“Safety Is Elusive:” A Critical Discourses Analysis of Newspapers’ Reporting of Domestic Violence During the Coronavirus Pandemic

Abstract

The COVID-19 pandemic exacerbated incidences of domestic violence (DV). The framing of DV within media sources contributes to the public's understanding of DV. Using critical discourse analysis (CDA), this paper explores representations of safety within newspapers’ reporting of DV during the pandemic. The sample included newspaper articles (n = 31) from U.S. newspapers. The analysis involved multiple rounds of coding and employing “structured questions.” These articles depicted limited courses of action for DV survivors and represented safety as unattainable. Safety was constructed in four ways: homes are unsafe, social services are overburdened, government failures, and the elusiveness of safety. These discursive formations provide insight regarding “idealized” social responses to DV.

When you’re being bombarded with messages like ‘stay safe at home,’ that can be really hard to hear when home is not a safe place.

—Kelly Starr, Washington State Coalition Against Domestic Violence (Article 16)

Introduction

Due to the rapidly evolving nature of COVID-19, the media became an important mechanism for disseminating up-to-date public health information about the pandemic and raised early alarm bells regarding the safety of survivors sheltering with abusive partners (Kaukinen, 2020). Emerging evidence suggests that social conditions during the coronavirus (SARS-CoV-2) pandemic exacerbated the incidence and severity of abuse for survivors of domestic violence (Bradbury-Jones, 2020; Fawole et al., 2021; Kaukinen, 2020; Mazza et al., 2020). The heightened rates of domestic violence (DV) led some scholars to frame this epidemic as a “shadow pandemic” (Parry & Gordon, 2020) or a “pandemic within a pandemic” (Evans et al., 2020).

Beyond being a conveyor of information, the media is one of the primary “cultural supports” (Stark, 2009, p. xiii) that constructs, reproduces, and reinforces dominant social scripts about the causes and consequences of social issues and about the collective social actions in response to them. In many ways, media outlets operate as contemporary storytellers in modern society and are powerful sources of messages about what is perceived to be typical or acceptable (Storer & Strohl, 2016). Historically, the media has been critiqued for its reporting and framing of DV. For instance, McManus and Dorfman (2005) found that DV was reported less frequently, less representatively, and in less depth than other types of violent crime in newspapers. Several studies found that the media attributed blame to survivors for remaining in abusive relationships (Bullock, 2007; Bullock & Cubert, 2002; Fairbairn & Dawson, 2013; Taylor, 2009), minimized the seriousness of DV (Thaller & Messing, 2014), diminished the responsibility of perpetrators for their actions (Lee & Wong, 2020; Lindsay-Brisbine et al., 2014), elevated the role of police for providing survivor safety without recognizing racial inequities resulting from police practices (Singh & Bullock, 2020), and failed to position DV as a broader social issue (Bullock & Cubert, 2002; Storer & Strohl, 2016).

This paper examines the discursive constructions of survivors’ safety in newspaper accounts of DV during the first year of the COVID-19 pandemic in the United States. Centering this research on safety as a sensitizing discourse emerged inductively, as it was one of the primary discursive formations we found in mainstream newspapers’ construction of how the pandemic was impacting DV survivors. At the same time, safety is an important discourse to examine because it is one of the core rights that is frequently denied to survivors of domestic violence (Morgaine, 2009) and because achieving survivor “safety” has been a discursive anchor in the missions of many mainstream DV agencies in the United States (Thomas et al., 2015). By examining discursive constructions of safety, this study brings into relief dominant messages about survivor safety and how it should be achieved during the pandemic.

Defining Domestic Violence

Throughout this paper, we will use the phrase domestic violence because it is the terminology most frequently used in the public discourse in the United States, including the media articles under examination (Kaukinen, 2020). We acknowledge that this terminology differs by country and cultural context. For the sake of clarity, we are referring only to abuse perpetrated within the context of intimate relationships (commonly called intimate partner violence in the academic literature in the United States), not the broader nexus of family violence that includes child maltreatment or elder abuse. Domestic violence within the context of intimate relationships comprises a pattern of coercive behaviors (e.g., physical, verbal, and/or mental abuse, digital and physical surveillance, and stalking) used to maintain power and control in current or former relationships (Bowen & Walker, 2015). Additionally, though we do not endorse gender as a binary construct, we are using the terms woman/women/female throughout this article to be consistent with the language used in the articles under review.

