Institutional Solidarity in The Netherlands: Examining the Role of Dutch Policies in Women with Migration Backgrounds’ Decisions to Leave a Violent Relationship

: In The Netherlands, women who experience domestic violence can rely on public policies that aim to support them, such as shelters. Drawing on the lived experiences, through 10 interviews and observations, of women with different cultural backgrounds and nationalities staying in a shelter, and on 37 interviews with social workers working with these women, we observed that this support falls short for them. We argue that immigration rules, together with policies on domestic violence and housing, (unintentionally) often work in tandem with violent partners to prevent women with migration backgrounds from leaving violent relationships. The paper draws on a perspective of institutional considerations of solidarity to unpack the relations between domestic violence, cultural constraints, and public policies but looks also at the positive experiences of women of migrant backgrounds with these Dutch policies. This research indicates that there is a lack of institutional solidarity towards women, especially those arriving as marriage migrants, who have experienced domestic violence. In exploring the intersections of domestic violence and often exclusionary state policies, the paper reﬂects on how The Netherlands can provide more support to those women and how intersectional justice and solidarity might be expressed.


Introduction
"My husband knew that if I leave [him], I might ruin everything, so he blackmailed me: "if you leave, you will lose your papers, it's me who decides to send you back to your country", and I was afraid.Because I was new here, I didn't know the rules of this country, I was afraid to go to the police and tell them I'm living in hell" -Patience Patience is a Moroccan woman who had a temporary residence permit to stay in The Netherlands that was dependent on her marriage with her Iraqi husband, who has Dutch citizenship.She eventually left him after experiencing two years of domestic violence (DV).In The Netherlands, domestic violence towards women is one of the most frequent forms of violence [1].DV has, for a long time, been seen as a private issue inside the household.Even if shelters existed, survivors of violence were not protected by Dutch law.This changed in 1984 when, after a sustained period of pressure from feminist advocates and policy makers, the government implemented wide-ranging proposals to fight domestic violence [2,3].This can be seen as a first act of state solidarity towards women who survived DV.
Since 1984, the group of survivors impacted by this policy has become more culturally diverse.Women with migration backgrounds who have experienced domestic violence are overrepresented in Dutch shelters [4]-due to, among other factors, a less extensive social support network [5]-and among them are also women, like Patience, who are not Societies 2023, 13, 243 2 of 16 Dutch citizens.When state policies on DV were developed, gender was foregrounded as a central concern within domestic violence.Cultural differences between women were not considered central to the fight against domestic violence [2].This raises the question of not only whether these women, who might have specific needs because of their cultural and migration backgrounds, receive adequate help from the Dutch state, but also, in the case of those who do not hold Dutch citizenship, whether the state takes responsibility for the wellbeing of this category of survivors.In addressing these concerns, this paper considers whether the Dutch state offers (enough) institutional solidarity to women, especially with migration backgrounds, after experiences of domestic violence.We understand solidarity as support offered to citizens in need [6].This includes support from laws and policies, as well as from professionals mediating between those policies and the citizens to whom they apply.We make use of solidarity as this is what Dutch feminists used to give visibility to, among others, violence against women on the political agenda [7].
This question of institutional solidarity is divided into three themes.The first theme considers the importance of policies and their implementation, focusing on the policy on DV itself and its relation to housing and sheltering: to what extent do women feel supported by these policies?The second theme concerns the role of the professionals, exploring whether women feel supported by their (re)acting.Lastly, we examine the intersection between policies on domestic violence and migration, considering how policies on migration influence the policy on domestic violence.For example, to what extent do women with migration backgrounds-such asmarriage migrants-experience additional barriers due to having a different cultural background and nationality?
The Netherlands is used as a case study as it was one of the first European countries that institutionalised the fight against domestic violence in policies [3].In this paper, we combine two research groups, drawing on interviews with ten women from different cultural backgrounds and nationalities staying in a women's shelter after experiences of violence, and on further interviews with thirty-seven social workers working with survivors.Before diving deeper into the intersections between marriage migration and domestic violence, we also consider how the policy regarding domestic violence impacts all women, including those born in The Netherlands.
In the next section, we discuss the existing literature around cultural expectations, domestic violence, and marriage migration policies, with the latter focusing on the context of The Netherlands.We then outline the methodology before considering three analytical themes that highlight the connections between domestic violence, migration policy, and the challenges of escaping.

