Population Composition, Migration and Inequality: The Influence of Demographic Changes on Disaster Risk and Vulnerability

Transcription

1 Population Composition, Migration and Inequality: The Influence of Demographic Changes on Disaster Risk and Vulnerability William Donner, Indiana University of Pennsylvania Havidán Rodríguez, University of Delaware The changing demographic landscape of the United States calls for a reassessment of the societal impacts and consequences oj socalled "natural" and technological disasters. An increasing trend towards greater demographic and socio-economic diversity (in part due to high rates of international immigration), combined with mounting disaster losses, have brought about a more serious focus among scholars on how changing population patterns shape the vulnerability and resiliency of social systems. Recent disasters, such as the Indian Ocean Tsunami (2004) and Hurricane Katrina (2005), point to the differential impacts of disasters on certain communities, particularly those that do not have the necessary resources to cope with and recover from such events. Ihis paper interprets these impacts within the context of economic, cultural, and social capital, as well as broader human ecological forces. The paper also makes important contributions to the social science disaster research literature by examining population growth, composition, and distribution in the context of disaster risk and vulnerability. Population dynamics (e.g., population growth, migration, and urbanization) are perhaps one of the most important factors that have increased our exposure to disasters and have contributed to the devastating impacts of these events, as the case of Hurricane Katrina illustrates Nevertheless, the scientific literature exploring these issues is quite limited. We argue that if we fail to acknowledge and act on the mounting evidence regarding population composition, migration, inequality, and disaster vulnerability, we will continue to experience disasters with greater regularity and intensity. This work was supported in part by the Engineering Research Centers Program of the National Science Foundation under NSF Cooperative Agrei ment No. EEC Any inquiries regarding this paper should be äirectcä to Havidán Rodriguez at the Disaster Research Center, University of Delaware, 6 HuUihen Hall, Newark, DE at havidan@udel.edu. The Universitv of North Carolina Press Social Forces 87(2), December 2008

2 1090. Soda/fore«87(2) Introduction Population growth, composition and distribution are perhaps the most important factors that have increased our vulnerability to disasters. Increasing population, particularly in coastal regions, has resulted in coastal development, which, in turn, results in the elimination or destruction of buffer zones (e.g., mangroves, trees, sand dunes, etc.). Among other factors, these demographic and environmental changes have systematically exposed greater numbers to natural hazards. This is further compounded by the fact that the minority population, which experiences significant socio-economic disadvantages, is one of the fastest growing populations in the United States. As a result of continuous migration from Latin America and Asia, these populations will continue to grow at a very rapid rate (see Rodriguez, Diaz, Santos and Aguirre 2006). Census projections show that the U.S. population will increase by 44 percent, or about 135 million people by 2050; immigration will be the primary reason accounting for this rapid growth. Consequently, it is extremely important for social scientists to focus on the relationship between population dynamics, vulnerability and disasters. Scholars widely acknowledge socio-economic inequality as one of the root causes of vulnerability to disasters (for a comprehensive overview, see Wisner et al. 2004). All the same, policymakers appear hesitant to embark on long-term strategies or initiatives to systematically address disaster preparedness, mitigation, response and recovery in poor communities. Even the devastation caused by Hurricane Katrina and the ensuing floods - as prolonged and visible as they were - faced by New Orleans' minority and poverty-stricken communities fell short of sustaining the political will to transform disaster policy in a truly meaningful way. It is, therefore, worthy to ask what the future might hold if the current social and political course is not altered as we continue to experience significant demographic shifts, leading to population growth, vulnerability and disasters. This paper focuses on some key issues in light of the changing demographic and socio-economic characteristics - namely, population growth, urbanization, migration and culture, poverty, race and ethnicity, gender and single-mother households, and age and physical disability. In contrast to the relatively common functionalist and collective behavior theoretical approaches in the sociology of disasters, we focus on notions of conflict, inequality and vulnerability. Vulnerability: A Sociological Perspective Within any given society, withstanding natural and technological disasters remains a challenge for disadvantaged groups. Scholars formally call this condition "vulnerability" and define it as "the characteristics of

3 Population Composition, Migration and Inequality io9l a person or group and their situation that influence their capacity to anticipate, cope with, resist and recover from the impact of a natural hazard." (Wisner 2004:11) These characteristics or social attributes, include socioeconomic status, race, class, gender and educational attainment. Disaster-resistant housing and affordable insurance are often beyond the financial grasp of the poorest groups within society, placing them at greater risk. Further, disasters aie hardest on those who are disadvantaged economically, politically and culturally. For example, in the United States, a history of slavery and educational disparities place blacks on the lower rungs of the socio-economic ladder. Also, as a result of military conquest, expansionism and colonialism. Latinos/as^ in the United States confront significant disparities and inequalities (see Rodriguez, Saenz and Menjivar 2008}, Vulnerability is intertwined with the sociological concepts of conflict and inequality. Income and wealth remain important here, but the common struggle over economic resources - the materialist crux of Marx's thought - s but one piece of the vulnerability puzzle. Sharing Weber's general criticisms of economic determinism, Bourdieu (1986) imagines different forms of capital beyond the command of private property as alternative axes of conflict in society. Those in possession of "cultural capital" have acquired and can avail themselves of the highstatus symbols - be they material or ideational - of a society. Education is one means of doing so. These forms of capital vary by social group. By way of a simple example, many Hispanics or Asians may lack cultural capital in the United States insofar as being unable to speak English proficiently. Deficits in "linguistic capital" - a common form of cultural capital - may lead to misunderstandings of warnings or difficulties in applying for disaster relief aid or insurance following a disaster. "Social capital" represents both the number and value of connections one has within a network. Recent immigrants may lack a certain community connectedness known to influence warning rectiption due to their recent arrival to the U.S. mainland. As helpful as they are, concepts of economic, cultural and social capital cannot fully explain vulnerability. This may be because human communities are "complex cooperative arrafigements," the goal of which is the survival of members (Hawley 19E.0:68), to which Duncan (1959) adds the notion of population and social organization. Thus, there remains a "human ecological" dimension in the problem of vulnerability, suggesting that the unique environment within which society is located must be explored together with forms of capital. Adaptation to a coastal environment through urbanization, for example, will have significant consequences in comparison to other forms of adaptation in similar environments.

