Responding to Domestic Violence in Disaster:
Guidelines for Women’s Services and Disaster Practitioners
Elaine Enarson, Visiting Scholar
Disaster Preparedness Resources Centre
University of British Columbia
November, 1997
Beatrice Lavelle and her daughters came to the battered women’s shelter in south Dade County after Hurricane Andrew, when conflicts with her brother-in-law in their crowded trailer escalated out of control and other housing was unavailable.[1] For Diana Sanchez, the violence started after six long months in a tent on the outskirts of the city where she and her boyfriend lived after migrating to Miami for construction work. When Hurricane Andrew destroyed their apartment, belongings, car, and workplaces, Lois Richards’ husband “just went berserk: “He really went crazy. Before, I would get beat up maybe once a month if I was lucky. Afterward, it was like every other day…I ran across a lot of women suffering, too, with their children—husbands beating them up and leaving them. It was pretty bad.”
Hurricane Andrew hit many people hard, but especially low-income women, single mothers, and recent immigrants. Long after city workers cleaned up the visible destruction, displaced families still camped out in half-repaired homes and overcrowded trailers. Residents discovered how little disaster grants and vouchers could replace and how fragile the spirit of shared struggle during crisis was. Women like Beatrice, Diana, and Lois were less safe than ever from violence in their own homes when Betty Hearn Morrow and I met them six months later while documenting the social impacts of the hurricane “through women’s eyes.[2] This US and Canadian survey of women surviving violence and disaster was inspired by their experiences.
Violence Against Women in Disaster: What are the Issues?
Disasters are not only powerful physical events but complex social experiences for individuals, households, and communities. Once considered the great “leveler” impacting poor and rich alike, we now know that floods, hurricanes, tornadoes, chemical spills, and other environmental and technological disasters hit some social groups more than others. [3] The poor, subordinated racial or ethnic groups, refugees and migrants, seniors, the disabled, and women are especially vulnerable to disaster losses and often recover from disaster more slowly.
This work was generously supported by grants from the BC Institute Against Family Violence and the Feminist Research, Education, Development & Action Centre. My thanks to Jill Hightower, Andrea Kowaz, Yasmin Jiwani, Greeta Smith, Victoria Constance, Sue Meuschke, Ruth Harding, and Tracy Porteous for their ideas and support, and to Maryvon Delano for her skillful French translation and interviewing. I am greatly indebted to the many women who shared their experiences and ideas, especially Bonnie Palecek and member programs of the North Dakota Council on Abused Women’s Services.
But which women are more vulnerable to disaster and why? Class and racial privilege protect some women more than others, when money in the bank means disposable income for hurricane shutters, extra food supplies, or home repairs. On balance, women in dual-income families are less vulnerable to the economic impacts of flood or tornado than single mothers and able-bodied women more able to prepare for and cope with disaster than women living with disabilities. Women and girls also die disproportionately because the daily routines of domestic work and caregiving keep them inside hazard-prone buildings. In some societies, daughters are too poorly fed to survive food shortages and epidemics and secluded women may not be able to access public shelters or public relief lines.[4]
Women living with the “daily disaster” of domestic violence are also highly vulnerable when disasters transform geographies, institutions, and relationships.[5] In the vicious dynamic of power and control, battered women live in a world of increasingly narrow social networks, often isolated, unable to take or keep paid work, lacking transportation, and financially dependent. Like their physical and emotional health, their sense of self-worth and efficacy diminishes in the face of continued violence. As one shelter worker noted, fragile support systems can make battered women even more vulnerable after disaster:
So many victims of battery have been isolated from the normal networks of support—family, job, things like that…Now here’s this person that’s holding on, just barely holding on—the disaster hits. It’s not just them, but everybody around them, they scatter. The little bit of support that’s been helping that victim hold it together is gone, and in fact, they may be forced into a situationwhich we saw hereof ending up in the home of the family of the abuser and actually having more to deal with less support than they’ve ever had before. I mean, it just mushroomsthe stress level of that victim.
Disasters have no single impact on women or women at risk of violence. Paradoxically, a family home destroyed by fire may loosen the ties binding women to violent partners; disaster relief money can buy a bus ticket out of town for women ready to leave; and responding to catastrophe may reduce abuse temporarily. More than simply “victims,” battered women develop survival skills to protect themselves and their children which may serve them well during disaster.
