The pursuit of standardization in domestic violence social work: A multiple case study of how the idea of using risk assessment tools is manifested and processed in the Swedish social services

Abstract

This multiple case study examines how the idea of using risk assessment tools is manifested and processed in Swedish social services. Based on the analysis of interviews with different stakeholders and of organizational documents in two social service organizations, we investigate the actors who control local risk assessment practices. The findings illustrate that a relatively small group of social workers in the organizations have been able to forward their claims and decide how risk assessment work should be carried out without much intrusion from local managers or politicians. The findings also validate other studies that found that increased standardization can strengthen social workers’ ability to perform their professional task rather than lead to de-professionalization. This article ends with a discussion of what risk assessment practices might mean for domestic violence victims.

Keywords

Domestic violence risk social work practice risk assessment jurisdiction control

Introduction

It is argued that professions are under siege (Ahlbäck Öberg et al., 2016). Professionals, such as doctors, teachers, and social workers, in the Swedish public services face external demands orchestrated by the state authorities to become more transparent and effective. If variation in practices and professional discretion has been viewed as the problem, standardization as a way to construct uniformities across time and space has been viewed as a solution (Timmermans and Epstein, 2010). Increased standardization of assessment and treatment work in public services reflects, it is argued, distrust in professionals and thus a breach in what is called the “professional contract” between the state and the professions (Ahlbäck Öberg et al., 2016). Another factor pushing towards standardization is the growing dominance of managing risk and the idea that the uncertainty of future events can be managed through evidence-based practice, and standardized risk assessment tools which are evident in many countries (Ponnert and Svensson, 2016). Detractors claim that this might increase tensions between different stakeholders and ultimately lead to a de-professionalized and alienated workforce (Ahlbäck Öberg et al., 2016; Lauri, 2016).

The way standardization affects social workers cannot, however, be prejudged (Evans, 2010); standardization might lead to positive professional outcomes (Robinson, 2003) and be used as a professionalization strategy (Barfoed and Jacobsson, 2012). Standardization is not carried out by its own power, but in organizational contexts by actors with intentions and expectations who diffuse, adopt, and adapt to standardization reforms for various reasons (Greenhalgh et al., 2004). In the wake of New Public Management (NPM) inspired technologies, formerly subjugated professionals have advanced and strengthened their positions (Hall, 2012). In this article, we will focus on the link between professionalism and standardization. Using a multiple case study analysis, we study how the idea of using risk assessment tools is manifested and processed in the social services work with domestic violence.

The responsibility to support victims of domestic violence has previously been a concern mostly for civic society. Clearly, public services, rather than civic society solely, need to engage in and take greater responsibility for combating domestic violence (Ekström, 2016). It is, however, an open question which actors in social service workplaces are successful in negotiating and claiming jurisdiction (Abbott, 1988) and thus controlling how work in this area is played out locally, an area of research that has received little previous attention (Kropp, 2008). Are domestic violence social workers also under siege—”entirely in thrall to the hegemony of the internal regime rather than a professional task” (Wastell et al., 2010: 319) becoming a deskilled ticker of boxes (Fitzgibbons, 2008; Gillingham, 2011)? Or could it be that the standardization trend and risk tools used by domestic violence social workers support their work?

Our research has been guided by the following research questions:

Which actors establish and control the use of assessment tools for risk assessments in the work with domestic victims in local social service contexts?

How is risk assessment practices regarding use of assessment tools filled with content? That is, how and on what grounds are the tools chosen?

Which mechanisms influence the consolidation of risk assessment tools in the social services?

The Swedish context for domestic violence risk assessment work

During the last decade, the Swedish government has worked actively to combat men’s violence against women (SOU 2014:49). New regulations and guidelines from the National Board of Health and Welfare (Socialstyrelsen, NBHW) have been developed and risk assessments are mandatory when clients are exposed to domestic violence. Specialized units and positions in the social services have emerged that focus either exclusively or partially on domestic violence (Ekström, 2016). Arguably, some progress has been made that conflict with previous notions of a client category that has not gained acceptance in the Swedish welfare system (Ljungwald and Svensson, 2007).

Differences still exist between municipalities regarding priorities, organization, staffing, resources, competencies, and routines, resulting in a considerable variance in capabilities and outcomes (SOU 2014:49). Legislation is not always followed and serious flaws exist when it comes to acknowledging and investigating needs of women and children experiencing and witnessing violence. Although risk tools have been put forward to deal with the problems, they are not used to a relevant extent by social workers according to a recent audit (IVO, 2014). Internationally, similar findings show how clients’ disclosures of domestic violence are assessed poorly or not at all by social workers (Lindhorst et al., 2010), that social workers lack relevant screening skills (Danis, 2003; Lindhorst et al., 2008), and that standardized risk assessment tools are used only to a limited extent in practice (Cattaneo and Chapman, 2011; Danis, 2003; Forgey et al., 2014). It has been argued that the management of social services needs to take greater responsibility in the implementation of domestic violence services as it is unrealistic for social workers to find the most reliable and valid services for their clients (Forgey et al., 2014), and social workers fear that the assessment tools they have access to may be harmful to their clients (Cattaneo and Chapman, 2011).

