinning" and note with interest that it is the only model of those reviewed which does not criminalise those who sell sex (Bindel and Kelly, 2004: 32).

In Scotland, the debate has polarized further, and more publicly, since 2002, when a private member's bill introduced by Margo McDonald MSP proposed that local authorities should be allowed to establish prostitution 'tolerance zones' in designated areas. Although the Bill fell at committee stage, it generated considerable debate. The Scottish Executive established an Expert Group on Prostitution to review the available evidence on effective responses to prostitution. The group met for the first time in August 2003, and has already heard evidence from different parts of Scotland. The group has also commissioned a piece of consultative research to gather the views of women working in prostitution, both indoors and outdoors.

Meanwhile, the Bill has been resubmitted and is currently being reconsidered by the Local Government and Transport Committee. Written evidence submitted to the committee reflects the polarisation of the current debate in Scotland about what constitutes an appropriate response to prostitution. Supporters of the Bill argue that the use of tolerance zones increases the safety of women working in street prostitution by enabling them to work within sight of each other. This view, they argue, is supported by a noted increase in the number of attacks on women working in street prostitution in Leith (Edinburgh) since the 'unofficial' tolerance zone was dropped. Prostitutes themselves, it is argued, support the introduction of tolerance zones, and as a matter of human and civil rights, their views should be given consideration (ScotPep, 2003).

Those opposed to the Bill argue that this view makes the assumption that women want to work in prostitution. If this is taken as a given, then obviously women will want any measure which promises a degree of safety. They also dispute the evidence on increases in attacks on women when tolerance zones are removed, citing a report by Base 75, an agency in Glasgow which provides support services for women working in prostitution, which shows no difference in the numbers of attacks reported to them by women before and after the introduction of 'sensitive policing' in Glasgow city centre (Routes Out Partnership, December 2003).

The Bill is still in progress. The Expert Group is expected to report on Stage 1 of its remit (on street prostitution) in Autumn 2004.

4.2.2 Supporting women abused in prostitution

Good practice in responding to women working in prostitution is identified in a guide originally published by the Franki Women's Support Project in Bolton, and subsequently adapted and published by the Women's Support Project in Glasgow. The guide highlights the need for realistic interventions, based on an understanding of the harm caused to women by prostitution, but recognising that for women, making the decision to leave prostitution may not be a simple or straightforward choice (Women's Support Project, 2002). A range of services and interventions may be required at different points in this process.

It has already been noted that prostitution impacts negatively on all aspects of women's health (Farley and Barkan, 1998). The medical literature, however, focuses predominantly on the prevalence and prevention of sexually transmitted infections in women working in prostitution, and pays scant regard to their wider health needs (Baker, Case and Policicchio, 2003). Health interventions, therefore, have tended to focus more on the threat to public health posed by women working in prostitution, rather than on the health needs of the women themselves. Although practice in this area is clearly beginning to change, with an increasing focus on the broader health concerns of women working in prostitution (Routes Out, 2003), there is still little to be found in the research literature which goes beyond "condoms and safer sex negotiation skills" (Farley and Kelly, 2000: 22).

There is a similar lack of research on what works in advocacy and support services for women working in prostitution. The Home Office Crime Reduction Programme funded 11 pilot projects addressing prostitution until March 2002. Three of the pilot projects focussed on young people, three on policing, and six on 'exiting and support'. Of the latter, two of the pilots identified a reduction in the number of people entering or an increase in the number of people exiting prostitution as a primary goal. The remainder primarily focussed on reductions in kerb crawling, soliciting and 'associated nuisance', or a range of other 'harm reduction' indicators (Home Office, 2001). An evaluation of the pilot studies is pending.

4.2.3 Developing a multi-agency response to prostitution

The Routes Out Partnership was established as a thematic social inclusion partnership in Glasgow in 1999, with the aim of improving and coordinating responses to women working in prostitution in the city. Membership of the partnership board reflects the diversity of agencies involved in responding to prostitution and women working therein - the police, the health board, the local drug action team, social work and other council services, and the women's voluntary sector.

Routes Out is based unequivocally on the principle that prostitution in any form is harmful to women, and works towards preventing women, particularly young women, from becoming involved in prostitution, and providing viable alternatives to women who wish to leave prostitution. It is, in effect, a demonstration project, testing an approach which involves crisis intervention and short term initiatives to support women who are still involved in prostitution, and longer-term interventions to support women who are leaving prostitution (Routes Out, 2003).

The Partnership is supported by a range of agencies, including Base 75, which provides a range of services at evening drop-in clinics, including methadone prescribing, and access to accommodation. Rape Crisis in Glasgow has been funded to develop a specific project 'Supporting Women Abused through Prostitution'. The Routes Out Intervention Team is funded through the Partnership to assist women to achieve some stability, before beginning to identify what they may need to do to be able to leave prostitution (Routes Out, 2003).

4.3 Responding to survivors of child sexual abuse

4.3.1 The criminal justice system response

There has been an increasing focus in the last two years on the prosecution of cases of child sexual abuse involving child witnesses 16. Changes in the procedures relating to child witnesses in such cases are to be welcomed, and will hopefully improve both the process and the outcome of sexual offences trials involving child witnesses. However, despite a noted rise in the numbers of adults reporting historical cases of child sexual abuse to the police (Scottish Executive, 2003), no research has been identified which examines the investigation and prosecution of historical abuse complaints.

What is known from practitioners, and from survivors, is that the increase in reporting is not matched by an increase in prosecution of historical cases. The chief reason for this is likely to be insufficient evidence. Lack of forensic evidence and the absence of witnesses are common barriers to prosecution across all forms of violence against women. Additional factors in historical abuse cases include the length of time which may have elapsed, and the implications of this for the reliability of the memories of witnesses. The age of the alleged abuser may also be significant, as the procurator fiscal must consider whether it is in 'the public interest' to prosecute if the accused is very elderly, and/or or very frail. Prosecution is most likely to proceed, and to result in conviction, in cases where there are two or more complainers reporting similar incidents involving the same accused. Even then, defence advocates may succeed in demonstrating inconsistencies in their narratives, making it difficult for a jury to convict 'beyond reasonable doubt'.

Broader measures designed to address the needs of all victims of crime obviously have relevance for adult survivors of child sexual abuse. For example, should a complaint survive the investigative process and be referred for trial, it is likely that the complainer will be defined as a 'vulnerable witness', given the nature of the offence and the levels of fear which they may still have in relation to the abuser. As such, there may be an entitlement to the range of special measures which have recently been introduced to ease the experience of giving evidence in such cases 17. This, however, will be of limited value until some of the prosecution difficulties are resolved.

4.3.2 Supporting adult survivors of child sexual abuse

In Scotland, as elsewhere, the earliest responses to adult survivors of child sexual abuse came from within the voluntary sector, and in particular from rape crisis services. From the early 1980s, increasing numbers of women making contact with rape crisis centres were disclosing childhood abuse by family members. Incest survivor groups and campaigns developed out of these early contacts, and by the 1990s, several new voluntary organisations had been established with a specific remit to provide services for survivors of childhood sexual abuse (Christianson and Greenan, 2001). In addition to providing support and counselling services for individual survivors, these organisations have also advocated for the development of a more consistent response from the statutory sector.

Most voluntary organisations provide a service which could broadly be described as client centred and non-directive. Support may be provided face to face and/or by telephone. In some organisations, all support is provided by paid staff; in others, volunteers may provide the support, with supervision from a paid staff member. Support may be open ended, or time limited. Some organisations also run support groups for adult survivors of child sexual abuse.

Evaluation of the service is generally carried out in-house, as part of the routine evaluation and monitoring of the organisation's work. Occasionally, organisations are funded to carry out a more formal evaluation of particular aspects of their work. More often, though, evaluation is seen as a necessary part of the evidence gathering required by funders, both private and public. At times, it may be limited to an assessment of satisfaction with the services provided, rather than a measure of outcomes or change for individual women (Riger et al, 2002).

4.3.3 The health service response

As might be expected from the documented impact of child sexual abuse on mental health, adult women survivors of child sexual abuse are significant users of mental health services, both as inpatients and outpatients (Mennen, 1990, cited in Thomson, 1998). Survivors also access general primary care services, both as a result of chronic physical health difficulties and emotional distress.

Despite this, relatively little attention has been paid in recent mental health policy developments to the specific needs of survivors of child sexual abuse (Short Life Working Group on the Care Needs of Survivors of Child Sexual Abuse (SLWG), 2004). The Framework for Mental Health Services in Scotland (1997) acknowledged that physical and sexual abuse are detrimental to mental health, and that responses to survivors of such abuse should therefore be within the scope of extended primary care services. Since then, more explicit attempts have been made to address how mental health services in Scotland might best respond to adult survivors of child sexual abuse.

