Domestic violence policies in the Netherlands: A regime of deficiency

In a number of countries, domestic violence is represented as a governable phenomenon that is amenable to policy interventions. Over the past 40 years in the Netherlands, however, this approach has not resulted in a reduction of domestic violence. Yet new policy strategies continue to be designed to improve existing interventions. In this article, we focus on a Dutch policy measure that aims to detect early signals of violence and abuse. We argue that this strategy, by approaching domestic violence as a technical problem, fails to take into account structural and symbolic violence. As a consequence, the impact of domestic violence policies on women, particularly poor women, and especially women with a migration background, is to intensify their difficulties. Moreover, these policies deploy a technology that shapes the subjectivity of professionals engaged in protection practices, while maintaining the status quo of inequality and violence against women. The connection between these two flaws of domestic violence policies leads us to claim that the current approach is constituted as a regime of deficiency.


Introduction
Domestic violence has been a high priority on the policy agenda in many postindustrial states over the past four decades, but nothing indicates that this results in a reduction of domestic violence. Nevertheless, new policies continue to be designed and introduced in order to improve existing interventions and their professional infrastructure. In this article, we seek to explain why interventions based on the same set of assumptions continue, despite their failure to reduce domestic violence. We shall do so by examining the development of early detection policies in the Netherlands. We will demonstrate how these policies are embedded into processes of neoliberal governance in which individual responsibilities are prioritized over structural factors underlying domestic violence, such as gender inequality and poverty.
As a consequence, these policy interventions have had a negligible effect on reducing violence against women, particularly poor women, and have been especially harmful to those with a migration background. At the same time, they have a significant impact on the subjectivity of social workers charged with protection practices, which effectively engages these professionals in maintaining the status quo of inequality and violence against women. We introduce and situate the concept of a regime of deficiency within theories of subjectivity and governmentality and regimes of truth, arguing that a regime of deficiency connects these dual flaws in domestic violence policies. We demonstrate how the current form of neoliberal governmentality in the Netherlands is translated into policy instruments that focus primarily on deficiencies in professional practices aimed at protecting vulnerable citizens.
We offer an ethnographic analysis of one recurring theme that is currently subjected to changing policy interventions: the detection of signals of violence and abuse. In this approach, domestic violence is rendered as a technical problem (cf. Li, 2007), which requires control and constant improvement in order to secure the protection of vulnerable citizens. These technologies simultaneously shape the subjectivity of social workers by disciplining them in techno-reporting systems that focus primarily on imperfections in professional practices, whilst neglecting structural features of domestic violence, such as gender inequality and social exclusion from mainstream society as a result of material hardship. Gender-neutral policies thus contribute to the disappearance of structural violence in policy concerns and development. As a consequence, new policies have a limited impact. This leads us to argue that Dutch policy approaches regarding domestic violence effectively aim at maintaining the existing social order and its gender inequalities.
We describe this approach as a regime of deficiency since it revolves around gender neutrality and the individualization of violence, even as it ignores the structure of violence against women, especially poor women, more particularly poor women with a migration background. This regime reflects a utopian modernism that can be traced back to the 19th century, when the conviction emerged that it is possible to eliminate all risks by 'articulat(ing) a standard of conduct that everyone must observe' (Ewald, 2002: 297). All kinds of tools, including campaigns, reporting codes, detection instruments and protocols, are put in place to ensure breaking the taboo in society about reporting domestic violence. Tools that have been introduced as neutral instruments to support professionals in detecting domestic violence, however, effectively contribute to a normative inscription influencing the conduct of individual professionals, which in turn reflects a deeply rooted form of distrust of clients and professionals. Thus, a regime of deficiency transforms professionals into tools themselves by narrowing their professional space, which is confined through the instruments they are prescribed to use and the protocols they have to follow. This move ensures that professionals also comply with a regime of truth and violence that entails risks and uncertainties predominantly resulting from a narrow definition of violence, while in practice they are confronting a structure of violence that is built on violence against women.
Domestic violence and child abuse are complex problems, ultimately shaped by both individual peculiarities and structural conditions. In a regime of deficiency, however, they are reduced to a technical issue, neglecting the role of the state in sustaining violence that is chiefly generated by factors such as poverty, gender inequality and/or racial discrimination. Within this delimited technical matrix, a variety of policies, quality programmes and schemes are designed and redesigned by the state, usually in collaboration with professional experts, to improve the performance of practices that aim at combatting domestic violence. The neoliberal tendency towards individualization in professional practices and the responsibilization of citizens opened up, we argue, a 'technical space' that corroborates the widespread conviction that the refinement of intervention practices and organizational structures can resolve the problem. Indeed, much effort and money are invested in mending the flaws in interagency collaboration, blocking out uncertainties by measuring and monitoring clients, and by standardizing and disciplining the conduct of professionals.
In the Netherlands, domestic violence encompasses child abuse, elderly abuse and intimate partner violence, against both men and women. In the past, however, this was not always the case. The first policy document on sexual violence was released in the 1980s and focused exclusively on women and girls by stating unequivocally that the state had an obligation to protect them from sexual and physical violence. Domestic violence was considered as the oppression of women through patriarchal dominance (Acker and Rawie, 1982). Furthermore, it was represented as being of a gendered nature: women are victimized because they are women.