The NBHW are in charge of developing FREDA assessment tools used by social services. The regional County administrative boards (Länsstyrelsen) have organized several short courses that educate social workers about FREDA. Included in FREDA is the Danger Assessment (DA), widely used by many nations and by various professions (Messing and Thaller, 2015). In addition, the Spousal Assault Risk Assessment Guide (SARA) is also frequently used (Kropp, 2008) in the Swedish social services to assess risk. The DA in FREDA is an actuarial tool whereas SARA is a structured professional assessment tool (Kropp, 2008). Regardless of the model (actuarial or structured professional judgment), they serve as examples of the increasing standardization of social work practice.

In addition to the NBHW and the County Administrative Boards, other significant players are university researchers, private companies that support implementation of assessment tools and provide education about violence, as well as shelters for women and other NGOs. We now find a mosaic of activities and actors involved at the national and regional levels contributing to the way risk assessment practices unfold locally.

Conceptual framework

Theoretically, our analysis is influenced by sociological writings on the professions, and recognition that experience at the “street level” is important to scrutinize when national top–down imperatives meet everyday practices (Evans, 2010). Significant here are the concepts of discretion and jurisdiction (Abbott, 1988; Evans, 2010; Freidson, 2001). Whereas discretion refers to one’s autonomous use of training, knowledge, and skills with few external decrees when making judicious decisions (Freidson, 2001), jurisdiction concerns the link between a profession and its work (Abbott, 1988). Our primary focus is jurisdiction. Jurisdiction concerns who has the control over a set of tasks. Traditional professional control implies that work processes and tasks conform to the standards set out by the profession and that these standards can be used with discretion and judgment by each professional. It also deals with how problems are defined and understood. This control is continually negotiated with other professions as well as with other stakeholders such as politicians and managers (Kouzes and Mico, 1979). According to Abbott, such negotiations are played out on three arenas: in the legal system, the arena of public opinion, and at the workplace. We focus on the workplace where claims for jurisdiction must compromise with organizational requirements for efficiency and rationality (Abbott, 1988) and sometimes with divergent and ambiguous goals (Lipsky, 2010). Such claims might also, over time, be renegotiated in ways that change the patterns of jurisdiction.

In a society cast in NPM, it would be hard to claim that professionals’ control over the principles and procedures of their work have not changed (Ahlbäck Öberg et al., 2016). When NPM is paired with a view of social work as lacking a solid knowledge base, poor practitioner decision-making, and unclear or even ineffective interventions and results (SOU 2008:18), it might make sense that social work is fertile ground for standardization reforms and a reduction of professional discretion. The construction and diffusion of assessment tools and other standardized manuals can be viewed as antidotes whereby social workers can access potential legitimate knowledge sources, providing them with a codified vocabulary that corresponds to a new professional style (Barfoed and Jacobsson, 2012) or a new professionalism (Evetts, 2010). It is, however, important to acknowledge that professions such as social workers are not uniform groups (Abbott, 1988). Specialization may evolve as new risks and new clients are established, and related to these new categories of experts or elite professionals might emerge with ambitions to govern and control the conduct of the rank and file.

Method

This article is part of a study about the mechanisms influencing interventions contesting domestic violence. The research examines the prevalence of interventions and how different forms of governance—non-profit organizations, government agencies, and local social work—affect interventions in a time of evidence and NPM. The empirical material uncovered in this article derives from a multiple case study where social service organizations from two Swedish municipalities were strategically selected. The choice of working with case studies was based on our interest in how actors reflect and act according to preconditions at the workplace (Yin, 2009).

The two cases had distinguished themselves for being driving forces in the development of social work for abused women. Their specialist social workers had come to be recognized as competent and experienced risk assessors in their respective organizations as well as across the country. As both organizations had been active in contesting domestic violence for over a decade, we hypothesized that the settings would provide us with material where negotiations over how risk assessments should be enforced had breached changing organizational preconditions. Differences between the two organizations exist.

