This chapter examines a number of factors identified in the conceptual framework in Chapter 2 as affecting whether and how climate events can lead to significant disruptions in societies or political systems. Understanding these factors—exposure and susceptibility to harm from hazards plus coping, response, and recovery after disruptive events occur—has long been a central concern of several communities of scholars and policy makers. For example, researchers and practitioners concerned with natural disasters have developed concepts, measurement methods, and data to improve understanding of disaster preparedness and response and of the social and economic consequences of floods, wildfires, storms, and various other climate-related events. Researchers and practitioners in such areas as public health, water management, famine and food insecurity, and humanitarian relief have also developed knowledge that is relevant for understanding how climate events may lead to significant social and political consequences. Knowledge from these fields is critical for understanding how and under what conditions climate events can disrupt societies. It has also been central to progress made in understanding vulnerability and adaptation to climate change in the work of the Intergovernmental Panel on Climate Change and the U.S. Global Change Research Program. In this chapter, we briefly review knowledge and insights from these fields that are relevant to security analysis and also review the implications for assessing the potential of climate change to influence social and political stresses.
We begin by noting that climate events can be socially disruptive both where they occur and elsewhere. The local disruptions are familiar: Storms, floods, heat waves, droughts, and the like have their most obvious consequences where they occur. However, as we noted in Chapter 2, climate change can result in events that create shocks to globalized systems that support human life and well-being and that can therefore affect populations far from where the climate events occur. Here we discuss the susceptibilities of several key globalized systems to harm from such shocks.
Global Food Systems
Under normal conditions the globalization of markets, access to humanitarian relief, and public health systems all tend to reduce the susceptibility of countries and their populations to local climate risks. For example, one of the first responses of governments to expected shortfalls in domestic production is to secure food imports (Timmer, 2010). Yet these global institutions have evolved in, and in some sense been calibrated to, a climate regime that may differ in important ways from the climate of the coming decades. For example, a key feature of commodity markets is the maintenance of stocks that buffer the impact of short-term fluctuations in supply or demand. The levels of these stocks are determined by several factors, including storage costs, interest rates, and the perishability of the commodity, but a key factor is the expected volatility of supply (Wright, 2011). If climate change were to increase the chance of relatively large shortfalls in global production, stocks based on historical expectations of supply variability could be insufficient. Similarly, the capacity of countries to provide humanitarian or public health assistance is related to historical experience with the level of aid needed around the world.
Relatively little peer-reviewed literature has focused on how climate changes in the coming decades and the ability or inability of institutions to adapt to these changes will affect the likelihood of global systemic shocks to food systems, such as rapid price increases. There is some basis for expecting that indicators that aggregate over large areas, such as global food production or global incidence of humanitarian disasters, will be more affected by global climate trends than outcomes in any single location simply because the “signal” of climate change relative to natural variability tends to be clearer at larger spatial scales (Intergovernmental Panel on Climate Change, 2007). For example, Figure 1-1 in Chapter 1 shows that the total fraction of land area experiencing extremely warm temperatures (more than three standard deviations above average) has risen sharply in the past three decades, even if some individual regions have not seen dramatic warming.
A shortfall in food supply that arose from multiple bad harvests around the world and that was large by historical standards would not necessarily result in rapid price increases, given that other important factors affect price fluctuations. If a sufficient number of preceding good harvests had helped to build up stocks, if growth in biofuel demand related to energy policies slowed or reversed, or if a global recession reduced aggregate food demand, supply shortfalls could have relatively little influence on global markets. However, when bad harvests occur in an already tight market, this will generally result in large increases in food prices, as analyses of recent episodes of high prices in 2007–2008 and 2010–2011 have emphasized (Abbott et al., 2008, 2011; Wright, 2011). Policy responses to the initial shortfalls, such as export bans designed to stabilize domestic markets, then often act to further amplify price changes.
In light of recent food price increases, there has been a renewed interest in the effects of high international food prices on domestic prices and social and political stresses. One clear finding is that domestic prices in many countries change substantially less than global prices, partly because of exchange rate variability and partly because of policies aimed at stabilizing domestic prices, such as tariff adjustments, export restrictions, and the use of government storage (Dawe, 2008; Naylor and Falcon, 2010). Nonetheless, in 2008 and 2011 most countries witnessed significant increases in prices versus historical levels, with consequences for local producers and consumers.
The susceptibility of national populations to global price increases depends in large part on the countries’ net trade positions: Major importers will generally be hurt, and exporters will benefit. The MENA (Middle East and North Africa) region is the main area of the world that relies on food imports for a large (more than 30 percent) fraction of calories consumed. Wheat prices are especially important in the MENA, given that nearly half of all calories consumed in some countries are from wheat (Food and Agriculture Organization of the United Nations, 2012). A recent World Bank study (Ianchovichina et al., 2012) found that MENA countries are highly vulnerable to global food price shocks. Sub-Saharan Africa is also relatively dependent on food imports, with roughly 40 percent of rice and 70 percent of wheat consumption derived from imports (Naylor and Falcon, 2010).
Because the prices of basic commodities such as bread or flour are often subsidized, demonstrations and even riots frequently occur in response to efforts by governments to reduce subsidies, for example as part of structural adjustment policies. In general these disturbances are contained without an impact on the regime, even if there may be significant violence or property damage. The issue with regard to climate change is whether that pattern could change and that the countries most vulnerable to food price increases could become vulnerable to severe social and political unrest. Unfortu-
nately, there is very little in the peer-reviewed literature concerning the links between food price increases and political unrest. One notable exception is a recent working paper (Bellemare, 2012) that presented an econometric analysis of global data since 1990 and found that high food prices were significantly correlated with political unrest related to food prices, with the latter measured by counting the number of news stories with at least five mentions of terms related to food and riots (or their synonyms).
