The “Problem” of Multispecies Families: Speciesism in Emergency Intimate Partner Violence (IPV) Shelters

Abstract

When a woman seeks emergency shelter from an abusive relationship, she may bring her children but rarely companion animals. Through a Critical Animal Studies (CAS) lens, this article qualitatively analyzes in-depth interviews with shelter workers in Ontario, Canada, exploring the place of multispecies families in intimate partner violence (IPV) shelters. The findings indicate that companion animals are viewed as problematic, as obstacles to their clients’ safe relocation, falling outside the scope of IPV shelters (who rarely take a co-sheltering approach), and as potential strains on an already resource-stretched social institution. Addressing a gap in the literature about the effects of companion animal policies in social housing on clients and staff, the results are relevant to social service providers and policymakers working with multispecies families, including insights about women and children’s reactions to separation from companion animals, contradictions in related policies, and institutional priorities. The article concludes that multispecies families are poorly accounted for in the IPV shelter system and suggests that researchers and shelters should collaborate with their communities to advocate for resources and policies that accommodate families with companion animals.

1. Introduction

Studies indicate that there are few options for women who wish to include companion animals as members of their family when fleeing from violence (Gray et al. 2019; Matsuoka et al. 2020; Stevenson 2009; Stevenson et al. 2018). Most emergency intimate partner violence (IPV) shelters are not able to accommodate multispecies families; companion animals are rarely, if ever, permitted. Initiatives that are alternatives to co-sheltering, such as the fostering of companion animals (Flynn 2000; Kogan et al. 2004), do not keep companion animals with their families (Krienert et al. 2012; Matsuoka et al. 2020; Stevenson 2009; Taylor et al. 2020).

Families include other species, yet these multispecies groupings are rarely the focus of sociological research. Taylor writes that, “since the social world with which sociologists concern themselves has always been a multi-species one, it is time for sociology as a discipline to reflect this” (Taylor et al. 2018, p. 465). Building on Taylor’s call, this article explores the place of companion animals in intimate partner violence (IPV) shelters. It focuses on the perspectives of shelter workers about multispecies families in emergency IPV facilities, asking the question: How do multispecies families fit in emergency IPV shelters? As a work of activist-scholarship (Meyer 2005) employing a qualitative methodology and the conceptual approach of critical animal studies (CAS), this article aims to contribute, broadly, to the sociological knowledge of multispecies families.

This article argues that the predominant impediment to co-sheltering multispecies families in intimate partner violence shelters is species. The constructed category of species is used as the basis for excluding companion animals from IPV shelters: an act of speciesism. Speciesism, a term coined by Richard Ryder (1989), “refers to an ideology which supports treating sentient and morally equivalent beings differently on the basis of species alone, rather than giving them equal consideration” (Matsuoka et al. 2020, p. 56). Companion animals in multispecies families are not accommodated, based on their constructed, nonhuman status (Irvine and Cilia 2017; Kirksey 2015; Matsuoka et al. 2020). Through a Critical Animal Studies (CAS) lens, in the broader shelter system, humans and nonhuman animals are sorted by such categories; species is used along with other categories such as gender and ability to oppress and divide. For example, there are different shelters and transitional spaces for homeless individuals, those exiting incarceration, companion animals, and IPV survivors.

This article makes three important contributions to the existing literature on intimate partner violence and nonhuman animals. First, the data are drawn from IPV shelter workers, providing a collective view of actors from this social service sector, instead of the voices of those utilizing the institution, such as IPV survivors. In particular, the data provide new information about how shelter workers handle multispecies families. Second, as this article can be considered activist-scholarship, where the interests of practitioners and activists drives the question being asked by the researcher (Reynolds et al. 2018), certain findings may be of direct use in the field of social work. Third, this article also offers significant empirical data related to social psychology about the intricacies of multispecies emotional bonds for recently relocated IPV survivors.

2. Literature Review

Sociological research on women help-seeking with companion animals reveals multiple challenges to safe relocation, even after the woman has decided to leave (Barnett 2000, 2001; Barrett et al. 2017). Over time, studies continue to find that many women delay or do not leave their abuser out of concern for their companion animals (Barrett et al. 2018; Crawford and Bohac Clarke 2012; Fitzgerald 2005; Fitzgerald et al. 2019; Giesbrecht 2021a; Gray et al. 2019; Stevenson et al. 2018; Taylor et al. 2020; Wuerch et al. 2021). Studies in the United States have found that women remain in abusive homes because of companion animals (Ascione et al. 2007; Faver and Strand 2003; Flynn 2000; Krienert et al. 2012; Strand and Faver 2005). In Canada, nearly half of women stay or delay leaving because they cannot bring their companion animals with them (Barrett et al. 2018; Fitzgerald 2005). In their Canadian sample, Stevenson et al. (2018) found that “the majority of shelter staff (74.8%) … were aware of abused women in the community who did not access the shelter because they could not bring their pets with them” (p. 242). Recent studies in Saskatchewan with human service providers and animal welfare workers find that not being able to co-shelter with a companion animal is contributing to women remaining with their abuser (Giesbrecht 2021a; Wuerch et al. 2020, 2021). Even if women leave the home for a shelter, they may return to care for their companion animals (Giesbrecht 2021b; Stevenson et al. 2018; Wuerch et al. 2021).

