Introduction

When women decide to leave a violent relationship, they can be influenced by their children’s exposure to domestic violence. Realizing the harm their children risk suffering because of the violence, they might try to protect their children by leaving the relationship and seeking support (Rhodes et al., 2010; Zink et al., 2003). Many women who have been abused move with their children to domestic violence shelters for protection and to escape the violence, but they also tend to need social support, health care, and legal support (Sullivan & Virden, 2017). Research stresses that the factors that women value the most are safety and support for their children (Jonker et al., 2014), and these are also among the main reasons women return to the shelters with their children (McFarlane et al., 2016). However, research has mainly focused on women’s experiences of shelters and their life situation and social relations while there (Jonker et al., 2014; Øverlien et al., 2009; Wessels & Ward, 2016). This suggests there is a need to investigate the situation for children specifically. Shelter staff are also sometimes asked about their work and how they experience the residents’ situation, but even in these cases it is mostly the mothers’ situation that is in focus (e.g., Fleckinger, 2020). Nevertheless, living with domestic violence affects children’s health (Annerbäck et al., 2012; Holt et al., 2008), social relations (Överlien & Hydén, 2009; Selvik & Øverlien, 2015) and preschool and school attendance (Selvik & Øverlien, 2015). Moving to a shelter could have further impacts on these areas, as for many children it involves leaving their social context and familiar environment (Selvik & Øverlien, 2015).

Selvik and Øverlien’s (2015) review of data on children in domestic violence shelters in the Nordic countries has contributed to a better understanding of children’s life situation in shelters. Although their 2015 review gives insight into the children’s own perspectives and about children in domestic violence shelters, the subject has gained more research attention in recent years. Given the limited research on the subject, there is also value in doing a broader search that includes countries outside the Nordic region. Children might experience similar challenges regardless of country of origin, or good examples might be identified that can be learned from.

Our aim is therefore to review existing research on children in domestic violence shelters, specifically in order to examine previous research on how shelters contribute to children’s life situation after leaving a violent home. We will employ a thematic analysis and discuss the results from a children’s rights perspective to identify strengths and limitations in existing research and social work practice.

Children’s Rights Perspective

According to the UN Convention on the Rights of the Child (1989), UNCRC, Article 19, every child shall be protected from violence, and according to Articles 2 and 3, all children have the same rights and value, and their best interest shall always be considered in cases involving them. All children also have a right to life and development (Article 6), as well as to express their opinions and feel that they are respected (Article 12). The convention further stresses the importance of children’s right to education, rest, relaxation, play and participation in cultural and creative activities (Articles 28 and 31). The responsibility to ensure that children receive the rights to which they are entitled is shared between the legal guardians (usually the parents) and the state. This means that if the parents cannot fulfill their obligations, the state shall act to guarantee that the child receives the protection and nurture that it needs. However, the convention has not been incorporated into law everywhere, and its standing in relation to national legislation can therefore differ between countries. As a result, there can be a certain degree of conflict between children’s and parents’ rights (Quennerstedt, 2009), and in many cases where decisions have been made concerning children, the emphasis has been on the parents’ rights. This suggests that a children’s rights perspective is needed to understand how living at a domestic violence shelter affects children’s lives, relationships, and well-being, and how professionals can support children fleeing violence and secure their rights.

Method

This review compiles research articles from two databases on children at domestic violence shelters in a systematic integrative review. An integrative review is exploratory in the sense that it integrates different kinds of studies, such as qualitative and quantitative studies, to gain a broader picture of the existing knowledge (cf. Booth et al., 2016). The main search was conducted in September 2020. A narrower search followed in February 2021 to complement the first one with a specific focus on children’s own perspectives. A further search was also conducted in February 2022 to update the searches with articles in the two databases that had been published since the previous searches.

Search Strategy and Procedure

The literature searches on children in shelters were done in the databases Social Services Abstracts (ProQuest) and PsycINFO (EBSCO). These databases were chosen for their focus on social work, criminology, psychology, and public health. These are all academic fields that may examine domestic violence and its effects on children, and especially children’s experiences of living in domestic violence shelters.