Interest in the topic has increased in recent years, particularly within the community concerned with food security, spurred on by the question of whether rising food prices played a role in sparking the unrest of the “Arab Spring” of 2011. It is worth noting that the rapid food price increases in the MENA during this period were not driven by local weather conditions, but by events around the world including a severe heat wave in Russia. A report by Lagi et al. (2011) notes that clusters of unrest in the MENA region in 2008 and early 2011 both began immediately after the United Nations Food and Agriculture Organization food price index passed a value of 210. Although they do not identify a causal link between high food prices and riots, the authors argue that a food price index value of 210 represents a simple potential predictor of increased unrest in food-importing countries. Breisinger et al. (2011) find that the unrest was preceded by a drop in food security across the MENA, and Ciezadlo (2011) emphasizes the role that food subsidies have played in popular attitudes toward regimes throughout the region. Johnstone and Mazo (2011) draw connections between climate events (which reduced global food production in the years preceding 2011) and the uprisings, describing climate change as a potential “threat multiplier” in the case of already unstable situations. All of these analyses are careful to note that drawing direct causal links between food prices and political instability is not possible, but they argue that food prices must be considered along with political and cultural factors in explanations of the uprisings.
Global Energy Markets
Like the food system, markets for energy commodities have become increasingly integrated globally over recent decades. In the case of petroleum, this integration is essentially complete: There is one global market that determines prices of crude petroleum, linking producers and consumers around the world (Yergin, 2006). Integration is also increasing, although lagging significantly behind petroleum, for other energy commodities such as electricity (Jamasb and Pollitt, 2005; Boëthius, 2012), natural gas (Siliverstovs et al., 2005), and coal (Wårell, 2006). Thus, possibilities for energy system shocks to have global impacts in the coming decade lie primarily in the petroleum sector.
The integration of petroleum markets was stimulated by desires to safeguard the supply of oil from manipulation by political actors in the wake of Organization of Petroleum Exporting Countries embargoes in the 1970s (Yergin, 2006). A consequence of this integration was that by the 2000s the petroleum system had become so complex and interconnected that, as one study concluded, “a disruption in one part of the infrastructure can easily cause severe discontinuities elsewhere in the system” (International Institute for Strategic Studies, 2011:21). Furthermore, the sensitivity of the system has increased because of a rapid growth in global petroleum consumption that has not been matched by a corresponding increase in production. The result has been an extremely tight market, with petroleum supplies not significantly greater than demand (Gupta, 2008). This “demand shock” (Yergin, 2006), led by the emerging economies in China and India, has left global markets volatile and very sensitive to disruptions in supply (Patrick, 2007; Gupta, 2008; International Institute for Strategic Studies, 2011).
In this tight, sensitive market, climate events that disrupt the production or distribution of oil could lead to price spikes across the global energy market. Several types of climate events could cause such disruptions. Tropical storms and the increased storm surges that result from sea level rise and, in some cases, land subsidence, can disrupt production, refining, and transport of petroleum. For example, one-third of U.S. petroleum refining and processing facilities are located in coastal areas vulnerable to storms and flooding (Schaeffer et al., 2012). Similar infrastructure vulnerabilities exist in Europe and China as well (International Institute for Strategic Studies, 2011). In addition, because offshore oil and gas platforms are generally not designed to accommodate a permanent rise in mean sea level, climate-related sea level rise would disrupt production (Burkett, 2011). The effects of Hurricanes Katrina and Rita in 2005 illustrate this potential. The storms disrupted oil and gas production from offshore rigs, refining at facilities in the coastal zone, and transportation via port facilities and pipelines, causing a spike in global prices (U.S. Department of Energy, 2005; Yergin, 2006; Schaeffer et al., 2012). The pattern repeated, although with a smaller magnitude, when Hurricanes Gustav and Ike hit the Gulf Coast region in 2008, destroying drilling rigs and disrupting refineries (Paskal, 2010).
Other climate events could also affect the global oil market. Oil refining requires large amounts of water for cooling purposes; hence, reduced water availability during a drought would reduce refining capacity. If drought is accompanied by increased temperatures, refineries will require more cooling water to operate, potentially exacerbating the situation (Schaeffer et al., 2012). Also, Arctic energy infrastructure (pipelines and drilling operations) is vulnerable to damage from subsidence caused by melting permafrost (Paskal, 2010; International Institute for Strategic Studies, 2011).
Climate change thus entails some increased likelihood of petroleum
supply disruptions, and such disruptions are likely to affect market prices, potentially causing price shocks. The likely magnitude and duration of these price shocks, however, has not been addressed in the research literature. There has been some analysis of their potential macroeconomic effects. Hamilton (2003, 2008), reviewing six decades of oil price and macroeco-nomic data, reported a very strong relationship between oil price shocks and recessions. To the extent that economic disruptions drive political instability (see, e.g., Alesina et al., 1996), it is plausible that an oil price shock could increase instability, particularly in a situation that is already politically sensitive. However, little research to date has directly addressed the political impacts of energy price shocks, whether caused by climate-related supply disruptions or other factors. These possibilities deserve more careful empirical analysis, particularly as energy markets continue to tighten with increased consumption from Asian nations and as risks increase of climate events disrupting energy supplies.