Collective Responses to Domestic Violence

Contemporary societal responses to DV emerged out of early feminist organizing nested within the larger women's civil rights movement of the 1960s (Ake & Arnold, 2017). The early politicization of this issue is important because it framed DV as the product of systemic, patriarchal inequities and institutionalized misogyny. In recounting the history of the DV movement (historically referred to as the Battered Women's Movement), Fleck-Henderson (2017) describes a slow “co-option” where the “demands of providing services to domestic violence survivors threaten[ed] to eclipse the larger social change purpose” (p. 477). In other words, feminist scholars highlighted the underlying structural causes of DV such as gender inequity, workforce discrimination, and pay inequity, but as DV organizations evolved into more professionalized organizations, they implemented more easily funded need-based responses to pressing survivor needs (Lehrner & Allen, 2009). Such services include robust national networks of emergency shelter systems and coordinated social services such as 24-h phone hotlines, crisis counseling, youth services, and legal advocacy (Ake & Arnold, 2017).

As the movement evolved, interventions with a basis in criminal justice and undergirded with neoliberal rhetoric of law and order—such as mandated arrest and criminal protective orders—became ubiquitous elements of coordinated community response systems in the United States (Bumiller, 2008). Indeed, neoliberal discourses of managerialism and professionalism have been implicated in transforming “the battered women's movement from a social change movement rooted in feminism to services focused on treatment” (Mehrotra et al., 2016, p. 154). The turn away from feminist organizing principles is evidenced by a study of domestic violence coalitions which found that less than 10% (n = 51) reported to self-identify as feminist in orientation or purported to be informed by feminist theory (Barrett et al., 2016). The trend of depoliticizing DV led Mehrotra et al. (2016) to conclude that “large-scale efforts to mobilize communities and end violence are no longer at the center of DV work” (p. 156). Therefore, the work of DV agencies became framed primarily in social service terms rather than as structural or political ones.

The constellation of services offered at DV organizations is essential for providing emergency support to survivors. At the same time, it is important to note the unintentional harm that social service agencies (including DV agencies) have had—particularly for those from socially excluded populations including those that identify as Black or African American (Richie, 2012) or LGTQIA+ (Ford et al., 2013; Overstreet & Quinn, 2013). While the mainstream domestic violence movement has espoused a commitment to survivor-centered approaches grounded in the empowerment tradition in social work (Kulkarni, 2019), scholars have noted an inherent contradiction between these commitments and organizational policies that infringe on survivor agency and self-determination (Goodman et al., 2020; Gregory et al., 2017). Within social service organizations, scholars have also identified an inherent power imbalance within DV agencies between “advocates ‘helping’ and survivors being helped” (Goodman et al., 2020, p. 226). These practices can include restrictive policies within shelter environments (Glenn & Goodman, 2015; Gregory et al., 2017), compulsory parenting or financial literacy classes, and required involvement in the criminal justice system through policies like “no-drop” prosecution (Durfee, 2020). Other scholars have noted that social service agencies have often invited surveillance by government institutions (such as child welfare and criminal legal systems) into survivors’ lives, leading to increased and disproportionate risk of incarceration or family separation among communities of color (Goodman et al., 2020; Rhodes et al., 2010).

Evidence suggests that the many rules and practices at DV agencies may positively impact some survivors by imposing supportive structure; however, they may simultaneously negatively impact other survivors’ decisions to remain in shelter (Kulkarni et al., 2019) and their overall psychological well-being (Gregory et al., 2017). Another study reported that the majority of DV survivors were unprepared for the numerous trade-offs they encountered on their journeys for achieving safety from DV. These opportunity costs included loss of support networks, heightened physical safety issues from abusive partners, financial stability, disconnection from home and feelings of “rootedness,” loss of control over their parenting, and loss of freedom due to needing to restrict their activities (Thomas et al., 2015).

Framing Theory and Media Representation of DV

Media and communications scholars have employed framing theory to elucidate discernable patterns of the media's reporting and categorization of social issues, including what information is consistently included or excluded (Bullock, 2007; Campbell & Townsend, 2011; Carlyle et al., 2014; Estes & Webber, 2021). Media framing influences the public's understanding of social issues and desired societal and policy responses to these issues (Kuypers, 2009). Thus, an examination of media frames can shed light on dominant societal explanations of DV and offer an opportunity to rewrite harmful dominant social scripts, such as those that perpetuate stereotypes of those who have experienced DV (Campbell & Townsend, 2011; Storer & Casey, 2021).

