Compounding Disasters in Gulf Coast Communities 2020-2021: Impacts, Findings, and Lessons Learned (2024)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

2 Hazards, Exposure, Vulnerabilities, and Disaster Risk in the Gulf of Mexico Region Driven predominantly by use of fossil fuels and associated carbon emissions, climate change is a global existential threat to communities. Climate change drives slow- and rapid-onset changes in the physical environment, prompting disruptive and deleterious social changes with lasting health, community, and economic impacts. In short, climate change is the overarching threat multiplier for risks to population health, safety, and well-being in the Gulf of Mexico (GOM) region. Moreover, social structures and networks, including systemic behaviors, prejudice, racism, and institutionalized policies, result in inequitable distribution of climate change impacts, resources, and delivery of services (Singh et al., 2023). A range of social, economic, geographic, demographic, cultural, environmental, historical, political, and governance factors (i.e., social determinants of health; see Box 2-1) influence exposure and vulnerability to hazards, including those associated with climate change, and poor health outcomes. Relatedly, inequities in education, income, access to transportation, affordable and safe housing and/or neighborhoods, and health care and social services drive disproportionate vulnerability and exposure to disruptive events. These inequitable circumstances, particularly among communities of color, Indigenous populations, and those living in poverty and areas of historical socioeconomic disadvantage, increase disaster risk and constrain opportunities to expand adaptive capacity, community resilience, and health and prosperity, further elevating the risk of future disaster impacts. 45 PREPUBLICATION | UNCORRECTED PROOFS 

BOX 2-1 Social Determinants of Health in the Gulf of Mexico Region Nested within the variables of vulnerability and exposure are multiple social determinants of health, defined as the conditions in the environments where people are born, live, learn, work, play, worship, and age that affect a wide range of health, functioning, and quality-of-life outcomes and risks (Office of Disease Prevention and Health Promotion, 2023). The U.S. Department of Health and Human Services groups social determinants of health into five domains: economic stability, education access and quality, health care access and quality, neighborhood and built environment, and social and community context (ODPHP, n.d., which represent dimensions of both vulnerability and exposure in the context of disaster risk. Baseline conditions for social determinants of health are generally poorer for Gulf of Mexico states than for the rest of the United States, with systematic disparities in mortality and other measures of well-being even across small areas that lie relatively close together (NASEM, 2017). These conditions worsen during disaster response and recovery, heightening inequities and sensitivity to subsequent disruptive events (Martin, 2015). The response and recovery phases of a disruptive event also elevate the risk of such outcomes as poverty, illness, mental health struggles, housing instability, social isolation, death, and exposure to threats or violence, limiting opportunities to recover effectively (Drabek, 2007; Smiley et al., 2018). Adverse social determinants of health thus can render communities unevenly and profoundly exposed and vulnerable to adverse impacts and heightened challenges at every disaster stage (Bikomeye et al., 2021). Population exposure to disasters has been documented as producing a range of consequences for physical and mental health, as well as the environment. Disasters impact public health by causing damage, destruction, disruption of health care services, displacement, debility, disease, and death. They can exacerbate preexisting mental health conditions for individuals, but on a broader population level, contribute to stress and distress that may lead in some cases to disaster event–related, new-onset mental disorders. When compounding disasters occur, public health impacts may be amplified by simultaneous or sequential exposure to harmful impacts along with the experience of loss and disruption of life in the aftermath of the events (Hahn et al., 2022; Leppold et al., 2022; Wells et al., 2022). 46 PREPUBLICATION | UNCORRECTED PROOFS 

Against this backdrop, this chapter explores the drivers of disaster risk in the GOM region: hazards, exposure, and vulnerabilities beginning with a brief background on climate attribution science and a description of the natural and technological hazards that contributed to compounding impacts in the GOM in 2020–2021. Next, it provides an overview of exposure associated with land use patterns, past and current, in the GOM region, that have placed populations, infrastructure, housing, production capacities, and other tangible human assets in hazard-prone areas. Finally, it describes extant physical and social vulnerabilities in the GOM region wherein disaster impacts are “piled on” and disaster risk is exacerbated. The broad impacts of disasters on public health are discussed as well as the disproportionate risks social vulnerability has created for population health in the GOM region. Impacts to health system performance and the psychological footprint of a disaster follow, with a discussion of medically vulnerable patient populations. Social capital and cohesion is also discussed, followed by GOM public infrastructure and building inventories. Next is a discussion on housing’s role in vulnerability, including the U.S. legacy of racial discrimination in homeownership opportunities; displacement following disasters; and disproportionate impacts on renters, rental homes, and manufactured housing units. Disasters’ impacts on the changing insurance market is followed by a discussion of economic instability and inequities before and during the period of a presidentially declared disaster. HAZARDS Hazards in the GOM region include extreme weather-climate events, land subsidence, and technological hazards related to fossil fuels, all of which were compounded during 2020– 2021 by the COVID-19 pandemic. Climatologically, the GOM region is susceptible to a variety of hazards, including winter storms, wildfires, drought, and extreme temperatures; however, hurricanes (or tropical cyclones), flooding, and severe local storms (including tornadoes) are the most common weather threats. Hurricanes produce significant impacts on coastal and inland communities, including such hazards as storm surge, heavy rainfall, inland flooding, and strong winds (Ashley and Ashley, 2008; Fussell et al., 2017; Muller and Stone, 2001; Rappaport, 2000). Storm surge occurs when hurricane winds push seawater onto shore, causing flooding, erosion, and property damage. 47 PREPUBLICATION | UNCORRECTED PROOFS 

Heavy rainfall can cause inland flooding, while strong winds can damage buildings, power lines, and other infrastructure (Henry et al., 2020). In addition, hurricanes can result in significant economic losses, including damage to crops, businesses, and transportation systems (Faber, 2015; Schmidt et al., 2009). Hurricanes and other hazards are occurring within a changing climate amid a dynamic demographic and economic landscape, driving numerous complex and interrelated challenges, ongoing community disruptions, and enduring adverse impacts. Attribution Science Attribution science attempts to disentangle the influence of human-caused climate change in individual extreme weather-climate events from other factors, such as the climate patterns El Niño and La Niña (NASEM, 2016). Attribution studies (Fowler et al., 2021; NASEM, 2016; van der Wiel et al., 2016) confirm that the influence of climate change is represented by contemporary heatwaves, heavy rainfall, consecutive dry days, droughts, tropical cyclones, and other weather-climate events. However, because the science is relatively new, 1 it is still an area of active research (NASEM, 2016), and attribution results are more robust for certain events (e.g., heatwaves) than for others (e.g., increased extreme precipitation). Recent studies suggest that the occurrence of rapidly intensifying and stronger tropical cyclones may be increasing as a result of climate change (Knutson et al., 2021; Walsh et al., 2016). In the Atlantic Basin, the number of major hurricanes (Category 3, 4, or 5) has increased, and there is some evidence that hurricanes are intensifying, slowing down, or stalling as they approach the coast (Hall and Kossin, 2019; Kossin, 2018; Walsh et al., 2016). Shepherd and colleagues (2007) document that rainstorms associated with lower-category hurricanes are the most prolific rain producers, with inland freshwater flooding being one of the deadliest aspects of landfalling tropical cyclones (Rappaport, 2014). Hurricane Harvey (in 2017) is a good example of how a “stalled” storm can produce disastrous flooding, winds, and other hazards. In Harris County, Texas, Smiley and colleagues (2022) showed that 30 to 50 percent of Harvey-flooded properties would not have flooded without climate change using climate 1 The first publication attempting to attribute an extreme weather-climate event to climate change was in 2004, analyzing the 2003 European summer heat wave (see Stott et al., 2004). 48 PREPUBLICATION | UNCORRECTED PROOFS 

attribution science paired with hydrological models, detailed land-parcel and census tract socioeconomic data. Witze (2018) demonstrates how climate warming is leading to “wetter” storms, which likely include tropical cyclones (and their remnants). The Clausius-Clapeyron relationship clearly establishes that a warmer atmosphere has more water vapor capacity (Martinkova and Kysely, 2020). Balaguru and colleagues (2018) and Bhatia and colleagues (2022) also found evidence of increasing trends in rapid intensification of tropical cyclones, a phenomenon that hampers the ability to prepare and evacuate adequately for landfalling storms. Some studies also suggest that climate change could lead to more polar vortex weakening events (Cohen et al., 2021), such as those that spurred Winter Storm Uri. Reed and colleagues (2022) demonstrated that human-induced climate change played a role in the full 2020 hurricane season by increasing storm rainfall rates and accumulated rainfall amounts for observed storms that were at least tropical storm strength. While climate is shaped by far more complex processes (Balaguru et al., 2023; IPCC, 2021; USGCRP, 2023), a simplification of the effects of climate change as related to wind, warming air, and water is offered here. As humans burn more fossil fuels, ocean and atmospheric temperatures grow warmer. The warmer the ocean temperature, the higher the volume of water and the more evaporation is available to the atmosphere. The warmer the atmosphere, the more moisture it can hold, and the more rain can be produced. More rain means more released heat, and more heat means stronger winds. Wind speed is nonlinearly related to potential structural damage. The damage potential from a Category 4 storm with 150 mph winds (the strength of Hurricane Laura at landfall) is 256 times greater than that from a Category 1 storm with 75 mph winds. Stronger storms also mean the potential for higher storm surge, due not only to stronger winds driving storm surge inland, but also to rising ocean temperatures that lead to an expansion of the sea’s volume and corresponding sea level rise (Barlow and Camargo, 2022). Despite the relative nascency of attribution science, valuable data can be gleaned from this research to inform future risk calculations pertaining to extreme weather events and their impacts on critical infrastructure, housing, food management, building codes, and insurance (NASEM, 2016). Studies have shown that these data can even be used to establish liability for adverse climate-related impacts for litigation purposes (Burger et al., 2021; Burton, 2010; CRS, 2023). 49 PREPUBLICATION | UNCORRECTED PROOFS 

Hurricanes and Tropical Storms (Tropical Cyclones) Since official records began in 1851, 216 hurricanes have made landfall between the Rio Grande River in Texas and the Florida Keys, 80 of which were rated as major hurricanes (NOAA, n.d.-a). Hurricanes and tropical storms have wide-ranging regional impacts, affecting both coastal regions and areas further inland. They are often rated on the well-known Saffir- Simpson wind scale, but an array of associated hazards is possible, including storm surge inundation, strong winds, tornadoes, intense rainfall, and pluvial (surface water) and fluvial (waterbody overflow) flooding (Andersen and Shepherd, 2013; Fussell et al., 2017; Rappaport, 2000). These hazards threaten life, safety, physical and mental health, private and public property, household incomes, residential stability, public infrastructure, and utilities. The GOM region houses vital energy, industrial, and transportation infrastructure. Wind damage, storm surge, and flooding compromise or disable power transmission, energy production, and road networks (Henry et al., 2020; Schmidt et al., 2009). Figure 2-1 shows the number of tropical cyclones per decade in each county and parish in the Southeast United States. Since 1980, the time between landfalling tropical storms in the GOM region has shortened (Xi and Lin, 2021) and this pattern may be escalating as a result of climate change (Knutson et al., 2010; Walsh et al., 2016). This phenomenon is particularly evident in the Atlantic basin, where the number of major hurricanes has increased in recent years (Walsh et al., 2016). Hurricane hazards at landfall (e.g., strong winds, heavy rainfall) are also expected to continue (Xi et al., 2023). Additionally, the dynamics of hurricanes are changing. Their rapid intensification reduces the time available for preparations and evacuations (Bhatia et al., 2022; Wang et al., 2019). Evidence also suggests that some hurricanes are slowing down, which results in increased rainfall and flooding (Kossin, 2018; Patricola and Wehner, 2018) and higher storm surge (Marsooli et al., 2019). These shifting climate conditions will likely contribute to continued changes in hazard characteristics: in combination with the heavily populated and built-up coast, an increase in disruptive events, subsequent compounding of disasters, and likely less time for disaster recovery before the next tropical cyclone makes landfall. 50 PREPUBLICATION | UNCORRECTED PROOFS 

FIGURE 2-1 Frequency of county/parish-level landfalling storms per decade since 1900: (A) tropical storms and hurricanes, (B) all hurricanes (Category 1–5), and (C) major hurricanes (Category 3–5). Saffir-Simpson scale ratings were determined by the storm intensity at landfall. SOURCE: Strader, 2023. 51 PREPUBLICATION | UNCORRECTED PROOFS 

