Final Project Bibliography Checklist

A Stop – Start Response: Social Services’ Interventions with Children and Families Notified following Domestic Violence Incidents

Nicky Stanley *, Pam Miller, Helen Richardson Foster, and Gill Thomson

Nicky Stanley is Professor of Social Work at the University of Central Lancashire. She managed the research study described in this paper and is currently undertaking research on services for

male perpetrators of domestic violence as well as leading on the national evaluation of social work practices. She is co-editor of Child Abuse Review and has published books on such issues as domestic violence and child protection, mothers’ mental health needs and inquiries in health and social care. Pam Miller, MSSW, was a Research Fellow on the project, ‘Children and families experiencing domestic violence: Police and children’s services responses’. Pam is

currently a Research Officer at the NSPCC and is undertaking research into services in London for children and young people who live with domestic violence. Helen Richardson Foster, MSc, was a Research Fellow on the project, ‘Children and families experiencing

domestic violence: Police and children’s services responses’. Helen is currently a freelance social researcher and her research interests include domestic violence, parenting and work

with vulnerable children and young people. Dr Gill Thomson is a Research Associate who works within the Maternal and Infant Nutrition and Nurture Unit at the University of Central Lancashire. Gill was employed on the study reported here for approximately ten months and

was primarily involved in the collection and analysis of the police data. Gill’s current research projects relate to socio-cultural influences of women’s experiences of maternity care and infant

feeding.

*Correspondence to Professor Nicky Stanley, School of Social Work, University of Central Lancashire, Preston PR1 2HE. E-mail: [email protected]

Abstract

The harm consequent on children’s exposure to domestic violence is recognised in legis-

lation in England and Wales. This paper reports on a study of the social work response to

184 families notified by the police to children’s services in two English authorities.

Families were tracked through case records over 21 months subsequent to the notifica-

tion. The perspectives of social services’ practitioners and managers were also captured

through interviews. Only a small proportion of families received a service in the form of

an initial assessment or further intervention; the notification triggered a service for just

five per cent of families. Families who received a warning letter only were just as likely to

# The Author 2010. Published by Oxford University Press on behalf of

The British Association of Social Workers. All rights reserved.

British Journal of Social Work (2011) 41, 296–313 doi:10.1093/bjsw/bcq071 Advance Access publication June 19, 2010

be re-referred as those who met with no response. Those families receiving a service

were likely to experience repeated notifications and assessments. The limited time

period for completing assessments contributed to initial assessment workers’ lack of

engagement with perpetrators of domestic violence. Current structures for assessment

and intervention contribute to a stop-start pattern of social work that seems ill-suited

to building the trust and engagement needed to challenge the complex and enduring

experience of domestic violence.

Keywords: children, domestic violence, social services, interventions

Accepted: May 2010

Introduction

In England and Wales, the 2002 Adoption and Children Act identified ‘impairment suffered from seeing or hearing the ill-treatment of another’ (s. 120) as a form of significant harm. This inclusion of children’s experience of domestic violence within the category of child abuse owed much to the accumulating body of evidence on the impact of domestic violence on chil- dren. Such research has included accounts of mothers and children (McGee, 2000; Mullender et al., 2002; Buckley et al., 2007) and reviews detailing the impact of exposure to domestic violence on children (Cleaver et al., 1999; Edleson, 1999; Holt et al., 2008). The legislation crys- tallised the shift from a conceptualisation of domestic violence as an adults’ issue to a view that children were intimately involved in the problem. However, it also had the effect of drawing a new and potentially vast group of children and families into the auspices of children’s social services.

In the 1980s and 1990s, statutory social work in the UK was much criti- cised for its failure to acknowledge the presence and impact of domestic violence in families (Farmer and Owen, 1995; Stanley, 1997). As social work has become more sensitised to the issue, criticisms have moved to focus on the nature of the response. International research has explored key questions such as what priority is allocated to domestic violence (Irwin and Waugh, 2007; English et al., 2005; Devaney, 2008) and whether the social work response to domestic violence is able to balance the needs of children with those of victims and perpetrators (Humphreys, 2007). The extent to which children’s social services work collaboratively with other key agencies operating in this field, especially third-sector dom- estic violence services, has also been explored (Banks et al., 2008; Cleaver et al., 2007).

These concerns link to wider discussions about social work’s capacity and readiness to engage with men and fathers (Scourfield, 2003; Featherstone, 2009). Domestic violence provides the focus for some of these debates, as a failure to engage with male perpetrators who are also fathers appears

A Stop – Start Response 297

to leave social workers in the difficult-to-defend position of demanding that mothers who are the victims of abuse protect their children from exposure to that experience (Lapierre, 2008). The study reported here provided the opportunity to examine social work practice in England in relation to families experiencing domestic violence and to explore some of these issues.

Methodology

This study was undertaken between 2007 and 2009 and utilised two research sites in north and south England that were well matched in terms of the size of the children’s population. While one was a metropolitan area with a very diverse population, the other contained a mixture of urban and rural areas with ethnic diversity concentrated in the towns.