Inclusion and Exclusion Criteria

In Social Services Abstracts, the search function “anywhere except full text (NOFT)” was used. To be included, the studies needed to be peer-reviewed articles and be written in English, Swedish, Norwegian, or Danish (i.e., languages the authors could understand, although only English search terms were used). No exclusions were made based on publishing date for the first two searches. In the third search, the publication date was restricted to between September 2020 and February 2022. We kept studies that focused on children in domestic violence shelters and their experiences of living there. We also kept articles that mainly focused on the mother or the shelter environment, for example, in order not to miss any articles that discuss the children’s life situation while living in a shelter. We further kept articles that emphasized the mothers’ or the shelter staff’s perspectives, if they focused on the children’s situation. Articles focusing on adults’ experiences of violence or living at a shelter, and discussing other reasons for living at a shelter than domestic violence (e.g., homeless shelters) were excluded. Similarly, articles that focused on various therapeutic or treatment methods used at a shelter, but did not discuss them in relation to children’s experiences of living there, were excluded.

First Search—Main Search

The first search was conducted on September 16, 2020. The search terms used were: (child* OR Young People OR Adolescent) AND (domestic violence OR Intimate partner violence OR IPV) AND (shelters OR Emergency shelters OR Shelters for battered women OR Sheltered residence). The search resulted in a total of 758 hits in Social Services Abstracts and 359 hits in PsycINFO. Based on the titles, these were reduced to 108 and 48 hits respectively. Further studies were then excluded based on the abstracts, and duplicates were removed. Thirty studies then remained from the search in Social Services Abstracts and 16 studies from PsycINFO, including one study in Danish. No studies that met the inclusion criteria were excluded based on the language criteria. The remaining 46 studies from both databases were read in full text. During this phase, the quality of the studies was checked in accordance with CERQual (SBU, 2020). After this step, 19 studies were left for the analysis. In addition to these, further research articles were included after searching the reference lists of already included articles, a method called pearl growth (Booth et al., 2016; Petticrew & Roberts, 2008). This method is used in systematic reviews, including integrative reviews, to widen the search and identify relevant articles that may have been missed in the main search. The pearl growth used the same inclusion and exclusion criteria as are described above. Based on their titles or how they were referenced in the texts, 28 articles were compiled in a list. Based on a reading of the titles and abstracts, two further articles were read in full text, and one was included for analysis, suggesting that the main searches did capture most of the studies on the subject in the two databases used in the review.

Second Search—Complementing with Research on Children’s Own Perspectives

To try and identify more studies that focus on children’s perspectives of living in domestic violence shelters, we undertook a narrower second search in Social Services Abstracts on February 4, 2021, which resulted in 191 hits, and in PsycINFO on February 11, 2021, which resulted in 98 hits. The search terms were (Adolescen* OR Child* OR Youth OR teenage*), AND Refuge, AND violence. Based on the titles, 25 studies were included from Social Services Abstracts and three from PsycINFO. On title level, two studies in German, two in Spanish, and one in Portuguese were excluded. Eight were selected to be read in full text and quality controlled using CERQual (SBU, 2020), and four of these were included in the analysis. No further articles were added through pearl growth based on the articles’ reference lists.

Third Search—Complementing with Research Conducted After the Previous Searches

A third complementary search was conducted on February 24, 2022 to search for further articles published in the two databases since the main search. In this search, the two previous search strategies were repeated separately. First, the search terms from the initial search were used for the period September 2020 to February 2022. This resulted in 24 hits in Social Services Abstracts and 17 hits in PsycINFO. Secondly, the search terms from the second search were used for the period February 2021 to February 2022. This resulted in four hits in Social Services Abstracts and five hits in PsycINFO. With the exception of restricting the publications to the period February 2021 to February 2022, the same inclusion and exclusion criteria were used. Based on titles from both databases, 19 abstracts were read, and of these 10 were selected for full-text reading. After reading the full texts and performing a quality control check using CERQual (SBU, 2020), four articles were added to the analysis. No further articles were added through pearl growth. In total, the three searches and the pearl growth together resulted in 28 included studies Fig. 1.