Strategic Product Supply Chains
Over the past few decades the globalization of many industries has been accompanied by a streamlining of their supply chains in order to reduce costs. However, as a 2012 World Economic Forum publication noted, “the focus on cost optimization has highlighted the tension between cost elimination and network robustness—with the removal of traditional buffers such as safety stock and excess capacity” (p. 10). Climate events can thus be a source of major disruptions in world markets for critical non-food commodities. Such events are counted as one of the major risks to be addressed in the U.S. National Strategy for Global Supply Chain Security, released in January 2012 (White House, 2012).
Although not attributed to the effects of climate change (Peterson et al., 2011), the floods in Thailand in 2010–2011 illustrate how an extreme climate event that stresses a government’s ability to respond can have global consequences. Much of Thailand, including portions of the capital Bangkok and its surrounding manufacturing districts, was flooded for extended periods between July 2011 and January 2012. The flooding resulted in more than 800 deaths, affected 13.6 million people, damaged 7,700 square miles of farmland, and caused more than $45 billion in economic losses (World Bank, 2011).
Thailand is a flood-prone country with an extensive system of dams, drainage canals, levees, and other flood-control systems, but a series of events in 2011 overwhelmed this system. The most immediate event was the abnormally high rainfall that year. In March 2011, for example, the rainfall in northern Thailand was more than three times the mean level. The abnormally low rainfall of 2010 was another contributing factor. The
initial response of dam managers to the intense rainfall of early 2011 was to store the water in the depleted reservoirs, building capacity and preventing early flooding. But when major rains unexpectedly continued, the reservoirs filled, and the dams had to release the water, resulting in flows too large for the downstream drainage canals and levees. This overflow was exacerbated by the many decades of deforestation that have taken place in northern Thailand, which allowed a greatly increased runoff from the rains and helped to overwhelm the downstream flood defenses.
The Thai government suffered significant criticism for what many saw as its mismanagement of the situation. The government was criticized for its forecasts that underestimated the scale of the flooding, for its management of the upstream dams that exacerbated downstream flooding, and for poor communications. Once the flood waters began to overwhelm the existing flood defenses, the government launched many emergency responses, including evacuations, the placing of sandbags, and the diversion of water from some channels to others. In one case the government placed hundreds of anchored boats in one river so that their propellers could help push water toward a second river. As the damages increased, many of these responses were criticized as inadequate. In addition, resistance appeared in some localities where flooding had increased due to barriers designed to protect neighboring communities. Some people ripped down the sandbags that they saw as unfairly diverting flood waters to their areas.
The floods also caused significant disruption to regional and global supply chains. Manufacturing parks located near Bangkok supply parts for the worldwide automobile and electronics industries. One-third of the world’s hard drives and high percentages of other key computer components are built there (Connor, 2012). Many of these Thai manufacturing areas were covered by up to 3 meters of water, causing parts shortages worldwide. Even the computer firms located elsewhere in Thailand that escaped the flooding found they could not get critical parts. Production is not expected to fully recover until 2013 (Mearian, 2011). In the meantime, component prices rose as suppliers attempted to stockpile what was available and manufacturers found they could not get the parts they needed. The flooding of automotive parts production facilities forced Honda and Toyota to slow production lines in many countries (Fuller, 2011).
Other Global System Effects
Climate events might also put stress on global health systems in various ways, most of them hard to predict. As discussed in the next chapter, climate change is expected to alter the ranges of disease vectors or pathogens in ways that expose large human populations to diseases to which they have not been previously exposed. This could lead to a rapidly increasing
demand for treatments and supplies that may not have been adequately stockpiled. If such health problems arise in combination with a disruption of supply chains for critical inoculations or medications, the potential for a severe health crisis could grow dramatically. Again, the effects might be felt far from the locations where the climate events occur. Climate events, especially when they occur in clusters, can also stress the capacity of international disaster response and humanitarian relief systems and thus cause harm in places that are not directly affected by the events but that need international assistance for other reasons.
Such shocks to integrated global social, economic, health, or technological systems are likely to have different effects in different places. It is reasonable to expect that they would be most disruptive in countries that are dependent on imports of the products of the global system that is shocked and in places or among populations that are particularly susceptible to harm if the availability of the outputs of those systems is restricted by price or policy.
Since the 1980s the number of recorded natural disasters related to weather and climate events has roughly doubled, while the number of those related to geophysical events, such as earthquakes, tsunamis, and volcanic eruptions, has neither increased nor decreased (Munich Re, 2012). Reported losses from global weather- and climate-related disasters also increased over the past few decades, mainly because of monetized direct damages to assets, with the amounts of losses varying greatly from year to year and region to region (Intergovernmental Panel on Climate Change, 2012). Since 1980 annual disaster losses have ranged from a few billion dollars to more than $200 billion (in 2010 U.S. dollars), with the greatest losses coming in 2005, the year of Hurricane Katrina. Loss estimates are lower bounds because many impacts, including the loss of human lives, cultural heritage, and ecosystem services, are difficult to monetize and so are poorly reflected in these estimates. Middle-income countries with rapidly expanding asset bases are particularly vulnerable to changes in the frequency, intensity, geographic range, and duration of extreme events. From 2001 to 2006 disaster losses were about 1 percent of gross domestic product (GDP) for middle-income countries, 0.3 percent of GDP for low-income countries, and less than 0.1 percent of GDP for high-income countries (Intergovernmental Panel on Climate Change, 2012). Most fatalities from extreme weather and climate events (95 percent) occur in developing countries.