The main stage of the research constituted a study of professional prac- tice that included retrospective analysis of agency records and interviews with practitioners and managers. Cases were originally identified as dom- estic violence incidents in police records and were tracked through to social services’ case files. Findings regarding police practice and inter- agency work between police and children’s social services are reported else- where (Stanley et al., 2010); this paper focuses on children’s social services’ response to notifications. The sample was created by selecting all incidents of domestic violence notified by the police in January 2007. This allowed follow-up of children and families in social services’ records over a period of twenty-one months following the original notification.

Identifying markers assigned to police records were used to locate files on notified cases in children’s services’ records. However, in common with similar research studies (Beeman et al., 2001), the researchers experienced difficulty in matching up police and social services records and the original sample of 251 domestic violence incidents was reduced to a total of 184 families who, between them, had 196 reported incidents (some had more than one incident) for January 2007. This process of attrition can be attrib- uted to a number of factors, including multiple notifications, files being lost or unavailable and the likelihood that some notifications were never received.

As only a small proportion of the notified cases received a service com- prising an assessment or intervention, the number of cases in which it was possible to examine social work interventions in depth was reduced to twenty-eight. In order to increase the sample size at this stage of analysis, a booster sample was constructed of notified cases from February 2007 that received an initial assessment or further intervention. This increased the total number of cases receiving a substantial social work intervention to forty-six, facilitating examination of sub-groups.

298 Nicky Stanley et al.

Since case records only convey what is recorded and since practice in our two sites in relation to notifications had developed between 2007 and 2008, a series of semi-structured interviews was completed with twenty-five social workers, administrative staff and managers who volunteered to be inter- viewed in both sites, asking them about information contained in notifica- tions, systems for managing notifications and their practice in relation to families experiencing domestic violence. Interviews were recorded with the participants’ agreement and informed consent procedures were adopted. Ethical approval for the study was given by the University of Central Lancashire’s Ethics Committee.

Other stages of the research included a survey of innovative practice in relation to the management of notifications of domestic violence incidents in England and Wales and interviews with young people, survivors and per- petrators of domestic violence. Participants were recruited separately through relevant services so they had no relationship with one another. Nineteen young people aged ten to nineteen participated in five focus groups held in settings in which they were familiar with discussing domestic violence in a group. Eleven survivors of domestic violence who were parents were contacted through specialist domestic violence services and interviewed individually, as were ten perpetrators (all fathers) recruited from both voluntary and mandated perpetrator programmes. These inter- views aimed to capture perceptions of domestic violence and its impact on children and young people as well as expectations and experiences of services. Some of these data are reported here to reinforce the findings on professional practice.

The analysis of file data entailed entering anonymised data into prede- signed data collection forms and from there into spreadsheets. Quantitative data were coded and analysed using SPSS. Qualitative data were extracted from the spreadsheets and thematically analysed; case vignettes that allowed interventions to be mapped across the follow-up period were also developed. All interviews were transcribed and NVivo was used to sort the data for analysis. Themes were identified both in accordance with the existing literature and as they arose from the data using Grounded Theory principles (Strauss and Corbin, 1990; Charmaz, 2006) with cat- egories being double-checked and refined in the course of analysis.

Findings Sample characteristics

The sample of 184 notified families located in social services files included 333 children of whom slightly over half (51 per cent) were aged under five. Police records noted the ethnic appearance of 277 children: 77 per cent were described as white European; 5 per cent dark European; 13 per cent

A Stop – Start Response 299

Black African; 5 per cent Asian; and less than 1 per cent Arab. Police records for the 196 incidents did not record or were unclear about where- abouts for 25 per cent of the potential child witnesses. Forty-five per cent of children were recorded as directly witnessing the violence in these inci- dents; 5 per cent were not at home at the time. Although police records did not always distinguish a single perpetrator, the majority (93 per cent) of identified perpetrators were men. Fifty-three per cent of the adults involved in incidents were identified as ex-partners, while 47 per cent were described as current partners. For the majority of children (62 per cent, n ¼ 208), both the adults involved in the incident were their birth parents.

Cleaver et al. (1999) emphasise that the risks for children are highest when domestic violence occurs in the context of other problems. Table 1 shows that seventy-five identified perpetrators (four of whom were women) were described by the police at the incident as being drunk or having been drinking or had a self-identified alcohol problem. Similarly, thirty-seven victims were identified as having alcohol problems and this is consistent with the considerable body of evidence highlighting the role of alcohol as a contributory factor in domestic violence incidents (Gilchrist et al., 2003; Galvani, 2004). Fourteen perpetrators and one victim were identified by the police as having a drug problem. Fourteen perpetrators and nine victims were described as having a mental health problem by either their partners or themselves.

Among those families receiving a service, one parent was recorded as having a learning disability and in two families, a child was identified as having a learning disability. Physical health issues were found to affect two parents and seven children in families receiving a service. Families where the child or parent had learning disabilities or physical health issues were over-represented in the cases receiving family support or safe- guarding services but this might have been a reflection of the fact that more was known about families that received a service.