Fig. 1
figure 1

PRISMA flowchart illustrating the inclusion process

Full size image

Analytical Framework

The review used a thematic analysis based on Braun and Clarke’s (2006) six steps. To get an overview of the 28 studies included in the analysis, a table was constructed that includes each study’s title, aim, method, main results regarding children’s experiences, and country of origin (see Table 1). The table also served as a summary to enable the authors to get an overview of potential themes for the analysis. The first step in the thematic analysis was for the authors to familiarize themselves with the included articles by reading them multiple times and taking notes (Braun & Clarke, 2006). Based on the reading, in the second step several initial codes were identified, which were then summarized into themes in the third step. The authors first conducted steps one to three separately; then they met to discuss the codes and themes that had emerged and the arguments for and against certain codes, which resulted in some revisions. This was done to reduce bias in the coding process of the analysis. Overall, however, the authors were in agreement about the codes and themes that they had initially identified. Based on these discussions the authors continued to the fourth and fifth steps. These involved reviewing and naming the themes, which are as follows: (1) safety and security; (2) health, behavior, and support; (3) schooling; (4) spare time and shelter environment; and (5) social relationships. Some of the results could fit into more than one theme, and hence the analytical procedure involved a back-and-forth movement between the studies, the coding process, and the analysis (e.g., Braun et al., 2006). In the sixth step, the themes were analyzed from a children’s rights perspective using the UNCRC (1989) to place the described children’s experiences in a larger context. The children’s rights perspective was the theoretical basis according to which the identified themes were interpreted. During the analysis process it enabled the authors to view the children as rights bearers with their own needs, and not just as people accompanying their mothers to the shelters.

Table 1 Children in sheltered housing
Full size table

Results

This section presents the 28 studies included in the review. The studies are from the USA (n = 10), Norway (n = 7), the UK (n = 3), Canada (n = 2), Denmark (n = 2), Australia (n = 1), Israel (n = 1), Portugal (n = 1), and the UK and Australia combined (n = 1). The studies span more than three decades (1984–2021). The analysis from a children’s rights perspective will follow in the next section—Discussion.

Safety and Security

Several studies mention safety and security at the shelter as being of great importance. Even so, some rules and restrictions that shelters impose to prevent further violence can be seen as difficult to accept (e.g., Jarvis et al., 2005). These can include no use of violence within the shelter (Mullender et al., 1998), keeping one’s stay at the shelter a secret from anyone outside the shelter (Øverlien, 2011a), respecting curfews and not leaving the premises (e.g., Chanmugam, 2011; Gregory et al., 2021; Øverlien, 2011b), and restrictions on teenage boys (Mullender et al., 1998; Theobald et al., 2021). Chanmugam (2011) highlights that security features in the shelter environment such as fences, locks, security windows and curfews are meant to protect the women and children living there, and hence are important for the safety and security of the residents. However, according to several of the children in that study, these features also give the shelter a ‘prison like’ feeling. In multiple studies, children also perceived the rules as confining and difficult to accept and adapt to (Chanmugam, 2011; Jarvis et al., 2005; Øverlien, 2011a).

Studies also show that shelter stays could be characterized by instability, unpredictability, and an overall sense of insecurity about the future, and that it can be hard to establish routines and a sense of normalcy (Bennett et al., 1999; Chanmugam, 2011; Gregory et al., 2021; Hauge & Kiamanesh, 2020; Øverlien, 2011b). Still, several studies emphasize that children describe feeling safe at the shelter and happy or relieved to be away from the abuser and the violence (e.g., Jarvis et al., 2005; Ornduff & Monahan, 1999; Øverlien, 2011b; Selvik, 2020; Vass & Haj-Yahia, 2020). It is likely that the security features and rules contribute to the children’s sense of safety (e.g., Theobald et al., 2021); however it is important to bear in mind what features and rules are needed, and how their negative effects can be minimized. To enhance the children’s sense of security as well as their sense of agency, it could be important to inform the children in an age-appropriate way about the shelter’s function as a safe haven from violence (Mullender et al., 1998). Not informing them about where they are and why can lead to confusion, as can be seen in Øverlien’s (2011a) and Vass and Haj-Yahia’ (2020) studies. If children feel confined at the shelter and do not understand the necessity of the rules, it is likely that they will find it difficult to adapt to life there.