Over the years, however, three counter-narratives have changed the campaign to stop violence against women in the Netherlands, to wit, minimization, gender neutralization and culturalization (R€ omkens, 2017). First, numbers on the nature and size of intimate partner violence against women are often downplayed by overstating violence against men, thus effectively discounting the violence as gender-based (De Vaan et al., 2013). Second, a legal discourse on sexual violence emerged in the 1990s, which entailed the introduction of sexual morality laws such as the criminalization of marital rape. These policies simultaneously shifted the focus from victims or survivors towards perpetrators. This move was supposed to strengthen the protection of women, but the juridification of domestic violence also involved a more gender-neutral approach in order to match the neutral objectives of law. Accordingly, the policy perspective changed dramatically to a more 'safety' oriented approach that decentred gender inequality as a root cause for intimate partner violence. Violence in intimate partner relationships came to be imagined as caused by individual pathologies that require care (Naezer and R€ omkens, 2008).
During the next decade, the culturalization of violence against women also emerged as a third counter-narrative in the Netherlands. In this narrative, violence against women is viewed as a problem of people with a migration background in non-western countries, especially Islamic societies, whereas violence against Dutch women (and children, for that matter) continues to be viewed as caused by individual incidents. Cases in which a man kills his wife and children are referred to as family tragedies in which so-called relational problems were involved. Those incidents are not necessarily understood as an intrinsic part of Dutch culture, but as individual cases. In this sense, the Dutch are supposedly without culture. Following R€ omkens (2017), such a narrative obscures intersectional analyses of intimate violence against women in which a complex of interrelated factors, such as migration, citizenship, economy, religion and inequality play an important role.
The current gender-neutral and individual approach to combat domestic violence has resulted in a range of policy measures since 2002. These policies are increasingly focusing on the realization of an infrastructure that operates in terms of this understanding of domestic violence, such as the creation of a nationwide network of domestic violence agencies, the merging of those agencies with child abuse agencies, the development of a reporting code for domestic violence and child abuse, a temporary home restraining order and, more recently, an advanced multi-agency approach to improve collaboration between the police, public prosecution and care services. These and other policy measures are based on the same problem analysis that has been dominating the debate in the Netherlands over the past 40 years. Based on a content analysis of policy documents over this period, we distinguish three interrelated themes in this discourse: the lack of detection of signals of domestic violence, the poor registration of domestic violence and failing multi-agency collaboration. All themes revolve around deficiencies in professional practices, the linchpin of the approach in which social workers have been tasked with care and control of victims and perpetrators.
Unfortunately, however, the policies that have been designed and implemented around these themes have not resulted in a significant reduction of domestic violence, not even of gender-based violence against women in intimate partner relations (R€ omkens et al., 2016;Tierolf et al., 2014). Nevertheless, the government continues to rely on the same approach. In order to improve previous interventions and their professional infrastructure, existing policy strategies are frequently refined, redesigned and re-introduced. In 2015, for example, the separate agencies for domestic violence and child abuse were merged into regional agencies called Safe at Home (Veilig Thuis) in order to refine multi-agency responses towards situations that involve both partner violence and child abuse. As such, a 'technical' lens has been developed for the resolution of domestic violence, which, in turn, triggered the demarcation of a boundary that marks a specific knowledge domain.
This article is based on ethnographic research that involved 100 days of participant observation within one Safe at Home agency in the Netherlands between 2015 and 2017. During this period, 75 people were interviewed, including survivors of domestic violence, professionals of Safe at Home, physicians, policy-makers, aldermen and police officers. In addition, more than 350 socio-materials, such as policy documents, minutes of parliamentary sessions and newspaper articles were collected and analysed. All places that are referred to in this article have been anonymized for obvious reasons, while pseudonyms are used to protect the privacy of our interlocutors. We begin with a further theorization of our conception of a so-called 'regime of deficiency'.

A regime of deficiency
Regimes are effects of the rationality of government, labelled by Foucault (1991) as 'governmentality', which he understands as a complex series of human and nonhuman power assemblages-instead of a single power assemblage-to govern a population. The shift from government to governance was facilitated by the invention of political technologies, including statistics, policy-making and audits. Foucault's concept has been fundamental for the emergence of an anthropology of policy (Shore and Wright, 1997;Wright and Shore, 1995), while it was also instrumental in the study of 'audit cultures' (Shore and Wright, 2015;Strathern, 2000). Exemplary for anthropological approaches is the ethnographic focus on multiple state and non-state practices aimed at modifying human conduct, selfunderstanding and self-management (Eckert et al., 2012). Indeed, policies and audits, as particular political technologies and forms of power, 'work upon the individual's sense of self' (Shore and Wright, 1997: 29;italics in original), while at the same time implying objectivity and 'recast(ing) a political problem in the neutral language of science' (Shore and Wright, 1997: 31).
Likewise, policies and audits actively constitute the subjectivity of professionals and rely on their collaboration to be effective. In the course of our fieldwork, a seminal moment occurred that illustrates this effect of audit cultures. During a staff meeting at Safe at Home, a senior staff member tried to motivate and prepare his colleagues for another audit by the Health and Youth Care Inspectorates (HYCI). 'I want to make sure we are not implementing the improvements solely for the Inspectorate, but I want it to be our will', he argued. With his statement, he is clearly offering a critique of audit practices that make governing from a distance possible, but which might produce 'improvements' that are intrusive to his daily practice. At the same time, he is showing his compliance with the audit by emphasizing his 'will to improve' (Li, 2007). This example illustrates that in the policy regime governing the domain of domestic violence, a will to improve is inherently linked to a will to purity and a will to control. Ironically, the interplay of these different wills produces a utopian practice that transforms the policy field into a regime of deficiency. For that reason, too, it is necessary to examine how exactly these different dimensions constitute the governing regime. We distinguish four different desires that may be considered characteristic of domestic violence policies: the will to protect; the will to improve (Li, 2007); the will to purity (see also Mellaard and Van Meijl, 2017) and, finally, the will to control.