In organization 1, a specialized unit addresses all aspects of domestic violence. The unit conducts investigations and assessments as well as provides treatment to victims and offenders. Clients apply for support directly to the unit or are referred to the unit by other professionals from inside or outside the organization. The second organization (organization 2) has also a specialized unit, but it works solely with treatment of victims and offenders and conducts risk assessments. Clients are provided support anonymously and they can apply for support directly or be remitted for treatment or risk assessment by other professionals. Here the social assistance unit also conducts investigations and are also more operative (e.g., acute placements of women in shelters). At both the specialized units, there are experienced social workers who have cooperated for several years. The primary research method was to interview stakeholders (Table 1) and to examine organizational documents (e.g., formal routines, applications for government funding, activity reports, etc.). Within the scope of this article, we have conducted interviews with 30 participants.

Table 1.

Sample.

Organization 1	Coordinator (n = 1, interviewed twice) Social workers, specialized unit (n = 3) Social worker, social assistance unit (n = 1) Head of department (n = 1) First line managers (n = 4) Politician (n = 1)
Organization 2	Coordinator (n = 2) Social workers, specialized in violence (n = 2) Social worker, social assistance unit (n = 4) Head of department (n = 1) First line managers (n = 3) Politician (n = 2)
Regional level	Coordinators, domestic violence (n = 3)
SALAR	Official (n = 1)
National Board of Health and Welfare	Researcher (n = 1)
N	30

SALAR: Swedish association of local authorities and regions.

The inclusions of managers and politicians were based on the premise that the social services include three domains with specific governing logics (Kouzes and Mico, 1979). It is assumed that control over how work should be carried out in human services are often governed by bureaucratic and managerial logics rather than professional logics as a result of various public sector reforms (Ahlbäck Öberg et al., 2016). The inclusion of actors from these domains enabled us, paired with the documents, to validate accounts made by social workers. In addition to the local actors, we have interviewed stakeholders at the regional and national levels who have specific interest and responsibilities when implementing risk assessment tools in the social services. Their voices have helped us understand how our cases are part of a larger institutional context.

Interview guides were adapted to correspond to the interviewee’s positions (manager, social worker, politician, coordinator, official, etc.). As our research interest in the project is broader than risk assessment practices, questions covered various aspects of the social services work with violence, including the following themes: history and emergence of work with violence locally, regionally, and nationally; organizing rationales; understandings, consequences, and handling of violence; examples of cases they have worked with; different work approaches and tools used; perceptions of control over work arrangements; and factors affecting how the organizations meet the needs of victims. All the interviews were transcribed.

Initially, our analysis of the material focused on the organization’s risk assessment stories. After carefully reading the interviews and documents, the first author constructed a timeline, noting events or situations that seemed significant for a thorough understanding of each case. The empirical material related to these events and situations were ordered in a coding matrix according to the following analytical questions: How have risk assessments been organized, by whom, and with what driving rationales? Interesting segments were underlined and labelled with orienting concepts (Layder, 1998) from the sociology of professions. These concepts included professionalism, technicality/indeterminacy, jurisdictional control, and discretion. Relevance and alternative interpretations of the material were discussed in the research group and at seminars (i.e., our community of inquiry). Following these discussions, the three themes presented below were derived in a further exchange between previous research, our orienting concepts, and the empirical material, making our analytical strategy close to that of abduction (Tavory and Timmermans, 2014).

The study conforms to the ethical standards for humanities and social science research in Sweden and was approved by the Ethical Review Board in Sweden.

Results

The following part of the article presents our analysis about how the idea of using risk assessment tools is manifested and processed in the two cases studied. The themes follow each other chronologically.

Conditions for establishment and control over the use of assessment tools for risk assessments

We found consensus across domains that government initiatives (such as strengthened regulations) and pressure from women’s shelters to take greater responsibility were important stimuli for acknowledging the problem of domestic violence around 2005/2006 in the two organizations. Our data highlights a conflation of disturbances in the institutional environment and local problem descriptions and needs, which reasonably would provide good conditions for practice development. However, the translation of general ideas—e.g., the imperative to conduct risk assessments—into concrete practices in the social services puts the issue of control over how this should unfold at the fore. A group of professionals were keen to annex this new and open area within their organizations, initially being successful in their jurisdictional claims without many perceived disputes.

Specific social workers in both organizations, often termed as enthusiasts, are significant. The chairman of the social welfare board in organization 1, for example, speaks of “a privilege to have an employee in the organization” with a special interest concerning domestic violence while a politician from organization 2 talks about how she “does not have a picture of it as a political initiative.” Rather it was, she said, a dialogue between the profession and the women’s shelters that started the development, the latter putting pressure on the social services to expand their services.