4 1092. Sociai Forces 87(2) In Contrast to Functionalism and Collective Behavior The approach offered in this article contrasts sharply with functionalist (see Fritz 1961 ) and collective behavior (Turner and Killian 1987) theoretical approaches to disaster. Both of these approaches share the fundamental view that a "disaster" occurs when society cannot function "normally" under the weight of severe environmental changes, an idea which takes root in functionalist notions of consensus. Applying this notion to disasters results in the simplistic assumption that members of society are able to cope with day-to-day problems and meet their everyday needs. The major flaw in the functionalist-disaster perspective is that many groups in society do not have access to the resources that will allow them to meet their everyday needs. When studying disasters, iiuman diversity must be considered alongside functionalist ideas of design and purpose, towards which society is assumed teleologically directed (Hewitt 1983). The so-called "normal" existence of an impoverished black community or a poor Hispanic migrant community along the coast of New Orleans is dramatically different than that of a white, upper-class community in New York. "Normality" is quite a different thing among these heterogeneous groups. For these black and Hispanic communities, everyday needs cannot normally be met or may only be met with great difficulty. Society's poor and disadvantaged find in the "normal" state of affairs an inability to meet basic needs. As mentioned previously, demographic shifts in the United States have resulted in greater population exposure and increasing vulnerability to disasters. Population growth and urbanization also affect the environment (e.g., deforestation, coastal erosion, etc.) thus magnifying the impact of atmospheric events, such as hurricanes and floods. Vulnerability and Population Distribution and Composition Population Growth and Urbanization Perrow (2007) has highlighted the ecological problem of population concentration in high-risk areas. Attention has been increasingly paid to changing demographic patterns and their impact on disaster vulnerability, as well as the impact and consequences of disasters as a result of population movements and settlements in regions with greater exposure to disasters (e.g., coastal communities and flood prone regions, among others). The changing size and structure of human populations across the United States may create new risks and worsen those already present. In the past 40 years, the U.S. population experienced significant growth, reaching the 300 million mark in 2007 (see Figure 1 ). Increasing population numbers as well as population density (see Figure 2) have transformed

5 Figure 1. U.S. Population Growth, Population Composition, Migration and Inequality ,000,000 - ^- ë 250,000,000 3 P 1 200, ,000, ') Year Source: both infrastructure (e.g., urbanization) and natural landscapes (e.g., deforestation, soil erosion, etc.). For example, since the 195ÜS, there has been approximately a 10 percent increase in the number of people living in urban areas in the United States, The past few decades, however, reveal only a limited cross section of a larger and mere sustained trend in U.S. urbanization. In the 18"" century, nearly 100 percent of the U.S. population lived in rural areas; now, nearly 80 percent of the population resides in urban areas (U.S. Census n.d.; U.S. Census 2007). Of particular concern are urban settings, which give rise to a host of vulnerabilities. Naturally, rural communities face serious threats attributable to "remoteness" resulting oftentimes in limited Hccess to a variety of state of the art technological services (e.g., "digital ine quality;" see Marks 2006). Cities, however, experience unique vulnerabilities rarely encountered in smaller communities: limited escape routes, dense infrastructure and high levels of poverty, to name a few. The physical structure of urban settings alone is dangerous and remains a fundamental human ecological source of vulnerability. Urban areas endure greater flood risks due to impermeable surfaces, reduction in the carrying capacity of river channels due to construction, and the inability of sewage systems to deal with massive urban-generated runoff (Smith 1995). Kumagai and Nojima (1999:67) reached similar conclusions in their observation of reduced run-off capacity in Tokyo, noting increasing flooding in upland tributary valleys as a result of "a well-known relationship between urbanization and increased flooding aï permeable undeveloped soils are replaced by impermeable roads, parking lots and buildings."

6 Social Forces S7 (2) Figure 2. Population Density Changes in the United States, Year Note: U.S. Census Population, Housing Units, Area Measurements, and Density, ( Epidemiological studies on earthquakes appear to support this argument. In China, researchers predicted that cities are more likely to experience deaths and injuries than towns, and towns more deaths and injuries than rural areas (Quijia, Xinling, Guomin and Rouwei 1992}. The Loma Prieta (1989) earthquake in the San Francisco Bay area and the Northridge Earthquake (1994) in Los Angeles, among others, serve as clear examples of the devastating impacts of earthquakes in urban areas with large population settlements and high population density. It is noteworthy that larger populations not only result in greater numbers of individuals being exposed to disasters, but when viewed within an ecological context, also create the conditions for greater exposure. For example, population changes also affect the environment (e.g., deforestation, destruction of mangroves or expanding the coastal zone to areas which were once occupied by water), which in turn, increases the likelihood of disasters. Taking this argument a step further, some researchers suggest that areas of high population density - particularly urban areas - are more likely to experience changes in meteorological conditions thus increasing the likelihood of severe storms such as tornadoes (Elsom and Meaden 1982; Snider 1977).

7 Population changes in Coastal Communities Population Composition, Migration and Inequalit)' 1095 Coastal development is another aspect of vulnerability that falls under the condition of human ecology. In the past few decades, the populations of coastal megacities have skyrocketed (Nichclls 1995). The increasing movement of populations towards coastal regions places populations living in such areas at greater risk to a host of natural disasters (Klein, Nicholls and Thomalla 2003). Despite these risks, Miller ( ) indicates that "most oceanfront property ownors have to be forced off their property by forces of nature that destroy the land and its buildability before they will leave." (author's emphasis) Even relocation or demolition benefits paid to owners (such as that mandnted by the 1987 Upton- Jones Amendnnent) in many cases do not "overcome individual, market, and regulatory incentives for oceanfront owners to remain in place," (Miller 1990:20) When the state requires real estate agents to inform buyers about potential hazards, this information has only minor effects on the decision to purchase (Palm 1981). But this is only part of the problem: some simplyca/7a7of leave dangerous coastal areas. Hurricanes Andrew and Katrina, as well as the Indian Oce.an Tsunami, illustrate the consequences of living in coastal areas, particularly for poverty-stricken communities (see Rodriguez, Wachtendorf, Kendra and Trainor 2006). There are some sobering facts about U.S. coastal populations that need to be highlighted, particularly given the implications that they have in the context of disasters, risk and vulnerability. NOAAs (2004) report on coastal population trends in the United Stales shows that: Coastal counties include 17 percent of the land area in the United States (excluding Alaska) In 2003, 53 percent (153 millioti people) Americans lived in coastal areas From 1980 to 2003, coastal populations increased by 28 percent Ten of the 15 cities with the highest population counts are in coastal counties Population density in coastal counties is about 300 persons per square mile compared to 98 for the whole of the United States Of the 25 coastal counties with the highest population density, 21 are in the Northeast The number of vehicles in coastal counties increased by 25.5 million or 43 percent from 1980 through 2000

