Hazard Vulnerability, Media Construction of Disaster, and Risk Management

Christine M. Rodrigue, Ph.D.

Professor and Chair
Department of Geography
California State University
Long Beach, CA 90840-1101 USA
rodrigue@csulb.edu
https://home.csulb.edu/~rodrigue/

Education and Training in Disaster Medicine and Major Incident Management
An International Working Conference of the
World Association of Disaster and Emergency Medicine
Brussels, 29 October-1 November 2004


==============================================================

I am a social scientist, a social geographer, interested in hazards and disasters. [ slide 2 ] There are three main concerns in my work across a variety of different case studies: equity issues in risk and vulnerability, the impact of media on hazards and disaster perception and vulnerability, and communications between risk assessment science and risk management policy. I'll organize my comments around these three themes, drawing on the case studies as needed. [ slide 3 ]

==============================================================

Equity Issues in Risk and Vulnerability

______________________________________________________________

The terms, "natural hazard," "technological hazard," "disaster," "risk," and "vulnerability" are among those listed in the WADEM working paper (Issue 1.2) as having overlapping and sometimes confusing meanings and usages in the disaster research and practitioner communities. A "natural hazard" necessarily combines two elements: an extreme event and a human society situated in time and space so that its people and assets are in the path of that event . Natural hazards require integration of the social sciences to understand them, prepare for them, and respond to them.

Adding to the complexity of the concepts, a "natural" disaster, such as a landslide, may be the result of human alteration of the landscape, such as deforestation, mining, or overgrazing, causing a particular "natural" event to shade into a sociogenic or technological disaster. And a purely natural disaster, such as an earthquake, can trigger complex cascades of interacting natural and technological hazards, such as the collapse of a dam and resulting floods, the release of toxic chemicals into air or water, the liberation of Coccidioidomycosis during quake-triggered landslides, or the rupture of lifelines and the eruption of fires that cannot be contained.

The term, "hazard," describes a "risky" or "dangerous" situation, a potential, while "disaster" is usually reserved for the realization of that potential in a large scale event, in which many "vulnerable" people become casualties. All these meanings shade into one another, but it is sometimes useful to create boundaries between terms to draw out a particular concept of theoretical importance. [ slide 4 ]

I had to differentiate one pair of such concepts, "risk" and "vulnerability," to analyze the peculiar social geography of one hazard in Southern California: chaparral wildfire hazard in the wildland-urban interface. Chaparral is a scrub vegetation common in California, very similar to the maquis vegetation in the Mediterranean borderlands. It is adapted to the recurrent fires that sweep through the mountains at the end of summer and, in fact, depends on fire for its reproduction and renewal. It actually promotes the fire on which it depends, with the result that, the longer it goes without fire, the higher the probability of a fire and the greater its magnitude. In California, this pyrogenic vegetation grows on scenic mountainous terrain that is highly valued by California residents, only the wealthiest of whom can afford to buy them. As a result, the geography of risk to chaparral fire coïncides to a large degree with the geography of wealthy neighborhoods.

This peculiar geography of risk seems to counter the expectations of structural hazards theory. Structural theory, originally developed in Third World situations, focusses on how the poorest and most marginalized groups in society are disproportionately vulnerable to natural and technological disaster. The expectations of structural theory would appear to founder on the First World case of chaparral fire hazard in California. In fact, they don't. In California, it is important to differentiate "risk" from "vulnerability." Risk is a statistical concept quantifying the probability of specified dangers; vulnerability is the inability to evade, withstand, recover from, or externalize the costs of a disaster or of preparing for it.

The montane suburbanites of California certainly incur greater risk to chaparral fire in their search for a home with a view, but their vulnerability is socialized to the rest of society through insurance and tax mechanisms. Obviously, fire-fighting costs are necessitated by residential development of the fire frontier; less obvious is the substantial insurance subsidy for such development through a mandatory assigned risk pool, which provides “affordable” insurance for those otherwise denied coverage. Other policyholders of more modest means, holding standard fire policies in suburban and urban neighborhoods far from the montane suburbs, thus subsidize the cost of wealthy residential development in the fire frontier. Risk is voluntarily assumed by well-to-do households with the means to pursue a home with a view and exclude others from their viewsheds; vulnerability is externalized and socialized to others less well-off than they are, who then enable the pursuit of a dangerous landscape aesthetic on the part of the hillside residents. The result is that social vulnerability to this particular hazard nearly reverses the distribution of risk. [ slide 5 ]

==============================================================

Impact of Media on Hazards and Disaster
Perception ... and Vulnerability

______________________________________________________________

A second focus of my work is the rôle of media in constructing public understanding of hazards and disasters and the risks faced by the public. Media coverage is often unsatisfactory due to its proclivity for sensationalism and for casting any event into a preset narrative of human drama with clashing opponents or the search for blame. [ slide 6 ]

______________________________________________________________

Risk Amplification in the Media

--------------------------------------------------------------

Risk amplification occurs when media sensationalism causes hazard perception and public concern to be magnified far beyond levels proportional to the risks estimated in risk assessment science. Such agenda-setting can create political activism or disruption more costly than the original hazard.