The individuals referred to as enthusiasts early were given positions as coordinators in the organizations with the aim to develop a consensus on how to work with domestic violence among different units as well as to establish good collaborative working conditions with the shelters. In addition, they were to start the creation and building units that would specialize in domestic violence within the organizations functioning alongside other specialized units such as child welfare and social assistance. Their influence on how the work should be carried out was regarded as high in both organizations. According to managers, they have been, together with their peers, relatively self-sustaining: “[T]here was no need for me to direct or tell them what to do, instead I have been able to listen and comment” (Head of Department, organization 2). Similar statements are found in organization 1:

The coordinator initiated it, applied for money to start projects quite a few years ago, and then she has been very active in applying for more money [state funds]… they have been self-funded like a small business. (Head of Department, organization 1)

The coordinator in organization 2 talked about some vague reflections about work-formation from the management and political domain and that it needed some sort of unit specialized in domestic violence:

Someone gets assigned and then it has to take care of itself, roughly speaking. We’ve [referring to the specialized unit] run everything ourselves, methods and how we work, what our approach is and all the practical matters. Doing risk assessments as requested from social workers or when we feel that it is needed. (Coordinator, organization 2)

Thus, at least at the local organizational level, there seems to have been perceptions of some sort of professional control over the establishment and development of the domestic violence work. However, it is not a matter of a larger collegial entity driving the work forward, but it could rather be interpreted as an “elite” of some kind, exemplified here by the coordinators and their closest colleagues. When a social worker in organization 1 talks about their coordinator as an “enthusiast” and “without her we wouldn’t be here,” one of his colleagues adds the following:

The national general guidelines and regulations are always there, of course. We have to evaluate and see if we are doing things right looking at the effects. I think the coordinator is a driving force and great, but we are a group [the specialized unit] and we also want to know what we are doing. We should not just bumble around, we should do good things, and I think that helps us to constantly push on and strive to be in the forefront.

Although the statement from the respondents from the different domains above is wider than just referring to risk assessment practices, they illustrate that in the shaping of practices institutional values that define a victim logic requires allocating resources and justifying actions in relation to domestic violence victims. These complexities are reflected in how the social workers view government mandates: “The national general guidelines and regulations are always there.” In addition, several social workers noted that importance of government funds. However, this work is also a bottom-up commitment as these professionals have relatively extensive control over both organization and work procedures and with ambitions to “not just bumble around.” Even if national governance is clear and uses regulations, economic means, and information to acknowledge domestic violence and place it on the political agenda, the implementation stage is ultimately bound to the municipalities. Here, local governance of risk assessment practices seems to be left to a relatively small group of professional without much local political or managerial interference in this regard. An interpretation is that as long as those specialists are initiating requirements that correspond to exogenous demands and thus reflect symbolic rationality in the organization, the translation of the risk assessment idea is ceded to them. In the following, we will describe how these professionals filled their practices with content.

Content in practice regarding assessment tools

One of the main strategies found in the empirical material is the benchmarking of others who are legitimate actors with legitimate working procedures in the field. In organization 1, this started in 2007 when they scanned the field after appropriate ways to work and took a liking in an integrated way of supporting victims. This approach would, according to our respondents, minimize the risk for fragmentation as too many parties were involved in supporting victims, a conclusion also acknowledged in previous research (Ekström, 2016). Victims of domestic violence experience powerlessness over authorities’ inability to coordinate services (SOU 2014:49). Thus, the social workers at organization 1 would do investigations, risk assessments, and provide treatment when necessary. Furthermore, inspired by other units, organization 1 implemented specific tools to assess risk and map violence. The following excerpt describes how this was pursued:

I felt we needed a benchmark. You cannot just go on your gut feeling. So I searched for things related to risk. Since I had been working for a long time with family law, I stumbled upon this man, who had designed a risk assessment with seven variables or questions to ask to judge the risk. We used this a while around 2007, but it still felt too vague. By chance, I saw an invitation for an event in [referring to a city]. I don’t know how I ended up on the webpage, but some man called NN was to come talk about the “safety program” and it felt like the right area. Three of us from here went to listen to him and we felt like this was it and we have to have this! It felt thorough, thought through; it [i.e. the safety program] was originally from Canada, where it was validated so there was a foundation. He had translated it and been in Canada for a year at Changing ways, so it felt solid somehow. (Coordinator, organization 1)

One interpretation of her statement is that the scanning of the field went by rather ad hoc. The use of the phrases “felt too vague,” “stumbled upon,” and “by chance” does not give the impression of a systematic scan but instead illustrates a kind of randomness. Somewhat paradoxically, the pursuit for replacing gut-feeling with a new professional style was initially based on “so it felt solid somehow.” However, concrete guidance regarding how risk assessment should be executed was at that time rather absent.