Wind Hazards (Unrelated to Hurricanes) The GOM states are part of the Southeast, a region of the United States exposed to tornadoes and other wind hazards predominantly during periods of active thunderstorms in late spring, with a secondary peak in late fall. While the Central Plains are often referred to as “Tornado Alley,” recent studies have found that tornado activity is increasing in the Southeast (Gensini and Brooks, 2018), which experiences a greater frequency of significant (EF2+ [Enhanced Fujita Scale]) tornadoes, as well as a larger overall tornado damage footprint (Ashley and Strader, 2016; Dixon et al., 2011). Tornado-related fatalities are also more likely in the Southeast than in the rest of the continental United States because of an array of factors, including storm motion (Strader et al., 2022), built-environment density, the nocturnal nature of storms (Ashley, 2007; Strader and Ashley, 2018; Strader et al., 2022) and greater social and infrastructure vulnerability (Ash et al., 2020; Cutter et al., 2003; Strader and Ashley, 2018). While having a smaller damage footprint compared with landfalling tropical cyclones, tornadic storms kill a greater percentage of the exposed population (Ashley, 2007; Strader et al., 2022). Tornadic activity is not the only source of wind damage in the GOM region. Murley and colleagues (2021) found that high-wind events, or HWEs, meeting National Weather Service criteria are most prevalent in the South and West North Central regions. The authors note that such events, which may include derechos, straight-line gust winds, downbursts, or microbursts, have caused nearly $300 million in property losses during the past 60 years and roughly 1,500 deaths since 1980. Knox and colleagues (2011) have documented the role of non-convective winds (not associated with a hurricane or convective storm) in property damage. Extreme Temperatures Since global temperature records began in 1880, Earth’s temperature has risen by an average of 0.14°F (0.08°C) per decade, or about 2°F in total, with the 10 warmest years occurring since 2010 (Lindsey and Dahlman, 2023). From 1981 to the present, the warming rate per decade has been more than double the rate of 0.32°F (0.18°C) as previous decades (Lindsey 52 PREPUBLICATION | UNCORRECTED PROOFS 

and Dahlman, 2023), and 2023 recently surpassed 2016 as the hottest year on record (Copernicus, 2024). Marine heat waves (when the ocean temperature is warmer than 90 percent of the previous observations for any given time of year) are occurring in GOM waters, stressing sensitive marine ecosystems (NOAA, 2023), raising heat and humidity on nearby land during the hottest times of the year and providing “jet fuel” for tropical cyclones (Harvey, 2024). GOM states currently lead the nation in numbers of days of perilous heat. Climate change, humidity, low elevation, and warm waters in the GOM will contribute to an average of 20 extra days of triple-digit heat per year (First Street Foundation, 2022). Residents of Texas and Florida can expect to see more than 70 consecutive days with the heat index surpassing 100°F (37.7° C)(Muyskens et al., 2023). Three GOM states have the second (Texas), third (Mississippi), and fourth (Louisiana) highest rates of extreme heat. Extreme temperatures and weather-climate events are among the leading causes of major power outages in the United States, prompting increased electricity demand for heating and/or cooling, which can stress and overload aging energy infrastructure (Climate Central, 2024). From 2018 to 2020, three GOM states had the highest (Louisiana), second-highest (Texas), and fourth-highest (Mississippi) annual average counts of 8+ hour outage events. For 1+ hour outage counts, these states rank highest (Texas), second highest (Louisiana), and third highest (Mississippi) (Do et al., 2023). Hazards associated with landfalling tropical cyclones are often exacerbated and compounded by power outages. Roughly 90 percent of major power outages in the United States are associated with hurricanes (Alemazkoor et al., 2020). Since tropical cyclones are inherently warm season hazards, the impact of heat in the aftermath of a storm is often understudied. In fact, very few studies are found in the literature. Reesman (2022), in a thesis focused on Hurricane Laura and defoliation, found evidence of increased apparent temperature during nocturnal hours after the storm. Guido et al. (2022) reported that heat index anomalies surged in the days following many hurricanes in the Caribbean region. 53 PREPUBLICATION | UNCORRECTED PROOFS 

Sea Level Rise and Coastal Flooding By 2050, sea level along U.S. coastlines is expected to rise by 10–12 inches (0.25–0.30 m; see Figure 2-2 for relative sea level rise estimates for 2050 and 2100 under the Intermediate Sea Level Rise Scenario), resulting in extensive erosion, land loss, and more frequent overland flooding in coastal communities (Dangendorf et al., 2023; NOAA, 2022c; Yin, 2023). Notably, sea level rise rate projections in the United States are highest along the western Gulf Coast (see Figure 2-2). FIGURE 2-2 Relative sea level rise along the U.S. coastlines under the Intermediate Sea Level Rise Scenario of the U.S. Interagency Sea Level Rise Task Force for 2050 (left) and 2100 (right). By both 2050 and 2100, sea level rise is projected to be highest along the western Gulf Coast. SOURCE: NCA, 2023. 54 PREPUBLICATION | UNCORRECTED PROOFS 

Land Subsidence Subsidence, the gradual sinking of the land surface, results from several natural and anthropogenic factors. The Mississippi River Delta Plain naturally subsides as sediments compact under the weight of overlying material. Levees have greatly reduced the rejuvenation of sediment to the delta surface, and consequently subsidence exceeds land building. Other factors include oil, natural gas, and water extraction, which contribute to the sinking of the land mass. Groundwater removal has been a primary factor in Texas. Faulting in the delta sediments also plays a role (Lane et al., 2018; USGS, n.d.). Recent research reveals that subsidence rates vary across the Gulf Coast, but some areas are experiencing subsisdence at a rate of 4 mm per year (Wang et al., 2020; Zhou et al., 2021). Although marshes can sustain themselves during moderate sea level rise, Törnqvist and colleagues (2020) report that the Mississippi Delta marshes are nearing a tipping point for long-term sustainability. In coastal marshes, land subsidence amplifies the effects of sea level rise. Sea level rise and land subsidence are phenomena that, like other chronic, slow-moving hazards such as “nuisance flooding” and the slow march of drought, can have significant agricultural, environmental, economic, health, and social consequences on their own, but in combination with more acute stressors and a changing climate, can be catastrophic. Technological Hazards: Chemical Releases The GOM region is the primary U.S. hub for the fossil fuel industry. Forty-seven percent of total U.S. petroleum refining capacity is situated in the GOM region (USEIA, 2024) with more than 90 percent of U.S. primary petrochemicals capacity located in Texas and Louisiana (Augustine, 2024). Together, Louisiana and Texas have the 10 largest petrochemical complexes in North America and produce more than a third of the United States’ methane (USEIA 2023; USEIA, 2024). Many residents, disproportionately communities of color and with low socioeconomic status live close to these and other high-risk chemical facilities, including transport facilities and/or carcinogen-laden hazardous waste (Superfund) sites. Many of these communities have been overburdened by the environmental harms and risks from exposure, cumulative impacts, disproportionate health impacts, and greater vulnerability to pollution for 55 PREPUBLICATION | UNCORRECTED PROOFS 

decades (Saha et al., 2024; Terrell and James, 2022; see “Exposure to Environmental Contaminants” later in this chapter for more information). There are more than 4,000 oil and gas structures (NOAA, n.d.-b) in the GOM region, interconnected by more than 26,000 miles of oil and gas pipelines (NOAA, n.d.-c), making the region vulnerable to contaminant spills, leaks, and explosions, (e.g., 2010 Deepwater Horizon disaster). Human health hazards are evident at every point along the petroleum production pathway; from extraction (drilling or hydraulic fracturing [“fracking”]) to refining to petrochemical production to burning to end-product production (e.g., plastics) to disposal and management of wastes. Liquefied natural gas (LNG) facilities, for example, emit several hazardous air pollutants (HAPs) that are known to cause a variety of human health problems (Saha et al., 2024). Some of these HAPs include “formaldehyde (a known carcinogen that causes myeloid leukemia and nasal cancers), benzene (also a known carcinogen), toluene (which can cause a range of reproductive harm including birth defects), ethylbenzene (a suspected carcinogen that can cause hearing and kidney damage), and xylene (which has wide-ranging effects” (Saha et al., 2024, p. 172). According to the Federal Energy Regulatory Commission, six LNG facilities located in Texas and Louisiana are anticipated to collectively emit 264 tons of these HAPs annually (Saha et al., 2024). Chemical releases occur on blue-sky days as well as in the context of a disruptive event. Natural Hazards Triggering Technological Accidents (Natech) events are incidences of natural hazards (e.g., flooding) that initiate events that challenge the safety and operation at hazardous installations (OECD, 2024). Technological failures and Natech events are not only increasing in the United States (Sengul et al., 2012) but are expected to continue to rise in both frequency and magnitude (Krausmann et al., 2017 as climate conditions worsen (Saha et al., 2024). Yet, they are typically overlooked in regional and national disaster risk management plans (Girgin et al., 2019). The COVID-19 Pandemic With extraordinary speed, COVID-19 circumnavigated the globe, moving from the Western Pacific to worldwide over the span of 1 month—March 2020 (Shultz, Perlin et al., 2020). The World Health Organization declared COVID-19 to be a “public health emergency of 56 PREPUBLICATION | UNCORRECTED PROOFS 

international concern” in January 2020 (WHO, 2020a) and then upgraded it to “pandemic” status in March 2020 (WHO, 2020b). In the United States, the pandemic was declared a national emergency under Section 501(b) of the Stafford Act (Robert T. Stafford Disaster Relief and Emergency Assistance Act, 42 U.S.C. 5121 et seq.) on March 13, 2020. The sudden emergence and ongoing evolution of the COVID-19 pandemic joined climate change as phenomena of such global scope and scale as to fundamentally influence concurrent disaster events at all levels: worldwide, nationally throughout the United States, and with disproportionate impact in the GOM region due to the confluence of the COVID-19 variants during the 2020 and 2021 hurricane seasons. As a hazard, the COVID-19 pandemic stressed the public health system, including those who work and/or volunteer in the sector at the local, state, and regional levels. The pandemic has also heightened the health care sector’s awareness that social conditions influence health (Gottlieb et al., 2021). EXPOSURE Exposure to extreme weather-climate hazards in the GOM region is increasing as a result of population growth and increased construction of public and private development in hazard-prone areas. In addition to potential increases in the temporal frequency or rate of return of extreme weather-climate hazard events, several studies have found that migration and development near the U.S. Gulf Coast are also contributing to escalating losses from these events (Hoffman, et al., 2023; Pielke and Landsea, 1998). Exposure is understood as land development that places people, infrastructure, housing, production capacities, and other tangible human assets in hazard-prone areas. The term encompasses the biophysical factors that may contribute to a disruptive event (e.g., residence in a coastal area or a river floodplain). Figure 2-3 illustrates how the expansion of the built environment can increase exposure to flood hazards and thereby increase risk associated with a flood disaster. Growing exposure along the Gulf and Atlantic Coasts is linked to mounting hurricane losses (Freeman and Ashley, 2017; Zhu and Quiring, 2022). Much of this construction is also inherently vulnerable because of inadequate construction standards, discussed later in this chapter. (See also Strader, 2023 for more information on how the GOM societal landscape is shaping cyclone and tornado disasters.) 57 PREPUBLICATION | UNCORRECTED PROOFS 

FIGURE 2-3 The “expanding bull’s-eye effect” is a conceptual model for a hypothetical metropolitan region characterized by increasing development spreading from an urban core over time. A sample flood scenario is overlaid to illustrate that flood risk increases as population and built environment expand into areas in or near flood zones. SOURCE: Ashley et al., 2014. Population and Population Exposure Of the three U.S. census-designated coastline regions (Pacific, Atlantic, and GOM), 2 the GOM is the smallest coastline region by area but the fastest growing in terms of overall population (U.S. Census Bureau, 2019). From 2000 to 2017, the GOM coastal population increased by 26.1 percent; by context, the U.S. population growth rate over the same time frame was 15.7 percent (U.S. Census Bureau, 2019). From 1940 to 2020, GOM states experienced significant population growth (432.7 percent, 34 million people) and corresponding exponential growth in housing units (2,818 percent, 17 million homes; Strader, 2023). Of the six GOM representative counties/parishes that are a focus of this study, Harris County, Texas, experienced the highest rate of growth in population (32.3 percent) and housing units (61.1 percent) from 1940 to 2020, driven largely by the Houston metro area. Baldwin County, Alabama, and Calcasieu Parish, Louisiana, also saw significant increases in population (20–30 percent) during that period, with corresponding increases in housing units (approximately 50 percent). Of note, vacation or secondary residences comprise many of the housing units in the coastal study area. Vacation homes account for 39.7 percent of the housing units in Baldwin County, Alabama, 63.2 percent of the housing units in 2 Counties or parishes adjacent to coastal water or territorial sea. 58 PREPUBLICATION | UNCORRECTED PROOFS 

Cameron Parish, Louisiana, and 40 percent of the housing units in Galveston County, Texas, as of 2022 (U.S. Census Bureau, 2020). Rapid population growth in GOM states has further contributed to extending and expanding persistent socioeconomic inequities (USGCRP, 2023), pushing many low-income residents toward low-lying areas with flood risks amplified by climate change and increased runoff from expansion of the built environment (Ueland and Warf, 2006). The 2020–2021 period was also marked by sharp population shifts. From April 2020 to May 2022, Texas and Florida saw among the highest U.S. population increases in terms of relative growth (first and fourth, respectively, in percent growth). Regionally, within the GOM states of Louisiana and Mississippi, for example, there are notable exceptions. Population estimates released in 2021 for Cameron and Calcasieu Parishes in Louisiana, show that these localities experienced some of the nation’s sharpest declines and the largest combined percentage decrease compared with metro areas nationwide (M. Smith, 2022). This decline is part of a wider trend in Louisiana, as well as in Mississippi, which ranked fifth and ninth, respectively, in relative decline (third and seventh, respectively, in percent decline) (U.S. Census Bureau, 2022c). It must be noted that, as populations depart the coast and parish/county budgets decline, inadequate funding will become an increasing concern. Selected census-derived 2020 demographics for the six representative counties/parishes that are a focus of this study and Marion and Jefferson Davis Counties, Mississippi, are shown in Table 2-1. TABLE 2-1 Selected Demographics of Counties/Parishes in GOM States in 2020 County, Population Race– Race– Ethnicity– Owner- Median Poverty State White Black or Hispanic occupied Income level (%) alone African or Latino housing unit ($) (%) American (%) rate (%) alone (%) Baldwin, 231,767 83.4 8.4 5.0 77 61,756 9.2 AL Mobile, AL 414,809 55.7 36.7 3.2 64.1 44,091 17.6 59 PREPUBLICATION | UNCORRECTED PROOFS 

Cameron, 5,617 88.1 3.9 4.6 88.5 56,902 13.1 LA Calcasieu, 216,718 67.2 24.8 4.4 68.5 54,530 17.4 LA Marion, 29,341 65 31.5 1.8 75.9 40,978 17.3 MS Jefferson 11,321 37.9 59.1 1.9 80.3 32,214 22.7 Davis, MS Harris, TX 4,731,145 69.6 20.1 43.7 54.9 63,022 15.6 Galveston, 350,682 80.3 13.3 26.8 67.5 74,633 10.9 TX U.S. 331,464,948 58.9 13.6 19.1 64.8 75,149 11.4 NOTE: Searches conducted for galvestoncountytexas, cameronparishlouisiana, calcasieuparishlouisiana, marioncountymissississippi, jeffersondaviscountymississippi/FIPS. SOURCE: U.S. Census Bureau, 2021a. Exposure to Wind Hazards A population’s exposure to wind hazards may best be characterized by the hazard maps in ASCE 7, an American Society of Civil Engineers standard that underpins the International Codes. 3 ASCE 7-22 establishes hurricane-prone regions as those areas with basic wind speeds of 115 mph or higher, which are subject to heightened design requirements (ASCE, 2022a). This region extends from the coastline of the GOM states to approximately 100–150 miles inland. Within this zone, ASCE 7-22’s wind speed contours signify increasing exposure to strong winds by progressively raising the basic wind speed a building must resist the closer it is to the coastline. For example, structures in ASCE 7-22 Risk Category II, which would include 3 The International Codes are developed by the International Code Council, which uses a governmental consensus process to develop International Codes (I-Codes) (ICC, 2024). I-Codes are the minimum standard for building design to ensure safety and wellbeing (ICC, 2024). 60 PREPUBLICATION | UNCORRECTED PROOFS 

noncritical, low-occupancy buildings such as single-family homes, built in Galveston, Texas, are expected to experience wind speeds of 150 mph, just slightly higher than the basic wind speeds in Mobile, Alabama (146 mph) and markedly higher than the 129 mph basic wind speed requirements in the inland city of Lake Charles, Louisiana (see Figure 2-4). These standards presume that hurricanes weaken as they pass over land. Yet fast-moving storms like Hurricane Laura challenged this assumption and delivered winds in excess of the basic wind speed assumed in ASCE 7 well inland of the coast, including in Lake Charles (Roueche et al., 2020). ASCE 7-22 also defines tornado-prone regions of the United States, which includes the entirety of all GOM states, highlighting that geographies within 100–150 miles of the GOM coastline face heightened exposure to all varieties of wind hazards compared with the overall continental United States. Thus, exposure to the damaging effects of wind can persist for miles inland from the coast. This wind hazard exposure includes not only the direct impacts of the wind pressure on the built environment but also heightened exposure to other related hazards, such as windborne debris and wind-driven rain, both of which contribute significantly to losses often outside of delineated flood zones. FIGURE 2-4 Basic wind speeds for Risk Category II buildings and other structures along the Gulf Coast. SOURCE: ASCE, 2022a, with permission from ASCE. 61 PREPUBLICATION | UNCORRECTED PROOFS 