For many women, companion animals are a key source of support. Of the women in Flynn’s (2000) study, for instance, forty percent stayed or delayed leaving, explaining they viewed their “pet as a source of emotional support” (p. 169). Bonds between women and their companion animals (Barrett et al. 2018; Faver and Strand 2003; Matsuoka et al. 2020; Stevenson et al. 2018; Strand and Faver 2005) are particularly strong and important when violence is present (Barrett et al. 2018). Further, research has found that companionship from animals may aid in feelings of security, or a sense of being “at home” (Kidd and Kidd 1994; Irvine 2013), an “attachment safe haven” (Lem et al. 2013; Kerman et al. 2019) that is a source of protection with emotional benefits.

Moreover, there is a recognized link between human and nonhuman animal abuse. We find that companion animals are similarly at risk of violence (Boat 1995; Fitzgerald 2005; Fitzgerald et al. 2019; Flynn 2011; Newberry 2017; Arluke et al. 2009). The co-occurrence of nonhuman and human abuse in the home is well documented (Barrett et al. 2017; Boat 1995; Fitzgerald 2005; Flynn 2011; Newberry 2017). In cases of IPV, the abuser may leverage the woman’s bond with her companion animals, and threaten, harm, or otherwise use these animals to manipulate and control her (Barrett et al. 2017; Fitzgerald 2005; Fitzgerald et al. 2019). The use of companion animals as bargaining tools in such situations is acknowledged in the literature (Mills and Akers 2002; Rook 2014), with laws to address it often modeled after child law (Rook 2014). Notably, companion animals are considered property under Canadian law (Sorenson 2010), significantly complicating their place and “rights” in cases of abuse.

Studies also demonstrate that leaving companion animals behind places them at risk (Flynn 2000; Gray et al. 2019; Stevenson 2009; Stevenson et al. 2018). This potential for further harm is another factor preventing women from leaving. Stevenson et al. (2018) found in their sample that shelter workers knew of women leaving the shelter to return to the home to care for their companion animals. For those that do leave their nonhuman companions behind, the limited research on this topic finds various fates for those companions. In Flynn (2000), more than half of these animals remained in residence with the abuser (55%), resulting in many respondents expressing deep concern for their companions’ safety (p. 169). The others were left with family or friends (25%), surrendered to an animal shelter, or worse (i.e., abandoned) (Flynn 2000, p. 169). Other arrangements that women and shelters may utilize include keeping their companions: with other family or friends, in temporary kennels, with veterinarians, or in foster care (Flynn 2000; Kogan et al. 2004). Studies of what happens to the companion animals who are turned away from IPV shelters are near absent in the sociological literature, to the knowledge of the author.

Knowing about companion animal abuse helps shelters better assess the risk to the women they serve (Fitzgerald et al. 2019; Gray et al. 2019). Fitzgerald et al. (2019) explain:

Specifying the factors that increase the probability of pet abuse is critical to inform the development of risk assessment measures that can identify pets that may be in elevated danger, as well as people who are being victimized and whose decision making regarding leaving the relationship may be shaped by the abuse of their pet(s).

(p. 2).

As such, knowing about companion animals in the home helps emergency shelters better develop safety and care plans. This last point mirrors recommendations that shelters ask about companion animals when performing risk assessments or safety plans (Barrett et al. 2018; Fitzgerald 2005; Gray et al. 2019; Stevenson et al. 2018), as “promoting the safety of pets is critical not only for the well-being of animals but also for the protection of the people who care for them” (Fitzgerald et al. 2019, p. 2).

IPV shelters often advocate for a trauma-informed approach (TIA) to service delivery. TIAs are grounded in the understanding that the psychological effects of trauma require specialized care that is individualized, centred around, and largely guided by the survivor, prioritizing her/his autonomy and unique needs, and the fostering of trust and feelings of safety, where the survivor decides the pace and scope of care (Kulkarni 2019). Further, research demonstrates that such a “survivor-defined advocacy” model (Stevenson et al. 2018, p. 247) also identifies and addresses systemic issues that affect the clients of IPV shelters (Davies and Lyon 2013; Kulkarni et al. 2015; Stevenson et al. 2018). Shelters and shelter workers may encounter difficulty in actually implementing a TIA approach, such as limited resources like labour and funding and competing time pressures (Burnett et al. 2016; Stevenson et al. 2018).

