Emergent Organizations and Women's Response Roles in the 1998 Central Florida Tornado Disaster

by Jennifer Wilson and Arthur Oyola-Yemaiel

Copyright 2000 by Jennifer Wilson. All rights reserved. Permission to reproduce material from the JOURNAL is granted for academic research, library or other archives, or classroom instruction provided the source of material is acknowledged by appropriate citation.


(1) What are the conditions under which women play vital roles in extra-organizational disaster response activities in the community? To explore the complex issues of women and emergent organization in disaster, we intended to investigate women’s roles in emergent, ad hoc groups which formed in a disaster affected community to meet otherwise unmet recovery needs of the population. We attempted to examine when such groups emerged, how many women were involved, the structural characteristics of the organization, and the individual characteristics of women in these roles in the context of a localized disaster in Florida.


(2) Disasters do not affect everyone equally—such phenomena are discriminatory (Blaikie et al. 1994, Neal and Phillips 1990, White and Haas 1975). Vulnerability is unequal among social groups, with disaster victims more likely to be groups with the least power and resources in the social system to escape from hazards. For example, the elderly are more apt to lack the physical and economic resources necessary for effective response, are more likely to suffer health related consequences, and may be slower to recover (Morrow 1998, Tobin and Ollenburger 1992). Poorer households more often live in substandard and inadequately maintained housing, increasing their vulnerability (Phillips 1993, Peacock et al. 1997, Bolin 1982, Bates 1982). Although the economic losses of the poor will be less in absolute terms, even minor losses can be devastating, viewed in the context of limited resources and assets. Ethnic differences affect ways in which people processes warnings and respond to disasters (Perry and Mushkatel 1986). Language barriers often limit minority group access to warnings and disaster information, such as that on available aid (Bolin and Bolton 1986, Aguirre 1988). Around the world women are the population most at risk (Enarson and Morrow 1998a, Blaiki et al. 1994). Women typically have fewer resources, less autonomy, and greater caregiving responsibilities, serving to increase vulnerability and victimize them disproportionately at all stages of disaster (Morrow 1998, Enarson and Morrow 1998a, Blaikie et al. 1994).

(3) Thus, an individual’s place within the social structure contributes to determining the likelihood of their becoming a disaster victim. Among victims, those individuals belonging to one or more less powerful groups will have more difficulty in the recovery process. In addition, research suggests that less-powerful groups are less likely to be part of existing disaster preparedness, response, and recovery efforts (Enarson and Morrow 1998a, Peacock et al, 1997, Neal and Phillips 1995, Phillips 1990, Bolin and Bolton 1986).

(4) However, there is evidence that groups that do not have their needs met through pre-existing social or organizational means will organize among themselves to satisfy these requirements. Emergent or ad-hoc organizations arise (see Dynes’ 1970 typology of organizational behavior in disaster) outside the structure of the official disaster relief network to link with insiders to acquire a fair share of the means for recovery. Although termed emergent, these groups often draw on existing networks for labor and resources.

(5) Women often become actively involved in their communities and neighborhoods during disasters. Neal and Phillips (1990) illustrated that women were key participants or leaders in emergent citizen groups in impact communities. Emergent citizen groups fit this traditional, local pattern of women’s activism based on perception of the cause as an extension of traditional gender roles. Enarson and Morrow (1998b) found that women’s formal and informal networks were central to both household and community recovery after Hurricane Andrew in 1992. Women’s experience as community workers, informal neighborhood leaders, and social activists propelled them to take the initiative in organizing a disaster recovery coalition. Other women have been highly involved as community workers and organizers in disaster-prone areas (Eade and Williams 1995), including neighborhood-based household preparedness programs (Faupel and Styles 1993).

(6) Conversely, Gillespie (1992) suggested network structures in already existing organizations shift functions and modify goals during disasters to adapt to the new environment. This adaptation is better suited and more cost-effective than the formation of new organizations. Bates and Harvey (1978) and Peacock (1991) labeled groups that form between existing organizations and coordinate the pooling of resources of many sources, interstitial groups. These groups link the various social units participating in a goal-oriented exchange relationship. An exchange interstitial group may, or may not, continue to exist contingent on the attainment of goals, expectations, and needs of the participants (Peacock 1991) as well as the scope of work, jurisdiction, and/or financial limitations of the parent organizations. Based on this concept, pre-existing women’s groups or networks may be a source for leadership in such interstitial groups.

