Introduction

In 2020, the COVID-19 pandemic upended life for the entire global community. In the early days of the pandemic, while everyone tried to navigate the panic and uncertainty, domestic violence (DV) shelter staff had to figure out how to respond to this unprecedented situation not only for themselves, but for the countless survivors they serve as well. With weeks to months of widespread lockdown restrictions, survivors faced increased challenges in reaching services. These unique circumstances forced both survivors and shelter staff to rely on their ingenuity and resourcefulness to connect survivors with the help they needed. With little guidance from public officials and national agencies, DV shelters were left to determine how to navigate the early stages of the pandemic largely on their own. The purpose of this study is to examine how DV shelter practices and policies, direct services shelter staff, and shelter residents were impacted by the pandemic.

Literature Review

Domestic Violence Shelters

DV shelters have been a refuge for countless victims of DV and their children since they first emerged in the United States in the 1970s (Richie, 2012). In addition to a safe and confidential place for survivors and their children to stay, shelters provide a multitude of other services and resources. Shelters represent a crucial piece of the network of services available to survivors, especially for the most vulnerable with limited resources to leave an abusive relationship. Shelters also act as a central location for multi-faceted services that typically include assistance or referrals for securing childcare, employment, long-term housing, legal advocacy, medical care, substance abuse treatment, mental health services, transportation, and financial assistance. It is also imperative that these residential services provide child-specific supports as many survivors enter the shelter with their children. Many shelters provide counseling and support groups for adults and children as they cope with the harmful effects of experiencing or witnessing abuse. Typically, DV shelters also engage in outreach activities, educating their local community about DV and the help available to victims (Johnson & Stylianou, 2020; Sullivan 2018).

According to a 2021 census, the National Network to End Domestic Violence (NNEDV) reported that 38,608 adult and child survivors received residential services in a single day in the United States. The same census found that 71% of DV programs in the United States (estimated to be 1,914 total programs) provide emergency shelter options (2022). These numbers show the incredible impact shelter has on survivors’ everyday lives; however, shelter is often an unavailable service due to lack of space. NNEDV found that over 6,000 requests for housing and emergency shelter in one day were unmet due to lack of resources (2022).

COVID-19 Pandemic

Impact on Domestic Violence. Domestic violence is defined as “a pattern of abusive behavior in any relationship that is used by one partner to gain or maintain power and control over another intimate partner,” and includes abuse that is physical, psychological, economic, or sexual (U.S. Department of Justice, n.d.). Both advocates (Figueroa & Torres, 2021) and researchers (Straka & Montminny, 2008) stress that regardless of the way(s) abuse is perpetrated in the relationship, DV involves an abuser seeking power and control over their victim. Maintaining control over someone is often accomplished at least in part by isolating the victim from friends and family, hiding the severity of the abuse from others, and preventing the victim from receiving help or support (Duron et al., 2021; Heron et al., 2022). Such isolation was exacerbated by the COVID-19 pandemic. To combat the spread of the virus, people were encouraged or even required to remain socially distant and isolated from people outside their household (Campbell, 2020). While essential to the pandemic response, implementing these lockdown measures was carried out without sufficient consideration of how these restrictions would impact the issue of DV (Kaur & Koli, 2020). At the same time, the pandemic created increases in other situational factors known to increase DV, such as unemployment and alcohol consumption (Das & Roy, 2020; vora et al., 2020). Under both these lockdown restrictions and other additional stressors, DV quickly became the “pandemic within a pandemic,” as victims were left with no choice but to quarantine with their abusers, increasing abuse and decreasing opportunities for seeking help (Dodd, 2021; Kofman & Garfin, 2020).

It is unsurprising then that rates of DV during the COVID-19 pandemic increased across the globe (Women, 2020). There were several early indicators that DV levels were increasing worldwide (World Health Organization, 2020). Reports of DV in many countries increased after lockdowns began, including a of 40–50% increase in Brazil, 30% in France, and 300% in China (Campbell, 2020). There was a 75% increase in DV Google searches in Australia after the first detection of COVID-19 there (Poate, 2020). Argentina, Cyprus, and Singapore all saw an increase of at least 25% in reported cases (United Nations, 2020). Prominent organizations addressing DV in Great Britain saw a 150% increase in web searches (Kottasova & Di, 2020). The United Nations estimates suggest that three months of quarantine likely increased global DV by 20% (United Nations Population Fund (UNFPA), 2020).