8 1096. Social Forces S7(2) The significant increase in coastal populations, not only in the United States but internationally as well, has occurred despite coastal regions suffering more significant risks than other regions, from storm surges, extreme winds and floods, among others. The salination of surface waters, the result of coastal flooding, also poses significant health risks to local populations. Coastal erosion, a result of unsustainable development, places local populations at greater risk. According to recent research on coastal erosion, "over 80% of the North Atlantic. Lower Mississippi, and California shorelines were found to be significantly eroding," a crucial threat to the economic security of regions in which coastal business constitutes a major force in the local economy (Platt et al, 1992:6). While mitigation techniques have reduced the potential damage caused by flooding, these nevertheless may create a false sense of security that comes with not having experienced a disaster for some time. For instance, in response to massive erosion in South Central Alaska, many have argued for the installation of a seawall to prevent flooding. Responding to this, however. Mason (1997:203) cautions that "[the seawall] may merely transfer the effects of wave action from one part of the bluffed coastline to the other" and will increase the rate of shoreline erosion, thus resulting in greater exposure to disasters. The Role of Migration and Culture Migrants might be more vulnerable to disasters due to a lack of economic, cultural and social capital. For example, many Hispanic immigrants may lack a crucial form of capital conferred by language, which continues to be problematic in tornado-prone areas where authorities may or will not issue warnings in Spanish (Donner 2007; Aguirre 1988). What is more, difficulties associated with relocation and adaptation to a new country may be more salient than the disaster risks encountered in the new region. New residents - documented or not - often face everyday challenges (e.g., finding work, housing, and so forth) that outweigh those of less routine events, such as earthquakes, floods or tornadoes. As new immigrants without strong social capital or community networks, overcoming these problems is extremely difficult. Even if immigrants do perceive that they are at risk as a result of a hazard event, they may be reluctant to seek help, particularly if they are undocumented immigrants. Enarson and Morrow (2000) indicate that immigrants and minorities in Florida, following Hurricane Andrew, often distrusted the aid workers and uniformed immigration officers that offered aid, most likely for fear of deportation, These perceptions of distrust and helplessness are further exacerbated by recent announcements by public officials in Rio Grande Valley, Texas that U.S. Customs and border patrol agents will pre-screen hurricane evacuees for citizenship.

9 Population Composition, Migration and Inequality 1097 Figure 3. Primary Language Spoken at Home in the United States Only English Spanish Other Indo- European* Asian Languiige* Other Source: Finaiiy, value differences betv^een nninorlty groups and institutions may also be a salient factor in this process. Intriguing is Anderson's (1968) analysis of the flooding of the Rio Grande, where he sees the cultural prioritization of honor and self-sufficiency among Mexican communities as blocking outside assistance from Anglo communities. Following the 1999 Loma Prieta earthquake, Phillips (1993:102) offered a striking example of the manner in which prior experiences - largely shaped by one's culture - affected the sheltering behavior of victims. After erecting a temporary shelter of fenced off tents, it was discovered that "Central American refugee families apparently found this image terrifying. Immigrants who had fled military and government-backed death squads in their native countries now faced similar imagery after the disaster" This all means that without an understanding of the cultural backgrounds of disaster victims, relief efforts are bound to fail. Poverty and Disasters In a comprehensive review of the literature, Fothergill and Peek (2004) show that poverty influences nearly all stages of the disaster process - including risk perception, preparedness, warning communication, physical and psychologicai impacts, emergency response, recovery, and reconstruction. Most striking is the overwhelming consensus that low-

10 1098. Sociai Forces»7(2) cost, affordable housing is more vulnerable to disasters - precisely the type of housing in which the poor are likely to reside. Examining flooding events in the United States and Scotland, Enarson and Fordham (2001:47) conclude, "Structural factors increasing women's poverty and economic insecurity placed women at higher risk than men in these floods. Like race and ethnicity, class differences among women also positioned women differently to withstand the material losses of a flood." Excessive scarcities of economic capital intersected with human ecological factors to produce Hurricane Katrina's devastating outcomes. At the time, a triune of vulnerability - race, poverty and geographic location - would set the stage for one of the worst catastrophes in American history. New Orleans has historically been home to a large black community that, according to the U.S. Census, made up nearly 70 percent of the city's total population prior to Hurricane Katrina. At the same time, the city was well above the national average for individual and family poverty rates. According to a report issued by The Urban Institute (2005), at the time Katrina struck. New Orleans bore a poverty rate nearly 16 percent greater than the national average together with one of the poorest black populations in the nation. That 91 percent of biack families in New Orleans lived in poverty prior to Katrina is indeed a stark indicator of the hardships a majority of the city's population faced (Wisner and Walker n.d.). These features would come to the forefront during all stages of the disaster: more than 80 percent of the city was flooded; there were more than 1,500 deaths; and parts of the city were re-flooded later by Hurricane Rita, which blocked repopulation efforts. Prior to the onset of Katrina, the city boasted a population of nearly 484,674; after Katrina. this figure was reduced by approximately two-thirds to about 181,400 as of Jan. 28, 2006 (Common Ground n.d.). Of those who remained, a majority were black often lacking the means to evacuate or sustain themselves during longterm evacuation. In the following months, the city began to recover, albeit very slowly and with limited success. The relationship between economic capital and human ecology is often a cyclical one: there exists a self-sustaining relationship between land use, poverty and subsistence. Wisner et al. (2004:57) offer an equally instructive and stark example of the challenges squatter settlements face as itinerant residents in Bangladeshi floodplains: The former landless people who inhabit this depression are there because of its proximity to Dhaka's vegetable market. Already the chain of explanation of their vulnerability can be seen at work: rural people who are landless have few alternatives, and many seek the economic opportunity provided by the urban vegetable