Two Scottish studies have explored the experiences of women survivors of child sexual abuse who have accessed mental health services and other support services. Thomson (1998) reported on the experiences of women who had disclosed abuse in a range of settings, voluntary and statutory. Nelson and Phillips (2001) focused on the experiences of women survivors of child sexual abuse who had been inpatients at one Edinburgh hospital.

Thomson interviewed 19 women about their experiences of service responses to their disclosures of child sexual abuse. Common themes included the importance of confidentiality, establishing trust, having the time and space to talk, and the need for professionals to have more awareness of the impact of sexual abuse on women's lives. Common difficulties encountered by women included long waiting lists, and consequently long delays between their initial disclosure of sexual abuse and being able to access a service. Negative and/or disbelieving responses from professionals were also a concern for women (Thomson, 1998).

Nelson and Phillips echo Thomson's findings to some extent, with similar negative experiences of responses to disclosure reported by many of the 22 women interviewed. However, where the Thomson study focuses primarily on the process of disclosure and the reactions women experienced, the more recent report provides a more detailed analysis of what women found helpful in responses from workers and services, and how far these responses assisted their recovery (Nelson and Phillips, 2001). In particular, this report highlights that for many survivors of sexual abuse, the levels of awareness and attitudes of staff, as characterized by their willingness to ask about child sexual abuse, are more important than any particular theoretical perspective (Nelson and Phillips, 2001).

A recent survey of clinical psychologists working in the Greater Glasgow area sought information about their awareness of the prevalence and impact of child sexual abuse and their practice in the detection and management of sexual abuse issues. The survey also addressed the level of training psychologists had undertaken and the degree of confidence they felt in working with clients who had been sexually abused (Biggam and Johnson, 2003).

Awareness of the prevalence and impact of child sexual abuse was high, although the proportion of survivors of child sexual abuse in the collective caseload was found to be somewhat lower than the national average. In response to questions about the range of therapeutic techniques being utilized, 57% of the respondents (n=74) indicated that they found cognitive behavioural therapy (CBT) helpful when working with this client group. Significantly, 95% of the respondents indicated that they were trained to use this model, compared with only 23% who had been trained in the use of other approaches (Biggam and Johnson, 2003).

The authors also reviewed the literature on psychological interventions with survivors of sexual abuse. Acknowledging the scarcity of studies on either individual or group therapy approaches, they provide a useful overview of some of the main methodological difficulties inherent in designing research in this field. Most of the studies reviewed used pre- and post-intervention testing to identify changes in rates of depression, trauma symptoms, self-esteem or other mental health indicators. Most found significant improvements across these indicators. However, citing Donaldson & Cordes-Green (1994), the authors comment that "inherent weak research designs mean that such changes cannot be definably attributed to the intervention" (Biggam and Johnson, 2003).

Only one of the studies they reviewed on group therapy used a comparison group (Alexander et al, 1989, cited in Biggam and Johnson, 2003) and this absence of a control group is one factor that weakens the other studies. However, they also acknowledge the inherent ethical difficulty of providing a therapeutic intervention to one group of clients and withholding it from another. In addition, they comment that the range of factors which may contribute to the efficacy of group psychotherapy interventions is extensive. This would include personal characteristics of the clients and/or the therapist, the level of competence and experience of the therapist, the clients' commitment to the process, and whether clients are also engaged in individual therapy.

A lack of reliable research on the effectiveness of individual psychotherapy is also identified, with much of the concern focused on similar weaknesses in research design. However, Biggam and Johnson note that some of the results are of interest, with an indication that "gains were maintained across global and specific measures" in three out of the seven studies of individual psychotherapy interventions they reviewed (Biggam and Johnson, 2003: p 39).

Recent guidance on treatment choice in psychological therapies suggests that cognitive behavioural therapies have the most demonstrable benefit in the treatment of depression, anxiety and post traumatic stress disorder, three of the more common mental health sequelae of child sexual abuse (Department of Health, 2001). However, the guidance is framed purely around a symptomatic approach, with no focus on the causes of mental health difficulty in adults. There is therefore no specific analysis of the efficacy of a CBT approach for survivors of child sexual abuse, beyond short-term symptom alleviation.

4.4 Responding to domestic abuse

4.4.1 Criminal and civil justice responses

Many aspects of the criminal justice system response to domestic violence have come under scrutiny in Scotland and elsewhere over the last two decades. Throughout the 1970s and 1980s, domestic violence advocacy organisations consistently identified the justice system's inadequate response as the most significant barrier to women achieving safety. Frustration at the perceived indifference of police, prosecutors and judges was gradually matched by the realisation that "the failure of the courts and police to protect women was not simply a matter of an attitude on the part of individual practitioners. It was a lack of legal tools to intervene in a legal system that did not take into account…the complexities of women's experiences in a society in which citizens' access to resources and social privilege is determined by their sexuality, race, gender and class position.". This 'lack of legal tools' contributed to low arrest rates, ineffective prosecution and infrequent conviction (Pence and Shepard, 1999:9).

Attrition in domestic violence cases is still remarkably high. A recent study in England found that of 869 domestic violence incidents reported to the police, only 291 were deemed to have a power of arrest attached. Of 222 actually arrested, 60 individuals were prosecuted for criminal offences. Ultimately, only 31 individuals were convicted, and only four of those received custodial sentences. Three police areas in Northumbria were studied, each with slight variations in the policing approach adopted. The authors note with concern that cases from the area which used a 'positive policing' approach, i.e. most likely to arrest, were most likely to fail at court. The authors suggest the need for a more consistent approach across police areas, but also for this to be reflected in the courts (Hester, Hanmer et al, 2003).

In Scotland, several aspects of the justice system response to domestic abuse have been reviewed since the publication of the National Strategy to Tackle Domestic Abuse in Scotland. The short-life working group established by the Scottish Executive to review legislation related to domestic abuse acknowledged the complexity of criminal and civil law in this area, and recommended bringing all of the legal remedies together in a single Act. The working group also recommended that a feasibility study be carried out into the practicalities of establishing a domestic abuse court to provide a more cohesive judicial response. A pilot domestic abuse court was subsequently proposed and is due to be implemented in Glasgow.

Many women experience ongoing harassment and abuse after leaving an abusive partner, and may resort to civil remedies in an effort to protect themselves. The Protection from Abuse (Scotland) Act 2001 (PFA Act) strengthened the existing provision around the use of interdicts by allowing powers of arrest to be attached to an interdict, regardless of the relationship between the two parties; this provision was previously only available as part of a matrimonial interdict. An early evaluation of the Act suggests that interdicts granted under the PFA Act are gradually replacing common law interdicts. This in turn implies that more people seeking protection by interdict are able to access powers of arrest. However, the authors note with some concern that awareness of the provisions of the PFA Act is variable. Few of the women who responded to postal questionnaires had heard of it, and some professionals were unclear about the detail of the provisions. Training for professionals is recommended, along with awareness raising for the general public. The evaluation commenced immediately after the Act was implemented, and the authors suggest that it should be seen primarily as a 'scoping exercise', with a recommendation that a more comprehensive evaluation should take place (Cavanagh, Connelly and Scoular, 2003).

For women with children, there may be further difficulties if their partner has a right to contact with the children. Research with survivors of domestic abuse in England found that contact with children by a violent ex-partner led to further abuse of the woman and/or the children in 92% of cases. In cases where the partner applied to the court for contact it was rare for this to be denied, and even in cases where there were allegations of physical or sexual abuse of the children, contact orders which allowed direct visiting contact were granted in 75% of the cases. The authors acknowledge that a broader piece of research would be valuable, involving the children, ex-partners and professionals involved in the administration and monitoring of contact orders (Radford, Sayer and AMICA, 1999). There is a lack of any empirical research in Scotland on the use of contact orders in cases where there has been domestic abuse, but the Scottish Executive has recently commissioned research which should address this gap.

The remainder of this section will explore some of the key themes identified in the research literature in relation to the policing and prosecution of domestic violence, and the development of coordinated justice system responses to domestic violence.

Policing domestic violence

In Scotland, there have been significant changes in the police response to women reporting domestic abuse since the publication, in 1990, of a Scottish Office circular to Chief Constables which provided guidance for Scottish forces on how to respond to domestic abuse. The guidance indicated clearly that allegations of assault by a partner should be investigated as thoroughly as any other assault, and that where sufficient evidence existed, an arrest should be made. The guidance also outlined steps which should be taken to ensure the safety and welfare needs of women and children were addressed as a priority (HMICS, 1997).