Domestic Violence, Culture, Gender, and Migration
The main research group concerns women with migration backgrounds.Although three of the women were born in The Netherlands, we use the heading "with a migration background" due to all of them having at least one parent born outside The Netherlands.By doing this, we follow the official definition of the Dutch Centraal Bureau voor de Statistiek [8] 1 .
DV is frequently linked to culture in the literature [5,[10][11][12].McCloskey et al. [10], for instance, conducted research on gender roles and their intersection with domestic violence in 24 Sub-Saharan African countries.The authors argued that in 17 countries, more than half of the interviewed men believed that a man had the right to be violent towards his wife if she disagreed with him, argued, or left the house without telling him.Those cultural expectations, however, differ based on educational level.And, of course, if these are the results for seventeen out of twenty-four countries, it means that in seven other countries, there are more men holding different views on gender relations.Also, it has been argued that notions of honour and shame may withhold women from leaving violent relationships.Aboulhassan and Brumley [11], for instance, found that Arab American women believed that divorce was a damaging factor to women's reputations and were therefore inclined to stay in violent relationships.
Recent studies in The Netherlands have begun to examine the interconnections between domestic violence and experiences of marriage migration.For example, De Hart et al. [12] emphasised that the migrating partner, often a woman, is dependent on the resident partner, which makes for an unequal relationship between the partners and increases the risk of DV to occur.Bartels [5] focused on Moroccan women living in The Netherlands as marriage migrants and experiencing violence.She argued that both culture and Dutch policies play a role for women in (not) leaving violent relationships.Returning to the country of origin may not be an option for a marriage migrant because of economic factors or because, as a separated woman, especially with a child, people could look down on her and she could be shamed-even if she received support from her family [5,12].
Building on such work, we consider DV the outcome of an intersection of different factors, including not only cultural expectations but also language barriers [13], family structures [14], personal reasons [15], and policy [5,12,16].Researchers in Canada [13] have argued that women coming from countries where they do not have knowledge of the language of the new country, experience language barriers on a structural and personal level.As an example, when they do not understand the language, they find it more difficult to explain their problems to professionals.The authors therefore plead for the need to deliver language-sensitive services so that professionals can reach all survivors of domestic violence.When diving deeper into family structures, Tur-Prats' [14] research states that the perspectives on violence against women depend on the family type, such as who lives with the family, where the members of the family grew up, and other factors.Furthermore, Tur-Prats argues, from a historical view, that there is a difference of perspective between family types-such as stem and nuclear families-and that this influences how violence against women is perceived.When looking at personal reasons to stay with the violent partner, we see that having children, love, and the energy of leaving the situation plays a role [15].Therefore, whilst we acknowledge the role of culture, we are wary of a culturalization of DV, and our critical emphasis is on the role of policy.

Domestic Violence: An Institutional Perspective on Solidarity 1.2.1. Domestic Violence and Institutional Solidarity
In this paper, we follow a feminist sociological analysis, highlighting that domestic violence impacts women especially due to patriarchal structures and unequal power relations [17][18][19].Legally, and culturally, what constitutes domestic violence varies between countries [20].The definition for domestic violence adopted in this paper is that provided by the United Nations [21], as it includes forms of coercive economic violence and is broader than operating only between married partners.The UN defines domestic violence as the following: "a pattern of behavior in any relationship that is used to gain or maintain power and control over an intimate partner.Abuse is physical, sexual, economic, or psychological actions or threats of actions that influence another person.This includes any behaviors that frighten, intimidate, terrorize, manipulate, hurt, humiliate, blame, injure, or wound someone".
The Netherlands has its own definition of domestic violence, which includes physical, sexual, and psychological acts.However, forms of economic violence, such as not having the right to work outside of the household and being denied access to finances, are not considered domestic violence in this context [12].Previous research from de Hart et al. [12] has shown that economic violence occurs in many violent relationships, not least because domestic violence is often multilayered, reflecting a combination of several types of violence.
In this paper, the concept of institutional solidarity in cases of DV is expressed in policies and laws, organisations where survivors can report violence, shelters for those who need them, appointing professionals with specific expertise (i.e., providing expertise enhancement), and developing protocols so that professionals know how to act in case of (a suspicion of) violence [22].

Discretionary Freedom and Domestic Violence
Importantly, such institutions are interpreted not as one body but as various interrelated actors such as housing municipalities, care organisations, or migration administrations.Laws and policies are of importance as they define the playing field in which these institutions operate.Policies, however, are carried out by policy officers and professionalspeople working in education, healthcare, childcare, social and youth support, and justice, or what Lipsky [23] calls 'Street-Level Bureaucracy'.Professionals in these fields are often described as having a discretionary freedom to interpret policy.How laws and policies are conceived may therefore differ from how they are implemented in practice [24].Professionals who deal with (suspicions of) domestic violence have a certain liberty to decide when acting on such suspicions.In The Netherlands, there is a reporting code for professionals in cases where domestic violence is known or suspected [25].According to this code, there are several steps to follow, such as discussing the situation with colleagues and the survivor, but professionals can also reach out to Veilig Thuis 2 if they choose.However, they are not obliged to report (suspected) violence; it is their decision to make, and this produces scope for discretion.General practitioners are an exception, as they have a duty of confidentiality, which means that they are obliged to keep confidential the information provided by patients.They are, however, strongly advised to start a conversation with the person that experiences violence and recommend institutional help [26].When children are involved, GPs have a duty to inform Veilig Thuis.This follows from their duty to protect children; experiencing or witnessing the violence is considered as harmful to children.