First, the will to protect expresses the political urge to offer protection to 'vulnerable' people, although it is highly diversified and also historically contingent. In the current Dutch context, for example, children are prioritized above the elderly, men and women or even families, as expressed in (local) policy documents through slogans such as 'children first' (Baartman, 2007).In policy protocols, child abuse is also singled out as a separate category within the general domain of domestic violence. Thus, it may be argued that the fundamental will to protect generates a debate about who deserves to be protected, even to the point that a hierarchy of victims emerges among victim categories (Adelman, 2017). The policy gaze of the will to protect is directed to the prevention of violence in individual cases and leads, amongst other things, to the development of early detection policies and risk assessment instruments. These policies and instruments trigger, in the second place, a will to improve which manifests itself in an 'endless deferral of the promise of development to the time when the ultimate strategy is devised and implementation perfected, (which) keeps the attention of many critics focused on the deficiencies of such schemes and how to correct them' (Li, 2007: 276). In this debate, in the third place, a will to purity emerges through an emphasis on efficiency that calls into being a distinction between pure and impure professional practices. Even though modern practices present themselves as pure, enacting a single logic of efficiency, they are usually made up of a messy patchwork of different logics rendering them impure beneath the surface (Latour, 1993;Law et al., 2013). As we will demonstrate, however, professional practices need to present themselves as pure when dealing with policy actors in order to obtain legitimacy, even though this requires social workers to navigate a discrepancy between representation and actuality. Fourth, and finally, the will to control emphasizes the degree of trust in professional practices. As a result of new public management rationalities that have been introduced over the past four decades, all kinds of audits, measurements and review techniques have been put in place to certify that policies are enacted as intended, thus tying professionals and organizations to policy objectives to ensure that their practices deliver 'quality' and keep finances 'in control' and that a range of risks are minimized (Shore and Wright, 2015).
These multiple 'wills' intersect in everyday practices, either reinforcing particular effects or counteracting one another, leading to ambiguities, discrepancies or even failures. Irrespective of the effects of these processes, the result is a particular assemblage that is enacted as a 'regime of deficiency' in professional practices. However, this configuration is by no means random. On the contrary, it is embedded in a 'regime of truth' (Foucault, 1977) and a 'regime of violence' (Schinkel, 2013).
A regime of truth is constituted by a specific territorially based set of state and non-state institutions and actors. Regimes of truth determine which discourses are regarded as truthful and which are not. Furthermore, they control not only which statements are true or false, but also how they and their associated actions are sanctioned. A regime of truth is therefore a strategic field within which truth is not only produced, but often it also becomes a tactical element in the functioning of power relations as it governs the differentiation between what counts as truth and what doesn't (Lorenzini, 2015). Indeed, these regimes govern conduct and as such they constitute a link between subjects and the state (Foucault, 1977). More importantly, regimes of truth also organize normalizing and disciplinary strategies that shape the subjectivity of people in order to bind them to the manifestation of truth (Foucault, 2014), such as the social workers caught between care and control of domestic violence.
Regimes of truth are reproduced through the daily enactments of the multiple 'wills' in a regime of deficiency. The materialization of those wills in policy, norms, quality frameworks, audit techniques, risk assessments and review practices as a manifestation of truth in professional practices links the subjectivity of professionals and their conduct to what are considered as 'pure' practices. However, what is regarded as 'pure', as we will show, may be highly ambiguous and frequently far removed from actual practices. Still, the continuous identification of ambiguity, deficiencies and failures in professional practices ensures that we remain with the promise of improvement (Li, 2007) rather than offering a critique of the problematization of domestic violence by the state that hardly has an impact.
This approach to governmentality assigns the state the general role of normative enforcement through violence, while understanding that various forms of public policy and institutional and professional practices are also vital parts of the regime of truth. Schinkel (2013: 8) has argued compellingly that 'the core of the state's violence lies in its very power to recognise and sanction violence'. The state's power is instrumental in the recognition of legitimate and illegitimate forms of violence (including violence committed by the state), even in the fundamental question of what counts as violence and what does not. According to Schinkel, violence used by the state itself is normally imagined as a necessary response to private violence. The re-recognition of domestic violence in the Netherlands as a private illegitimate form of violence is therefore an important event (see Ministerie van Justitie, 2002) since it legitimizes state intervention. More importantly, the state also defines what domestic violence is.
In the Netherlands, domestic violence is defined as 'physical, mental or sexual violence or to threaten with this by someone from the household' (Wet maatschappelijke ondersteuning (Wmo), 2015). Forms of violence that are distinguished include child abuse, intimate (ex-)partner violence, forced marriage and honour violence, elder maltreatment and female genital mutilation. Thus, the state frames violence as a more or less private event between two (or more) people who are involved in an intimate relationship. This definition makes private violence visible, but at the same time it makes other forms of violence invisible, such as structural and symbolic violence, or how private violence intersects with these forms. In this article, we highlight the link with structural violence especially, and how this is both absent and present in professional practices.
The concept of structural violence was introduced by Johan Galtung (1968). Although notoriously difficult to define and delineate, structural violence is generally understood as the violence of 'injustice and inequity' that is embedded in social structures and normalized by institutions, bureaucracies and experience (see Rylko-Bauer and Farmer, 2017: 1). Social structures intersect with the whole range of economic, political, cultural, religious and legal domains that shape the lives and conduct of individuals and groups, which in practice includes patriarchy, slavery, poverty and all kinds of discrimination. Sokoloff and Dupont (2005: 4) point out that 'structural issues, particularly socioeconomic factors, are significant in contributing to and perpetuating domestic violence'; they 'constrain and shape the lives of battered women, albeit in different ways' (3).
