IRCD Researchers Meeting Abstracts

Sudha Arlikatti, University of North Texas
Michael Lindell, Texas A&M

Social Perceptions of Flash Floods and Motivations for Protective Action Behavior: A Case Study of the 2013 Flash Floods in Uttarakhand, North India

Flash floods are some of the most dangerous weather-related disasters, as they are involuntary risks where individuals are unaware of the onset of the event, have not chosen to associate with the event, and have no control over how the event will affect them (Knocke & Kolivra 2007, pp. 159). Consequently people’s immediate response is very likely to determine whether they survive the event. Unfortunately, there is nonexistent literature on this topic in developing countries of the world where official flash flood warning systems are lacking. This study funded by a NSF-RAPID grant aims to fill this gap using variables from the Protective Action Decision Model to guide data collection about 316 villagers’ responses to the June 17th flash-floods in Uttarakhand, India. Seeing the flash flood coming (37.3%), understanding that torrential rains would cause flash flooding (23.1%), and observing unusual changes in the mountains (5.4%) were people’s primary environmental cues. However, the results show most respondents received their first information about the flood from social sources such as friends, family and other villagers (45.9%) and themselves (42.1%), rather than from the mass media. The content of these warnings included what the threat was (82%) and what actions to take (49%). Most respondents reported their homes were either moderately damaged (14.6%), severely damaged (25.0%), or completely destroyed (48.4%), but only 9.2% had a family member killed or injured. These findings clearly demonstrate that although over 98% villagers had never experienced a flash flood in the past 10 years, their social perceptions motivated them to take adequate protective actions that saved the lives of their families and themselves. Despite a lack of a well-established flash flood warning system or educational campaigns and assistance from the government and the mass-media, these villagers’ traditional knowledge about their environment helped them survive.

Return to Index

Eric Best, Jacksonville State University

Examining the Efficiency of Disaster Aid Policies Using Mortgage Performance Data

This paper examines United States mortgage data from Alabama, Florida, Louisiana, and Mississippi between 1999 and 2013. Using a dataset that contains performance information from hundreds of thousands of area mortgages, it is possible to examine the effects of major disasters including Hurricane Katrina.

There are significant mortgage performance differences between urban and rural and rich and poor areas, regardless of disaster declaration status. Using the mortgage performance data from the years after Hurricane Katrina, it is possible to examine the success or failure of disaster-related financial aid policies. Data suggest that aid is frequently distributed to more resilient areas where disaster victims are not experiencing post-disaster financial hardship, while at the same time not sufficiently assisting homeowners in vulnerable areas who experience significant financial hardship.

Preliminary analysis shows that over 60% of mortgages in some areas of New Orleans were delinquent in the first two months after Hurricane Katrina, while mortgage delinquency was virtually unchanged in inland communities that received access to the same disaster aid. In the most vulnerable communities, over 10% of mortgage holders missed six straight months of payments after Katrina, indicating that aid programs were not working as intended half a year after the storm. Meanwhile, many inland communities within the disaster declaration area experienced historically low delinquency rates after the storm, indicating that aid was inefficiently sent to highly resilient areas with significant political influence while highly vulnerable areas experienced unprecedented hardship.

This paper suggests the use of long-term financial data such as mortgage and consumer credit performance data to study the efficiency of past disaster aid to make recommendations to improve the understanding of how to distribute future aid. By focusing on reducing hardship, we can counteract the political inefficiencies that result in disaster aid being distributed to communities that do not need assistance

Return to Index

Hsien-Ho Chang, University of Delaware

The Analysis of Incident Command System (ICS)

Since its establishment in the 1980’s, many ICS discussions have focused on its pros and cons. These discussions are actually related to the benefits and limitations of using a mechanistic system versus an organic system.

On the one hand, using a mechanistic system enables Incident Commanders to command all responders; however, this means that every ICS supervisor has to maintain reasonable span of control, which can be problematic when responding to disasters that impact broad areas.

Organizational theorists say that mechanistic and organic systems are not dichotomous. It is consequently possible that the ICS have both mechanistic and organic characteristics. Therefore, one research question is generated: to what degree is the ICS mechanistic or organic?

To explore this research question, the researcher uses direct coding method to analyze two ICS official documents and three online courses. This presentation will demonstrate his results of content analysis, which indicate that ICS incorporates both organic and mechanistic design elements and the natures of this system.

Return to Index

Zhen Cong, Texas Tech University
Ali Nejat, Texas Tech University
Jian Luo, Texas Tech University
Dann Liang, Texas Tech University

Family Structures, Relationships, and Relocation Decisions after Hurricane Sandy

This study examines how family bonds affect rebuilding/relocating decisions after hurricanes. The sample was from a survey of 129 individuals recruited from Staten Island after it was seriously damaged by the 2012 Hurricane Sandy. Multinomial logistic regression was used to investigate how respondents’ family structures before Sandy and whether their relationships improved with family members after Sandy influenced their plans for rebuilding/repairing their homes or relocating to other places. Multinomial logistic regression was also used to examine whether those factors affected respondents’ suggestions to a vignette family concerning rebuilding and relocating. Results indicated that respondents who lived with family members before Sandy were less likely to plan for relocating than those who lived alone. More detailed examination showed that this effect was driven by those who improved on their relationships with family members; those who did not improve on their family relationships were not significantly different from those who lived alone concerning rebuilding/relocating planning. Those who improved on their relationships with family members were also less likely to suggest the vignette family relocate. This study supports the general hypothesis that family bonds reduce the endorsement for relocating and provides empirical evidence that family mechanisms are important for the rebuilding/relocating decision making.

Return to Index

Jacqueline W. Curtis, Kent State University
Andrew Curtis, Kent State University
Steve Smith, Missouri Southern State University

Using Geonarratives to Understand Environmental Perception in Post-Disaster Recovery: A Pilot Study from the May 22, 2011 Joplin Tornado

Background: People’s perception of characteristics in the urban environment influences a range of behaviors which then have implications for their health, as well as for the vitality of their neighborhood. Geographic Information Systems (GIS) is often employed for examining this relationship. In particular, GIS has been used to map characteristics of neighborhood environments that are hypothesized to have negative impacts on health outcomes and other measures of well-being. However, a recent call has been made to move from a focus on purely observable features of neighborhoods to how these features make people feel. This project extends the existing research through use of the geonarrative in a post-disaster environment.