Although IPV shelters are a fixture in Canadian social services, there is limited literature on how applicable policies and protocols affect service delivery (Burnett et al. 2016). Responsibility for social housing, such as IPV shelters, was downloaded from the federal government to the provinces in 1993 (Johnstone et al. 2017, p. 1444). Beginning in 1995, this social service became further altered in Ontario as oversight was passed to the municipalities (Johnstone et al. 2017, p. 1444). In the province, IPV shelters fall under the umbrella of Ontario’s Ministry of Children, Community, and Social Services (MCCSS) (Ministry of Children, Community, and Social Services 2022), “but they operate in a complex web of federal, provincial, and municipal policies” (Johnstone et al. 2017, p. 1444).

Non-profit agencies primarily run the IPV shelters in Ontario (Johnstone et al. 2017, p. 1444; Laforest 2013). Research finds that the added responsibility of compliance with various governmental legislation is costly for shelters, in time and labour (Burnett et al. 2016). Beyond the requirements of funders, shelters also have to consider regulations related to employment standards, building codes, water and fire safety, and disability accommodations (Burnett et al. 2016, p. 523). Research has found that “shelter workers become astute system knowledge brokers” (Burnett et al. 2016, p. 517) in response to the complex needs of their clients and the regulations they are subject to. In other words, in order to fulfil their mandate of keeping abused women and their children safe, staff learn to navigate multiple policies and priorities related to recovery and life after the shelter, all in a short time and during a period of great upheaval for their clients, in tandem with the effects of such trauma.

Applicable then is research that examines the larger social housing system in Canada. The populations known to access public housing, in general, are those at risk for homelessness, low income, seniors, or those with disabilities (McCabe et al. 2021, p. 292). Much of past (Blau 1988; Hill 1991) and current scholarship and policy debate about social housing is divided by the believed causes of poverty and homelessness (Johnstone et al. 2017, p. 1445). The discourse either posits that blame lies in personal failings and is thus an individual responsibility, or in systemic inadequacies, such as governmental policies and shortfalls in funding (Johnstone et al. 2017, p. 1445). What these policies and priorities are and how they impact the work of shelter employees and thus indirectly affect clients is an important area of research. Work carried out on companion animal policies in IPV shelters reveals inconsistencies in policies and even confusion or a lack of discussion within shelters about how they help clients with companion animals (Gray et al. 2019). One study suggested that workers do what they can in each case to keep women safe, gesturing to ad hoc problem-solving around companion animals, but also that some workers may not mention companion animals at intake due to a lack of clarity or solid options for their care (Stevenson et al. 2018).

Thus, research continues to demonstrate the importance of knowing about companion animals in the lives of IPV clients, and to provide access to multispecies emergency shelter from intimate partner violence. Attention to the companion species in abusive homes remains limited and as such there is an “urgency of better understanding and mitigating the unique barriers to leaving an abusive relationship faced by women with companion animals” (Barrett et al. 2017, p. 2). This article builds on the literature to broaden our understanding of how multispecies families fit and intersect with IPV shelter policies and priorities. More generally, this article moves the sociological conversation forward about what does or could occur during emergency relocation of a multispecies family, with insights from shelter workers who are closely involved in these situations.

3. Methods

This article draws on data collected for a larger research project on multispecies families in intimate partner violence shelters in Ontario. For the purpose of this paper, the focus is on semi-structured interviews with shelter staff about their experiences with multispecies families.

I consider the umbrella project, and this article, activist-scholarship. A dichotomy is often supposed between those working in academia, and those in the field, “on-the-ground” (Reynolds et al. 2018). As a vegan activist-sociologist working in the area of social inequality, I have found myself confronted with inhabiting both fields numerous times in my career. The motivation for studying shelters—specifically IPV-responsive and companion animal facilities—and the needs of multispecies families is more than interest. I am a survivor of both family and intimate partner violence, and I have experienced help-seeking with a child and companion animal, with a low income and few social, financial, or familial supports. My research in general, and this article, can thus be described as activist-scholarship (Meyer 2005), which denotes work by individuals in academic institutions that leverages teaching, research, and related activities in support of social movements or community-based initiatives (Reynolds et al. 2018).

3.1. The Umbrella Project

The purpose of the larger, “umbrella” project was an investigation into the policies and regulations surrounding the place of companion animals in emergency IPV shelters in Ontario, Canada. This study sought to reach as many English-language based, non-Indigenous-focused IPV shelters in Ontario as possible to produce a good cross-section of these facilities across the province, with results being contextual to the region. Eighty-one of these shelters were located on the Ontario section of the Canadian IPV shelter directory website “Sheltersafe”, a “project” of the national organization, Women’s Shelters Canada (Women’s Shelters Canada 2020). These facilities are funded and governed provincially by the Ministry of Children, Community, and Social Services (MCCSS) (Government of Ontario 2021) and run as non-profits. A few of the shelters in my survey were affiliated with other shelters in their area. For instance, one first-stage shelter worked closely with two others in the same geographical area to collectively manage client care, and two other first-stage facilities mentioned their organization included a second-stage shelter, at a different location in their region. All but one shelter in this study is a first-stage facility (short-term); the exception is a second-stage (longer term) shelter. Ontario’s IPV shelter system is not unified, with each (or a group of) shelter(s) operating on its own and adhering to provincial requirements (Women’s Shelters Canada 2020).