The vast majority (86 per cent) of incidents took place at the victim’s home, emphasising the fact that domestic violence occurs in the place where children should expect to feel safe. The range of incidents was wide, with the researchers using the information contained in police records to classify 54 per cent as low-level or purely verbal incidents, 20 per cent as medium-level incidents that involved verbal abuse with some physical abuse but without any injuries being sustained and 26 per cent as high-level incidents involving a high level of violence or a verbal argument with threats to kill. Injuries were inflicted on adults in 30 per cent of the incidents; children were injured in three incidents.

The notification forms received by children’s services contained rela- tively little information about the extent of children’s involvement in the incidents. However, the original police records revealed more about the context and it was clear that children were intimately involved in many of the incidents. Thirty-six incidents occurred after a former partner was

300 Nicky Stanley et al.

denied access to the house or children (usually because he had been drink- ing) and sought to enter the home by violent means. Twenty-six incidents occurred in the context of contact visits or handovers and fifteen incidents appeared to be related to arguments about parenting or disciplining the children. Children made the call to the police on nine occasions.

No further action cases

File data from the 184 cases were analysed to determine the service pathway cases followed in the first instance. As Table 2 shows, 60 per cent of cases resulted in ‘no further action’. Three-quarters of these ‘no further action’ cases had no or low (i.e. previous notifications or referrals that had been closed on receipt with no further action) previous levels of involvement with children’s social services in that authority. This contrasted with the finding that nineteen of the twenty-eight cases that received a family support or safe- guarding assessment or intervention were already open cases. The pattern was therefore one by which families most likely to receive a service were those with whom children’s services were already involved. Notifications triggered a new service for only 5 per cent of the sample.

The failure of so many of these notified cases to reach children’s services’ threshold for intervention was attributed by social workers to the low level of seriousness of the incidents and 60 per cent of the cases following a ‘no further action pathway’ were classified by the researchers as verbal alterca- tions only. However, 55 per cent of the cases in the sample in which an adult

Table 1 Victims’ and perpetrators’ alcohol, substance misuse and mental health problems as ident- ified by police at the incident

Perpetrators Victims

Alcohol problems 75 (45.7%) 37 (22.6%) Drug use 14 (8.4%) 1 (0.6%) Mental health problems 14 (8.5%) 9 (5.5%)

Table 2 Service pathways taken by sample cases

Case pathway Research site Total

North South

No further action 43 (23%) 68 (84%) 111 (60%) Letters only 42 (41%) 0 (0%) 42 (23%) Visit/phone call only 3 (3%) 0 (0%) 3 (2%) Family support assessment/service 6 (6%) 3 (4%) 9 (5%) Safeguarding assessment/service 9 (9%) 10 (12%) 19 (10%) Total 103 (100%) 81 (100%) 184 (100%)

A Stop – Start Response 301

was injured followed a ‘no further action pathway’ (although all those that involved a child being injured did receive a service) and, similarly, eight of the thirteen cases in which the notification conveyed information about the use of weapons in an incident followed a no further action pathway.

Table 3 shows the subsequent involvement of the ‘no further action’ families with children’s social services over the following twenty-one months. Just over half were re-notified or referred by other agencies; some families were re-notified on multiple occasions: two families were notified on six occasions but did not receive an assessment. Seventeen per cent of these originally ‘no further action’ families received an assess- ment consequent to re-notification/re-referral, suggesting that for some families, increased contact with social services increased their chances of receiving a service. Staff noted that managing these re-notifications rep- resented a considerable demand on the service:

I think we spend a lot of time trying to assess whether or not we should be involved . . . that is very resource intensive regarding the actual resources I have to provide a front line service (Senior Manager 2).

Letters

Although some areas in England and Wales are now utilising particular tools and protocols for screening notifications, such approaches were only implemented in our two sites after 2007. At the point of data sampling, letters were being used in one site as a means of managing the heavy flow of notifications. The forty-two families who received a letter in lieu of any other intervention tended to be those perceived as ‘low-level’ and most had no or low previous involvement with children’s social services. Templates of letters used for this purpose were available in both sites, although practitioners reported that these were tailored to the needs of par- ticular situations and families. Letters seen by the researchers included information about relevant services, but the tone was invariably threatening or warning and the language was sometimes technical and obscure, as illus- trated by the examples below taken from templates designed for cases that evoked low or high levels of concern:

This referral does not meet the threshold for the directorate’s involvement but you need to be aware that this information is now held on our records in respect of this incident.

Low Level of Concern Letter

Children’s Services have been notified by the police of an incident that occurred at your family home which falls under Domestic Violence proto- cols . . .. As you may be aware, the emotional impact of domestic violence is well researched and can be extremely harmful . . . any further reports of this nature will likely result in a home visit being completed in order to discuss the next appropriate course of action.