The thematic inspection by Her Majesty's Inspectorate of Constabulary in Scotland which resulted in the publication of Hitting Home in 1997 18 reviewed the progress made since 1990 and made several recommendations for further improvement in the Scottish police forces' response to domestic violence. Some of the recommendations would now be seen as standard in any service review, on almost any issue - for example, awareness training, the development of written guidelines, and participation in local multiagency partnerships. Others referred more specifically to organisational issues within forces. The use of nominated officers was clearly identified as beneficial, to improve the monitoring of individual incidents and the provision of follow-up contact to victims, and also in improving liaison with other agencies. It was noted that, even in forces which had designated domestic violence liaison officers, there was a lack of clarity about who had responsibility for deciding on further investigative action, and it was recommended that this should be addressed. The use of statistics on repeat victimisation as an aid to monitoring the effectiveness of interventions was commended as an example of good practice by one force, and the report recommended that this should be adopted more widely. Concern was expressed about the lack of information sharing in most forces between officers dealing with domestic abuse and those dealing with child abuse cases. This echoed a concern about the lack of formal procedures in some forces for sharing information about possible child protection concerns with relevant agencies, in particular the social work department and the Reporter to the Children's Panel (HMICS, 1997).

On the whole, whilst noting the areas which merited further attention, the HMICS report commended the Scottish police forces for the improvements which had already been made in response to the 1990 circular on domestic violence. A subsequent study of police forces in England and Wales reviewed the impact of organisational structure on police responses to domestic violence, and focused specifically on how Domestic Violence Officers (DVOs) were being deployed. The report concurs broadly with the HMICS report on the main areas requiring improvement, but is couched in more critical terms, and notes in particular that:

"No single structure emerged as either more or less problematic than any of the others. The problems related less to the structure than to the status of domestic violence work within forces, the level of commitment of headquarters and divisional commanders, the clarity with which responsibilities were defined and the effectiveness of management arrangements." (Plotnikoff and Woolfson, 1998:41).

The authors suggested a range of measures which might be adopted by police forces to overcome this, including the establishment of performance indicators for domestic violence, with associated reporting requirements from divisions to force headquarters; the introduction of a standard format for reports; clarification and documentation of the responsibilities of key staff in relation to domestic violence, and clarification of the role of DVOs.

Both of these studies were designed to identify gaps in existing police response to domestic violence. Based largely on analyses of force policies and procedures, examination of organisational structures and interviews with serving officers, they provide a broad perspective on the police response to domestic violence at the time. Both studies take the position that responses to domestic violence should be consistent with responses to other offences, in terms of the standard of investigation, recording and monitoring approaches, and identification of the "lines of accountability" (Plotnikoff and Woolfson, 1998: 41). Although acknowledging that assaults by partners are 'different', the route to improving the policing of domestic violence in both of these studies is predicated on ensuring that structural and procedural gaps are filled.

A different approach was adopted in Leeds, where a 1997 study set out to evaluate a 'tiered' response to domestic violence. This 12 month pilot project took place in the Killingbeck division of West Yorkshire police, and aimed to reduce repeat victimisation by providing a graded response to domestic violence, with the police intervention intensifying in response to repeat callouts to the same woman, or the same perpetrator. The level of response was determined by the number of times the offender had attended in the previous 12 months, either in relation to his current partner, or another woman. A first attendance warranted a Level 1 response; one previous incident triggered Level 2, and two or more previous incidents triggered a Level 3 response. Although no formal risk assessment approach was adopted as part of the project, some men were assigned immediately to level 2 or 3, despite no prior attendance, if the level of violence used warranted a more intensive response (Hanmer, Griffiths and Jerwood, 1999).

The range of interventions with the women included issuing information letters; 'police watch' i.e. assigning additional patrols in the neighbourhood following an assault; 'cocoon watch' which involved neighbours, family and friends in actively contacting the police to report incidents; target hardening, i.e. improving the physical security of the woman's home, and issuing panic buttons and mobile phones. Simultaneously, a series of official warnings and information letters would be sent to the offender, and a graded response activated via the Crown Prosecution Service (CPS) and the magistrates' court (Hanmer, Griffith and Jerwood, 1999).

Responsibility for allocating the interventions rested with a DVO, a sergeant who was appointed specifically to progress this project. The volume of work increased during the pilot study and a second DVO was appointed to cover the last three months. It is worth noting that that this increase in the volume of work for the project did not represent an increase in the number of incidents reported, but rather an increase in the number of incidents which were accurately recorded as domestic violence incidents. Recording was acknowledged to be poor at the beginning of the project, with only 50% of domestic violence incidents being accurately coded; this increased to 80% over the lifespan of the project (Hanmer, Griffiths and Jerwood, 1999).

The twin focus on both women and men created "an interactive crime prevention approach" (Hanmer, Griffiths and Jerwood, 1999: 5) which sought to protect women by "demotivating the offender" (Ibid: 6). The removal of any responsibility from the woman for 'pursuing charges' gave a clear message to both parties that domestic violence was a crime and would be dealt with as such. Women interviewed after receiving an intervention commented favourably on the value of the pro-arrest policy, and gave feedback about the effect it had on their partners. Women who received Level 2 or 3 interventions also noted that a strong verbal warning, including the threat of arrest, could have a significant effect, but that this was lost if not acted on upon at a subsequent attendance. Level 2 and 3 women felt that the 'warning letters' sent out by the project were less effective, easily dismissed by the men, although level 1 women felt they provided an authoritative condemnation of their partners' actions.

The police watch component attracted some of the most positive comments from women, who felt that the patrols enhanced their safety. Some women reported favourably on the presence of police patrols in their area, even where the records indicated that this aspect of the response had not been activated in their case. Very few women remembered the 'cocoon watch' component, although those who did generally regarded it as beneficial.

The sought for outcome in this project was a reduction in repeat victimisation, and this was achieved. The project was able to demonstrate that early intervention reduced repeat offences, with only 25% of the Level 1 entrants to the project (i.e. those for whom this was a first offence) requiring a second attendance during the lifespan of the project, and only 9% requiring a third attendance. By comparison, of those who started at Level 2 19, 46% had a repeat attendance; and of those who started at Level 3 20, 64% had a repeat attendance. Predictive factors for repeat offences included a history of domestic violence offences predating the project; arrest, with men who were arrested being 51% more likely to re-offend 21, and the beat area the victim lived in, with women living in high crime areas more likely to be repeat victims. The issue of whether the couple was living together at the time of the offence was not identified as having statistical significance in determining the likelihood of repeat attendances. However, it was noted that offences were likely to be more serious (as evidenced by a much higher rate of criminal charges being made) when the woman was separated from the offender (Hanmer, Griffiths and Jerwood, 1999).

The Killingbeck project represented an important step forward in the development of a consistent evidence based police intervention in response to domestic abuse. Factors which contributed to its success included the focus on both victim and offender; the involvement of all police officers, rather than just specialist DVOs; improved communication and co-ordination between the police and related agencies, and 'low additional resource implications' (Hanmer, Griffiths and Jerwood, 1999: 40). It established reduction in repeat victimisation as a performance indicator for tackling domestic abuse which was both achievable and measurable, and demonstrated that proactive policing could enhance women's safety. Crucially, however, the project also demonstrated that arrest played a less significant role in reducing repeat victimisation than early and repeated police interventions (Hanmer and Griffiths, 2001).

In the U.K., pro-arrest policies have been adopted increasingly by police forces since the Home Office and Scottish Office circulars of 1990. Such policies were in adopted in North America from the early 1980s onwards, and were seen as integral to a response which placed responsibility for challenging abusive men with communities rather than with individual women. The earliest proponents of pro-arrest policies were agencies advocating for women abused by their partners, acting on the rationale that domestic violence should be treated no differently from any other crime - where there was a sufficiency of evidence, arrest should follow.

The police department in London, Ontario became the first in Canada to adopt a pro-arrest policy in 1981, implementing a recommendation by the London Co-ordinating Committee on Family Violence (F/P/T Ministers Responsible for Justice, 2003). Subsequently, a series of six studies in the U.S., known as the Spouse Assault Replication Program (S.A.R.P.), produced some conflicting results. The first study, which took place in Minneapolis, was widely cited as providing evidence that arrest was an effective deterrent which reduced repeat offending. However, analysis of the data from some of the subsequent S.A.R.P. studies suggested that there was no, or minimal deterrent effect, with some even suggesting that there might be an increase in violence following arrest (Miller, 2003).