This research explored the perspectives of workers and affiliates about co-sheltering policies and practices, opinions about co-sheltering, companion animals that had or were currently allowed in-shelter, and what happens to multispecies families in these shelters. Interviews were conducted with those closely involved and witness to multispecies families in emergency IPV shelters. They included shelter workers, and individuals at a provincial shelter association and a companion animal foster organization (the latter two groups are considered “affiliates”, hereafter).

Email invitations to interview of were sent to these shelters in 2018 and generated 14 individual acceptances from 10 shelters. Two affiliates—one employed in the shelter sector and the other working with a companion animal foster organization—also agreed to interview, bringing the total semi-structured interviews performed to 16.

Respondents in the study are involved with IPV emergency shelters, either as shelter workers, or as affiliates working with IPV shelters. All but one of the staff in this study identify as cisgender women; this individual describes themselves as non-binary. The age of participants ranges from early 20s to mid-60s. The individuals identify mostly as “Canadian” or White/Caucasian; two participants are bi-racial. The majority of participants describe their relationship status as married or living with a partner; three are single and two are widowed, and a few also have children. All participants have some form of post-secondary education. The employment type held by these workers include part-time/casual (“relief”), full-time, and positions range from client counsellors to management/leadership roles, and executive directors.

In-person or telephone interviews were held over the Summer and Fall of 2018. The option to interview in-person or via telephone allowed for flexibility and possibly more participation by both the staff and researcher, due to the sensitive nature of the topic of study, and considerations of personal and professional anonymity (Esterberg 2002). A semi-structured question guide was used to ensure comparability across interviews, which were recorded and transcribed. I included questions that would allow my respondents to talk about their experiences with nonhuman animals, such as “Do you have companion animals?” and “Are companion animals allowed in your shelter?”

3.2. Analysis

The interviews were transcribed and qualitatively analyzed, guided by the arguments in the existing sociological research about multispecies families, as reviewed above. The researcher’s field notes, taken during and after the interviews, were also reviewed. The interview data were coded thematically, and findings relevant to social service providers and policymakers about multispecies families and intimate partner violence are highlighted next.

4. Findings and Discussion

All of shelter workers in this study (n = 14) recognize the close, kin-like relationships some of their clients have with companion animals, echoing the literature on multispecies families (Irvine and Cilia 2017; Kirksey 2015; Taylor et al. 2018). In agreement with the literature on co-sheltering (Barrett et al. 2018; Gray et al. 2019; Stevenson 2009; Stevenson et al. 2018), the majority of facilities represented here do not co-shelter companion species (n = 7). However, two first-stage and (the) one second-stage stage shelter in this study offer some form of co-sheltering on-site. Several shelter workers also report working with alternative care providers (such as foster homes or programs) to find temporary housing for clients’ companion animals; this is also reflected in the literature (Barrett et al. 2018; Stevenson et al. 2018) and on Sheltersafe.ca’s online directory (Women’s Shelters Canada 2020). These exceptions and alternatives warrant discussion and study on their own and are the focus of current analysis by the researcher.

Findings relevant to social service providers and policymakers about multispecies families are included here, offering insights about women and children’s reactions to separation from companion animals, contradictions in related policies, and institutional priorities.

4.1. Priorities of the Institution

The dataset consists of accounts from workers at IPV shelters. During analysis of these conversations and the researcher’s field notes, some priorities of the shelters became evident, in both explicit and more veiled ways. The central priority of these institutions was occasionally identified outright as female and/or female-identifying survivors of IPV, but this focus was always implied as the known orientation of this social institution. When overtly mentioned, workers often described this priority as their “mandate”; for example, one worker said, “we’re mandated to remove barriers… for women to get out [of abusive situations]”. Literature on the policies that govern and effect Canadian IPV shelters and their related impacts is scarce (Burnett et al. 2016). However, it is found that this institution’s chief mandate is the provision of safe housing for women and often their children (Burnett et al. 2016, p. 520). I discuss institutional oversight below (Section 4.3).