Øverlien (2011b) and Selvik (2020) show that some children felt less safe outside of the shelter premises due to fear of meeting the abuser. Children who do not want to leave the shelter for this reason are at risk of becoming isolated. A pair of studies highlight strategies that can be used to prevent this. In Chanmugam and Hall’s (2012) study, shelter staff stressed the importance of conducting safety planning together with the children to prevent future violence, and Selvik (2020) describes the importance of doing the same with school personnel and involving the children to make them feel safe at school. This can lead to a stronger sense of security outside the shelter as well, which is needed to make the children’s leisure time and schooling secure. In summary, safety and security are of importance, and to minimize the negative effects on children of the stay, it is important that they receive adequate information and are involved in decision-making processes concerning them as much as possible.

Health, Behavior, and Support

Many children staying at domestic violence shelters exhibit behavioral problems. These behaviors may initially increase when they arrive at the shelter only to later decrease. According to Fredland et al. (2014), this might be related to the shelter being a safe place where the children can externalize their emotions related to the violence (see also Copping, 1996; Troensegaard, 2014). Similarly, Pinto et al. (2019) show that children living in shelters exhibit more internalizing and externalizing symptoms than children who continue living with the abuser, suggesting that the symptoms are a result of the loss of their social network. There also seems to be an association between mothers’ depression and children’s problematic behavior both during and after their stay at a domestic violence shelter. Long-term support seems to benefit children, either directly or because the mother’s improved health and parenting skills increases the children’s well-being (Jarvis & Novaco, 2006).

The support available at shelters mainly seems to focus on helping the mothers, who in turn support their children (e.g., Bennett et al., 1999; Bunston et al., 2020). Bowyer et al. (2015) show that some children view their mothers as vulnerable, which makes them reluctant to talk about their own emotions to protect their mothers. Therefore, interventions directed towards children have been stressed for quite some time as a way to prevent emotional and behavioral problems (e.g., Copping, 1996; Jarvis & Novaco, 2006). Structured counseling has been shown to be appreciated by children in several studies (Bowyer et al., 2015; Mullender et al., 1998; Øverlien, 2011b). Also, Theobald et al. (2021) describe that shelters can collaborate with other organizations to provide tailored support for specific children. So, when it comes to improving the life of children at domestic violence shelters, the results from the studies show the importance of offering long-term support and different forms of support, for both children and their mothers.

Schooling

From as early as 1984 until as late as 2021, studies have reported barriers to children’s schooling when living in domestic violence shelters. These barriers concern arranging transportation to schools in other districts, difficulty getting the children admitted to new schools because of a shortage of available school places, enrolment delays due to missing documents, and concerns about children’s safety at school. Obstacles like these have caused children to have periods of absence (e.g., Bracewell et al., 2020; Chanmugam et al., 2015; Johnson, 1984; Mullender et al., 1998; Øverlien, 2011b, 2012; Selvik et al., 2017; Vass & Haj-Yahia, 2020). Most children that moved to shelters had to leave their school of origin and reported being absent from school for up to 9 months at a time, and many were subjected to multiple shelter stays that led to repeated periods of absence (Bracewell et al., 2020; Chanmugam, 2011; Øverlien, 2012; Selvik et al., 2017). Taken together, this means that in the worst case, children could miss out on years of schooling.

Some children can continue to attend their school of origin, as usually it is their wish even when it involves having to travel a long distance (e.g., Bracewell et al., 2020; Selvik et al., 2017; Theobald et al., 2021). However, this means that the children’s time is taken up by school and traveling, and that they have little time for themselves or for doing homework (Selvik et al., 2017). Children who cannot participate in regular school activities can sometimes receive distance education, education at the shelter, or a personal tutor for a short period of time (Bracewell et al., 2020; Chanmugam et al., 2015; Selvik et al., 2017). On one hand, Chanmugam et al. (2015) study, which is based on the perspectives of shelter staff in the USA, shows that staff are pleased with the education the children are offered while living at the shelters. On the other hand, Selvik et al. (2017) study from Norway and Bracewell et al. (2020) study from the UK show that children are less positive about their schooling, with many describing that they received no or only insufficient support. It is uncertain whether the differences in the staff’s and children’s experiences are affected by differences between shelters, countries or perspectives.