Their pursuit resulted in the implementation of tools associated with “the safety program” in Sweden. These tools included risk assessments, the DA, and to map the violence the Psychological Maltreatment of Women Inventory and Severity of Violence Against Women Scale. These tools were developed in other national contexts and in the case of DA also by the nursing profession. The three tools became very influential in developing FREDA, and the social workers in organization 1 were thus an attractive ally to co-opt when the NBHW developed FREDA: “We already did risk assessments according to FREDA; we just changed the name from the safety program to FREDA. It’s 90% the same” (Coordinator, organization 1). The social workers at the unit were part of a reference group at NBHW and provided “input from the field.” In association with the County Administrative Boards, two of the social workers trained more than 100 other municipalities on how to conduct risk assessments with FREDA. According to the officials at the County Administrative Boards, not only do these professionals control their own work content, but they also have “had much to say about how they want to launch it [nationally] because they have been part of it.”

In organization 2, the story of choosing tools took a different path. As in organization 1, it still is a task for the profession to choose tools. Indeed, statements from managers from both organizations were rather vague when talking about which tools were in play; in some cases, they did not even know which tools were used at the time of the interview. As mentioned earlier, it has mainly been the social workers at the specialized unit together with the coordinator that has driven the development of conduct risk assessments. As in organization 1, the social workers found that risk assessment needed to become more structured and standardized as the guidance to professionals was initially rather vague. The organizational action plan for domestic violence cases from 6 June 2005 instructed social workers rather vaguely on how to make risk assessments: “How critical is the threat? Develop an emergency plan with the woman so that she knows where she can turn if it becomes acute.”

An expert (from psychiatry) provided a lecture about risk and safety planning, and the social workers also received training. However, eventually they chose to opt out of both the DA (used previously) as well as FREDA and instead invested time and efforts to make SARA stick as a standard in the organization. The same expert taught them how to use SARA. He has been highly critical of the development of FREDA stating that “[w]omen could clearly die because of these follies” because one might get the impression that “a social worker does not need, or is even allowed, to think for themselves,” referring to the actuarial risk assessment method in FREDA (Belfrage, 2011).

Social workers at organization 2 were not as critical and had somewhat other motives for choosing SARA instead of FREDA. The final “nail in the coffin” for the possibility to use FREDA came after a meeting with a group of social workers from different units. The extent of FREDA (over 80 questions) could not be managed with social workers because there was limited time. This could mean that “the client is left alone in difficult and unprocessed emotions” and that “the danger assessment is felt inadequate and does not address the issues that need to be asked to make an adequate risk assessment for future violence” (meeting protocol, dated 080415). In addition, when they simulated a risk assessment situation and compared FREDA and SARA, they felt that the SARA assessment was more accurate, as confirmed by all the social workers we interviewed. Another rationale for using SARA is that it “is much easier to understand, we speak the same language as the police” (Social worker, organization 2).

Both cases above illustrate some of the processes and considerations when social workers choose risk assessment tools. This seems rather random and without a systematic and critical review of existing options and previous research. Furthermore, it seems as if the development of available risk tools is mainly in the hands of others rather than the professionals who will use them. That is, the development of DA and SARA are not primarily based on knowledge from social work research or the experiences from social workers. In other words, if standardized risk assessment practices are a way to forward jurisdictional claims and control work processes (Abbott, 1988), the knowledge legitimizing this, materialized in the assessment tools, is found outside the own profession. Furthermore, the choices made in organization 2 seem to be based also on an adaption to a language that corresponds with the language police use.

The social workers’ pursuit for standardization seems to be rooted in a conviction that there exist restrictions in their own “gut-feeling” and would thus rather work in accordance with this new professional style influenced by knowledge developed by other professions. However, they do not equate this with loss of discretion or a devaluation of their own professional knowledge. Rather, the latter, especially concerning violence, is, as we shall see, considered to be a prerequisite for using the tools.

Mechanisms influencing the consolidation of risk assessment tools

Moving from gaining control over construction and content of practices, we now focus on what happens when practices over time confronts with the everyday conditions of risk assessment work and organizational requirements for efficiency and rationality.

In organization 1, social workers at the specialized unit use FREDA to investigate cases, perform risk assessment, and treat domestic violence problems. FREDA is formalized in various types of local organizational rules and procedures, developed, not by management as usually assumed (Lipsky, 2010), but by the coordinator in agreement with social workers. The rules and procedures set the premises of how to act in cases where suspicion of or exposure to violence arises and it becomes evident that all roads lead to the specialized unit: “If I have indications that this is a case where we need to make a risk assessment, then yes, I call the coordinator” (Social worker, social assistance unit).