Research Question: Do geonarratives provide understanding of how residents perceive the post-disaster environment? If so, how should this approach be conducted and data analyzed for broader use?

Methodology: The geonarrative is an approach that enables participants to be prompted by their immediate environment. In essence, they are geographically immersed and are asked about the meaning of locations or characteristics of locations in which they are present. Geonarratives were conducted as part of a pilot study in spring 2014, three years after the Joplin tornado. The commentary provided in the geonarrative is collected with spatial video, so the location of the environmental features being mentioned and the transcription of the comments can be mapped in a Geographic Information System (GIS) for spatial analysis and content analysis.

Findings: By overlaying multiple such maps from participants, weighted surfaces of importance to the residents are created. Recovery planning can be informed by identification of those features with high agreement.

Conclusions: Findings from this study argue for the need to investigate how resident perceptions of the post-disaster environment influence how they feel and behave. Implications are presented for use of this approach to inform recovery planning.

Return to Index

Chris Emrich, University of South Carolina
Eric Tate, University of Iowa
Seth Spielman, University of Colorado

Improving the Validity of Social Vulnerability Indicators

In recent years, social vulnerability has received increasing attention as states and communities seek to reduce adverse impacts of natural hazards. Yet far too little is currently understood about the reliability of the leading vulnerability indicators, both in terms of the approaches used to build them and their ability to adequately represent real world conditions. There is an unmet need for theoretical and methodological advancements to improve the validity of social vulnerability indicators. Without such advancements, the expanded use of social vulnerability indicators in disaster preparedness, mitigation, and recovery could mislead decision-making.

The academic literature is replete with case studies that build a quantitative social vulnerability index, but few have sought to examine the accuracy or precision of the vulnerability model. This presentation outlines a conceptual framework for a new generation of social vulnerability indicators designed to improve both internal and external validity. The framework addresses internal validity through reduction of uncertainties in demographic data inputs and indicator construction methods. This enables an examination of external validity through comparison with observed disaster outcomes. Preliminary implementation of the framework has begun through support from the National Science Foundation (Award #1333190), and examples will be provided.

Return to Index

Y. Gurt Ge, North Dakota State University
Shih-Kai Huang, Texas A&M University
Michael Lindell, Texas A&M University
Carla Prater, Texas A&M University
Hao-Che Wu, Sam Houston State University

Vulnerability and Hurricane Evacuation Expectations in the Lower Rio Grande Valley

The Lower Rio Grande Valley (including Cameron, Willacy, and Hidalgo counties) is a fast growing region in the southernmost tip of Texas with a long coastline in the Gulf Coast. Hurricanes and the resulting floods remain critical threats to the traditionally agricultural area. Having a large population of Hispanic migrants, the Valley has substantially higher poverty rates, lower education level, larger household size, and more health challenges. Environmental and social vulnerabilities affect efficient and effective evacuations from hurricanes in terms of limited evacuation routes, shortage of evacuation resources, and needs for evacuation information dissemination.

This study conducted a behavioral survey of urban areas and unincorporated colonias in the Valley, examining residents’ information sources, expected evacuation impediments, expected hurricane impacts, perceived structural vulnerability, and demographic characteristics. Urban data collected in a mail survey of 1,198 households yielded 254 valid questionnaires (response rate = 23.3%), and 227 completed questionnaires were obtained via personal contact interviews with 450 colonia households (response rate = 50.4%). By combining the two surveys, potential factors impacting residents’ evacuation expectations were examined through correlational analyses and multiple regressions. Relying on information sources such as National Hurricane Center, local TV and radio was more likely result in evacuations from major hurricanes, while the strong community network in the Valley elicited a higher likelihood of evacuations from minor storms. Expected evacuation impediments (economic impediments, traffic impediments, and border checkpoints) did not significantly affect evacuation intentions. Expected personal damage and casualties had positive correlations with evacuation expectations for the minor hurricanes. Similarly, perceived structural vulnerability also had positive correlations with evacuation expectations for minor hurricanes. Some demographic factors are negatively associated with evacuation expectations for minor storms, such as level of education, household income, and number of registered vehicles.

The results of this study suggest that emergency managers need to make aggressive efforts to implement hurricane evacuation plans and allocate evacuation management resources for the vulnerable population.

Return to Index

Dana Green, University of North Carolina, Chapel Hill

Evacuation Planning around Nuclear Power Plants: Is the United States Prepared for What Happened in Japan?

The earthquake, tsunami, and ensuing nuclear crisis in Fukushima, Japan has sparked a renewed interest in the topic of evacuation planning around nuclear power plants. Questions of safety, whether what happened in Japan could ever happen in the United States, and whether or not the American people are prepared to evacuate were raised, not only by high level politicians and emergency managers, but also by the American people who, sadly, watched the events unfold in Japan as the Fukushima Daiichi Nuclear Power Plant went into crisis. While the events in Japan were horrific, they have prompted emergency planners in the United States to re-evaluate issues of preparedness and whether the US population would be a) aware, b) armed with the necessary tools with which to protect themselves (e.g., correct dosages of potassium iodide (KI)), and c) would know where to go (based on shifting wind patterns) in the event of a nuclear or radiological event. Given this, this paper begins with an examination of the preparedness and planning efforts for nuclear power plants in the United States, and pays particular attention to the vulnerability of nuclear reactors in the country, noting that vulnerability of a reactor to an earthquake depends on two primary factors: 1) proximity to seismically active areas, and 2) the ability of the reactor and associated infrastructure to withstand an earthquake. Taking the United States as a case analysis, this paper focuses on seismically active regions of the country (West Coast, Midwest, and Eastern states), and evaluates both the efficacy of the evacuation plans that have been put into place by the nuclear power plants themselves, and how readily evacuation procedures are known (based on fliers, drills, and siren tests sent out per Nuclear Regulatory Commission (NRC) mandates and identified to the general public. Recognizing that the NRC identifies two Emergency Planning Zones (EPZs) – one within a ten (10) mile radius and the other within a 50-mile radius of the plant – this paper concludes with an evaluation of whether the general public in the United States are aware of evacuation procedures and would know what to do in the event of a nuclear or radiological event.