High Level of Concern Letter

302 Nicky Stanley et al.

While in one site, letters were addressed to both parents if they were living together, in the other, the letter was usually addressed to the mother. Receipt of such a letter may have the effect of triggering further abuse (Humphreys et al., 2001; Stanley and Humphreys, 2006). Some social workers were aware of this risk and described contacting victims by other methods but they noted that notifications rarely provided them with victims’ telephone numbers. Where the perpetrator was not resident, letters were usually sent to the mother. Such messages can convey unfeasi- ble expectations (particularly if the couple have already separated) and can act to allocate the victim responsibility for controlling her partner’s abusive behaviour (Farmer, 2006; Hester et al., 2006). Practitioners were aware of how these letters might be received:

I’ve had a phone call in the past where the woman I had written to was quite . . . frustrated with the letter . . .. Because clearly she . . . had tried very hard to keep her child safe and felt that it was the husband or the ex-partner’s behaviour, that he should be the one that we should be addressing (Initial Assessment Worker 3).

In contrast, some social workers considered that these warning letters might be appropriate in some situations:

Sometimes it is appropriate because . . . women particularly are notoriously good at moving from one abusive relationship to another, and sometimes, unfortunately, children are witnessing and do get caught up in the flack, and in those cases it is appropriate to say to Mum, ‘look, you know, you can make your decisions but children they don’t . . ., they have to sort of go along with everything else. This isn’t fair to them, you know’ (Initial Assessment Worker 2).

However, the re-referral rate over the subsequent twenty-one months showed that these letters were ineffective as a means of managing demand on the service. Fifty-five per cent of families receiving letters returned to the attention of children’s services—slightly more than the 52 per cent of the ‘no further action’ group who were re-referred. Twenty-one families, nearly half of those re-referred (most of these were repeat

Table 3 Children’s social services’ subsequent involvement over twenty-one months with no further action families in January 2007 sample

Level of subsequent involvement by children’s social work services Number Percentage

None 52 47 Low—re-notified/re-referred: NFA/letter/phone call/visit 39 35 Medium—re-notified/re-referred: initial assessment

followed by family support or NFA 16 14

High—re-notified/re-referred: Section 47 assessment/ safeguarding intervention

3 3

Unknown 1 1 Total 111 100

A Stop – Start Response 303

notifications), were again provided with a minimum-level response, which could be either ‘no further action’, another letter or, rarely, a telephone call or visit.

Cases receiving assessment or intervention

As noted above, most of the twenty-eight cases that received an assessment or intervention were already open to children’s social services at the time of notification. Those nine cases not already open were distinguished by the fact that the notification contained information about injuries either to the child or the victim. A further nine of these twenty-eight cases received assessment or intervention at the family support (Section 17) level. Prac- titioners noted that they frequently ended up working with families on a safeguarding basis because families struggled to acknowledge and disclose domestic violence. Such difficulties were attributed to the stigma surround- ing disclosure of hidden abuse and to fears of having children removed, but also to the fact that abusive behaviour was so heavily embedded in some families’ culture that it went unacknowledged:

. . . it’s very difficult when you have a professional coming into your home saying, ‘well actually, you are not parenting your child’, and if that’s what they have always done, they have always been brought up with, Dad’s done it and Granddad’s done it, they say, ‘Why are you jumping all over me?’ (Child Protection Social Worker 1)

The interviews undertaken with survivors and perpetrators confirmed the difficulties that parents experienced in acknowledging the impact of dom- estic violence on children. Parents varied in their capacity to recognise that children had been exposed to abusive behaviour and might be affected by it. While some perpetrators argued that ‘he was upstairs when this was going on’, others were more ready to accept that children were affected by what they witnessed:

My little girl’s told me, it scares her, she’s scared when we argue (Mark, perpetrator).

McGee’s (2000) study of the views of mothers and children in refuges found that acknowledging the extent of children’s exposure to domestic violence often only happened after women had left a violent relationship. Guilt accompanied this recognition for some survivors in our sample:

I think that was part of the breakdown as well, ’cos I was watching my chil- dren suffer . . . and I felt guilty, then guilty inside and I’m thinking why am I letting them go through this? But, at the time I couldn’t find a way out (Pearl, survivor).

The use of the booster sample to augment the cohort of cases receiving an intervention allowed for patterns of intervention to be studied in forty-six cases. Most of these cases were already open to children’s services. The

304 Nicky Stanley et al.

immediate response to the fourteen cases that were not already open is shown in Table 4.

Over the following twenty-one months, six of the fourteen families did not return to the attention of children’s services. For some families, the original intervention had acted to connect them with relevant services. However, the other eight families returned to the attention of children’s services, usually through notifications of domestic violence incidents and, in some cases, multiple notifications were recorded. Five of these re-notifications received a ‘no further action’ response that appeared in most cases to be influenced by the fact that the couple had separated and mother and children were perceived to be ‘safe’. The family in this triggered group that received the most intensive intervention was given numerous assessments and is shown in Example 1.

Example 1

This family with three children, aged two, six and sixteen, had a history of domestic violence and the father had a history of drug and alcohol abuse. The children witnessed the January 2007 incident where the father punched and assaulted the mother and verbally abused her. An initial assessment was instigated but closed because father received a prison sen- tence and mother was engaged with domestic violence services. Later in 2007, an anonymous referral informed children’s social services that the father was living back at home. This triggered a core assessment and support was offered. The family did not engage with services but, sub- sequent to a child protection case conference, the family did engage with voluntary sector family support services; however, there was another dom- estic violence incident following which the mother required medical treat- ment. The two youngest children were put on the Child Protection Register for seven months and the parents continued to be supported. After the case was closed in 2008 there was one further notification when the father threatened to burn down the house with mother in it: this notifi- cation was stamped ‘no further action’ (Incident/Case No. 101).