The Dutch Policy on Domestic Violence
In 1984, The Netherlands was, as a result of feminist activism and advocacy, one of the first European countries to frame domestic violence as a state issue [2,3], compared to, among others, France in 1994 [27], England in 1996 [28], and Spain in 1998 [3].Since 2009, an Act has been implemented wherein, in cases of acute threats of domestic violence, the police have the right to impose a restraining order of ten days on the violent partner [3,29].This restraining order exists so that the situation can be assessed.This period can be protracted to a maximum of four weeks if necessary.Moreover, in cases where violence already has taken place, a contact ban can be applied.When it is impossible for a woman to stay in the house, it is recommended that they remain in a safe place.In cases where there is a lack of safe accommodation, cities have the right to give housing priority to survivors of violence, a policy called 'housing urgency'.The requirements for such a priority differ between cities [30,31].In principle, every person experiencing domestic violence should be able to apply for 'housing urgency' independent of where they are located [32].Research from Bargellini [29], however, states that this is often more complicated, as there are additional requirements that not every woman in danger can meet, such as a secure income.Moreover, because of the wider housing shortage in The Netherlands, women will more probably first spend some time in a shelter.
When there is no safe place and no possibility to find a new place to live in (due to, among others, being unable to ask for housing priority), and there is no network to rely on, but there is a high likelihood of continued violence, survivors can go to shelters.Shelters are country-wide temporary emergency accommodations, and depending on the city, survivors can stay between six and nine months [29].Research shows that women cannot always immediately go to a shelter.The shelter might not have the capacity to accommodate the women (and their children).Undocumented women and marriage migrants, in particular, may experience rejection from shelters due to both a lack of capacity within the shelters and due to their insecure immigration statuses.In these cases, we see the first examples of how the implementation of domestic violence policy may work out differently for those women born in The Netherlands [5,29], as will be elaborated below.

Domestic Violence and Immigration Policies
A marriage migrant is a person that migrates to the country of the partner with whom they are legally bound, the referent [5,12].De Hart et al. [12] indicated that in 2020, 12,585 marriage migrants received a residence permit in The Netherlands.The majority (62%) were women.This permit is temporary and is recognised as a "dependent residence permit".Until 2012, as a marriage migrant, one could apply for an independent residence after three years of marriage.The Dutch government, however, has raised the period to five years.In addition, the migrant must have a positive result on most of their integration exams completed within the first three years of arrival.If the migrant leaves their partner, or the referent legally ends the relationship, the marriage migrant may have to go back to their country of origin [5,12].As a result, the temporary residence permit leads to a dependency relation between referents and marriage migrants in which the latter is dependent on the former.In the case of domestic violence, such a dependent position can make it more difficult to end or escape the violence.One of the forms that domestic violence can take is coercive control, such as controlling the partner's actions [33,34].Due to the additional dependency that leads to an unequal power relationship because of the temporary residence permit, this risk of control is increased.Dependency in marriage migration in The Netherlands can be found in three areas: legal, financial, and social.Legally, as has been the case since 2013, it is the responsibility of the referent to see to the integration of the migrant and to inform the IND (immigration and naturalisation service) when there are any changes in the residence status of the marriage migrant.These changes can lead to increasing inequalities in agency and power between referents and marriage migrants as the former can more easily withhold information about the latter's residence status [12,35].
During this period, marriage migrants may also be financially dependent on the income of the referent.Whilst they are permitted to work, due to Dutch language barriers and unfamiliarity with the regulations of the country, starting employment is often initially complicated.Since 2013, there has also been more autonomy and responsibility placed on both referents and migrants in relation to integration, as they have the choice of how to prepare for the integration exam testing Dutch values, norms, and language.Integration courses need to be paid for by the referent and the migrant, but they can make use of a loan.As integration courses are not mandatory, neither the referent nor marriage migrant can be punished (e.g., by receiving a fine).It is, however, in most cases, mandatory to participate in the integration exams; otherwise, the referent will receive a fine of a maximum of 4.080 euros [36].There are examples of referents who refused to let their partners take the exams even if this led to fines [12].The consequences for the marriage migrant are that she will experience more difficulties not only in learning the language but also in meeting other people.The lack of Dutch language skills and knowledge about the country can hinder women from gaining knowledge about leaving violent relationships and may increase social isolation and dependency on their partners [12,35].Moreover, without passing the exams, it is almost impossible to obtain an independent residence permit.