The United States saw similar trends in many areas. Available data shows an increase in many cities in both DV related arrests and 911 calls (Boserup et al., 2020). One review of 18 empirical studies found that not only did DV increase globally at a mean effect size of 0.66, but that effect size grew to 0.87 when only studies from the U.S. were considered (Piquero et al., 2021). In Georgia, DV calls, including calls from first-time reporters, increased 42–79% at the onset of the pandemic (Braverman, 2020; Burns, 2020; Evans, 2020). A study across 22 states estimated that DV increased by 5% between March and May of 2020 alone as restrictions forced more people to stay home (Hsu & Henke, 2021). While reports of DV decreased in other areas, this is unfortunately not indicative of a decrease in levels of DV. In many cases, victims did not have opportunities to report abuse and may have felt safer staying in an abusive environment and not risking exposure to the virus. The pandemic also coincided with a severe uptick in firearm purchases across the United States, and a subsequent rise in DV related firearm injuries (Schleimer et al., 2021). Additionally, although DV cases declined in some cities, statistics for the most violent forms of DV in all three big cities analyzed in one study remained constant (Li & Schwartzapfel, 2020).

Impact on Shelters. With increased difficulties for survivors and worsening rates of abuse, this pandemic has called attention to the already existing need for further DV prevention efforts and the expansion of current preventive measures (Ertan et al., 2020). DV shelters have had to overcome numerous obstacles to continue to provide services to survivors throughout the pandemic. For example, in Arizona, 100% of programs experienced shortages in fundraising, 79% did not have the space for recommended physical distancing, and 62% reported staffing shortages caused by pandemic-related resignations or staff illness (Arizona Coalition to End Sexual & Domestic Violence (ACESDV), 2020).

Previous research has identified several factors that make it especially difficult for DV shelter staff to control the spread of various coronaviruses. Given the communal nature of most shelters, illness prevention has long been especially challenging for these agencies (Gross & Rosenberg, 1987). It is common for shelter residents to utilize services at multiple locations, increasing exposures between residents (Farrell et al., 2020). Furthermore, public health recommendations are not always feasible for those living in shelters (Leung et al., 2008). Physical distancing can be impossible due to the sleeping arrangements and supplies such as hand sanitizer and face masks can be difficult for shelters to obtain (Farrell et al., 2020; Leung et al., 2008). Communication is another challenge, and respondents in one study identified the needs for a more effective system to disseminate updates and recommendations for disease control (Leung et al., 2008).

Further challenges for shelters included a lack of structural support. With the globe in crisis, DV was not a priority for officials. As non-profit organizations, DV shelters rely heavily on fundraising, almost all of which had to be halted during the pandemic (ACESDV, 2020; Domestic Violence Awareness Project, 2020; Merryman, 2021). Many were left wondering how their agencies would fund the increased demands brought about by the pandemic, such as personal protective equipment (PPE), hotel stays for those unable to keep all clients in the shelter, and additional food and personal items for survivors mandated not to leave the shelter (Nnawulezi & Hacskaylo, 2021). Additionally, a review of DV literature in times of disaster found that a lack of structural supports directed explicitly at DV services leads to increased victimization (Medzhitova et al., 2022).

NNEDV, a national organization offered recommendations for DV shelters during the COVID-19 pandemic. Confidentiality, an important regulation in DV shelters, still applies. Shelters cannot release and identify information or health status information to anyone outside of a legal mandate, meaning if their states do not require shelters to report cases of COVID-19, DV shelters will not disclose this information. NNEDV encouraged shelters to take precautionary steps to prevent the illness’s spread whenever possible and not house people in dormitory-style settings. However, unlike recommendations from other organizations, NNEDV states that DV shelters should not screen clients for COVID-19 and health status should not prevent survivors from accessing any program resources (NNEDV, 2020). While likely to be the best scenario for survivors actively fleeing their abusers, these recommendations create health risks for DV shelter staff and other residents.

Impact on Shelter Staff. Before the COVID-19 pandemic, DV shelter workers were already at an increased risk for compassion fatigue and secondary trauma (Bell, 2003; Voth Schrag et al., 2021). The terms compassion fatigue and secondary trauma used inter-changeably, although there are some distinctions (Bride et al., 2007). Compassion fatigue is defined as a provider’s reduced ability or interest in responding empathetically to their clients (Figley, 1995). Secondary trauma is essentially a form of exposure-induced posttraumatic stress disorder, and can lead to fear, trouble sleeping, intrusive thoughts, or avoidance of things related to the trauma (Bell, 2003; Stamm, 2005). Both can be caused by repeated exposure to vicarious trauma or hearing about traumatic events others have experienced, thereby being regularly confronted with the harsh realities of the world (Bell, 2003).

