This article discusses the factors that impact upon evacuation behaviour. Social issues are crucial to this topic as individuals will want to coordinate with others, in particular their families before acting. Some individuals may also resist evacuation orders, whereas others may evacuate from unaffected areas (termed an ‘evacuation shadow’). The use of reception centres is discussed, as are the implications for the communication of clear and timely instructions.
In the event of a CBRN incident it may be necessary to evacuate the entire population of an affected area. In such an event, many of the issues discussed in our other articles are relevant. For example, panic is unlikely (although fear of panic may have its own consequences) and social networks are crucial. Families will not want to leave as individuals, but will attempt to contact other family members and arrange to leave together (Perry, 1994).
This will have implications for an already overloaded communications system. Some individuals may not know of the threat or warning, particularly if it occurs at night. Moreover, evidence from natural disasters suggests that some are likely to refuse to evacuate because they fear their property will be unsafe and will be looted, or they have infirm or ill relatives. In the case of a CBRN incident people may refuse to evacuate due to fear of greater exposure to the agent if they do so (Fischer, 1998).
The population needs to be convinced that their property is safe, or at least that their lives are more important than their property, that evacuating is the safest option for them and that provision will be made for ill family members.
However, as well as concerns about individuals complying with instructions to evacuate, previous incidents have also highlighted the need to be aware that individuals may evacuate from areas outside of the evacuation zone (Tierney, Lindell, & Perry, 2001a). This evacuation by people living outside of the risk zone is termed an ‘evacuation shadow’ and is likely to occur if individuals perceive themselves to be at risk.
For example, following Three Mile Island, individuals defined by the authorities as not at risk, nevertheless defined themselves as at risk due to a number of factors including:
- confusing and conflicting information from the authorities;
- geographical proximity to the plant;
- similarity to the demographic groups targeted in the warning messages.
Therefore, when an evacuation is announced for a specific area, it should be expected that residents outside this area, but nearby, will also evacuate, and that group-specific evacuations (e.g. pregnant women and children in the case of Three Mile Island) will generate additional evacuations, in particular by other family members (Perry, 1983a). Some individuals near Three Mile Island also reported evacuating in order to avoid later forced evacuation (Barkun, 2002).
These evacuation shadows will serve to increase the scale of evacuation, causing increased traffic congestion and confusion about which is a safe area. In the event of an airborne CBRN agent, wind direction is likely to play a crucial part in determining evacuation areas and given the low confidence many people hold in the accuracy of weather forecasts, evacuation shadows may be significantly increased.
Once individuals or groups have taken the decision to evacuate it is unlikely that they will leave immediately; instead evacuation will straggle out over a period of hours (Fischer, 1998). If the authorities have given specific route instructions some will follow these, but many will not. It has been observed that in emergency situations, individuals tend to move in the direction of the familiar, excluding new information that may lead to less risky behaviours (Sime, 1992; Robert J. Ursano & Fullerton, 1990).
If possible, evacuees will go to stay with friends and family who live outside the evacuation zone, rather than to public congregate care facilities or reception centres set up by the authorities. Reports suggest that evacuees tend to underutilise public congregate care facilities, which are usually seen as the last resort for accommodation (R. W. Perry, M. Lindell, & M. R. Greene, 1981).
A report after the Flixborough disaster stated that evacuees would arrive at reception centres, but then depart for friends’ houses without their presence or departure having been recorded (Explosion Nypro (UK) Ltd. Chemical Factory Flixborough, 1974). It has been suggested that use of these facilities will rarely be by more than one quarter of evacuees, although this will vary as a function of population, event and community characteristics.
One review suggests that higher levels of public congregate care use are associated with agent characteristics such as rapid onset, large scope of impact, high levels of destruction and also with evidence of chemical and nuclear agents (Drabek, 1986). Community characteristics influencing take-up include greater community preparedness and lack of individual access to private transport (Tierney, Lindell, & Perry, 2001a).
In summary, although previous research suggests that some individuals may resist evacuation, it is also possible that, in a CBRN incident, evacuation shadowing may in fact increase the scale of evacuation response. However, authorities can attenuate these problems by making instructions clear, timely and specific and by providing resources through which families and friends can communicate and act together.