11 Population Composition, Migralion and Inequality 1099 market. But this means living in an unsafe location. As newcomers, and extremely poor, the squatters in these low-lying areas had no access to i he structures of power that control marketing. They also had insecure.; title to land in the depression, and therefore no access to credit to allow them to increase their productivity and compete with better-established market gardeners (A.! Ali 1987). This situation meant that they had to grow rice rather than vegetables on their land, and thus the poor were forced into low-income pursuits. The above passage illustrates that poor populations (whether in Bangladesh or New Orleans) are often given little choice about where to live. In many cases, where people live is a measure of how little society values them, leading to unbearable immobility and poverty. Racial and Ethnic Minorities The United States is rapidly becoming a more diverse society as a result of a gradual increase in the proportion of the minoiity group population (see Figure 4), which, historically, has faced significant barriers to education and higher rates of poverty (see Figure 5). It is noteworthy that, between July 2005 and July 2006, about one of every two people added to the U.S. population were Latinos, reaching a total of 44.3 million Latinos. Moreover, Latinos have become the largest and fastest growing minority group in the United States and they will continue to increase at a fairly rapid rate as a result of migration and fertility patterns (see Rodriguez, Saenz and Menjivar 2008). However, beyond the grasp of many minonty communities are forms of cultural and economic capital enjoyed by pow^îrful social groups. Recent evidence also suggests that racial and ethnic minorities tend to hold lower trust in public institutions when compared to non- Hispanic whites (Barnshaw 2006). In further support of this argument, other research documents racial and ethnic differences in confidence in health care institutions (Boulware et al. 2003) and insurance settlements (Peacock and Girard 2000), Accordingly, it is reasonable to hypothesize that these groups may not be willing to seek help beyond a limited informal community when confronted with disasters (see for example. Nelson and Perry 1991; Pern/and Mushkatel 1986), Persistent problems with chronic poverty, unemployment or underemployment, as well as their experiences with discrimination and racism in the United Stages, adversely impact the trust of minority groups in institutions that are oesigned to provide them with assistance both generally and in times of disasters specifically. The complexities of the insurance industry, moreover, may be impenetrable to

12 1100. Social Forces S7(2) Figure 4. Growth of Minority Populations in the United States, % Black Hispanic Asian Native American Source; Census Summary Files 1 and 3 (SFl and SF3) Figure 5. Persons Living Below the Poverty Threshold by Race, % I White (not Hispanic) Black Asian Hispanic origin (any race) Race Source: US. Census. "Income, Poverty, and Health Insurance: Coverage in the United States: 2006." those without sufficient education. Affordability may also be a barrier to minority groups in acquiring insurance. Indeed, minorities are less likely to be insured. Following a disaster, health insurance is oftentimes necessary to deal with the acute and chronic health issues victims suffer. However, minority groups are far more likely to be uninsured or underinsured relative to their white counterparts. For example, in 2005, 11.3 percent of whites were uninsured compared to 19.3 percent for blacks and 32.7 percent for Hispanics. When one considers that undocumented immigrants are not included in these figures, the disparities become even more

13 Population Composition, Migralinn and Inequalít)' UOl Striking (DeNavis-Walt, Proctor and Lee 2006). C5iven these differences in health coverage, it follows that minorities, on average, will face greater difficulties dealing with the health and economic costs of disasters, leading to slower rates of recovery. The specific manner in which minority communities are organized may also place them at risk. Gladwin and Peacock i2000} show that family size is indirectly related to the likelihood of evacuation: that is to say, larger families will tend not to evacuate or take longer to do so. Recent population trend surveys show that while overall family size for all groups seems to be on the decline, minority groups in the United States tend to have, on average, larger households in comparison to whites (Population Reference Bureau 2003). For Hispanics, the average household size was about 3.5 in 2000 compared to 2.5 for whites, It is important to highlight that larger average family sizes among Hispanics is both a function of higher fertility rates among some Latino-subgroups in the United States, but also a function of migration patterns to the mainland. Despite their origins, these structural differences impact vl.lnerability to disasters, particularly if the nature of the disaster requires rapid evacuation. We might seek answers to the question of minority vulnerability at the level of cognition as well. Minority and majority groups recognize and interpret risk differently (see Perry and Greene 1982; Slovic, 2000; and Rodriguez, Diaz, Santos and Aguirre 2006). At the level of social psychology, Vaughan and Nordenstam (1991) reveal differences in how groups perceive risk. Likewise, Perry, Lindell and Green (1982) show that the likelihood of response varies between different groups. What seems more troubling, however, is that weather forecast information delivery systems are primarily oriented to powerful groups, not necessarily minorities (NRC 1999:86). Lending support to this notion is Mitchem's (2003) research finding that blacks often failed to grasp the significance of tornado warning messages. In many cases, warning systems are simply not attuned to the cultural and social demands of minority groups. Behaviorally, racial and ethnic minorities appe.ir less likely to evacuate in times of disasters {Gladwin and Peacock 2000; Riad, Norris and Ruback 1999). Moreover, a study on tropical Storm Allison reveals that racial minorities generally live in "relatively flood-prone landscapes with low base-flood elevation levels relative to Caucasians and people of higher socio-economic status." (Adeola 2003) Here, as with poverty, vulnerability rests at the intersection of race and ecological CDnditions. Women, Single-Mother Households and Disasters The social science disaster and risk literature suggests that women are more likely to take warnings or threat notifications more seriously relative

14 1102. Sociai forces 87(2) Figure 6. Single-Mother Households in the United States, % Source: U.S. Census Summary Files 1 and 3 (SFl and SF3). to men (see Slovic 2000; O'Brien and Atchison 1998), This might lead us to conclude that women are less vulnerable to disasters as a heightened sensitivity to risk should, in all likelihood, result in a response to v\/arnings, such as seeking shelter or evacuating. Scholars, however, have painted a much more complex picture of the problem in the recognition that age, race, ethnicity and socio-economic status may also shape perceptions of risk and their capacity to take protective action. Demographic data, for example, show that single mothers tend to be less educated and poorer than the general population, thus placing them at greater risk to disasters. A greater psychological sensitivity to risk means little if one lacks the cultural, social and economic capital to act on that sensation. Although the number of female headed households has continued to increase in the United States (see Figure 6), minority females are more likely to be female householders. In 2000, approximately 45 percent of black households were headed by single mothers compared to 22 percent for Latino ^and 12.5 percent for white households (Saenz 2004). Furthermore, although women in general may be more likely to perceive situations as risky, such risks may become irrelevant or less pressing relative to the burdens of single motherhood and other immediate needs, such as providing food, shelter and other basic necessities to their families. Like race and ethnicity, class differences among women also position them differently to withstand the material losses of a flood and rebuild their homes and engage in daily routines after these major community floods.

