research paper

Community resilience and vulnerability to disasters: Qualitative models and megacities – a comparison with small towns BORIS PORFIRIEV*

Risk and Crisis Research Center at the Institute of Economics, Russian Academy of Sciences, Novocheriomushkinskaia, 42a 117418 ,

This paper contrasts the resilience to disasters of megacities with small towns, using vulnerability as a key variable. As an instrument of this comparison a formal model of the communities’ vulnerability, further deconstructed into a set of specific modules, is developed and used. It is argued that the megacities’ high resilience capacity in the main ensures only a debilitating (although undoubtedly major) effect on them by disaster agents. Meanwhile, the impact on the small towns is often disastrous and sometimes turns into a real catastrophe with some communities totally devastated. However, this observation does not preclude some notable exceptions. To corroborate and highlight the key findings above, empirical data from the Russian experience of the late twentieth to early twenty-first centuries are provided, supplemented by some international illustrations.

Keywords: natural hazards; disasters; vulnerability; resilience; megacities; small towns; models

1. Introduction and to the perception of the disaster/hazard by different communities can show important It is argued in the disaster literature that metropo- differences. litan regions with their large and rapidly growing This paper attempts to compare the resilience populations and concentration of wealth, to disasters of megacities and small towns using especially those located on the coastline or vulnerability as a key variable. To substantiate seismic zones, are most susceptible and worst the investigation and the proposition about hit by disasters (see, for instance, Sylves and differences, it starts with an interpretation of resi- Waugh, 1990; Institute of Civil Engineers, 1995; lience. This involves (a) resistance and adaptive Mitchell, 1999; Pelling, 2003). Incidents that capacity as key conceptual elements important would seem to corroborate this include the to comprehending the specificity and complexity havoc of 1903, 1948 and 1995 when major earth- of resilience as a community’s attribute, and (b) quake disasters struck San Francisco (USA), Ash- vulnerability as its integral metric. The latter pro- khabad (Turkmenistan) and (Japan), vides a conceptual and evaluation tool to under- devastation by the more recent floods in Central stand how and why a hazardous impact on a in 2002, in Indonesia and UK in 2007, community such as a big city develops generally let alone the pictures of , Madrid, only into a major emergency and rarely into a dis- Moscow and New York after terrorist attacks. aster, yet it turns into a major disaster and often However, historical and research experience into a catastrophe in a small community (i.e. shows that the most obvious and ready-at-hand hamlet, town and even small island states like ‘evidence’ often proves to be incorrect. A com- the Solomon Islands, severely hit by a tsunami parative approach to both the actual damage in April 2007).

B *E-mail: b_porfi[email protected]

ENVIRONMENTAL HAZARDS 8 (2009) 23–37 doi:10.3763/ehaz.2009.0004 # 2009 Earthscan ISSN: 1747-7891 (print), 1878-0059 (online) www.earthscanjournals.com 24 Porfiriev

This is followed by a model illustrating recipro- comprehensive, incorporating the elements of city between resilience and vulnerability as criti- both a community’s resistance and adaptation cal characteristics of a community, which capacity with an emphasis on its self- precipitate the escalation of a disaster or the pre- organizational ability to develop and manage vention of a catastrophe. The last section inte- this capacity. grates the earlier findings of the paper by Whatever its merits, resilience is still a concept providing and interpreting a qualitative model not yet operational for disaster policy and man- of vulnerability of big cities that are contrasted agement or used by grassroots communities.3 with small settlements.1 No less important, this and many other defi- nitions of resilience, however apt, are not accepted as universal by social science scholars, 2. Community resilience to disasters: who have developed and from time to time keep theoretical issues – vulnerability as an producing a gamut of their own conceptualiz- integral metric of resilience ations.4 Nevertheless, given the above recog- nition of the comprehensiveness of resilience as One academic definition implies that resilience a defining characteristic of a community’s has the capacity for collective action in the face coping capacity, one should consider its perspec- of unexpected extreme events that shatter infra- tive for crisis and disaster management policy.5 As structure and disrupt normal operating con- to its definitions, broader conceptualizations are ditions. Resilience involves the processes of worth adding which link resilience and ‘sense making’ and creative problem solving in vulnerability. socio-technical systems over time. It incorporates The point here is that whatever the formu- concepts of business continuity, effective govern- lation of vulnerability,6 its key parameters are ance and self-organization in complex, social the impact to which a system is exposed, its systems (Comfort et al., 2009). This definition exposure (sensitivity) to that impact and its resist- resonates with that which is recently and widely ance/adaptive capacity. Thus, vulnerability and used by the international disaster practitioner resilience have common elements, which community, which believes resilience is ‘the involve the hazardous impact experienced by a ability of a social system to organize itself and social or socio-technical system, the resistance increase its capacity for learning and adaptation of the system and the capacity for adaptive to withstand or change in a way to keep the accep- action. Clearly, the points of the two concept’s table level in functioning and structure’.2 convergence could hardly be overestimated Within the framework of crisis and/or disaster when considering directly contrary attributes of policy both definitions emphasize the capacity the system’s coping capacity: while resilience of a community as a socio-technical system effec- stresses a community’s ability to withstand the tively to prepare, respond to and recover from an hazardous impact and maintain development, impact that would otherwise disrupt its critical vulnerability puts to the fore the inability or functions (both social and technical infrastruc- failure to do that. ture), incur casualties and suffer significant econ- This predetermines reciprocity between a com- omic damage and/or completely devastate the munity’s vulnerability and resilience to disasters. community as a socio-technical system. The On the one hand, vulnerability is influenced by goal of enhancing resilience is to provide a the build-up or erosion of the elements of resili- system management capacity and flexibility to ence, primarily the ability to self-organize and keep following its development path (strategy) act collectively and creatively to cope with the in conditions of a changing environment. If hazardous impact (Adger et al., 2005; Adger compared to ‘resistance’ and ‘adaptation’, the 2006). On the other hand, resilience depends concept of resilience proves to be the more on a set of underlying socio-economic and