In Case of Domestic Violence
If a marriage migrant experiences violence before the end of the five years and wants to leave her partner, she can apply for an independent residence permit based on humanitarian grounds [5].To gain this permit, evidence of violence is necessary.A doctor needs to send (recent) medical data, and a social worker is required to submit a statement of proof of the survivor staying in a shelter [5,12].Research from Vergaert [37] indicates that general practitioners in Belgium struggle with filling in such statements.If they state that the injuries are not bad (enough), women might not receive the permit on humanitarian grounds.On the other hand, they also do not want to submit false reports.This is an area of considerable discretionary freedom [23], as general practitioners and social workers have a significant influence on the future of the women, as they have a certain liberty to fill in the statement.
Alongside data from professionals, recent records from the police are needed.These could be reports of domestic violence, but could also be statements from the police or prosecutor that a prosecution against the violent partner has been started.Reporting to the police itself is insufficient; it is necessary that the police acted with the report or found findings of their own to support a report [12].Previous research has shown that this requirement for action on the part of the police makes addressing abuse more complicated.It is easier to receive the documented proof when there is existence of physical or sexual violence.This can lead to problems for women who experience(d) other types of violence, such as isolation, economic dependency, and psychological violence [5].De Hart et al. [12] compared The Netherlands with other countries where, in case of DV, a survivor can obtain an independent permit.As an illustration, in Germany, domestic violence includes physical and psychological violence.Furthermore, survivors do not need to report the violence; the onus of proof can be of any means.A statement by a friend or family member is sufficient.Compared to Germany, gaining evidence to prove violence is more complicated in The Netherlands.Next to challenging requirements, Bartels [5] shows that marriage migrants do not always know that they can gain a residence permit on humanitarian grounds.Sometimes this is due to language barriers, but also-again-because it is the responsibility of the referent to inform the marriage migrant about their rights.It is thus notable that due to this 2013 policy change, coercion and control by the referent has become easier.In the following sections, the paper explores how these policies put women in vulnerable positions after experiences of violence and how this can change in the future.

Methodology
This research was carried out by a combination of narrative and semi-structured interviews and participant observation.The first author met the women in a shelter in the province of South Holland, where she spent a year and a half undertaking participatory fieldwork.All participants received an information letter and provided their informed consent before each interview.All interviews were open-coded using Atlas.TI.Subsequently, interview fragments were clustered using axial coding.Non-English quotes have been translated.The women and social workers could choose their own pseudonyms at the ends of the interviews.

Narrative Interviews
The narrative interviews with ten women were carried out between June 2022 and March 2023.Each interview started with a question about the participant's life story before arriving in the shelter.More questions were then asked about the support and difficulties women experienced to leave their partners.The aim of using narrative interviews was to give participants the space to tell their stories in a way that made them feel comfortable and able to retain control.The participants differed in age, religion, level of education, and ethnicity.The women were, at the time of the interviews, between 27 and 46 years old, and six of them were born outside of The Netherlands (Table 1).Their cultural backgrounds were recorded based on their own choosing 3 .Regarding educational level, some started working around 14 years of age, and others graduated with a Bachelor's or Master's degree in their country of origin.Four of the women were Hindu, four Muslim, and two non-religious.Most women had voluntarily chosen their partners, except for one case where there was mention of an arranged marriage.Next to this, one woman had been married twice: once an arranged marriage in Suriname and once voluntarily in The Netherlands.In both marriages she experienced violence, but we only focus on the last marriage as it was in The Netherlands.The interviews were carried out in three languages, depending on which the woman preferred, namely Dutch, English, and French.Other languages, such as Turkish or Arabic, were common in the shelter, but the resource constraints of this project precluded their inclusion.As a result, women with little knowledge of Dutch, English or French were unreachable.We decided not to use an interpreter due to the sensitivity of the topic.

Semi-Structured Interviews
Most of the data used in this article have been derived from interviews with the women noted above; however, they are underpinned by additional interviews with social workers working with women experiencing domestic violence.Thirty-seven semi-structured interviews were conducted with these social workers.Two of them were managers, and thirty-five were social assistants of different types: child(ren)-mother, personal, residential, and ambulatory.Child(ren)-mother assistants coach children and their relationships with their mothers.Personal social workers work more individually with women around legal, social, and psychological issues.As an example, they accompany women to visit houses if they receive 'urgency'.Residential assistants work in the shelter, providing first-line support and activities.Ambulatory social workers mostly assist the women from the latter's own places.
All these social workers were women between 22 and 66 years of age, consistently divided.Most of the social workers were working in the shelter where fieldwork was carried out.The other interviewees were working in the provinces of Gelderland, North Brabant, North Holland, Friesland, and Utrecht.This wider pool of participants was contacted through LinkedIn and snowball methods.All the semi-structured interviews were in Dutch.The social workers had different cultural backgrounds, mainly Dutch but also Surinamese, Moroccan, Kurdish, Turkish, Maluku, Tunisian, Brazilian, and Indonesian (Figure 1).Twenty-six of them had degrees in social work.Other common degrees were in psychology and pedagogy.
The information from the social workers for this article is based on questions on the role of the Dutch institutions asked in interviews with a broader purpose.The first author's assumption was that social workers had knowledge of the practice of the policies, as they worked closely with those affected.As such, social workers were asked what they thought should change in the policy regarding domestic violence, especially considering the specific social groups they were working with.Personal and ambulatory social workers had more knowledge about policies because they were closer to the implementation and effects of these policies.
thought should change in the policy regarding domestic violence, especially considering the specific social groups they were working with.Personal and ambulatory social workers had more knowledge about policies because they were closer to the implementation and effects of these policies.