Exposure to Coastal Hazards More than 18,130 km2 (7,000 mi2) of the Gulf Coast is 1.2 m (4 ft) or less above current sea levels. This area includes critical infrastructure (e.g., ports, airports, hospitals, schools, petrochemical facilities, interstates, railroads) essential to local and national economies and large urban centers (e.g., Houston and New Orleans; Strader, 2023). Approximately 25 percent of health care facilities and public safety assets and 20 percent of schools in all six counties/parishes that are a focus of this study are flood-prone (within 100 m of a 100-year floodplain; Strader, 2023). In Galveston County, Texas, more than 60 percent of hospitals are flood-prone, as is the only hospital in Cameron Parish, Louisiana, where 99.7 percent of surface roadways are in a flood hazard zone (Strader, 2023). Exposure to hazards is heightened by the fact that substantial portions of the housing inventory are flood-prone. Of the six counties/parishes, Cameron Parish (93.3 percent) and Galveston County (49.2 percent) contain the highest percentages of their building footprints in flood-prone areas (Strader, 2023). Sea level rise and flood exposure will significantly impact businesses and residents adjacent to or dependent upon the bayous, coastlines, and coastal waterways, and also inland areas subject to riparian flooding (Colten, 2021), as people migrate away from the rising water and subsiding land. Figure 2-5 illustrates physical factors that can shape the potential for exposure to flood risk. FIGURE 2-5 Physical factors contributing directly to flood exposure. SOURCE: NOAA, 2022e. 62 PREPUBLICATION | UNCORRECTED PROOFS 

Exposure to Extreme Temperatures Extreme cold temperatures are decreasing in frequency and intensity globally (Hu et al., 2020), while extreme heat temperatures are increasing (Song et al., 2022). But both heat and cold have measurable health effects, particularly among vulnerable populations (USGCRP, 2023; Adams-Fuller, 2023; Jay et al., 2021) and are the deadliest weather-climate-related phenomenon in the United States (NOAA, 2021n; Adams-Fuller, 2023). In the United States, extreme heat kills approximately 1,300 people every year—more than hurricanes, tornadoes, and floods combined (Adams-Fuller, 2023). Thirty-year projections for the number of days with a heat index of 100°F (37.78°C) or higher show a worsening situation in the coming decades (Adams-Fuller, 2023). Within the next 30 years, the dangers of extreme heat are projected to be the most widespread in the southern United States. As climate change continues to amplify exposure risk to extreme heat and hot weather, the associated health stress is escalating mortality, morbidity, adverse pregnancy outcomes, and adversely affecting mental health and constraining the ability to work outdoors (Ebi et al., 2021). Despite the known risk, this hazard receives minimal dedicated federal support and funding for planning, education, mitigation, and recovery (Wickerson, 2023). Decision-makers and policymakers and the public tend to focus on more episodic or telegenic events (e.g., landfalling hurricanes, tornadoes), in part because extreme heat events are ineligible for major disaster declaration under the Stafford Act, precluding deployment of resources and coordinated action, and as such, the consequences of extreme heat are difficult to respond to. Extreme temperatures (thus far) largely do not damage property or cause as much physical destruction as severe storms yet are due far more attention in the face of global temperature warming trends. With planning, education, and action, heat-related impacts are preventable (NIHHIS, n.d.). Localized and well-communicated heat action plans can help elevate public discourse on the ways in which extreme heat is and will continue to interrupt daily activities. Plans that address behavioral approaches and biophysical adaptations as well as include surveillance, monitoring, and evaluation can reduce the interruptions and adverse health consequences of current and future extreme heat (Errett et al., 2023; Ebi et al., 2021; Jay et al., 2021). 63 PREPUBLICATION | UNCORRECTED PROOFS 

Exposure to Environmental Contaminants One of the most consistent findings in the social science literature is that place matters. More than two decades of research shows a relationship between the location of environmental hazards (e.g., petrochemical production and transport, by-products associated with hydraulic fracturing, landfills, hazardous waste sites) and racial (e.g., African American) and/or economic (e.g., low-income) status. GOM populations are disproportionately and unequally exposed to environmental contaminants, toxicants, and other hazards that not only exacerbate negative physical and mental health outcomes on blue-sky days (Domingue, 2022; Donaghy et al., 2023; Prochaska et al., 2014), but also pose increased risks to human health when hurricanes, flooding, and other weather-climate events mobilize these toxicants to enter communities. Furthermore, because many industrial activities often require access to surface water, industrial land often overlaps with flood zones (Flores, Castor et al., 2021), increasing the susceptibility of adjacent neighborhoods to hazards associated with the petrochemical industry (Bernier et al., 2017). This increased exposure occurred, for example, during and after Hurricane Harvey, when many low-income communities were disproportionately exposed to chemical contaminants (Karaye, Stone et al., 2019) wherein a U.S. Environmental Protection Agency Office of Inspector General report (U.S. EPA, 2019) concluded that neither the state of Texas nor local authorities adequately monitored air quality during or immediately following the storm. Nor did these entities effectively communicate with residents about the risks posed by the chemical releases and explosions. Further, many industrial facilities were caught off guard and were forced to shut down and restart, resulting in the release of an estimated 2,750–4,150 tons of excess emissions (Phillips, 2018). Flooding prompted more chemical releases, including an explosion at the Arkema chemical plant in a suburb of Houston that burned for 4 days after a backup generator failed (CSB, 2017). Increased exposure also regularly occurs in Louisiana, which ranks among the bottom five states for air pollution and mortality (Yu et al., 2023). In a region commonly known as “Cancer Alley,” where more than half (378 of 671) of the state’s industrial facilities are spatially clustered along a 184-mile stretch of the lower Mississippi River, communities of color have “7- fold to 21-fold higher emissions, depending on the pollutant, than predominantly White 64 PREPUBLICATION | UNCORRECTED PROOFS 

communities” (Terrell and St. Julien, 2023). Chemical manufacturing facilities are disproportionately responsible for pollution emissions in Cancer Alley, accounting for about one- third of emissions of particulate matter (PM10 and PM2.5), nitrogen oxides (NOx), sulfur dioxide (SO2), carbon monoxide (CO), and volatile organic compounds (VOCs; Terrell and St. Julien, 2023). Exposure to these chemicals can contribute to a wide range of negative health impacts such as chronic obstructive pulmonary disease (COPD), asthma, bronchiolitis, lung cancer, cardiovascular events, central nervous system dysfunction, and cutaneous diseases (Manisalidis et al., 2020). Residents of some Cancer Alley towns were forced to abandon the tradition of porch sitting in the evening due to nighttime chemical releases (Baurick, 2019). Despite the higher levels of exposure to environmental risks and health inequities, these communities have fewer resources to address the adverse effects of environmental contaminants (U.S. EPA, 2021; Shepherd and KC, 2015; USGCRP, 2023). Understanding of these inequities resulted in the concept of environmental justice and fenceline communities—those located alongside environmental hazards—which experience disproportionately higher rates of cancer (Raun et al., 2013), asthma (Byrwa-Hill et al., 2023), cardiovascular diseases (Motairek et al., 2023), and COVID-19 deaths relative to the general U.S. population (Fos et al., 2021; Terrell and James, 2022). Fenceline communities’ proximity to industrial facilities means that they are not only exposed to potentially toxic emissions but face higher risk of exposure to a Natech event (Nicole, 2021; see “Technological Hazards: Chemical Releases” earlier in this chapter for more information on Natech events). VULNERABILITY Vulnerability is shaped by complex forces, and is a result of the range of institutional, political, economic, social, cultural, and psychological factors and systems that configure people’s lives and the environments in which they live (UNDRR, n.d.-a). Disparate yet interdependent social stratification processes (Kuran et al., 2020) influence the degree or sensitivity to which an individual or community is vulnerable to and affected by disruptive events. Sensitivity encompasses the social factors that lessen the ability of a person, family, or community to cope with a disruptive event, and can determine the degree to which they are affected by such an event. 65 PREPUBLICATION | UNCORRECTED PROOFS 

Adaptive capacity can mitigate the potential for harm by reducing sensitivity to a disruptive event. Adaptive capacity enables a human community to adjust to environmental conditions and effectively offset vulnerabilities (Smit and Wandel, 2006). The identities, conditions, and demographics of vulnerable individuals can be intersectional, operating together to increase disaster risk. As Tierney (2019, p. 127) states, people “are not born vulnerable, they are made vulnerable.” Vulnerable groups include, but are not limited to, pregnant people (Sharma et al., 2022), women (Nguyen et al., 2023), ethnic/racial minorities, children, elderly adults, Indigenous people, low-income communities (Bullard and Wright, 2012), undocumented immigrants, unhoused populations, prisoners, people with disabilities (Shultz et al., 2018), LGBTQIA+ 4 populations (Haworth et al., 2022), unprotected occupational groups (e.g., workers exposed to extreme weather; Kiefer et al., 2016), medically high-risk patients (Balbus and Malina, 2009; Espinel et al., 2022; Shultz et al., 2024), and the politically marginalized (Joseph et al., 2021). Vulnerabilities and the composite burden of multiple, interrelated sensitivities are place based, and the resulting consequences are contingent on the particular disaster or disasters. Actions to address them must be tailored to the complexities of particular situations that arise from local populations and social, economic, and political conditions. Many people who are vulnerable along one dimension (e.g., experiencing housing instability) are vulnerable along others as well (e.g., having difficulty accessing health care or employment opportunities). These individuals experience intersecting, overlapping, clustered, and at times compounding vulnerabilities even before a disruptive event occurs (Stanley, 2017). The impacts of the event are then “piled on” to these extant vulnerabilities. Some factors that contribute to vulnerability are based on nonmodifiable risk factors (e.g., age is unalterable, yet children and elderly adults bear disproportionate disaster burdens). This same concept applies to discrimination based on gender, race, and/or ethnicity; women, for example, are 14 times more likely than men to die in disasters (CDP, 2023; Howe, 2019; Neumayer and Plümper, 2007). Persistent social vulnerabilities, disparities, inequalities, and inequities are dominant social forces that shape population health, local economies, the education system, and infrastructure quality, and ultimately exacerbate disaster risk. 4 LGBTQIA+ is an abbreviation for lesbian, gay, bisexual, transgender, queer, intersex, asexual, and other identities (Haworth et al., 2022). These terms are used to describe a person's sexual orientation or gender identity. 66 PREPUBLICATION | UNCORRECTED PROOFS 

Many vulnerabilities relate directly to sensitivity to hazards. For example, the density of mobile/manufactured housing can lead to greater odds of dying in a tornado, particularly in the Southeast United States; young and elderly populations are more likely to be killed; and those living below the poverty line have less ability to withstand impacts and losses (Strader, 2023). High Social Vulnerability in Gulf of Mexico States Communities throughout the GOM region lead the nation across a spectrum of vulnerability measures. Since 2000, social vulnerability has increased in the region, especially for the percentage of persons or households that are unemployed (+3.7 percent), aged 65 or older (+3.6 percent), and/or members of minority groups (+5.5 percent; Strader, 2023). Several GOM states rank among the highest in the nation on, among others, the Centers for Disease Control and Prevention’s (CDC) Social Vulnerability Index (SVI). 5 Table 2-2 presents county-level SVI vulnerability measures for GOM states in 2000 and 2018, as well as the percentage change for each variable for the GOM region and the United States. TABLE 2-2 County-Level Vulnerability Measures from the Social Vulnerability Index (SVI) for the GOM States and Percent Change, 2000–2018, GOM States and United States 2000 to 2000 to Year Year Year Year 2018 2018 Theme Variable 2000 (%) 2000 (%) 2018 (%) 2018 (%) Change Change (GOM) (U.S.) (GOM) (U.S.) (%) (%) (U.S.) (GOM) Socioeconomic Below Poverty 19.0 14.2 19.9 15.6 0.9 1.1 Status Unemployed 3.7 3.4 7.4 5.8 3.7 2.1 Per Capita Income ($) 16,027 17,513 24,045 27,036 8,018 9,515 5 The CDC’s SVI is a place-based index, database, and mapping application designed to identify and quantify communities experiencing social vulnerability (CDC, 2024). 67 PREPUBLICATION | UNCORRECTED PROOFS 