When used, FREDA seems to help social workers accomplish their tasks. The questions in FREDA help social workers pinpoint the needs of their clients: “[I]f I have conducted a Freda, I almost have the entire investigation” (Social worker, specialized unit, organization 1). With regards to the risk assessment, FREDA functions as a checklist that together with expertise about violence helps create the conditions for an overall assessment of risk: “[T]he latent violence, normalization, you have to ask some follow-up questions to get the whole picture” (Social worker, specialized unit, organization 1). Lastly, the use of standardized questions in FREDA seem to help the victim gain insights in and challenge the process of normalization, which could improve the victim’s ability to remain safe and to benefit from treatment.

Between 2014 and 2016, the specialized unit conducted between 40 and 50 risk assessments with FREDA each year. The risk assessment idea as well as the tool to execute this seems to have gained a foothold and practical significance for the social workers at the specialized unit. The empirical material, from the point of view of the social workers, conveys a consistency between how the risk assessment practices have been developed, standardized, and executed in the organization and their own view of what constitutes good professional social work with victims of domestic violence. However, it is possible to distinguish a managerial interference in how risk assessment practices are to be organized since “the expertise is not spread across the organization; instead they have become a small island, a small enterprise of their own” (Senior manager, organization 1). The specialized social workers’ expertise in domestic violence is used by the manager to question knowledge-cementation in the organization. If domestic violence cuts across all problem areas, should all expertise on risk assessments be gathered in one place?

In organization 2, risk assessment practices are formalized to a significantly higher degree than earlier. The local rules and procedures (also here constructed by professionals) are more detailed, more prescriptive about how to act, and now also state clearly that SARA is to be used when appropriate. In contrast to organization 1, where assessments are allocated to the specialized unit, organization 2 allows several other units to perform assessments. However, at the time of our data gathering, there seemed to have emerged a kind of catch-22 situation. As in organization 1, there is a common understanding that knowledge about domestic violence is crucial for performing a risk assessment properly; however, such knowledge is not evenly distributed among the employees in the organization but is concentrated to the specialized unit. When other social workers in the organization estimate that a risk assessment is needed, they remit the client there since they doubt their own competence performing assessments.

At the specialized unit, risk assessments are placed in a queue due to the heavy work load (high frequency of clients, severity of client situations, etc.). According to our respondents from the specialized unit, knowledge about violence and risk assessment competence is limited and strategies for action are not anchored well enough outside their unit: “[If] someone fires a gun through the window of a car, we need no risk assessment, but then it is pretty bad, common sense is all that is needed to judge that as dangerous” (Social worker, specialized unit, organization 2). This social worker refers to an actual request of a risk assessment from social workers from a different unit.

The consolidation of risk assessment tools in our cases is contingent on mechanisms such as consistency between professional norms and the tools in question, the level of knowledge about the practice of risk assessment (including knowledge about domestic violence and its mechanisms), as well as how work is organized. The practical effects of the unequal knowledge spread in organization 2 have meant that risk assessment is “ordered” from the specialized unit even though the unit cannot perform the assessment.

Discussion

This study examines how the idea of using risk assessment tools was manifested and processed in the Swedish social services work with domestic violence. The findings from our cases support conclusions from other studies: Increased standardization or what Robinson (2003) refers to as technicality, does not necessarily alienate social workers, but can strengthen their ability to perform professional tasks.

The result could be interpreted as a bottom-up professional drive, successfully pursuing the quest for standardization, removing assessments based on gut-feeling. Consequently, it becomes problematic to explain standardization of social workers’ assessment practices solely with reference to bureaucratic and hierarchic managerial governance. We have illustrated how a relatively small group of social workers in both organizations seems to have been able to forward their claims, execute their intentions, and decide how risk assessment work should be carried out without much intrusion from the management or political domain, at least in the first years. In the first years, there were a professional freedom to try things out and experiment with different methods and tools to use and also to connect with other social workers and experts in the domestic violence field.

In our cases, the pursuit for standardization by the professionals is at the same time connected to a larger movement where the inherent uncertainty in social work is supposed to be eliminated or reduced by the implementation of risk tools such as FREDA or SARA and the striving towards EBP. The social workers at the specialized units have been successful in anchoring their operations in this larger movement. In other words, the “opportunity-structures” possible for the social workers to take grasp on, conditions which paths towards increased professionalization that actually are possible to follow. That is, not all paths forward are possible if one is to be viewed as legitimate and be successful in claiming jurisdiction.

In this field, we witness the pursuit for standardization of assessment practices both from the outside (i.e., the state through legislation and government funds) and from the inside (i.e., the social work profession). We view the implementation both as a way to seek legitimacy and to claim jurisdiction and as a way for the social workers to increase their professionalism, that is, an enhanced quality in the execution of tasks. It is not simply a question of symbolic actions decoupled from practices. The coherence between exogenous and indigenous mechanisms is likely an important factor in understanding the discretion the professionals have had when developing practices in their respective organizations. Drawing on Waring (2014: 701), it is possible to view the insiders identified in this article as practice elites in their contexts, reflecting a “growing sub-specialization of professional work” and as such they “shape organizational change” by pushing for standardization.