ENVIRONMENTAL HAZARDS Community resilience and vulnerability to disasters 25

environmental conditions which provide for perceived from a perspective of survival and not increasing or decreasing the community’s resilience (Heijmans, 2001). exposure and sensitivity to impact. However, Given this and the paper’s focus on the contrast this reciprocity between the two concepts is not between the resilience and vulnerability of small linear: for instance, strengthening a small com- towns and those of large metropolitan areas, our munity’s resilience does not lead automatically further analysis will involve the disaggregating to reducing its vulnerability to a disaster in the of vulnerability into its basic elements and their case of an extremely severe impact overcoming qualitative description rather than using detailed all protective barriers and turning into a cata- risk indicators. This will provide for a better strophe (for evidence of this, see below). understanding of the tendency of a hazardous Reciprocity between a community’s vulner- impact on big cities developing generally into a ability and resilience to disasters provides for major emergency and rarely into a disaster important theoretical and practical implications. whereas in the small communities it escalates Among the latter, the use of these concepts for the into a major disaster and often into a catastrophe. assessment and measurement of a social or socio- To add transparency and illuminate the dynamics technical system’s coping capacity is worth of this tendency, a generic model of escalation emphasizing. Timmerman (1981) was the first to from disaster to catastrophe, with a description link resilience and vulnerability and consider of its driving forces and indicators, is introduced the former as the measure of a system’s or subsys- below. tem’s ability to absorb and recover from a hazar- dous event impact. Building on this, other scholars have added to the development of the 3. Community vulnerability and resilience: proposition, introducing new insights. Particu- a development model of escalation from larly worth mentioning is the work by Pelling disaster to catastrophe (2003), which interpreted resilience as a determi- nant of vulnerability and stresses the role of In terms of crisis management policy, resilience planned preparation and insurance in anticipat- and vulnerability reveal a community’s prepared- ing a potential hazardous impact (Timmerman, ness, protection, exposure and non-adaptation to 1981; Blaikie et al., 1994; Pelling, 2003). In tune a hazardous impact. This precipitates the aggrava- with these conceptualizations is our earlier tion of a minor crisis into a major crisis, or an paper, which assumed vulnerability as an integral emergency turning into something much worse, metric of resilience (Porfiriev, 1999), thus invert- even into a catastrophe. Figure 1 shows two differ- ing the above interpretation of this dichotomy ent paths or models of such development. by Pelling. One of these paths involves a bifurcation These metrics include a set of both objective directly from the hazardous impact on a commu- (qualitative and statistical) and subjective (social nity, with an emergency immediately transform- construct) indicators that are actually organic ing into a catastrophe with no time and room for elements of risk (hazard) assessment and risk disaster. Otherwise, a catastrophe may arise from (hazard) perception and evaluation. The issue of a disaster, which in turn may either evolve from perception and subjective interpretation of vul- an acute crisis extending to a protracted crisis, nerability is particularly important; it is often or flare up from a minor crisis or emergency ignored despite the fact that the concept of risk, expanding and aggravating to a major which is at the core of vulnerability, has been con- fast-burning crisis (Table 1). sidered a social construct for many years. The situ- The former ‘bifurcation’ model provides for ation is further exacerbated by many lay people catastrophe factors associated either with the believing recurrent adverse occasions and events overwhelming devastating effect of a natural or a part of ‘normal life’, with rare or new disasters man-made event given that whatever resistance

ENVIRONMENTAL HAZARDS 26 Porfiriev

FIGURE 1 Crisis development: factors and path stages. Source: Porfiriev, 1998, p. 38 (edited). or resilience is present a community can in no social differentiation. These two pressures are way cope, or with the underlying set of poor socio- cumulative in disasters. economic conditions and/or major failures in a Vulnerability and resilience are associated with community’s preparedness, response to and the two sets of the community’s basic character- recovery from disaster, i.e. vulnerability. The istics: first, the development level (often associ- latter ‘evolutionary’ model implies that the disas- ated with the level of well-being and/or life ter impact on a community (not a historical quality) and type (social and geographical con- record) unequivocally deals with the commu- ditions) of an impacted community. An equally nity’s vulnerability and resilience. This echoes severe impact on relatively similar protected the ‘pressure and release’ model of hazards pro- communities could produce qualitatively differ- posed by Blaikie et al. (1994), who assumed that ent political, social and economic implications, hazards represent one pressure and characteristic which would require different disaster policies, of vulnerability and that a further pressure comes in particular in terms of planning. For instance, from the cumulative progression of vulnerability, one could easily imagine the contrast between from root causes through to local geography and the actual impact of a unique or series of

ENVIRONMENTAL HAZARDS Community resilience and vulnerability to disasters 27

TABLE 1 Development of emergency into disaster and catastrophe: interpretation of the stages