No High School 29.0 22.6 18.3 13.4 -10.7 Diploma -9.2 Aged 65 Years or 14.0 14.8 17.6 18.4 3.6 3.6 Older Household Aged 17 Years or 26.0 25.5 22.8 22.4 -3.2 -3.1 Composition + Younger Disability Civilian w/ Disability 24.0 20.9 17.0 15.9 -7.0 -5.1 Single-Parent 10.1 8.2 9.9 8.3 -0.2 0.1 Households Minority 36.0 18.7 41.5 23.5 5.5 4.8 Minority Status Speak English “less + Language 2.6 1.6 2.7 1.7 0.1 0.1 than well” Multi-Unit Structures 4.0 4.1 4.4 4.7 0.4 0.6 Mobile/Manufactured Housing Type 24.0 14.9 22.0 12.9 -2.0 -2.1 Homes + Crowding 5.9 3.6 3.3 2.4 -2.6 -1.2 Transportation No Vehicle 9.1 7.6 6.8 6.4 -2.3 -1.2 Group Quarters 4.3 3.4 4.6 3.5 0.3 0.1 NOTES: County-level SVI data for the years 2000 and 2018 show how individual vulnerability measures and the four primary SVI categories have changed over the 18-year period. Although 2020 SVI data are available, the categories used were slightly different from those used in 2000. Data from 2018, before the categories changed, are therefore used for the comparison with 2000. Pink cells indicate increasing vulnerability, and blue cells indicate decreasing vulnerability. SOURCE: Strader, 2023. Social vulnerability in the GOM region continues to be approximately 25 percent higher than that in the rest of the continental United States (Strader, 2023). GOM states rank high on poverty measures and low on household income and educational attainment (see Box 2-2 for information on the role of education in disaster risk and recovery). This substrate of concentrated disadvantage has substance, depth, and constancy that reinforces a negative feedback loop and elevates risk for the subset of GOM communities as they grapple with disruptive events. Because social vulnerability is socially constructed (Trivedi, 2023) context-specific, and an outcome of demographic and historical legacies intrinsic to a place (Bankoff, 2004; Kuhlicke, et al., 2011), some have expressed caution as to how single composite indexes, such as the SVI, 68 PREPUBLICATION | UNCORRECTED PROOFS 

are used (Roy and Berke, 2022). Marino and Faas (2020, p. 33), write, “There is a growing discomfort that categorizing the ‘vulnerable’ acts to flatten and simplify diverse communities, as well as discursively nullify the everywhere-visible ‘resilience,’ toughness, and genius that exists in communities, and subsets of communities, that are habitually exposed to risk.” Quantitative indexes are only one metric for building an understanding of the social impacts of compounding disasters, and pairing indexes with empirical validation is one way to make them more robust (Rufat et al., 2019). Gathering information on people’s lived experiences goes beyond the demographics of the people and places at risk and can elucidate the ways in which social marginalization—the root force that drives most forms of vulnerability—can prolong the impacts of compounding disasters (Priest and Elliott, 2023). BOX 2-2 The Role of Access to and Quality of Education in Disaster Risk and Recovery Studies show that individuals with higher education experience healthier lives and longer lifespans (Raghupathi and Raghupathi, 2020) and are better positioned to build generational wealth (Kent and Ricketts, 2021). However, access to and quality of education take on increased importance in the context of disaster risk and recovery. In one longitudinal study (Thiede and Brown, 2013) on race, socioeconomic status, and evacuation behavior during Hurricane Katrina (in 2005), researchers found that people with less than a high school education were less likely to evacuate. Comparison of low-education respondents with respondents having at least some college education showed that the low-education respondents were more than twice as likely to have been unable to evacuate because of a lack of money, transportation, a place to go, or job requirements. In another study (Tracy et al., 2011), conducted several months after Hurricane Ike, the researchers interviewed 658 adults who had been living in affected areas during the storm. The study showed that depression was more likely to occur among those with lower annual household income and fewer years of education (a high school degree or equivalent versus than those with some college or more years of education) (Tracy et al., 2011). Nationwide, almost 90 percent of students complete high school, and many go on to higher education; however, all Gulf of Mexico (GOM) states have graduation rates well below the national average. After California (the state with the lowest high school graduation rate in the nation), three GOM states, that is, Texas, Mississippi, and Louisiana, have the second-, third-, and fourth-lowest rates, respectively, while Alabama has the seventh-lowest rate. Here too, there are disparities between 69 PREPUBLICATION | UNCORRECTED PROOFS 

racial and ethnic minorities and the non-Hispanic White population. While graduation rates continue to improve among racial/ethnic groups, the historical injustices and lack of access to quality education persistently place these groups at a disadvantage with respect to improving and increasing generational wealth. Limited education, low literacy, and a lack of broadband access can also hinder people’s ability to understand and access disaster and other types of federal assistance. Many GOM residents are marginalized in terms of income; education; age; ethnicity/race; representation in policy, governance, and recovery planning; gender; sexuality; and/or being medically high-risk. These individuals and groups bear a disproportionate risk for and burden from compounding disasters (Bullard and Wright, 2012; Lichtveld, 2018; Machlis et al., 2022). Their marginalization reduces their individual and collective adaptive capacity and magnifies their sensitivity to the consequences of future compounding disasters. Comprehensive community participatory planning for future disruptive events is one way to help address these vulnerabilities before an event occurs. Participatory planning processes seek to engage and empower community members, especially marginalized people and those disproportionately affected by disasters to proactively engage in planning processes that build close partnerships with local and regional governments and strengthen community leadership and local adaptive capacities. Participatory planning can build and reinforce trust and enhance transparency and inclusivity via multiple levels of information gathering and sharing, while also strengthening social cohesion. As a component of participatory planning, there are tools and frameworks that communities can use to evaluate their adaptive capacities (see, e.g., Sherrieb et al., 2010; Masterson et al., 2014; Notre Dame Global Adaptation Initiative) This evaluation is one way to understand a community’s strengths, assets, resources, and deficits to better inform community-based planning. Providing flexible resources—including training and financing—directly to community-based organizations and communities will help them assess their adaptive capacity, prepare and plan for, absorb, recover, and ultimately reduce impacts from ongoing and future disruptive events. The recent Climate Vulnerability Index highlights where flexible resources and action are urgently needed for communities. It offers climate risk and vulnerabilities data at the census tract and county/parish level and incorporates input from community leaders. It comprises 184 indicators grouped under four baseline vulnerabilities (health, social/economic, infrastructure, 70 PREPUBLICATION | UNCORRECTED PROOFS 

and environment) and three climate change risks (health, social/economic, extreme events) to explain cumulative effects on neighborhood-level stability. Seven of the 10 most at-risk U.S. counties/parishes are in GOM states; of those, half are in Louisiana (EDF, 2023). Three GOM states—Texas, Louisiana, Alabama—and one southern state—Tennessee—account for 87 of the highest 100 scores of at-risk census tracts (Lewis et al., 2023). Of those, many are located near concentrations of industrial facilities and are known environmental justice communities (Lewis et al., 2023). Material losses experienced by vulnerable communities in hazard-susceptible areas (e.g. health care, education, community infrastructure, housing, transportation) trigger and magnify existing social and spatial health stratifications (Weden et al., 2021), creating a vicious cycle that many cannot escape, and leaving some communities unable to prepare for the next disruptive event because of insufficient recovery from previous ones (Ingham et al., 2022). Public Health Disasters have broad and profound impacts on public health (Nomura et al., 2016; Shoaf and Rottman, 2000; Shultz, 2019; Shultz, Espinel, Galea et al., 2007), producing harms to individuals and communities in such forms as injury, disease, psychological distress and trauma, and death (UNDRR, n.d.-b; Shultz, Espinel, Galea et al., 2007; Shultz et al., 2013, 2017). The complexities and/or accumulation of multiple concurrent or sequential disaster exposures produce risks to population health that exceed those associated with single disaster exposures (de Ruiter et al., 2020; Ebi et al., 2021; Leppold et al., 2022; Lowe et al., 2020 Pescaroli and Alexander, 2018; Zscheischler et al., 2018). Disasters also may damage health care facilities and disrupt access to health care services, placing heavy demands on frontline health professionals (Shultz and Forbes, 2014) 6 and social support services (Heagele and Pacquiao, 2019). Each disaster has unique impacts on public health related to its distinguishing features (Shultz, 2019; Shultz and Neria, 2013; Shultz et al., 2015) and the underlying vulnerabilities at every social level—individual, family, social network, community, city, state, nation, and beyond. Frontline health professionals include first responders, public health professionals, and volunteers tasked 6 with medical and public health–related disaster response and recovery responsibilities. 71 PREPUBLICATION | UNCORRECTED PROOFS 

Many residents of GOM communities suffer from poor health and health disparities that increase their vulnerability to disasters. Among key vital statistics, compared with those in the rest of the United States, life expectancies in the GOM region are low, while infant mortality and maternal mortality rates are high (CDC, 2023a, 2023b). Box 2-3 describes a study examining the relationship between compounding disasters and adverse health outcomes. The region also has some of the highest rates of incidence, prevalence, and mortality for the major noncommunicable diseases that represent the leading causes of death in the United States, including heart disease, stroke, cancer, COPD, diabetes mellitus, and Alzheimer’s disease (CDC, 2022; CMS, 2018). The clustering of very high mortality rates for these diseases in the GOM states of Alabama, Louisiana, Mississippi, and at times Texas is evident and generally consistent (CDC, 2022). The region also has a high prevalence of prominent lifestyle risk behaviors associated with these diseases, including physical inactivity, obesity, and cigarette smoking. Population Health Disparities and Social Vulnerability Research teams have documented how social vulnerability has created disproportionate risks for population health in the GOM region. Smiley (2020) examined racial inequalities in “flood extent,” primarily outside the 100-year floodplains, during the unprecedented inundation associated with Hurricane Harvey. Using attribution science, Smiley and colleagues (2022) were able to demonstrate that 30 to 50 percent of flooded properties would not have flooded without climate change, and flooded properties were concentrated in low-income Latina/x/o neighborhoods primarily located outside of the Federal Emergency Management Agency’s (FEMA) 100-year floodplain. Responding to the call to prioritize the integration of research into public health emergency response (Lurie et al., 2013), Horney and colleagues (2018) investigated residential contamination with polyaromatic hydrocarbons (PAHs) in household dust and outdoor soil in a Houston environmental justice neighborhood prior to and after Hurricane Harvey. Corroborating Horney et al.’s finding of redistribution of PAHs in low-socioeconomic status (SES) neighborhoods in the aftermath of Hurricane Harvey’s flooding, Lieberman-Cribbin et al. (2021) found sharp socioeconomic disparities when examining 83 toxic waste site releases during 72 PREPUBLICATION | UNCORRECTED PROOFS 

Harvey, with most toxic releases concentrated in neighborhoods in the lowest SES index quintiles. Sansom et al. (2023) and Atoba et al. (2023) documented poorer health indicators in residents living close to toxic release facilities and areas prone to flooding contamination. Collaborators have also conducted in-depth research on the intersection of social vulnerability, health status, and disproportionate disaster impacts, with multiple studies focusing on Hurricane Harvey. Chakraborty, Collins et al. (2019) and Flores, Collins et al. (2021) bring an environmental justice lens to these analyses, illuminating that, prior to Harvey, racial/ethnic minority and low SES populations were less prepared for disaster impacts and had limited resources to mitigate hurricane and flood hazards. This is important because better-prepared residents who were able to engage in pre-event mitigation had fewer physical health problems, post-traumatic stress symptoms, and adverse experiences (Grineski et al., 2020). This team conducted detailed analyses of disparate flood, toxic waste, and hazardous contaminant exposures during Harvey, linking these findings to pervasive social vulnerability (Chakraborty et al., 2021; T. W. Collins et al., 2019; Flores, Castor et al., 2021). During Winter Storm Uri, Grineski, Collins et al. (2023) documented social disparities in the duration of power and piped water outages throughout Texas. Flores et al. (2020a) documented disparities in physical and mental health and health care access among Houston-area residents following Hurricane Harvey, prompting them to prioritize the need to ameliorate public health disparities resulting from climate change–related disasters. The team has published extensively on the intricate relationship between social vulnerability and harms to population health in the Houston metro area (Flores et al., 2020b; Griego et al., 2020), leading to adverse event experiences that prolong recovery. Health System Performance Compounding these high rates of disease and risk behavior are limited access to health care and social services due to factors such as income, systemic racism and discrimination, and geographic location. These limitations delay treatment and increase the risk of poor health outcomes. Not only are GOM populations vulnerable to illness and chronic health conditions but their health systems are also vulnerable. Composite rankings for 2019–2021 across multiple 73 PREPUBLICATION | UNCORRECTED PROOFS 

measures of “health system performance”—including health care access, quality, use of services, costs, health disparities, reproductive care and women’s health, and overall health—rank GOM states in the lowest 30 percent, with most in the bottom 20 percent: Alabama (42nd), Louisiana (43rd), Texas (48th), and Mississippi (51st), lowest in the nation (Commonwealth Fund, 2023). Another indicator of health systems performance is an adequate health care workforce, assessed by Health Professional Shortage Areas (HPSA). HPSAs are defined as “geographic areas, populations, or facilities” that “have a shortage of primary, dental, or mental health care providers” (HRSA, 2023). HPSA calculates scores that identify areas of greatest priority in need of improved health care services and workforce availability, of which the GOM, particularly Texas, Florida, and Mississippi, have a sizeable amount (Rural Health Information Hub, 2024). During and oftentimes in the aftermath of a disaster, it becomes increasingly challenging to access social services, including services such as housing, food, and education provided by government, and private, profit, and nonprofit organizations for the benefit of the community and to promote social well-being (NASEM, 2019b). Health care services are also limited and people may deprioritize seeking care due to more pressing needs such as food, water, and/or shelter. Barriers also include insurance coverage and the availability of and access to culturally appropriate, high-quality care, including preventive, primary, specialist, dental, and vision care; chronic disease management, mental health treatment; and emergency services (NASEM, 2022c). These barriers can lead to increased morbidity and mortality rates. Similarly, people lacking access to mental health services or experiencing social isolation may face severe mental health crises. Disaster Mental Health The psychological footprint of a disaster is larger than its medical footprint—more people are affected psychologically than medically—because of the compelling nature of the event and the network of connections that extends geographically and socially beyond the scene of the event (Shultz, Espinel, Galea et al., 2007; Shultz et al., 2017). Populations exposed to disasters are known to experience elevated risks for psychological distress and psychopathology, including not only post-traumatic stress disorder (PTSD) but also other common mental disorders, including major depression, generalized anxiety disorder, and panic disorder, as well 74 PREPUBLICATION | UNCORRECTED PROOFS 