A study carried out by Chris Brown and colleagues (2020) before the pandemic found that while 94% of shelter workers believed their work had a positive impact, 36–39% reported compassion fatigue or secondary trauma because of their work. Just over half of shelter workers in this study reported feeling frustrated by their job, and challenges identified included lack of funding and housing, communal living issues, and mental health (Brown et al., 2020). The pandemic then not only created additional risks and complexities for victims of DV, but also added new frustrations and challenges for an already chronically overburdened workforce.

In addition to forcing them to face potential exposure to COVID-19 through their work and the corresponding health risks, the pandemic exacerbated the already existing struggles for shelter workers (Nnawulezi1 & Hacskaylo, 2022). For many, their typical strategies for engaging in self-care were no longer possible with mass business closures and lockdown orders, increasing emotional burdens (Burd et al., 2022). For example, compassion fatigue increased, with 94.6% of shelter workers in one study reporting that their work had become more stressful since the pandemic began (Wood et al., 2020). Being forced to impose restrictions on people fleeing a controlling abuser out of necessities for public safety, many workers experienced feelings of guilt. (Burd et al., 2022). As many began to work from home, shelter workers not faced only the common technology-related challenges many did, but also a unique issue from this blurring of boundaries between work and home: doing trauma work from home. This created problems for client privacy in addition to mental health struggles resulting from the new lack of separation between these two worlds (Trudell & Whitmore, 2020).

The work of implementing safety measures itself created significant burdens for staff, both physically and mentally. Staff were tasked with staying abreast of constantly changing rules and guidelines from public officials, adjusting their own rules based on those recommendations, and communicating changes to both other staff and clients (Wathen et al., 2022). The regular cleaning of shelter spaces to prevent the spread of infections also added an additional physical demand that many found tiring (Women’s Shelters Canada, 2020). For some, their work duties now expanded to include not only cleaning, but also things like cooking and meal delivery for sick or otherwise isolated clients (Burd et al., 2022). Despite the increase in the amount of work and work-related stress, 73.6% of shelter workers reported that their clients’ safety had decreased since the beginning of COVID-19 (Wood et al., 2020). There is a need for research that examines not only these challenges, but also the strengths and resiliency of DV shelter workers who continued to perform this already taxing work as it was made even more daunting and demanding by the challenges of the COVID-19 pandemic.

Current Study

The unprecedented circumstances of the COVID-19 pandemic and its substantial impact on DV survivors warrants a closer look at experiences of both survivors and those who serve them. Currently, the academic literature lacks information on the challenges faced by DV shelter workers and leadership during the COVID-19 pandemic, including action steps to overcome those challenges. This study sought to understand the experiences of DV shelter staff and leadership and how they navigated the early stages of the pandemic, including the consequences of actions on survivors and staff. This study seeks to inform future actions through policy recommendations and prepares shelters for future unforeseen events.

Methods

The research team collected study data using an online Qualtrics survey administered to DV shelter staff and administrators. Participants were recruited using a modified census sampling method distributed in two separate ways. Initially, in June of 2020, the research team sent the survey to the 51 DV coalitions in the United States (50 states and District of Columbia) for distribution to the programs in their state. After one month, in July of 2020, the research team contacted the states’ shelters directly if a state had less than ten responses at that time. The survey closed in August of 2020.

Measures

The online survey contained multiple choice, Likert scale, and open-ended questions. Respondents answered demographic information about themselves and their agency, such as the number of beds available for survivors and their state. They were then asked about their agency’s preparedness and response to the pandemic, including policies for staff and survivors, and both feared and actual changes to staff schedules. Finally, shelter staff was asked about the challenges faced by both themselves and their clients, and how they were able to overcome them.

Data Analysis

The research team used SAS OnDemand for Academics for statistical analysis. Univariate and bivariate analyses were conducted to identify patterns in the data. Thematic analysis was conducted using Virginia Braun and Victoria Clarke’s six-step process (2006) After the data was collected and cleaned, initial codes were created. Themes were then identified among those codes, and the data was revisited to revise and finalize these themes. This paper focuses on the analyses related to either staff overall or specifically direct services staff (rather than leadership-specific analyses), and the impacts on the survivors served by the responding participants.