Newspaper accounts of DV frequently omit important contextual information about how structural and systemic barriers contribute to DV (Bullock & Cubert, 2002; Fairbairn & Dawson, 2013; Gillespie et al., 2013; Ryan et al., 2006; Seely & Riffe, 2021; Sellers et al., 2014; Singh & Bullock, 2020). For instance, one study found that only 10–34% of newspapers included such context (Richards et al., 2011), and another reported that newspapers rarely framed DV as stemming from the perpetrator's coercive control and male privilege (Lindsay-Brisbine et al., 2014). Singh and Bullock (2020) found that 99% of newspaper portrayals of the 2013 reauthorization of the Violence Against Women Act did not mention structural causes of violence, such as economic marginalization or gender inequality, and articles that discussed law enforcement equated it with survivor safety, despite disparities in the treatment of survivors by law enforcement. Such framing divorces DV from its broader social milieu and minimizes the importance of coordinated societal responses to prevent domestic violence (Campbell & Townsend, 2011).

As of this writing, there has been limited discussion of the framing of DV in media during the COVID-19 pandemic. One exception is the content analysis of help-seeking messages in Univision, the largest Spanish-language media platform in the United States (Alvarez-Hernandez et al., 2021). The authors found that these messages were overly reductionist, lacked culturally sensitive context, tacitly reinforced messages that survivors are responsible for leaving abusive relationships, bolstered paternalistic notions that survivors need protection, and communicated that professional social service providers are available to provide support and tools to facilitate safety (e.g., protection orders, shelter, etc.).

Study Purpose

An analysis of mainstream newspapers is important because it provides an important lens for unpacking dominant societal representations about domestic violence. The media, like members of the general public, harbor significant societal stereotypes and misinformation about DV (Worden & Carlson, 2005). As will be discussed in further detail in the methodology section, we are less interested in the veracity of the reporting or whether these accounts are "right" or "wrong." An analysis of discourse is about identifying the shared understanding of a phrase that hovers “above the sentence or above the clause” (Stubbs, 1983, p. 1). By exploring representations of safety within newspapers’ reporting of DV during COVID-19, this study presents new opportunities to reflect on current understandings of DV and dominant DV response paradigms. Thus, the purpose of this paper is to examine how mainstream U.S. newspapers represented survivors’ safety during the pandemic, described pathways for achieving safety, and idealized societal responses to support survivor safety.

Methodology

This study employs critical discourse analysis (CDA), which examines the constructive function of language-in-use (Gee, 2011). As is consistent with CDA methodology, our approach is informed by a poststructuralist epistemology in which discourse is understood as collective “systems of representation” (Greckhamer & Cilesiz, 2014), and shared meaning-making is undergirded by systems of power and privilege (authors, masked for review). Through an analysis of language-in-use, CDA is an iterative process of uncovering both collective meaning-making and the impact of dominant constructions on the understanding of complex social issues such as domestic violence. The everyday use of language serves to reinforce social practices, privilege dominant ideas or beliefs, reify social norms regarding what is considered “right” or “normal,” and center dominant knowledge systems while rendering others invisible (Gee, 2011). Thus, a primary goal of CDA is to make clear how the discursive functions of language maintain systems of inequality and epistemic injustice. Though underutilized, CDA is an important tool for investigating DV because enduring social norms regarding survivors have contributed to societal-level stigma and victim-blaming that influence societal-level responses to this social issue (authors, masked for review).

Data and Sample

Data were drawn from articles (n = 31) in the four newspapers with the largest circulations in the United States: The New York Times, The Los Angeles Times, The Washington Post, and USA Today. We initially included the Wall Street Journal but found that it contained no substantive coverage of domestic violence in the context of the pandemic. To be included in the sample, articles had to be published from March 1 through December 2020 and needed to include content directly related to both COVID-19 and DV. Articles were retrieved using the NexisUni media search engine and cross-checked in Google News to ensure no inadvertent exclusions. We used the following search terms: domestic violence (this term will identify all articles containing the word violence, including intimate partner violence) and either coronavirus or COVID-19. Op-eds, weekly news summaries, and letters were excluded because they represent contextually different types of documents. A total of 291 articles were screened and excluded because they did not include substantive content on both domestic violence and COVID-19 or because they were duplicate articles. Our total sample (n = 34) is summarized in Table 1.