Crisis type/stage Qualitative (verbal) description

Emergency (acute crisis) Involves temporary break of normal social routines, relatively limited number of affected people (casualties) and economic damage in a given social system, which could be almost fully recovered (compensated) within a comparatively short time. Disaster (chronic and/or Involves long-term and overall break and substantial rupture of the social routines persistent crisis) (communications) and structures within a social system including human losses, health and/or environmental deterioration, considerable distress load on the affected community, huge material damage which may be restored, rehabilitated and compensated to a significant degree within a comparatively long-term perspective only. Catastrophe (devastating Involves long-term, complete break and rupture of the social communications and structures overwhelming crisis) within a given social system including numerous human losses and casualties, mass health deterioration and morbidity, huge distress load on the affected social community and stress over neighbouring and more distant communities, enormous and practically irreversible social, environmental and material damage that may be partially covered somewhere in a distant future only

Source: Porfiriev, 1998, pp. 58–59 with improvements.

devastating disasters like the Pinatubo volcanic mitigation, coping and alleviation procedures eruption in 1991, the 2004 South-East Asian and measures; and leadership, competence and tsunami or the Katrina, Rita and Wilma hurricanes accountability of management and personnel on a US and that of Singapore carrying these out as critical components of leaving alone the other much less developed business continuity and community/national small island states. In the former, the extreme situ- resilience. Failure of one or a few of these crisis ation would force the total evacuation of people to management elements may facilitate or trigger the neighbouring regions, which would mean a an escalation from disaster to catastrophe. The major national disaster. In the latter, such risk of such a failure is never zero, given the het- occasions would be considered a total catastrophe erogeneity of a community to be protected and with most of the nation disappearing. the complexity and uncertainty of disaster miti- This highlights the difference between resist- gation and alleviation to prevent escalation to ance and resilience. While one could fairly con- catastrophe. However, as social theory and sider a given US metropolitan area and policy/management practices show, the type of Singapore comparatively equally resistant to dis- political regime, social order and organization asters, the example above proves that the same pattern (at a corporate level), which creates the is not true for resilience. This illustration also framework and environment of a particular ignores the difference in public perception of crisis decision support and decision-making both the disaster risk and the salience and effec- arrangement, plays a major role in reducing tiveness of disaster policy. The contrast is no less such risk. impressive between the hazard impact on urban Worthy of special consideration is the issue of areas in developed and developing countries, openness, or in a broader sense the democratic and between that produced on a small remote nature of a given organizational or social town and on a metropolitan area (see below). system. Repeatedly, evidence reveals the applica- Second, one should mention the comprehen- bility and utility of a command-and-control siveness, consistency and timeliness of disaster approach towards crisis management by

ENVIRONMENTAL HAZARDS 28 Porfiriev

deploying and using an incident command As a result, control, manipulation and covering system. This creates the temptation to consider up information – including in the media – are it a universal crisis management tool to be employed, which distort the real picture of a dis- employed everywhere. However, its efficiency in aster and create favourable conditions for it esca- reducing the risk of a disaster’s escalation to a cat- lating into a catastrophe. One could immediately astrophe is constrained by the relatively short cite the radiation disaster in Chernobyl in 1986 ‘hot’ or ‘burning’ stages of crisis response and that badly affected unwarned communities immediate alleviation. Even at this stage, as including big cities and small hamlets in and experience of using the army as rescuers during outside the former Soviet Union; or the more the 1995 Hanshin-Awaji major earthquake disas- recent ‘Kursk’ disaster; or the toxic spill in ter in Kobe (Japan) shows, efficiency could not China in 2005 that was covered up for some always be guaranteed. Primarily this is because it weeks, thus allowing the chemicals to contami- precluded or constrained would-be improvisa- nate vast areas to the North including those tion; this is particularly important in the con- in the Russian Far East (see Porfiriev, 1998; ditions of uncertainty and complexity that are Mezhenkov, 2000; RBC, 2006). fertile grounds for the escalation from disaster In addition, it should be specially noted that to a major disaster and eventually to a information distortion and shortage, the major catastrophe. obstacle to efficient management that stems Moreover, in times of so-called normalcy (i.e. from the above-mentioned command-and- routine life), when a community or nation control model of disaster policy and institutional should prepare for crises and disasters, relying rigidity, could be precipitated by the low level of on a command-and-control model would mean economic development associated with resource excessive centralization, isolation and insulation scarcity, further aggravated by the geographical of a social and/or organizational system,7 and a remoteness of a specific area (Porfiriev and more technocratic and bureaucratic mode of Svedin, 2002). As our analysis below shows, this decision making and implementation. is particularly important for the small towns Such types of culture, worldview and oper- and hamlets in large countries like Russia, ational arrangements are dominated by the expanding over 11 time zones. Eventually, all superiority of ‘political expediency’ over human these factors contribute to the communities’ values, of means and methods over the goal, of increasing vulnerability and lessening resilience tactical aims over the strategic mission and of to disasters. symbolism over reality. Human conscience, morale, empathy, social responsibility and even , the value of human life are considered less or 4. A qualitative model of a community s least important. A set of relevant examples vulnerability to disasters: megacities vs. would bring to light the poor welding of small towns reinforcement bars and loosening of cement by adding too much sand to concrete mixtures in The conceptual model above shows resilience and the construction of residential and other vulnerability as serving a key role in understand- buildings in Armenia and Turkey that were then ing the sequence and dynamics of disaster escala- completely destroyed by major earthquakes in tion into a catastrophe. However, it does not 1988 and 1999 respectively. Or one could high- reveal the critical thresholds that a social (or light the case of the ‘Kursk’ submarine in Russia socio-technical) system should reach or cross to in 2000 when a pronounced delay in appealing pass from the state of emergency to disaster or cat- to the international sea rescue community cost astrophe. The proposition of critical thresholds the lives of 118 servicemen (Porfiriev, 1998; implies the introduction into the normative Mezhenkov, 2000; Gulkan, 2001). concept of resilience of a specific set of empirical