Interventions in the thirty-two cases already open at the point at which the notification was received were also examined and the pattern of repeated assessments seen in the case in Example 1 was identified as characteristic. Substance abuse or mental health problems were found in one or both

Table 4 Initial response to cases triggered by notification in augmented sample

Initial response Number

Initial assessment only 4 Section 47 enquiry only 8 Section 47 enquiry and package of support services 2 Total number of cases not previously open to social services 14

A Stop – Start Response 305

partners in most cases in this group; such difficulties were much more likely to be found in these families than in those that did not receive an interven- tion. Cases tended to be closed at the point at which couples separated (in two cases, reports of a separation proved later to have been misleading), the perpetrator was imprisoned, was seen to address his substance misuse pro- blems or, in some cases, the child or children moved to live elsewhere. The case in Example 2 encapsulates this pattern of intervention.

Example 2

The police notification in February 2007 followed the perpetrator breaching bail conditions (imposed as a consequence of domestic violence during pregnancy) which prevented him contacting or approaching his partner. The notification form noted that the victim was pregnant and very scared and that three children (aged between 20 months and 13 years) were present at the incident and distressed. A Section 47 enquiry was undertaken with a joint visit from the police and social workers; the victim said that she was no longer in a relationship with the perpetrator. The health visitor stated she had placed a ‘cause for concern’ in the youngest child’s file. The case was closed as mother was ‘protecting herself’; two months later an initial assessment was undertaken following another notification and the couple stated they had never in fact separated. A family support service was offered to the mother, the perpetrator was seeing a probation officer. The case was closed in December 2007 because the mother was con- sidered not to have engaged with support services. The family was notified again in April 2008 when a letter was sent (Incident/Case No. 267).

Eleven cases in this group of ‘already open’ cases had a child protection plan in place. Domestic violence appeared to have been central to the decision to allocate the case for eight of these families, suggesting that dom- estic violence was acknowledged as a risk factor. In eight families in this group of ‘already open’ cases, children were placed away from home in the twenty-one months subsequent to the notified incident. Homelessness was an additional feature of these families, which may have contributed to this outcome.

Table 5 shows that when all forty-six cases that received an intervention are examined together, children were living away from home in just over a fifth of families at the point of follow-up.

Since social workers identified the extent of families’ engagement with services as key to achieving change, the relationship between outcomes at follow-up and families’ engagement with both supportive services and chil- dren’s social services was examined. Families in which children were living at home at the point of follow-up in 2008 were those in which mothers were most likely to have engaged either fully or partially with a range of suppor- tive services. Ten of the sixteen mothers in the group in which children remained living at home with their mothers had engaged fully or partially with services, while nine of the seventeen mothers in those families in

306 Nicky Stanley et al.

which children remained living at home with both partners had engaged fully or partially with services.

Fathers/partners were substantially more likely to have been involved with support services in those cases in which children remained at home with both partners. Eleven of the fathers in families in which children remained living at home with both partners had engaged fully or partially with support services compared with one father who had partially engaged with services in the group of families in which children were living at home with their mothers. However, it cannot be assumed that chil- dren remained living at home because fathers/partners engaged with ser- vices; an alternative explanation is that where fathers/partners were motivated to engage with services, they were also motivated to stay in the family home and to change their abusive behaviour. Those families in which the children were living away from home in 2008 were characterised by the fact that neither parent had engaged fully with services.

Social workers’ engagement with perpetrators

Findings regarding perpetrators’ engagement with services raise the ques- tion of the extent of social workers’ engagement with them, which has been queried elsewhere in the literature (Scourfield, 2003; Featherstone and Peckover, 2007). Analysis of recorded social work practice in the forty- six cases receiving an assessment or intervention addressed the issue of which family members were the focus of social workers’ attention. In forty- three cases, there was evidence that social workers talked directly to chil- dren or, in the case of younger children, had observed and interacted with them. Mothers were the focus of intervention in the majority (forty- four) of cases. However, social workers were less likely to engage with fathers or male partners who were the perpetrators of domestic violence in forty-three of the forty-six families in this group. Evidence of social workers talking directly with fathers/male partners was found in twenty- nine cases (63 per cent). The extent of this interaction varied considerably: in some cases, it was minimal; in others it was extensive.

Table 5 Family outcomes for forty-six intervention cases twenty-one months after original notifica- tion, January/February 2007

Family situation Number (%)

Children living at home with mother and partner/father 17 (37) Children living at home with mother 16 (35) Children living away from home 10 (22) Children living at home with father 3 (6) Total 46 (100)

A Stop – Start Response 307

The interviews with practitioners confirmed this pattern of intervention with perpetrators. Initial assessment team workers were much less likely to see themselves engaging with male perpetrators than their counterparts in safeguarding teams:

As a general rule, I personally don’t ever get involved with the perpetrator. Not at the time that the domestic violence has gone on (Initial Assessment Family Worker 2).