Results

Participants

Three-hundred and sixty-eight (368) DV agency staff completed the study survey. Participants included 167 direct services staff (45.38%), 180 agency leaders and administrators (48.91%), and 21 other positions such as outreach, fundraising, or technology support (5.71%). As the survey was anonymous and did not collect identifying information, it is possible that multiple staff from the same agency responded. Respondents were overwhelmingly female (95.20%) and predominantly white (75.45%). Participants were overall highly educated, with 76.95% (n = 257) of those who reported their education having a bachelor’s degree or higher. Participants also had an overall high level of experience working at their agency, with 63.15% (n = 233) of those who reported their length of employment at their agency for three years or more (see Table 1 for full participant demographics). The location of agencies included rural (45.65%), urban (31.79%), and suburban (22.55%) areas. Shelter size varied: 14.13% had ten beds or fewer, 44.60% had 11–25, 28.53% had 26–50, 8.03% had 51–75, 2.77% had 76–100, and 1.94% had 100 beds or more. The distribution of participants in this sample across the United States Department of Health & Human Services regions can be seen in Fig. 1.

Table 1 Participant Demographics
Full size table
Fig. 1
figure 1

Distribution of Study Participants across U.S. Department of Health & Human Services (HSS) Regions (n = 368)

Full size image

(Office of Intergovernmental and External Affairs (IEA), 2021)

Preparedness for the Pandemic

Only 44.44% (n = 160) participants indicated knowing that their shelter had an existing policy for infectious diseases prior to COVID-19. Some (16.11%, n = 58) were unsure if their shelter had an infectious disease policy prior to the pandemic. Overall, most staff (53.46%) felt that their agencies were “somewhat prepared” for the pandemic, with 6.65% feeling very prepared, 34.63% feeling “somewhat unprepared”, and only 5.26% feeling “not prepared at all.” In some cases, this lack of preparation was frustrating for staff members, one of whom reported that she had notified her agency leadership months prior to the virus’ emergence in her state, but that “they did nothing until it was too late.”

Preventive Methods

With the onset of the pandemic, most DV shelters took measures to prevent the spread of the virus within their shelter to clients and their staff members. Respondents indicated varied preventative actions implemented early in the pandemic (see Table 2 for most common prevention methods). In addition to the methods shown below, several shelters (n = 18) prevented COVID-19 by reducing the capacity of their shelter to allow space for social distancing. This included housing residents in separate rooms with only themselves and their children if they had any, rather than some having roommates as is often ordinarily the case. Other strategies included requiring masks, increasing remote and outreach services, halting group programming and collective meals, providing disposable dishes and utensils, placing mini refrigerators in each room to limit shared kitchen use, scheduling blocks of time for shared spaces such as showers to prevent overlap, educating clients about the virus, and regulating clients leaving the shelter.

Table 2 COVID-19 Prevention Methods (n = 362)
Full size table

Out of the 89 shelters that relocated residents, 97.75% (n = 87) relocated at least some of them to local hotels. Others (n = 14), including some who utilized hotels, relocated people to separate spaces within the facility or in unoccupied transitional housing units. A few (n = 3) required survivors to find their own housing and supported them financially through the relocating process. One shelter reported relocating at least one resident to a local hospital. Shelter staff stayed in contact with relocated survivors through face-to-face meetings (n = 10), food and supplies drop-offs (n = 9), phone calls (n = 3), and other remote services such as video calls (n = 2).

Response to Positive Cases

When preventive measures failed, shelters responded to positive cases with varied methods. Between June and August of 2020, 120 participants in this study reported at least one resident housed by their agency either testing positive (n = 41) or exhibiting COVID-19 symptoms (n = 79). After a resident tested positive, some responded by isolating the positive client to a section of the shelter (n = 68), increasing cleaning (n = 72), closing the shelter to new residents (n = 24), closing the shelter completely and relocating all residents (n = 7), and relocating COVID-positive residents (n = 51). Those that relocated positive residents utilized hotels (n = 46), separate areas within the shelter or transitional units (n = 12), survivor homes (n = 2), and a respite facility (n = 1) to do so.

Shelter Policies

Length of Stay Policies. The implementation of preventive measures required the implementation of new policies for both residents and staff. Many shelters have a length of stay policy, which enforces a maximum time clients can receive residential shelter services (e.g., 30 days). The COVID-19 pandemic impacted survivors’ ability to secure employment and stable housing, resulting in many shelters adjusting or temporarily eliminating their length of stay policies in response to the pandemic (see Table 3). Some also reported that they “paused” the survivor’s stay during lockdowns, meaning they maintained their usual length of stay policies, but did not count any time the survivor was sheltered during COVID lockdowns towards their length of stay.