An example of risk amplification is the controversy that erupted over the plutonium dioxide on the Cassini-Huygens mission. The mission was to launch in October 1997 and swing by Earth again in August 1999. Internet activists expressed concern that the 33 kg of Pu-dioxide carried on the spacecraft could cause anywhere from 200,000 to 40 million deaths if there were a launch or flyby accident. Conventional risk assessments conducted by independent organizations for NASA calculated the launch risk as less than 1% for between 0.55 and 1.50 health effects over 5 decades. For inadvertant entry failures during flyby, there was a 1% probability of 450 surplus deaths being exceeded in 50 years.

Opposition to the mission began to organize itself by 1995 on the web and, even more effectively, on e-mail lists and newsgroups. This campaign obviously did not stop the launch or flyby, but it did generate much controversy and pressure on Congress, the White House, the courts, and even local governments.

I found that 937 people expressed their views on the subject in newsgroups. Of these, 21% were mission opponents, and these generated 31% of the 8,020 posts I tracked. Of these 2,508 oppositional messages, 24% simply passed on others' messages. I followed those forwarded messages back to their original authors and found they originated with about half a dozen individuals! These few people were able to generate a major controversy through use of the Internet, because of the exponentially expanding number of people who can be reached as people forward messages to their friends, who then resend the messages to their friends. This kind of exponential progression represents a significant democratization of access to the means of public communication in society away from the traditional print and broadcast media. It also represents a significant increase in the potential for demagoguery as this media power can be used for risk amplification, as, I believe, it was in this particular case. [ slide 7 ]

______________________________________________________________

Risk Attenuation in the Media

--------------------------------------------------------------

Sometimes the media acts, instead, to attenuate perception of a significant hazard and thereby blunt concern and political pressure to do something about it. Chaparral wildfire hazard again provides an example. The pyrogenic nature of chaparral has been appreciated for decades by researchers and the insurance industry in California. It became more widely known when Mike Davis, a caustic observer of Los Angeles' social and environmental scene, began to publicize the wildfire hazard in a series of inflammatory articles in the alternative press and then as a chapter in his book, The Ecology of Fear.

The original publication of this book in 1998, with its chapter, "The Case for Letting Malibu Burn," together with Davis' MacArthur Foundation genius grant, was quickly followed by an astounding ad hominem attack on his character in October 1998. Leading the charge against Davis was one Brady Westwater, who began the controversy by posting a large web page accusing Davis of "error, deception and mistakes" in his research. Westwater then argued that, because of “residential development in the wildfire-urban interface, the fire damages will actually decrease over the years" in Malibu. "Brady Westwater" turned out to be a pseudonym for a realtor in Malibu, whose living depends on selling those structures in the chaparral. The "Westwater" web page then became the basis for print media attacks in many newspapers around the world.

Lost in the invective has been any sense of the risk to which wealthy home buyers in the montane suburbs of California are subjecting themselves, their families, and their prized possessions, not to mention the social inequities involved in paying for their inevitable losses. An understanding of the fuel cycle had been delivered forcefully to a broader public audience by an able and entertaining author, which set off an explosion of protest by a handful of individuals with direct interests in continuing suburban development of the fire frontier. The media treatment of Mike Davis by those with a stake in his message may well discourage others from engaging in popular discourse to inform policy toward development on the fire fringe and public perception of this and other hazards. Agenda-setting by the media in such cases amounts to removal of a significant hazard from the social and political conversation. [ slide 8 ]

______________________________________________________________

Media Social and Political Bias during a Disaster

--------------------------------------------------------------

Not only can media amplify or attenuate hazards perception and political action, it can bias the representation of an actual disaster, with negative impacts on the equity and efficiency of response to a disaster and public understanding of the hazard realized in it. Faulty response can thereby accelerate polarization tendencies in a society in the wake of a disaster. I have encountered these phenomena in two of my case studies, one involving the Northridge earthquake of 1994 and the other the September 11th attacks on the United States. [ slide 9 ]

..............................................................