This lack of engagement with male perpetrators was attributed to a number of factors, including concerns about staff safety, the accessibility of men in the limited period of seven days within which initial assessments had to be completed, the extent of their involvement with the children and the lack of available services to offer them. However, practitioners in safeguarding teams were more likely to work with perpetrators, especially if they contin- ued to be part of the household:

I’ve heard it said we don’t work with perpetrators in social work and I struggle with that really, you know, and I don’t think you can ever say we don’t work with perpetrators of anything, if they’re part of the family unit and if that risk can be managed and if that person is open to change (Child Protection Manager 2).

When work was undertaken with perpetrators, it entailed including them in assessments and referring them to other services such as anger management programmes, counselling or substance misuse services. Practitioners con- sistently identified a shortfall in services working with perpetrators of dom- estic violence on a voluntary basis (as opposed to the probation-run courses mandated by the courts) and argued that this gap in provision contributed to pressure on mothers to protect children from domestic violence:

Not for perpetrators, no, not at all. There are no services out there to sort of try and minimise the risk. As far as children’s social care goes, it is a case of saying to Mum: ‘it’s your job to protect and you will have to get on with it’ (Family Support Manager).

Discussion

This study has a number of limitations: a proportion of cases was lost in the process of tracking families from police to children’s services records and this may have impacted on the extent to which the sample was representa- tive. The sample was collected retrospectively and this, together with the mobility of the workforce, meant that it was not possible to interview prac- titioners about cases included in the sample; instead, their views on practice generally in relation to families experiencing domestic violence were cap- tured. However, the use of a retrospective sample allowed for cases to be followed up over time and for outcomes to be examined.

308 Nicky Stanley et al.

The low proportion of notified cases receiving a social work service is clearly cause for concern and practitioners and managers participating in this study acknowledged this problem. Some children’s services depart- ments have attempted to address this by using letters in the absence of any other intervention. However, the impact of such letters on family dynamics cannot be controlled and this study has shown their ineffective- ness in reducing the demands on children’s social work services: families that received a letter but no further intervention were just as likely to be re-notified or re-referred as those who followed a no further action pathway. Letters alone did not therefore seem to reduce children’s experi- ence of domestic violence and it is doubtful whether the threatening tone of some of these letters contributed to families’ willingness to disclose the extent of domestic violence and to engage with services. Since social workers identified families’ readiness to acknowledge the domestic vio- lence and to engage with children’s services as key factors shaping the pattern and level of intervention, it seems important that nothing is done to reduce motivation to contact and use services.

Follow-up of those cases that did receive an intervention over a period of twenty-one months revealed a ‘stop – start’ pattern of social work interven- tion in which families were assessed, perhaps referred or signposted to other services and cases were closed, then re-opened and re-assessed following another notification or referral. When families’ histories were examined both prior to the original notification and afterwards, this ‘stop – start’ approach was discernible in over a third of the forty-six cases receiving a service. In some cases, it appeared that social workers were too readily assured of children’s safety by the fact that the perpetrator was reported to have left the household. Since it is well established that domestic violence frequently continues or may even accelerate following separation—over half the sample incidents involved perpetrators who had left the household—it is hard to justify this definition of safety.

Some of the social workers and managers interviewed were aware of this stop – start pattern of intervention and saw it as a consequence of the shift in their role to assessing and case managing families rather than providing ser- vices directly themselves (see Parton, 2009):

. . . the government want us to go in, assess and get out again, they only want us to be involved in statutory cases (Senior Manager 2).

This ‘stop – start’ approach appeared inadequate to effectively address the complex web of long-term problems that often involved mental health pro- blems and/or substance misuse and for some families, behavioural or mental health problems in children as well as long-established patterns of domestic violence. Managers highlighted the need for longer-term involve- ment with such families:

A Stop – Start Response 309

. . . these families need a lot longer, more targeted over time work because it’s years and years and years that you’d need to get to them (Child Protec- tion Manager 1).

A Californian study (Jones et al., 2002) identified a similar pattern of recy- cling families experiencing domestic violence through child protective ser- vices. Cleaver and Nicholson’s (2007) study of social work intervention with parents with learning disabilities has drawn attention to the need for longer- term, low-level support and monitoring of such families. Domestic violence is similarly a long-term problem, often deeply embedded in relationships, families and communities. Change in such behaviour is not achieved by short-term interventions and the harm done to children accumulates over time (Rossman, 2001) and extends to include children’s relationships with their mothers (Humphreys et al., 2006).