Research in Ontario following the adoption of the pro-arrest policy found that charging men who had assaulted their partners did reduce the violence (London Family Court Clinic Inc, 1991, cited in F/P/T Ministers Responsible for Justice, 2003). More recent analysis of the S.A.R.P studies suggests that the impact of arrest may be different for different perpetrators. It is more likely, for example, to act as a deterrent to future offending if the perpetrator is employed or has some status in the community than if the perpetrator has nothing to lose. Individual offender characteristics, such as age or previous criminal record, may also be more significant than arrest as predictors of repeat offending (Maxwell et al., 2001, cited in Miller, 2003).

Pro-arrest policies are not without critics. Some women's advocacy organisations are concerned about the removal of all control from the woman, although others have argued that pro-arrest policies bring police response to domestic violence into line with responses to other crimes. Women themselves may have ambivalent feelings about the efficacy or the desirability of arrest. Some women fear retaliation, or that the arrest will have no effect on their abuser; a belief which may be strengthened if he is arrested then released with no further action. Arrest may be seen as unhelpful by women who want the violence to stop, but not the relationship. For others, it is the case that the violence does indeed increase following an arrest (Hoyle and Sanders, 2000). Black and minority ethnic women, and others from "over-criminalised communities", may be reluctant to involve the police if they fear racism against themselves or their partner (F/P/T Ministers Responsible for Justice, 2003: 18).

Other women welcome the removal of any suggestion of responsibility for making the decision on whether their partner should be charged. In a 1996 study in the Yukon, 85% of women whose partners had assaulted them believed the pro-charge policy was positive, and 68% indicated increased confidence in reporting future incidents (Department of Justice Canada, 1996, cited in F/P/T Ministers Responsible for Justice, 2003). A recent review of Canadian policies and legislation on domestic abuse has reaffirmed a commitment to the pro-charging policies currently in place in Canadian provinces, in all cases where there are reasonable grounds to believe that an offence has been committed (F/P/T Ministers Responsible for Justice, 2003).

Studies in the 1990s examined the impact of coordinated approaches to domestic violence, and found that arrest was more likely to reduce repeat offending in those areas which had adopted more integrated criminal justice approaches (Steinman, 1990, cited in Shepard et al, 2002). This might include proactive prosecution, consistent advocacy and support for women through the whole criminal justice process, and access to mandatory treatment programmes for offenders (Tolman and Weisz ,1995, cited in Shepard et al, 2002).

Prosecuting domestic violence

In Scotland, the decision to prosecute is made by the procurator fiscal, and is based on two key factors - whether there is a sufficiency of evidence, and whether it is in the public interest to proceed to trial. Two independent sources of corroborative evidence are required - in cases of assault, the statement of the complainer is one source of evidence. Other sources might include forensic evidence, medical reports, photographs of the injuries sustained or the scene of crime, or the testimony of other witnesses who saw or heard the assault. Taken together, the evidence must demonstrate beyond reasonable doubt that an offence was committed, and that the accused person was the perpetrator of that offence.

In cases of assault by a partner, as in other cases of violence against women, the presumption has been that the woman herself is the main witness in the Crown case. The nature of the crime means that eye witnesses are rare, or reluctant to come forward. Children and young people are often the main witnesses to assaults on their mothers, but it may not be considered appropriate to call them as witnesses in court. Corroboration may therefore be difficult to find. Even where women have welcomed the arrest and charging of their abuser, they may be unwilling or unable to co-operate with the prosecution. The time delay between arrest and trial can mean that they have reconciled with their partner in the hope that things will now change; or they may be pressured to 'withdraw the charges' by the accused or his family. Whatever the reason, the procurator fiscal must then decide whether to proceed to trial with a 'hostile' chief witness (Barry, 2000).

In other jurisdictions, prosecution may proceed without the woman's co-operation. In Canada, a 'pro-prosecution' policy has applied in all provinces since 1986, i.e. cases will be prosecuted where there is sufficient evidence, regardless of the wishes of the individual complainer. In practice, this means that the Crown must consider whether to lead the case without the woman's testimony, taking into account the strength of any other available evidence. It is not generally considered appropriate to compel her to testify, or to charge her with contempt if she doesn't (F/P/T Ministers Responsible for Justice, 2003).

By reducing the number of withdrawals of charges, the pro-prosecution policy aimed to reduce the attrition rate. Early studies suggested that it was successful. 38.4% of charges were dismissed or withdrawn prior to the introduction, in 1981, of the pro-prosecution policy in London, Ontario. By 1990, this rate had decreased to 10.9% (F/P/T Ministers Responsible for Justice, 2003).

The policy also aimed to improve the co-operation of women survivors with the criminal justice process. The non-co-operation of survivors of domestic violence continues to be a source of frustration for prosecutors (Brown, 2000, cited in F/P/T Ministers Responsible for Justice, 2003) but there are ongoing developments to improve this position. A recent study identified the availability of witness support and the use of video taped evidence as the two most influential elements in determining the co-operation of women (Dawson and Dinovitzer, 2001, cited in F/P/T Ministers Responsible for Justice, 2003).

In the U.S.A. the use of 'no-drop' policies has a similar history, with different jurisdictions adopting variations on this theme since the 1980s. The most rigidly applied policies include the option of arresting and jailing women who do not co-operate with the prosecution of their abusive partner. More flexible approaches are similar to the Canadian model, intent on prosecution where there is sufficient evidence, but acknowledging the difficulties some complainers may have in testifying (Ford, 2003).

As with mandatory arrest policies, pro-prosecution or 'no-drop' policies can be contentious. Although predicated on the assumption that prosecution protects women, Ford (2003) suggests that there is a lack of empirical evidence to support this. Where there is evidence, as in a 2001 study which found that 'no-drop' policies did increase convictions, he notes that further analysis identified a significant degree of 'screening out' of cases which were unlikely to gain convictions (Smith et al, 2001, cited in Ford, 2003). Overall, he argues, 'coerced victim participation' in domestic violence prosecutions has no demonstrable impact on the safety of women in the wider community, and may actively jeopardise the safety of the individual woman it seeks to protect (Ford, 2003).

Work with domestic violence offenders

It is not within the scope of this report to review the substantial body of literature on offenders. However, in the context of domestic abuse, it is important to acknowledge that 'making men visible' 22 is central to the effectiveness of the multi-agency initiatives which are recognized as examples of best practice. How far domestic violence offender programmes actually improve the safety of women and children continues to be the subject of debate. Research suggests that a significant number of men attending probation programmes refrain from violence in the short term but that this impact diminishes over time (Burton, Regan and Kelly, 1998; Dobash et al, 1996). Attrition is high (Burton, Regan and Kelly, 1998) and sanctions for breaching probation orders are not always implemented, although where they are, the completion rate improves and recidivism appears to reduce ((Mullender and Burton, 2000).

The content and ethos of domestic abuse offender programmes may vary considerably. Some programmes take a therapeutic focus, others an educational focus. Most involve a combination of inidividual work and groupwork. In the U.K., most programmes are psycho-educational, and make use of cognitive behavioural techniques combined with an analysis of the gendered nature of domestic abuse (Mullender and Burton, 2000). Minimum standards of practice have been adopted by RESPECT, the National Association for Domestic Violence Perpetrator Programmes and Associated Support Services. Key principles for RESPECT would include work with the partners of abusive men as an essential part of any programme, and the need for programmes to be linked to the overall community response to domestic abuse (RESPECT, 2000).

Men's programmes in the U.K. are almost exclusively court-mandated probation programmes. An exception is 'Working with Men', a project based in north Edinburgh and initially funded by DAPHNE 23 to carry out a feasibility study into the development of a voluntary intervention for men who have assaulted their partners, or are at risk of doing so. A model of good practice was developed as part of the initial study. This involved practitioner training programmes on how to recognise and respond to abusive men, the adaptation of an existing men's programme, and the design of a suggested referral pathway for a men's service. An overarching theme was the integration of any voluntary men's programme into a multi agency response to domestic abuse. The project is now funded by DASDF 24 to develop the model further (City of Edinburgh Council, 2002).

Coordinated criminal justice responses to domestic abuse

Probably the best-known example of a coordinated criminal justice response to domestic abuse was developed in Duluth, Minnesota from 1980. The Duluth model, as it has become known, is sometimes assumed to apply only to the development of education groups for abusive men, or even to the 'Power and Control wheel' diagram which is used to explain the dynamics of domestic abuse. In fact, it is a pragmatic and methodical approach to developing an interagency response to domestic abuse, involving both individual advocacy and institutional advocacy. The rationale for engaging in both of these approaches is encapsulated in this definition given by an advocacy worker from Duluth:

"When I advocate for an individual woman, I am trying to help her overcome the many obstacles on her path to effectively using the courts and police to protect her. When I do systems advocacy, I am trying to build a new path. I come to understand what I need to do in systems advocacy by my work with individual women." (Pence, 2001, in Renzetti et al, 2001: 329).