(7) Today women are increasingly incorporated into the official disaster relief network. Women are more often found in official emergency management positions at the federal, state, and local levels (Wilson 1998, Enarson and Morrow 1998a, Drabek 1986). Women also enter other emergency services, such as police and fire departments, in greater numbers (Chetkovich 1997, Martin 1980). These conditions, along with traditional participation in human services agencies such as the American Red Cross, make their presence more common in the emergency operations center (EOC) as representatives of important functional areas. Women’s greater participation in the interstitial group of the EOC may provide less reason for outside ad hoc groups to form. Thus, women’s needs may primarily be met through existing organizations.


(8) Storms that swept across Central Florida in the early morning hours of 23 February 1998 spawned the deadliest round of tornadoes on record in Florida. Historically, ninety percent of Florida’s tornadoes have winds under 72 miles per hour (Tornado Project 1999). However, due to the effects of El Nino atmospheric disturbances, the several tornadoes that struck Florida on 23 February contained wind speeds from 210 to 260 miles per hour. James Lushine of the National Weather Service stated that in only two other cases, in 1958 and 1966, have Florida tornadoes generated wind speeds of more than 206 miles per hour. Both were El Nino years, and both times the storms impacted Central Florida (SUN-SENTINEL 1998).

(9) Tornadoes in Florida most often occur as by-products of hurricanes. As a result, Florida does not have alerting sirens, such as those found in Texas, Oklahoma, and Kansas, which sound to warn residents of approaching tornadoes. Despite Tornado Watches and Warnings throughout the day (see, for example, Tornado Watch 58 issued at 8:13 pm on 22 February by the Storm Prediction Center), many Central Floridians went to sleep on Sunday night with little apparent concern for severe weather.

(10) Tornadoes touched down in Brevard, Dixie, Manatee, Nassau, Orange, Osceola, Seminole, Sumter, and Volusia Counties. Forty-two people were killed and more than 250 injured throughout Central Florida (ST. PETERSBURG TIMES 1998) in the state’s worst documented tornado outbreak. Osceola County experienced the most severe impact with 25 people killed and 148 injured. The Osceola County Office of Emergency Management estimated that the county sustained more than $37 million in major damage to 150 homes, 200 mobile homes, 15 recreational vehicles, a strip mall, and approximately 30 businesses. An additional 225 homes, 60 apartments, and 25 mobile homes suffered less severe damage (THE OSCEOLA SENTINEL 1998).

(11) Osceola county is located immediately to the south of Orange County. The county’s population is approximately 130,000 (Pierce 1995). The county contains two large lakes, West Lake Tohopekaliga and East Lake Tohopekaliga, surrounding the county’s two largest cities, Kissimmee and St. Could. These two large lakes and many smaller lakes to the south form the Upper Kissimmee Waterway Basin. With the lakes, conservation areas, and farmland in the southern part of the county, most of the population is concentrated in Kissimmee and St. Cloud. In addition, much of the Walt Disney World theme park, located directly west of Kissimmee, falls within the county boundary.

(12) The storm crossed Osceloa County, tracking from southwest to northeast. It first caused minor damage to homes in a restricted, elders-only subdivision near the Poinciana Office and Industrial Park, and then damaged homes in the Campbell area southwest of Kissimmee. As it continued northeast, the tornado collapsed one wing of The Shops at Kissimmee strip shopping mall, and then passed over West Lake Tohopekaliga severely damaging a neighborhood of lakeside homes on the northeast shore.

(13) After crossing Highway 441 and leaving the Osceola County Stadium and Sports Complex unscathed, the tornado touched down at approximately 2:00 am in the Ponderosa Pines mobile home park near Boggy Creek Road. This resulted in the highest concentration of fatalities; rescue workers subsequently recovered 10 bodies in the Park. Nearly all of the community’s 200 mobile homes and recreational vehicles were destroyed. The tornado then continued northeast across the Florida Turnpike and the Lakeside Estates subdivision of single-family homes in the Buena Ventura Lakes area, damaging approximately 400 homes and the Cypress Creek Elementary School.


(14) The research for this project was qualitative in design, including document analysis, ethnographic interviewing, and participant observation. We visited Osceola County on four occasions for two days each to study the community’s coordinated response to the tornado disaster. Data was primarily collected through semi-structured, open-ended interviews of emergency management organization staff, government and non-profit disaster relief organization personnel, and citizens/victims. In total, we conducted 10 interviews.