ENVIRONMENTAL HAZARDS Community resilience and vulnerability to disasters 29

metrics, which the community should perceive as these also could be treated as resilience prefer- ‘red lines’ or at least benchmarks. ences to maintain this level of risk). This fails to Following the earlier suggestion which consider the impact on an exposed community, assumed vulnerability as an integral of metrics being mentioned earlier as an intrinsic attribute of resilience, these criteria should involve accep- of vulnerability. table risk standards with specific indicators char- These and other complications reveal the acterizing the elements of the community’s serious challenges experienced in the search for vulnerability to disasters. However, given its com- an adequate set of resilience and vulnerability plexity, vulnerability is not easily quantifiable metrics, which so far have not been developed and could hardly be reduced to a single metric and used operationally, and ensure that scholars like the decrease of well-being of individuals keep persistently seeking them (see Adger, shown in equation (1) or its derivatives in 2006). Undertaken below is an attempt to join Table 2 proposed by Adger (2006, Appendix 1). in with this effort and to propose a model which considers vulnerability in a comparative context a Va ¼ 1=n½S(W0 Wi =W0) (1) and contrasting criteria indicators of megacities i¼1 and small towns. This model should be seen as where Va is the vulnerability indicator, Wi the generic and a preliminary (and clearly disputable) well-being of individual i; W0 the threshold version limited to communities’ vulnerability to level of well-being representing danger or vul- disasters and designed at this early stage to inte- nerability; n the total number of individuals grate in a concise way the findings in the earlier (whether households, farms, settlements or sections of this paper. whatever); q the number of individuals above The set of factors which precipitate and cata- the vulnerability threshold; a the sensitivity lyze a hazardous impact on a community, taking parameter and individuals are ordered from the magnitude and severity of disaster and then bottom to top (W1 is more vulnerable than W2 escalating into a catastrophe and eventually and so on). determining vulnerability (or conversely the resi- Following a neoclassical economic approach, lience) of the community in a formal way, could these models are based on an individual function be written as of well-being, which blurs collective preferences (assumed here are preferences to reduce vulner- 1 ability to an acceptable risk level; however, Vc ¼ f ½Si, Ec,(AC)c (2)

TABLE 2 Vulnerability measures and their interpretation

Measure Formal Verbal interpretation interpretation

Proportional V0 ¼ q/n The proportion of the relevant population (individual components of a system) that are vulnerability classed as vulnerable. This is a ‘headcount’ indicator and does not account for the degree of vulnerability of the individual.

Vulnerability gap V1 ¼ 1/n The aggregate scope of vulnerability measured in well-being terms: the summed q 1 [Si¼1(W0 2 Wi/W0) ] distance of well-being for each individual from the vulnerability threshold level of well-being. Action to reduce vulnerability could focus either on reducing the number of individuals that cross the threshold or the scale of their vulnerability. q Vulnerability V2 ¼ 1/n [Si¼1 The severity of vulnerability is measured by weighting the distribution of the vulnerability 2 severity (W0 2 Wi/W0) ] gap within the vulnerable population. The greater the vulnerability is skewed towards the most vulnerable, the greater the severity.

ENVIRONMENTAL HAZARDS 30 Porfiriev

where Vc denotes vulnerability of a community; Si economies, except in cases with special reser- severity or magnitude of hazard impact, Ec, com- vations. As elsewhere, , denotes less than, munity’s exposure, and (AC)c community’s adap- less than or equal to; . more than and more tive capacity. than or equal to. The right-hand part of equation (2) could be The data in Tables 3 and 4 show that a typical big further unpicked in the following way: city – in contrast to a small town – is more exposed to industrial accidents (the exception being small Ec ¼ f (Nc, Cc) (3) and medium towns/factories), less or equally exposed to natural hazards with usually a higher where N symbolizes the number or proportion of c probability of destruction. However, a megacity, people and C the tangible values (assets) or pro- c when it is exposed to and actually affected by disas- portion of these in an exposed community, and: ter agents, often experiences much less risk of huge 0 economic damage, which if it occurs covers a (AC)c ¼ f (Sc, Wc, E , Ac, Fc, Lc, M) (4) c minor part of the metropolitan area (see Table 4). where Sc denotes the structural adjustments (con- This stems from their relatively higher adaptive struction quality of residential buildings, indus- capacity to disasters, which involves not only trial and protection facilities, e.g. dams, etc.) in more widely used structural adjustment measures, a given community or area; Wc the availability more available and better developed warning 0 and efficiency of hazard warning; Ec the emer- systems, better logistics and management gency services’ response capabilities (SAR, support systems, much more efficient emergency medical, transportation services, evacuation response services and funds available for response planning and early evacuation); Ac the public to and recovery from disasters, but also much and infrastructure facilities’ personnel awareness higher political and media support (Tables 5–10). and capacity to cooperate; Fc the availability of At the same time, small towns lack the political funds for response and recovery; Lc the avail- influence and economic power of megacities that ability and efficiency of logistics and manage- determine resilience capacity no less than the ment support for response and recovery; and M availability of the funds. In addition, in contrast the availability of the local, regional and national to megacities, in relation to the absolute number media support (publicity and official recognition of people at risk, a far higher proportion of the of a disaster). living area and population of small towns are A closer look at the effect of some variables often exposed and affected (Cross, 2001; and see which make up components or elements of the Table 4). communities’ vulnerability to disaster agents Of course, the finding above is too generic to and eventually determine their resilience, and consider it a universal rule. One could easily cite contrasting those in the megacities to those in a few examples showing small towns to be more small towns, reveals a complicated and often con- resilient than big cities against specific hazards troversial picture, as shown in Tables 3–10 below. and in a specific social domain. For instance, a These tables assume a ‘typical’ megacity and a small or medium town in a developed nation small town in developing and transitional with the notable exception of the capital city or