15 PopulalionCümposítion, Migration and Inequality 1103 Morrow and Enarson (1996} observed, during the Ic ng4erm recovery period following Hurricane Andrew, that poor and minority women experienced significant difficulties in accessing relief and recovery aid given that policy programs were reportediy set up oniy with small, nuclear househoids in mind. Moreover, Morrow and Enarson argue that women were victims of exploitation and fraud during the recovery period, in another study, Donner (2003) shows that census tracts in the United States with iarger percentages of singie mothers are more likely to experience a tornado death and/or injury. Rodriguez and Russeli (2006) also report that women and children were disproportionately affected by the 2004 Indian Ocean Tsunami and were more iikely to be victims (e.g., fataiities) of this catastrophic event relative to their male counterparts. We should highlight that O'Brien and Atchison (1998:176) found that women are significantly more likely "to seek information about both earthquakes and aftershocks, to beiieve a damaging event was iikeiy to occur, to make their household safer, to develop an emergency response plan, and to prepare for aftershocks." However, we should not confuse mere perceptions or attitudes of risk with actual response behavior. Women - especiaiiy singie mothers - may indeîed share in common a greater likeiihood of perceiving risk but, as we have documented in this paper, the conversion of these perceptions into behavior is likely to be affected by socio-economic and demographic characteristics (e.g., ciass, race/ethnicity, and age, among others). As Indicated by Rodriguez and Russell (2006), women experience higher vulnerability to disasters as a consequence of gender-based inequalities that characterize societies throughout the worid, which iimit their access to important financiai, political and other societal resources. Age and Physical Disability Census data shows that the median age is steadiiy increasing (see Figure 7) in the United States and that the population ihat is 65 and oider aiso continues to increase (see Figure 8), a trend likely to continue as a result of low ievels of fertility and increasing iife expectancy. Moreover, the regions where many senior citizens are choosing to live generate serious concerns regarding disaster risk and vulnerability. For example, Florida alone has an elderly population of 16.8 percent, more than 4 percent higher than the national average (U.S. Census 2007). The net result is a rate of exposure far exceeding that previousiy experienced by a group aiready at risk. When we combine a vulnerable elderly population in high-risk areas, we create the potential for a major disaster or catastrophe, as the case of Hurricane Katrina iüustrates (see Hartman and Squires 2006; Brunsma, Overfelt and Picou 2007). Age and disability form what Collins (1975) calis

16 1104. Social Forces S7i2) "status communities" who share certain characteristics. In this case, these characteristics place them at great risk during a disaster. That seniors are more vulnerable to disasters is a proposition that is supported by a growing body of literature. Experimental research confirms that the elderly and disabled confront unique difficulties during periods of evacuation. In a laboratory simulation sponsored by the U.S. Department of Commerce and the U.S, Department of Health and Human Services, Pearson and Joost (1983) uncovered, through the analysis of several fire scenarios, that in common activities associated with evacuation (e.g., calling the fire department, searching for the young, and donning a robe), blind participants took 2.47 times longer and wheelchair participants took 2.36 times longer than non-disabled participants to complete these activities. Older adults in the second set of scenarios took 27 percent longer and individuals with arthritis took 44 percent longer to complete these activities relative to younger individuals and those without arthritis, respectively. Also, other studies show that physically disabled persons may have fewer opportunities to access personal medical items after a disaster (Rahimi and Azevedo 1993), Moreover, shelters aimed at providing essential services to evacuees following a disaster may not have the necessary resources to provide adequate care to individuals with severe disabilities or chronic illnesses, Also, persons with developmental disabilities may require special training and needs in cases of fire evacuation (Ramey-Smith and Fechter 1978). However, in an assessment of the long-term effects of the 1966 tornado on elderly residents of Topeka Kansas, Kilijanek and Drabek (1979) found that older victims were allotted a smaller portion of community resources and experience insurance problems with respect to recovery. A later study by Bolin and Klenow (1983) reproduced similar findings. These studies accurately reveal what one might expect: the elderly are more likely to be injured or die as a consequence of hazard events. Findings from a study on Japanese earthquakes demonstrated that "twenty-three percent of casualties were aged 75 years or older. Another 30 percent were between 60 and 74 years. In other words, more than half of the casualties {53 percent) were over 60." (Wisner 1998:27) More recently, medical records and survey analysis after the Northridge earthquake found that older people were more likely to be killed or hospitalized (Peek-Asa, Ramirez and Shoaf 2001). Residential fires are more iikely to kill seniors (Gulaid, Sacks and Sattin 1989), and severe weather is not an exception. Following the Birmingham, Alabama, killer tornados that occurred in April of 1998, Legates and Biddle (1999) observed that of the 32 victims, more than 40 percent were 66 years of age or older. Quick response research repeatedly demonstrates that the age of those killed and/or injured during a tornado event tends to be higher than those that were not injured or

17 Population Composition, Migration and Inequality 1105 Figure 7. Median Age of the U.S. Population, en < Source; U.S. Census Summary Files 1 and 3 (SFl and SF3). Figure 8. Percentage of Total Population 65 Years of Age and Older in the United States, Source; U.S. Census Summary Files 1 and 3 {SFl and SF3). killed (Schmidlin and King 1998; Schmidlin and King 1995). Finally, the case of Hurricane Katrina (see Hartman and Squires 2006) and even the 1995 Chicago heat wave show the vulnerability of the elderly to these types of disasters (see Klinenberg 2002); the majority of victims in this disaster were 65 or older (Whitman et al. 1997).