Table 1.

Newspaper Articles Analyzed.

Title	ID	Date	Periodical
With fewer eyes on kids, abuse reports plummet	1	4/25/20	Washington Post
“The worst is yet to come”; Isolation could raise the risk of domestic abuse; Experts say seclusion, stress from pandemic may worsen already strained relationships	2	3/25/20	LA Times
Home isolation is fertile ground for domestic abuse	3	3/19/20	USA Today
Recovering addicts find they’re suddenly alone	4	3/20/20	Washington Post
For abused women, a pandemic lockdown holds dangers of its own	5	3/24/20	New York Times
Police try to protect public while battling spread among own ranks	6	3/29/20	Washington Post
Restrictions trap many victims of domestic abuse with their abusers	7	4/2/20	Washington Post
State court leaders OK new rules for virus	8	4/7/20	LA Times
Murder rates were staggering. The virus has brought some quiet, for now	9	4/11/20	New York Times
Murder rates see steep decline: “It's taking people off the streets”	10	4/12/20	New York Times
A new COVID-19 crisis: Domestic abuse rises worldwide	11	4/6/20	New York Times
In pandemic's grip, Russia sees spike in age-old bane: Drinking	12	4/14/20	New York Times
Vodka is no disinfectant, but self-medication soars	13	3/15/20	New York Times
Cuomo and Trump spar over coronavirus aid to New York	14	4/17/20	New York Times
Danger being trapped at home	15	3/20/20	LA Times
Sheltering in peril	16	4/5/20	Washington Post
Crime in streets drops sharply but disturbance calls from homes on the rise	17	4/7/20	USA Today
Locked down, and more vulnerable to abuse	18	4/7/20	New York Times
Why a drop in domestic violence reports might not be a good sign	19	4/17/20	New York Times
Rihanna donates $2.1 million for domestic violence victims in coronavirus lockdown	20	4/10/20	LA Times
Uber donates rides to abuse victims; Pandemic has added hurdles for survivors	21	12/21/20	USA Today
DV survivors say housing group betrayed them	22	11/8/20	Washington Post
When staying home, or trying to leave, carries its own risk	23	5/17/20	New York Times
Abuse survivor: “I got my kids”; Escaping domestic violence in a pandemic	24	5/5/20	USA Today
Domestic abuse worsens in pandemic; The incidence and severity of injuries increased sharply, a Boston study finds	25	8/19/20	LA Times
It's been “such a weird year.” That's also reflected in crime statistics	26	7/6/20	New York Times
Failings in Britain leave victims of domestic violence in peril	27	7/3/20	New York Times
For abused women, a pandemic lockdown holds dangers of its own	28	6/18/20	New York Times
Witness to a hidden epidemic	29	6/19/20	New York Times
Virus deaths in New York hit lowest level since April 1	30	6/1/20	New York Times
Chinese city to unveil domestic violence database for those getting married	31	6/25/20	New York Times
Superior Court expands ability to hold virtual hearings	32	5/18/20	Washington Post
Doctors are COVID's first historians	33	6/14/20	New York Times
Domestic violence calls mount as restrictions linger: “No one can leave”	34	12/14/20	New York Times

Analysis

The analysis involved an inductive process rather than an a priori coding scheme or previously identified discursive formations. After the articles were accepted for inclusion in the sample, they were uploaded in Dedoose Analytic Software version 8.0.35. Each article was inductively coded by two analysts using an open coding process (Braun & Clarke, 2006). Examples of the kinds of codes initially identified included the following: community responsibility to help survivors, government responsibility to support survivors, home is a dangerous place, public health measures have exacerbated DV, and safety is achieved outside the home. All codes were clustered into more substantive categories comprising similar constructs (Saldaña, 2009). Throughout the coding process, we created a codebook and collectively discussed our application of these codes and categories to ensure interpretive agreement. Throughout the initial analytical process, we used interpretive memos both to reflect on our impressions of the data and to practice reflexivity, giving particular attention to how our personal experiences might have influenced the analysis process (Stige et al., 2009). The memos were also used to surface some of the emerging discursive constructions in the data.