Based on eight key components (see table, below), the Duluth model has been successfully replicated, with some adaptation, in several countries, including New Zealand, Australia, and the U.K. (Balzer, 1999; Holder, 1999).

Eight Key Components of Community Intervention Projects

  1. Creating a coherent philosophical approach centralizing victim safety
  2. Developing "best practice" policies and protocols for intervention agencies that are part of an integrated response
  3. Enhancing networks among service providers
  4. Building monitoring and tracking into the system
  5. Ensuring a supportive community infrastructure for battered women
  6. Providing sanctions and rehabilitation opportunities for abusers
  7. Undoing the harm violence to women does to children
  8. Evaluating the coordinated community response from the standpoint of victim safety

(Shepard and Pence, 1999: 16)

The effectiveness of the Duluth model, in its original form, can be attributed largely to the coordinating and facilitating role played by the Domestic Abuse Intervention Project. This independent non-profit agency which was set up in 1980 with the specific goal of providing that coordinating role (Pence, 1999). The size of the city was also significant - with a population of just 85,500, this was a relatively small area, with a correspondingly small network of criminal justice workers and agencies. Finally, some of the key players in that network were supportive of the initiative - in particular, the police chief and the city attorney (Shepard and Pence, 1999). In the United States, these positions function with considerably more autonomy than would be the case in this country, and that too, undoubtedly facilitated the adoption of new polices and procedures to support the development of a coordinated response.

Elsewhere, use of the Duluth model has involved a degree of adaptation to account for cultural as well as structural differences. In New Zealand, the model informed the development of the Hamilton Abuse Intervention Project, with some changes to address the needs of Maori women and men - in particular a commitment to service provision by Maori staff, using materials which reflect the culture and history of Maori people (Balzer, 1999).

In the Australia Capital Territory (ACT), the Interagency Family Violence Intervention Programme, based on the Duluth model, was initially established as a 12 month pilot, and has subsequently achieved impressive results, including substantial increases in reporting, arrests and guilty pleas 25. More importantly perhaps, 75% of 'victims of family violence' who were contacted 12 months after proceedings were finalized reported that they felt safe/fairly safe. Only one person had been further physically assaulted in that period (Humphreys and Holder, 2002).

The Australian project has involved significant levels of institutional change, and its success is attributed largely to effective change management processes. The Duluth model is acknowledged as a significant influence, although a note of caution is sounded about the need to avoid a 'one size fits all' assumption (Holder, 1999).

In the U.K., the Domestic Violence Intervention Project in Hammersmith took on elements of the Duluth approach in the development of twin services focused on 'supporting women and challenging men'. The project evaluation is positive about the benefits of both the Women's Support Service and the Violence Prevention Project which provided education groups for abusive men. However, project staff in the Women's Support Service acknowledged that they had had less of an impact in the area of 'institutional advocacy' (Burton, Regan and Kelly, 1998).

The Women's Safety Unit (WSU) in Cardiff is central to a recent development in the U.K. of a criminal justice intervention which aims, like the Duluth model, to address all facets of the criminal justice system. The goals of the WSU include increasing the number of women seeking help; increasing the numbers of arrests, charges and convictions; the extension of appropriate services to women and children, and a reduction in repeat victimisation. The Unit provides advocacy and support for individual women, and works closely with the police and the CPS as part of an interagency approach designed to improve the safety of women and children. 1150 women with 1482 children were referred to the project over the first 14 months. An evaluation of the project included interviews with 222 women who had attended during that period (Robinson, 2003).

South Wales Police have adopted a pro-arrest policy, along with a Police Watch initiative based on the Killingbeck three tier intervention in West Yorkshire. The combination of the pro-arrest policy, Police Watch and liaison with the Women's Safety Unit are credited with a 36% reduction in repeat victimisation. Multi-Agency Risk Assessment Conferences (MARACs) have also been initiated, and these are seen as further enhancing the potential for co-operation between agencies and reductions in repeat victimisation of women (Robinson, 2003). An evaluation of the MARACs has been commissioned, and publication is imminent.

Initially, the CPS in Cardiff agreed to designate a specialist domestic violence prosecutor. This subsequently proved to be unworkable (due to the workload) and it is now agreed that all prosecutors must be able to effectively prosecute domestic violence cases, and that they will work closely with WSU staff. Domestic violence cases are heard at Pre-Trial Review court on Mondays, and a WSU staff member is always in court at this time. The presence of a worker from WSU (usually the seconded police officer) is seen as providing a valuable source of supplementary information which can inform the decision making of prosecutors about how to proceed with a case (Robinson, 2003).

Court procedures have also been reviewed. The court process for domestic violence cases dealt with by the magistrates' court has been reduced from 14 weeks to seven weeks as a result of the streamlining which has taken place. Cases which go to the Crown Court are also now dealt with on Mondays, and since January 2003, there has been agreement that only experienced full time judges will hear domestic violence cases (Robinson, 2003).

Overall, there has been a steady decrease in the number of cases discontinued, or in which the woman retracts. It is surprising, then, to note that cases in which the WSU is involved are more likely to involve retractions or be discontinued, despite the assumption often made that increased support for women will increase the likelihood that they will stay with the criminal justice process. The evaluation report speculates that this outcome may be due in part to the severity of the cases being referred to the WSU, but it is also noted that it is not clear whether the WSU was contacted before or after women retracted - i.e. prosecutors may be referring women to the WSU because they (the women) have decided not to proceed. It may also be that contact with the WSU means that women feel supported in a decision not to proceed with prosecution, for a variety of reasons, not least of which might be that prosecution may not be in the best interests of the women concerned (Robinson, 2003).

Justice system responses to domestic abuse continue to evolve in response to changing perceptions of what is required to protect the women who experience it and to challenge the men who perpetrate it. There has been a general trend towards greater understanding of the effects and dynamics of abuse, and the implications of this for women's ability to seek protection from the justice system. The research reviewed suggests that pro-arrest and pro-prosecution policies give a clear message that domestic abuse is a criminal act, and have some impact on reducing recidivism. However, the overall tenor of the research literature is that these shifts in policy are most effective when located in the context of a coordinated justice system response to domestic abuse.

4.4.2 Support and advocacy services

In Scotland, as elsewhere in the U.K., Women's Aid is acknowledged as the lead organisation providing support for women experiencing domestic abuse, and several studies have affirmed the value women place on the services offered by Women's Aid (Hague and Malos, 1996; Sissons, 1999). The best known aspect of the service offered by Women's Aid is undoubtedly refuge provision. Recent research commissioned by the refuge provision working group 26 notes that women who had accessed refuge were particularly positive about the practical support provided and the non-judgmental, empowering approach taken by Women's Aid workers. The researchers also note, however, that some women indicated that they would value a more proactive approach by refuge staff - "You say if we need support, just ask. But not all women are strong enough to ask, so you should ask more often." (Fitzpatrick et al, 2003: 49).

Women's Aid in Scotland has tended to define the service provided to women as either 'practical' or 'emotional' support. Elsewhere, particularly in North America, 'advocacy' is more routinely referred to. It has been noted by some researchers that there is a degree of confusion about the distinction between 'support' and 'advocacy' for women experiencing domestic abuse. Some service providers use the terms interchangeably; others draw a clear line between the provision of a direct support service and the use of advocacy to ensure that an individual woman receives the service she needs and to which she is entitled (Sullivan and Keefe, 1999). This might involve providing women with information about entitlement to services, and liaising between women and the services they require (Kelly and Humphries, 2001). A further distinction is made between individual advocacy, which aims to facilitate access to services for one woman, and systems or institutional advocacy, which works at a more strategic level to address the failure of institutions to respond appropriately to survivors of domestic abuse as a group (Kelly and Humphries, 2001; Pence, 2001; Riger et al, 2002)). Systems advocacy may be framed around an issue of broad concern to survivors, or it may be a strategy adopted as the result of examining the situation of an individual woman in some detail - "chasing an individual story down" (Burton, Regan and Kelly, 1998:5).

This latter approach may be of particular value in achieving change at a local level (Burton, Regan and Kelly, 1998), but can also be used to achieve broader policy shifts. The campaigns run by Southall Black Sisters and Justice for Women in support of Kiranjit Ahluwahlia, Sara Thornton and Amelia Rossiter, women who had been convicted of killing their abusive partners, played a significant role both in raising awareness of the long term effects of living with domestic abuse, and in facilitating the acceptance of the defence of 'provocation' (Gupta, 2003). Southall Black Sisters have also advocated consistently since the late 1980s for changes in immigration rules. In particular, they campaigned for changes to the 'one year rule' 27 to ensure that women coming to the U.K. to marry are not forced to remain with a partner who is violent (Joshi, 2003).