When talking about who was allowed in the shelter, a few workers identified the populations who their shelter could and could not accommodate, based on the priorities their institution set. One of worker explained that men are not permitted on the premises at all, “because we can’t brand who’s the right one and who’s the right one”. However, this same worker said that clients’ sons 18 or older are allowed to stay, since “that’s part of the family and that’s different”. In contradiction, a few workers mentioned that if a client’s child was male and over age 18, he would not be able to live with his mother in the shelter for reasons such as client wellbeing and safety. One worker provided tangible examples of why they restrict the access of men:

We have had men dress up as pizza delivery, as a woman, as different occupations, looking for women and putting things in her mailbox… There was more abusers sitting out in the parking lot [before they replaced a solid fence with a see-through metal gate system]… they would wait for the person and then would either assault them or [other things]”.

Several other workers explained that they have security measures in place such as video cameras, locked doors, unlabeled buildings, and separation between the living quarters and main office areas. Safety concerns are essential to IPV shelters, given the nature of intimate partner violence.

Although, in some cases, women feel that companion animals provide security and safety (Hardesty et al. 2013), a recent study raised the important point that abusers often use women’s bonds with companion animals to control them, and thus, “care for animals controlled victims by creating fear and increasing barriers to safely exiting the relationship” (Giesbrecht 2021a, p. 18). Likewise, some of the workers interviewed for the present study knew of situations where companion animals had been harmed or killed by women’s partners. As in the literature (Barrett et al. 2017; Boat 1995; Fitzgerald 2005; Flynn 2011), these workers’ accounts support “the Link”, where nonhuman and human abuse are concurrent. Further, the workers in this study identify the animal victims in their clients’ lives as companion animals, usually cats and dogs, suggesting that knowing about companion animals in the home may be an additional safety measure to consider.

Overall, my data and the literature imply that safety is an institutional priority; this seems a given, as the population served is expected to have been the victim of violence, often in the same community, but my data provide more nuance as to how client safety is linked to bonds with companion animals.

Recovery is also a priority of all the shelters in this study. Most workers either explicitly identified or gestured to their use of a “trauma-informed approach” (TIA) to client care. Data gathered in this study were fairly limited regarding the specific details about the care provided, as this was not the central focus of the interviews. However, one worker spoke about a requirement for the client to want to “stay safe” and recover, as, at her second-stage shelter, women must apply to move in, and that “the team” will reach out to “Children’s Aid” and/or the first-stage shelter where she is staying to ask:

Do you see this person as being able to do the safety piece here? What is it that you feel, because you referred her here, what do you think we could offer this woman that would help her move forward?

Further, women are only accepted into this facility if they agree to follow “the program”, which seems to refer to her shelter’s priorities and guiding philosophy. The worker describes it this way:

The program, first and foremost, is to help women who have left violence and have children or not, move on with their lives. That is first and the biggest priority. When you come in here, you have to also understand that you’re part of the safety system, so will your lifestyle, will your decisions keep the community safe? And are you able to know about keeping yourself safe as well as your community? Because something that one person may do may open up [something] very unsafe for others. You have to buy into the program, you also have to buy into the program that you’re here to work on yourself. While you’re here you’re going to be asked to come to some kind of program. It might be a group. It may be one-to-one counseling. It may be whatever it is that person wants to work on. They drive their own vehicle as their own life, we just accompany, but they must be working on something so that when they do move out of here they have the tools in their basket to make it successful for them.

Unpacking this quote, this worker describes a necessity to “buy into” the shelter’s prescriptive method, to use their tools for recovery. This interviewee also identifies her shelter’s requirement that clients be “working on something”. This worker’s sentiments stand out in the data as somewhat contradictory to the trauma-informed approach mentioned by other interviewees. Although the worker mentions “they drive their own vehicle as their own life”, gesturing to their autonomy, the client must also be willing to “work on” themselves and that their “lifestyle” and “decisions” must align with a desire to “stay safe”. Reading these last requirements through the literature on TIAs to care, this seems responsive to the finding that “IPV survivors may self-medicate physiologically uncomfortable trauma symptoms with substance use in a manner that interferes with their safety or ability to parent effectively” (Kulkarni 2019, p. 58). It is notable that the unique finding comes from staff at the only second-stage (longer term) IPV shelter in the study. Mandatory participation or use of shelter services beyond the “basic” housing, food, and living materials provided were not explained as being required of clients by the other shelter workers in the study.

Several workers mentioned that they follow a “harm reduction” approach, which translates to an acceptance of clients’ substance use, and a focus on reducing or avoiding associated harms. In Burnett et al. (2016), a frontline shelter worker explained:

People aren’t going to stop using something they are very addicted to while they are in a middle of a crisis. So, we have a lot of new harm reduction policies and I think it has a gentler tone with women that use substances. We’re not a zero tolerance environment anymore… it’s not without controversy, that’s a big shift.