Return to Index

Alex Greer, University of Delaware
Su McNeil, University of Delaware
Joseph Trainor, University of Delaware
Israt Jahan, University of Delaware

Understanding the Relationships between Household Decisions and Infrastructure Investment in Disaster Recovery: Cases from Superstorm Sandy

In the wake of disasters, catastrophes, and repeat flooding events policymakers, researchers, and the media are all devoting more attention to the recovery phase of disaster management, specifically whether resettlement is a better option than rebuilding in situ, and whether to invest in mitigation and repair or just repair the current transportation infrastructure.

While many have discussed these issues in passing, research devoted to household relocation and resettlement decision-making is relatively sparse, and research on the impact of transportation infrastructure on these decisions is even sparser. Most scholarship in this area only tangentially relates to resettlement, or merely offers “best practice” recommendations.

What we do not have is a developed, empirical understanding of what factors households consider when deciding where to live after a disaster, and how important each factor is in the decision-making process. As part of an effort to understand how households decide to either resettle in a new location or rebuild in situ following a disaster, this presentation highlights research findings from a survey of two communities affected by Superstorm Sandy. Using two-tailed sampling of extreme cases, this study examines Sea Bright, NJ, a community that is rebuilding in situ, and Oakwood, NY, a community that is resettling.

Factors in the survey include perceptions of risk, community functioning, disruption, attachment to place and demographic factors compared to their residential decisions following Superstorm Sandy.

Return to Index

Sara Hamideh, Texas A&M University
Michelle Meyer, Texas A&M University
Marccus Hendricks, Texas A&M University
John Cooper, Texas A&M University
David Bierking, Texas A&M University
David Bierking, Texas A&M University
Shannon Van Zandt, Texas A&M University
Walter Peacock, Texas A&M University

Perceptions of Poverty in Disaster Recovery and Aid Distribution

Long-term recovery is “the differential process of restoring, rebuilding, and reshaping the physical, social, economic, and natural environment through pre-event planning and post event actions” (Smith & Wenger, 2006). This process is unique for every community therefore successful recovery requires plans carefully tailored for the unique features of the affected community.

In this presentation, we reflect on how the unique features of a marginalized neighborhood in Texas and recovery assistance network affected the recovery process from a 2013 tornado. Using a qualitative case study approach during the first year of recovery, we conducted 31 interviews with organizational representatives and 23 interviews with residents. We also completed participant observation of 15 community meetings and events related to the recovery. To explore and connect major themes of recovery in our case study, we used a grounded theory approach for data analysis and present our ongoing analysis based on the first two rounds of coding.

The predominantly poor and minority residents of the affected neighborhood lacked material assets and were eligible for charity-based recovery assistance. Our preliminary findings highlight that, on the one hand, recovery for many residents was symbolically and materially strengthening. For example, some became eligible for new Habitat for Humanity homes or received new modular homes for free. Also residents indicated that social ties within the neighborhood grew stronger. On the other hand, this neighborhood was invisible to many local citizens and organizations prior to the tornado. While the charities located in the county rushed to assist recovery, residents were not incorporated in recovery decision-making processes. The charitable impetus of the recovery exacerbated the social distance and subordinate situation of the affected residents to the rest of this wealthy county, highlighting how negative perceptions of the poor affect disaster aid distribution.

Return to Index

Sara Hamideh, Texas A&M University
Shannon Van Zandt, Texas A&M University
Walter Peacock, Texas A&M University

Community Planning for Long-Term Recovery: When and How to do it?

How can we improve recovery planning at the local level? Galveston, following Hurricane Ike, provides a unique case to examine the challenges and successes of a local community in recovery planning. It represents one of the few recovery experiences ostensibly driven by a broad-based and active local community involvement.

There is a general consensus on the value of planning for disaster recovery and its timing as a critical factor (Smith, 2011; Smith & Birkland, 2012, Johnson & Hayashi, 2012, Olshansky et al., 2006). While disaster and planning researchers emphasize the need for planning for long term disaster recovery, experiences and challenges of smaller communities in this area have rarely been studied. In our paper we discuss the preliminary findings of a qualitative case study of planning and policy making for recovery, post hurricane Ike, in Galveston. Our findings highlight conflicts in recovery process as well as opportunities for making changes in the community. Despite a seemingly inclusive post-Ike recovery planning process, competing recovery agendas resulted in prolonged conflicts among local leaders, community members, and state and federal agencies stifling recovery on many fronts. In many respects, preexisting conflicts and competition intensified during the recovery process. Our findings reinforce the assumption that deciding on post disaster change and recovery goals is a challenge that communities should address through collaboration before disasters. Galveston´s experience suggests and reemphasizes the potential value of collaboration between local policy makers in using recovery resources to deal with preexisting issues and to envision possibilities for community betterment through recovery process.

Return to Index

Shih-Kai Huang, Texas A&M University
Michael K. Lindell, Texas A&M University
Carla s. Prater, Texas A&M University