The Dutch policy approach, however, excludes structural violence and, instead, stresses individual pathologies and responsibilities. This manifestation of a neoliberal governmentality also had its effects, as Bouagga (2015: 159) noted, on all dimensions of social work: 'Tasked with "responsibilizing" their publics, social workers have seen their duties shift more towards control rather than direct intervention.' Indeed, not only poor people, especially impoverished immigrants, but also social workers are confronted with a structure of violence that is sanctioned by the state and built on violence against women. This strategy that excludes structural violence informs a policy approach that simply renders the problem to be resolved as a 'technical' problem. Li (2007) describes this process as the gradual construction of a boundary around a knowledge domain and associated policies that is viewed as knowable and open to improvement. Essential to her understanding of 'rendering technical' is that the approach to knowledge is partial, both 'skewed and incomplete' (Li, 2007:154).
Regarding domestic violence, substantial knowledge about the impact of poverty on child abuse or the gendered nature of partner violence is widely available, but it is currently not included in the development of policies. As a consequence, the Netherlands state has been repeatedly criticized by the United Nations Committee on the Elimination of Discrimination against Women (CEDAW) regarding its gender-neutral approach to domestic violence, leaving out genderbased causes and contexts, such as differences in power between men and women (Bijleveld et al., 2017;Hearn et al., 2016). Since 2010, the response by respective governments to these criticisms has been that a gender-neutral approach in policy does not necessarily hinder a gender-specific approach in practice. De Vaan et al. (2013) have criticized this response for being unclear in its message and have demonstrated, moreover, that a gender-specific approach in practice is an illusion.
Part of the 'technical space' surrounding domestic violence is concerned with the detection of signs. In the 1980s, the detection of signs was mainly improved by enhancing the expertise of professionals through training, but in the 1990s knowledge about domestic violence gradually changed. It became clear that the relatively low number of reportings of domestic violence and child abuse is not in balance with estimates about their prevalence. A simple calculation led to the conviction that large numbers of people who might need protection are being missed (Tweede Kamer, 2012). A number of fatal incidents of predominantly child abuse also led to demands for earlier detection of such cases (Kuijvenhoven and Kortleven, 2010). This intensified the discussion about the need for professionals to become aware of domestic violence and to act when noticing signs of violence, abuse or neglect (Tweede Kamer, 2008). In this sense, the gap between the number of reportings and the estimates of domestic violence, which we refer to as the 'reporting gap', was and is regarded as a deficiency to be addressed.
The process of improving the detection of signals is supported by an expanding hinterland of knowledge about domestic violence. This knowledge is translated into training and risk assessment tools for different kinds of professionals and interagency networks. These 'products' are also under scientific scrutiny for improvement, which is characteristic of what we describe as a regime of deficiency: the constant isolation, evaluation and innovation of the 'technical space' in approaches to resolve complex societal problems. The focus is on the further refinement of policies, programmes, infrastructures and the tools that are used to combat domestic and child abuse each time when it becomes clear that statistical indicators of domestic and child abuse are not diminishing or when the detection system fails to keep people safe from harm, even though there is no evidence that this will improve practices or, worse, may even be detrimental to them (Kuijenhoven and Kortleven, 2010;Munro, 2004). Still, when an innovation in policy does not deliver what is expected, it is generally attributed to a lack of knowledge about working processes and methods by professionals and their organizations (Bosselaar, 2018).
At the same time, more critical approaches, such as the intersectional approach by feminist scholars and the women's movement, have been mainstreamed. In recent policy documents, references to gender inequality or to other structural causes of domestic violence are missing. Instead, intra-and inter-personal dynamics are emphasized. The entire infrastructure that has been built around domestic violence and child abuse, including reporting agencies, interagency collaboration and the tools used to administrate and assess individuals, is founded on this approach, which disciplines social workers without having a significant impact on the reduction of gender-based violence. In the next sections, we elaborate how a regime of deficiency aims at improving professional conduct within such an infrastructure but, instead, produces ambiguity and new deficiencies to be improved.

Taboos: Detection of warning signs of abuse and violence
The main thrust of domestic violence and child abuse policies is the manner in which the public in general, and professionals in particular, detect and report early warning signs. In view of the reporting gap, health care inspectors investigating tragic, often fatal cases commonly draw the conclusion that professionals often suspected abuse, but that they did not act upon their suspicion. In 2003, during a parliamentary debate about the first policy document on domestic violence, the secretary of the Health Department formulated this deficiency as follows: There is still a taboo on signalling domestic violence. This fact can only be addressed when we are prepared to talk about how we deal with one another. We must equip professionals for this task. The threshold to talk about parenting and bad parenting should be as low as possible. (Tweede Kamer, 2003: 13) Here, domestic violence and the need to take note of it in one's own environment is framed as a taboo. A general appeal is made to become more sensitive to signs of domestic violence as it should be 'normal' to talk about this, while professionals should be 'equipped' to do the same. This frame, building on the historical notion of domestic violence as a private affair, is still dominant in government policies. The perceived taboo is often described in terms of 'shyness to act' (handelingsverlegenheid), indicating an inclination of professionals to avoid domestic violence. Both bystanders and professionals are assumed to 'look away' when suspecting that violence is taking place in certain situations, which they subsequently rationalize as 'a private affair' or argue they have no evidence of the suspected abuse.