Even if we have not closely observed what kind of knowledge that is used in relation to clients and what importance the tools play in such a work, it is somewhat paradoxical that the knowledge claimed to constitute and define the risk assessment part of this new social service field is gathered from outside the profession. The knowledge materialized in the tools is developed not by their users (the social workers in our cases) but by outside actors. This might, as others have shown (cf. Herz, 2012), impact on how the problem of domestic violence can be understood. Paired with the relative randomness we have identified with respect to which tools to use, an interpretation is that the establishment of this new professional task can be designated as instable. Standardization and risk tools are indeed intertwined with a professional project in the social workers’ statements and in external impositions, but the type of tool and their implications for professional judgment are debatable, as we have shown, not only by our professionals but also by researchers and authorities. For clients, this means that they may face different tools depending on where they live, and the consequences of this inconsistency are unknown. We therefore wish to emphasize the need for future research to examine the consequences of the knowledge materialized in these various tools on social workers’ understanding of domestic violence as a social problem and the consequences of such an understanding for the interactions and interventions vis-a-vis the social workers’ clients. Incorporating social service users experience’s in such research is essential.

According to our participants, a thorough theoretical understanding of domestic violence and experience of meeting victims and offenders are prerequisites for performing assessments with risk tools such as FREDA and SARA. They warn against losing the practical glance at the expense of the high confidence in the tools. Others (cf. Ponnert and Svensson, 2016) have connected the use of, and increased confidence in, manuals and tools to a paradoxical increased professional uncertainty and anxiety, ultimately reducing professional self-esteem if other knowledge sources and aspects of practice are under-emphasized. Perhaps, this is what we have witnessed when social workers not specialized in domestic violence express a need to consult the specialized unit when a man fires a gun through a window of a car. Arguably, with some confidence in one’s professional judgment, they also could have concluded that he was potentially dangerous to others. At the same time, as the specialists could be said to have increased and promoted occupational professionalism when implementing the tools, among other segments within the profession, we can possibly witness a less reflective and more passive stance, which would imply a reduction in professionalism.

Conclusion

In conclusion, we have (like others such as Ponnert and Svensson, 2016; Robinson, 2003) illustrated the possibilities to balance increased standardization with what our respondents equate with high quality social work and professional judgment. This balance, however, seems not to fall out equally among different segments of the social work profession. Furthermore, what has not been penetrated in this article, but is of high importance, is what consequences the professionalization of support and services such as risk assessments and safety planning might have for survivors of domestic violence. Hoyle (2008), for example, has pointed out that “with risk assessment and management techniques in place, victims may be more likely to be blamed for their future victimization” (p. 333). The possible change from a feminist structural understanding anchored in the women’s movement towards an individualized and professionalized social service practice might put significant pressure on domestic violence victims rather than on the violent men who commit these violent acts to change their “risky” behaviors.

There are some limitations with this study. First, the study has not included everyone in the organizations; it is possible that we would have found other stories regarding risk assessment practices not heard with the current design, for example, from people highly critical of risk tools. Second, we do not claim that our results are generalizable in a more statistical sense. However, our theoretical analysis and comparisons with other studies suggest possibilities for analytical or theoretical generalizations.

Footnotes

Acknowledgement

The authors would like to thank senior lecturer Maria Alm for comments on the earlier drafts of this article.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Swedish Crime Victim Fund.

References

Abbott

(1988) The System of Professions: An Essay on the Division of Expert Labor, Chicago: University of Chicago Press.

Ahlbäck Öberg

Bull

Hasselberg

et al. (2016) Professions under siege. Statsvetenskaplig tidsskrift 118: 93–126.

Barfoed

Jacobsson

(2012) Moving from “gut feeling” to “pure facts”: Launching the ASI interview as part of in-service training for social workers. Nordic Social Work Research 2(1): 5–20.

Belfrage H (2011) Fler kvinnor kommer dö [More women will die]. Available at: www.aftonbladet.se/debatt/article12656493.ab (accessed 13 October 2017).

Cattaneo

Chapman

(2011) Risk assessment with victims of intimate partner violence: Investigating the gap between research and practice. Violence Against Women 17(10): 1286–1298.

Danis

(2003) Social work response to domestic violence: Encouraging news from a new look. Affilia 18(2): 177–191.

Ekström

(2016) Carriers of the troublesome violence – The social services’ support for female victims of domestic violence. European Journal of Social Work. Available at: www-tandfonline-com.web.bisu.edu.cn/doi/abs/10.1080/13691457.2016.1255930?journalCode=cesw20 (accessed 13 October 2017).