18 1106. Social Forces S7i2) Apart from physical disabilities, psychological factors may offer some insight into the observed vuinerabiiity among seniors. Previous studies have shov^n that older persons are less likely to personalize warnings (Hodge, Sharp and Marts 1981; Flynn 1979; Mack and Baker 1961) and they are also less likely to receive warning messages (Turner et al. 1979). The former may, in part, be due to their extended experience with false alarms and inaccurate warnings. With respect to the latter, the elderly may experience a significant loss of social ties often associated with the aging process. Unfortunately, little consensus is found in more recent studies as to whether the elderly are more or less likely to respond to warnings relative to their younger counterparts (Mileti and Darlingon 1997; Edwards 1993; Sorensen 1991; Mileti and O'Brien 1992). It appears that age seems to influence the probability of engaging in frequent kinship and community activities. The lack of community interactions or involvement may, in turn, place limits on the number of warnings people receive (Perry 1979) or respond to. Discussion and Conclusion Events such as Hurricane Andrews and, more recently. Hurricane Katrina, illustrate how disasters can be amplified by demographic and socioeconomic factors. The devastating results of these disasters - and, to be sure, any disaster - were a result of the intersection of poverty (in some cases, chronic poverty), decades of neglect and years of discrimination and racism, exacerbated by the changing composition of coastal communities in the region. Economic, cultural and social capita!, as well as human ecological trends, held equal hands in placing some population groups at greater risk. Theoretically, however, the problem is much more complex and, at times, vulnerabilities are a result of multiple forms of social capital. The forms of capital discussed earlier seem to interact and the distinctions, at times, seem more analytical than empirical. One such example involves poorer groups' relegation to at-risk areas - near levees, the inner cities and coastal industry. Living in such areas is an immediate source of risk; therefore, we might be inclined to trace this to human ecological dynamics. Capitalism and urbanization, driving the need for new industrial and commercial sectors, forces people into dangerous areas. But it is often the poor who live in such regions, making the problem as much an issue of human ecology as economic capital. It is also important to note that it is only in rare cases that a single dimension of the social structure is responsible for vulnerability. In the context of disasters, conflict and inequality are social forces that are multidimensional. For example, despite overwhelming support regarding the relationship between gender and risk perception, this evidence

19 Population Composition, Migration and Inequality 1107 becomes less clear when considered in the context of race (Johnson 2002; Flynn, Slovic and Mertz 1994). A poor black woman will face different - and arguably more severe challenges - than a middle class non-hispanic white woman during periods of disaster. The complexities of these interactions must be taken into account in future studies, possibly applying the concept of "intersectionality" in such analyses (see Andersen and Hill Collins 2004). Vulnerability is a multifaceted phenomenon. As such, solutions, too, must be multifaceted, addressing the r.ange of social, cultural, demographic and economic conditions - often interacting in complex ways - that culminate in population vulnerability. Population changes also require the frequent and thoughtful revision of disaster mitigation, preparedness and response policies and plans. As such, emergency managers, planners, and other policymakers should they endeavor to develop effective disaster policy, must keep their fingers on the demographic pulse of their communities. Failing rodo so will undoubtedly have significant and long-term deleterious public health effects as well as devastating sociai and economic consequences for the populations that are vulnerable to disasters and are living in high risk areas. Individual preparedness and response to disasters is iienerally influenced by factors that have little to do with the hazard agtînt or the disaster event itself, such as social class, education, gender, race/ethnicity, cultural background and language proficiency, among others. More than 50 years of disaster research in the- social sciences provide ample evidence on the intersection of these demographic and socioeconomic factors and disasters. It is, therefore, imperative that local and national officials engaged in the development cf disaster preparedness and mitigation policies, also take these factors into account if our aim is to ameliorate the devastating impacts of disasters on human populations. Population growth, composition and distribution will be the primary driving forces in the context of exposure to disasters and disaster vulnerability. As such, social science disaster researchers need to pay closer attention to population shifts and their implications for disaster risk and vulnerability. Notes 1. The terms "Latinos/as" and "Hispanics" will iie used interchangeably throughout this paper to denote the population oi Hispanic, Latin American, or Caribbean descent that reside in the United States. 2. We should note that important differences exist amc ng the Latino/a households in the United States in regards to single-mother households. For example, according to the 2000 U.S. Census, about 16 percent of Cuban households were headed by single females, compared to 17.5 percent for Mexican Americans, 37 percent for Puerto Ricans, and 39.5 percent for Dominicans.

20 1108. Socio/Forces 87(2) References Adeola, Francis "Flood Hazard Vulnerability: A Study of Tropical Storm Allison Flood impacts and Adaptation Modes in Louisiana." Quick Response Research Report #162. Boulder, CO: University of Colorado, Natural Hazards Research and Applications Information Center Aguirre, Benigno "The Lack of Warnings before the Saragosa Tornado." International journal of Mass Emergencies and Disasters 6(1 ): Andersen, Margaret, and Patricia Hill Collins. Editors Race, Class, and Gender: An Anthology. Wadsworth, Barnshaw, John "Beyond Disaster: Locating Hurricane Katrina within an Inequality Context." Pp Learning from Catastrophe: Quick Response Research in the Wake of Hurricane Katrina. Boulder, CO: Institute of Behavioral Science, University of Colorado. Bolin, Robert, and Daniel J. Klenow "Response of the Elderly to Disaster: An Age- stratified Analysis." International Journal of Aging and Human Development 16(4): Boulware, L. EBony, Lisa A. Cooper. Lloyd E. Ratner. Thomas A. LaVeist and Neil R. Powe "Race and Trust in the Healthcare System." Public Health Reports n8(july-august): Bourdieu, Pierre "The Forms of Capital." Pp J.G. Richardson, editor The Handbook of Theory: Research for the Sociology of Education. Greenwood Press. Brunsma, David L., David Overfelt and Steven J. Picou. Editors The Sociology of Katrina: Perspectives on a Modern Catastrophe. Rowman b Littlefield. CNN "Nursing Home Owners Face Charges." Accessed Oct. 4, 2007 at: Collins, Randall Conflict Sociology. Academic Press. Common Ground, n.d. "New Orleans in Numbers." Accessed April 10, 2007 at: DeNavis-Walt, Carmen, Bernadette D. Proctor and Cheryl H. Lee "Income, Poverty, and Health Insurance Coverage in the United States: 2005." Current Population Reports: Consumer Income. \NasW\r\QXor\. DC: U.S. Department of Commerce, Economics and Statistics Administration, U.S. Census Bureau. Department of Health and Human Services "Nursing Home Emergency Preparedness and Response During Recent Hurricanes." Washington, D.C: Office of the Inspector General. Accessed Nov. 7, 2008 at: hhs.gov/oei/reports/oei pdf.