In this context, taboos serve as symbolic boundaries around unwanted or displaced matters in society that have the potency of disorder and risk (Douglas, 2002(Douglas, [1966). Domestic violence entails such disordering and risk. In the past, this type of disorder was considered a risk for the unity of the community, or as undermining the patriarchal status of men in families. At present, it is regarded more as a risk for the safe and healthy development of individuals and a violation of their human rights. Before the 1980s, state intervention in domestic violence was not even contemplated. Respect for the private lives of citizens was predominant and family boundaries were relatively sacred. Since the 1980s, however, this view has gradually changed into a perspective that gives the state an important role in protecting the rights of individuals in private settings. Thus, the taboo frame in policy discourses aims at breaking the taboo that domestic violence is a private affair. The government attempts to do so, first, by introducing campaigns to raise public awareness and by appealing to good citizenship, and, second, by inscribing professionals. In the next section, we first examine the strategy that aims at normalizing the detection of signs of violence and abuse.

Public campaigns, policy reform and waiting lists
Over the years, different campaigns for the prevention of domestic violence and child abuse have been held, usually reflecting the policies at the time. The shift in the focus of these campaigns is significant. In 1991, for example, a campaign entitled Nu is het genoeg (Enough now) aimed at changing common ideas about masculinity, but over the years campaigns have begun urging individuals to call an agency and make a report. In 2012, three types of violence were simultaneously highlighted in a campaign-child abuse, intimate partner violence and elder maltreatment. Since 2015, this campaign refers to the new Safe at Home agencies, thus expressing a will to improve inter-agency collaboration and the processing of reports that include both intimate partner violence and child abuse.
The message of the campaigns to take action, however, is not always in line with a professional infrastructure that can handle the requests for advice and process the reports. A discrepancy exists between, on the one hand, the goal of recent campaigns to raise awareness and increase reporting about concerns, and, on the other hand, how those reports are handled by the agencies. Indeed, in practice, the will to protect conflicts with the will to control and the will to purify. Professionals in Safe at Home, for example, observe an increase in reported concerns after each campaign, but they hasten to add that these also generate enormous waiting lists. In practice, it is always a combination of different factors that leads to waiting lists.
As a result of severe budget cuts, the newly formed Safe at Home agency was understaffed from the outset. The reorganization of child abuse agencies and domestic violence agencies preceding the establishment of Safe at Home also made many employees highly unmotivated for the new challenges, such as the implementation of a new 'protocol of conduct' and a new 'triage instrument' that covered different kinds of intimate violence. Several employees confessed that they did know about the new protocol, but had not read it. They were too busy to keep up with the workload or unmotivated because of their precarious employment.
Furthermore, many professionals at Safe at Home experienced the new triage instrument as too 'obsessed' with safety and (extreme) danger instead of being sensitive to the healthy development of children, which was the main concern behind most reports. This caused the triage instrument to impede their work since processing a report took considerable time. As a consequence, most experts did not use the 'triage instrument' to evaluate the reports as systematically as required. On the contrary, in most cases staff already figured out how to proceed with a report after a preliminary reading. Nevertheless, they had to 'please the system' and 'tickbox' their way through the new, but heavily criticized, triage instrument.
The newly formed Safe at Home organization was furthermore tasked with decreasing the number of investigations in response to reports. Instead, Safe at Home was requested to refer more cases to the so-called 'social teams' that were formed in the municipalities at the same time to integrate social services for residents on a range of welfare topics. Social teams are part of a strategy to organize services 'closer' to residents. Municipalities hoped that referral of 'lighter cases' of domestic violence to the social teams would compensate for the budget cuts of Safe at Home. In this context, however, the Safe at Home organization in which we did fieldwork still developed a 'waiting list'.
Wiege, an experienced senior staff member of Safe at Home, observed that 'waiting lists' create an interesting dynamic within the policy assemblage. He described a 'culture of reporting and supply', indicating that the number of reports increases as a result of the campaigns. He feels the pressure to act upon all reports made by citizens, but he also expresses his concern: 'the risk is that we are working on the "wrong" people'. He explains that the normal procedure for handling child reports is to listen carefully for 'facts'-for example, child-related signals, such as inappropriate sexual behaviour, or parent-related signals. Wiege argued, however, that some reports disclose few facts, but instead are steeped in emotions that he associates with, for example, divorces or family quarrels. Routinely, such reports are referred to local social teams or, even more likely, 'downgraded' to advice to the reporter. The advantage of 'downgrading' is that a formal assessment of the situation is not required so that report does not contribute to the waiting list.
Another strategy to deal with the paradoxical situation of policies urging people to report concerns and budget cuts for the agencies dealing with those reports is to mask the existence of waiting lists. Wiege appears to be on his guard when waiting lists are discussed since in his experience they attract negative media coverage that frames the organization as dysfunctional, with political upheaval as a result. In 2016, this situation actually eventuated in most Safe at Home agencies after the combined national Health and Youth Care Inspectorates (HYCI) qualified the long waiting lists and the fact that 80% of the agencies could not meet the strict deadlines for initial safety screening (triage) as 'unacceptable' (Inspectie Jeugdzorg, 2016: 16). The media coverage of the waiting lists was intense throughout the country. As a consequence, anxiety also emerged in the agency in which we conducted research. Political discussions about domestic violence in municipal councils, however, were generally restricted to the technical blueprints that are offered in policy documents. Indeed, the HYCI reports never resulted in an in-depth reflection on domestic violence and related questions regarding, for example, the intersectionality of interpersonal and structural violence. Members of councils, and also aldermen and policy planners, generally translate their 'will to protect' into a constant improvement of instruments and organizational structures as well as by attempting to resolve the problems that are produced by those structures, such as 'waiting lists'.