Advocacy, by its nature, presents a challenge to the status quo. It is no surprise therefore, that the voluntary sector has largely led the way in developing advocacy responses to women experiencing domestic violence. Indeed the earliest services for women experiencing domestic abuse were those provided by support and advocacy organisations in the voluntary sector, often acting to 'fill the gap' in existing service provision. As an extension to this 'gap filling' role, women's support and advocacy organisations also took on the task of challenging the practice of statutory agencies, demanding services and legal protection for survivors of domestic violence (Pence, 2001). Kelly and Humphries note that although individual social workers, healthcare staff and even police officers take on the role of 'advocate' for women, their ability to challenge may be compromised by their role within a statutory agency (Kelly and Humphries, 2001).

There has been relatively little research of either support or advocacy interventions which examines outcomes or efficacy (Kelly and Humphries, 2001; Sullivan and Bybee, 1999). A review of outreach and advocacy approaches to women experiencing domestic violence identified only three evaluated outreach or advocacy projects in the U.K. (Kelly and Humphries, 2001). The picture is similar in the U.S. (Abel, 2000; Sullivan and Bybee, 1999). Abel suggests that this should not be surprising given the nature of the work, which is largely based on crisis intervention and often focused on women in refuge. She acknowledges that services must inevitably be fairly open ended to meet the needs of women in this situation, but notes that this creates difficulty in designing outcome evaluations (Abel, 2000). This is echoed in a recent evaluation of the Scottish Domestic Abuse Helpline, which acknowledges that the short-term transient nature of the contact with women callers, as well as the focus on crisis intervention, makes it difficult for helpline volunteers to ask women questions to support an evaluation of outcomes. The evaluation therefore focuses primarily on the processes involved in running the helpline, and a statistical analysis of calls received (Brown, 2004).

The Michigan-based Community Advocacy Project sought to reduce women's risk of violence from male partners through the use of a structured intervention programme delivered by trained volunteer advocates. 278 women participated in the study; all had been resident in the local refuge for at least one night. Women were interviewed six times over a two year period. The study was evaluated using a control group approach - at the first interview, women were given a sealed envelope which randomly assigned them either into the group which would receive the intervention, or into the control group. Women assigned to the control group were not contacted again until the second interview, 10 weeks after this initial contact. Women who received the intervention (n=143) were supported by a volunteer advocate for 10 weeks; on average this involved two meetings per week, an average of 6.4 hours per week. During that time, they identified what was needed to achieve positive change, including safety planning and accessing appropriate protection from the criminal justice system, but also unmet health, social and economic needs (Sullivan and Bybee, 1999).

All of the women, including the control group, were interviewed again at 10 weeks, and then at 6-, 12-, 18- and 24 months. A range of formal test measures were used, assessing the levels of physical or psychological abuse women experienced, their quality of life, depression, degree of social support available, effectiveness in obtaining resources/accessing services, and women's perceptions of their difficulty in obtaining resources. Overall, the results were perhaps not surprising - women who received the intervention experienced less physical violence, and reported improvement in their quality of life, levels of social support and feelings of depression. 25% of the women who received the intervention experienced no further violence over the two years of the study, compared with only 10% of the women in the control group; both groups of women, however, experienced significant incidences of further violence - 79% of the intervention group and 89% of the control group (Sullivan and Bybee, 1999).

Although the Michigan project undoubtedly demonstrates that the provision of advocacy increases women's safety, the research design raises some questions about the ethics of using a control group approach to evaluate interventions to reduce violence against women. The random nature of the control group selection implies that no risk assessment was undertaken, thus excluding women in high risk situations from a potentially life saving intervention.

Domestic Violence Matters (DVM) adopted a different approach to evaluation. This pilot project ran in Islington and Holloway (police divisions) in London from early 1993 for 32 months. The project aimed to provide crisis intervention to women, enhance the criminal justice system response, and promote interagency links and co-ordination. Five civilian workers - four support workers and a co-ordinator - provided crisis intervention between 10am and 2am every day, including holidays. They were located within a police station, and aimed to follow up all domestic violence incidents within 24 hours of the incident being reported to the police. They achieved this in 90% of the 1542 incidents they responded to (Kelly, 1999).

DVM adopted a proactive approach to intervention, maintaining contact with women following the initial referral, rather than waiting for women to make contact themselves. Although most (70%) of the referrals came via the police, women also referred themselves. Over two thirds of the referrals came outwith normal office hours, confirming the need for out of hours services. As well as general support, legal advice was provided in 86% of cases, housing advice in 60%. Crisis planning was provided in 30% of cases, and accompaniment to safe accommodation in 15%. (Kelly, 1999).

DVM defined 'crisis' as "any point at which routine coping mechanisms break down and the need or potential for change is present" (Kelly, 1999: 15). Kelly suggests that effective crisis intervention is about enabling change, and explains,

"…change was not conceptualised by DVM solely in terms of leaving or taking legal action…Rather it was much more fluid and variable; the basic requirement being only that it shift the dynamics of power and control which underpin domestic violence in the woman's favour; ensuring that she had more resources after intervention than before it. This could be strengthened resolve, accurate information, access to other agencies, or a firmer alliance with the criminal justice system; often it was a combination." (Kelly,999:16).

Thus the potential 'performance indicators' for DVM were broad, qualitative and largely assessed by means of feedback from service users. Questionnaires were sent out to 789 people; 221 women responded, and two men 28. These initial questionnaires were completed either within a week of first contact with DVM (32%), within a month (30%) or more than a month later (38%). A further 23 women participated in a follow up interview - 14 of these were interviewed 12 months after first contact with DVM, and nine were interviewed six months after first contact.

In the initial questionnaire, women identified the immediacy of the response as one of the most positive aspects of the service. 151 of the women gave specific examples of how this had helped them, including comments that it gave them space to explore options, information and support with the practicalities of leaving, and affirmation that their partner's behaviour was unacceptable. The immediacy of the support was credited by some women as having been crucial to their ability to talk about the abuse - "If more time had lapsed I probably wouldn't have talked to anyone about it", or to safeguard themselves - "If the response had not been immediate I would probably have taken him back." (Kelly, 1999: 29). The majority of the women (62%) indicated that what they wanted most from DVM was someone to talk to about what had been happening. Few women, by comparison, wanted to discuss the implications of arrest (13%) or prosecution (17%). Only 15% of the women were still living with the abusive partner at the point of completing the questionnaire, although 50% had been living with him at the time of the incident. Of the 23 women who took part in the follow up interviews, only five were still living with their partner, and there had been no further violence or threats of violence against these five women since their contact with DVM (Kelly, 1999).

The DVM approach was characterized by proactive crisis intervention, with staff taking responsibility for making and maintaining contact with women, rather than waiting for women to initiate contact. Summarising women's views on this approach, Kelly notes that:

"The importance of home visits and follow ups illustrate very clearly that pro-active responses are neither resented by women nor ineffectual; rather they appear to accelerate a process of change in a manner which both at the time and retrospectively are valued positively." (Kelly, 1999: 34)

The evaluation of Domestic Violence Matters has at the heart of it women's perceptions of the service they received, how it helped them to change their situation, and what could have made it better. Service user evaluation can go some way towards informing the future development of individual services. Wider consultation with women survivors of domestic abuse is rare, and where it exists, there may be sharp distinctions between service providers' perceptions of the degree to which they consult with women, and women's perceptions of the extent to which their views actually influence policy or service development (Hague, Mullender and Aris, 2003). It is worth noting that at present, only five of the local multiagency partnerships on violence against women in Scotland indicate that they have or are planning to establish some form of consultation with women survivors 29.

4.4.3 The health service response

The involvement of women service users in the design of services was one of the key recommendations of the Scottish Needs Assessment Programme (SNAP) report on domestic violence published in 1997. Identifying domestic violence as "a significant public health issue" (SNAP, 1997: i), the report noted that "between 260,000 and 700,000 women may be experiencing domestic violence" (SNAP, 1997:i). The report made several recommendations, covering the main policy and practice developments which the authors identified as necessary in order to improve the health service response to domestic violence. These included staff training, the development of a monitoring and recording framework, and the development of guidelines specific to each health service setting.

The recent publication by NHS Scotland of guidance for healthcare staff on responding to domestic abuse incorporates some of the recommendations of the SNAP report. The guidance highlights the need for awareness training for staff as a central component in developing a more effective health service response to domestic abuse. Recognising signs and indicators of abuse, supporting disclosure by women, risk assessment and safety planning, and the need for accurate record keeping are also addressed (Scottish Executive, 2003).