21 Population Composition, M igration and Inequality 1109 Donner, William R, An Integrated Model of Risk Perception and Protective Action. Doctoral Dissertation. Newark, DE: University of Delaware, Disaster Research Center The Human Ecological Model and Its Application to Severe Weather Phenomena: Predicting Tornado Morbidity. Masters Thesis, Newark, DE: University of Delaware, Disaster Research Center Duncan, Dudley "Human Ecology and Population Studies." Pp The Study of Population. PH. Häuser and O.T Duncan, editors. University of Chicago Press. Edwards, Margie L "Social Location and Self-Protective Behavior: Implications for Earthquake Preparedness." Internationa! Journal of Mass Emergencies and Disasters 11 (3): Elsom, Derek M., and G. Terence Meaden "Suppression and Dissipation of Weak Tornadoes in Metropolitan Areas." Monthly Weather Review n0(7): Enarson, Elaine, and Maureen Fordham "Lines that Divide, Ties that Bind: Race, Ciass, and Gender in Women's Flood Recovery in the US and UK." Australian Journal of Emergency Management 15(4): Enarson, Elaine, and Betty H. Morrow "A Gendered Perspective: The Voices of Women." Pp Hurricane Andrew: Ethnicity. Gender, and The Sociology of Disasters. W.G. Peacock, B.H. Morrow and H. Gladwin, editors. Miami, FL: International Hurricane Center, Laboratory for Social and Behavioral Research. Flynn. Cynthia B "Three Mile Island Telephone Survey: Preliminar/ Report on Procedures and Findings." Tempe, AZ: Mountain West Research. Flynn, James, Paul Slovic and C.K. Mertz "Gender, Race, and Perception of Environmental Health Risks." Risk Analysis 14(6»: Fritz, Charles E "Disasters." Pp Contemporary Social Problems. R.K. Merton and R.A. Nisbet, editors. Harcourt. Fothergill, Alice, and Lori Peek "Poverty and Disasters in the United States: A Review of Recent Sociological Findings." Natural Hazards 32(1): Gladwin, Hugh, and Walter G. Peacock "Warning] and Evacuation: A Night for Hard Houses." Pp in Hurricane Andrew. W.G. Peacock, B.H. Morrow and H. Gladwin, editors. Miami. FL: International Hurricane Center Gulaid. Jama, Jeffry J. Sacks and Richard W. Sattin. 1989, "Deaths from Residential Fires among Older People, United States, 1984." Journal of the American Geriatric Society 37(4): Hartman, Chester and Gregory D. Squires, Editors There is no Such Thing as a Natural Disaster: Race, Class, and Hurricane K,itrina. Routledge.

22 1110. Socía/Forces 87(2) Hawley, Amos H i-iuman Ecoiogy: A Ttieory of Community Structure. New York: The Ronald Press Company. Hewitt, Kenneth. Editor interpretations of Calamity from the Viewpoint of Human Ecoiogy. Allen and Unwin. Hodge, David. Virginia Sharp and Marion Marts "Contemporary Responses to Volcanism; Case Studies From the Cascades and Hawaii." Pp Voicanic Activity and Human Ecoiogy. PD. Sheets and U.K. Grayson, editors. Academic Press. Johnson, Branden B "Gender and Race in Beliefs about Outdoor Air Pollution." Risi( Analysis 22(4):725-38, Klein, Robert J.I, Robert J. Nicholls and Frank Thomalla. 2003, "The Resilience of Coastal Megacities to Weather-Related Hazards." Pp Buiiding Safer Cities: The Future of Disaster Risi<. A. Krcimer, M. Arnold and A. Carlin. Washington, DC: The World Bank Disaster Management Facility. Klinenberg, Eric Heat Wave. University of Chicago Press. Kilijanek, Thomas S., and Thomas E. Drabek 'Assessing Long-Term Impacts of a Natural Disaster: A Focus on the Elderly." The Gerontoiogist 19(6): Kumagai, Yoshio, and Yoshiteru Nojima "Urbanization and Disaster Mitigation in Tokyo." Pp Crucibles of Hazard: Mega-Cities and Disasters in Transition. J.K. Mitchell, editor. Shibuya-ku, Tokyo: United Nations University Press. Legates, David R., and Matthew D. Biddle "Warning Response and Risk Behavior in the Oak Grove-Birmingham, Alabama. Tornado of 08 April 1998." Quick Response Report #116. Boulder. CO: Natural Hazards Center, University of Colorado. Mack, Raymond W., and George W. Baker "The Occasion Instant: The Structure of Social Responses to Repeated Air Raid Wamings." Publication No Washington. DC: National Academy of Sciences. Mason, Owen K., William J. Neal and Orrin H. Pilkey Living with the Coast ofaiaska. Duke University Press. Marks, Daniel "Digital Inequality and the Implementation of New Technologies: Problems with Technological Diffusion among Oklahoma Emergency Managers." Preliminary Papers 354. Newark, DE: University of Delaware, Disaster Research Center. Mileti, Dennis S., and JoAnne D. Darlington, "The Roie of Searching in Shaping Reactions to Earthquake Risk Information." Soc/a/Proô/ems 44(1 ):89-l03.