(p. 521)

Some of the “controversy” noted includes the additional challenges of working with clients under the influence, and the possibility of that this policy could result in other clients being exposed to these behaviours, affecting their ability to “stay safe” (Burnett et al. 2016, p. 521). More targeted research is needed to determine how IPV shelters may differ in their service delivery, and if this is tied to their structure as either a first- or second-stage facility. The author is unaware of specific or scoping studies on the participation requirements for clients in IPV housing; this is an area in need of further research.

However, the shelter workers in this study describe routinely engaging in a myriad of services for their clients, such as counselling in-house or brokering connections to other supports in the community. Indeed, the literature shows that women in IPV shelters, “face time-limited stays in which they must reconstruct their lives” (Burnett et al. 2016, p. 517). These women and families are often in the shelter in the first place due to a lack of choice or access to other options, as the populations served are often low-income and facing multiple barriers to independent, safe living environments (Johnstone et al. 2017; Kidd and Kidd 1994; Kulkarni 2019). Orchestrating this is an enormous and stressful task; the shelter workers spoke of a sort of dance they performed individually and as an organization, where they had to “do everything” with few resources, and quickly. Indeed, the literature reflects the high-stress, low resource nature of these environments (Stevenson et al. 2018), and effects on workers, such as burn-out, secondary stress disorder, and high staff turnover (Kulkarni et al. 2015, p. 60). The ongoing challenges faced by shelters and their workers, such as resource shortages and policy navigation, have been found to be systemic barriers to successful service delivery (Burnett et al. 2016). It should be noted that it is these institutions, with too few resources, that are charged with caring for people who also lack resources; the irony has been explored in the literature (see, e.g., Kerman et al. 2019; Kidd and Kidd 1994).

Many of the workers also spoke about clients who would come to or call the shelter falsely claiming to be fleeing violence, because of the “housing crisis”, which the interviewees explained reflected the critical shortage of affordable housing in the area. Decades ago, Hill (1991) noted that stays at shelters are meant to be temporary, yet the potential of abused women to move on is limited by a lack of affordable alternative, rental accommodations, and other factors such as unemployment, insufficient social assistance, and a lack of nurturing and supportive environments to gain confidence and independence. Hill (1991) also found that women who end up in the shelter routinely come from impoverished families and cannot turn to them for financial support or housing, essentially rendering them homeless. All shelter workers in the current study indicated that their shelter was always at or near, and sometimes over, capacity, and that this limited the number of clients they could help. This finding agrees with current literature, which finds that shelter capacity is a concern further amplified by housing cost and availability in the community more generally (e.g., Stevenson et al. 2018).

The workers further described that their shelter was only available to women who were survivors of IPV, not to women who were “just homeless”. This speaks to IPV shelters’ unique connection to class, and also gender. Only women/women-identifying individuals who are also victims of IPV are permitted to stay in the shelters interviewed for this study. In line with literature around the clientele of IPV shelters, the data demonstrate that these facilities identify and prioritize their clients’ identities as women who are IPV survivors, rather than women who are need of shelter due to being homeless, poor, or for other circumstances.

As the focus of the study is families that include companion animals, much of each session was spent talking to the shelter workers about their experiences with companion animals personally and in the workplace. All of the interviewees had personal connections to companion animals and knew that some women in abusive homes would not leave without these animals. Several workers observed close relationships between their clients and their companion animals. It has been established in the literature that many homeless individuals have close bonds with CAs, and that sometimes that companion animal is the person’s only family (see, e.g., Hill 1991; Irvine 2013; Kerman et al. 2019). Adding to the literature on co-sheltering is this study’s finding that, for some shelter workers, companion animals’ levels of need and maturity are similar to that of children, suggesting a need for close supervision and care.

For all interviewees, companion animals were problematized in some way. The majority of workers felt that including companion animals in their shelter was unlikely, citing resource limitations, such as labour, funding, and space, and reinforcing that their facility was meant to support women and their children. Several of these workers also mentioned that having CAs in the shelter may create conflict between clients and be sources of stress, fear, worry, or additional work. The literature also reflects this species-based mandate as well as how this view leaves out multispecies families (e.g., Matsuoka et al. 2020; Stevenson et al. 2018).

The workers at the three co-shelters in this study identified companion animals as a low priority in the sector generally, yet a persistent concern for vulnerable women in their community. Several mentioned that CAs fall under the umbrella of “animal shelters” or other community agencies, not IPV shelters. Even the interviewees working at co-shelters viewed CAs as barriers to accessible emergency shelter in their community. This is consistent with current literature on co-sheltering (e.g., Giesbrecht 2021a)