Table 3 Changes to Length of Stay Policies (n = 346)
Full size table

Movement Outside of Shelter. With nationwide and statewide lockdowns and government-implemented shelter-in-place orders, shelter leadership had to determine their own survivor policies for movement outside of the shelter. Of the 342 respondents who answered this question, the majority (58.48%) recommended that residents not leave except for essentials but did not make that recommendation mandatory. Others reported that their programs restricted survivors from leaving the shelter except for essentials (35.09%) or restricted survivors from leaving for any reason (2.33%), allowing residents to leave the building to spend time outside, but not to leave the premises. Some shelters (9.65%) did not use any restrictions or recommendations. Some shelters that allowed clients to leave enforced nightly curfews, did not allow staying out over the weekend, or did all necessary shopping for residents to reduce their need to leave. One respondent’s answer exemplifies how preventing COVID-19 exposures was uniquely difficult for DV shelters in particular: “We are empowerment based so by rule we cannot restrict movement.” Still, other staff reported that clients who wished to visit “high-risk” areas were forced to exit the shelter.

Further challenges were created by survivors who were essential workers and were still working in the community during lockdowns. Staffs in this situation used a variety of methods to contain virus spread, including not allowing survivors to work during lockdowns and requiring removal of clothing upon returning from work. Some reported that their agency took extra precautions to isolate survivors employed as essential workers from other residents, and one respondent’s shelter even implemented a protocol for confirming residents’ work schedules and work-site safety environment (see Table 4 for common restrictions).

Table 4 Restrictions Placed on Employed Survivors (n = 344)
Full size table

Cleaning Responsibilities. Most staff reported that their agency (n = 135) split the responsibility for cleaning and disinfecting the shelter between staff and clients. One-third (33.33%) said their shelter split the duties evenly, and 35.56% put the responsibility primarily on staff, with some placed on residents. Only 7.41% placed the responsibility mostly on residents, and 17.78% placed all the responsibility on staff. A few (2.96%) hired outside cleaning services or relied on hotel staff for residents placed in hotels (1.48%). In the majority of cases then, whether they were completely or only partially responsible for these tasks, the cleaning required meant additional labor performed by this already overburdened workforce.

Satisfaction with New Policies. Most direct service workers were satisfied with the policies implemented for both staff and residents. These results can be seen in more detail in Table 5. When responses were converted to a numerical Likert scale, with “extremely” dissatisfied” set at one up to “extremely satisfied” at five.

Table 5 Staff Satisfaction with Shelter Policies (n = 135)
Full size table

Effect on Staff

The early stages of the pandemic proved to be a difficult time for most DV shelter staff. One participant believed the pandemic had been even more difficult on staff than on survivors. While navigating the pandemic for themselves and their own families, shelter workers also bore the responsibility of caring for the survivors in their shelter. “We had families in all our shelters and in motels throughout two counties. Feeding families and picking up and delivering their laundry and groceries was exhausting work for our staff.” This increased workload was not always accompanied by the kind of recognition other professions saw: “I feel that front-line workers at DV shelters have not been acknowledged to the same level as essential workers in other fields such as healthcare. Especially in the early days of the pandemic, it was stressful coming to work and worrying about possible exposure to covid-19.” And some shelter workers did in fact become infected with COVID through their work, which then rippled into further challenges for their agency as well: “Staff getting sick posed a different kind of challenge.” Despite these negative consequences for many, most shelters made every effort to protect their staff while continuing to serve clients.

Personal Protective Equipment (PPE) Availability. Most frontline workers were satisfied (77.04%) with the amount of PPE offered at their shelter. This included 26.67% who were “somewhat satisfied” and 50.37% who were “extremely satisfied.” Of those who did not report being satisfied, 6.67% were “neither satisfied nor dissatisfied,” 7.41% were “somewhat dissatisfied,” and 8.89% were “extremely dissatisfied.” PPE commonly provided includes masks, hand sanitizer, cleaning supplies, and face shields. Interestingly, one participant identified the “option of meeting with a licensed therapist” as PPE. While most shelters were satisfied with PPE provided, one respondent did not indicate that for their agency, “obtaining the necessary PPE products was difficult and an unexpected problem.”