TABLE 3 Comparative severity of disaster impact

Vulnerability component Criteria Indicator Megacity Small town

Si Probability of destruction Frequency/sq km More often Less often Degree of destruction Percentage of the total affected area Small/medium Large/enormous

ENVIRONMENTAL HAZARDS Community resilience and vulnerability to disasters 31

TABLE 4 Comparative exposure to disaster impact

Vulnerability Type of disaster Indicator Megacity Small component agent town

Ec Floods Probability of event (Pe) and proportion of a people (Nc) and facilities (Cc) , Earthquakes experiencing it (%) Hurricanes Industrial . accident

TABLE 5 Comparative structural adjustments to disasters and publicity of a disaster event

Vulnerability Type of disaster Indicator Megacity Small component agent town

Ec All types of Number and equipment of SAR personnel, firemen, hospitals and medical . hazards personnel (both absolute and per number of residents)

Mc Visibility and publicity of a disaster event .

TABLE 6 Comparative emergency response capacity

Vulnerability Type of disaster Indicator Megacity Small town component agent

Sc Floods The number and height of flood levees, dams, proportion of . Earthquakes reinforced and retrofitted structures and redundant safety . Hurricanes systems . Industrial . accident

TABLE 7 Comparative availability and efficiency of early warning of disasters

Vulnerability Type of disaster Indicator Megacity Small town component agent

Wc Floods Number of river gauge stations, radio and automatic warning . Earthquakes systems and their functionality . Hurricanes . Industrial . accident

at the centre of hazardous industries is less vulner- developing or transitional societies in a state of able to international terrorist attack than a big war or armed conflict, as the cases of Afghanistan, city, which seems a more attractive target to ter- Iraq, Russia (North Caucasus region) and Sudan rorists. However, the same is not true for demonstrate.

ENVIRONMENTAL HAZARDS 32 Porfiriev

TABLE 8 Public and infrastructure facilities’ personnel awareness and capacity to cooperate

Vulnerability component Type of disaster agent Indicator Megacity Small town

Ac All types of hazards Awareness . Capacity (willingness and readiness) to cooperate ,

TABLE 9 Comparative availability of funds for disaster response and recovery

Vulnerability component Type of disaster agent Indicator Megacity Small town

Fc All types of hazards Amount of response and recovery allocations: – in absolute terms (US$) . – per household affected

TABLE 10 Comparative availability of logistics and management support for response and recovery

Vulnerability component Type of disaster agent Indicator Megacity Small town

Lc All types of hazards Amount response and recovery L&M support: – in absolute terms (US$) . – per household affected

In addition, certain elements of the small already disaster practitioners should bear in towns’ adaptive capacity are more developed mind the salience of the above-mentioned con- and can cope better with disasters than those ditions as factors precipitating specificity of disas- of the big cities. Particularly worth mentioning ter planning in different types of communities. here are the informal channels of information, In general, the above tendency should be per- which make up an important part of Wc and Ac ceived as being correct in all social and geo- components of communities’ vulnerability graphic conditions, and as such one might find (and resilience) to disasters. Or the readiness it consistent with the realities of Russia. A and willingness to cooperate during response handful of Russian megacities, with slightly to and/or recovery from disasters, thus revealing more than a quarter of the total population, con- the social capital accumulated in a given com- centrate economic power in terms of capital munity, which are critical to the Ac component. assets and even more in financial resources (see The ‘superiority’ of these elements of the small Table 11). towns’ adaptive capacity is an important Leaving aside human capital, which is embo- characteristic in both developed and develop- died in modern construction and organiz- ing economies, and also in transitional ational technologies and in those who economies. develop and use them to provide more efficient These reservations provide important findings disaster risk reduction and which is also highly about the non-linearity of the tendency of the concentrated in metropolitan cities, this major cities to be more resilient to disaster implies more funding for strengthening mega- impacts than small towns. In future in-depth city communities’ resilience. In turn this research will be needed to test these matters, but implies less vulnerability and much less risk