as increases in alcohol dependence and substance use (Pietrzak et al., 2012; Shultz et al., 2016). Further, preexisting mental health conditions can be exacerbated by the stress and trauma caused by a disaster, and even by the prospect of a disaster. Psychological consequences occur on a spectrum of severity, from distress to detrimental behavior change to diagnosable mental disorders, generally in relation to the severity of exposure to a hazard when it occurs and hardships in its aftermath (Davidson and McFarlane, 2006). It is well known that disasters result in high levels of stress, which is possibly the most persistent, harmful, and universal adverse health consequence of disasters. Allostatic load, which is a concept used to physiologically measure the cumulative burden of chronic stress and life events or the wear and tear on the body, contains elements that can help elucidate increasing exposure to disaster-induced stressors (Guidi et al., 2021; McEwen and Stellar, 1993; Rodriguez et al., 2019; Sandifer et al., 2022). Because pervasive toxic stress as a result of disasters can lead to adverse health effects in the aftermath of the event, Sandifer and colleagues (2022) developed a framework that combines aspects of psychosocial and physiological allostatic load to estimate its burden in people who have experienced disasters and other traumatic events. The framework can be used to gauge the short- and/or long-term health effects of disasters and to predict and mitigate related illnesses or conditions, and can be deployed as a disaster-focused human health– observing system. Psychological consequences are strongly dependent on the defining features of a disaster to which a person is exposed (i.e., the specific hazards, harms, losses, and changes, or combination thereof associated with the disaster; Shultz and Neria, 2013) as well as the collective consequences of and response to the disaster (Kirmayer et al., 2010). Psychological consequences also extend over a prolonged duration. When a disruptive event occurs, people may have traumatic experiences that for some may progress to PTSD. Disaster-related PTSD is unique in that it occurs in large population groups at once, often overwhelming community mental health care systems (Espinel, Kossin et al., 2019). Long after the event has occurred, psychological consequences can persist as a result of ongoing hardships; losses (of resources and of loved ones, which can lead to traumatic bereavement; life changes (e.g., displacement); and risks for depression and anxiety disorders (Davidson and McFarlane, 2006) and collective trauma. Collective trauma can undermine a collective sense of security and can extend into second and third generations of survivors (Hirschberger, 2018). It is described by 75 PREPUBLICATION | UNCORRECTED PROOFS 

sociologist Kai Erikson as a “blow to the basic tissues of social life that damages the bonds attaching people together and impairs the prevailing sense of community” (Erikson, 1976, p. 153). GOM residents, one study (Lorenzini et al., 2024, p. 3) notes, “experienced a cascade of collective traumas in recent years.” Current and ongoing explorations of compounding disasters provide opportunities to advance understanding of the field of disaster mental/behavioral health, a discipline that has evolved rapidly over recent decades. The GOM region has been a primary locale for exploration of how compounding disasters affect mental health. Studies have shown how frontline health professionals experienced exhaustion as they dealt with their own storm recovery while deploying to the sites of multiple other storms (Herberman Mash et al., 2013), and how successive hurricanes severely affected their sleep patterns and provoked hyperarousal responses (McKibben et al., 2010). Fullerton and colleagues (2013) found that 8 percent of the statewide public health workforce experienced new-onset, hurricane-related PTSD or depression, along with associated increases in both smoking and alcohol use. A dose-response relationship was evident. The likelihood of PTSD and depression increased with each additional storm that directly affected a worker’s home community and with each additional storm to which the worker was deployed. BOX 2-3 Compounding Disasters and Adverse Health Outcomes A study by Hahn and colleagues (2022) examines the relationship between recent and compounding disasters and the incidence of acute and chronic health outcomes among residents in 500 U.S. cities at the census tract level from 2001 to 2015. Communities that had recently (within 5 years) experienced a disaster reported higher incidences of high blood pressure and asthma and worse mental health than communities that did not recently experience a disaster. The incidence of these poor health outcomes increased by 1–2 percent for each additional year a community experienced a disaster. Seminal disaster mental health research derives from studies conducted on GOM storms, most notably starting with Hurricane Katrina. Galea and colleagues (2007) explored exposures to hurricane-related stressors during Katrina in relation to mental health outcomes. Of particular 76 PREPUBLICATION | UNCORRECTED PROOFS 

concern, treatments were disrupted for people with preexisting mental illness (Wang et al., 2008). In 2008, Hurricane Ike moved across the GOM, aiming directly for Galveston Island, and residents were strongly advised to evacuate to the mainland. Some did not, and they experienced the full force of the storm. Tracy and colleagues (2011) examined new-onset PTSD and depression in survivors of Hurricane Ike. Describing PTSD as “a disorder of event exposure,” the researchers found that those most likely to develop hurricane-related PTSD had stayed on Galveston Island, experiencing Ike’s wind and flooding hazards directly. The subset of Ike survivors who presented with clinically significant PTSD self-reported experiencing life threat (“I thought I was going to die”), losing a family member or close friend, or being physically injured in the storm. Stressors for depression included displacement from the home for a week or more, severe home damage, and significant financial hardship. Sophisticated geospatial analyses identified clusters of individuals with hurricane-related psychopathology, concentrated in particular on Galveston Island in the aftermath of Ike (Gruebner, Lowe, Tracy, Cerdá et al., 2016; Gruebner, Lowe, Tracy, Joshi et al., 2016). Cerdá and colleagues (2013) followed Hurricane Ike survivors to examine “post-disaster mental health” after the event in relation to the stressors experienced. Lowe et al. (2013); Lowe, Fink et al. (2015); Lowe, Joshi et al. (2015); and Lowe et al. (2016) conducted a series of analyses that also looked at poststorm stressors and use of mental health services in relation to measures of both common mental disorder symptoms and wellness indicators. These studies of survivors of GOM storms were pivotal in teasing apart the effects of direct traumatic exposure as distinguishing predictors of those who would be diagnosed with PTSD as a result of experiencing loss and life change, diagnoses that were closely tied to diagnoses of depression, generalized anxiety, and substance use. Climate change is predicted to increase the frequency and severity of hurricane- related mental disorders (Espinel, Galea et al., 2019; Espinel, Kossin et al., 2019). In 2010 the Deepwater Horizon disaster was a technological disaster involving release of a hazardous material (oil) on an expansive scale. Unlike a hurricane, this disaster posed more unknown and uncertain risks to residents. GOM researchers immediately began to document its effects on mental health. Osofsky and colleagues (2011) were among those at the forefront of this research, examining negative mental health effects for populations most heavily exposed. A 77 PREPUBLICATION | UNCORRECTED PROOFS 

concerning finding was that symptoms of depression in women increased over multiple time points, although symptoms of distress diminished (Rung et al., 2019). Given the unique and unexpected nature of the widespread population exposures due to the Deepwater Horizon disaster, Shultz and colleagues (2015, p. 58) conducted a “trauma signature analysis” of the event, with “psychological risk characteristics of this event includ[ing] human causation featuring corporate culpability, large spill volume, protracted duration, coastal contamination from petroleum products, severe ecological damage, disruption of Gulf Coast industries and tourism, and extensive media coverage”. Studies of residents in the GOM region have shown links between exposure to severe hurricanes and PTSD, psychological stress (Cohen et al., 2023; Raker et al., 2019) and its comorbidities, substance abuse, and suicide (Gradus et al., 2010; Ouimette and Read, 2014). Recent experiences with Hurricane Harvey became a focal point for research on the mental health of disaster-exposed populations. Karaye, Ross et al. (2019) and Karaye et al. (2020) examined self-rated mental health for Texas GOM residents who were exposed to Hurricane Harvey compared with persons without hurricane exposure. They found that GOM residents have poorer self-rated physical and mental health than the overall U.S. population. Analyses also showed the close interrelationship between repeated exposures to hurricanes and adverse effects on both mental and physical health. When comparing hurricane-exposed and nonexposed samples, those who had experienced a hurricane exhibited poorer self-rated mental health. Poorer mental health was reported by women, younger adults, mobile home residents, and persons with lower levels of educational attainment. Authors indicated that the self- assessment findings support the “enhanced provision” of mental health services coupled with educational programs and economic supports. Using the same 12-item Short Form (SF-12) Health Survey as Karaye, Ross et al. (2019), Sansom and colleagues (2022) found progressively lower scores on mental health on the SF-12 in relation to increasing numbers of direct hazard exposures over the past 5 years. Relatedly, an international team of investigators (Li et al., 2021) conducted exploratory analyses on the role of neighborhood green space on coping with mental distress—and mitigating PTSD—following exposure to Hurricane Harvey. Perceived higher quality of green space, mediated through emotional resilience, was associated with lower PTSD. 78 PREPUBLICATION | UNCORRECTED PROOFS 

Grineski et al. (2022) and Grineski, Scott et al. (2023) examined PTSD symptoms (anxiety and depression) in relation to the “cascading disasters” of Winter Storm Uri during the peak of the COVID-19 pandemic. A poststorm preliminary analysis of the data (Grineski et al., 2022) found an incidence rate of 18 percent. Uri-associated, new-onset PTSD was more common in minority race/ethnicity respondents than in White, non-Hispanic respondents. Guided by a “cascading disaster health inequities” approach, Grineski, Scott et al. (2023) examined anxiety (using GAD-2) and depression (using PHQ-2) 7 in eight Texas metropolitan areas 6 months after Winter Storm Uri. Experiencing more adverse events and living with a disability increased odds of depression and anxiety. The detailed analyses present a more nuanced picture. First, minority racial/ethnic status was associated with greater odds of depression, but not anxiety. Second, a series of “doubly impacted” subsets—those who were Black and disabled, Hispanic and disabled, disabled and Uri-impacted, Black and Uri-impacted, or Hispanic and Uri-impacted—experienced elevated odds of depression compared with the applicable “doubly privileged” reference category. COVID-19 Impacts on Frontline Responders A number of studies were conducted that examined the mental health impacts of working on the frontlines during the COVID-19 pandemic. As a globally declared pandemic and a U.S. national public health emergency, first responders, hospital-based “first receivers,” and public health preparedness coordinators were called upon to lead aspects of the COVID-19 response. Mendez and Horney (2023) conducted qualitative interviews to examine the nature of the COVID-19 response from the vantage of Emergency Medical Services personnel in Texas. The pandemic set in motion a novel set of personal and professional stressors that included fear of transmission to friends and family, increased workloads, operational changes, and fatigue, along with impediments associated with their usual repertoire of coping skills. Authors indicate that there is a need to prioritize and implement evidence-based interventions to safeguard the health and well-being of frontline response personnel. 7 GAD-2 is an initial screening tool to identify symptoms of “Generalized Anxiety Disorder.” PHQ-2 or “Patient Health Questionnaire-2” is an initial screening tool to identify symptoms of depression. 79 PREPUBLICATION | UNCORRECTED PROOFS 

Pfender and colleagues (2022) conducted survey research to explore the dynamics of anxiety and depression among public health workers charged with responding to COVID-19. In a pandemic, it is the public health workforce that operates the frontlines. Investigators noted the dearth of knowledge about the prevalence of anxiety and depression within the public health emergency preparedness workforce and possible applications of social support to safeguard these workers. This cross-sectional survey revealed high rates of anxiety (40 percent) and depression (29 percent), and symptoms and diagnoses associated with burnout and suicide among frontline personnel. These findings require rapid implementation of robust supports at both the individual and organizational levels. Medically High-Risk Patients Medically high-risk patients (MHRPs)—those whose health conditions require them to have ready access to health services, systems, and often social services—face elevated risks from hazards (Balbus and Malina, 2009; Espinel et al., 2022; Heagele and Pacquiao, 2019; Shultz et al., 2024). They include people who overtly qualify as having a disability, such as those living with spinal cord injury (Shapiro et al., 2020), and whose health conditions are characterized by functional impairments (Kruger et al., 2018; Mitra et al., 2022). Medically high-risk patients also include individuals with chronic diseases who generally do not define themselves as disabled but whose medical conditions require a regular, dependable connection to critical medical treatments, such as cancer therapies (Nogueira et al., 2020; Shultz et al., 2024). MHRPs also include individuals whose special medical needs are transient but risk elevating in times of disaster. Examples include pregnant persons and those in postsurgical recovery. The physical, psychological, and socioeconomic challenges associated with living with chronic medical conditions can make it particularly difficult for medically high-risk patients to cope with the added stressors experienced in a disaster (Espinel et al., 2023). Medically high-risk patients are at disproportionate risk for physical harm, psychological distress, and disruption of their care systems during disasters and other extreme events. They are also at elevated risk for such environmental hazards as extreme heat, humidity, heavy precipitation, and pollution. Since access to health services and supplies is frequently disrupted when a disaster occurs (Adams et al., 2019; Alnajar et al., 2021; Balbus and Malina, 2009; Espinel et al., 2022; Kruger 80 PREPUBLICATION | UNCORRECTED PROOFS 

et al., 2018; Mitra et al., 2022; Shultz et al., 2024), medically high-risk patients are likely to experience interruption of care routines, aggravation of symptoms, increased health care needs while evacuating and sheltering, increased susceptibility to injury, increased risk of cardiorespiratory events, and elevated stress levels. For example, exposure to hurricane hazards was associated with higher mortality among patients with lung cancer whose radiotherapy treatments were disrupted (Nogueira et al., 2019) and patients with end-stage kidney disease whose hemodialysis treatments were delayed (Blum et al., 2022), compared with patients who were not exposed. Yet although medically high-risk patients have special needs during a disaster, they tend to be underprepared compared with their healthy, disease-free counterparts (Adams et al., 2019). The specialized needs of medically high-risk patients in the face of compounding disasters are inadequately incorporated in disaster planning, preparation, and policies. On a broader scale, the subset of medically high-risk patients with chronic and life- threatening diseases who must be safeguarded during a disaster is particularly concentrated in the GOM region (Achenbach et al., 2023). Even in the absence of compounding disasters, life expectancies in the GOM region declined steeply during the COVID-19 pandemic. Life expectancies have been extremely slow to rebound postpandemic (Achenbach and Keating, 2023), and markedly so in the GOM region. This means, in part, that medically high-risk patients in the region are at particular risk due to (1) the burden of chronic, life-threatening disease; (2) continuing disproportionate risk of severe illness and death from COVID-19; and (3) elevated risk during compounding disasters when the health care and social services and support systems, on which so many with chronic diseases rely for their survival, are disrupted. Specific to the GOM region, Chakraborty and colleagues published a series of papers outlining elevated risks for “people with disabilities” during Hurricane Harvey (Chakraborty, Grineski et al., 2019), Winter Storm Uri (Chakraborty et al., 2023), and the COVID-19 pandemic (Chakraborty et al., 2024). Chakraborty, Grineski et al. (2019) make a series of compelling points. First, people living with disabilities are vulnerable to disasters, yet environmental justice studies have generally not focused on this subpopulation of MHRPs. Second, disabled populations were overrepresented in Hurricane Harvey–flooded areas, and people with cognitive and mobility impairments were especially at heightened risk. Third, the authors advocate for inclusion of persons with disabilities in disaster planning and environmental justice research. 81 PREPUBLICATION | UNCORRECTED PROOFS 