The second-stage shelter in this study did not make companion animals a priority, but instead relied on the municipal bylaw restriction on animals in a residential setting, which allows a limited number and species to reside in these spaces. The other two first-stage shelters did intentionally prioritize the inclusion of clients’ companion animals; one quietly helped find care or temporally allowed companion animals in-shelter when possible “for years” and lobbied “the board” (of directors) until her shelter was allowed to accept clients’ CAs openly and officially, beginning about four years prior to our interview in 2018. The other co-shelter, consulting with a veterinarian who was also an intern, found support internally and sought board approval and capital funding to construct a dedicated CA space in the shelter. In all of these cases, companion animals were viewed as a problem, as they identified these beings as not included in the institution’s mandate, yet a concern of and obstacle for their clients to enter emergency housing. The difference is in how two first-stage shelters actively made CAs a priority in their facility, while the other facilities report only occasional discussions or the formation of a “pet committee” to think about the problem of clients’ companion animals, but no reported prioritization in their facility for inclusion. A recent review of companion animal restrictions in public housing for low-income seniors found a narrative that CAs are problematic or risky (Matsuoka et al. 2020). The data in the present study support a narrative that companion animals are experienced as problems by IPV shelters, and that they are not institutional priorities.

4.2. Insights for Workers in the Field of IPV and Family Social Services

This article can be considered activist-scholarship, as certain findings may be of direct use in the field of social work. The study finds that children may resent a parent who re-homes (even temporarily) or leaves a companion animal behind, and that women may prioritize companion animals over self-care.

Several of the shelter workers in my study knew of clients placing their companion’s care and health above their own. One worker said:

I’ll see women who’ll give up stuff. They’ve given up stuff so that their pets have it. I’ve seen it for years and they treat them better than themselves at times and it’s that unconditional piece that quite often humans can’t give to humans, right?

The finding that some women prioritize their companion animals’ wellbeing also relates to another respondent’s comment, that “[it is hard] having women not leaving relationships because of, and I use the term pets, family members”. Several other shelter workers articulated that companion animals are not just “nice to have” or luxuries, but “actual” family members. In the same vein, another worker commented on the result of non-prioritisation of co-sheltering, that the impact is not felt by a particular “type” of multispecies family:

I think that’s a big message to get out… it’s affecting families as a whole, not just some woman who lived with her cat, or whatever, right? It’s affecting everyone, it’s affecting families, it’s affecting moms, it’s affecting dads, it’s affecting a lot of people to not have these things in place, because people aren’t going to leave their pets because, well, most people aren’t going to leave their pets if they view them as family… and I feel a lot of people do view their pets as family. It’s always been kind of there, but I’m feeling it’s louder. Maybe it’s just said more.

Corroborating the recent literature on multispecies families (Irvine and Cilia 2017; Taylor et al. 2018), these workers point to the close bonds in these groupings and identify the impact of separation, much like this worker’s comment:

And that loss… not being able to bring the pets can be such a significant loss. And… a barrier to physical safety but also mental, emotional safety as well. It can definitely be detrimental to someone’s mental health to lose the pet that’s like part of their family, their only source of support.

Another says, “single women that are having to leave their dogs are being [negatively] affected”; another worker stated that companion animals are sometimes their clients’ “only family”. In all these cases, the shelter workers express empathy for clients who “lose” their companion animals during their relocation/recovery journey, gesturing to the kin-like relationships in multispecies families (Irvine and Cilia 2017; Kirksey 2015).

A related finding is how children may be resentful toward a parent who leaves their companion animal behind in an abusive home. One shelter worker explained:

When you have a family of a mom with five children that they all of a sudden had to leave their dog that they’ve grown up with every single day, they can have resentment towards mom: “Why would you take us out of our house? We don’t have our dog. We don’t have all the things that are comforting to us. And especially in these times when you know when I’m upset I cuddle with my dogs? That’s my thing.” If a child’s used to doing that, and all of a sudden you take that away, it’s affecting the children.

Connected to this insight is this study’s finding that this resentment happens even when a foster home or other arrangement is found for their companion animal. The objectionable situation, from a child’s point of view, is that the companion animal is no longer with them; it may not matter where they are instead.

4.3. Species-Related Policy Contradictions

Lastly, the data reveal policy contradictions in terms of how protocols are applied in spaces where oversight is provided by more than one governing body. As illustrated in the literature (Burnett et al. 2016; Johnstone et al. 2017), IPV shelters in Ontario are subject to multi-level governmental oversight. When speaking about companion animals, the shelter workers in this study had inconsistent and conflicting understandings of the regulations that apply to companion animals in their facility. Some said they believed CAs were not permitted due to “public health” regulations. When pressed for specifics, these workers were largely unsure exactly what these policies were, although some mentioned issues such as allergies. A worker at one of the few co-shelters disagreed, saying she had not experienced complaints or concerns from public health inspectors, but that she tended to ask that CAs be kept out of the shelter’s kitchen to avoid such issues.

Two other areas of policy mentioned by workers were municipal restrictions on companion animals as well as accommodation of “service” animals, specifically dogs. When asked about policies around companion animals, a worker from the second-stage shelter said, “we go with the [municipal] by-law. So, if you’re allowed three cats, you’re allowed three cats”. In this case, the shelter defers to regional guidelines about keeping companion animals.