Studies also indicate that having to attend a new school, and especially multiple new schools in connection with multiple shelter stays, can be difficult and tiring for children because they must adapt to new environments, teachers, and classmates (Bracewell et al., 2020; Selvik et al., 2017; Vass & Haj-Yahia, 2020). However, returning to one’s school of origin after a long absence can also be challenging (Vass & Haj-Yahia, 2020). The missing school time, switching between schools, and different curricula at different schools can lead to knowledge gaps that impact children’s academic achievement and social relationships in the school context (Bracewell et al., 2020; Selvik et al., 2017; Vass & Haj-Yahia, 2020). Selvik et al. (2017) and Bracewell et al. (2020) stress the need for more support, private spaces in shelters where children can do homework, and educational resources such as computers, internet access, and books. Similarly, they argue that teachers rarely acknowledge the children’s situation or offer support to help them succeed at school. However, some children take the initiative themselves to seek emotional and educational support from teachers and other school staff, and experience a sense of relief when they receive it. Still, many children are not comfortable talking about their living situation at school, which leaves them dependent on adults recognizing their need for support and offering it (see Selvik et al., 2017; Bracewell et al., 2020; Vass & Haj-Yahia, 2020).

None of the included studies report on the parents’ contact with the school; however, studies show that there is little or no communication between the shelter and the school (Bracewell et al., 2020; Mullender et al., 1998; Selvik et al., 2017). According to Mullender et al. (1998), the shelter staff find it important not to interfere with the mothers’ communication with the school. They stated that it was difficult to establish and maintain relations with all the schools the children received placements in. Still, several respondents had experiences of shelters that had good collaborations with local schools and specific individuals who facilitated them (see Mullender et al., 1998; Chanmugam et al., 2015). Good collaborations can improve children’s chances of having their educational needs fulfilled. To summarize, the children’s school situation is of great importance, and when their educational needs are not met, their social relations and academic achievement might be adversely affected.

Spare time and Shelter Environment

Living in a shelter can make it difficult for children to engage in after-school activities (Øverlien, 2011b), and activities offered by the shelter are therefore important for the children’s ability to have meaningful spare time. For example, studies show the importance of good recreational and play areas (Bennett et al., 1999), and of suitable activities or recreational areas specifically for teenagers, such as a quiet TV room or private space (Bowyer et al., 2015; Chanmugam, 2011; Mullender et al., 1998; Øverlien, 2011b). Studies further show that children, especially those with multiple shelter stays, appreciate the playrooms and playgrounds (Chanmugam, 2011; Mullender et al., 1998; Øverlien, 2011b). Children also appreciate structured activities that shelters arrange for them, such as trips to swimming pools, go-cart tracks, ski slopes, and summer camps (e.g., Mullender et al., 1998; Øverlien, 2011b; Vass & Haj-Yahia, 2020). However, such activities do not occur every day. Therefore, activities offered in the shelter environment are important, and some shelters also invite children who have recently left the shelter to participate (Mullender et al., 1998). Øverlien (2011b) highlights the need for secure outdoor spaces so that children do not have to spend all their spare time indoors for security reasons. In summary, the studies show that the activities offered at the shelters are appreciated and important, but that there is a need for more suitable activities, play areas and recreational areas for children, especially teenagers.

Social Relationships

Moving to a shelter can lead to a loss of contact with family members, relatives, and friends (Bowyer et al., 2015; Chanmugam, 2014; Gregory et al., 2021; Ornduff & Monahan, 1999; Øverlien, 2011a; Selvik & Thjømøe, 2021; Vass & Haj-Yahia, 2020). Both Mullender et al. (1998) and Øverlien (2011b) show how siblings are sometimes separated, with some being left with the abuser, some choosing to stay with the abuser, or some choosing to live with relatives. Families with teenage boys seem to be especially affected by this, because some shelters have rules excluding them from being placed at the shelter (e.g., Theobald et al., 2021). The children’s feelings towards their abusive father tend to differ; some miss him, while others do not. However, this does not need to reflect on their wish to stay at the shelter. Some are happy to be away from the abusive father even though they miss him, while others want to move back home, or at least be allowed to have contact with their fathers (e.g., Bowyer et al., 2015; Henze-Pedersen, 2021; Ornduff & Monahan, 1999; Øverlien, 2011b, 2012). The relationship between mother and child is also affected, as studies show that mothers are sometimes unable to take care of their children when they arrive at the shelter, and this can make teenage children adopt a care-taking role for their mother and their siblings (e.g., Chanmugam, 2014; Mullender et al., 1998). Henze-Pedersen (2021) describes how the shelter environment can enable children and mothers to build on their relationship. However, the same study also shows that the unfamiliar atmosphere, and the lack of possibilities for children and mothers to do activities together such as cooking, playing, or going to the park, can make it challenging to deepen the relationship. This may be related to the mothers being in a state of crisis during the early part of their stay at the shelter, but it can still be possible for mother–child relationships to develop because of the support and parenting help that shelter staff provide and activities that the mothers and children do together.