Participant Observation
Several observations made during the year and a half of fieldwork have also been drawn upon in this paper, anonymously.The conversations during observations were in the aforementioned languages with the addition of basic Spanish.Twice a week during the period of fieldwork, the first author helped in organising activities for children and the women, which also gave her the opportunity to have ongoing informal conversations with the la er.

Results
All the interviewed women had experienced policy-related difficulties that had added barriers in leaving their partners.To explore these factors, we focus on three themes.The first issue concerns the policy in practice, with a focus on housing and sheltering.The second topic is about the role of professionals.The final issue is about the intersection of domestic violence and migration policies and how policy change resulted in the state (unintentionally) working in tandem with violent partners to prevent marriage migrants from leaving violent relationships.

Women
The policy regarding domestic violence is broad and multifaceted.A key concern for the women interviewed was accommodation and sheltering.Yasmina, a woman with a Hindu-Dutch background, shared that it had taken her several times before definitely leaving her partner.She had been separated from him before, but had decided to start over because they had already had a child and culturally, it was expected that they should stay together.After several abusive incidents, however, she decided to leave with her children to stay with her mother.As she could not stay there with three children, she asked for housing priority from her mother's place.This, however, was impossible as she did not have a job and therefore had no salary to prove that she could pay the rent, despite having shared a house before.She recounted being strongly advised by social services to stay in a shelter first and afterwards apply for 'urgency', and discussed the negative effects of this policy rule:

Participant Observation
Several observations made during the year and a half of fieldwork have also been drawn upon in this paper, anonymously.The conversations during observations were in the aforementioned languages with the addition of basic Spanish.Twice a week during the period of fieldwork, the first author helped in organising activities for children and the women, which also gave her the opportunity to have ongoing informal conversations with the latter.

Results
All the interviewed women had experienced policy-related difficulties that had added barriers in leaving their partners.To explore these factors, we focus on three themes.The first issue concerns the policy in practice, with a focus on housing and sheltering.The second topic is about the role of professionals.The final issue is about the intersection of domestic violence and migration policies and how policy change resulted in the state (unintentionally) working in tandem with violent partners to prevent marriage migrants from leaving violent relationships.

Women
The policy regarding domestic violence is broad and multifaceted.A key concern for the women interviewed was accommodation and sheltering.Yasmina, a woman with a Hindu-Dutch background, shared that it had taken her several times before definitely leaving her partner.She had been separated from him before, but had decided to start over because they had already had a child and culturally, it was expected that they should stay together.After several abusive incidents, however, she decided to leave with her children to stay with her mother.As she could not stay there with three children, she asked for housing priority from her mother's place.This, however, was impossible as she did not have a job and therefore had no salary to prove that she could pay the rent, despite having shared a house before.She recounted being strongly advised by social services to stay in a shelter first and afterwards apply for 'urgency', and discussed the negative effects of this policy rule: "I am holding this place for another woman with a family who really needs to leave now instead of tomorrow.But the place is occupied because I am here.Whereas I could have also waited for urgency with my mother"-Yasmina Alongside staying in a shelter, there are other suggested requirements as well.As an example, in the policy, it is strongly recommended that survivors fill in a report, as this will make accessing a shelter easier.However, a lot of women shared, during interviews and observations, that they were afraid to report their partners.The women realised that their partners could have temporary restraining orders imposed on them but feared what would happen after this period.They were afraid that their partners would react aggressively after the end of a restraining order.They were also afraid of the consequence of being sent back to their country, which intersected with being a marriage migrant and their lack of knowledge about the migration policy.To be able to receive an independent residence permit on humanitarian grounds, they were obliged to fill in a report.For women with the fear of reporting, this was very conflicting.This was the case for Espoir: "When I had to go to the other shelter, they said I need to report first.I said: "I can't report for my safety", so they said that I will have to go back to Morocco"-Espoir Though women were grateful for the possibility to stay in the shelter and to be able to receive support, most of them shared that they were angry that they had to leave everything behind.Often this involved their partners remaining in their houses whilst the women were in shelters that, as Sija expresses it, "could feel like a prison".The need to stay in a shelter especially affects women of migrant backgrounds as they often have no other place to go due to having less extensive social networks and due to other reasons, such as their families not being willing to have them back, as was the case for Patience: ". ..my father told my mother: "I only accept her dead."Because the honour of the family, it's in a sinking ship".-Patience