Households' Evacuation Decision in Response to Hurricanes Katrina and Rita

Although evacuation has been recognized as an effective protective action in responding to a hurricane emergency, it is still not clear why some people leave but others do not. In order to better understand this issue, this study began with a statistical meta-analysis (SMA), which is a procedure that has never been conducted previously in the field of disaster studies. The SMA indicates that homeownership, official warning, risk area, seeing peers evacuating, expected hydrological impacts, and expected wind impacts have strong and consistent effects on evacuation decisions whereas female gender, black ethnicity, presence of children in the home, reliance on news media for storm information, reliance on peers for storm information, and hurricane intensity have weaker effects that might be due to mediation through psychological variables. Next, this study collected data from the Hurricane Katrina and Rita evacuations and extended the results of the SMA by testing the Huang et al. (2012) abbreviated protective action decision model (PADM). The results show that (1) a household’s evacuation decision, as predicted, is determined most directly by expected wind impacts and expected evacuation impediments. In turn, expected wind impacts and expected hydrological impacts are primarily determined by expected storm threat and expected rapid onset. Finally, expected storm threat, expected rapid onset, and expected evacuation impediments are determined by households’ personal characteristics, their reception of hurricane information, and their observations of social and environmental cues. (2) Surprisingly, expected hydrological impacts did not have as much of an impact on evacuation decisions as wind impacts—which are associated with expected injuries, job disruption, and service disruption. (3) Official warnings and risk area also had direct effects on households’ evacuation decisions, which can be explained as the peripheral route to persuasion that bypasses messages about the personal impacts hurricane impact (Petty & Cacioppo, 1986). (4) Unlike other hurricane evacuation studies, this one found that expected rapid onset had a significant effect on households’ evacuation decisions, perhaps because both Hurricanes Katrina and Rita had late-changing tracks that might have caused residents to be concerned being caught on the road by a rapidly approaching storm. (5) Supplemental information, such as environmental cues, risk area, and hurricane experience, have effects on individuals’ expectations of storm threat, wind impacts, and hydrological impacts that are similar to those of National Weather Service information that is disseminated through multiple information channels (e.g., news media and official warnings). This implies that households used other sources to place the National Weather Service’s hurricane information into an appropriate context. Nonetheless, some of the results conflict with the model presented by Huang et al. (2012), so further research is needed to determine whether the conflicting results can be replicated and, consequently, require revision of the model.

Return to Index

Emma Hudson-Doyle, Massey University
Douglas Paton, University of Tasmania
David Johnston, GNS Science

Exploring scientific uncertainty through table top emergency management exercises

Successful decision-making during natural hazard events is fundamentally dependent upon the situational awareness of the emergency management officers (EMOs), both as individuals and as a team. This includes their selection, interpretation, assessment and understanding of the available information which defines the problem and identifies solutions. Developing this initial and ongoing situational awareness is critically dependent upon the information and advice from external experts. This makes the quality of shared mental models (developed in training and through critical incident management and simulation) crucial to how external information is used during natural hazard crises. However, this advice is subject to many levels of uncertainty including the stochastic uncertainty (the variability of the system), and the epistemic uncertainty (lack of knowledge). This uncertainty can block or delay action and decision making, and while behavioural decision theories have identified heuristics for coping with such uncertainty, little is known about how these operate in emergency settings. We explored how provision of science advice influenced emergency decisions during natural hazard events using real time group exercises (3 full day exercises with 4-6 practitioners) to investigate how personnel create Incident Action Plans based on a range of injects for a volcanic eruption, with a particular focus on uncertain science advice and the communication of non-consensus advice. Data were obtained from recordings of group activity, questionnaires and debriefs about the processes they went through, the key issues identified, what information they were looking for during the hypothetical event, and what they would do differently. We discuss the development and conduct of these exercises and present findings covering how participants self-organised in exercise settings, and how they coped with the prevailing scientific uncertainty and corresponding disagreement amongst the group. The implications of the findings for emergency manager use, understanding and utilisation of scientific information are discussed.

Return to Index

Kanako Iuchi, Tohoku University

Understanding the Relationships of Risk, Land Use and Time for a Planning Decision after a Major Disaster

Keeping communities away from large-scale future risk is often considered a key measure in early stages of recovery. Planning policies and programs are thus created and/or modified with such shared intention.

However, other pressing issues emerge in the rebuilding process, and resulting resettlement patterns years later tend to be of a similar form as before the disaster. This is especially apparent in places where land use enforcement is lenient; for example in post-Indonesian Ocean Tsunami Aceh, changes in legal framework for risk-averse spatial control were not particularly effective in keeping new inhabitants away from the devastated coastal areas 10 years later.

Focusing on the post-Yolanda Typhoon affected area of the Philippines, this paper aims to understand how decisions to manage risks using land use regulations are initially made, and how key factors influence these to change over time. Fieldwork was conducted between March 4th and 13th, 2014, utilizing semi-structured interviews with national and local government officers, focus group interviews with community (Barangay) leaders, and interviews with local residents.

At four months after the typhoon, keeping rebuilding out of coastal zones was still the primary concern of the government and communities. However, there is speculation that the implementation of risk-averse land use planning will soon become complicated. At the national level, an initial 40-meter setback from the coast prohibiting any construction in “no build zones” is currently softening to a “no dwelling zone”. At the local level, the City of Tacloban is revising their land use plan to incorporate the original national-level designated “no build zone”, and affected communities, regardless of formality of land title, are willing to relocate inland. Nonetheless, the prospective number of temporary and permanent housing units is much smaller than the demand for them, pushing residents to rebuild homes ‘temporarily’ on their original land along the shore.

Return to Index

Nabil Kamel, Arizona State University

Against the Odds: Three Cases of Alternative Post-Disaster Recovery Pathways

Within the field of post-disaster recovery, numerous studies have documented and empirically tested the relationship between pre-disaster conditions and post-disaster outcomes, whereby positive post-disaster recovery outcomes are likely to be associated with favorable pre-disaster conditions and vice versa. The theoretical explanations for this phenomenon can be grouped into two categories. The first category, explains variations in post-disaster recovery outcomes as a function of endogenous conditions. Here, areas with high levels of human capital, social capital, and/or natural capital fare better than their counterparts that lack such assets. The second category, explains uneven recovery outcomes as a function of urban and supra-urban structural conditions. Here, pre-existing political economic interests embedded in socio-spatial urbanization processes influence recovery processes and outcomes. In the US, under a neoliberal regime, class, race, and citizenship determine access to political, financial, and institutional resources and, by extension, determine post-disaster recovery outcomes.

This paper presents three cases where favorable post-disaster recovery outcomes emerged despite unfavorable endogenous and structural conditions. These are cases of marginalized communities recovering from large-scale disasters in major cities: West Oakland and the 1989 Loma Prieta earthquake that hit the Bay Area, Canoga Park and the 1994 Northridge earthquake that hit Southern California, and Village de l’Est and the 2005 Hurricane Katrina that hit New Orleans. The cases illustrate the historically unfavorable structural conditions of discrimination, isolation, and marginalization that these communities faced before the disaster and how these conditions restricted local recovery and development opportunities in the aftermath of each disaster. The cases also illustrate how these communities were able to overcome these structural obstacles and not only restore the livelihood of their communities, but also introduce transformative progressive change in their reconstruction and recovery efforts through various forms of political organizing, creative institutional arrangements, and community solidarity.