The Northridge Earthquake. -- In the case of the Northridge earthquake, my colleagues and I collected address-specific information on the conditions of more than 100,000 inspected buildings, that is, whether they were condemned, seriously damaged, or not seriously damaged. I grouped the data to generate a community-level geography of damaged buildings. My colleagues and I went through the media coverage of the disaster in the Los Angeles Times and in La Opinión, counting place name references in articles about the quake to generate a community-level geography of media attention.

We then related the two geographies to one another through simple linear regression, finding that the geography of damage accounted for only about 34% of the geography of media attention. We then focussed on the communities that departed markedly from the regression line, extracting their Census demographics. The overcovered communities were predominantly non-Hispanic white at 61% and prosperous, with per capita incomes of about $26,000, while the undercovered communities were predominantly minority (only 22% non-Hispanic white) and of more modest means (with per capita incomes of about $14,000). Media coverage, then, appears skewed by race and income.

We then conducted a random survey of Los Angelenos to get their collective mental map of the disaster. The place names they mentioned as the hardest-hit resulted in a geography that coïncided almost perfectly with the geography of media attention and relatively little with the actual geography of damages.

Rather disturbing is the association between media attention, itself skewed by race and income, and rates of recovery. Among the 8 overcovered communities, the crude recovery rate was 42% from April to August 1994. Among the 9 undercovered communities, however, there was a much slower recovery rate of 34%. The overcovered communities recovered much more quickly than average in Los Angeles, while the undercovered communities recovered rather more slowly than average.

While it might be argued that wealthier neighborhoods have more resources with which to recover more speedily from an earthquake, it was interesting to note that two of the undercovered communities were very wealthy white neighborhoods -- Woodland Hills and Chatsworth -- and that these two communities also lagged in their recovery rates. These two cases suggest that media may have an independent effect in slowing recovery rates by obscuring the extent of the damages. [ slide 10 ]

..............................................................

American Print Media Coverage of the September 11th Attacks. -- In the case of the 9/11 attacks on the World Trade Center and the Pentagon, I did a literature content analysis of the front-screen or home page coverage of the online Los Angeles Times, one of the three dominant newspapers in the United States. Three main themes dominated the coverage of these events in the three months following. The military response was the largest of these, with nearly 20% of the 558 stories analyzed falling in this category, and this was the only category that remained prominent for all 12 weeks of coverage, completely dominating coverage for the last 9 weeks of these. Stories about the crime investigation and about the impacts of and reactions to the disaster itself each comprised about 10% of the coverage. The crime story never dominated coverage, while the disaster coverage only dominated the first couple of weeks after the events.

Sensationalism is a common criticism leveled at media during disasters, and it is evident here in the form of obsessively repetitive imagery of United Flight 175 striking the South Tower on the front page graphics of the L.A. Times. Also, the coverage of the anthrax bioterrorist incidents amplified public concern far above the actual numbers of people exposed, sickened, and killed, leading to pressure on physicians for wanton prescription of Cipro.

Poor contextualization is a common feature in the coverage of any disaster. In this case, only 6 stories appeared on the front screen of the L.A. Times about the geopolitical background that produced such homicidal and suicidal men. These comprised only 1% of the stories on the front screen! This is a new hazard for Americans, and context is key to their understanding and preparation for this probably permanent new element in their lives. The media aversion to full discussion of context does not serve them well, and American coverage of 9/11 and events following on it, such as the wars in Afghanistan and Iraq, are quite parochial.

One surprise in this case study is that the Times actually did a fairly equitable job in treating different classes of victims. To its credit, the L.A. Times covered impacts on businesses and impacts on workers in roughly equal numbers in front screen stories. The balance of Times coverage was quite unexpected in light of the skewed coverage of the areas victimized by the Northridge earthquake. [ slide 11 ]

==============================================================

Interactions between Risk Assessment Science
and Risk Management Policy

______________________________________________________________

I have increasingly become interested in how risk amplification and risk attenuation impact public discourse about hazards and how that discussion affects political decision-making about risk management. This concern with communication about risk is now beginning to lead me into studies of institutional culture and communication within a hierarchy with risk management responsibilities. [ slide 12 ]

______________________________________________________________

Risk Amplification and Attenuation and Elected Decision-Makers

--------------------------------------------------------------

Risk amplification can result in considerable public pressure on elected decision-makers to do something about the perceived hazard, as was seen in the scare about plutonium-dioxide on the Cassini-Huygens spacecraft. Elected decision-makers may be confronted with a marked disparity between the level of risk emerging from conventional risk assessment science and the level of risk perception and political concern being expressed by constituents. The dissonance between the two can present a unique hazard to the elected officials. That is, if they opt to support the results of conventional risk assessment, they may be in danger of courting public outrage that could result in eviction in the next election. If, on the other hand, they decide to act in accordance with amplified risk concerns to please the electorate, there may be opportunity costs in the economic benefits foregone. The elected risk managers, thus, face their own Type I and Type II error framework in addressing risks to their own careers!