Notwithstanding these coping mechanisms, the daily realities of living with violence make attending to the threat of disaster unlikely and stabilizing life in a disaster-stricken neighborhood and community very difficult. Already in emotional crisis, battered women in shelter are focused on the relationship they have just left; as one shelter worker observed, they often lack “a strong awareness of what’s going on around them.” Severe weather events like mudslide or blizzard isolate women at home in unsafe environments without working telephones or accessible roads; contact with crisis counselors may be cut off and court-ordered protection unavailable when major disasters disrupt or destroy lifeline services, including law enforcement agencies.
Relationship stress factors increase when families struggle to replace lost possessions, housing, jobs, and peace of mind; a counselor working with men after the Loma Prieta earthquake noted also that “many men used the quake as a way to get themselves back into an old relationship.” [6] Men emotionally invested in the role of family provider and protector will struggle with uncertainty and feelings of inadequacy:
The whole issue of not being able to protect his family, not being able to provide for his family, was devastating for him, just devastating. He does not like to ask for help…He said, ‘You know, I’ve been taking care of myself since I was 12 years old…I don’t know what it’s going to be like when I get back there. I don’t know what kind of work is going to be there.’ He was overdrawn on his checking account by $2,000 because he’d just finished a house and was supposed to get paid and that house got washed away. Went to the Red Cross and they said ‘No, we can’t give you any financial support.’ He got $7. He doesn’t want to take money from us.
For women and children housed in shelter, mandatory evacuation following an industrial accident or in advance of wild fire is a second-order evacuation. Designated evacuation centers may not protect their privacy or be safe space for battered women. The advocate quoted below works in a rural area where abusers can easily locate their partners, but she raises a broader issue:
And the Red Cross shelters, those type of shelters, are not safe for them…[The abusers] are just going to put two and two together and say OK, well where is she going to go? And so I think this really shows that we need to have a plan of action ahead of time…[I]f something happens and our shelter’s unavailable, is there another shelter, across the state even, where we could transfer her? Because they’re there for a reason. It’s specifically because they’re in danger.
When the dust clears or the waters recede, women coping with physical and/or emotional abuse must access bureaucratic disaster relief systems and compete with other impacted residents for housing, child care, employment, education, transportation, and health services. Private and public relief funds are rarely sufficient to meet all costs of repairing homes and replacing possessions; relief funds may be more available to the abuser at home than to women living in shelter. Arguably the most vital lifeline for battered women, affordable housing stocks are likely to decline after disaster, when cheap housing on hazard-prone land is damaged or destroyed by flood water, earthquake, or tornado.
Before, during, and after disaster, the emotional, economic, and housing insecurity of women living with violence exposes them to greater risk and hence imposes new demands on shelter staff and advocates. How well prepared are they to meet the challenge? How do communities plan for battered women’s needs during disaster?
Study Design and Sample
This report summarizes the findings of a study investigating disaster impacts, preparedness and response in US and Canadian domestic violence programs.[7] Mail and/or telephone surveys were completed between April and November, 1997, by 77 domestic violence shelters, transition houses, and state or provincial coalition offices, including among them 35 surveys completed by members of the British Columbia/Yukon Society of Transition Houses. Most respondents reported no or relatively minor disaster events (e.g., minor flooding, localized toxic spill); however, the survey also included 13 programs severely impacted in the l990s by major flooding in the US Midwest, Quebec, and the Red River Valley, a southern California earthquake, and Hurricanes Andrew and Iniki. A case study of the April 1997 Red River Valley flood was developed through participant observation with the North Dakota Council on Abused Women’s Services and telephone interviews in Manitoba. This event flooded Grand Forks, North Dakota and East Grand Forks, Minnesota as well as two rural communities south of Winnipeg, and put the provincial capital on evacuation alert for several weeks.
As a pilot study for future research, the survey provides baseline data and identifies emergent issues. The survey results cannot be generalized to all shelters in either the US or Canada; however, they do raise significant questions about the degree of protection afforded battered women when communities prepare for and respond to disaster.[8] Survey findings are incorporated into action guidelines for shelter managers, coalition staff, and local emergency practitioners (see Appendices A and B).
Getting on the Agenda: Awareness and Preparedness in Battered Women’s Programs
“What we give them is all that they have,” one worker said of her shelter. It will not surprise battered women’s advocates to learn that pets, tourists, and cultural artifacts receive more attention than battered women from disaster planners. Although less visible to emergency managers than facilities housing the frail elderly, disabled, or mentally ill, for whom emergency preparedness guidelines are available,[9] battered women’s shelters, safe houses, and transition homes also need comprehensive disaster planning. As they are not generally recognized as priority facilities housing and serving an especially vulnerable population, their self-reliance through disaster preparedness is critical.