The major causes of the long-term increase in economic losses from weather- and climate-related disasters have been the increasing exposure of people and the increased value of economic assets in exposed regions. Cal-
culations on the long-term trends in economic disaster losses adjusted for wealth and population increases, which have been conducted in an effort to separate the effects of change in the frequency or intensity of damaging climate events from the effects of increased exposure and vulnerability, have not attributed the increase in losses to climate change alone, but neither has a role for climate change been excluded (Neumeyer and Barthel, 2011; Intergovernmental Panel on Climate Change, 2012). These studies have not accounted well for vulnerability or for adaptation efforts, and they are limited by poor data availability. Settlement patterns, urbanization, and changes in socioeconomic conditions have influenced the observed trends in the exposure and magnitude of harm from climate events (Intergovernmental Panel on Climate Change, 2012). In particular, rapid urbanization and the growth of megacities, especially in low-income countries, have led to the emergence and growth of highly exposed and highly susceptible urban communities.
As discussed in Chapter 3, projections for the next few decades indicate that there will likely be a continuation of current trends, with greater changes in the frequency, intensity, duration, and spatial extent of some extreme events by the end of the century (Intergovernmental Panel on Climate Change, 2012). Who and what is exposed to an extreme weather or climate event depends on the event. For example, many regions are susceptible to flooding following heavy precipitation events, although the flooding and resulting damage can take a number of forms. The people and places most susceptible to harm when there are changes in the frequency, intensity, duration, and spatial extent of extreme events depend on the event and on local factors. For example, a typhoon coming ashore in the Philippines has very different consequences from one of similar strength striking Japan (United Nations International Strategy for Disaster Reduction, 2009b).
It is important to consider the possibility of compound events (see Chapter 3), such as what occurred in South Australia in January 2009 (Murray et al., 2012). An unprecedented heat wave occurred during a multi-year drought, exposing the area to some of the highest temperatures on record. In central Victoria the 12-year rainfall totals were approximately 10 to 20 percent below the 1961–1990 average (Australian Government, 2009). In Victoria, during the week of the heat wave there was a 25 percent increase in total emergency ambulance dispatches and a 46 percent increase over the three hottest days. There were 980 deaths during the four days of the heat wave, compared with an average of 606 per year over the previous five years. A few days after the heat wave, temperatures spiked again, and the forest fire danger index reached unprecedented levels. High winds caused a power line to break, sparking a wildfire that became one of the largest, deadliest, and most intense firestorms in Australia’s history; 173 people died. The bushfires also destroyed almost 1,660 square miles of
forests, crops, and pasture as well as 61 businesses. The Victorian Bushfires Royal Commission conservatively estimated the cost of the fire at AUS$4.4 billion (Parliament of Victoria, 2010).
In sum, the frequency of certain kinds of potentially damaging climate events has changed over the past half century, with additional changes in the same direction expected in the coming decade and beyond. The increasing exposure of vulnerable populations to climate and weather hazards has been the most important reason for the impacts. The exposure of people and economic assets to some climate and weather hazards (e.g., coastal storms and valley floods) is expected to continue to increase in coming decades (Intergovernmental Panel on Climate Change, 2012).
As discussed in Chapter 2, susceptibility refers to the likelihood of harm to a population as the result of either direct or indirect exposure to a climate event, such as a drought or hurricane. In the climate change literature susceptibility is sometimes used as a synonym for vulnerability (see Kasperson and Kasperson, 2001; Adger, 2006; Eakin and Luers, 2006; Gaillard, 2010). Adger, for example, defines vulnerability as “the state of susceptibility to harm from exposure to stresses associated with environmental and social change and from the absence of capacity to adapt” (Adger, 2006:268). We consider susceptibility as one component of vulnerability that becomes evident when or immediately after an exposed population experiences an event, and we distinguish it from actions taken following exposure to an event in order to reduce or alleviate harm, which we discuss in terms of coping, response, and recovery. We also include in our definition of susceptibility the likelihood of harm from exposure to the effects of climatic shocks and stresses that may occur in other regions, such as exposure to a spike in food or energy prices as the result of a climate-induced drought or energy supply disruption.
A large body of climate change and hazards literature explores the characteristics that influence the susceptibility of a population and its life-supporting systems to both direct and indirect harm from climatic risks and hazards (e.g., Liverman, 1990; Kasperson and Kasperson, 2001; Adger, 2006; Eakin and Luers, 2006; Leichenko and O’Brien, 2008; Adger et al., 2009b; Keskitalo, 2009; Gaillard, 2010). Factors that are widely agreed to influence susceptibility include economic, demographic, social, cultural, and environmental conditions; the form and quality of the infrastructure and the built environment; the presence of social capital; the effectiveness of institutions and governance; and the presence or a recent history of violent conflict (Intergovernmental Panel on Climate Change, 2007, 2012). These factors, briefly described below, are often correlated and interrelated. Many
are also directly influenced by ongoing climatic and environmental changes (Paavola, 2008) as well as by such non-climatic processes as globalization and urbanization (O’Brien and Leichenko, 2007; Leichenko and O’Brien, 2008). Bensen and Clay (2004), for example, found that extreme storm events have long-lasting negative impacts on economic growth and development and that these effects are particularly acute in poorer regions. A study by Dell et al. (2012) found that a 1°C rise in temperature in a given year increased the probability of “irregular” leadership transitions (such as coups) in poor countries but had no effect on leadership transitions in rich countries (p. 86). Keskitalo (2009) documented the influence of the globalization of renewable resource–based industrial sectors, including forestry, fishing, and reindeer herding, on local decision making in the area of climate adaptation within Arctic communities in Finland, northern Norway, and Sweden. These examples suggest that factors influencing susceptibility are often in flux and subject to both the direct and indirect effects of other stresses (O’Brien and Leichenko, 2007).