The NHS Scotland guidance acknowledges the difficulty women may experience in voluntarily disclosing that they are experiencing domestic abuse. It also acknowledges the value of early intervention. The question of how proactive healthcare professionals should be in asking women whether they are being abused by their partner is perhaps less clear in the guidance, and this may be a reflection of the ambivalence within the healthcare community as a whole about the issue of 'screening' or 'routine enquiry' for domestic abuse.

In the United States, screening has been recommended by the American Medical Association and other professional bodies since 1992, although professional groups may differ in the particular approach they advocate (Family Violence Prevention Fund, 1999). Despite the general support for screening expressed in professional guidelines, resistance to implementing screening protocols in the U.S. is still widespread. A survey of 2,400 doctors from a range of medical settings found that only 6% of the 1103 respondents screened all their female patients (Elliott et al, 2002). Barriers to screening included a perceived lack of appropriate interventions (45%), concerns about offending patients (33%), and simply forgetting to ask (41%). Lack of time was cited by only 21% of the respondents (Elliott et al, 2002).

The inconsistent application of a screening approach in an emergency room setting is reported in a study from Ohio, and a range of reasons for this were identified by the researchers. Some members of staff felt it was not their responsibility, others said they were not aware that they were supposed to be screening. Some were clearly uncomfortable about 'asking the question', while others asserted that they "didn't not know where the forms were kept". In addition, the researchers note that no data was collected during night shifts. The researchers report that they had gone to some lengths to ensure that staff were aware of the study and the reasons for it, and that they had spent some time preparing staff for the study through the provision of films and written material. They suggest that continuing education and training may assist with a more consistent approach in the future (Heinzer and Krimm, 2002).

The findings of these studies are consistent with experiences in Scotland and the rest of the U.K. Where attempts are made to implement screening protocols, no matter how selective, there is still a significant degree of resistance from practitioners. Studies in Scotland of GP responses to women concur with much of the North American literature, with GPs identifying lack of time, lack of appropriate services and concerns about offending women with intrusive questions as significant barriers to asking about domestic abuse (Cosgrove, 1998; McKie, 2002; Munro, 2001).

Similar concerns were raised by some of the participants in a London study of midwives' perceptions of routine enquiry in relation to domestic violence. A three hour training session was provided to 145 midwives. They were given information about the prevalence and consequences of domestic abuse, how to administer a screening tool, and how to refer women to local agencies. Although participants were enthusiastic about the study, and saw it as relevant to their work, they acknowledged practical difficulties with implementing it. Time constraints and competing priorities, both for the women (e.g. financial or social difficulties) and for the midwives (e.g. the numbers of women to be seen in a clinic, and the quantity of other information which had to be shared with women during appointments) were all identified as barriers to applying a consistent approach to screening. Midwives expressed concern about possibly placing women at greater risk as a result of attempting to exclude a partner from a consultation. They also raised concerns about their own safety (Mezey et al, 2003).

Many practical problems associated with implementing screening for domestic violence were identified. The authors note that screening is time consuming, and that there are logistical problems associated with creating an appropriate and safe environment in which to 'ask the question'. They also note that many of the participants disclosed personal experiences of domestic abuse, and comment on the impact this had on the way they engaged with the study. For some it created an additional barrier to implementing the screening tool; for others, their personal experience appeared to enhance their ability to identify signs of domestic abuse and respond appropriately. Much appeared to rest on how far workers had resolved their feelings about what had happened in their own lives (Mezey et al, 2003).

The training provided before the study gave a clear message that midwives were not expected to take on a counselling role. They were asked only to identify and assess the needs of women experiencing domestic abuse before referring on to other agencies. In practice, boundaries are harder to maintain than this. The process of disclosure often involves more of a personal commitment from workers than is implied in simply following a protocol, and the reality is that 'asking the question' may often involve more work than can be anticipated. The authors of this study raise a question as to "the practicality of domestic violence screening by NHS staff within a busy clinical service", observing that as soon as this study was ended, most of the midwives stopped asking women questions about domestic violence (Mezey et al, 2003:751).

A great deal of the research on health interventions has focused on the use of screening tools for domestic violence, and in particular, barriers to implementation by healthcare staff. More recently, two systematic reviews have considered whether there is sufficient evidence that screening leads either to appropriate intervention, or to improved outcomes for women (Ramsay et al, 2002; Wathen and MacMillan, 2003).

The first review, conducted by a U.K. based team of researchers, sought to "assess the evidence for the acceptability and effectiveness of screening women for domestic violence in healthcare settings" (Ramsay et al, 2002: 314) and to this end reviewed three groups of research studies. The first group explored the attitudes of women and health professionals towards screening for domestic abuse in healthcare settings. The second group compared identification rates for domestic abuse between health settings which used screening and those which did not. The third group measured the outcome of interventions with women who had experienced abuse. Studies included in the third group were limited to those which provided a comparison with a group of women who had received no intervention. The inclusion criteria were strictly applied, and a total of 20 papers were eventually reviewed, from a starting sample of 2520 papers identified from a search of three databases (Ramsay et al, 2002).

In relation to the acceptability of screening, across the five studies reviewed in this category, 43-85% of all women surveyed thought screening for domestic abuse was acceptable (with women who had experienced abuse at the higher end of that scale). Much lower percentages of health professionals favoured screening, giving many or the reasons already discussed above. On the whole, the nine studies which assessed 'numbers of women identified' found that more women experiencing domestic violence were identified as the result of screening. However, the increase tended not to be substantial, and there was some evidence that improved identification was not sustained beyond the period of the study (Ramsay et al, 2002).

The six studies which examined interventions with women who had experienced domestic abuse were the most contentious. The review is critical of the design of studies, highlighting weaknesses in the methodology of most, and in particular the absence of randomized control trials, or any focus on qualitative outcome measures such as 'quality of life' or improved mental health. Overall, the reviewers found little evidence that any of the interventions they considered were effective, and the main conclusion of the review is that there is insufficient evidence to support the use of screening at this time. They suggest that more research is required to identify effective healthcare interventions with abused women (Ramsay et al, 2002).

A more recent review by two Canadian researchers looked more specifically at primary health care interventions designed to prevent further abuse of women. The review identifies two main options open to primary care professionals - identification of women, and referral to appropriate agencies. In relation to the former, the reviewers reach the same conclusion as Ramsay et al, i.e. that no studies to date have demonstrated that screening improves the outcome for women. The authors also note that no research has been conducted to assess potential harm to women as a result of screening (Wathen and MacMillan, 2003).

Overall, Wathen and MacMillan conclude that there is a dearth of evidence based interventions in response to domestic abuse either within a primary care setting or beyond. They concur with Ramsay et al that there is a need for further research, in particular to determine whether a combination of screening by healthcare workers and effective intervention would reduce physical and emotional abuse of women (Wathen and MacMillan, 2003).

Neither of these reviews suggest that clinicians should never ask women about domestic violence - the Canadian review in particular stresses the importance of asking questions to elicit information about domestic abuse when there are indicators of it during routine history taking (Wathen and MacMillan, 2003). Both reviews acknowledge the importance of training for health professionals in raising awareness of the impact of domestic abuse, and supporting the development of effective interventions (Ramsay et al, 2002; Wathen and MacMillan, 2003).

Despite the acknowledged impact of domestic abuse on women's health, the response of health services beyond the treatment of immediate injuries is still largely confined to identification and referral on to other agencies. Furthermore, there is little quantitative research to indicate whether either of these approaches are effective responses to women. However, it is worth noting that, despite the lack of a clinical evidence base for screening, and despite the reticence of healthcare staff in implementing screening protocols, the majority of women want health professionals to ask about domestic violence. In particular, women who have experienced domestic violence want health professionals to 'ask the question' (Ramsay et al, 2002; Taket et al, 2003).

4.4.4 Multiagency responses to domestic abuse

The development of a multiagency response to domestic abuse is now widely acknowledged as the most effective way both to support and protect women and children who have experienced domestic abuse, and to challenge male perpetrators (Hague and Malos, 1996; Hague, 2001; Pence and McDonnell, 1999; Scottish Executive, 2002). Some of the features of multiagency criminal justice responses to domestic abuse are addressed elsewhere in this report. However, multiagency responses to domestic abuse may encompass a much more diverse range of agencies, including social work, housing, health, education and the voluntary sector, in particular Women's Aid (CoSLA, 1998; Scottish Executive 2002).

Individual practitioners from different agencies may work very successfully together at an individual case level without any formal interagency agreement. However, the development of a consistent coordinated response is more likely to be achieved when agencies engage in more formal strategic partnerships (Moelwyn-Hughes, 1999). It is crucial that multiagency partnerships on domestic abuse move beyond simply defining the problem and identifying the gaps in services. The development and implementation of shared policies and procedures, the provision of information and training for staff, and the establishment of effective monitoring systems are all essential steps in the process of ensuring that multiagency strategies to tackle domestic abuse move from being statements of intent to supporting the development of good practice (Gamache and Asmus, 1999; Hague and Malos, 1996; Hague, 2001).