During the massive power outages that accompanied Winter Storm Uri (Chakraborty et al., 2023), people with disabilities, and most notably those living in federally assisted rental housing, sustained more severe and prolonged utility service disruptions, colder temperatures, and delayed recovery compared with nondisabled people—experiences that jeopardized their health. Authors call for the formulation of remedies that provide equitable protections. Detailed analyses (Chakraborty et al., 2024), revealed marked disparities in adverse impacts of the COVID-19 pandemic by disability status in metropolitan Texas in five areas of life: mental health, physical health, living conditions, health care access, and social life. Additional disparities were uncovered when analyses were disaggregated by disability type. Respondents experiencing cognitive and independent living difficulties were the most disadvantaged by COVID-19 in all five areas of life. Authors once again emphasize the need for increasing research and for enacting disability-inclusive policies to protect MHRPs. Social Capital and Cohesion The multidimensional concept of social capital can be defined as “features of social organization, such as networks, norms, and trust, that facilitate coordination and cooperation for mutual benefit” (Putnam, 1993, p. 2). Social capital evolves through relationships between people, and between people and institutions (e.g., decision-making bodies, elected officials, emergency managers, nonprofit organizations) and is linked to the size and type of an individual’s personal networks as well as other types of capital possessed by individuals in those networks, (i.e., financial, educational, cultural, and experiential resources; Bourdieu, 1985; Elliott et al., 2010). These relationships are central to a well-functioning and resilient society (NASEM, 2021). Within communities, social capital constitutes the social connections formed between and among community members (Putnam, 1993; Putnam and Goss, 1995). Social capital is sustained through social memory (Adger et al., 2005; Colten et al., 2012) and activated via social and civic networks (Aldrich, 2012; Putnam, 1993). These networks can help individuals cope with stress and trauma both in daily life and in times of a disaster. Given the likelihood for increased compounding climate hazards, broader support, including financial (Elliott et al., 2010) and mental health services (Ritchie and Long, 2021), is needed, if social capital is going to be effective for all populations in times of disasters to build adaptive capacity. 82 PREPUBLICATION | UNCORRECTED PROOFS 

Social capital also fortifies health at the individual level through access to resources and information and functional support (e.g., borrowing money; Clay and Abramson, 2021). Research has documented the linkages among compounding climate hazards, social capital, and community health resilience (i.e., the ability of a community to withstand, adapt to, and recover from public health risks; Roque et al., 2022). Social cohesion is “the glue that bonds people to one another, in families, groups, organizations, and communities” (Rodin, 2014, pp. 61–62). It builds shared values, reduces community disparities, and can help communities thrive by fostering well-being through safety, social inclusion, self-determination, and preservation of cultural and traditional ceremonies. Being able to physically gather together is an important aspect of social capital and cohesion, as well as a major source of community health resilience and recovery in the wake of a disruptive event. Compounding disruptive events can make gathering together, and hence social cohesion and access to local social support networks, far more difficult by reducing physical connectedness; they can also disrupt intergenerational knowledge transfer through increased outmigration (Roque et al., 2022). Social capital and cohesion, while intangible and difficult to quantify, are critical in all phases of a disaster (Broderick, 2023). These connections and networks are the infrastructure within which trusted relationships and solidarity among community members, or social cohesion, are nurtured (NASEM, 2024). During and after a disruptive event, many people report turning to their social networks for assistance (Adger et al., 2005; Aldrich, 2012) in well-being and livelihood support regarding safety, social inclusion, self-determination, and maintenance of traditional ceremonies. These networks warn their neighbors of impending disruptive events; lend assistance in their wake; and assist with rebuilding even in the absence of formal, government-sponsored assistance (Airriess et al., 2008; Aldrich and Meyer, 2015; Colten et al., 2012). They enable informal arrangements to direct mutual aid, foster creative problem solving, deliver trusted risk communications, and facilitate swift responses to emerging challenges. This is a common, long-standing response in GOM communities (Colten et al., 2015). The role of social cohesion in disaster phases has received attention, but its influence on the speed of postdisaster recovery remains understudied (Bergstrand and Mayer, 2020; Elliot et al., 2010; Sobhaninia, 2023). One recent study (Sobhaninia, 2023) attempts to compare social 83 PREPUBLICATION | UNCORRECTED PROOFS 

cohesion and recovery rates in four different Puerto Rican communities affected by Hurricane Maria (in 2017), which was experienced by more than 3.4 million people. More than 200,000 individuals were displaced or left their homes, potentially impairing their social networks. With multiple hurricanes impacting the island since 2017, research showed that social cohesion plays a large role immediately after a disaster and provides many benefits, but that over time, its effect in advancing disaster recovery diminishes. Social cohesion may also be a double-edged sword in certain instances, with some pointing to the two-faced nature of certain types of social capital (Aldrich, 2012; Aldrich and Meyer, 2015; Howell and Elliott, 2019; Priest and Elliot, 2023; Smiley et al., 2018). The close ties of social capital can also perpetuate collective trauma and prolong recovery (Priest, 2023) as people tend to experience disasters not as isolated individuals but as members of localized social networks (Priest and Elliott, 2023). As discussed in this chapter, disasters have a profound effect on the mental health of community members. In the aftermath of a disaster, social capital and cohesion can become even more important as community members deal with acute individual and collective stress during disaster response and recovery. Social capital can also provide social and material resources that shield against mental health stressors associated with both chronic and acute disruptive events (Torres and Casey, 2017). People with weaker support systems, such as those who are socially isolated or lack a close-knit community, may experience particular struggles with emotional and mental health challenges. Neighborhood and Built Environment Codes and standards are intended to minimize the vulnerability of buildings and infrastructure, but they face challenges in practice. First, these requirements are generally not retroactive, meaning that older structures, even if well maintained, do not comply with the latest technical guidance on hazard-resistant design, including evolving hazard characteristics. Because the U.S. building inventory is slow to turn over and building code reform remains a contentious political issue, the majority of buildings and other infrastructure are not designed to the latest construction standards. The second challenge is the evolutionary nature of hazards, which is particularly problematic in an era of climate change. Codes and standards are statistically regressive in 84 PREPUBLICATION | UNCORRECTED PROOFS 

nature, reflecting an attempt to predict the demands posed by various climatological stressors based on historical data. This means that buildings and other infrastructure designed to the latest technical guidance do not anticipate how the demands of even routine hazards (e.g., tropical storms in the GOM) may intensify over the lifetime of those structures. Furthermore, codes may not consider or may discount “gray swan” events—those with significant consequences but low probability of occurrence. It is unclear, in fact, that science presently has the ability to project future hazards reliably (Williams, 2023). This pervasive vulnerability to an evolving hazard landscape is only compounded by issues unique to how specific building classes and infrastructure is designed, built, and maintained. Public Infrastructure The vulnerability of public infrastructure has been well established by the assessment and advocacy efforts of the American Society of Civil Engineers, which publishes national- and state-level Infrastructure Report Cards, presented for five GOM states—Alabama, Florida, Louisiana, Mississippi, and Texas—in Table 2-3. In short, Alabama, Florida, and Texas have overall infrastructure grades in the C or “mediocre” range, with “general signs of deterioration” and “significant deficiencies in conditions and functionality, with increasing vulnerability” for some elements (ASCE, 2021a, 2021b, 2022b). Meanwhile, Louisiana and Mississippi are graded in the D or “poor, at risk” range, “with many elements approaching the end of their service life.” In these GOM states, issues of “significant deterioration” and “condition and capacity” are prevalent, “with a strong risk of failure” (ASCE, 2017, 2020). The specific infrastructure sectors driving these grades, listed in Table 2-3, highlight glaring vulnerabilities in many of the systems intended to manage the use of, and mitigate risks associated with, water—a critical asset and hazard for GOM states. TABLE 2-3 Overall Infrastructure Grades from the American Society of Civil Engineers (ASCE), with Grades by Sector Infrastructure Alabamaa Floridab Louisianac Mississppid Texase Element 85 PREPUBLICATION | UNCORRECTED PROOFS 

Overall C- C D+ D+ C Bridges C+ B D+ D- B- Dams Unknown D- C+ D D+ Drinking C- C D- D C- water Energy B C+ N/A C B+ Inland D N/A D- D N/A waterways Levees N/A D+ C D D Roads C- C+ D D- D+ Wastewater D C C- D D a ASCE, 2022b. b ASCE, 2021a. c ASCE, 2017. d ASCE, 2020. e ASCE, 2021b. These deficiencies are associated largely with underinvestment in the maintenance and replacement of aging infrastructure, particularly those elements that have deficiencies, have reached the end of their service life, or have capacities insufficient for current demands. Underinvestment by state and local authorities illustrates the direct link between poverty and physical vulnerability. 8 Steady growth in population and industrial and manufacturing industries has only further strained the capacity of many aging GOM infrastructure systems and has left states such as Texas struggling to maintain the sheer volume of infrastructure its population 8 A subtext of infrastructure vulnerabilities is “do more with less.” Public infrastructure is funded by tax dollars. Many localities and states lack the revenue to complete the necessary inspection, maintenance, and upgrades. A recurring theme is the gas tax, used to maintain roads and bridges. These taxes have been frozen for decades in some states because of the low economic capacity of citizens, even though inflation has doubled the cost of doing business during that time. Meanwhile, demands are rising, with more people and more commerce in the GOM region during that same time, and now even greater demands from a changing climate. Governments are thus straining to do more with half the funding today, a situation that creates deferred maintenance and infrastructure that is well past its design life, posing safety threats and increasing vulnerability to future hazards. 86 PREPUBLICATION | UNCORRECTED PROOFS 

demands. Box 2-4 provides an example of underinvestment in infrastructure in Texas and the vulnerability caused by such underinvestment. BOX 2-4 The Texas Power Grid and Winter Storm Uri In 2021, the vulnerability of the Texas power grid came into stark relief during the blackouts that occurred during and after Winter Storm Uri. While Uri’s extreme cold temperatures were uncommon in Texas, they are not entirely unprecedented. Estimates show that a 1989 cold snap in Texas would have exceeded the per capita energy demand of the 2021 blackouts (Doss-Gollin et al., 2021). Extreme cold events in Texas in 1951, 1962, and 1983 are also estimated to have resulted in about 90 percent of the per capita energy demand for heat as the 2021 blackout. These examples highlight the need for a power grid that is more resilient to extreme weather events (Doss-Gollin et al., 2021). (For additional discussion on the realized impacts of this widespread power grid failure in Texas, including the death toll, see Chapter 3.) GOM states rely on their natural waterways and coastlines for commerce, which leads to increased concentration of bridges. The high vulnerability of these infrastructure networks not only places local economies at risk but also heightens the vulnerability of neighboring communities to floods and isolation. Finally, the public wastewater system vulnerabilities across the GOM region are notable, creating a significant threat to public health, as well as the potential for environmental impacts. Private water systems can also have vulnerabilities, as private wells and septic systems are at increased risk of microbial contamination during disasters (Gitter et al., 2023; Pieper et al., 2021). Building Inventory Building inventory provides a critical complement to public infrastructure in enabling economic activity; facilitating the delivery of education, health care, and government services; and housing the population. Like public infrastructure, building inventory (publicly and privately owned) has pervasive vulnerabilities driven by deferred maintenance and replacement of older buildings and the incompatibility of existing buildings with evolving hazard demands. Unlike infrastructure, however, buildings shelter and protect human life, and most are privately owned, 87 PREPUBLICATION | UNCORRECTED PROOFS 

creating additional challenges in addressing vulnerabilities to damage. The primary mechanism for doing so at scale—building codes—is based on a life safety approach aimed at maintaining affordability at the expense of potentially substantial losses from a disruptive event. Facilities such as housing are not intended for sheltering-in-place or being rapidly reoccupied after a disaster in which they are likely to sustain damage. Thus, while lives may not be lost, particularly if populations evacuate promptly, the physical damage to these buildings will disrupt functionality, displace populations, strain insurance and recovery assistance programs, and exert a significant financial and emotional toll on building owners. The potential for mounting losses is only heightened in the GOM region by inconsistent adoption and enforcement of the latest model building codes. As noted in the Insurance Institute for Business and Home Safety (2018) report Rating the States: 2018. An Assessment of Residential Building Code and Enforcement Systems for Life Safety and Property Protection in Hurricane-Prone Regions, Atlantic and Gulf Coast States, while Florida has led the nation in advancement of building codes to limit hurricane losses, earning the top score of 95/100 in their rating system, other GOM states such as Texas, Alabama, and Mississippi were among the lowest-rated states nationally immediately preceding 2020–2021, largely as a result of the lack of adoption and enforcement of mandatory codes statewide. Even when adopted, codes are generally not regressive, governing only new construction and major renovations. Thus, while Louisiana made great strides in adopting and enforcing the latest building codes after Hurricanes Rita and Katrina, the building inventory is slow to turn over, so the vast majority of existing buildings—including those exposed to the weather-climate events of 2020–2021—were likely not constructed in accordance with modern building codes. This is particularly significant for GOM states, where population and housing booms during 1950–1960 resulted in what is now an older housing inventory. Table 2-4 summarizes the percentage of housing units likely built to modern building codes in five GOM states. Florida is a recognized leader in the early adoption and enforcement of a statewide building code, leading to nearly 30 percent of its housing inventory likely built to model codes. Louisiana started enforcing modern codes after Hurricanes Katrina and Rita, so a larger proportion of its inventory is built to modern standards, but that proportion is still less than 5 percent. In contrast, Texas has undergone rapid growth in recent decades, resulting in over 13 percent of its housing inventory being built after 2010, when modern codes began to penetrate 88 PREPUBLICATION | UNCORRECTED PROOFS 

coastal and metro areas despite the lack of statewide mandated code adoption and enforcement. Note that Alabama does not enforce statewide building codes but references the International Residential Code, so estimates in Table 2-4 should be viewed with caution; the situation is even more dire in Mississippi, where no homes could be reliably assumed to be built to modern codes. TABLE 2-4 Percentage of Housing Units Built in a Single Year and in the Modern Code Era in Alabama, Florida, Louisiana, Mississippi, and Texas Alabamaa Floridab Louisianac Mississippid Texase (%) (%) (%) (%) (%) Built in a Single Year 2.25 2.90 1.91 1.66 3.66 (2020–2021) Likely Built to 8.24 29.96 4.67 0.00 13.08 International Building Code/International Residential Code (IBC/IRC) a Alabama does not have a mandatory statewide building code system. Since state codes reference the 2015 IRC for voluntary adoption, and coastal counties and major metropolitan areas were likely building to such standards, the census reference category “built 2010–2019” through “2020 or later” are adopted therein, assuming half the properties the first term met the new standards. b The Florida Building Code, or FBC, was enacted in 2002 and draws upon national model building codes and national consensus standards, with modifications for the Floridian context. Therefore, the census reference categories “built 2000–2009” through “2020 or later” are adopted therein, assuming 80 percent of the properties the first term met the new standards. c The Louisiana State Uniform Construction Code, or LSUCC, referencing the IRC became effective in 2007 with some exclusions and modifications; thus, the census reference categories “built 2000–2009” through “2020 or later” are adopted therein, assuming 20 percent of the properties the first term met the new standards. d Mississippi adopted a building code law in 2014 that governs construction of most residential buildings, but with opt-out provisions. As a result, adoption and enforcement are currently uneven across the state. e In 2001, the Texas legislature adopted the 2000 IRC as the standard for residential construction; however, the state does not mandate its adoption and enforcement. Given the variable uptake and delays in enforcement in coastal and metro areas that have adopted modern codes, census reference categories “built 2010–2019” through “2020 or later” are adopted therein. SOURCE: U.S. Census Bureau, 2022b. 89 PREPUBLICATION | UNCORRECTED PROOFS 