While some social workers interviewed for this study, notably those working in initial assessment teams, were reluctant to engage with male per- petrators, others argued that since many victims did not want to leave violent relationships, work with perpetrators needed to be a part of the sol- ution. Previous research has identified men’s inaccessibility to social work and social work’s preoccupation with women as carers as well as perpetra- tors’ threatening or intimidating behaviour as factors underpinning this failure to engage with perpetrators of domestic violence (Stanley, 1997; Littlechild and Bourke, 2006). However, this study highlights the structural constraints imposed by the requirement in England and Wales (DES, 2006) to complete an initial assessment within seven days. This, in combination with men’s reduced levels of accessibility, acted to exclude male perpetra- tors from assessment. Broadhurst et al.’s (2010) study found structures for assessment similarly restrictive and identified a comparable pattern of repeated assessments. The National Safeguarding Delivery Unit’s (2009) proposal to extend the timescale for an initial assessment to ten days suggests increasing acknowledgement of the limitations of the current requirements.

Conclusion

While statutory social work has acknowledged the presence of domestic violence in families and the risk of harm it poses for children, this study demonstrates that defining exposure to domestic violence as a form of sig- nificant harm is not on its own sufficient to ensure that children and families receive support. In the absence of any additional allocation of resources to respond to families living with domestic violence, the majority of families notified to children’s services are failing to meet the threshold for interven- tion. Many of these families may not require the involvement of statutory services (Edleson, 2004), but the re-notification of a high proportion of

310 Nicky Stanley et al.

these families indicates that children in these families continue to be exposed to domestic violence.

For those families that do receive a service response, social work inter- vention has to acknowledge the fact that domestic violence is a hidden and stigmatised experience and that the barriers to disclosing it and seeking help are considerable. Conveying threatening messages to families through bureaucratic letters is unlikely to contribute to dissolving such bar- riers. Social work also needs to acknowledge that separation has limitations as a means of ensuring children’s safety. It may not be an option for all families and separation is not an assurance of safety. While they may be invisible to social workers, many fathers do not conveniently ‘disappear’ from families and social work therefore needs to engage with male perpe- trators more fully; practitioners might benefit from training to build skills and confidence in this respect.

However, this study also raises questions about the extent to which the current structure of assessment supports intervention with families with long-term entrenched difficulties such as domestic violence. Repeated assessments in response to recurring notifications will not contribute to building the consistent and trusting relationships between social workers and families that promote disclosure of need and engagement with relevant services. If social work is to act effectively to reduce the harm domestic vio- lence inflicts on children, structures of assessment in England and Wales will need to be revisited and rethought.

References

Banks, D., Landsverk, J. and Wang, K. (2008) ‘Changing policy and practice in the child

welfare system through collaborative efforts to identify and respond effectively to

family violence’, Journal of Interpersonal Violence, 23(7), pp. 903 – 32.

Beeman, S. K., Hagemeister, A. K. and Edelson, J. (2001) ‘Case assessment and families

experiencing both child maltreatment and woman battering’, Journal of Interpersonal

Violence, 16(5), pp. 437 – 58.

Broadhurst, K., Wastell, D., White, S., Hall, C., Peckover, S., Thompson, K., Pithouse, A.

and Davey, D. (2010) ‘Performing “initial assessment”: Identifying the latent con-

ditions for error at the front-door of local authority children’s services’, British

Journal of Social Work, 40(2), pp. 352 – 70.

Buckley, H., Holt, S. and Wheelan, S. (2007) ‘Listen to me! Children’s experiences of

domestic violence’, Child Abuse Review, 16(5), pp. 296 – 310.

Charmaz, K. (2006) Constructing Grounded Theory, London, Sage.

Cleaver, H. and Nicholson, D. (2007) Parental Learning Disability and Children’s Needs:

Family Experiences and Effective Practice, London, Jessica Kingsley.

Cleaver, H., Nicholson, D., Tarr, S. and Cleaver, D. (2007) Child Protection, Domestic

Violence and Parental Substance Misuse: Family Experiences and Effective Practice,

London, Jessica Kingsley Publishers.

A Stop – Start Response 311

Cleaver, H., Unell, I. and Aldgate, J. (1999) Children’s Need: Parenting Capacity, the

Impact of Parental Mental Illness, Problem Alcohol and Drug Use, and Domestic Vio-

lence on Children’s Development, London, The Stationery Office.

Department for Education and Skills (DES) (2006) Common Assessment Framework for

Children and Young People: Practitioners’ Guide. Integrated Working to Improve Out-

comes for Children and Young People, London, Department for Education and Skills.

Devaney, J. (2008) ‘Chronic child abuse and domestic violence: Children and families

with long-term and complex needs’, Child and Family Social Work, 13(4), pp. 443 – 53.

Edleson, J. L. (1999) ‘Children’s witnessing of adult domestic violence’, Journal of Inter-

personal Violence, 14(8), pp. 839 – 70.

Edleson, J. L. (2004) ‘Should childhood exposure to adult domestic violence be defined

as child maltreatment under the law?’, in Jaffe, P., Baker, L. and Cunningham, A.

(eds), Protecting Children from Domestic Violence: Strategies for Community Inter-

vention, New York, Guildford Press.

English, D. J., Edleson, J. L. and Herrick, M. E. (2005) ‘Domestic violence in one state’s

child protective caseload: A study of differential case dispositions and outcomes’,

Children and Youth Services Review, 27, pp. 1183 – 201.