Return to Index

Hans M. Louis-Charles, University of Delaware

Predicting Public Support towards Environmental Migrants: Natural Disasters and Xenophobia

“Environmental Migration”, including migration caused by natural disasters, has shown significant resurgence in recent decades amongst academic researchers, policy and law makers. Ongoing policy debates and recent academic publications have focused on the consequences of singling out this form of migration, methods to mitigate climate change, and how a legal protection framework could address concerns regarding both state sovereignty and human rights. Presently there is no coherent multilateral governance framework or legal protection for future Environmental Migrants (EM). Concurrently, there has been limited research analyzing public attitudes towards EM within potential receiving countries. Using multivariate analysis of the 2011 Transatlantic Trends Immigration survey I deciphered the attributes that best determine individual support towards immigration caused by natural disasters and compared them to the attributes that determine support for other forms of immigration. The Transatlantic Trends: Immigration is a public opinion survey that addresses multiple aspects of the immigration and integration debate. The survey is a project of the German Marshall Fund of the United States (GMF), the Compagnia di San Paolo (Italy), and the Barrow Cadbury Trust (U.K.), with additional support from the Fundación BBVA (Spain). My findings show that anti-foreigner sentiment will be a significant hindrance to those forced to immigrate, despite the push factor of a natural disaster. I conclude with some possible policy implications from these findings.

Return to Index

John McClure, Victoria University of Wellington
Justin Velluppillia, Victoria University of Wellington

How Different News Media Reports Shape Judgments that Earthquake Damage Can Be Prevented: Mixed Messages About the Canterbury Earthquake

Previous research has shown that different types of news media reports have contrasting effects on citizens’ perceptions about different risks and hazards. For example, messages that stated that earthquake damage was distinctive led people to see the damage as more preventable than messages that said the damage was universal.

This research examined the effects of two different types of message appearing simultaneously in the news media in the weeks following the February 2011, Canterbury New Zealand earthquake: fatalistic and analytical messages.

Fatalistic messages portrayed widespread, generalized damage and made little reference to the contrasting performance of different types of buildings. In contrast, analytical messages conveyed the distinctiveness of damage and the flawed design of buildings that were damaged.

The study examined the effects of these different messages on judgments of the cause of the damage and the preventability of the damage, fatalism about earthquakes in general, and estimates of the percentage of buildings that were damaged. Participants who were already familiar with reports about the earthquake read either the fatalistic messages or the analytical messages. Participants reading the analytical reports attributed the damage to controllable causes and saw it as more preventable than those reading fatalistic reports. The reports had no effect however, on overall fatalism about earthquakes. These findings show that the different messages in the news media have contrasting effects on judgments about damage in a recent, local, earthquake, despite competing real world information. These results clarify which messages are likely to facilitate preparedness for earthquakes and other hazards, and have several implications for risk communication strategies.

Return to Index

Sisi Meng, Florida International University
Pallab Mozumder, Florida International University

Estimating the Effects of Utility Disruption on Household Well-being During Hurricane Sandy

Critical infrastructure and public utility systems are often seriously damaged by natural disasters like hurricanes. The disruptions in the supply of public utility services result in direct losses of households’ well-being. On October 2012, Superstorm Sandy ripped across 24 states and caused severe damages particularly in New Jersey and New York area. The New Jersey Transit and other public transportation agencies were forced to shut down their services in preparation for Sandy’s arrival. More than 8 million homes were reported to lose power. High winds, heavy storms and flooding brought widespread interruptions to public utility services, internet, wireless connections and other communication devices along Sandy’s path.

The International Hurricane Research Center (IHRC) at Florida International University conducted a household survey to investigate and quantify the economic impacts of utility disruption on household well-being in the aftermath of hurricane Sandy. Using household survey data, we attempt to provide an initial estimate of the total household damage due the disruption in utility services, including electricity, water, gas, telecommunication and public transportation. Furthermore, we also study the household recovery process from the shock inflicted by hurricane Sandy.

Our study contributes to the literature by investigating the relative importance of each utility service by estimating each utility’s contribution to the household’s total economic loss and its role in the recovery process. Understanding the major determinants of damages caused by disruptions in public utility services and the process through which it affects household’s well-being can provide key inputs for disaster management. We discuss policy implications of our findings especially the role of restoring public utility services for expediting household recovery and for promoting community resilience in the aftermath of a natural disaster.

Return to Index

Michelle Meyer, Texas A&M University
Lori Peek, Colorado State University

Longitudinal Analysis of Anti-Islamic Hate Crime Before and After September 11, 2001

Following the September 11, 2001 terrorist attacks, Muslim Americans, Arab Americans, immigrants, and many others who were perceived to be Muslim or of Arab descent became victims of hate crimes and other forms of discrimination and bias (Bakalian and Bozorgmehr 2009; Cainkar 2009; Peek 2011; Welch 2006). Yet, few studies have linked the disaster, criminology, and hate crime literatures together theoretically or methodologically (Peek and Meyer Lueck 2012).

Our paper speaks to how hate crime following a disaster conforms to (or diverges from) the general types of hate crime--thrill-seeking, defensive, and mission--identified by Levin and McDevitt (2002). Thrill-seeking crimes are committed by young men, often involve drugs or alcohol, and are crimes of convenience. Defensive crimes occur in response to a perceived threat such as demographic or social change. Mission crimes are often carried out by members of hate groups who wish to destroy a certain population.

We use FBI data available through the Uniform Crime Reporting Program from 1992 through 2011 to understand how the 9/11 disaster affected hate crime occurrence across counties in the U.S. We build upon our prior descriptive analyses, which show anti-Islamic hate crimes increased dramatically in the immediate aftermath of 9/11 and have never returned to pre-9/11 levels (Peek and Meyer Lueck 2012). For example, in 2011, the mean occurrence of anti-Islamic hate crimes across U.S. counties was five times higher than in 2000. In the present work, we analyze hate crime at the county level using predictors includingMuslim population, race and gender demographics, rate of religious adherents, presence of hate groups, age distribution, poverty, alcohol and drug offenses, and immigration. With these predictors, we show what factors likely influence anti-Islamic hate crime in a county, pre- and post-9/11, and identify how the effects of these predictors changed following the 9/11 attacks.