Housing While the underlying vulnerabilities discussed above affect the entire building inventory, they are particularly salient for housing. Housing is an oft-neglected though central pillar of societal infrastructure. Stable, long-term housing is a foundational element for physical, social, and economic well-being, both individual, and collective (NASEM, 2018) and can provide a sense of belonging in a community. Housing has also taken on new roles since the pandemic, enabling productive activities such as e-learning, telehealth, and telework to continue during lockdown periods (Mendonça et al., 2023)—activities that were reinstated when schools, businesses, and health care facilities were physically damaged in disasters. Thus, access to affordable and hazard-resistant housing is essential to the existence, flourishing, and resilience of communities. Beyond the aforementioned issues of an aging housing inventory, the nation was plagued by a crisis in affordable housing well before 2020, which was exacerbated in the GOM region by the extreme weather-climate disasters and pandemic that subsequently occurred. Research suggests that disasters “disproportionately affect poor households living in exposed areas, often leading to displacement” (Norwegian Refugee Council, 2019, p. 41). This phenomenon is driven by a lack of safe, affordable housing options that forces lower-income populations into older or deferred-maintenance properties that are more likely to sustain losses even in more frequent, lower-intensity events such as Hurricane Sally or Winter Storm Uri. The intersection of socially vulnerable households living in physically vulnerable housing creates vast segments of the GOM population who are especially ill equipped to recover from compounding disasters. Despite its vital role, housing is treated like a commodity: nonengineered, privately developed, minimally regulated after its initial construction, and seen by many as a pathway to building wealth rather than core infrastructure that all residents have a right to access. However, physical vulnerabilities within the built environment and the infrastructure supporting it (e.g., water networks) are the primary drivers of material losses in disasters. Although housing recovery is certainly challenging for all disaster survivors who sustained housing damages or loss altogether, communities of color residing in lower-income areas tend to sustain more damage and recover more slowly compared with those in higher- income areas (Lee and Van Zandt, 2018; Peacock et al., 2014; Smiley et al., 2018). Moreover, 90 PREPUBLICATION | UNCORRECTED PROOFS 

while it is difficult to quantify the loss of cultural identity or cultural traditions important to populations, recovery processes typically overlook and undervalue historical significance and community cultural identities in favor of rebuilding newer, up-to-code, or safer alternatives for wealthier individuals. The damage to housing stock resulting from a disaster opens up new opportunities for redevelopment or “climate gentrification” (NASEM, 2022c), which often excludes or “outprices” generational community residents. The physical loss of housing after a disaster drives up the cost of housing, rent, insurance, and property taxes. It also can contribute to loss of mental health stability, safety, sense of and attachment to place, and community cohesion, affecting all sectors of society. Homeownership and a Legacy of Discriminatory Housing Practices In the GOM region, as in other parts of the United States, the legacy of discriminatory federal housing policies and practices relegated many Black Americans and other minority residents to neighborhoods and regions with disproportionate exposure to climate and industrial hazards and environmental contaminants. Practices broadly known as redlining, 9 which was not outlawed until the Fair Housing Act (42 U.S.C. 3601 et seq.) was passed in 1968, included racial (or exclusionary) zoning 10 where, as early as 1910, local governments in GOM states with large Black populations (and in many other U.S. cities) instituted racial zoning ordinances to preserve segregation where it existed (Gray, 2022, Seicshnaydre et al., 2018), restricting where Black and other potential homebuyers of color could purchase homes. The exclusion of opportunities for federal housing loans, 11 resulted in households of color receiving only 2 percent of all government-backed mortgages between 1934 and 1962 (Swope and Hernández, 2019). 9 Prior to 1968, the Federal Housing Administration (FHA) did not insure mortgages in and near low- income urban neighborhoods where the vast majority of urban Black Americans lived (Fishback et al., 2022). Concurrently, the FHA subsidized construction of subdivisions for White Americans with the requirement that none of the houses be sold to African Americans (Rothstein, 2017). 10 Prior to the Supreme Court’s 1917 Buchanan v. Warley decision, city zoning ordinances across the United States legally forbade minorities from residing on blocks where a majority of residents were White (Rigsby, 2016). 11 The Federal Housing Administration, established in 1934, enacted federal loan underwriting guidelines that explicitly supported racially homogeneous, suburban developments, effectively subsidizing loans for White homeowners, while systematically excluding racially mixed or predominantly minority neighborhoods that largely prevented homeownership opportunities and access to federally insured mortgages to non-White residents and expressly contributing to the concentration of poverty in urban areas (Taylor, 2019). 91 PREPUBLICATION | UNCORRECTED PROOFS 

These discriminatory practices resulted in long-term disinvestment and stigmatization (Keene and Blankenship, 2023) of these regions and neighborhoods that have residual effects into the present-day, including residence in areas more prone to disruptive events, entrenched racial segregation, generational poverty, an inability to reside in safer areas, and overall housing affordability challenges, which disproportionately affects communities of color, compounding existing inequities in wealth and opportunity (Taylor, 2019). In combination with existing patterns of segregation, redlining has resulted in “indelible patterns of social and environmental inequalities” (Grove et al., 2018, p. 524) with sweeping implications for health outcomes, including higher rates of cancer, asthma, poor mental health, and people lacking health insurance (Nardone et al., 2021) as well as residence in high-risk floodplains and more physically vulnerable structures (Bidadian et al., 2024) and reduced present-day greenspace (Grove et al., 2018), which all can be exacerbated in compounding disasters. The disruptive events of 2020–2021 largely reinforced the persistently unequal access to homeownership (Joint Center for Housing Studies, 2021). Displacement from Home Disaster-related housing shortages prompt both short- and long-term population displacement. As populations are displaced, the loss of a viable tax base can remove economic revenue that further erodes communities lacking intentional investment in recovery. When disasters destroy subsidized housing in disaster-prone areas, the failure to rebuild leaves displaced residents without viable options for recovery beyond relocation (Mehta et al., 2020). For example, the loss of government-subsidized housing replaced with condominium developments eliminates the possibility of recovery in that area for those who are displaced, contributing to community erosion and a new normal that is not representative of the historical significance of what has been lost. Whether the negative effects of redlining and flooding, inequitable insurance practices, inflation of property values, or policies and processes that result in loss of property, displacement, and relocation are significantly more common among vulnerable communities. 92 PREPUBLICATION | UNCORRECTED PROOFS 

Large-scale housing losses are often accompanied by the absence of (or a significant lag in) access to temporary housing close to the household’s place of work or children’s schools, forcing low-income families to choose among a set of undesirable options that include remaining in an otherwise uninhabitable damaged home, making onerous daily commutes, permanently relocating, living in a vehicle or overcrowded housing, or experiencing homelessness. Following the disruptions associated with Hurricane Katrina, assessments revealed that one-third of displaced children were at least one grade behind in school for their age (Baussan, 2015). For more information on well-being outcomes related to Hurricane Katrina, the Resilience in Survivors of Katrina, or RISK, Project, a longitudinal, multimethod, multidisciplinary study, focuses on “the recovery of people, rather than place.” The project has published more than 40 studies by following vulnerable individuals and their families if/when they relocated from New Orleans after the storm (Waters, 2016). Yzermans and colleagues (2005) note that psychological symptoms persist for more than 2 years after a disaster, and the displaced are twice as likely to have such symptoms. Prolonged or permanent displacement, such as that created by experiencing compounding disasters has significant impacts on security, mental health, social connectedness, well-being, and more. Disasters interrupt society, and in some instances, cultures. Long-term displacement ranks among the highest of the societal costs of disasters. Baussan (2015) notes that “evidence indicates that the climate displaced, particularly those who are low income, can suffer from greater hardships than they did prior to evacuation” resulting from challenges with housing, employment, education, and health (p. 2). Among the many losses experienced by these populations is the loss of a sense of and attachment to place, as discussed in Box 2-5. BOX 2-5 Sense of and Attachment to Place The concept of place can be understood as a structure that includes both human experiences and the material world in which those experiences occur, or, as Casey (1997, p. 9) describes it: “an embodied experience—the site of a powerful fusion of self, space and time.” Place is related to physical and mental functions or “nested collections of human experience” (Hess et al., 2008, p. 468) 93 PREPUBLICATION | UNCORRECTED PROOFS 

that engender intimate and personal knowledge of living in a specific place. Identities are rooted in a sense of and attachment to place and are geographically connected to the places at risk (Adger et al., 2011) in the GOM region. As the physical landscape changes in a place (e.g., hurricane-damaged coastline) peoples’ identities can be deeply affected (Burley, 2010; Simms, 2017). Sense of place and place attachment are terms used to depict how people create, organize, and relate to places where social networks are rooted and both material and cultural flows coalesce. Postdisaster, sense of place can play a key role in evaluating the meanings of long-term community recovery, as it did in a study (Schumann, 2018) with coastal Mississippi residents after Hurricane Katrina that compared quantitative and qualitative techniques. The author concludes that place-based qualitative assessments can enrich quantitative analyses and, together, achieve more equitable, holistic recovery appraisals. Sense of and attachment to place can also assist residents in place (re)making after a disaster has changed their familiar physical surroundings (Zavar and Schumann, 2019). The dynamic nature of place attachment was highlighted in a study (Nelan and Schumann, 2018) that examined the interactions and limitations of informal and formal gathering places that emerged in the aftermath of Hurricane Harvey. Researchers found that, in part due to lack of formal mental health support resources, displacement of residents, and an influx of aid workers (many nonresidents), the gathering places were limited in their ability to foster a communal recovery among the residents. Renters and Rental Homes For renters, high demand for and low supply of housing are accompanied by a rising cost burden; between 2017 and 2022, rental prices increased at a rate faster than that of inflation (Schaeffer, 2022). Furthermore, between 2019 and 2021, the affordable housing shortage in the United States increased, with the number of affordable and available rental homes for extremely low-income renters 12 decreasing by more than 500,000 units, or 8 percent. Two GOM states, Texas and Florida, were among the top five states facing the most severe shortages (National Low Income Housing Coalition, 2023). Mississippi, while among the five states with the lowest relative shortage of affordable and available units, had the highest concentration of renters (27 12 Households with incomes at or below the federal poverty guideline or 30 percent of American Median Income, whichever is higher (National Low Income Housing Coalition, 2023). 94 PREPUBLICATION | UNCORRECTED PROOFS 

percent) in the United States who are behind in their housing payments and facing eviction in 2021. Louisiana is second with 25 percent (Joint Center for Housing Studies, 2021). Renters are particularly vulnerable to disaster-related housing disruptions as a result of past and ongoing discriminatory housing practices, increased hazard exposure, physical vulnerability of structures, and socioeconomic disadvantage (Best et al., 2023). Weather-climate events tend to reinforce patterns of racial and class inequalities that were in place before a disruptive event, both in direct effects to housing and on household financial stress (Brennan et al., 2021; Lee and Van Zandt, 2018; Rumbach and Makarewicz, 2017). Exclusionary practices and policies prioritize disaster aid for homeowners rather than renters, particularly wealthier ones. This has the effect of reinforcing discriminatory practices that lead to gentrification and isolation (Aune et al., 2020; Brand and Seidman, 2008; Ellen et al., 2016) and increases generational poverty, leaving communities with decreased capacity to recover from future disasters. After a disaster, renters often have difficulty finding permanent housing near their original home because of rent increases, while the cost of newly rebuilt permanent housing often exceeds the ability of low-income disaster survivors to pay (Fothergill and Peek, 2004). This pattern also applies to residents of public housing development, who are also disproportionately affected (Chakraborty et al., 2021, Lee and Van Zandt, 2018). When public housing is destroyed in disasters, there is not typically a previously agreed-upon action for permanent reconstruction or recovery of the units (Hamideh and Rongerude, 2018). The former residents often find themselves with a reduced ability to participate in postdisaster decision-making due to displacement and a diminished number of residents able to advocate for reconstruction of government-subsidized housing units. Alongside the stigmatization of public housing and a lack of economic incentive to rebuild public housing and (Hamideh and Rongerude, 2018), a failure to rebuild government-subsidized housing in disaster-prone areas removes viable options for recovery beyond relocation (Mehta et al., 2020). These conditions can exacerbate marginalization and leave residents more vulnerable to subsequent disruptive events. Evictions increase after disasters and are associated with significant disruptions for renters in the years following a disruptive event (Brennan et al., 2021). Accelerated gentrification, challenges in accessing federal disaster recovery assistance, lack of clear title to the land, and overall lower supplies and higher demands are contributing factors to increased 95 PREPUBLICATION | UNCORRECTED PROOFS 