Funding Challenges. One of the elements that made acquiring necessary supplies difficult was the lack of funding. For many shelter staff, funding was reported as huge issue for many reasons, including increased hotel usage for quarantining clients, increased food/supply shopping to reduce need for clients to go out, PPE for clients and staff, and loss of fundraising opportunities. In some instances, key sources of revenue themselves were harmed by the pandemic, with most fundraising efforts and events coming to a sudden halt. As one participant stated, “Non-profits survive on donations and events.” In other cases, it was not a lack of funding overall, but rather a lack of flexibility in the terms of grant funds to allow them to be spent on the agency’s most pressing needs. As another participant shared, “Having no flexibility with our grant funding has been the most problematic.”

Staff Schedules. Many direct service workers (n = 135) were “worried” (32.59%) or “very worried” (11.85%) about the reduction of their hours due to the pandemic. Others were “not worried” (17.78%) or “not worried at all,” (23.70%) and a few (14.07%) were “neither worried nor not worried.” Despite this initial worry, very few workers faced a reduction in their hours, as can be seen in Table 6. In fact, many direct service workers saw increased hours due to increased challenges and demands. Some workers chose to take time off because of the pandemic or were forced to take time off after contracting COVID. Between June and August of 2020, 40% of respondents (n = 135) had someone working in their agency take time off out of fear of contracting COVID, and 36.30% of shelters had at least one staff member who had tested positive. It is worth noting that this was higher than the number of participants whose shelter had any client who tested positive (33.61%).

Table 6 Impact of COVID-19 on Staff Schedules (n = 136)
Full size table

Challenges for Staff. Staff were tasked with maintaining a clean shelter environment, keeping the shelter running, and enforcing new policies. Keeping the shelter space clean and sanitized posed the least challenge for staff and was reported to be “not difficult at all” by 26.16%, “not difficult” by 30.23%. Only 5.52% reported that it was “very difficult” and 18.90% said it was “difficult.” With schools and daycares closing across the country, DV shelter staff also faced added challenges in keeping survivors’ children safe and occupied. 26.53% found this task “difficult” and another 11.37% found it “very difficult.” Others found it “not difficult” (20.41%) or “not difficult at all” (12.24%). One shelter advocate noted, “Finding space for 50 children to do online school daily is a huge hurdle.”

As shelters enacted movement restrictions for survivors both within and outside the shelter, direct services staff were responsible for enforcing such policies. For policies restricting movement within the shelter, 14.62% found this “very difficult,” 30.70% “difficult,” 14.62% “neither difficult nor not difficult,” 17.25% “not difficult,” and 9.65% “not difficult at all.” Workers found enforcing restrictions on movements outside of shelter even more difficult. 26.20% reported this was “very difficult,” 25.44% “difficult,” 18.71% “neither difficult nor not difficult,” 11.70% “not difficult,” and 6.43% “not difficult at all.”

One of the most difficult tasks for staff proved to be managing the anxieties of residents in the shelter. The COVID-19 pandemic produced anxieties for everyone, but this was especially true for survivors of DV who were already in the midst of one of the most unstable and chaotic times of their lives after fleeing their homes and moving themselves and their children into a shelter. Many (37.72%) staff found managing the anxieties of survivors “difficult” and another 15.79% found it “very difficult.” Only 6.73% found it “not difficult at all,” 17.25% “not difficult,” and 18.71% “neither difficult nor not difficult.” One shelter advocate shared that their agency required masks, which meant having to threaten to exit residents who repeatedly did not wear a mask, even though many of their clients had been victims of strangulation at the hands of their abusers. This advocate noted that they felt this was “a horrible thing to have to do to survivors.” In situations such as this, domestic violence shelter workers were forced to take on the emotionally intensive tasks of enforcing rules antithetical to their usual trauma-informed procedures. The alternative was to accept increased risks of infection within the shelter.

Another participant further explained how the empowerment model important to so many DV shelters in some instances proved counterintuitive to preventing COVID-19 exposure:

It has been difficult to ensure that residents are social distancing because we work from an empowerment model and therefore we are not telling residents that they absolutely cannot leave the shelter unless it is essential. As a result, some shelter residents would leave the shelter for non-essential reasons which could have made us more vulnerable to COVID-19. It has been difficult to enforce any kind of restrictions on residents coming/leaving shelter for essential or non-essential reasons without disempowering them.

Effect on Survivors

Challenges for Survivors. DV shelter staff work closely with the survivors they serve. In our sample, even when residents had to be relocated outside of the shelter, staff still maintained regular contact with them. We asked all staff what factors of the pandemic they thought created the most difficulties for their clients. Most staff reported that lack of childcare, economic insecurity brought on by the pandemic, and loss of social connections had the greatest impact on their clients. The full results can be seen in Table 7.