Early in 2016, a discussion was held to promote Safe at Home in the region. Wiege also participated in that meeting and he insisted that they should be prepared for more reporting in response to more exposure in the media. He had even been asked by some policy planners to postpone or cancel a promotion of Safe at Home since they feared that extra attention to domestic violence would generate more reports. An alderman in another region also recommended postponing all national campaigns, because they confront municipalities with bills they cannot pay. Needless to say, this suggestion received a negative response from a wide range of actors. So, although in official policy discourses detection of domestic violence is framed as important, in practice these efforts are tempered when the number of reports is expected to rise steeply and administering to such numbers exceeds the available budget.
At some stage, Safe at Home also attempted to reframe its waiting lists by referring to them as 'work load'-for example, in communications with the HYCI during the first audit. Unassigned cases on a longlist were divided among all staff members in an effort to convey the impression that domestic violence was 'under control' and their practice efficient and thus pure. We also observed anxiety among domestic violence experts that they might be questioned by the HYCI about their use of the 'triage instrument' or knowledge about the new 'protocol of conduct'. They wondered what to say to the inspectors since they did not really 'use' the triage instrument but, instead, relied on their years of experience with work procedures in the former domestic violence agency. Towards the end of the audit, when the HYCI shared its preliminary and moderately positive review, the staff were immensely relieved. The presentation of their practices had been staged with enough conviction to ward off extra pressure from the inspectorate and the media to improve procedures. For the time being, Safe at Home had satisfied the 'will to purity' of the inspectorate. Subsequently, the unattended cases were taken off each staff member's caseload again and put back on the longlist of unassigned cases.
At the same time, additional measures were taken to manage the waiting list. Extra sessions were scheduled to catch up on the pile of reportings, while alterations were also made to the protocol for assessing safety in the reported cases. This appeared a proven method to process reports faster, but still the waiting lists did not disappear. Since almost all Safe at Home agencies in the Netherlands were facing the same problem, the organization that we studied felt confident enough to inform the responsible municipalities about their waiting lists a few months after the audit had taken place. Political negotiations followed, resulting in extra budget to take care of the waiting lists and to prevent them from reappearing. Towards the end of 2017, when staff numbers had almost doubled and reached the same levels as before the budget cuts of 2015, the waiting lists were finally dissolved.
In sum, improvement in communication, such as public campaigns, and technological improvements, such as a new protocol and the introduction of a 'triage instrument', may be viewed as a genuine will to protect vulnerable individuals and a serious attempt to normalize 'a culture of reporting'. Accordingly, a range of actors seeks to break the taboo on domestic violence, including Safe at Home agencies. The 'will to protect' and the improvements that were implemented, however, were attuned inadequately, as the 'will to control' increased reportings in the context of a slashed budget. Nevertheless, to ward off further problems it was vital to enact a pure practice and show the inspectorate and municipalities that all reports were under review and new working standards were being used. It demonstrates that a regime of deficiency is not only driven by a will to improve, but paradoxically also contributes to other, more or less hidden deficiencies that compromise the will to control and, ironically, corrupt the will to purity. The techno-reporting system is driven by different motivations that counteract each other in practice and therefore preoccupy policy-makers and professionals without making a significant contribution to changing the structural conditions that underlie domestic violence. In the next section, we focus on how disciplining professionals through the use of new 'tools' further contributes to an increase in reporting.

Managing distrust: Inscribing and tooling professionals
A second path that is pursued to break the taboo surrounding domestic violence and to increase reports of possible abuse is to inscribe professionals normatively and to 'equip' them with a range of specific tools. These may vary for different types of professionals, but they all share one 'tool': the Reporting Code on Domestic Violence and Child Abuse.
For various politicians it was a major concern that professionals did not always report their suspicion to official agencies. The discussion focused on the boundaries of professional and medical confidentiality and whether the reporting of concerns should become compulsory (Tweede Kamer, 2008). After years of debate, a law made it mandatory for organizations to implement a reporting code. Such a code standardizes the conduct of professionals in case they suspect violence or abuse. It 'guides' professionals from the moment of suspicion to the decision to arrange help for the people involved, or to report the case to what is currently labelled Safe at Home. This was implemented on top of the continuous training of professionals in detecting signs of domestic violence and child abuse. However, the reporting gap did not diminish, let alone disappear, and professionals still seemed unaware of or reluctant to use the reporting code 1 . For that reason, another effort to close the reporting gap has been pursued by changing part of the reporting code. Now it is mandatory for professionals to report suspected cases of domestic violence and child abuse that are identified as 'severe' to Safe at Home (Tweede Kamer, 2016). Thus, a more refined assessment framework became part of the reporting code in order to encourage professionals to report more frequently to Safe at Home.
The reporting code has been developed from a will to improve the reporting rates and to substantiate both a will to protect and a will to control, in this case in the conduct of professionals. After all, the code is not simply a 'tool' for professionals, but it simultaneously inscribes a norm into their working practices. They are expected to be sensitive to signs of abuse and violence and to act upon their suspicion in order to protect people from harm. As we have discussed elsewhere (Mellaard and Van Meijl, 2017), however, in daily practices professionals have many obligations and are not constantly focused on the battle against domestic violence. As a consequence, policies do not always achieve the changes that are desired. Subsequently, a further improvement of instruments and the 'technical' aspects of the policy chain seeks to address the shortcomings and to control the conditions of professional practices. This approach is set in motion to 'purify' unruly practices and to discipline the conduct of individual employees, but the focus on statistics rather than the structural conditions that engender domestic violence and child abuse creates a double bind for professionals who are working under this regime of deficiency.