Although there appears to be a degree of agreement about the key principles which underpin effective multiagency work, the specifics will vary from one area to another. The geography and demographics of an area, the number and range of services involved, and the gaps in local service provision will all play a role in shaping the development and delivery of a multiagency response to domestic abuse (Hague and Malos, 1996). The commitment of key personnel within the partner agencies will also be significant. Senior managers may not attend partnership meetings, but their support is crucial to ensuring that proposed policy and practice changes are adopted and implemented (Hague, 2001). The involvement of women's activist organisations, including Women's Aid, in multiagency partnerships on domestic abuse ensures that the safety of women and children remains central to the process, and the impetus for change is not diluted (Kelly, 1999).

Good practice examples of multiagency work to tackle domestic abuse can be found in many areas, both in the UK and internationally. The 'Duluth model' is widely acknowledged to have provided a benchmark against which to measure coordinated criminal justice responses. A central component in the success of the Duluth initiative was the establishment of an independent non-profit organization, the Domestic Abuse Intervention Project (DAIP), to take on the role of coordinating and monitoring the criminal justice response to domestic abuse. The case tracking approach adopted by DAIP to support this work is undoubtedly very effective in identifying problems within the criminal justice system.

The Duluth Model is resource intensive, and this has been identified as a barrier to the adoption of the model in other areas, particularly those with larger populations (Hague, Kelly and Mullender, 2001). The London Coordinating Committee to End Woman Abuse (LCCEWA), for example, has been at the forefront of developing interagency responses to domestic abuse, but has not gone down the case tracking route. Instead, LCCEWA has developed an action research approach, with a focus on short life projects with achievable goals (Hague, Kelly and Mullender, 2001). Membership of LCCEWA is diverse, and women's advocacy organisations have played a central role in the development and leadership of the Committee. Research interviews with women who had sought assistance from services in London suggests that the model is working - the majority were satisfied with the response they got when making the initial approach to services, regardless of which service they approached first. They also reported that appropriate referrals were made to other agencies. The researchers note that with one or two exceptions, there appeared to be a high level of awareness among service providers about the range of relevant services which might provide additional support to women and children (Grasely et al, 1999).

A review of multiagency initiatives to tackle domestic abuse which was carried out in the mid 1990s found that multiagency fora and partnerships had some difficulty in identifying ways to evaluate the effectiveness of the work they were doing. In part this may be due to a lack of clarity about what, exactly, is being evaluated. It may also be difficult to identify appropriate baseline data against which to measure progress. Although there appears to be a general concensus that the goal of any multiagency response to domestic abuse should be improved safety for women and children, in practice it seems that it is difficult to demonstrate whether this has been achieved (Hague, 2001).

Since then there has been considerable development in the understanding of what is required to overcome some of the barriers to effective multiagency work. In Scotland, the commitment to working collectively and collaboratively to tackle domestic abuse was reaffirmed with the publication of the National Strategy on domestic abuse in 2001. Significantly increased resourcing of local domestic abuse partnerships through the Domestic Abuse Service Development Fund (DASDF) has accelerated the development of local strategies and action plans. These have been supported to a large extent by the increased resourcing across Scotland of services for women and children affected by domestic abuse. A recent evaluation of the impact of the DASDF involved a postal survey of local partnerships. Several respondents identified problems within their multiagency groups, including lack of a shared agenda or a shared understanding of the problem; differences in the capacity of partner agencies, and therefore in their ability to contribute at times to the work; and power differentials between the partners, particularly between statutory and voluntary sector members (Reid-Howie Associates, 2003). It was not within the remit of the DASDF evaluation to carry out evaluation of the individual projects funded. It would seem, however, that some research on the effectiveness of the current multiagency response to domestic abuse in Scotland would be worth considering.

4.4.5 Responding to violence against women - conclusions

There is a substantial body of literature on how services respond to violence against women. However, there is a significant quantitative difference between research which focuses on domestic abuse and that which considers other aspects of violence against women. In part, it must be acknowledged that this is reflective of the much greater numbers of women reporting domestic abuse, compared with other crimes of violence against women. It may also reflect the greater impact which domestic abuse has on service provision, since a wider group of service providers have a statutory responsibility to respond to domestic abuse than, for example, to rape or sexual assault. This is not to say that the needs of survivors of rape or sexual assault, or survivors of child sexual abuse, are necessarily seen as less valid than survivors of domestic abuse.

In theory, it should be possible to adapt responses to women who have experienced domestic abuse to meet the needs of survivors of other forms of male violence. However, the possibility of adapting and replicating interventions is rarely addressed in the literature, and only one research study was identified which attempted to evaluate the use of the same intervention with survivors of different types of violence.

In summary:

  • The research literature repeatedly acknowledges the role of women's NGOs in the development of services which are responsive to women's needs.
  • Research focuses on discrete issues, although it is recognised that women may experience more than one form of abuse at more than one point in their lives.
  • There is a lack of evaluative research in most areas. More evaluation of responses to domestic abuse has been carried out than of responses to other forms of violence against women.
  • This is perhaps reflective of the general picture of services, which can be described as "patchy and inconsistent", with substantially more service development worldwide in response to domestic abuse than there is to other forms of violence against women.
  • Different aspects of violence against women are addressed in different areas of the literature.
  • Research on interventions with rape survivors is primarily focussed on medico-legal responses, with some literature on therapeutic interventions, but little on interventions by primary care workers. Although there have been recent improvements in the treatment of rape and sexual assault complainers there are still areas of concern, particularly in relation to conviction rates. Rape crisis provision is still poor across Scotland, and there is a lack of independent evaluation of the approach. There is a similar lack of evaluation of sexual assault referral centres (SARCs), although a forthcoming report from the Home Office should address this.
  • There are different degrees of understanding or acceptance of prostitution as 'violence against women'. Some would make a distinction between 'forced prostitution', including trafficking, and prostitution as an active choice by women. Others assert that the harm caused to women by prostitution should define it as a form of violence against women. Much of the literature on women working in prostitution is taken up with questions of definition and agency, and in this, it reflects early debates about how far women 'choose' to stay with violent partners. There is also a significant body of literature which considers crime management interventions. There is little on interventions which support women abused in prostitution, or assist them in leaving.
  • Research on interventions with adult survivors of childhood sexual abuse is primarily found in the mental health literature. Cognitive behavioural therapy (CBT) appears to achieve the most consistent outcomes. However, this may be partly due to the relative ease with which a CBT approach can be measured, compared with other therapies. Research with survivors of childhood sexual abuse suggests that what they value above all is the warmth and empathy displayed by practitioners, and that they are less concerned with the particular therapeutic approach being used. Survivors of childhood sexual abuse also praised the service provided by voluntary sector support organisations. It is acknowledged that this type of service provision is poorly distributed across Scotland, and that there is a lack of evaluative research.
  • Although some work has been carried out which explores the links between childhood sexual abuse and chronic physical health problems, no research was identified which addressed how healthcare staff should acknowledge this or respond to it. No research into the criminal justice response to adult survivors of childhood sexual abuse was identified.
  • By comparison, research on interventions with domestic abuse survivors cuts across several sectors, including criminal justice, acute and primary care health services, social work services and outreach and advocacy services.
  • Research on the justice system response to domestic abuse suggests that pro-arrest and pro-prosecution policies give a clear message that domestic abuse is a criminal act, and this has some impact on reducing recidivism. However, the overall tenor of the literature is that these shifts in policy are most effective when located in the context of a coordinated justice system response to domestic abuse.
  • Evaluation of support and advocacy services for domestic abuse survivors is limited, but the research which has been done shows that such services are much valued by women, and that they may have a significant role in improving women's safety.
  • Despite the acknowledged impact of domestic abuse on women's physical and mental health, research into healthcare responses to domestic abuse has not gone much beyond issues of identification and assessment. The literature on screening or routine enquiry for domestic abuse is contentious. There is an emerging concensus among health researchers in the U.K. and Canada that there is limited evidence as yet to support a routine enquiry approach; this is at odds with the position in the U.S., where screening is well established and supported by all the major professional bodies.
  • The scarcity of research on interventions for black and minority ethnic women, women with disabilities, lesbian women, and older women affected by male violence against women is perhaps a reflection of the dearth of services for these groups.
  • Across all of the literature, across all aspects of violence against women, there is a noticeable absence of attempts to engage women survivors of violence in the development, design or evaluation of services or policy initiatives.

« Previous | Contents | Next »

Page updated: Thursday, March 24, 2005