Return to Index

Qing Miao, Syracuse University
David Popp, Syracuse University

Necessity as the Mother of Invention: Innovative Responses to Natural Disasters

How do innovators respond to the shock of a natural disaster? Do natural disasters spur technical innovations that can reduce the risk of future hazards? In this paper we take a worldwide view in investigating how innovation, as an economic and scientific endeavor, responds to the shock of natural disasters. Drawing on the induced innovation hypothesis, we examine the impact of three types of natural disasters—floods, droughts and earthquakes—on the innovation of their respective risk-mitigating technologies (including flood control, drought-resistant crops and quake-proof buildings). Using patent and disaster data, our study is the first to relate natural disasters to technological innovation and also presents the first attempt to empirically examine adaptation responses across multiple sectors at the country level.

Particularly, we construct a panel of up to 28 countries over a period of 25 years, and investigate the innovation responses using a Poisson fixed-effects model with a distributed lag of recent disaster impacts. Considering the potential endogeneity of disaster damages, we use meteorological and geophysical data to create hazard intensity measures as instrumental variables. We find that all three types of natural disasters have a significant and positive impact on patenting of their corresponding technologies, while the degree of influence varies across different types of disasters and technologies. This result implies that the private sector is adapting by innovating, but in a more reactive than proactive manner. It thus suggests government has a particularly important role to play in developing technologies necessary for mitigating risks so they are in place before a disaster occurs. We also explore whether domestic innovation is spurred by foreign disasters, and find such evidence in the flood case.

Return to Index

John Nichols, Texas A&M University
K Tan, Southwest University of Science and Technology, Mianyang, Sichuan

Mortality in the Sichuan Earthquake

The 2008 Sichuan earthquake occurred on 6:28:01 UTC on 12 May 2008. The event had a moment magnitude of at least 7.9, which places it close to the range of Great earthquakes as defined by Richter. The death toll is set at about 68,000. This paper investigates the change in the mortality statistics with distance from the epicentre. From recent work on a number of fatal earthquakes, it is postulated that the fatality rate of decreases in a logarithmic fashion. The postulated loss rate model has controversial aspects as the type of building is assumed to impact on the constant element of the logarithmic model and not the logarithmic rate. Sichuan provides a well-established data set to further test the loss model.

Return to Index

Walter Peacock, Texas A&M University
Shannon Van Zandt, Texas A&M University
Yang Zhang, Virginia Tech
Wesley Highfield, Texas A&M University

Long-Term Housing Recovery after Disasters: Findings from Miami and Galveston

Disaster impacts are due to interactions between hazard exposure, physical vulnerability, and social vulnerability. Natural disasters can both magnify and accelerate processes already occurring in communities. The authors report empirical work from 1992’s Hurricane Andrew in Miami-Dade, Florida and 2008’s Hurricane Ike in Galveston, Texas to assess and characterize long term trends in housing recovery. We examine trajectories of housing recovery as well as obstacles to building back better to answer the question, “Recovery for whom?”

Longitudinal, parcel-level data on both housing units and households, along with neighborhood socio-demographic data permit modeling and analysis of the pace of housing recovery for different neighborhoods, populations and, most importantly, housing types (single/multi-family; rental/owner occupied). The data clearly indicate that housing recovery is a highly uneven process. Not surprisingly damage has major consequences for recovery; even after four, almost 5 years, the consequences of damage are evident in the rebuilding process. Reaching restoration levels is highly dependent upon the initial levels of damage. Rental housing in general, and housing often associated with rentals, multi-family and duplexes, are much slower to recover. Furthermore, while social vulnerability may find play in different ways, in general housing in socially vulnerable neighborhoods can be slower to recover. Our findings suggest that in Miami income and race/ethnicity were more critical factors with much slower recovery rates for predominantly black neighborhoods, while in Galveston income was the more critical factor with housing in lower income areas lagging significantly in the recovery process. Over time, the differences in these recovery trajectories in terms of tenure, housing type, and neighborhood socioeconomic factors leads to very different housing opportunities within a community and may result in the permanent displacement of more vulnerable residents and redevelopment of previously affordable housing into less affordable housing types, favoring affluent stakeholders but limiting the availability of much-needed housing.

Return to Index

Laurie Pearce, Justice Institute of British Columbia

Exercise Target Red: Establishing Protocols to Encompass Psychosocial Considerations for Casualties and Families in Mass-Casualty Incidents (MCI) involved an Active Shooter

Unaddressed, the psychosocial consequences of being involved in an Active Shooter MCI can increase the risk of short- and long-term negative psychological outcomes and post-traumatic stress, both for those directly involved in the incident and family members. A review of recently published After Action Reports following Active Shooter MCIs in the United States and Canada has indicated that little attention has been paid to mitigating the psychosocial aspects of these events while the event is unfolding and immediately afterwards.

The Simulation Training and Exercise Collaboratory (SIMTEC): Enhancing CBRNE Psychosocial Capacity and Capability Management research project is multi-faceted and includes the development of a series of tabletop exercises to test training materials and psychosocial protocols to reduce the negative impacts of being involved in a traumatic event. Interviews were held with over 20 responders, family members and casualties directly involved in MCIs to identify areas where protocols could assist first responders and better serve families and victims. Working with police, Victim Services, psychologists, paramedics, 911 dispatchers, crown prosecutors and other subject matter experts, protocols were developed to reduce the anxiety and stress of family members while casualties may be held hostage and/or trapped during an active shooter MCI. In addition, protocols were developed to assist police in obtaining clearer and more coherent statements from witnesses who are extricated from the MCI and reduce their trauma and anxiety while waiting to be interviewed without contaminating potential evidence.

A table top exercise, Exercise Target Red, was developed to test these protocols. This presentation will present the protocols, the developmental elements of the exercise to test the protocols and the preliminary findings following the exercise.