Farmer, E. (2006) ‘Using research to develop practice in child protection and child care’,

in Humphreys, C. and Stanley, N. (eds), Domestic Violence and Child Protection:

Directions for Good Practice, London, Jessica Kingsley.

Farmer, E. and Owen, M. (1995) Child Protection Practice: Private Risks and Public

Remedies, London, HMSO.

Featherstone, B. (2009) Contemporary Fathering: Theory, Policy and Practice, Bristol,

Policy Press.

Featherstone, B. and Peckover, S. (2007) ‘Letting them get away with it: Fathers, dom-

estic violence and child welfare’, Critical Social Policy, 27(2), pp. 181 – 202.

Galvani, S. (2004) ‘Responsible disinhibition: Alcohol, men and violence to women’,

Addiction Research and Theory, 12(4), pp. 357 – 71.

Gilchrist, E., Johnson, R. T., Weston, S., Beech, A. and Kebbell, M. (2003) Domestic Vio-

lence Offenders: Characteristics and Offending Related Needs, Findings 217, London,

Home Office.

Hester, M., Pearson, C., Harwin, N. and Abrahams, H. (2006) Making an Impact—

Children and Domestic Violence: A Reader, 2nd edn, London, Jessica Kingsley

Publishers.

Holt, S., Buckley, H. and Whelan, S. (2008) ‘The impact of exposure to domestic violence

on children and young people: A review of the literature’, Child Abuse & Neglect, 32,

pp. 797 – 819.

Humphreys, C. (2007) Domestic Violence and Child Protection: Challenging Directions

for Practice, Issues Paper 13, Australian Domestic & Family Violence Clearinghouse,

Sydney, University of New South Wales.

Humphreys, C., Mullender, A., Lowe, P., Hague, G., Abrahams, H. and Hester, M.

(2001) ‘Domestic violence and child abuse: Developing sensitive policies and gui-

dance’, Child Abuse Review, 10(3), pp. 183 – 97.

Humphreys, C., Thiara, R., Mullender, A. and Skamballis, A. (2006) ‘Talking to Mum’,

Journal of Social Work, 6(1), pp. 53 – 63.

Irwin, J. and Waugh, F. (2007) ‘Domestic violence: A priority in child protection in New

South Wales, Australia?’, Child Abuse Review, 16(5), pp. 311 – 22.

Jones, L. P., Gross, E. and Becker, I. (2002) ‘The characteristics of domestic violence

victims in a child protective service caseload’, Families in Society, 83, pp. 405 – 15.

312 Nicky Stanley et al.

Lapierre, S. (2008) ‘Mothering in the context of domestic violence: The pervasiveness of

a deficit model of mothering’, Child and Family Social Work, 13, pp. 454 – 63.

Littlechild, B. and Bourke, C. (2006) ‘Men’s use of violence and intimidation against

family members and child protection workers’, in Stanley, N. and Humphreys, C.

(eds), Domestic Violence and Child Protection: Directions for Good Practice,

London, Jessica Kingsley Publishers.

McGee, C. (2000) Childhood Experiences of Domestic Violence, London, Jessica Kings-

ley Publishers.

Mullender, A., Hague, G., Iman, U., Kelly, L., Malos, E. and Regan, L. (2002) Children’s

Perspectives on Domestic Violence, London, Sage.

National Safeguarding Delivery Unit (2009) Safeguarding Indicators and Statutory

Targets: Consultation Paper, London, DCSF.

Parton, N. (2009) ‘Challenges to practice and knowledge in child welfare social work:

From the “social” to the “informational”’, Children and Youth Services Review,

31(7), pp. 715 – 21.

Rossman, B. (2001) ‘Longer term effects of children’s exposure to domestic violence’, in

Graham-Bermann, S. and Edleson, J. (eds). Domestic Violence in the Lives of Chil-

dren: The Future of Research, Intervention and Social Policy, Washington, DC, Amer-

ican Psychological Association.

Scourfield, J. (2003) Gender and Child Protection, Basingstoke, Palgrave/Macmillan.

Stanley, N. (1997) ‘Domestic violence and child abuse: Developing social work practice’,

Child and Family Social Work, 2(3), pp. 135 – 46.

Stanley, N. and Humphreys, C. (2006) ‘Multi-agency and multi-disciplinary work: Bar-

riers and opportunities’, in Humphreys, C. and Stanley, N. (eds), Domestic Violence

and Child Protection: Directions for Good Practice, London, Jessica Kingsley.

Stanley, N., Miller, P., Richardson Foster, H. and Thomson, G. (2010) Children and

Families Experiencing Domestic Violence: Police and Children’s Social Services

Responses, London, NSPCC, available online at www.nspcc.org.uk/Inform/

research/Findings/children_experiencing_domestic_violence_wda68549.html

(accessed 30 April 2010).

Strauss, A. and Corbin, J. (1990) Basics of Qualitative Research: Grounded Theory Pro-

cedures and Techniques, Newbury Park, CA, Sage.

A Stop – Start Response 313

Copyright of British Journal of Social Work is the property of BASW Trading Limited and its content may not

be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.