A worker at one of the first-stage co-shelters explained that they developed their own CA policy, with includes an intake procedure and understanding. During intake, and until the CA has been vetted and assessed as “safe”, they are kept in a holding area (an accessible washroom or meeting room). She explained that staff complete an internal wellness assessment:

The system is that the woman brings it in. We do an investigation, just with her, a sheet of, … about the animal, about the temperament, about the age, and health issues… And then, once the wellness check is done [and the veterinary assessment is complete], the animal can go up and move into the woman’s bedroom with her; animal or animals.

Interestingly, both staff and clients at this shelter are also advised that the facility is a co-shelter before they begin working or come to stay, and that there are other shelters in the area they can go to if they do not wish to work or stay in a shelter that accommodates companion animals. The third co-shelter in this study had yet to accept CAs at time of interview and were still developing internal procedures. In all other shelters in this study, there were no internal, organizational policies guiding workers when handling clients’ companion animals. The literature also finds that companion animals are unevenly accounted for in the IPV shelter system (Stevenson et al. 2018; Gray et al. 2019).

Finally, all shelter workers in this study were aware of and claimed to be compliant with accessibility requirements for “service” animals (SAs). There seemed to be clarity that clients’ “service” animals are to be permitted in-shelter, although there was very little experience with this scenario amongst the interviewees. The workers also made a distinction between companion animals and SAs; this is guided by legislation in the Accessibility for Ontarians with Disabilities Act (AODA), which establishes that shelter workers know that “service” animals are required to be admitted (Accessibility for Ontarians with Disabilities Act 2019; Government of Ontario 2022).

5. Conclusions

This article considered the question: How do multispecies families fit into IPV shelters? Through the lens of shelter workers, we learn that the answer, in many cases, is not well. Echoing the existent literature on co-sheltering, this article finds that companion animals remain outside of these shelters’ mandate of assisting women and children recover, reorganize, and restart their lives after abuse. The findings also demonstrate that companion animals are known to be a source of support, may be abused as well, and that separation from them is detrimental to the family, creating stress and distress as children and women experience this loss in a time of great upheaval. The companion animals involved are shown to be a low priority for IPV shelters, and also problematized as a barrier for victims to successfully escape abuse.

As a work of activist-scholarship, this article argues that the predominant impediment to co-sheltering multispecies families in intimate partner violence shelters is species. As a survivor of multiple abuses who was turned away by three local IPV shelters in central Ontario with my young daughter “because” of my cat Silver, I am an example and reminder that multispecies families need help escaping abuse, together. From a Critical Animal Studies perspective, it is unacceptable that something is not in place, system-wide, as a social safety net in the name of gender, class, and species equity, after so many years of research finding the same thing: the shelter system is inadequate and speciesist, and the result is that multispecies families are often, as I was, “out of luck”.

I do not advocate for species erasure in the sense that all animals “become” a homogenous group but point to how species is used along with other socially constructed categories such as gender and ability in the emergency shelter system, and that this division has its own set of impacts.

Collectively, the findings of this article provide important insights for policymakers and social workers who are trying to overcome barriers related to parent-child relationships, as well for those working with multispecies families. Moreover, these are critical insights for people working in the fields of intimate partner violence and family social services. These findings are also significant to social psychology, when considering the emotional, psychological, and social impacts of the non-prioritization of multispecies families by IPV shelters.

This study also adds to what is known about the multi-level oversight IPV shelters are subject to. The workers describe numerous policies connected to decision making about companion animals, and thus multispecies families. In agreement with the literature, the shelter workers identify that their facilities and practices fall under multiple governing bodies, and that this affects their decision making and priorities. Overall, the multi-level oversight of IPV shelters, and the interaction of these policies, directly impact how the workers work with clients in these spaces. As Burnett et al. (2016) indicate, sociological knowledge about policy and its impacts on workers and clients is a gap in the research, one that this study begins to address.

In conclusion, multispecies families are poorly accounted for in the IPV shelter system. As suggested in the literature (see, e.g., Brewbaker 2012; Kulkarni 2019), as a stop-gap and impetus for change, researchers and shelters should collaborate with their communities to advocate for resources and policies that accommodate families with companion animals. As noted in the literature (see, e.g., Matsuoka et al. 2020), social housing institutions could contribute to stabilizing multispecies families after abuse or incidents of homelessness by prioritizing companion animals. Indeed, attention is warranted to the nonhuman animals who are involved. Despite recognition of multispecies families and some patchwork “solutions” to species-divided sheltering, companion animals are drastically affected by the lack of co-sheltering in emergency IPV facilities. Urgent areas of future research include studies on the integration of companion animals into IPV shelters as well as work on the various fates of companion animals in situations of emergency relocation due to violence in the home.