ENVIRONMENTAL HAZARDS Community resilience and vulnerability to disasters 33

TABLE 11 Population distribution and financial flows in point a set of tables in the Appendix to this Russia paper provides comparative data that contrast % total % total capital % total the impact and implications of specific disasters population1 assets2 financial in the Moscow metropolitan area and the flows2 smaller communities in Russia. Another important piece of evidence of the Rural 27 15 5 latter being more prone to disasters deals with Urban 73 85 95 the Mc component of vulnerability, i.e. with visi- (Megacities3) (18) (48) (85) bility and publicity of a disaster event, and Total 100 100 100 especially the political and economic salience of the impacted community. It discloses the cases 1 Rounded numbers from official statistical sources. 2 of the small towns, hamlets and villages where Author’s assessment. 3 12 cities with more than 1 million people each including Moscow safety is ‘exchanged’ for – or even sacrificed (11 million) and St Petersburg (5 million). for – the sake of the big cities. This practice may be a part of, first, the national security and/or civil defense policy of locating some critical facili- of escalation from disaster to catastrophe. To ties such as nuclear power or industrial units the contrary, the small towns and hamlets, par- (including military bases) in Russia, the UK or ticularly those in the Urals and remote areas of USA outside the big cities so as to reduce the risk Siberia and the Far East, lag considerably of damage (say by vandalism) and/or to ensure behind their ‘big brothers’ in every or almost the secrecy and concealment of these facilities. every respect considered above. The notable This policy indirectly increases vulnerability of exceptions are a limited number of relatively the small towns to incidents such as terrorist more resilient small communities, mostly bombing compared to big cities. Second, it may those of research centres in close proximity stem from a specific disaster policy, in particular or, on the other hand, those distant from that of flood management during crises. From metropolitan areas, but these examples do not time to time in many parts of the world the practice change the general picture. of opening dams or locks to allow massive flood- It is sufficient to cite one example, that of the water streams to miss the major cities and flood fire services. Fire safety in metropolitan areas in smaller communities down-river has been used as Russia with a total population of some 16 a planned operation. These small towns, hamlets million people, primarily living in Moscow and and villages receive an early warning, the people St Petersburg regions and seven more big cities are evacuated before the waters swamp the streets with more than a million inhabitants, is main- and houses, and households receive compen- tained completely or almost completely by sation.9 However, sometimes in Russia this practice municipal and industrial fire brigades using has not worked satisfactorily. During the major mostly adequate equipment, although some of floods in the Krasnodar region in summer 2002 it is obsolete and not effective. However, this some settlements were washed away with over a should be contrasted to the 37 million people hundred inhabitants killed (see Appendix 1B). living in small towns and industrial hamlets, Five years later, in July 2007, opening the spillways who lack any fire service at all. Moreover, seven of the dam at the huge hydropower station on the million of these people lack any line of communi- Zeya River in the Far East caused inundation of the cation including telephones, which means that adjacent communities, which in turn caused the they have no chance of seeking help from destruction of 63 houses and the evacuation of outside their immediate community. The impli- over 400 people from the small town of Zeya and cations of these conditions could be and actually Ovsianka village, and caused economic damage of are disastrous.8 To corroborate this important almost US$ 2 million.10

ENVIRONMENTAL HAZARDS 34 Porfiriev

5. Conclusions and comparative explorations of communities’ vulnerability and resilience. In policy terms The analysis above leads to two major con- this implies more attention of the federal and clusions. One of these implies the validity of the regional authorities and media to the small resilience concept as a useful analytical and towns and villages in hazard-prone areas. This policy tool of disaster risk management, in par- in turn implies an elaboration and implemen- ticular for a comparative assessment of small tation of a framework that would marry disaster towns and metropolitan areas’ exposure to, pre- risk reduction and development programmes, paredness for, response to and recovery from dis- and enable local population participation, asters. Its salience should not be underestimated, which is critical to address the root causes of with the analytical potential of the concept vulnerability and the basics of resilience policy. further strengthened by considering vulner- More specifically, it means political, financial ability and its key components’ integral metrics and organizational assistance to build resistant of resilience. The proposed qualitative model of communities at the pre-crisis stage, efficiently vulnerability provides a set of criteria and specific rescue affected people and reduce the damage at indicators to measure particular components of trans-crisis stage, and to recover from disaster adaptive capacity as a key element of both vulner- after the crisis impact subsides. Meanwhile, the ability and resilience in a comparative perspective megacities should rely more on intercity, intra- (megacities vs. small towns). It should help to and interregional cooperation to share more enhance the merits of the concepts of resilience equally the federal resources involved in disaster and vulnerability as its metrics suggest. management, especially for mitigation. These However, the model itself needs empirical verifi- measures, if put together, will provide for a signifi- cation using statistical data from different geo- cant contribution to reduce the risk of disasters and graphical and social areas, creating a future prevent their escalation to catastrophic research agenda. proportions. Secondly, the comparative analysis of megaci- ties and small communities’ vulnerability and Appendix 1. risk of escalation from a disaster to catastrophe, although undoubtedly far from exhaustive, Appendix 1A. Hurricane in Moscow, 1998 reveals a much more complicated picture than typically portrayed. One of these complications involves a non-linear reciprocity between vulner- TABLE 1A.1 A snapshot of Moscow city ability and resilience: an increase of the former 2 does not automatically imply the same for the Area 900 km latter and vice versa. No firm evidence exists Population 8.5 million 2 that metropolitan areas are more vulnerable to Population density 9,444 people/km disaster impacts. The other complication reveals Share in GDP 20% a tendency of the megacities to be comparatively more resilient than small towns, which, however, is true only where the environmental and social conditions of both communities are equal or similar. TABLE 1A.2 Hurricane profile This provides a lesson and calls for a more Timing 22:00, 20 June 1998 balanced approach to both practical policy and Wind speed 20–30 m/sec research in the area of disaster risk management Rainfall 150 mm/day and crisis policy. From an academic perspective Hail 5–15 mm hailstones it pinpoints more focus on regional studies