Social Workers
Social workers confirmed that while it is possible to ask for housing priority when not in a shelter, this rarely works in practice because of housing shortages.When discussing housing priority, four social workers mentioned having mixed feelings about the possibility of receiving 'urgency' if women are not in a shelter.On the one hand, they fear that it would be too easy to ask for a house and raised concerns over the possibility of abuse of the system.On the other hand, they also acknowledged that there is almost never enough space in shelters for new women who are in danger.Because of this, some needed to wait before leaving or stay in a hotel until there were places for them and their child(ren).Several social workers believed that the best solution was for a woman to stay in the house and for the violent partner to leave but that this rarely happened.The main reason for this is that at the end of a restraining order, there is too great a risk that the partner returns and has access to the house, as Sofia outlined: ". . .eventually the perpetrator goes back to the house and the wife and children have to leave anyway.What you often see is that it is very acute, that they must leave at once and leave everything behind.So, a restraining order, it's good that it's there.It is for a short period, it can be extended if necessary, but it turns out that it is not enough".-SofiaMoreover, social workers pointed out that shelters are not accessible quickly due to bureaucratic processes, such as professionals first assessing the level of danger, a waiting list before there are spaces in shelters, or stimulating women to report to the police.These factors, especially the latter, also have effects on the fears of the women.
"That fear is valid, though.Because even with the first report, you are not immediately placed in a shelter.There is still research to be done"-Cloë Despite these concerns, Cloë noted that there was no obvious solution to these challenges because without clear guidelines and referral processes, abuse of the shelter would be possible.In reality, though, she highlighted that she had almost never experienced women abusing this opportunity.

Women
Professionals had an important role to play for the women who had left their partners and arrived in the shelter.Most of them had found a listening ear for their stories and received advice on how to act.This was the case for all the women, regardless of migration background and civil status.Espoir, for instance, recalls that during her pregnancy, nurses came to her house to change transfusions.She had not been able to speak freely about her situation because her husband had installed cameras in the house, but recalled that the nurses had shown concern: "People came for my infuse and asked why I was always here alone, where my husband was."I find it weird, why are you so sick?Are your in-laws good with you?"They asked"-Espoir Several women told professionals about the violent situations at home but that they wanted to stay with their partners.Mary, a Pakistani woman in an arranged marriage, told her children's nurse about the isolation and psychological violence she experienced from her husband and family-in-law.Afterwards, she asked the nurse to keep this a secret because she wanted to try again, to be a 'better wife', as she explained it.For Mary, it was a matter of gendered cultural expectations that she wanted to follow.She even cut ties with the nurse.The latter respected that choice until several months later, when she heard that the situation had become worse.She then decided to step in and warned Mary that she should leave the house or could lose custody of her children.
"And then the nurse, she called me, and she said: "you must take a step for your children.If you won't, I will go to child custody, and we will take the children"."-Mary

Social Workers
It is thus notable that how professionals (re)act, as mediators of policy, can have a great impact on women's decisions to stay in or leave violent marriages.In cases such as Mary's, several social workers confirmed that many women had arrived in the shelter out of the fear of losing their children.For example, Mona noted: "Look, most of the women are here because Veilig Thuis told them: "If you don't leave that man, we will take your children"."-MonaAlongside cases such as Mary's, social workers argued that most women had arrived in the shelters due to someone else referring them to these institutions after experiences of violence.

Leaving a Violent Relationship as a Marriage Migrant Women
The requirement to be legally bound to the referent for five years [5,12] was known by all the women that had arrived in The Netherlands with this temporary permit, not least because, in many cases, their partners had used this as a threat to keep them in the relationships.For the women, this was one of the main reasons they had endured the violence longer than they would otherwise have chosen.For example, Prija stated: "When I was finally allowed to naturalise for my Dutch citizenship, he [told the IND that they had to] stop my residence permit [suggesting to the IND that the relationship was ending].So, everything then lapsed.It's been lost violent years." Prija can now stay in The Netherlands as her child was born in the country.However, she cannot receive a permanent residence permit anymore on grounds of marriage migration.For a certain period of time, she can stay in The Netherlands due to her being a parent to a child born in the country.Prija also shared how surprised she was that the authorities had not asked for her side of the story.This was not only the case for her, but can also be applied to other marriage migrants, indicating how much power the referent has as the figure with a responsibility to inform the authorities about the status of their spouse.
Fulfilling an integration exam is an additional mandatory requirement to obtain Dutch citizenship [12].However, four of the women interviewed had not been allowed by their partners to take civic integration courses or do other activities.This was notable, as they had arrived in the shelter with little or no knowledge of the Dutch language and of how the country works.None of these women had been in The Netherlands for three years, and so, they were not yet obliged to make an integration exam.In the cases of Mary and Espoir, they had only been allowed to go grocery-shopping or leave the house for medical visits, and their mothers-in-law had to accompany them.They realised that these were ways to control them and served to restrict their ability to learn about their rights.Next to not being able to go out on their own, they had also not been allowed to have contact with people they knew or people that they may have encountered in their everyday lives in or around their own houses.As an example, Espoir shared that after a while, she felt guilty for opening the door to the mailperson, as her partner had accused her of flirting with him.Also, five of six marriage migrants had not been allowed to open their own bank accounts.As some had been forbidden from attending integration sessions, they did not know how to do so for themselves.The excessive control experienced by these women meant that they had not met new people or gained knowledge of how The Netherlands is structured, further restricting the possibility of developing an awareness of DV and how to escape it.
In such a context, it is not remarkable that the women did not have any knowledge of the possibility to apply for a special permit on humanitarian grounds.For five of the women arriving as marriage migrants, the fact that it is the partner who is the gatekeeper to this knowledge had led to more control over them.It was only in the shelter that most of the women heard, for the first time, that they could apply for a humanitarian permit.It brought them even more stress when they heard of the requirements to receive a residence permit on humanitarian grounds because proof of physical and/or sexual violence might already have disappeared.All of them had proof of violence on their phone, such as videos and messages, but this does not count as valid proof as it could have been edited.For several of the women, such as Mary and Prija, they were subjected to mainly, but not exclusively, psychological violence and isolation, experiences that were hard to provide this form of physical proof for.All of this led to extra stress and insecurity for the women.