Further, staying at a shelter seems to put a strain on social relationships outside the family, such as friendships (e.g., Selvik & Thjømøe, 2021). Øverlien (2011a) describes how children find it difficult to make new friends outside the shelter, as they are not allowed to tell anyone they live in a shelter. A lack of understanding of their situation and feelings of shame are also reasons why older children in particular find it hard to make new friends outside the shelter (Chanmugam, 2011; Øverlien, 2011a; Vass & Haj-Yahia, 2020). This can lead to feelings of loneliness. Still, social media and internet can be helpful resources, facilitating socialization with friends (Vass & Haj-Yahia, 2020), and without them, children may miss out on possibilities to socialize (see Øverlien, 2011b). However, several studies also report that children in shelters find friends in each other, and that this reduces feelings of isolation (e.g., Chanmugam, 2011; Mullender et al., 1998; Øverlien, 2011b). While this is a positive experience, Øverlien (2011a) points out that it makes moving out of the shelter emotionally difficult as well, as the children must experience another loss of friendship. Based on the studies presented in this section, the children’s social relationships seem to be affected by living at a shelter, and their social life appears to be characterized by multiple separations as well as by difficulties in creating new relationships outside of the shelters. Still, they establish relationships with the people at the shelter, which can be experienced as positive.

Discussion

The aim of the study was to review existing research on children in domestic violence shelters, and more specifically to examine previous research on how shelters contribute to children’s life situation after leaving a violent home. Based on 28 studies, five themes were identified relating to children’s situations in domestic violence shelters. Safety and security emerged as an important factor and main theme that influences other aspects of the children’s lives, such as schooling, spare time, and social relationships. Even though the domestic violence shelters offer a safe space, free from violence, they have safety rules that the children find difficult to live with, such as not being allowed to tell anyone they live in a shelter or to bring friends to visit. The rules are there to protect the residents and make them feel safe, but the children’s social relationships can be adversely affected by them, especially when it comes to the older children. Similarly, it was common for the children to be absent from school for shorter or longer periods for several reasons ranging from their need for protection to a lack of transportation or available school places. For safety reasons, spare time activities outside the shelter were also sparse, and structured activities organized by the shelter, such as trips to a swimming pool, became important instead.

From a children’s right perspective, safety and protection from violence (Articles 19, UNCRC, 1989) are important for children who flee from violence to a domestic violence shelter. However, research shows that in many cases children are viewed as an accompaniment to their mothers rather than people with their own rights and needs (cf. Bunston et al., 2020; Øverlien, 2011a). Children’s needs regarding, for example, social relationships, interventions to help them process the violence, and education often come second to their mothers’ needs (cf. Chanmugam, 2011; Johnson, 1984; Mullender et al., 1998; Øverlien, 2011b; Selvik et al., 2017). It is therefore important to take the children’s perspectives and opinions about the situation into account (Article 12, UNCRC, 1989), especially as several studies show that children’s situations and relationships are affected by staying at a shelter (e.g., Chanmugam, 2011; Øverlien, 2011a; Vass & Haj-Yahia, 2020). In relation to this, it is also important to acknowledge all children’s need for a supportive and safe environment, as in some cases teenage boys are prohibited from joining their mothers and siblings at the shelter (Mullender et al., 1998). This is especially the case for women, as a safe and supportive environment for their children is one of the most important factors in the decision to leave a violent partner and seek help (Jonker et al., 2014). It is accordingly important to highlight that the children’s right to protection from violence (Article 19, UNCRC, 1989) also includes social and legal support, as well as treatment and follow-ups, regardless of factors such as gender and age.