(15) We employed snowball sampling techniques. During each interview, we asked the respondent if he or she recommended someone else in the community as important to interview regarding coordination of the tornado response. To avoid bias, we did not reveal our primary interest in women’s roles in coordination of Osceola County’s response. Our respondents suggested more female participants in human services and more male participants in response activities, suggesting a traditional division of labor pattern within disasters that has been documented elsewhere (Wilson 1998, Enarson 1997, Enarson and Morrow 1998a, Phillips 1990).

(16) To supplement ethnographic data (Erlandson et al. 1993, Lincoln and Guba 1985), we gathered documents, including organizational reports, media accounts, weather reports, and other potentially useful materials (Webb et al. 1981, Plummer 1983).


(17) We found no evidence of formation of emergent organizations in Osceola County, Florida, following the February tornado disaster. The response was managed using pre-established organizational channels. Conditions for the formation of emergent organization do not appear to have been present. Thus, the opportunity for women to participate in these groups was nonexistent.

Why Were There No Emergent Groups?

(18) The Osceola County emergency operations center Operations Manager stated: “Initially it was overwhelming. The sheer volume of … needs was tremendous …. This county has never experienced anything like this …. Were we prepared? No, we weren’t.” However, all indications from other respondents were that the official response was immediate and thorough. Victim respondents with whom we talked were highly satisfied. Furthermore, the Salvation Army and the American Red Cross respondents believed that the response went extremely well despite the perceived lack of preparation.

(19) Although this county has rarely experienced disasters of the same intensity as this tornado, very innovative techniques were used during the response. The Osceola County Office of Emergency Management (OEM) established a storeroom for citizens whose homes were destroyed to use to store their possessions until they could find replacement housing. The OEM also established two warehouses, one each for incoming and outgoing donations. Osceola OEM was aware the County would receive large amounts of unusable donations that could then be forwarded to other agencies that would be glad to have them. The OEM Operations Manager said “at one point we had nearly twenty semi's coming in and twenty semi's going out each day” with donations received and then subsequently forwarded.

(20) Another effective response by Osceola OEM was the coordination of volunteers. The OEM arranged to have photographic identification badges made for each volunteer. This was accomplished through the development of a database which tracked volunteers’ names, skills and equipment they were able to provide, and assignments. The database also tracked volunteer needs within the community. The OEM covered volunteers with accident insurance and workman’s compensation insurance during their volunteer work. According to the Osceola County OEM Operations Manager, approximately 3,000 volunteers did 19,000 hours of work in the county in response to the tornado disaster.

(21) In all, respondents had very few complaints concerning the response. Rather, the respondents to whom we spoke praised the coordinated efforts among the community’s organizations. Indeed, according to our respondents, there was a high degree of coordination among existing agencies and organizations. For example, town meetings for city-county coordination were instituted the day after the event to facilitate communications among all the players. Respondents reported only minor communications problems that were resolved quickly. This may be due to the emergency operations center being expressly used for coordination among the players.

(22) Coordination was further facilitated by the involvement of most Osceola County departments and offices in the response in some way. These departments included the human resources department, parks and recreation department, road and bridge department, collections office, solid waste department, billing office, and others. These offices provided labor, equipment, and communications to the response effort.

(23) Another reason for the quick and thorough response was the involvement by key response agencies of members of their regional or state counterparts in the county to facilitate the response. The Florida Division of Emergency Management had a representative on-site almost immediately after the tornado's impact to work with Osceola County OEM in instituting the state response plan. This same response occurred with the American Red Cross and Salvation Army, which both had members of their regional disaster response teams arrive within 12 to 36 hours of impact. These teams’ expertise in disaster response was evident in the coordinated and swift response for sheltering and donations (American Red Cross) and feeding (Salvation Army).

(24) There was substantial evidence that the responders were concerned with making recovery from the disaster as easy as possible for individual victims in the community. For example, the local bus system, LYNX, established separate routes to transport tornado victims to the Disaster Relief Center (DRC). The DRC contained representatives from the Red Cross, Salvation Army, Federal Emergency Management Agency, Small Business Administration, Small Business Administration, some state agencies such as the Insurance Department, and united Methodist Disaster Relief, all in the same building. Such integrated and localized coordination of relief is significant in improving community recovery because it allows one-stop shopping. Cooperation and communication between agencies’ staffs in the same location resolves victims’ problems and requirements in a timely and relatively convenient fashion. In this case, it was possible to coordinate in such a way because of the localized nature of the damage; in other disaster events the magnitude of the impact may require relief efforts over an extended area. In other cases, lack of space or an unwillingness to work closely together may hinder this type of integrated relief coordination (Watson et al. 1998).