We used the themes constructed from our initial coding process as entry points for our CDA analysis. To identify the discursive formations in the texts, we adapted LeGreco and Tracy's (2009) use of “structured questions” to help explore the functions of language within these texts, what Gee (2011) calls the building tasks of language. These building tasks included interrogating how language-in-use privileged and constructed various identities and social practices by reinforcing taken-for-granted understandings of what is expected, appropriate, typical, or the way things are or should be. The structured questions provide a tangible tool to sensitize the analysts to how systems of power are constituted and reinforced within the texts. Versions of this process have been described in other published CDA analyses (authors, masked for review), and elements of this staged approach to analysis have been outlined in a CDA methodology paper (authors, masked for review).

The structured questions employed in this project (Table 2) were inspired by Gee's (2011) building tasks of language and the lead authors’ analytical memos. By assessing the patterns across our answers to these structured questions, the domains of home and safety became the primary discursive anchors to conduct a more fine-tuned analysis. Lastly, we performed a final focused coding of all of the articles, concentrating on how home and safety were constructed in these texts.

Table 2.

Discourse of Home and Safety: Structured Questions.

What is the ideal role homes should serve in society?

How do survivors achieve safety? What are the barriers to achieving safety?

How do survivors achieve safety?

How has the global pandemic compromised survivors’ safety? How did the pandemic exacerbate DV?

How are survivor/perpetrator identities constructed in these texts?

What do these texts say about the root causes of DV?

Who is responsible for supporting survivors to achieve safety?

Reflexivity

Reflexivity is an important benchmark for rigor in qualitative research (Stige et al., 2009). The lead author is a qualitative researcher and former practitioner in mainstream domestic violence agencies; undoubtedly, her professional experiences sensitized her to dominant models of identifying and responding to DV and the limitations of these approaches. The second and third authors on the paper were Social Work doctoral students at the time of this writing. The second had limited experience in the field of DV and provided fresh insights into examining these data. The third author brought both practice and research experience to this project. Having worked on issues of DV in Nepal (where community-based strategies were used to address DV), she was attuned to ways in which DV could be addressed outside of the formal service systems. Both the first and third authors are mothers and were particularly cognizant of the disproportionate effects of the pandemic on women and primary caregivers.

Results

Across this sample, the majority of newspaper articles on domestic violence in the context of COVID-19 described how homes have grown considerably less safe during the pandemic. Safety was discursively constructed in four primary ways: (a) the home is not a safe place for survivors during a pandemic; (b) strained social services are unable to mitigate survivors’ risk; (c) inadequate government responses have contributed to survivors’ risk, and (d) safety is elusive for abuse survivors. It is important to note that throughout these articles, there was a notable absence of how particular demographic segments of DV survivors (e.g., BIPOC or homeless survivors) were impacted during the pandemic. One notable exception was an article that added this caveat: “Certain populations are especially vulnerable [to social isolation], experts say, including rural people, communities of color and people who identify as LGBTQ, who research shows experience domestic violence at equal or even higher rates” (Article 3). However, across the sample, very little attention was given to survivors’ intersectional identities.

Home Is Not a Safe Place for Survivors During a Pandemic

In contrast to the dominant representations of home as a place of refuge to wait out the pandemic, the articles in this sample described the home as a particularly dangerous place for survivors of DV. One stated, “As the nation struggles to contain the spread of the coronavirus, vast numbers of Americans are being told to stay home. But what happens when home isn’t safe?” (Article 3). Safety was framed as a privilege that many survivors cannot enjoy. language like “trapped” (Articles 27, 15, 24), “stuck” (Articles 2, 5, 11, 18, 21, 28, 33), and “no way out” (Article 7) recast home as a place of confinement rather than a place of refuge. For example, Article 15 quotes a source as stating:

People like her [abuse survivors] are the ones I’m thinking about during this pandemic—those for whom the world outside might be safer than their own home….What happens when the demands of social distancing mean the most vulnerable in our society might be distanced right out of existence? (emphasis added)

The articles also contained examples of abusers using the conditions of the virus as tools of coercive control. In one instance, “a husband was making [a survivor] scrub her hands until they bled” (Article 16). In another, a survivor recounted, “I started coughing, he was throwing me out in the street and said that I could die alone in a hospital room” (Article 28). While there were numerous instances of egregious physical abuse represented in these articles, these passages are distinctive because they illustrate how partners who used coercive tactics intentionally weaponized the virus. Furthermore, several articles described less access to informal supports (such as relatives and friends) as well as to services not directly related to DV (such as work and school).