The types of economic factors associated with increased susceptibility to harm from climate events generally include low levels of per capita income, a lack of livelihood assets and opportunities, poor functioning of local markets, and a high degree of dependency on agricultural food imports to meet basic needs (O’Brien et al., 2004; Eakin and Luers, 2006; Paavola, 2008). As discussed above, dependence on food imports can make a region susceptible to harm from systemic shocks to global food markets as the result of climatic events that affect grain-producing regions.
Key social, cultural, and demographic factors associated with increased susceptibility include low levels of education and low literacy rates within the population, high degrees of gender inequality, and large shares of elderly or dependent individuals (Intergovernmental Panel on Climate Change, 2007). High rates of population growth, particularly in urban areas, that are the result of either natural increase or immigration (see the migration discussion in Chapter 5) also increase susceptibility.
The quality of the infrastructure and the pattern and form of the built environment play a role in the susceptibility of a population to certain types of climate hazards. Within arid or semi-arid agricultural regions, the presence of irrigation and the reliability of irrigation water supply influence susceptibility to harm from drought. Within urban and coastal areas, and particularly in cities located in the developing world, poor quality and maintenance of building stock; inadequacy of water, sanitation, and energy infrastructure; and the presence of extensive areas of unplanned settlement contribute to an increased susceptibility to harm from extreme storm events and flooding (Satterthwaite et al., 2007). The condition of the housing stock in informal settlements often significantly increases the susceptibility of populations to disasters. Stories of hillsides denuded by squatter develop-
ments collapsing in the event of sudden storms or slow erosion are common. In 1975, for example, a landslide destroyed one-third of the homes in the El Agustino district of Lima, Peru; in 2000 a garbage slide in an area of Manila occupied by urban squatters killed more than 300 people and destroyed 500 homes (United Nations International Strategy for Disaster Reduction, 2009b). While a few coastal cities in high-income countries, such as London and Rotterdam, have made extensive investments in protection against coastal storm surge and sea level rise, including construction of sea walls and barriers (see London Climate Change Partnership, 2006; De Graaf and Van Der Brugge, 2010), these types of investments are currently financially infeasible for many cities in low-income countries, most of which already have significant deficits in basic water infrastructure (Parry et al., 2009).
Environmental factors affecting susceptibility to various climate-related hazards include the abundance and quality of natural assets such as forests, wetlands, and freshwater and also how well ecosystems function, which affects such things as water supply and quality, flood control, soil conservation, and biodiversity. The loss or degradation of natural assets and ecosystem services, which may occur as the result of climatic events (e.g., Carter et al., 2007), increases future susceptibility to extreme climate events of all types in different regions of the globe.
Other critical facets of susceptibility are associated with governance, institutions, and social capital. The level of public spending, the quality of the public health infrastructure, access to health care, the transparency and legitimacy of governing institutions, the presence of social networks, and the level of social cohesion all influence preparedness for extreme events and the coping, response, and recovery capacities following those events (Adger, 2003, 2006; Adger et al., 2005, 2009b; Brooks et al., 2005; O’Brien et al., 2009; Termeer et al., 2012; Wamsler and Lawson, 2012).
A final factor that influences susceptibility is the presence of conflict or political or ethnic strife. Conflict can damage the infrastructure and life-supporting systems and can undermine the capacity of institutions to prepare for and respond to climatic hazard events (Barnett, 2006; Barnett and Adger, 2007; Brklacich et al., 2010). Populations living in regions where conflict is present are highly susceptible to harm from climate risks and hazards.
Many of the above factors are generic to all types of climatic shocks and stresses, while others apply to specific types of climatic hazards, such as floods or heat waves. For example, a higher level of per capita income generally means lower susceptibility to all types of climatic stresses, while the type and quality of housing stock is more relevant to the susceptibility of coastal populations to hurricanes and other storm events. Some characteristics may increase susceptibility to specific types of climate hazards while
reducing susceptibility to others. For example, within northern European and U.S. cities, brick housing stock and the lack of air conditioning have been implicated in heat wave mortality (Kovats and Hajat, 2008).