23 Population Composition, Migration and Inequality 1111 Mileti, Dennis S., and Paul W. O'Brien "Warnings During Disaster: Normalizing Communicated Risk." Soc/a/Proö/e^'5 39(1 ): Miller, H. Crane "On the Brink: Coastal Location and Relocation Choices." Unpublished Manuscript. Washington, D.C. Mitchem, Jamie D "An Analysis of the September 20, 2002, Indianapolis Tornado: Public Response to a Tornado Warning and Damage Assessment Difficulties." Ouick Response Report #161. Boulder, CO: Natural Hazards Research and Applications Information Center, University of Colorado., Morrow, Betty H., and Elaine Enarson "Hurricane Andrew Through Women's Eyes: Issues and Recommendations." International Journal of Mass Emergencies and Disasters 14( 1 ) : National Research Council Making Climate Forecasts Matter. RC. Stern and W.E. Easterling, Editors. National Academy Press Nicholls, Robert J, "Coastal Megacities and Climate Change," GeoJournal 37(31: O'Brien, Paul W., and Patricia Atchison "Gender Differentiation and Aftershock Warning Response." Pp, , The Gendered Terrain of Disaster. E. Enarson and B.H. Morrow, editors. Praeger Publishing. Palm, Risa "Public Response to Earthquake Ha, ard informaron." Annals of the Association of American Geographers 71(3}: Peacock, Walter G., and Christopher Girard "Ethnic and Racial Inequalities in Hurricane Damage and Insurance Settlements." Pp Hurricane Andrew: Ethnicity, Gender, and The Sociology of Disasters. W.G. Peacock, B.H. Morrow and H. Gladwin, editors. Miami, FL: international Hurricane Center. Laboratory for Social and Behavioral Research. Peek-Asa Corinne., Marizen Ramirez and Kim Shoaf "Population-Based Casecontrol Study of Earthquake Related Deaths and Hospitalized Admissions Sustained During the 1994 Northridge CA Earthquake." Proceedings First Workshop for Comparative Study on Urban Earthquake Disaster Management. Available at: httpv/ Ramirez.pdf. Pearson, Richard G., and Michael G. Joost "Egress Behavior Response Times of Handicapped and Elderly Subiects to Simulated Residential Fire Situations." Washington, DC: U.S. Department of Commerce and U.S, Department of Health and Human Services. Perrow, Charles The Next Catastrophe. Princeton University Press. Perry, Ronald W "Evacuation Decision-Making \n Natural Disasters." Mass Emergencies 4(1 }:25-38.

24 1112. Sociai Forces 87{2) Perry, Ronald W., and Alvin H. Mushkatel Minority Citizens in Disaster. University of Georgia Press. Perry, Ronald W., and Marjorie R. Greene "The Role of Ethnicity in the Emergency Decision-Making Process." Sociological Inquiry 52(4): Perry, Ronald W., Michael K. Lindell and Marjorie R. Greene "Threat Perception and Public Response to Volcano Hazard." Journal of Social Psychology 116(2): Perry, Ronald W., and Lisa S. Nelson "Ethnicity and Hazard information Dissemination." Environmental Management 15(4): Phillips, Brenda "Cultural Diversity in Disasters: Sheltering. Housing, and LongTerm y." IntemationalJournal of Mass Emergency and Disasters 11(1):99-110, Platt, Rutherford, H. Crane Miller, Timothy Beatley, Jennifer Melvilie and Brenda Mathenia Coastal Erosion: Has the Retreat Sounded? Boulder, CO: Natural Hazards Center, University of Colorado. Population Reference Bureau Number of People per Household, by Race/ Ethnicity, Accessed Sept. 28, 2007 at: WhileHouseholdsContractHomes Expand.pdf. Rahimi, Mansour, and Gienn Azevedo "Building Content Hazards and Behavior of Mobility-Restricted Residents." Pp The Loma Prieta California. Earthquake of October 17, Public fíesponse. Boiton, PA: U.S. Geological Survey Professional Paper 1153-B. Ramey-Smlth, A.M.,andV Fletcher "Group Homes for the Deveiopmentally Disabled." Washington, DC: Center for Consumer Product Technology, National Bureau of Standards. Riad, Jasmin K., Fran H. Norris and R. Barry Ruback "Predicting Evacuation in Two Major Disasters: Risk Perception, Social Influence, and Access to Resources." Journal of Applied Social Psychology 29(5): Rodriguez, Havidán, Rogeiio Sáenz and Cecilia Menjivar. Editors Latino/as in the United States: Changing the Face of América. Springer. Rodriguez, Havidán, and Carla N. Russell "Understanding Disasters: Vulnerability, Sustainable Development, and Resiliency." Pp Public Sociologies Reader Judith Blau and Keri lyall-smith, editors. Rowman & Littlefield. Rodriguez, Havidán, Walter Diaz, Jenniffer Santos and Benigno Aguirre "Communicating Risk and Uncertainty: Science, Technology, and Disasters at the Crossroads." Pp Handbook of Disaster Research. H. Rodriguez, E.L. Quarantelli and R. Dynes, editors. Springer.

25 PopiUation Composition, Migration and Inequality «1113 Rodriguez, Havidán, Tricia Wachtendorf, James Kendra and Joseph Trainor "A Snapshot of the 2004 Indian Ocean Tsunami: Societal Impacts and Consequences." Disaster Prevention and Management 15(1): Saenz, Rogelio, "Latinos and the Changing Face of America," The American People. Census Russell Sage Foundation and The Population Reference Bureau. Schmidiin, Thomas W., and Paul S. King "Risk Factors for Death in the February 1998 Florida Tornadoes." Quick Response Report ^106. Boulder, CO: University of Colorado, Natural Hazards Center. Schmidlin, Thomas W., and Paul S. King "Risk Factors for Death in the 27 March 1994 Georgia and Alabama Tornados." Disasters 19(2): Slovic, Paul "Trust, Emotion, Sex, Poiitics and Sciences: Surveying the Risk-Assessment Battlefield." Pp The Perception of Risk. R Slovic, editor. Earthscan Publications Ltd. Smith, Keith. ^9Q2. Environmental Hazards: Assessing Risk and Reducing Disaster. Routiedge. Snider, C,R "A Look at Michigan Tornado Statistics." Monthly Weather fîewew105(10): Sorensen, John H "When Shall We Leave? F.îctors Affecting the Timing of Evacuation Departures," International Journal of Mass Emergencies and Disasters Q(2)Ab3-6b. Turner, Ralph H.. and Killian, Lewis M Collective Behavior. Third Edition. Prentice-Hall, Inc. Turner, Ralph H., Joanne M, Nigg, Denise H. Paz and Barbara S. Young "Earthquake Threat: The Human Response in Southern California." Los Angeles, CA: University of California, institute for Social Science Research. Urban institute Katrina: Demographics of Disaster. Washington, D,C.: Office of Public Affairs, U.S. Census. N.d. "Table 4. Population: " Accessed Oct. 4, 2007 at: U.S, Census "The Foreign-Born Population in the United States: 2003." Current Population Reports by Luke J. Larson Available at: census.gov/prod/2004pubs/p pdf American Factfinder. Available at: home/saff/main.html?jang=en "State and County Ouick Facts: Florida." Accessed Nov. 21, 2007 at: 2OOO.html.