When an entire society shuts down when children are home all day from school, when sports and gyms and social activities are all canceled when friends can’t leave their own families to help, when places of worship are shuttered, when everything that ever tempered a violent situation is suddenly, terrifyingly, no longer available. What happens then? (Article 15)

Despite some recognition of the importance of such day-to-day social networks to support DV survivors, they were rarely presented as part of the solution.

Strained Social Services Are Unable to Mitigate Clients’ Risk

Given that home was depicted as a dangerous place for survivors, external social service providers (particularly domestic violence agencies, hotlines, and court-based services) were positioned as the primary mechanism and ideal societal response to help survivors and their children attain safety. Article 28 equated DV advocates with “first responders.” One DV service provider stated:

If anything, this pandemic has made our services much more critical…With the pandemic going on, with people being mostly indoors and at home, it limits the opportunities survivors have to leave or call for help. There is a heightened sense of vulnerability and danger. (Article 16)

However, despite the expressed acknowledgment of these services, social service and criminal justice systems were portrayed as “overwhelmed” (Article 27), “nearing capacity” (Article 20), “overstretched” (Article 27), experiencing “overcrowding” (Article 5), and “straining” (Article 11). Article 15 asked:

Who keeps them safe in a world of unmitigated dangers? In times of natural disasters and great social upheaval—hurricanes Katrina and Sandy, for example—the systems we’ve created to aid domestic violence victims can get easily overwhelmed.

This passage is also noteworthy because phrases like “who keeps them safe” underscore paternalistic notions that only external entities like professionalized agencies can support client resiliency.

In addition to overburdening social service networks, the pandemic was seen as further limiting clients’ ability to access needed services. Article 2 stated that “The strict measures put in place to curb the spread of the virus in the city have raised hurdles for people seeking help…. For someone who is in an abusive relationship, this is kind of a worst-case scenario…”

Although most articles reinforced the perspective that agencies were struggling under the weight of the pandemic, a handful represented agencies as resilient and innovative in their approaches to meet client needs. For example, Article 5 stated, “Shelters across the country are adapting as best they can while trying to keep pace with constantly changing virus regulations.” Another reported, “Organizations are working hard to ensure help is there if and when a victim needs it… Domestic violence shelters and programs are creatively and safely doing their best to make sure they are there for survivors” (Article 3). The emphasis in these articles encouraged potential survivors to seek services, with the implicit assumption that this approach would best meet clients’ needs.

Inadequate Government Responses Have Contributed to Survivors’ Risk

About half of the articles critiqued government responses across the world for not fully taking into account the needs of survivors of abuse, despite calls from the United Nations encouraging “all governments to put women's safety first as they respond to the pandemic” (Article 11):

But governments largely failed to prepare for how the new public health measures would create opportunities for abusers to terrorize their victims. Now, many are scrambling to offer services to those at risk. But, as with the response to the virus itself, the delays mean that irreparable harm may already have occurred. (Article 11)

While some of the articles clarified that “the government has long ignored” the problem of domestic violence (Article 13), the overarching implication was that governments should have anticipated the needs of survivors of abuse but did not. Article 18 stated, “there was every reason to believe that the restrictions imposed to keep the virus from spreading would have such an effect”), and Article 11 reported, “Now, with families in lockdown worldwide, hotlines are lighting up with abuse reports, leaving governments trying to address a crisis that experts say they should have seen coming.”

Some articles provided piecemeal examples of proactive government responses. For example, Article 7 stated that “France is paying for hotels and other accommodations as shelters exceed capacity," and Australia “has made it easier for victims to receive immediate police protection orders against their abusers.” In the United States, one county was praised for distributing informational cards “that list a phone number to call for help” in retail locations (Article 1). These actions were framed as isolated examples of positive steps that government entities were taking rather than coordinated and institutionalized efforts to support survivors of abuse.

Safety Is Elusive: The Virus Poses Dangers Both Within the Home and in Community Settings

Throughout many of these newspaper articles, survivors of abuse were positioned as being stuck between facing abuse in the home and risking being exposed to COVID-19 in community settings—thus constructing the notion that safety was elusive. Some of the articles described survivors as choosing not to seek services for fear of the virus: “Many victims of domestic violence decide not to leave…[they] refuse to seek medical treatment for fear of contracting the virus at a hospital or clinic” (Article 16). As described above, other articles framed social service agencies as too overburdened to protect their clients from COVID-19 exposure. For example, Article 11 stated, “lockdown began in early March. Soon after that, domestic violence reports began to rise, but there was nowhere for newly desperate women to go. Shelters could not take them because the risk of infection was too great.” Thus, these discourses present limited viable courses of action for survivors to achieve safety.