Is There Enough Institutional Solidarity towards Women?
Is there institutional solidarity towards women that have experienced domestic violence in The Netherlands?Yes.The fact that there is a policy about it proves that the state wants to help survivors of violence.What we learn here in the context of The Netherlands is that the policy has been developed out of gender concerns and feminist activism.By doing so, the country was one of the first European countries to show their support to women experiencing violence.Over the years, the policy has been broadened by including the possibility of receiving an independent permit on humanitarian grounds for, among others, women arriving as marriage migrants.These groups of women can thus also apply for support from Dutch government institutions in the case of DV.There are, however, still points for improvement that are experienced by the women.
First, in the study, women felt discouraged to leave their partners as they had to leave their homes behind or found it extremely difficult to meet the requirements of urgent housing from a temporary safe place, a point also noted in Bargellini's research [29].For women with migration backgrounds, leaving the house can be even more difficult if their families do not want to take them back, in case of cultural gendered expectations, or if they have less of a network in The Netherlands.A restraining order can be imposed upon their partner, but there are requirements for this, such as reporting the partner, which a lot of women fear to do.Such restraining orders are also often limited to ten days (with exceptions involving four weeks, if necessary) and are designed to allow for an assessment of the situation, and can lead to insecurity if the partner will come back.This brings additional barriers as there is also a lack of capacity in shelters.Because of this, some women must wait before they can be accommodated, notwithstanding the temporary nature of shelter accommodation.Furthermore, when adding the component of marriage migration, those women can experience even more difficulties.
The women shared that they received support from professionals that worked for the state and who played an important role in putting policy into practice.Professionals made use of their discretionary freedom [23] by, as an example, questioning situations where there were suspicions of domestic violence, which was the case with Espoir during her pregnancy.This discretionary freedom was also used to warn women to leave their abusive partners by saying that they could lose their child(ren).Whilst a challenging practice, this was discussed by the social workers as being carried out with the best intentions and out of an idea of solidarity; however, in cases involving women with migration backgrounds, this practice might not always respect their cultural values.Moreover, as some women feel coerced to make the 'right' choice, it also disrespects their autonomous agency.Nevertheless, the professionals were generally appreciated by the women for being very supportive and making the most of the discretional freedom to apply the policy.
Due to migration policy, marriage migrants must legally be bound with their partner at least five years before receiving independent Dutch citizenship [5,12].This already leads to a power inequality in which the referent is the one being able to control the permit, but in the case of violence, this can be even worse, leading to contradictions in the solidarity that the state offers to survivors of DV.Hence, the dependency that comes from the policy presents a way for the referent to further control their partner and coerce the woman to stay with them.As integration became the responsibility of the referent and the marriage migrant, when the former does not want to pay, the migrant has fewer chances of meeting people and learning about the rules of the country.What we see here is that by fully placing the responsibility for integration on the referent, the state ignores any caring responsibility towards marriage migrants arriving as newcomers.
On the one hand, The Netherlands has enacted a possibility to receive an independent residence permit on humanitarian grounds, demonstrating, in theory at least, an act of solidarity.However, in practice, the requirements are complicated as physical or sexual proof of abuse is necessary, which is not always easy to establish [12].Social workers and general practitioners can indicate that violence has taken place to reinforce their possibility of obtaining an independent permit, but, as Vergaert [37] has discussed, it can also be challenging for them to establish exactly what has happened.Furthermore, women do not always know about the possibility of such a permit and of participation in civic integration courses as access to this knowledge is the responsibility of the referent.Here, referents may act as gatekeepers to such knowledge.In this context, even if there are possibilities for survivors of DV to ask for independent permits on humanitarian grounds and the integration courses and exams exist, the state can be said to (unintentionally) be empowering the partner at the expense of the marriage migrant.
As it stands, despite the Dutch state's intention to offer protection to survivors of DV, if these survivors are not (yet) Dutch citizens, such policies fail in their ambition.This is especially the case for women arriving as marriage migrants, where the control by referents is made possible by the Dutch integration policy, further entrenching potentially damaging power inequalities.Alongside this, the state's abdication of any responsibility to check on the wellbeing of newcomers further allows referents to use the lack of knowledge of marriage migrants to strengthen power over them.