ENVIRONMENTAL HAZARDS Community resilience and vulnerability to disasters 35

TABLE 1A.3 Impact and damage TABLE 1B.4 Impact and damage to lifeline utilities1

Number of people affected: 6.5 million Suspension of water supply All houses in Neftegorsk for Killed 9 24 hours Injured 173 Impact on electrical grid lines 200 km Hospitalized 122 Impact on communications 300 km Number of houses damaged (roofs and 100 (telephone) windows) Impact on oil pipelines 45 km Direct economic losses 3.5 US$ Impact on gas pipelines 1 km million Suspension of oil and gas 3 terminals Direct economic losses 100 US$ million TABLE 1A.4 Recovery data (days) Indirect economic losses 300 US$ million Rehabilitation of electricity lines and roads 0.12–0.15 1 Repair of damaged houses 3–4 Porfiriev, 1998, p. 176. Clearing away fallen trees 150–180

Appendix 1B. Earthquake disaster in Sakhalin, TABLE 1B.5 Comparative vulnerability to Kobe and 1995 Neftegorsk earthquake disasters Indicator Neftegorsk1 Kobe1 TABLE 1B.1 A snapshot of Sakhalin Island Number of people killed 2000 6500 2 Territory 76,000 km As percentage of the total 72 0.4 Population 700,000 population 2 Population density 9.2 people/km Direct economic losses 100 US$ 114,000 US$ Population of Neftegorsk 3,000 million million Recovery potential Zero Full

TABLE 1B.2 Quake profile 1 Rounded numbers. Time of occurrence 1:04 a.m. (local time) 28 May 1995 1:04 p.m. (GMT) 27 May 1995 Magnitude 7.1–7.6 (Richter scale) Appendix 1C. Catastrophic floods in Yakutia, May–June 1998

TABLE 1B.3 Impact and damage in Neftegorsk

Number of casualties Damaged facilities TABLE 1C.1 A snapshot of Yakutia Instantly killed 1,989 Buildings1 26 Missing 350 Houses totally destroyed 17 Territory 3,103,000 km2 Injured 375 Houses severely damaged 0 Population 1.07 million Houses partially destroyed 9 Population density 0.4 people/km2 Other buildings 9 Temperature: January 2288C/2508C

1 Municipal houses only. More than 1,500 private houses are not July þ28C/þ198C included.

ENVIRONMENTAL HAZARDS 36 Porfiriev

TABLE 1C.2 Impact and damage Appendix 1D. Catastrophic floods in the south of

Districts affected (total) 20 Russia, 2002 Districts affected (% of total) 58 TABLE 1D.1 Impact and damage Communities affected 171 Territory affected Nine regions – total Number of people affected: 500,000 Killed 19 population of 11 million Evacuated 51,000 people Homeless 47,773 Number of communities 303 Buildings and facilities damaged: affected Residential (houses) 15,245 Number of people affected: .500,000 Industrial units 914 Killed 114 Social infrastructure units 456 Evacuated 106,000 . Boiler-houses 130 Homeless 400,000 Bridges 312 Houses damaged 40,463 Dams 213 Houses totally destroyed 7,703 Worst affected small town (Lensk)1 Direct economic losses 470 US$ million Population 38,000 Recovery potential estimate (months) Area of town inundated 72% Rehabilitation of the main 1.0–1.5 Houses partially damaged 92% lifelines (water, electricity, gas) and roads1 2 1 In Spring 2001 catastrophic floods of the same severity completely Rehabilitation of houses 12–18 destroyed Lensk. Full rehabilitation of property 48–60 including, belongings, livestock3

Notes 1 Estimate based on official EMERCOM data on the pace of recovery works. 1. This section of the paper builds and expands upon 2 The order of the President of Russia. propositions and empirical evidence provided in 3 Estimate based on the ratio between actual compensation allowances solid contributions by Handmer (1995), Parker and value of property loss. (1995) and Cross (2001). Added to these are some recent data including that on Russia, which have rarely if ever been cited in crisis and disaster complex systems like human settlements. The key research literature. attribute organic to such complex systems is self- 2. See UN/ISDR, 2002. In this paper definitions of resi- organization, which provides for their contempla- lience other than social science definitions, for tion considering binary processes ‘resilience– instance those of ecology and environmental manageability’ and ‘resilience–development’, science, are deliberately not considered. However, missing in interpretations of robustness. one can easily locate them in a very useful overview 6. An overview of these formulations is provided in by Longstaff (2005). Adger, 2006. 3. Given this, Klein et al. (2003) cast doubt on the resi- 7. On isolation and insulation effects of disasters on lience concept’s usefulness as such. the economy, see Albala-Bertrand, 2006. 4. Some of these that tackle communities’ resilience to 8. EMERCOM Minister’s TV comment on the impli- natural hazards are overviewed by Klein et al., 2003. cations of the major fire that killed 62 people in a 5. Some scholars believe a concept of robustness has retirement house at Kamishevatskaia village by no less merit than resilience. From my perspective, the Azov Sea, Krasnodar region. robustness corresponds more with technical and/or 9. As far as we know, such a practice was employed by socio-technical systems at microeconomic level regional authorities in the UK during disastrous (organization, facility, etc.) rather than big floods in late June 2007.