Discussion

Despite the national discourse of “healthy at home,” this sample of DV-specific newspaper articles constructs an important counternarrative portraying home as a place of violence, confinement, and heightened abuse during the first wave of a global pandemic. While this depiction is critical for raising awareness regarding the seriousness of DV (and a noteworthy departure from previous representations of DV in mainstream newspapers), it was accompanied by descriptions of severely compromised social services, profound government inaction, a frayed community response system strained by years of neoliberal retrenchment, a lack of accessible emergency housing, and community spaces that were vectors of COVID transmission. Together, these discursive themes functioned to construct a double bind, making safety elusive—and ultimately unobtainable—for survivors of domestic violence. In this worldview, survivors are presented with a limited set of options: either enduring abuse and isolation in the home or facing diminished service capacity and possible COVID-19 exposure in community spaces. It is important to note it is not that these representations were not necessarily inaccurate but rather incomplete as they offered a limited set of possible actions.

While this body of articles stands in contrast to more deficit-based and overtly victim-blaming depictions of survivors found in previous media analyses of DV (Bullock, 2007; Bullock & Cubert, 2002; Thaller & Messing, 2014), there is still a marked absence of discussion of larger systems of gender inequality as root causes of abuse, even though the disproportionate effects of the pandemic fall on women and are likely to have an adverse impact on the options available to survivors into the future. For example, throughout the pandemic, The New York Times published several articles on how the public health orders associated with the pandemic disproportionately impacted women, especially primary caregivers (see Grose, 2021), but included no mention of societal-level drivers of DV (e.g., hypermasculinity, gender inequality, patriarchy, structural racism). This omission is noteworthy because these structural factors provide a critical context for understanding and preventing the root causes of DV and for dismantling systems that replicate intersecting systems of inequality (authors, masked for review).

Despite the limitations of shelters, such as the enforcement of rules that can impede job searching and social support (Glenn & Goodman, 2015), the societal response to DV most commonly discussed in these articles was professionalized social services, including criminal justice interventions such as protection orders. While social services are critically important for addressing survivors’ immediate needs, there was a limited analysis of upstream solutions to supporting survivors facing childcare shortages or job insecurity.

The privileging of formal sources of help creates three critical problems. First, it elevates clinical services as the desired response to DV rather than correcting the structural inequalities that are the upstream determinants of DV. Second, it overlooks the reality that DV agencies have historically underserved underrepresented communities (who, thus, do not see these agencies as a viable source of help). Third, it supports the idea that the best pathway to safety is to leave home and seek external support but does not acknowledge the significant financial and emotional costs of doing so (Thomas et al., 2015). Thus, analysis of these articles shows that we still live in a society that regards survivors’ leaving their homes as the ideal solution to DV rather than putting in place the necessary supports to make homes a safer place for survivors and their children. In other words, significant emphasis has been on providing the supports, so survivors are "capable" of leaving home (through a coordinated community response to DV) rather than on making it safer for survivors and their children to stay in their homes and keep their existing support networks intact.

While achieving survivor safety has long been a primary goal of the DV movement (Sullivan, 2011; Thomas et al., 2015), there is a tacit assumption that social services agencies are a “safe” place for abuse survivors. While social services entities were depicted as spread thin, they were not seen as potentially harmful, even though institutional violence is a legitimate risk for communities of color. As Mehrotra et al. (2016) note, "the violence experienced within intimate relationships may be more predictable and manageable than the risks of institutional violence" (p.158). Furthermore, it is important to recognize that the majority of survivors do not seek formal services for their abuse (Voth Schrag et al., 2020), to honor the resilient ways that survivors resist their abuse and keep themselves safe (Thomas et al., 2015) and to acknowledge the myriad ways that survivors’ safety can be compromised when leaving abusive relationships (Fleury et al., 2000).