As we have already noted, many of these susceptibility factors are correlated and interrelated. A wealthier population will typically have higher levels of education, better quality building stock and infrastructure, better protection of natural assets, and more effective governing institutions, all of which reduce susceptibility. One of the main messages is the degree to which the poor suffer more and recover more slowly. As Kim (2012) recently noted:
Globally, the poor are much more exposed to [and susceptible to the effects of] natural disasters than the non-poor, regardless of measurement methods. The poor are almost two times more exposed to natural disasters than the non-poor when measured in terms of the total number of affected people per decade. When measured by the number of disasters, the poor are 20 per cent more exposed to natural disasters than the non-poor. The time trend varies across regions, with the poor in East Asia and Pacific being most exposed to natural disasters, followed by those living in South Asia and sub-Saharan Africa. The exposure of the poor in East Asia and Pacific has started to decrease in recent years, whereas it is rising in South Asia and sub-Saharan Africa. (p. 208)
A popular but superficial image of a disaster is that a community, city, region, or even an entire nation is struck by a highly damaging event more or less evenly and that the reaction to the event is also a more or less evenly paced sequence of coping, response, and recovery. The reality is much more complex (Cannon, 1993, 1994; Wisner et al., 2004). As the United States learned with Hurricane Katrina, although the storm affected a wide swath of national territory, not everyone was affected equally, nor did everyone recover to the same level or at the same rate, for a wide variety of socioeconomic, cultural, political, and geographic reasons (Adams et al., 2006; Finch et al., 2010; Gotham and Campanella, 2011). Because susceptibilities and initial coping capacities are not evenly distributed across an affected area, loss and damage patterns are highly differentiated even within a single community, with some neighborhoods or sub-areas (or even states for that matter) devastated, but others only slightly damaged, if at all (Bankoff et al., 2004; Wisner et al., 2004; Kahn, 2005). All communities have certain stresses and problems even before a disruptive event occurs, and an event’s effects will interact in various ways with those pre-existing stresses (De Sherbinin et al., 2007; Leichenko and O’Brien, 2008; Reser and Swim,
2011; Weisbecker, 2011). Box 4-1 illustrates some of these differences with the impact of cyclones in Bangladesh and Myanmar.
Highly differentiated loss and damage patterns may then be exacerbated as some parts of a community cope reasonably well with the damage and receive timely emergency assistance during response and then support for recovery, while other parts, with higher loss and damage levels and lower initial coping capacities, receive less than proportional help during
Cyclones in Bangladesh and Myanmar
In 1970, in what was then East Pakistan, between 300,000 and 500,000 people died from the multiple effects of Cyclone Bhola. The disaster contributed to pre-existing tensions between West Pakistan and East Pakistan and the eventual violent attempt at secession by East Pakistan, intervention by India, and the creation of the new nation of Bangladesh. With lessons learned and an innovative cyclone shelter program combined with improved public awareness, alert and warning systems, evacuation planning, hazard mitigation measures, and an understanding of local knowledge and norms, Bangladesh is now widely credited with having significantly reduced the potential for cyclone-related fatalities (Alam and Collins, 2010; Collins, 2011; Haque et al., 2012). Although no two storms are the same and all have different “signatures,” making it necessary to be cautious when making comparisons, the most recent major cyclone—Sidr in 2007—killed many fewer people (only about 4,000) than Bhola, which led Haque and colleagues to conclude that there had been a 100-fold reduction compared with 1970.
In 2008 Cyclone Nargis caused the worst disaster in Myanmar’s recorded history, with 130,000 dead or missing. Half of the people living in the affected areas—2.4 million out of a total of 4.7 million—were severely affected. Because the people in the region were extremely poor and had few functioning radios able to hear the only channel available, the warnings that the government issued about the impending storm were never received by most of the people there. The poor quality of the housing stock also contributed to the losses; less than 20 percent of the rural homes were able to withstand even normal monsoon rains. The storm was so strong that it overcame the local community infrastructure and social capital that had developed in large measure to compensate for the limited government presence in the region. The Myanmar government’s initial rejection of numerous offers of international aid—although it eventually did allow aid into the country— almost certainly increased the loss of life. In an interview with the New York Times several years later, Myanmar president and former general U Thein Sein called the poor government response a “mental trigger” for moving the country from decades of military rule toward democratization (Fuller, 2012).
SOURCE: Information for these examples comes from United Nations International Strategy for Disaster Reduction (2009b) except where cited otherwise.
response and recovery. As recent works on emergency management note, transitions between disaster phases are never clearly demarcated on the ground, where it counts; response capacities vary significantly (as the U.S. federal government saw in the aftermath of Hurricane Katrina, with state capabilities in Florida versus those in Louisiana); and recovery is usually very uneven (Phillips, 2009; Coppola, 2011; Phillips et al., 2011). Developing countries are particularly prone to sharply different disaster impacts and then to sharply different coping, response, and recovery patterns, which in turn can lead to new or exacerbated social and political stresses.
In a recent article that ties susceptibilities to post-impact issues, Wamsler and Lawson (2012) note that situations are especially problematic when “poor mechanisms and structures [are] in place for response and recovery by individual residents, households and communities, or the institutions serving them” (p. 31). They emphasize location-specific vulnerabilities, which are important because disasters (except in very small nations) are almost always local or, at most, affect only a part or parts of a country directly. Thus, different effects of a disaster and then different coping, response, and recovery in affected local areas may have national effects on social and political stability, either because of the importance of the local areas or because coping, response, and recovery may reflect or influence relations among groups or regions within a country.
That the poor suffer more in disasters, both absolutely and in relative terms, has long been established (e.g., Blaikie et al., 1994; Cannon, 2000; Juneja, 2008; United Nations International Strategy for Disaster Reduction, 2009b; Kim, 2012), as has been the disproportionately severe impact of disasters on women (Agarwal, 1995; Enarson, 1998, 2012; Denton, 2002; Neumayer and Plümper, 2007; Osman-Elasha, 2009; Arora-Jonsson, 2011; Kim, 2012). Indeed, after a disaster the poor, women, and other disadvan-taged groups often remain in emergency or “relief” conditions for extended periods—and sometimes permanently—while the rest of an affected community or nation moves into recovery; this confounds the simple model of disasters as following an orderly series of stages from coping to response, relief, and recovery. Thus a major lesson is that post-impact social stresses derive not only from the total of disaster losses in a community or nation, but also—and of more concern—from how those losses are differentially experienced across groups, classes, races/ethnicities, genders, and other categories. These stresses are exacerbated if disaster response and recovery efforts are seen as inadequate, inefficient, corrupt, or characterized by favoritism.