Similarly, if risk attenuation takes place because of the lack of appropriate media agenda-setting, politicians aware of a serious risk may endanger their own careers by imposing regulations that create opportunity costs on their constituents when those constituents don't believe in the seriousness of a hazard the regulations address. An example would be the imposition of fire-safe roofing standards in wildfire country or a tax measure to rationalize the firefighting system. [ slide 13 ]

______________________________________________________________

Risk Communication within Organizations: Disaster by Management

--------------------------------------------------------------

I've become interested in how communication about risks moves along a hierarchical bureaucracy and how people embedded in that hierarchy make decisions about whether to manage these risks themselves or to move the problem to another level. I have recently looked at two spectacular institutional failures: the crash of the spacecraft Columbia in February 2003 and the failure of FBI Headquarters to respond to the inquiries of field offices concerning the men who would later hijack the four airliners on September 11th, 2001. [ slide 14 ]

Both the Columbia accident and the FBI handling of field office concerns before 9/11 seem to validate normal accident theory. Normal accident theory argues that accidents are "normal" in large, complex systems, because of unexpected interactions among components as much as from failure of single components. In these two case studies, communication about risks appears to have been hog-tied in complex bureaucracies. Unpredictable external political constraints acted on both agencies and led to a shift in risk managers' perception of the relative importance of the precautionary principle and the opportunity costs its application can impose.

In NASA's case, the failure in communication can be traced to its external political environment and funding base, its geographically ornate and hierarchical structure, and the lower status and timidity of risk assessors compared with managers. In the FBI's case, the most egregious failure of communication was between the most senior levels of the Bureau and the lower-ranked personnel there at Headquarters, which affected their decision-making concerning the distant field offices' inquiries.

In both agencies, there were, additionally, parallel chains of command and communication. At NASA, individuals may find themselves wearing hats as engineers, as technical staff within the Shuttle Program, and as employees within the line structure of a NASA center, and it may not be clear to them which chain they should jerk to call attention to a safety-of-flight issue. At the FBI, intelligence and criminal investigation functions are kept strictly separated and compartmentalized, but no-one at the top seems able to collate findings from one function that are relevant to the other. There is little effective risk communication down the FBI hierarchy, such that the decisions of low-level Headquarters personnel to dismiss field office inquiries are quite defensible and reasonable in light of what they knew, despite the horrific outcomes.

The consequence of these barriers to communication along hierarchies, between chains of command, and across space was an imbalanced focus on the managerialist concerns of efficiency, budget, scheduling, and rules and regulations, instead of on the risk to human life. Managers had normalized anomaly and resisted data that contradicted their biases in perception, leading to what one NASA engineer called "worlds of pain." [ slide 15 ]

==============================================================

Conclusions

______________________________________________________________

In conclusion, my work can be distilled into three major concerns: differential vulnerability to disaster, media impacts on risk perception and policy, and communication between risk assessment and risk management. Certain kinds of people are likelier to live in more risky locations, usually, but not always, the poorest and most marginalized groups. Personal resources and the political power for commanding common resources to cope with hazard and disaster are not evenly held. This ensures that any existing social polarization is accelerated after a disaster, even in situations where the wealthiest and most powerful people choose to locate themselves in the riskiest places.

Media sensationalism, shallowness of coverage, and story framing result in risk amplification in some cases and risk attenuation in others. Media agenda-setting can thus structure risk management public discussions around faulty hazards perception, leading to inappropriate public pressure on risk managers. Too much pressure, and opportunity costs are exacted. Too little pressure, and the precautionary principle is completely ignored.

Risk communication also goes on within an institutional hierarchy through such internal media as memos, reports, telephone calls, face-to-face discussions, e-mails, and instant messaging. Risk assessment messages typically originate at political levels subservient to risk managers and spatially remote from them, so urgent risk communications about danger to human life may fail to evoke appropriate responses out of managers or be subsumed under a barrage of competing managerialist concerns at their level.

This kind of obstacle or breakdown may be quite familiar to emergency health workers and medical doctors in disaster situations, and health care professionals in ordinary medical milieux. How many times do doctors find that a hospital's financial bottom line is constraining their freedom of judgment to respond to patients' needs? How many times do nurses find that urgent communications about a patient's situation in hospital fail to provoke immediately needed attention from their supervisors, attention that they are not allowed to provide on their own in a hospital framework? How many times does an entire medical complex find itself undermined by a political system with an excessively attenuated understanding of the very real risk of underfunding emergency services?


==============================================================