Evans

(2010) Professional Discretion in Welfare Services, Burlington, VT: Ashgate.

Evetts

(2010) Reconnecting professional occupations with professional organizations: Risks and opportunities. In: Svensson

Evetts

(eds) Sociology of professions: Continental and Anglo-Saxon traditions, Borås: Daidalos, pp. 123–144.

10.

Fitzgibbons

(2008) Deconstructing probation: Risk and developments in practice. Journal of Social Work Practice 22(1): 85–101.

11.

Forgey

Allen

Hansen

(2014) An exploration of the knowledge base used by Irish and US child protection social workers in the assessment of intimate partner violence. Journal of Evidence-Based Social Work 11(1–2): 58–72.

12.

Freidson

(2001) Professionalism: The Third Logic, Cambridge: Polity Press.

13.

Gillingham

(2011) Decision-making tools and the development of expertise in child protection practitioners: Are we “just breeding workers who are good at ticking boxes”? Child & Family Social Work 16(4): 412–421.

14.

Greenhalgh

Robert

Macfarlane

et al. (2004) Diffusion of innovations in service organizations: Systematic review and recommendations. Milbank Quarterly 82(4): 581–629.

15.

Hall

(2012) Quality improvement reforms, technologies of government, and organizational politics: The case of a Swedish women’s clinic. Administrative Theory & Praxis 34(4): 578–601.

16.

Herz

(2012) Från ideal till ideologi [From Ideal to Ideology], Örebro: Örebro University.

17.

Hoyle

(2008) Will she be safe? A critical analysis of risk assessment in domestic violence cases. Children and Youth Service Review 30: 323–337.

18.

IVO (The Inspectorate of Health Care and Social Welfare) (2014) Våldsutsatta kvinnor och barn som bevittnat våld. Slutrapport från en nationell tillsyn 2012-2013. IVO Stockholm.

19.

Kouzes

Mico

(1979) Domain theory: An introduction to organizational behavior in human service organizations. The Journal of Applied Behavioral Science 15(4): 449–469.

20.

Kropp

(2008) Intimate partner violence risk assessment and management. Violence and Victims 23(2): 202–220.

21.

Lauri

(2016) Narratives of governing: Rationalization, responsibility and resistance in social work, Sweden: Dissertation, Umeå University, Umeå.

22.

Layder

(1998) Sociological practice: Linking theory and social research, London: Sage.

23.

Lindhorst

Casey

Meyers

(2010) Frontline worker responses to domestic violence disclosure in public welfare offices. Social Work 55(3): 235–243.

24.

Lindhorst

Meyers

Casey

(2008) Screening for domestic violence in public welfare offices. Violence Against Women 14(1): 5–28.

25.

Lipsky

(2010) Street-Level Bureaucracy, 30th Ann. Ed.: Dilemmas of the Individual in Public Service, New York: Russell Sage Foundation.

26.

Ljungwald

Svensson

(2007) Crime victims and the social services: Social workers’ viewpoint. Journal of Scandinavian Studies in Criminology and Crime Prevention 8(2): 138–156.

27.

Messing

Thaller

(2015) Intimate partner violence risk assessment: A primer for social workers. British Journal of Social Work 45(6): 1804–1820.

28.

Ponnert

Svensson

(2016) Standardisation – The end of professional discretion? European Journal of Social Work 19(3–4): 586–599.

29.

Robinson

(2003) Technicality and indeterminacy in probation practice: A case study. British Journal of Social Work 33(5): 593–610.

30.

SOU 2008:18. Evidensbaserad praktik inom socialtjänsten – till nytta för brukaren. Stockholm: Socialdepartementet.

31.

SOU 2014:49. Våld i nära relationer: En folkhälsofråga. Stockholm. Justitiedepartementet.

32.

Tavory

Timmermanns

(2014) Abductive analysis. Theorizing qualitative research, Chicago: University of Chicago press.

33.

Timmermans

Epstein

(2010) A world of standards but not a standard world: Toward a sociology of standards and standardization. Annual Review of Sociology 36: 69–89.

34.

Waring

(2014) Restratification, hybridity and professional elites: Questions of power, identity and relational contingency at the points of “professional–organizational intersection.”. Sociology Compass 8(6): 688–704.

35.

Wastell

White

Broadhurst

Peckover

Pithouse

(2010) Children's services in the iron cage of performance management: Street-level bureaucracy and the spectre of Švejkism. International Journal of Social Welfare 19(3): 310–320.

36.

Yin

(2009) Case Study Research: Design and Methods, 4th edn. London: Sage.