The constant improvements of the reporting code must be viewed within the context of a dense network of norms and instruments that govern the conduct of professionals and sustain a 'culture of reporting and supply'. The assemblage of domestic violence policies that is driven by this techno-reporting regime also produces a range of other instruments and interventions. Schools, medical and social work organizations, for example, assigned specific professionals, so-called 'dedicated officers', with the task of acting as expert for their colleagues and implementing extra tools, such as detection lists or risk assessments. These instruments are deployed to create a preventative gaze among professionals and to stimulate a 'follow-up'.
Although campaigns, protocols and detection instruments do improve the reporting rate, they also raise questions. Nora, a confidential doctor, 3 explained that this is a shifting aspect in her work on which she reflected regularly. What do we consider 'the limit' for strong intervention in families, such as the separation of children from their parents? But the definition of a limit, she continued, changes over time. This is exactly what has made her work more difficult in recent years. She gives an example to clarify the changes.
Nora mentioned a report to Safe at Home proceeding from a new alcohol intoxication protocol for 15-year-olds. Instead of paediatricians, Safe at Home has now also been assigned to speak with parents about reported cases. On one occasion, Nora received a report both from the ambulance service and from the emergency room of the hospital to which the intoxicated minor had been brought-in some instances she might even receive a third report from the police. This can be challenging, especially when the reports are slightly different. In this case she read in one of the reports that the father was shocked by the incident since he had already contacted a psychology practice to seek help for his daughter earlier that week. Subsequently, Nora quickly finished her assessment and consulted with a colleague about her intention to downgrade the report to an advice. Against this background, she expressed a concern that Safe at Home may be turning into a kind of surveillance institute, which in her opinion is not an appropriate task. However, this seems exactly the direction in which Safe at Home is developing since ambiguity in the definition of child abuse and domestic violence increasingly causes confusion in professional practices (Baartman, 2009).
Some professionals indicate that Safe at Home is rapidly turning into a last resort for a range of family issues that are often too easily viewed as abuse, neglect or violence. A rising number of reportings may contribute to closing the 'reporting gap', but all those reports also need to be assessed and processed. When staff numbers and other resources for follow-ups are limited, however, it cannot always be guaranteed that those in trouble receive the appropriate attention. This problem is even compounded by the development of detection instruments that generate many 'false positive findings', such as the SPUTOVAMO-checklist that was introduced in 2011 by the Health Care Inspectorate for after-hours primary care services and emergency departments. This instrument made the screening of all children for child abuse mandatory for physicians. Two studies evaluated the use and results of SPUTOVAMO. One study made clear that abuse was likely to be captured through the use of the checklists, but with very high false-positive rates (Sittig et al., 2016). The checklist was considered inaccurate in comparison with criteria for diagnostic accuracy (Sittig et al., 2016). Another study showed that the generic use of the SPUTOVAMO checklist had a low predictive value, since a screening instrument cannot capture the entire complexity of child abuse (Schouten, 2017).
One medical doctor even confided that he has become less attentive to child abuse because of the screening instrument that automatically pops up on his screen when he is seeing a patient after hours in the primary health clinic. He routinely ticks the 'no' box when an automated pop-up screen questions whether he suspects child abuse, because he experiences the completion of the test for each underage patient as disruptive during busy hours (Verstappen, 2018). Still, the GPs we interviewed did feel responsible for finding a solution, and budgeted for a triagist to follow up suspicions. Nevertheless, the use of screening instruments is characterized by great ambivalence.
In the debate about their use, the inspectorate defends the controversial detecting tools. A senior officer of the inspectorate argued: 'The past has taught us that medical doctors normally don't ask questions regarding children's well-being and screening forces them to be sensitive to child abuse. That is better than doing nothing.' 2 In light of this statement, we argue that taboos surrounding domestic violence and child abuse are also rooted in distrust. This is also underlying the adjustments made to the reporting code, in which new 'norms' have been designed from an audit perspective. Reporting norms, as the Health Care Inspectorate declared in 2008, are criteria formulated by the field that allow for an audit of professional conduct. For that reason, too, we contend that detection instruments emerge from a will to control and discipline professional conduct. The distrust of professionals that underscores these policies is deepened when they do not respond as desired to disciplinary instruments. A new round of 'improvements' will begin in order to generate even more refined ways of inscribing professionals and restricting their autonomy. For many professionals this is rather disconcerting as the neoliberal regime of deficiency increasingly ties them to technical policy procedures which do not alleviate the structural conditions that often trigger domestic violence in the first place, as appears from an analysis of the question of who actually is reported.

Who is reported?
Staff members of Safe at Home estimate that most of the reports they follow up with an investigation are about families that belong to lower socio-economic classes. Nora, for example, says: 'domestic violence does occur within upper and low classes, but . . . we visit people from lower classes more often'. Samira, a social worker, added that certain neighbourhoods are more visible in the reports. Ben, a colleague of hers, elaborates: 'underprivileged people in social housing neighbourhoods have less meat on their bones and therefore a lower level of tolerance. Those families deal with increasing tensions.' In contrast, Nahid, another colleague who previously worked in a women's shelter, pointed out that middle-class women may occasionally call the domestic violence agency for advice, but because they have more resources they are not often found in women's shelters.
Since Safe at Home does not keep a registration of socio-economic indicators, it is a challenge to substantiate their staff experiences as described above. A recent exploratory study by Statistics Netherlands (CBS), however, confirms, at least partially, their impressions. They examined all reports that were made to 16 child abuse agencies in 2014. CBS (2017) concludes that almost 40% of the reports come from families in which both parents are unemployed, whereas only 10% of all children in the Netherlands grow up in such families. Furthermore, a staggering 44% of the reported children come from a single parent family, whereas only 15% of all children in the Netherlands life with a single parent. Single parent families are generally among the poorest of all families, while they are also highly gendered. Single mothers are financially more disadvantaged in comparison with single fathers (Nederlands jeugdinstituut, 2014: 9).