Table 7 Pandemic Impacts on Residents (n = 333)
Full size table

Helpful Responses. All participants were asked what staff responses to the pandemic were most helpful for the survivors in their shelter. The most common answer to this question was support and unwavering commitment from the shelter staff. One advocate said it was “The overwhelmingly strong show of staff support the entire time.” Another said that despite all of the changes in both shelters and the world in general, the most helpful thing their staff did was “letting [survivors] know our support for them remained unchanged.” This included maintaining contact despite the closing or changes of operations in physical spaces. One respondent said, “Our staff was never inaccessible to survivors. Even when we temporarily closed the office. We were still available by phone, email, or video calls.” Staff availability was enhanced by the shift to virtual services as advocates “still provided services from home when the state was locked down.”

Many staff reported that this continuation of services was most important, citing the most helpful response being “that we continued to provide trauma informed, supportive services to individuals and families.” This included continuing to provide survivors with a safe place to stay. Many felt that the precautions taken to prevent the spread of COVID-19 within the shelter even provided reassurance for survivors. As one participant put it, “Most of our guests were concerned about the spread of COVID-19 inside the shelter and it was comforting to know that our staff was serious about their health and safety.” Another said simply, “They knew they were still safe with us.”

Another helpful response from staff included providing survivors in shelter with both information and transparency. In many cases, shelter staff were able to provide shelter residents with accurate information as it evolved. This was helpful for them in tuning out all the social media misinformation that was circulating. Transparency about the changes that were occurring was also “key in keeping the anxiety down,” or “[taking] anxiety out of the equation.” When survivors did feel negative mental health impacts from the pandemic, staff felt they were able to be helpful by “[validating] their fears and exhaustion,” and letting them “[know] they were not alone in their fears and frustrations.” Mental health services and support groups were also cited as being particularly helpful for survivors during this time. Support groups, many of which were offered virtually, provided a space “where we could talk about the pandemic and how [survivors] were feeling.”

Other things staff felt were helpful for clients included the use of hotels and spacing out residents to allow for separate living spaces with only one client or family per room, extending the length of stays, and providing financial support. Just as PPE was provided for staff, many provided it to residents as well. One person said, “The most helpful [thing] was probably providing all sorts of cleaning supplies, hand soap, and hand sanitizer to everyone living in shelter.” Financial support was also cited as being important to clients, both from shelters themselves and from stimulus checks. Finally, staff felt it was helpful for survivors when they were able to provide activities to keep clients and their children busy and entertained, which required staff “becoming much more innovative with creating ways to engage and remain busy.”

Innovative Practices & Lasting Impacts

As this paper has shown, when the pandemic placed obstacles in their way, DV shelter staff developed innovative practices to overcome them. One participant said, “The early days were very trying as things changed so quickly. Staff rose to many challenges and had great creativity in meeting survivors’ needs.” Staff showed incredible ingenuity and flexibility in adapting to the ever-changing needs of their clients. Another shelter worker stated, “We had to be nimble and respond very quickly to things that changed from day to day.” The earlier sections of this paper detail a variety of different responses to the various challenges posted by the pandemic, and without a unified source of information during this time, shelter workers are to credit for these innovations that allowed services for survivors to remain available throughout the entirety of this harrowing time.

Lessons from the pandemic will likely change the way many DV shelters operate day to day well into the future. As one participant put it, “We expect the effects of the pandemic to last for us long after it is over.” Some staff reported that their program will keep some of the changes brought about by the pandemic because they have proven to be helpful. In some ways, COVID sped up innovations to shelter operations. One shelter advocate said, “It prompted us to launch a text hotline, web chat hotline, and virtual support groups…all things that may not have happened for years otherwise.” Lessons learned from the COVID-19 pandemic will likely lay the groundwork for responses to future crises as well. Most participants (89.52%) felt that their agency was prepared for a similar situation in the future, with 46.11% feeling “somewhat prepared” and 43.41% feeling “very prepared” for another pandemic or similar event. Only 5.09 felt “somewhat unprepared,” 4.49% felt “neither prepared nor unprepared,” and 0.90% felt “not prepared at all.”