The steps of coping with and then responding to a disaster begin literally moments after the event starts, particularly in the most affected areas, and they are multilevel and multiactor. At the earliest stages very little coping involves the government or formal disaster response institutions.
Affected individuals, nuclear families, extended families, and neighbors in the affected communities react almost immediately, with their effectiveness largely determined by their levels of training, equipment, and social capital. Later, designated emergency response personnel arrive, with their effectiveness largely determined by their numbers, training, equipment, and logistical capabilities and support.
Initial coping is supplemented by the more formal response by various and complex combinations of emergency personnel (“first responders”); by local, subnational, and national agencies, organizations, and institutions, possibly including military organizations; and by community, religious, and other local nongovernmental organizations, and civil society in general. If disaster losses and disruptions are judged to exceed domestic coping capacity and response, the international community (bilateral, international, or national nongovernmental organizations) offers a potential additional response level. Surge capacity, the ability of assistance organizations to move needed supplies and people to the affected area in time to meet needs, is a critical factor in determining the effectiveness of the initial and early-stage response.
The coping and response levels or “waves” are not operationally distinct, because as the reaction to a hazard event deepens, the levels interact, with the optimal outcome being efficient synergy across levels and actors. The least optimal and most socially and politically damaging outcomes occur when the formal or official response conflicts with the initial coping efforts and appear to be characterized by misallocation, duplication, competitiveness, favoritism, and interagency conflict. In extreme cases, when initial coping with, and popular response to, a disaster is perceived by the leadership of the state as fundamentally threatening and is met with repression, the result can be escalating violence, as was captured classically by Cuny (1983) and recently updated and elaborated upon by Garrard-Burnett (2009) and Gawronski and Olson (2013). Thus in disaster situations, public authorities in many countries are faced with three different types of event-response tasks or challenges. First the government, broadly defined, is expected in most countries to respond using its own resources; it is also expected to cooperate with, if not support, initial popular coping efforts; and, finally, it is expected to coordinate the responses of other actors, including at times the international community. The extent of specific group and general public dissatisfaction with government disaster response or coordination is shaped both by expectations and by the perceived performance deficits in those task areas.
Given that affected and observing populations—and, in many countries, the mass media—are unusually attentive to the performance and probity of their leaders and institutions during the immediate aftermath of disasters and in the response phase, the political stakes are often quite high,
including the public support or tolerance of authorities, administrations, governments, and institutions in general. In extreme cases even the legitimacy of a regime or the viability of a national society may be brought into question, a situation that can be a harbinger of possible state breakdown or failure. The processes by which social and political stresses resulting from climate events may visibly manifest in political instability or state breakdown, however, are likely to take months or even several years.
The conceptual framework presented in Chapter 2 is useful for organizing available knowledge on the potential links between climate events and political and social stresses.
Conclusion 4.1: The overall risk of disruption to a society from a climate event is determined by the interplay among several factors: event severity, exposure of people or valued things, and the vulnerability of those people or things, including susceptibility to harm and the effectiveness of coping, response, and recovery. Exposure and vulnerability may pertain to the direct effects of a climate event or to effects mediated by globalized systems that support the well-being of the society.
Because risk reflects the interactions among these factors and not only the magnitude of climate events, events of a magnitude that has not been disruptive in the past can cause major social and political disruption if exposure and susceptibility are sufficiently great and response is inadequate or is widely seen as such. The other side of this coin is that unprecedentedly large climate events do not necessarily lead to security threats if actions have been taken to reduce exposure or susceptibility or increase coping capacity and if authorities are seen to be actively responding to events.
Insights About How Climate Events Can Create Stresses
Available knowledge on the factors linking climate events to social and political stresses supports several general conclusions and points to a number of needs for further research and analysis. We note first that each of the nonclimate factors linking climate events to social and political stresses is changing. Many of these conditions have been changing more rapidly than climate is changing, and this situation is likely to continue for at least the next decade or two. This suggests that the net effect of climate change in the coming decades may be determined in the near term more by social, economic, and political conditions and their interactions with climate events than by climate factors alone. In particular, combinations of
increased exposure and increased susceptibility are likely to have a multiplier effect on risk.
Several social, economic, and political factors that contribute to exposure and susceptibility to harm from climate events can be projected with some confidence for a decade or more at the country level or below. These socioeconomic and political factors include total population, population age distribution, level of social and economic development, land use patterns, compliance with building standards, governance capabilities, corruption levels, urbanization, certain changes in the physical infrastructure, and integration with global markets for key commodities.
Many other social, economic, and political factors that connect climate events to security threats cannot be projected with confidence at this time. However, the dynamics of climate-induced stress are well enough understood to establish several prudent expectations for anticipating climate-related security threats1:
- Many societies will encounter climate-induced disruptions that stress their capacity to adapt to, cope with, or respond to the disruption.
- The victim profile for climate events in the next decade will remain largely as it is: primarily the poor and the socially disadvantaged or marginalized. The absolute numbers of potential victims of climate events will increase with increased exposures and static or increasing susceptibility, particularly in low- to middle-income countries.
- Harm is likely to be greatest in low- and middle-income countries characterized by high levels of corruption and weak institutions and governance, because of high susceptibilities and ineffective response.