evictions postdisaster (Brennan et al., 2021; Schuessler et al., 2022). Rental units that remain tend to be older, poorly maintained, and not built to the latest building codes. Weather-climate hazards such as hurricanes, tornadoes, or winter storms will disproportionately damage such subpar properties. Manufactured Housing Units Manufactured Housing Units (MHUs) and MHU parks are different from rented or owned houses in ways that can render the units and their inhabitants more vulnerable to climate- weather disruptive events. Not only are MHUs typically flood-exposed at higher rates than other housing (Tate et al., 2021), they are especially vulnerable to high winds (Gabriel, 2023). MHU parks are also largely divided-tenure communities wherein residents own their housing unit, but rent the land beneath, a setup that restricts the rights of park residents (Rumbach et al., 2020). The socioeconomics of the GOM region and the lack of affordable site-built housing have created increasing reliance on manufactured housing units. MHU parks, where many MHUs are situated, are the largest unsubsidized source of affordable housing in the United States (Sullivan et al., 2021). The populations residing in MHUs are more likely to have higher social vulnerability than the general population (Fothergill and Peek, 2004). Of the 200,162 new MHUs shipped to states in 2020 and 2021, 77,242, or 39 percent, were shipped to the five GOM states (U.S. Census Bureau, 2023a). The six counties/parishes that are a focus of this study (with the exception of Harris County) all have a higher concentration of MHUs compared with the average for the continental United States (Strader, 2023). Cameron Parish, Louisiana, has the highest concentration (29.7 percent in 2000, 38.1 percent in 2018), 7 times the average of the continental United States. Of all MHU parks across all six parishes/counties, 23.6 percent are within special flood hazard areas; 13 highest percentages among these are in Cameron Parish (71.4 percent) and Galveston County (47.5 percent). 13 Special flood hazard areas (SFHAs) are areas that “will be inundated by a flood event having a 1 percent chance of being equaled or exceeded in any given year (FEMA, n.d.). The 1 percent annual chance flood is also referred to as the base flood or 100-year flood. Furthermore, “lenders must require flood insurance on improved real estate or manufactured homes that are, or will be, located in a SFHA in communities that participate in the National Flood Insurance Program. [Federally] regulated lenders are prohibited from making, increasing, extending, or renewing any designated loan unless the real estate or manufactured home securing the loan is covered by flood insurance for the term of the loan (Tobin and Calfee, 2005, p. 43). 96 PREPUBLICATION | UNCORRECTED PROOFS 

Insurance As discussed in Javeline et al. (2021), mounting catastrophic risk has prompted the insurance industry to not only raise premiums and deductibles, but eventually to deny coverage and avoid underwriting policies in Atlantic and Gulf Coast states (Hartwig and Wilkinson, 2016). Residual markets have naturally emerged in response to these market pressures, e.g., Fair Access to Insurance Requirements (FAIR) Plans and Beach and Windstorm Plans (Kousky, 2011; Randlett, 2010), but their explosive growth in recent decades (Hartwig and Wilkinson, 2016) reveals that they have now become the primary insurance market rather than the “market of last resort” in many hurricane-prone states (Laird et al., 2017; Randlett, 2010). When coupled with the heavy regulation of premiums, the shift into these residual markets has left these plans highly vulnerable to a catastrophic hurricane, with the majority of the plans already recording at least one underwriting loss (Hartwig and Wilkinson, 2016), with similar concerns over sustainability expressed regarding the National Flood Insurance Program (NFIP) (Ahmadiania et al., 2019; Arnold, 2011, pp. 235–236; Craig, 2019). Critiques further argue that these government-subsidized wind/hurricane and flood plans are actually working in opposition to their stated goal of reducing losses in these storms (Craig, 2019; Union of Concerned Scientists, 2014). While the growth of residual insurance markets as the primary insurer in coastal zones is a clear signal of escalating risk, these states continue the process of “adverse selection” (Rowell and Connelly, 2012; Wagner 2019) by pooling insurers into coastal underwriting associations required to insure the highest risk properties, even repetitive loss properties, at actuarially low premiums inconsistent with their actual levels of risk (Randlett, 2010). The consequences in the GOM region have been significant. For example, Louisiana insurers paid out more than $23 billion during the historic 2020 and 2021 hurricane seasons, causing 11 major insurers to declare insolvency (Barry, 2023; Finch, 2022). Meanwhile, the uninsured and underinsured often look to FEMA assistance, which is not only insufficient and slow to reach survivors (Kousky et al., 2021) but has similarly been depleted by excessive payouts in recent years (Kimball, 2023). 97 PREPUBLICATION | UNCORRECTED PROOFS 

Economic Stability Economic stability ensures that individuals and families have access to financial resources, job security, healthy foods, safe housing, and health care, all factors that improve quality of life and health outcomes and reduce vulnerability to disruptive events. Median household income provides a metric for examining the distribution of wealth— and poverty. Median household incomes are generally low for families in GOM states. Mississippi has the lowest median household income in the nation, followed by Louisiana with the third lowest, and Alabama with the fifth. Among GOM states, only Texas has just above the average median household income for the nation (U.S. Census Bureau, 2021b). GOM states also lead the nation in the percentages of their populations living below the federal poverty line. Louisiana leads the nation with the highest percentage of persons in poverty (19.6 percent), followed closely by Mississippi (19.4 percent); Alabama ranks fifth at 16.0 percent (U.S. Census, 2023b). The five GOM states are also home to 123 of the nation’s 341 persistent poverty counties (counties with extended periods of high poverty rates). Mississippi has the highest percentage of any state (53.7 percent), encompassing 34.8 percent of the state’s population (U.S. Census, 2023c). People who reside in persistent poverty counties experience more acute and systemic problems, including limited or lack of access to medical facilities and services, healthy and affordable food, quality education, and civic engagement opportunities (U.S. Census, 2023c), all important resources in the event of a disaster. GOM states have larger populations of racial/ethnic minority populations, lower aggregate incomes relative to the national average, and a higher percentage living in persistent poverty compared with non- Hispanic White populations in those states (Elder et al., 2007; Islam et al., 2022). Disasters and Economic Instability and Inequity Economic instability in GOM communities has cascading and compounding effects when disasters occur. During the response and recovery phases of a disaster, individuals may struggle to continue their employment because of the lack of paid time off and transportation challenges; low wages, a lack of savings, and insufficient insurance may result in their never recovering 98 PREPUBLICATION | UNCORRECTED PROOFS 

financially. This result in turn is a resource-loss spiral that leads to prolonged stress and poorer health outcomes (Sattler et al., 2002); research shows that poverty rates increase in the year following damage from natural hazards (Cutter et al., 2014; Gotham and Greenberg, 2014; Smiley et al., 2018). Economic access to resources is one of the most significant factors related to inequitable recovery for frontline communities. Disaster recovery is arguably one of the best potential generators of improved outcomes given mitigation funding requirements, rebuilding processes, and insurance implications. However, the benefits of the mobilization of recovery funding rarely manifest among the vulnerable. Instead, disaster recovery processes often intensify vulnerabilities. These effects are further exacerbated by compounding and cascading events that place additional burdens on communities and continuously strain the capacity of struggling systems. A longitudinal study (Smiley et al., 2018) examining the interaction of U.S. disasters (1998–2015), poverty, and various types of organizations involved in disaster recovery efforts found that poverty tended to increase the most where the number of “bonding,” or more inwardly oriented organizations, increased. The robust, in-group solidarity these types of organizations rely on may unintentionally produce inequitable recoveries by restricting resources to group members while excluding members outside the group, particularly those most marginalized, and leaving them in comparatively worse situations than they may otherwise have been (Aldrich, 2012; Aldrich and Meyer, 2015; Howell and Elliott, 2019; Priest and Elliot, 2023; Smiley et al., 2018). Generally, current disaster recovery approaches fail to address the root causes of social vulnerability and instability; thus, they often exacerbate preexisting vulnerabilities in tandem with losses from the disaster itself. Disaster-related inequality can be uniquely characterized by the unequal distribution of resources, power, and economic opportunity across societies in the United States (Bowdler and Harris, 2022). This diversion of resources is a key contributor to disparities seen across vulnerable communities. Research shows that frontline communities, which are often, though not limited to, communities of color and low-income communities, are unlikely to benefit as much as their White, wealthier counterparts when qualifying for federal disaster assistance and home buyout programs (Billings et al., 2019; Emrich et al., 2022; Han et al., 2024; Loughran and Elliot, 2019; 99 PREPUBLICATION | UNCORRECTED PROOFS 

NPR, 2021). The Federal Emergency Management Agency also has reported that disaster funds are not reaching those most in need (FEMA, 2020). Of 4.8 million aid requests analyzed, the U.S. Government Accountability Office (U.S. GAO, 2020) found that ● the poorest renters were 23 percent less likely than higher-income renters to receive housing assistance; ● FEMA was roughly twice as likely to deny housing assistance to lower-income than to higher-income survivors because their damages were judged “insufficient”; and ● additional research indicates that people of color are less likely to receive adequate disaster assistance. Following Hurricane Harvey, 40 percent of racial/ethnic minorities were denied assistance (Mosbergen, 2017). Additionally, 45 percent of households with annual incomes of less than $15,000 were denied individual assistance or received an average payout of approximately $4,300 (Texas Housers, 2018). Box 2-6 provides details on another barrier to receiving FEMA disaster assistance. BOX 2-6 The 25 Percent Local Match Requirement Access to Federal Emergency Management Agency (FEMA) public assistance funds requires that a financial match of 25 percent of the requested amount be provided by the requesting locality. Many smaller (by population) and more rural communities often have less capacity to navigate the funding application process and provide the dedicated “local match” required to qualify for the FEMA funding (U.S. GAO, 2022). Further, these capacity limitations constrain a community’s ability to build the expertise needed to apply for funding that could help mitigate the effects of disasters and build adaptive capacity (Seong et al., 2021). Without adequate financial resources to support disaster recovery and mitigation, sensitivity to future disasters increases. Policies such as redlining, deed covenants, heirs’ property (see Box 2-7), flood zone designation, land-use planning and plans, credit score ratings, and discriminatory insurance practices have contributed to inequities. These inequities can be exacerbated during disaster 100 PREPUBLICATION | UNCORRECTED PROOFS 

recovery (Clark-Ginsberg et al., 2021; Domingue and Emrich, 2019; Finch et al., 2010; Finucane et al., 2020; Masozera et al., 2007; Whytlaw et al., 2021). BOX 2-7 Heirs’ Property: A Significant Disaster Recovery Policy Change Heirs’ property is property passed to family members by inheritance, usually without a legal document proving ownership. It is typically created when land is passed on from someone who dies without a will to those legally entitled to their property, such as a spouse, children, or other relatives. These policies have contributed to the dispossession of 4.7 to 16 million acres over the last hundred years (USDA, 2023), disproportionately affecting African American families throughout southern states and accounting for an estimated one-third of all African American–owned land in the United States. The U.S. Department of Agriculture reported that “the lack of title severely limits property owners’ ability to access credit, to sell natural resources, or to participate in land improvement programs offered by the Federal Government, resulting in land and wealth loss for affected families across the South,” adding that it contributed to a 90 percent decline in Black-owned farmland nationwide between 1910 and 1997 (Gilbert et al., 2002; USDA, 2017, p. vii). In Mississippi alone, the current value of the land lost in that period is roughly $6.6 billion (Deen, 2021). Until 2021, the Federal Emergency Management Agency’s (FEMA) policies required applicants for disaster aid to provide a deed or other formal proof of homeownership, which effectively excluded thousands of families seeking grants for repairs after disasters because of heirs’ property restrictions (Dreier and Tran, 2021). However, new administrative policies announced by FEMA in 2022 offer flexibility in demonstrating proof of property ownership (FEMA, 2022). As growth in population and the built environment leads to greater exposure along coastlines at risk of experiencing extreme weather-climate events, economic instability, and other vulnerabilities to these high-impact hazard events is also increasing (Frazier et al., 2010). The convergence of repeated exposure to disruptive events and heightened vulnerability results in prolonged impacts and elevated sensitivity to future disruptive events (Smit and Wandel, 2006). 101 PREPUBLICATION | UNCORRECTED PROOFS 

SUMMARY OF KEY FINDINGS The GOM region has a well-documented history of experiencing extreme weather- climate events, disruptions, and disasters. Scientific studies suggest that occurrence of rapidly intensifying and stronger tropical cyclones and bouts of extreme temperatures are increasing as a result of climate change. Shifting climate conditions will likely increase the frequency and type of hazards that GOM populations are exposed to, affording less time to recover before the occurrence of additional disruptive events. Exposure to extreme temperatures (heat and cold) are the deadliest weather phenomenon in the United States, though these hazards receive insufficient dedicated federal support and funding for planning, education, mitigation, and recovery. Evidence shows that many GOM residents experience high levels of social and physical vulnerability. GOM populations experience a range of poor health outcomes and health disparities and are marginalized in terms of income, education, age, ethnicity/race, political power, gender, sexuality, and/or being medically high-risk. Many residents are disproportionately exposed to environmental contaminants and toxins, which exacerbate negative physical and mental health outcomes under blue-sky conditions and also increases risks to human health when hurricanes, flooding, and other weather-climate events mobilize these toxins into communities. Vulnerabilities also extend to the built environment of communities. Codes and standards are statistically regressive in nature, reflecting an attempt to predict the demands posed by various climatological stressors based on historical data. This means that even buildings and other infrastructure designed to the latest technical guidance do not anticipate how the demands of routine hazards (e.g., tropical storms in the GOM) may intensify over the lifetime of those structures. Vulnerabilities and exposure combine to produce a community’s sensitivity to experience the occurrence of one or more disruptive events as a disaster that exceeds its capacity to absorb, respond, and recover. Communities with high vulnerability and exposure are more likely to experience long-term adverse effects of compounding disasters. Current disaster management and recovery practices fail to address the root causes of vulnerability and instability; thus, they often exacerbate preexisting vulnerabilities in tandem with the losses and damages directly associated with the disaster itself. Strengthening the adaptive capacity of the systems and 102 PREPUBLICATION | UNCORRECTED PROOFS 

functions that underpin communities can mitigate the potential for harm by reducing sensitivity to a disruptive event. 103 PREPUBLICATION | UNCORRECTED PROOFS