Definition and introduction

The empirical study of public communications in emergencies has been ongoing for almost 50 years (Perry and Mushkatel, 1986, 1984; Leik et al., 1981; Quarantelli, 1980; Baker, 1979; Mileti and Beck, 1975; Drabek and Stephenson, 1971; Lachman et al., 1961). These studies, when viewed collectively, have compiled an impressive record about how and why public behavior occurs in the presence of impending disaster or threat. For example, it is well documented that emergency warnings are most effective at eliciting public protective actions like evacuation when those warnings are frequently repeated (Mileti and Beck, 1975), are confirmatory in character (Drabek and Stephenson, 1971), make specific recommendations, and are perceived by the public as credible (Perry et al., 1981). Informal warning mechanisms (friends or relatives) are also at times very effective. In many evacuations, people leave the area at risk before an official warning is announced. Evacuation behavior is also influenced by other factors such as personal or family resources, age, and social relationships including social networks, level of education completed, experience with previous emergencies, social and environmental cues of immediate hazard, physical or psychological constraints to evacuating, as well as other more specific circumstances (such as time of day, weather conditions, etc.). Table  1 provides a list of those factors and how they have covaried with decisions to respond (Mailman School of PH @ Columbia, annual preparedness survey, focuses on why parents may not heed evacuation orders).

Communicating Emergency Information*, Table 1 Factors associated with warning response
Full size table

Studies that have used surveys of random samples of people living in or near-disaster areas have been conducted for a variety of hazard events. For hurricanes, these include Elena and Kate (Baker, 1987; Nelson et al., 1988), Eloise (Windham et al., 1977; Baker, 1979), Camille (Wilkinson and Ross, 1970), David and Frederick (Leik et al., 1981), Carla (Moore et al., 1964), Floyd (Dow and Cutter, 2002; HMG, no date), Andrew (Gladwin and Peacock, 1997), Bertha and Fran (Dow and Cutter, 1998), Georges (Dash and Morrow, 2001; Howell et al., 1998), Brett (Prater et al., 2000), Bonnie (Whitehead et al., 2000) Ivan (Howell and Bonner, 2005), and Lily (Lindell et al., 2005).

Studies of flood include Denver, CO (Drabek and Stephenson, 1971); Rapid City, SD; (Mileti and Beck, 1975); Big Thompson, CO (Gruntfest, 1977); Sumner Valley, Fillmore, and Snoqualmie, WA (Perry et al., 1981); Abilene, TX (Perry and Mushkatel, 1984); Clarksburg and Rochester, NY (Leik et al., 1981); and Denver, CO, and Austin, TX, (Hayden et al., 2007).

Studies of chemical accidents include Mississauga, Ontario, Canada (Burton, 1981); Mt. Vernon, WA; and Denver, CO (Perry and Mushkatel, 1986); Confluence and Pittsburg, PA (Rogers and Sorensen, 1989); Nanticote, PA (Duclos et al., 1989); and West Helena, AR (Vogt and Sorensen, 1999). Graniteville, SC (Mitchell et al., 2005).

Other protective action studies include the Hilo, HI, tsunami (Lachman et al., 1961); the Mt. St. Helens, WA, volcanic eruption (Perry and Greene, 1983; Dillman et al., 1984); the Three Mile Island nuclear accident, PA (Cutter and Barnes, 1985; Flynn, 1979); the World Trade Center bombing, NY, in 1993 (Aguire et al., 1998); the World Trade Center collapse, NY, in 2001 (Averill et al., 2005); SoCal wildfires in 2003 and Australian bushfires in 2005 (Proudley, 2008, AJEM) and in particular 2009 (Haynes et al., 2008, J. Volc. Geotherm. Res., on volcanic risk perception; Wray et al., 2008, Am. J. Pub. Health, on communicating with public about health threats). The National Env. Health Assn published an excellent review of risk comm., risk perception, and loss of trust in “authorities” re post-collapse risk (Lyman, 2003, Messages in the Dust).

Excellent summaries of this research currently exist (Lindell and Perry, 2004; Drabek, 1986; Mileti and Sorensen, 1990; Tierney et al., 2003; National Research Council, 2006) and will not be repeated here.

Summary

Empirical studies and summaries have done much to further social scientific understanding of how people process and respond to risk communications in emergencies; it has also served to inform practical emergency preparedness efforts in this nation and abroad. Relevant research on human response to risk communications derived from the empirical research record can be summarized as follows.

Research indicates that people’s decisions to respond to emergency communications are influenced by:

  • The frequency and channel of communication of the warning. The most important dimensions of the warning frequency/channel are the number of different channels people hear the warning from, hearing from personal channels, and the frequency that people hear the warning.

  • The content of the warning message. The most important dimensions of content are a description of the hazard and impacts, the predicted location of impacts, what actions to take, and when to take those actions.

  • Observing cues. These include social cues (i.e., seeing neighbors evacuating) and physical cues (i.e., seeing flames or a smoke cloud).

  • Aspects of individual status. These include socioeconomic status (i.e., income level and education completed), age, gender, and ethnicity.

  • The role(s) an individual holds in society. These include having children at home, family size (i.e., larger vs. smaller), extent of kin relations, being a united family at time of the event, and greater community involvement.

  • Previous experience with the hazard. People are inclined to do what they did in a previous situation.

  • People’s belief in the warning. Belief is not determined by the credibility of the source issuing the warning but by the frequency the message is heard.

  • People’s knowledge about the hazard. This includes previous information and data gained in the event or by cues.

  • People’s perceptions of risk. This includes perception of the threat before the event and perception of risk from the specific event.

  • The extent of social interactions during the event. This includes efforts to contact others about the event, being contacted by others, and being able to confirm the message as accurate and credible.