Recommendations
Several issues that the women raised through this work are fundamental social concerns that extend beyond DV policy alone, such as a country-wide housing shortage.There are, however, small changes that are possible within DV policy that will make a positive difference.An example is by offering urgent housing to a survivor of violence that can stay at one of her network's places.On paper, this should be the case, but we see in practice that it is not always enacted.This would lead to a faster transit in the shelter so that there is more space for other women in need of leaving a violent situation.Next to this, restraining orders should be stricter and longer so that the survivor can stay in the house for a longer time and feel secure.Meanwhile, both survivors and perpetrators (and any other included in the violent situation) should receive adequate social support and help.This requires, at a minimum, that all parties are aware that such support exists, which, as we saw, is not always the case with marriage migrants.Therefore, it is the duty of the state to inform the marriage migrant, at their arrival, about their rights.Practically, this could take the form of a letter in multiple languages.
All the interviewed women had received support from professionals.In cases of marriage migrants, this was necessary as they had not been aware of the help they could get from the state; since it is the responsibility of the referent to tell them, as such, there is no control of this key information from the state.This lack of control is exacerbated because integration courses are not mandatory.A recommendation is to make the lessons mandatory again and paid for by the state, and, in case of migrants having children, to organise childcare to facilitate attendance.Another possibility is to ensure that social workers are able to support marriage migrants when they are not attending integration courses and to address potential barriers to their attendance.The courses would increase the likelihood that marriage migrants learn the language and learn about their rights.Another improvement would be, as suggested by the Dutch platform of Self-Determination and Residence Rights ("Platform Zelfbeschikking and verblijfsrecht") [38], that there is easier access to the permit on humanitarian grounds when there is any question of domestic and sexual violence.A concern might be that this leads to false claims about domestic violence; yet, the protection of survivors of violence should be prioritised.Not only would this be positive for migrant women if there were negative situations at home, but it would also make the Dutch state's ambition more substantive to show institutional solidarity to women of migrant backgrounds who are victims of domestic violence.
Taking such debates forward not only involves questions for Dutch policies on integration, domestic violence, and the contemporary welfare state but also poses challenges for Dutch feminist organisations.As noted, these groups were central to shaping policies towards domestic violence in the 1980s and beyond, and today face the critical challenge of addressing the intersection between domestic violence and migration.It is only by facing the tensions between policies of protection for women of all backgrounds and immigration statuses and policies of migration control and citizenship that solidarity can be meaningfully enacted and expressed, both in policy and in practice.

Limitations and Further Research
This article also has limitations that we believe could be researched more deeply in the future.First, the scale of this qualitative research was small, focusing on a small group of women in one shelter.However, this type of data cannot be developed by quantitative methods, such as surveys, because these lack the ability to build the personal trust that is essential in working through the sensitivities and complexities of DV.A second limitation is that women without knowledge of French, Dutch, and English during interviews, and additionally Spanish during conversations in the shelter, were not able to participate, as the first author does not have knowledge of other languages.Recently-arrived marriage migrants, or women who had been isolated and had never learned the Dutch language, were therefore excluded.However, we anticipate that they are likely to have experienced the problems indicated in this paper; indeed, it is likely that these problems may have been exacerbated by their increased isolation and lack of mastery of the Dutch language.A third limitation is that our research focused on the experiences of women while still in the shelter and gained no knowledge of the (lack of) institutional solidarity they experience after leaving the shelter.
Considering these limitations, it would be valuable for future work to repeat this research on a wider scale and with a broader range of common languages included, such as, alongside the aforementioned languages, Turkish or Arabic.Similarly, research that investigated the forms of institutional solidarity towards marriage migrants after leaving the shelter, including how the legal procedure for receiving an independent permit on humanitarian grounds proceeds and which factors may bring additional stress due to legal uncertainty, would offer additional and important insights to further these discussions and inform future policy development.

Figure 1 .
Figure 1.Self-identified cultural background of the social workers.

Figure 1 .
Figure 1.Self-identified cultural background of the social workers.

Table 1 .
Women's nationalities, self-identified cultural backgrounds, ages, and durations of stay in The Netherlands.