Critical Discourse Analysis as a Tool for Social Change Within Social Work and the DV Movement

Though underutilized in social work and health sciences, CDA is a critical analytical tool for shedding light on unexamined belief systems, social practices, and taken-for-granted ways of being (Gee, 2011; Willey Sthapit et al., 2020). Within the context of DV, the COVID-19 pandemic continues to be a watershed moment where our field can reflect on historical practices that may heighten inequity for abuse survivors. While this collection of articles paints a dire landscape for survivors of DV, it also presents a critical opportunity to reflect on what a more supportive and survivor-centered model of coordinated community response to DV could and should look like. The pandemic only amplified existing inequities. CDA can help us reimagine an ideal societal response to DV and work to implement it rather than quickly returning to “normal.”

In the 1980s, the mainstream DV movement shifted from focusing on altering the macrostructures that were contributing to and exacerbating inequities associated with DV to focusing on more “professionalized” and “fundable” models of social service delivery and criminal justice system interventions (Durfee, 2020). These models create a system that sees trained workers as “experts” and survivors as traumatized “clients” in need of clinical interventions while also considering other members of the population to be “off the hook” unless they have a family member or friend in need of support when that person—inevitably—leaves their partner. Reimagining models to prevent and intervene in DV could include promoting programs that challenge the intersecting upstream drivers of DV and repositioning DV as a societal problem that we all are responsible for dismantling. Examples of such programs include gender-transformative community-based bystander training, antiracist and gender-justice social policies, emancipatory models of community development, and/or inviting support networks to participate in safety planning. Rather than solely privileging service-based responses to DV, a structural approach to DV would prioritize support (such as financial remuneration, childcare, and educational resources) to survivors and families, alleviating the financial burden many survivors face in moving forward independent of their current or former partners. For example, the Washington State Coalition Against Domestic Violence has proposed survivor-driven models that offer flexible funding systems, immediate and affordable housing, and accessible childcare as the first steps in supporting survivor resiliency (see wscadv.org). Fifty years into the DV movement, there is much more the social work field can do to support survivor well-being and emancipation.

Limitations

There are several limitations to this study. First, rather than a reproducible process, CDA is an interpretive tool where meaning is largely ascribed by analysts. Thus, the interpretations found in this article are, in part, a reflection of the authors’ existing positionalities, ontologies, and worldviews. Another important critique of CDA is that it can exaggerate the influence of dominant discourses on courses of action (Finn & Jacobson, 2003). Despite these critiques, a standard of rigor in CDA is not objectivity but plausibility (Alvesson & Skoldberg, 2009). Thus, to enable readers’ evaluation of the plausibility of our findings, we outlined our stance, positionalities, and process, and we identified counterexamples of the discourses that we identified (authors, Masked for review).

Conclusion

This critical discourse analysis of newspaper reporting on DV within the context of the coronavirus pandemic challenges the broad notion that home is a place of safety. These findings provide important insight into prevailing social scripts of the “idealized” community and societal responses to DV. Within the context of the COVID-19 pandemic, safety was framed as elusive, with survivors deciding between heightened abuse at home or possible exposure to COVID-19 in community spaces. Although it is noteworthy that these articles did not hold survivors responsible for the abuse perpetrated by their partners, as other media representations have, the macrostructures driving DV were still underrepresented. For the DV movement to evolve, it is important for the field to grapple with the role that mainstream social service agencies have played in causing harm against survivors from historically excluded groups. In this unique historical moment of reflection regarding prevailing modes of action, CDA provides a critical tool for reimagining how homes could be restored as places of refuge and peace for abuse survivors and their families.

Author Biographies

Heather L. Storer is an Associate Professor at the Kent School of Social Work and Family Science at the University of Louisville. Her research investigates prevention-based approaches to ameliorating dating and domestic violence, particularly among disproportionality-impacted communities. Dr. Storer's work stands at the intersection of gender-based violence, technology, and social justice. A significant focus of her work focuses on addressing the upstream determinants of dating violence and challenging social norms that stigmatize survivors. Her most recent scholarship explores the processes by which digital infrastructure influences survivors’ lived experiences and reinforces dominant discourses regarding the causes and consequences of abuse.

Brandon D. Mitchell is a university fellow and doctoral candidate at the Kent School of Social Work and Family Science at the University of Louisville. His scholarship is aligned at the intersection of education and social work. He focuses on the role of the school social worker, uprooting deficit-based ideology, and the impact of the media on educational structures.

Claire Willey-Sthapit is an Assistant Professor at the University of Kansas School of Social Work. Her scholarship centers on translations between domestic violence research, policy, and practice, both within the United States and across national borders. Claire is also currently involved in research related to men's gender-equitable attitudes and behaviors.