Discussion

The results of this study show that DV shelter staff reacted to the challenges brought about by the COVID-19 pandemic in a variety of different ways. In the face of enormous challenges and unprecedented frustrations, workers showed incredible resilience and resourcefulness throughout these early stages of the COVID-19 crisis. Across the board, agencies did whatever was needed to keep their services available to those who needed them. Staff worked through uncertainty, risks to their own health, and both increased and ever-changing job responsibilities. Their actions allowed services for survivors of DV across the United States to continue, which was crucial especially given the rising rates of DV brought on by the pandemic and resulting lockdowns. Disasters such as this often leave their mark on DV rates long after the crisis is over (Sety et al., 2014). It is our hope that what DV shelter staff have learned from the pandemic can inform policies and practices that help enhance both the day-to-day operations of shelters, and their preparedness for crises in the future.

Implications & Recommendations

Given the already taxing nature of DV work, we recommend more research which focuses on the well-being of staff to ensure their overall well-being and longevity doing this work. This is even more true in the face of extenuating circumstances such as a global pandemic. In addition to increased research, we offer five recommendations for DV shelters and their funders. Our first recommendation is fewer restrictions on the use of funding. Previous research has shown that flexible funding shows promise in preventing homelessness among survivors (Sullivan et al., 2019), and our results support its utility in allowing shelters to adapt to the specific needs of their clients. For example, shelters in our sample spent funds on unexpected needs such as COVID tests, hotel stays, cleaning supplies, PPE, mini fridges, and mental health support for staff.

Our second recommendation is increased transparency for both direct services staff and clients on the reasoning behind all existing policies. Our results indicated that being more open with survivors and staff led to increased satisfaction with policies and less anxiety. This openness is also key to an empowerment model of DV services by providing survivors with all the information they need to best make decisions for themselves and their children.

Thirdly, we recommend shelters increase accessibility of their services through technology. This can include virtual support groups and one-on-one meetings through platforms like Skype or Zoom, crisis text lines or online chats, and keeping up-to-date information available on websites and social media. Chat lines allow survivors who cannot talk on the phone, either because the abuser is present or monitors their phone records, to reach out for help and support. Text lines also allow those unable to talk on the phone to reach out for support, and both options may be easier and less nerve-racking to use when reaching out for help given the personal and sensitive nature of DV victimization. While we do not mean to suggest that online services can, or should, ever fully replace those provided in person, they offer an opportunity to provide some services to survivors who would not otherwise be able to receive them.

Our fourth recommendation is housing individual families separately whenever possible. Many staff found that housing families separately during the pandemic was one of the most helpful things they were able to do for survivors. While these measures were likely a temporary solution to prevent the spread of the Coronavirus, we recommend keeping this practice wherever possible. This may require seeking out additional and diversified funding to design or re-design shelter spaces to accommodate this. In addition to the reduced health risks and increased feelings of safety found here, separate rooms would allow for greater privacy for survivors during a vulnerable time in their lives, prevent the spread of infectious diseases throughout shelter, and lower conflict between residents; a common issue in DV shelters (Gross & Rosenberg, 1987; Lyon et al., 2008; Vawnet, 2009).

Finally, we recommend enhanced wellness support for direct services staff. As this study showed, the COVID-19 pandemic introduced new and unprecedented stressors for DV shelter workers. However, the pandemic also exacerbated mental health challenges that shelter workers were already facing prior. DV shelter workers should have access to mental health services, be encouraged to set boundaries between their work and their personal lives, and have opportunities for taking time off without penalty for mental health-related needs. This is crucial not only to the well-being of staff, but also to the effectiveness of DV agencies’ services, as staff longevity allows shelters to maintain a trained and knowledgeable staff to serve survivors. Without mental health support DV shelter workers are likely to burn out and leave the work altogether (Merchant & Whiting, 2015). One salient example from these results that highlights the importance of staff wellness to both their well-being and work performance is the participant in this study who listed the ability to meet with a counselor or therapist as PPE, or in this case, something meant to protect staff from work-related hazards.

Limitations

It is possible that shelter workers did not have all the most up to date information on rapidly shifting policies during the early months of the pandemic. While the survey was anonymous, as with any self-report survey, it is possible that respondents’ answers are altered by social desirability bias, especially given the sensitive nature of the topic of DV. The survey was only in English, only distributed to federally funded shelter programs. It is also important to note that the impact on survivors reported here is based on the perceptions of DV shelter staff, and not direct reports from survivors themselves.

Conclusion

Overcoming the challenges brought about by the pandemic required immeasurable resiliency and strength from DV shelter staff members across the country and will likely continue to create unforeseen hurdles for years to come. Agencies should take this crisis as an opportunity to learn and to better their operations for both their staff and their clients. More research should also focus on staff perspectives, as they work closely with survivors every single day, and critical DV services would not be possible without their hard work and dedication.