Two of the fundamental rights of children are to receive information and, as mentioned above, to be able to express their opinions in decisions involving them (Articles 3 & 12, UNCRC, 1989); nevertheless, the children might not know why, or even that they are living in a domestic violence shelter (Øverlien, 2011a; Vass & Haj-Yahia, 2020), indicating that information is being withheld from them. This might be for their protection, but it can cause them to be anxious and unsure about their future (Överlien & Hydén, 2009). Giving children information about the shelter and its purpose can help them better understand the shelter environment as well as its rules and regulations, which can have a positive effect on their shelter stay. As Quennerstedt (2009) mentions, this also points to a conflict between children’s and parents’ rights, and specifically to the question of whose rights should take precedence when compromises are necessary. Accordingly, based on the compilation of research in this review, more studies are needed on children’s perspectives on different aspects of living in domestic violence shelters, including the process of leaving home, arrival at and life in the shelter, and the information and planning around moving out of the shelter. Regarding the time after leaving the shelter, we do know that shelters sometimes offer support for residents who have left, but Øverlien et al. (2009) emphasize that it was mainly the mothers who received support, not the children.

Although the existing research is of great value and shows the importance of shelters in creating a safe place free from violence, questions about how children’s lives are affected while living at a shelter have not received adequate research attention. For example, Överlien and Hydén (2009) highlight that collaboration between shelters and preschools/schools is important for making the children’s school situation easier, or in some cases for enabling them to attend school at all (see also Johnson, 1984; Mullender et al., 1998). The question of how this cooperation works remains unanswered, however. Acknowledging the importance of children’s leisure activities and schooling during their shelter stay is also in line with the UNCRC (1989, Articles 28 and 31), which highlights children’s right to education, rest, relaxation, play and participation in cultural and creative activities. Hence, this is an important area to investigate further.

Limitations

One limitation of the review is that only two databases were used. Even if they are quite wide and cover several disciplines, it is possible that relevant studies were missed because they are not indexed in these databases. In addition, the focus on searching in specific databases and the use of peer-review as an inclusion criterion limit the types of publications identified. For example, reports or book chapters on the subject have not been included. The fact that the second search was narrower is also somewhat of a limitation, as it did not follow the same search strategy and therefore captured different studies. Despite that, we believe that this search process is a strength, because important studies would have been missed without it. Another limitation is that although 28 studies have been included, several of them build on the same datasets, meaning that the number of research projects studying domestic violence with a focus on children’s own experiences and voices is even lower. A final limitation concerns the age of some of the articles. Five of the 28 articles included are from the 1980 and 1990s, which means that the implications in this review are constructed on both older and newer data. However, the use of older articles could also be a strength, as it shows how the research on children’s situation at domestic violence shelters has developed over time. For example, earlier studies stressed the importance of interventions directed towards children, while later studies showed that children received interventions, such as counseling, and appreciated them. However, when it comes to children’s opportunities to attend school while living at a domestic violence shelter, not much seems to have changed. Articles from as late as 2021 show similar hindrances as the oldest article from the 1980s. This means that although some studies are older, they can still contribute important knowledge, but their age needs to be kept in mind.

Implications

Based on the studies included in this review, we can conclude that greater emphasis is needed on children’s experiences and perspectives in both research and social work practice. However, by examining previous research we find that the use of a child-centered perspective has increased over time, and the situation for children in shelters has improved. This is also confirmed by Henze-Pedersen and Järvinen (2021). Still, we wish to stress the responsibility of governments to ensure that the rights of all children are respected in accordance with the UNCRC (1989). This also means that resources must be allocated to working with domestic violence (e.g., shelters, judicial system, social services, research) to make sure that children are both protected from violence and receive the help and support they need. Neither children, their families nor the domestic violence shelters should be left to work with the consequences of domestic violence on their own. Instead, governments, including public authorities and support systems such as the social services, need to take responsibility for the children and their families. Staying at a shelter should have as few adverse effects as possible on children’s lives. For example, guaranteeing well-functioning and safe education for children living in shelters must be a priority. Children must not, in the worst case, have to go without education for months or years (cf. Bracewell et al., 2020; Chanmugam, 2011; Selvik et al., 2017), as this can have long-term harmful effects.

Concluding Remarks

In conclusion, this review shows that the shelters are good at protecting children from further violence, and most shelters acknowledge the children’s situation. However, the children’s life situation is affected by moving to a shelter in various ways, and it is important that their perspectives are not excluded, diminished, or forgotten. While there is research focusing on children’s experiences of living in domestic violence shelters, and many shelters are striving to improve their work with children, further work is still needed in research, policy, and practice.