Evacuation Behavior and Problems: Findings and Implications from the Research Literature.

Abstract

Our task was threefold: (1) to examine what is and is not known about evacuation in peacetime disaster situations; (2) to systematically order and organize the literature and other research data; and, (3) to make recommendations from our findings and observations. About 150 literature sources and other newly gathered as well as previously gathered research data were examined. An analytical model of evacuation behavior was developed positing a relationship between community context, threat conditions, social processes, patterns of behavior including the withdrawal movement, and consequences for community preparedness for evacuation. Policy, planning, operational and research implications were derived suggesting future actions and efforts. Our study did find that we do currently have some research-based knowledge and understanding about evacuation phenomena in disasters. The literature and research data give us a comprehension beyond common sense notions, and in fact, at times, the evidence suggests that citizens in general and officials in particular may be working with incorrect assumptions and beliefs about the phenomena. On this topic, as is true of many other matters about disaster behavior, mythologies and misconceptions about evacuation abound. For example, to the extent that there are research observations, they show that the withdrawal movement itself usually proceeds relatively well. The flight tends to be orderly, reasonable from the perspective of the evacuees, and generally effective in removing people from danger. The problems with evacuation occur before and after the flight behavior itself. Organizational preparations for and initiation of mass evacuation efforts tend to be poor. Planning is often unrealistic, assumes that evacuees have to controlled and generally does not address the distinctive and special problem which can be involved in mass evacuations. Little consideration in plans or in actuality is given to the fact that evacuation involves going to some area, as well as from some locality, and almost always returning to the original place o f departure. A number of implications and recommendations follow form our analysis of the literature and research data. Thus, we argue that evacuation should be approached as a proactive policy matter important in itself rather than being treated primarily as reaction to warning activities. In some ways, peacetime evacuation ought to be viewed as distinctive and separate phenomena parallel to the treatment of crisis relocation in the literature on wartime emergencies. Planning might visualize evacuation not as an outcome, but as a flow process with different emergent stage involving various kinds of contingencies. Evacuation does not always develop in a singular and linear path, but may involve multiple and disjunctive paths. Operational activities in connection with evacuation must consider the full range of the patterns of behavior that are involved, from the warning to the withdrawal to the shelter and to the return stage. The heterogeneity of the population at different stage requires different organizational actions. We also found that the research base about evacuation phenomena is not strong. Evacuation has not been a major focus of systematic study, and knowledge of the phenomena is often surface and very uneven. Theoretical treatments of evacuation are even fewer and less informative as a whole that the descriptive and case study literature which provides the bulk of the findings and impressions about the topic. Priority in the future ought to be given to in-depth research on unexplored topics (e.g., the non-movers), little systematically examined areas (e.g., the shelter stage) and selected operationally important subjects (e.g.., the evacuation of institutionalized populations). At a more theoretical level, study needs to be done on understanding the meshing of individual and organizational behavior in mass emergencies.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 169 citations exist upstream.