Research in Human Ecology Miscommunication during the Anthrax Attacks: How Events Reveal Organizational Failures Karen M. O’Neill1 Jeffrey M. Calia Caron Chess Department of Human Ecology Rutgers University New Brunswick, NJ Lee Clarke Department of Sociology Rutgers University Piscataway, NJ Abstract This study of the anthrax attacks of 2001 treats risk communication as a series of events that can be analyzed to discern the strengths and weaknesses of organizations charged with responding to emergencies. To investigate how organizational practices shape risk communication, we use a method developed primarily for comparative-historical case studies called event-structure analysis. We analyze events leading to false media reports of anthrax infections in one New Jersey town soon after an infection by a potentially lethal strain of anthrax was confirmed in a nearby postal facility. This analytic method highlights the failures of organizations to institutionalize public health practices, which allowed contingent events to determine risk messages and responses. Keywords: risk communication, event-structure analysis, organizations, institutions, bioterrorism Introduction The risk communication field has drawn heavily on case studies to derive suggestions for practice and directions for research. Some of the field’s seminal works are in the form of case studies. For example, one of the earliest works in the field Environmental Hazards: Communicating Risks as a Social Hazard (Krimsky and Plough 1988) is comprised of case studies that explore communication between government agencies and communities. Wynne (1989) illustrated, through a case study of farming after Chernobyl, problems resulting from scientific experts’ failure to consider indigeHuman Ecology Review, Vol. 14, No. 2, 2007 © Society for Human Ecology nous knowledge. The National Academy of Sciences report Understanding Risk: Informing Decisions in a Democratic Society illustrates, through an appendix of case studies, failures in risk communication and the promise of integrating risk analysis and deliberation (Stern and Fineberg 1995). More recently, scholars used case studies to develop the concept of the social amplification of risk (Pidgeon et al. 2003), the idea that communication practices and other social variables may amplify or attenuate societal responses to risk. Although interactions among actors are often clearest in the narratives of case studies, researchers have only recently used case studies to analyze how organizational dynamics affect the ways that actors communicate about risk. In this paper, we apply a systematic method for analyzing case studies that can identify how organizational factors and institutional expectations affect communication. Our aims are to introduce a method that can be applied to a broad range of questions about risk communication and to show how organizations enact, or fail to enact, institutional expectations about risk communication as they confront a particular risk. Our analysis of institutions contributes to the study of organizations and risk by showing how expectations about proper organizational behavior emerge as organizations interact. Our focus on improving methods in this field is important because while the risk communication field has been advanced by case studies, the studies themselves are not notable for their methodological rigor. Some case studies include sections about the selection of interviewees or the interviewing process, but they skimp on explaining the methods used for data analysis. In fact, it is not unusual for risk communication case studies to fail to include any information about analytic methods. Although the social sciences include dif- 119 O’Neill, et al. ferent kinds of case studies (e.g., Ragin and Becker 1992) and different approaches for analyzing qualitative data (e.g., Miles and Huberman 1994), risk communication case studies have not drawn from the full range of available analytic methods. In short, risk communication case studies are rich, but the methods sections, if not the methods themselves, are weak. In an effort to improve risk communication case study methodology, this paper applies a systematic form of data analysis, event-structure analysis, to failed risk communication about anthrax. To our knowledge, there are no other studies that analyze risk communication as a series of events, which is the approach taken in event-structure analysis. Advocates of systematic social science approaches to history argue that narratives of case studies—such as those typically written about risk communication—are limited in their power to explain events (McCullagh 1978; Tilly 1981, 8; Griffin 1993), particularly when the narrative is complicated or when the sources report inconsistent details. A narrative serves as rich raw material for explanation (Thompson 1978, 199), but more “information and insight” is needed, as well as a method that makes the analyst’s assumptions and generalizations explicit (Griffin 1993). Event-structure analysis allows the researcher to systematically assess the time order and causal connections between events in an episode to discover which circumstances and events were critical in determining outcomes (Griffin 1993; Hawthorn 1991). Our goal was to see if applying event-structure analysis, by using the software program ETHNO, improved our understanding of a specific risk communication problem: an erroneous televised report on Cable News Network (CNN) about two workers from an Eatontown, New Jersey postal facility having become “ill” from anthrax. Our study asks whether ETHNO improves our deconstruction of the events of this small but revealing episode during the anthrax attacks so we can better understand the organizational dynamics behind this flawed risk communication. The following sections outline research on the organizational and institutional factors that affect risk communication and explain how event-structure analysis can be used to analyze these factors. We describe the events in New Jersey leading to the false reports and then outline the purposes and processes of applying event-structure analysis, showing how this method allowed us to reconsider and then systematically test our hypotheses for outcomes. We conclude by discussing the specific implications of our findings and the potential usefulness of event-structure analysis for the risk communication field. Risk Communication and Organizations 120 While the meaning of the term “risk communication” is debated among practitioners (e.g., Lundgreen and McMakin 1998) and academics, an oft-cited definition is offered by the National Research Council’s report: “Risk communication is an interactive process of exchange of information and opinion among individuals, groups, and institutions. It involves multiple messages about the nature of risk and other messages, not strictly about risk” (National Research Council 1989, 21). Event-structure analysis is a method that examines these processes of interaction step by step to reveal the constraints these actors face. The NRC definition mentions the interactions of groups as well as individuals, but because of risk communication’s roots in psychological theory, studies in this field often focus on the individual level of analysis. Such studies have particularly failed to explore in detail organizations that are the sources of messages, including the processes that affect organizations, and thereby messages. Thus, many influences on the purposes, timing, form, and content of the message are left relatively unstudied. Organizational theory is one of the most appropriate frameworks for explaining risk communication, allowing us to consider what is communicated, when, and by whom (Chess et al.