Hazard awareness and risk assessment
As might be expected, programs with direct experience indicated a higher awareness than non-impacted programs of local hazardous conditions or occurrences, citing risks ranging from avalanche, gas explosions or hazardous materials transportation accidents to wild fire, flooding, tornado and earthquake. Few reported regularly receiving any official information on disaster preparation, depending instead upon mass media; rural programs in Canada and the US were more likely to receive direct communication. In the Yukon Territory, for instance, the local Dawson City program reported receiving annual flood response plans from city officials.
Few programs are represented on local, regional, or provincial disaster planning groups. Among programs without experience of regional disasters (n=36), which may influence domestic violence programs without resulting in direct service impacts, four participate in local emergency networks; the great majority (31 of 36) were either not certain or reported that their facility was not specifically included in local disaster plans (e.g. for priority evacuation assistance, communications, or emergency power). Programs with some regional disaster experience (n=41) were also unlikely to participate in local planning efforts (4 of 41). More disaster-experienced programs did report being included in local response plans (13 of 41 programs, or 32%), in some cases, developing “stronger relationships with emergency managers” through the disaster (9 of 41, or 22%).
How safe are the buildings housing battered women? Overall, most responding shelters reported their physical facilities to be “relatively safe.” Many, however, are located in older buildings affordable to women’s services or are centrally located in communities built up in hazardous coastal or flood plain areas. “He just shook his head,” one shelter director in Texas reported of a conversation about safety with their building’s architect. A number of British Columbia programs in a known earthquake zone reported that their facility was “relatively safe” but added “not safe in the event of earthquake.” Assessing the structural integrity of shelters requires worst-case scenario planning, specialized knowledge, and adequate funding for analysis and follow-up renovation.
Shelter preparedness
Disaster planning is not a priority for domestic violence programs working hard at “securing basic needs for women and children, e.g. safety, housing, etc.” As Table l (below) indicates, fewer than half reported taking steps internally to protect staff or shelter residents during and immediately after a disaster. Programs with disaster experience reporting service changes of some kind were considered “disaster impacted” and generally reported higher levels of preparedness, as might be expected. However, six of 25 impacted programs (24%) and 19 of 52 non-impacted programs (37%) reported having taking no action of any kind.
Preparedness measures often appeared oriented toward the routine (e.g. minor flooding) or contained accident (e.g. house fire) rather than the catastrophic (major earthquake) though a shelter might be at risk of both. In British Columbia and the Yukon Territory, a region at risk of flooding, mudslide, transportation accidents, and severe weather events, with major metropolitan areas at risk of earthquake, 80% of the 35 responding programs reported either no preparedness steps (12 of 35) or only minimal steps (16 of 35), for example reporting stored food or evacuation plans “in case of fire but without follow-up plans.” Among these are six programs which each sheltered between 500-600 women and children during the last fiscal year. Of the 35 BC/Yukon programs, seven reported having taken three or more preparedness measures.
Table l Program preparedness*
Non-impacted programs Impacted programs
Action taken (n=52) (n=25)
Structural protection like earthquake
bracing or hurricane shutters? 8 15% 4 16%
Stockpiled emergency food, water
batteries, radios? 13 25% 10 40%
Emergency drills with staff? 17 33% 5 20%
Evacuation plans for shelter resident? 23 44% 11 44%
Disaster training for staff? 6 12% 5 20%
*Includes programs with and without shelters, and coalition offices
Overall, few programs in Canada or the United States were likely to have secured their building and equipment as recommended by local emergency specialists. Programs with residential shelters are, by their nature, likely to store extra food, water, and first aid supplies; like all organizations, coalition offices and non-shelter programs will also want to meet the emergency needs of residents, staff, and volunteers on site during a disaster. Programs are most likely to have evacuation plans and emergency drills in place, but these were often described as simple fire drills. Evacuating battered women from shelters is complex. They may be unable to safely utilize designated evacuation centers and need reliable transportation to alternate evacuation sites. Like other evacuees, they will be advised to take portable radios and back-up batteries, flashlights, and other supplies. If feasible and safe, some clients may also want to return to their own homes to gather critical documents or irreplaceable items when evacuation advisories are issued.
What factors facilitate or inhibit internal preparedness in battered women’s shelters? Shelters with very low levels of preparedness were characterized by lack of past disaster experience, low levels of hazard awareness, lack of information about preparedness, and organizational constraints. Asked about the primary barriers to greater disaster readiness, British