(25) Coordination took on a new meaning when most of the local social services agencies decided to establish a disaster relief fund for the donations they received. Every agency was then able to draw from this fund to practice their individual disaster relief work. This procedure balanced the disaster relief donations received by individual agencies, as some received large amounts of funds, and some received little or no donations. This was a significant component of the coordinative effort as non-profit agencies often have difficulty obtaining adequate operating funds.

Where Were The Women?

(26) Following Gillespie’s (1992) thesis, a few women were key participants within their existing organizational positions. For example, the director of the county personnel office was key in organizing the county’s volunteer program. A horticultural agent of the Department of Agriculture designed the computer database program used to track volunteers. Kissimmee’s assistant city manager was essential in facilitating the working relationship between city workers and the Osceola County Office of Emergency Management’s response plans. The director of the local chapter of the American Red Cross played an integral part in shelter provision. One of the co-directors of the local organization of the Salvation Army was crucial in providing feeding to both victims and rescuers. And the sheriff’s office emergency operations center representative served to link her department’s response with the OEM.

(27) In total, we identified six women as being prominent in the community tornado response in Osceola County. Although the total number of important female responders is small, only half of these women (n=3) occupied traditional female working roles in their official positions. Three worked in social services, but the remaining three occupied less-traditional working roles: a law enforcement officer, a horticulture agent, and an assistant city manager (public official). Even though these women regularly occupied less-traditional working roles, two of them fulfilled more traditional female roles during the disaster response. One woman was responsible for coordinating volunteer workers, and the second handled telephone communications at the emergency operations center. The third woman took on a much more substantial non-traditional working role during the disaster response as the mediator between city workers and the county OEM response process. Thus, we would argue that women were vital in the response process within the pre-existing organizational structure.


(28) In this relatively small community, the coordinative effort of local agencies was supported and assisted by the convergence of outside experts. The tornado, although severe and devastating for some, was localized and did not have a catastrophic effect in which the entire social structure and institutional fabric ceased to operate (Bates 1982). In contrast, a disaster of the scope of Hurricane Andrew in 1992 caused devastation so widespread that the social-organizational structure of everyday life was virtually dismantled (Peacock et al. 1997). In Osceola County the tornado left clusters of localized heavy damage, but the majority of the community was left intact and able to concentrate relief efforts on the damaged areas. This resilient community was able to respond effectively, solving the problems at hand that otherwise could have created unmet needs for sectors of the population. In doing so, unmet needs and the resulting emergent organizations never appeared.

(29) Among the reasons that emergent organizations did not form in Osceola County is that already existing organizations adapted to meet the basic needs of the affected community. The exchange relationship was conducted both within existing organizations and between these organizations, so that outside or ad hoc groups were not needed. This may have been accomplished in part because of a high degree of flexibility, ingenuity, cooperation, and communication by community organizations, evident in such innovative practices as free bus transit to the Disaster Recovery Center and insurance coverage for volunteer workers. Moreover, women adapted their organizational roles to become integral parts of the coordinative interstitial group (the emergency operations center) that handled the disaster response and recovery. As a result of this, we foresee that more diverse and adaptive emergency response organizations represented in the emergency operations center are more likely to be sensitive and respond to the needs and concerns of all members of an affected community.


(30) This research was supported by the Natural Hazards Research and Applications Information Center at the University of Colorado-Boulder through the Quick Response Grant Program which is supported by the National Science Foundation under Grant No. CMS-9632458. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the Natural Hazards Research and Applications Information Center or the National Science Foundation.

Article Links

Sources and Review Comments
supporting references and peer reviewer comments not included in the text

Journal Links

Electronic Journal of Emergency Management Cover Page
Journal policies and procedures and index to articles

Comment and Review

Articles published in THE ELECTRONIC JOURNAL OF EMERGENCY MANAGEMENT are intended to stimulate thought, comment, and additional research. Readers are encouraged to submit comments and questions to the Editor at wgreen@richmond.edu for possible inclusion in ongoing on-line peer review.