In this context, it is interesting to draw a comparison with a study by Dubois (2009: 236) on control practices of welfare agencies in France, which shows that 'bureaucratic control is becoming a tool of (social) control of the poor'. Welfare policies used to aim at alleviating poverty, but in recent decades the goal has shifted to pushing people into work. To achieve this goal, 'control' has become an important tool: The production of moral judgements (on 'a willing attitude' or how to organize one's private life), practical injunctions (do not lose your papers, get up in the morning), penalty decisions (to reduce or cut off benefits according to the behaviour observed) are, in this perspective, more than the side effects of the bureaucratic treatment of the poor. They are now its very objectives, and are increasingly orientated towards improving behaviour. (Dubois, 2009: 236) Detection policies to protect people from harm in the Netherlands have a similar policy effect. People who belong to lower social classes are more frequently reported than people who belong to middle or upper social classes and who are better equipped to avoid interventions by the state. People who belong to lower social classes, especially the most deprived, such as single mothers, not only have to deal with controlling practices of social welfare agencies and their push towards responsible citizenship. They also have to navigate the protection practices of family violence agencies as well as social work teams and their urge to become 'good' mothers. Those who have to deal with the intersections of intimate partner violence and structural violence are also forced to navigate their multiple encounters with a range of state agents and to negotiate the boundaries of self-determination.
Thus, states not only develop policies to battle domestic violence, but they are also part of it. Structural conditions of poverty that foster domestic violence are neglected, while the state also constructs a professional infrastructure that is built on violence against women and that ties social workers to a bureaucratic reporting system that shifts attention away from structural problems and structural violence. Accordingly, Adelman (2017) and McDonald (2005) both argue that domestic violence is embodied in a political economy that understands domestic violence as an individual deviance based on poor choices or stemming from a backward cultural heritage, which both require treatment of the victim and the perpetrator. Framing intimate violence as a more or less private event between two (or more) people 'assigns responsibility for violations to an individual . . . rather than to systemic inequalities or structural violence' (Merry, 2016: 208). It is on the basis of these assumptions that instruments, tools and reporting codes are developed to detect and to assess how individuals can be educated and improved.

Concluding remarks
Domestic violence and child abuse is a highly complex problem that is shaped by both individual and societal factors. It is the result of a range of intersections between an individual and her or his socio-economic and cultural circumstances. This implies that domestic violence and child abuse should be met with a public policy response that reflects this complex interplay and adequately addresses the full range of causes underlying domestic violence and child abuse. In the Netherlands, however, policy responses focus primarily on the individual dimension of this complexity. A trend towards individualization combined with techniques of responsibilization is undeniable; in fact, these have become buzzwords in the welfare bureaucracy throughout Europe (Dubois, 2009).
Needless to say, identifying deficiencies in professional practices is not necessarily objectionable. This form of governance, however, becomes questionable when professionals and their practices are turned into a 'tool' themselves. This 'tool' always must be more productive in a normative and performative sense and can be blamed when policy objectives are not met or tragic incidents occur. Such a regime keeps its focus on the policy approach of intimate violence within the auxiliary professional context, while intersections with structural violence remain absent, thus contributing to maintaining the existing social order. Professionals express critical views about these practices by posing questions and by not, or only partly, applying prescribed protocols in their practice. They indicate that an obsession with detection and reporting leads to 'automatic' reportings that hamper their working practices by distracting them from serious cases that require proper attention. Indeed, as our study of detection instruments shows, a high rate of false positive findings is inevitable, yet undesirable.
Detection policies render domestic violence as a technical problem and may be regarded as a neoliberal, if not a utopian way of dealing with 'uncertainty'. In this context, 'uncertainty' refers to the genuine difficulty to establish with any certainty whether someone is victimized in a violent intimate relationship and, as a corollary, which people are missing in the records. Standardizing the conduct of citizens and professionals through campaigns, a reporting code and protocols must reduce that uncertainty. In practice, however, the interplay of different 'wills' within a regime of deficiency produces ambiguities and problems that subsequently require additional solutions. More and more technical in nature, these solutions seem inevitable in order to manage and control that uncertainty.
For many years now, policy-makers have considered early detection as the 'ultimate solution for coping with risk and uncertainty' (Samimian-Darash and Rabinow, 2015: 4). The exclusion of structural violence from the domestic violence policy assemblage, however, creates a regime of deficiency in which detection policies function only as cultural symbols of a normative approach that aims to control citizens belonging to lower social classes and professionals. This dual role of domestic violence policies in disciplining social workers and their effects on women, particularly poor women and women with a migration background, is characteristic of what we have described as a regime of deficiency. Policy regimes that do not acknowledge let alone address structural violence in society, such as gender-based aspects of intimate violence and intersections with poverty, race or discrimination, and which focus instead on the continuous improvement of professional practices, unfortunately entail gendered effects and racialized injustices, such as the overrepresentation of single mother families and deprived families in registration numbers (cf. Lee, 2016). Effectively, they do not decrease the incidence and prevalence of domestic violence and child abuse.
Arne Mellaard is a PhD student in the Department of Anthropology and Development Studies at Radboud University, working on a project entitled 'Assemblages of Domestic Violence Policies in the Netherlands'. He holds an MA degree in anthropology and an MA degree in public management and policy.
Toon van Meijl is professor and chair of cultural anthropology in the Department of Anthropology and Development Studies at Radboud University. He has published widely on land issues, indigenous rights and cultural identity. At present, he coordinates a research program on the relationship between diversity and inequality, with special attention for issues of citizenship, democracy and dialogue.