- Harm is likely to be greatest to populations living in countries and regions where conflict and political or ethnic strife is present or has recently been present. Such countries are more likely to have government structures that are intermediate between democratic and authoritarian and that practice inequitable allocation of public resources.
- In some instances of ineffective response and recovery, there will be impacts on the coherence of affected states.
- In a few instances, the consequences will be severe enough to compel international reaction.
- If global-scale disruption does begin to occur, it is likely to appear first at especially sensitive locations and in the initial stages is likely to be interpreted as a local or regional phenomenon.
1 In addition to the evidence discussed in this chapter, the research associated with the interactions of climate with some of these factors may be found Chapter 5.
- The consequences of climate change for human societies will interact with the process of economic globalization. The known features of both processes give strong reason to expect that the conjunction of climate change and globalization will increase the risk that climate events in one location will have adverse impacts in other parts of the world.
Conclusion 4.2: To understand how climate change may create social and political stresses with implications for U.S. national security, it is essential for the intelligence community to understand adaptation and changes in vulnerability to climate events and their consequences in places and systems of concern, including susceptibility to harm and the potential for effective coping, response, and recovery. This understanding must be integrated with an understanding of changes in the likelihoods of occurrence of climate events.
Knowledge from several scientific fields provides useful general insights about the components of vulnerability and how they shape the effects of climate events on social and political systems. Much remains to be done, however, to advance this knowledge and to make it operational for assessing the risks of climate change to social and political systems in particular places.
A Strategy for Advancing Vulnerability Research
The intelligence and national security communities are not the only parts of the U.S. government that need to improve understanding of vulnerabilities to climate change in order to achieve national goals, and the U.S. government is not the only actor that needs improved understanding. Such improved understanding is among the objectives of the many federal scientific agencies concerned with climate change and will be valuable to the various federal, state, local, private-sector, and international organizations concerned with improving adaptation to climate change, reducing potential damage from climate events, and exploiting potential opportunities related to climate change. These shared needs for knowledge suggest that knowledge development is best pursued as a cooperative activity involving many organizations.
A recent report of the Defense Science Board (Defense Science Board, 2011) emphasized the need for federal interagency cooperation in dealing with issues of adaptation to climate change. The report notes the need for sustained attention by many federal agencies to “assisting vulnerable regions in adapting to climate change” (Defense Science Board, 2011:xiv) and calls for “a structure and process for coordination to more effectively leverage the efforts to address global problems” (pp. xiv–xv). It recom-
mends that “the President’s National Security Advisor, in conjunction with the Council on Environmental Quality, should establish an interagency working group to develop…a whole of government approach on regional climate change adaptation with a focus on promoting climate change resilience and maintaining regional stability” (p. vxii). The report emphasizes the need for information systems, including the translation of information into societal benefit metrics.
The analysis in this chapter clearly indicates that effective U.S. government efforts to facilitate adaptation to climate change in important regions will require knowledge about changing regional vulnerabilities as well as about climate trends. Developing fundamental knowledge about climate vulnerabilities is a major objective of the U.S. Global Change Research Program (USGCRP), which leads federal efforts to develop scientific understanding of climate change and its implications for humanity. One of the five scientific objectives in the USGCRP’s strategic plan for 2012–2021 is to “[a]dvance understanding of the vulnerability and resilience of integrated human–natural systems and enhance the usability of scientific knowledge in supporting responses to global change” (U.S. Global Change Research Program, 2012:29). The intelligence community is an obvious potential beneficiary of this effort.
The USGCRP, however, faces significant challenges in advancing this research area, as noted in a recent review of the strategic plan by the National Research Council (National Research Council, 2012a). These include expanding it within a declining budget and dealing with the limited capacity of many USGCRP agencies to integrate the social sciences with climate science. Historically, the USGCRP has devoted the vast majority of its resources to understanding climate processes and only a very small portion to understanding the “human dimensions” of climate change, including vulnerability and response to disruptive climate events. This weakness of the program has been identified repeatedly in program reviews by the National Research Council (e.g., National Research Council, 1992, 1999, 2009), but the challenge remains. It might be addressed in part by improved collaboration between the USGCRP and agencies in the intelligence and national security communities that have not previously been engaged in its efforts in the domains of vulnerability and adaptation but that need the knowledge that such efforts could provide.
Conclusion 4.3: Many of the scientific needs of the intelligence community regarding climate change adaptation and vulnerability are congruent with those of the USGCRP and various individual federal agencies. Intelligence agencies and the USGCRP can benefit by joining forces in appropriate ways to advance needed knowledge of vulnerability and
adaptation to climate change and of the potential of climate change to create social and political stresses.
A whole-of-government approach to understanding adaptation and vulnerability to climate change can advance the objectives of multiple agencies, avoid duplication of effort, and make better use of scarce resources.
Recommendation 4.1: The intelligence community should participate in a whole-of-government effort to inform choices about adapting to and reducing vulnerability to climate change. It should, along with the USGCRP and other relevant science and mission agencies, develop priorities for research on climate vulnerability and adaptation and consider strategies for providing appropriate research support. The interagency effort on vulnerability and adaptation should include agencies responsible for community resilience and disaster preparedness and response domestically and internationally.
Engagement of the security and intelligence communities could bring considerable additional resources to this effort. Establishing such an interagency process does not imply that climate change should be defined as a security issue. Rather, it indicates that security issues are among those that should be considered in developing and executing a research agenda on climate change adaptation and vulnerability.