This project is funded by the Canadian Safety and Security Program, Defence Research and Development Canada’s Centre for Security Sciences and our Project Champion is Health Canada’s Employee Assistance Services Bureau.

Return to Index

Ashley Ross, Sam Houston State University

Measuring Disaster Resilience as a Set of Adaptive Capacities: Insights from the Application of Cutter's Resilience Index to the Gulf Coast

Resilience has emerged as the defining frame to our nation’s emergency management in the past decade. While the idea of resilience as a quality that enables communities to “bounce back” from disasters is a promising one, defining and measuring resilience to disasters in a meaningful way that can guide research and practice is more difficult. This paper focuses on measuring disaster resilience as adaptive capacities or the strengths that communities possess enabling effective response and recovery from disaster events (Ross 2014). To quantify adaptive capacities, the baseline resilience index by Cutter et al. (2010) is replicated and applied to 75 counties and parishes across the Gulf Coast. The five sets of capacities Cutter et al. measure are replicated and explored: social capacity, community capital, economic capacity, institutional capacity, and infrastructure capacity. In addition, measures of ecological capacity are considered. Patterns among the index scores are examined and particular attention is given to the robustness and reliability of the index as applied to this sample. While the findings point to disaster experience and the size of the county as statistically significant correlates of adaptive capacity scores, the analyses of robustness and reliability highlight the problems associated with applying a measurement model to a smaller sample with distinct geographical characteristics and shared history of disaster experience. The paper concludes by outlining the major challenges to conceptualizing and measuring disaster resilience and offering ideas for moving forward.

Return to Index

Andrew Rumbach, University of Colorado at Denver

Planning, Informality, and Uneven Geographies of Risk: A Study of Bidhannagar, Kolkata

This paper describes how urban planning shapes geographies of disaster risk in developing cities. I mean to complicate an oft-repeated claim in the urban disaster literature, that so-called “unplanned” or “informal” urbanization is the primary driver of disaster risk in the Global South. Instead, I argue that formal governance institutions like planning are at the roots of risk because they guide urban growth into hazard prone places but exclude poor and working class households from the infrastructure and amenities that mitigate hazards.

I ground these arguments in a detailed case study of Kolkata, India. Kolkata is one of the largest and most at-risk cities in the world. Like many cities in the global South, disaster and climate risk in Kolkata is highly uneven. Nowhere is this more true than in the satellite cities that dot the urban fringe, where high rise condominiums and luxury hotels overlook sprawling slums and squatter camps that house many of the laboring poor. These informal settlements are vital to the everyday functioning of satellite townships, but are at high risk from monsoon rains, fires, tropical cyclones, and numerous other environmental hazards.

My paper is based on two years of field research in Kolkata and West Bengal. I draw on both qualitative and quantitative data including the results of two original surveys (N=597 and N=414), semi-structured and focus group interviews, archival research, and oral histories.

I conclude that Kolkata represents a growing environmental crises in the developing world, where the fracturing of the urban landscape between spaces for the wealthy and spaces for the poor will soon place hundreds of millions more people at-risk of disaster. If cities like Kolkata are to become resilient, they will require planning and urban governance strategies that are sufficiently broad, inclusive, and forward-looking.

Return to Index

Douglass W Shaw, Texas A&M University
Mara Thien, University of Padua
Riccardo Scarpa, University of Waikato

Perceived Risks of Mountain Landslides in Italy: Stated Choices for Subjective Risk Reductions

Throughout the history of Italy, mountain landslides have occurred regularly, often resulting in fatalities. Because of this history policies that would reduce landslide fatality risk need to be well informed. As a first step in the exploration of such preferences , we examine public perceptions of risk for landslides and related events. Subjective probabilities for others who might die in a landslide, as well as one’s own subjective probability of death are elicited for a sample of visitors and residents of a region in Italy prone to landslides. We also present one portion of the sample with scientific information and allow them to update their risk estimate if they so choose. Subjective probabilities are used to construct risk attributes in a pivot-design version of a conventional stated choice model and larger risk changes away from the baseline are found to be significant.

Return to Index

Laura M. Stough, Texas A&M University
Elizabeth McAdams Ducy, Texas A&M University

The Social Networks of Individuals with Disabilities in Disasters

It is estimated that 23 percent of those affected by Hurricane Katrina were individuals with disabilities (National Organization on Disability, 2006). An already frayed social support system for many of these families unraveled as households were uprooted, homes were lost, and families were separated. Separation from established social networks is particularly detrimental considering the importance of social supports in facilitating recovery from disaster (Norris et al. 2004).

Social support is an asset that promotes preservation or recovery of physical and psychological resources needed for successful coping and is often associated with both mental health and post-disaster mental health status (Kaniasty & Norris, 1995).

The purpose of our presentation will be to report on the social support of individuals with disabilities impacted by Hurricane Katrina. Specifically, we will examine how participants described that their social support changed from before the storm and how their long-term recovery was impacted.

This paper will report results from a larger study designed to evaluate the impact of Hurricane Katrina on individuals with disabilities (IWD) and individuals with diabetes. Thirty-nine hurricane survivors participated in face-to-face interviews designed to elicit information on their experience before, during and after Hurricane Katrina. Interviews were conducted with individuals that had an intellectual disability or diabetes that interfered with at least two areas of daily functioning.

Qualitative grounded theory methodology (see Glaser & Strauss, 1967; Strauss & Corbin, 1990) was used to collect and analyze the data. Social support emerged as the most prominent category and had the largest number of quotes arising from the data. For each category, the corresponding theoretical properties and dimensions were identified.

Results indicated that individuals with disabilities experienced a significant and negative change in the proximity, frequency, level of intimacy, formality, cohesiveness and diversity of social supports after the disaster. We will address how the changed configuration of survivor social support negatively affected the recovery of these individuals with disabilities.

Return to Index

Subrina Tahsin, Florida International University
Pallab Mozumder, Florida International University

Endogenous Risk Perception, Geospatial Characteristics and Hurricane Evacuation Behavior

The focus of this study is to improve our understanding regarding how geospatial factors influence hurricane evacuation behavior. Previous studies indicate that under extreme events, people take preemptive action to avoid risk in a way consistent with their risk perception and evacuation behavior can be seen as a risk averting measure across different socio-economic and demographic strata.