ENVIRONMENTAL HAZARDS Community resilience and vulnerability to disasters 37

10. Zeyskaya GES obviniayetsia v uscherbe na 40 millionov Leskov, S., 2007. Chem vstretim stikhiyu? (What do we rublei (Zeyskaya Hydropower Station is convicted in meet a disaster with?) Izvestia, 16 August 2007. causing damage worth 40 million rubles). www.vz. Longstaff, P., 2005. Security, Resilience and Communi- ru/news/2007/7/26/96814.html; Leskov, 2007. cation in Unpredictable Environments Such as Terrorism, Natural Disasters and Complex Technology. Harvard University Program on Information Resources References Policy Research Report, November 2005. Mezhenkov, V., 2000. Atomnaia Podlodka ‘Kursk’: Adger, W. N., 2006. Vulnerability. Global Environmental Khronika Gibeli (Nuclear Sub ‘Kursk’: Chronicle of Change, 16. 268–281. the Wreckage). Pushkinskaia Square Publishers, Adger, W. N., Hughes, T. P., Folke, C., Carpenter, S. R. Moscow. and Rockstrom, J., 2005. Social-ecological resilience Mitchell, J. K., 1999. Crucibles of Hazard: Mega-cities and to coastal disasters. Science, 309. 1036–1039. Disasters in Transition. United Nations University Albala-Bertrand, J.-M., 2006. The Unlikeness of Economic Press, Tokyo. Catastrophe Localization and Globalization. Presen- Parker, B., 1995. Hazard transformation and hazard tation at the SEMA Conference on Future Challenges management issues in the London megacity. Geo- of Crisis Management in Europe, Stockholm, 3–5 Journal, 37(3). 313–328. May 2006. Pelling, M., 2003. The Vulnerability of Cities: Natural Blaikie, P., Cannon, T., Davis, I. and Wisner, B., 1994. Disaster and Social Resilience. Earthscan, London. At Risk: Natural Hazards, People’s Vulnerability and Porfiriev, B., 1998. Disaster Policy and Emergency Disasters. Routledge, London. Management in Russia. Nova Science Publishers, Bobneva, M. (ed.), 1992. The Footprint of Chernobyl: New York. Socio-psychological Processes in the Post-disaster Com- Porfiriev, B., 1999. Uyazvimost’ territorii k vozdeistviyu munity (Chernobylskii Sled: Sotsialno-psikhologichesiye opasnikh pripodnikh I techno-sotsialnikh processov protsessi v post-karastrofnom obschestve). Votum-C, i yavlenii: voz-mozhnaia metodika otsenki (An area’s Moscow. vulnerability to the impact of hazardous natural, Comfort, L., Demchak, C. and Boin, A. (eds), 2009. technological and social events and processes: an Designing Resilience for Extreme Events: Socio-technical assessment methodology). Bezoposnost pri Chrezvi- Approaches. University of Pittsburgh Press, Pittsburgh, chainikh Situatsiyakh (Safety Issues in Emergencies),1. PA. 47–49. Cross, J., 2001. Megacities and small towns: different Porfiriev, B., 2005. Managing crises in the EU: some perspectives on hazard vulnerability. Environmental reflections of a non-EU scholar. Contingencies and Hazards, 3(2). 63–80. Crisis Management, 13(4). 145–152. Gulkan, P., 2001. Rebuilding the Sea of Marmara Region: Porfiriev, B. and Svedin, L. (eds), 2002. Crisis Manage- Recent Structural Revisions in Turkey to Mitigate Disas- ment in Russia: Overcoming Institutional Rigidity and ters. Issue Paper for a Wharton–World Bank Confer- Resource Constraints, CRISMART Series Vol. 17. Elan- ence on Challenges in Managing Catastrophic Risks: ders Gotab, Stockholm. Lessons for the US and Emerging Economies. RBC (Ros Business Consulting), 2006. China Will Spend Washington, DC. US$ 3.28 Billion to Clean Up Sungari River. http://top. Handmer, J., 1995. Managing vulnerability in Sydney. rbc.ru/index.shtml?/news/policy/2006/01/08/ GeoJournal, 37(3). 355–368. 08100439_bod.shtml. Heijmans, A., 2001. ‘Vulnerability’: a matter of percep- Sylves, R. and Waugh, Jr., W., 1990. Cities and Disaster. tion. Disaster Management Working Paper No 4. Ben- Charles C. Thomas, Springfield, IL. field Greig Hazard Research Centre, London. Timmerman, P., 1981. Vulnerability, Resilience and the Institute of Civil Engineers, 1995. Megacities: Reducing Collapse of Society: A Review of Models and Possible Cli- Vulnerability to Natural Disasters. Thomas Telford, matic Applications. University of Toronto, Toronto. Trowbridge, UK. UN/ISDR (United Nations International Strategy for Klein, R., Nichols, K. and Thomalla, F., 2003. Resilience Disaster Reduction), 2002. Living with Risk: A Global to natural hazards: how useful is this concept? Review of Disaster Reduction Initiatives. UN/ISDR, Environmental Hazards, 5(1–2). 35–45. Geneva.

ENVIRONMENTAL HAZARDS