Earlier studies also suggest that people’s risk perception is affected by spatial characteristics in addition to socio-economic and demographic conditions. This gives an indication that hurricane evacuation behavior may be better explained by explicitly incorporating geospatial characteristics.

In this study we have included some novel geospatial variables such as distance from shoreline, distance from storm tract, wind speed and precipitation during landfall, evacuation zone and distance from evacuation route based on the locational information obtained from a geocoded household survey conducted after Hurricane Ivan.

We exploit these geospatial variations to account for endogenous nature of risk perception in analyzing evacuation behavior of more than eight hundred households from Louisiana, Mississippi and Alabama. Empirical analyses based on a set of Bivariate Ordered Probit models suggest that geospatial variables have had significant influences in explaining hurricane evacuation behavior during Ivan.

Return to Index

Eric Tate, University of Iowa
Cristina Muñoz, University of Iowa

Uncertainty Analysis of HAZUS

Risk assessment plays an important role in decision support for reducing flood losses. In recent years, economic loss estimation has emerged as a primary approach to quantify flood risk. Loss estimation models allow users to create varying scenarios of flood severity and built environment characteristics, to determine the effects on physical damage and economic losses. Nearly a decade ago, the US Federal Emergency Management Agency (FEMA) released HAZUS-MH, a geospatial flood loss estimation model for flood mitigation planners. HAZUS users frequently supplant the default datasets the come with the model, with higher resolution local data to develop more precise economic loss estimates. However, this can be a lengthy and costly process. There is a need to understand which types of investments in local data development will lead to the greatest refinements in flood loss estimates.

Return to Index

Jennifer Tobin-Gurley, Colorado State University
Lori Peek, Colorado State University
Robin Cox, Royal Roads University
Cheryl Heykoop, Royal Roads University

Youth Creating Disaster Recovery: A Canadian-U.S. Community-Based Participatory Research Project

Youth have historically been understudied and excluded in disaster research and practice. Yet, a growing body of evidence suggests that they want to be actively engaged, and when they are, can contribute in myriad ways to preparedness, response, and recovery processes. Youth Creating Disaster Recovery (YCDR) is a cross-border, community-based research and youth empowerment program working with youth ages 13-22 in the United States and Canada. Our project is currently located in the communities of Joplin, Missouri and Slave Lake, Calgary, and High River, Alberta. YCDR will soon expand to other disaster affected cities, including: Boulder, Colorado Springs, Estes Park, Evans, Fort Collins, and Greeley, Colorado in the United States and Medicine Hat, Exshaw, Canmore and Bragg Creek Alberta, in Canada. Participatory action research with youth offers unique opportunities and challenges. This is even more true in the post-disaster environment where youth and their communities are struggling to recover and rebuild their lives. This presentation will provide an overview of the YCDR project and our iterative approach to engaging both adults and youth in the focal communities. We will discuss the particular opportunities and challenges associated with using creative and participatory methods--such as Photovoice, Videovoice, and other approaches--to research youth in an extended cycle of exchange in post-disaster settings. Specifically, we will draw from the research findings from the first year of fieldwork and explore those findings in the context of the current literature on participatory methods with children and youth. We will also share some of what we have learned about how these approaches are suitable (or not) when working with young people in a post disaster context.

Return to Index

Joseph Trainor, University of Delaware
Caitlin Gruber, University of Delaware

TV Media and the Communication of Severe Weather Risk

Research often describes how important the media generally and television stations specifically are in disseminating hazard risk messages. Even so, few analyses have looked specifically at the techniques and processes media outlets use to communicate real-time severe weather risks. The purpose of this study is to provide a preliminary understanding of how the media performs that role. Our analyses focus on a broadcast from one of ABC’s local affiliate stations in Oklahoma as they provided coverage of tornado risks on the night of May 13, 2009. Based on in-depth analyses of that data we inductively developed a number of insights into the actors responsible for interpreting the risk and communicating it to the public, as well as the specific techniques used by those actors. The ultimate contribution is to provide a more detailed first look at the very important, but under studied area of warning delivery on television.

Return to Index

Michio Ubaura, Tohoku University

Planning Aid by University Teachers in Recovery Planning Process in Tohoku Region

The damages caused by the Great East Japan Earthquake and the following tsunami were overwhelmingly beyond the response capability of the affected local governments which should be in charge of recovery planning. Most of the local communities in the affected area also did not have sufficient experience of community planning either. It was, therefore, expected that experts should support these planning activities. This paper describes the role of an expert team of university in the recovery planning process of Ishinomaki city after the GEJE and shows its effects and problems.

Ministry of Land, Infrastructure and Transport supported the formulation of recovery plan of the affected municipalities by organizing planning committees in the first year after the disaster. Ishinomaki city government uniquely organized recovery planning conference regularly from the second year, chairman of which was a university professor. Three working groups on three themes were organized under this council, whose chair was university professors, to change information, adjust each projects and to share common recognition of problems. University teachers also support the reconstruction activities on the community level.

The characteristics and effects of these trials are as follows; a team of experts in architecture, civil engineering and urban planning work together exceed their specialized field. They forge a relationship of trust with local government on one hand, and they contribute to consensus building in local community on the other. They carry out a function as an intermediary between local government and community through their technical support. However, some advice was not accepted by local government because of the lack of reality or limited administrative capacity. It constitute another issue that the similar movement does not spread to other regions.

Return to Index

Guochun Wu, China Earthquake Administration

Bottom-up Disaster Management in China: a case study on One Foundation

A variety of disasters have afflicted China with large trauma and losses over the past decade. These disasters promoted the development of China’s disaster management system. We call the governmental disaster management “top-down disaster management” and the participation from communities and NGOs “bottom-up disaster management”. After SARs in 2003, new emergency management was promoted and volunteers' participation became common after Wenchuan Earthquake in 2008.

Based on participant observation, this study focus on Ya'an Earthquake in 2013 and analysed the activities of One Foudation, a specialized disaster relief NGO . We found that corporative mechanism among the local government and NGOs played a role. The private sector for NGOs was developed.

Return to Index