Full Terms & Conditions of access and use can be found athttps://www.tandfonline.com/action/journalInformation?journalCode=tcld20

Climate and Development

ISSN: 1756-5529 (Print) 1756-5537 (Online) Journal homepage: https://www.tandfonline.com/loi/tcld20

Making sense of resilience planning and policyin the pursuit of sustainable development anddisaster risk reduction

Hyun Kim & David W. Marcouiller

To cite this article: Hyun Kim & David W. Marcouiller (2019): Making sense of resilience planningand policy in the pursuit of sustainable development and disaster risk reduction, Climate andDevelopment, DOI: 10.1080/17565529.2019.1613215

To link to this article: https://doi.org/10.1080/17565529.2019.1613215

Published online: 17 May 2019.

Submit your article to this journal

View Crossmark data

RESEARCH ARTICLE

Making sense of resilience planning and policy in the pursuit of sustainabledevelopment and disaster risk reductionHyun Kima and David W. Marcouillerb

aChungnam National University, Daejeon, Korea; bUniversity of Wisconsin Madison, Madison, WI, USA

ABSTRACTSocio-ecological resilience and community vulnerability to climate change impacts are temporally andspatially unique concepts. Spatial models at the county-level within the US Mississippi River basin overthe past 20 years (1990–2009) were developed to explain losses associated with flooding events. Weapplied integrative indices, spatial clustering, and spatial regression models to estimate the role ofplanning and policy effort in disaster risk reduction and sustainable development. Results suggest thatengaged social capital and social justice characteristics combined with local proactive planning andpolicy in place before a disaster result in lower disaster losses and serve to enhance communityresilience. Further, while our findings are informed by a case study within the US, they are applicableto other less-developed regional and country contexts.

ARTICLE HISTORYReceived 25 March 2018Accepted 25 April 2019

KEYWORDSClimate change; communityresilience; planning andpolicy context; spatialanalysis; vulnerability

1. Introduction

Many communities are ‘becoming ever more vulnerable tonatural hazards while simultaneously becoming less disasterresilient’ (Masterson et al., 2014, p. 3). This trend suggeststhat efforts of communities are essential to reduce vulnerability,enhance response and recovery, and strengthen resilience to cli-mate risks (Frazier, Thompson, Dezzani, & Butsick, 2013; Pea-cock, Zandt, Henry, Grover, & Highfield, 2012; Schipper,Ayers, Reid, Huq, & Rahman, 2014). Natural disasters have sig-nificant impacts on socio-ecological status at diverse spatialscales and through time.

Prior research of natural disaster effects (in particular, Cut-ter et al.’s 2008 Disaster Resilience of Place model or BaselineResilience Index for Communities) attempt to capture thespatial and temporal characteristics of socio-economic situ-ations with ‘place-based differences in the level of social vulner-ability’ (Cutter, Ash, & Emrich, 2014; Cutter & Finch, 2008;Cutter & Derakhshan, 2018; Frazier et al., 2013; Oulahen,Mortsch, Tang, & Harford, 2015; Parry et al., 2018, p. 127).In the context of developing countries, analogous researchevaluates the dynamics of disaster vulnerability. For instance,Mohan and Sinha (2016) investigated spatial vulnerabilitytrends with a combined macro-and micro-level approach onthe impacts of climate change and non-climatic stressors inthe Ganges Basin of India.

As described in the works of Buggy and McNamara (2016,p. 272) and McNamara and Buggy (2017), recent research onnatural disaster reduction and climate change considered thenature of ‘community’ as geographically determined basedupon socio-political elements; important contexts of power,inequality, and marginalization. Connecting with pivotal deter-minants of vulnerability, adaptive capacity, and resilience, criti-cal elements of ‘community’ involve an ability to achieve

collective action, social networks, exclusion, social capital,and social justice (Adger, 2003; Ensor & Berger, 2009; Kim,Marcouiller, & Woosnam, 2018; Smit & Wandel, 2006). Fur-thermore, such an integrated approach can be informed bythe 2030 agenda for sustainable development adopted by theUnited Nations in 2015. This urgent action call supports globalstakeholders in improving health and education, reducinginequality and tackling climate change.

Our work reported here explores the socio-ecological com-ponents of community resilience that can contribute to redu-cing disaster vulnerability and improved disaster recoverywith an emphasis on spatial and temporal scales. Vulnerabilityand resilience concepts as central to disaster risk are uniqueconcepts (Kim & Marcouiller, 2016). They represent differingfacets of a community’s ability to prepare for, respond to,and recover from disasters and the community’s predispositionto incur losses.

Based on the application of socio-ecological resilience con-cepts within less-developed regional or developing country con-texts,1 we set out to accomplish the following research objectives:

. To describe an integrative spatial and temporal frameworkin disaster vulnerability and resilience

. To integrate various community capacities with thedynamics of short-run and long-run disaster losses to assessdeterminants in a case study region (the Mississippi Riverbasin)

This manuscript is organized into five subsequent sections.Following this introduction, we include two sections that out-line a spatio-temporal vulnerability and resilience model withan overview of the recent literature using a community capacityframework that explicitly captures planning and policy effort.

© 2019 Informa UK Limited, trading as Taylor & Francis Group

CONTACT Hyun Kim hkim9129@gmail.com Chungnam National University, Daejeon, Korea

CLIMATE AND DEVELOPMENThttps://doi.org/10.1080/17565529.2019.1613215

This involves both an acknowledgment and discussion of exist-ing knowledge that significantly contributes to the formation ofdisaster risk reduction and sustainable development in the con-text of less-developed regions and/or developing countries. Thenext two sections outline our approach and summarize ourempirical findings in investigating the role of communitycapacity to enhance natural disaster resilience using a spatialmodel, an integrative vulnerability index, and spatial clusteringtechniques. The final section provides conclusions and relevantpolicy implications.

2. Community capacity, community resilience, anddisaster risk reduction: connections andapplications for sustainable development

Current community development studies use a livelihood fra-mework to understand the quality of life, social and economicsecurity, and economic opportunities.2 This holistic frameworklooks at the strategic use of a variety of ‘community capitals,’‘community assets,’ ‘community capacities,’ and ‘communitycapability or capacity development,’ in securing sustainablelivelihoods over time (Brinkman, Seekamp, Davenport, &Brehm, 2012; Chaskin, Brown, Venkatesh, & Vidal, 2001;Donoghue & Sturtevant, 2007). Ludi et al. (2014, pp. 40–41)reviewed the local adaptation capacity framework developedby ACCRA (Africa Climate Change Resilience Alliance) (e.g.Jones, Ludi, Jeans, & Barihaihi, 2019) and community-basedadaptation to climate change (e.g. Reid & Schipper, 2014)and found that the interface between development and adap-tation ‘implies the ability [to adapt] to a constantly changingcontext, starts with assessing what people and communitiescan do for themselves to maintain or improve their well-being, and the context in which they make their living’.

Among the many terms that describe community develop-ment in relation to risk, vulnerability, and resilience, weselected ‘community capacity’ since it has been used to accountfor the collective ability of residents in disaster risk reduction(e.g. Berkes & Ross, 2013; Frazier et al., 2013; Magis, 2010;Newman, Carroll, Jakes, & Higgins, 2014; Ross, 2014; Ross &Berkes, 2014; Sherrieb, Norris, & Galea, 2010; Wilson, 2012)and community-based adaptation to climate change (Reid &Schipper, 2014). Community capacity can be appropriate inconnecting the notion of disaster resilience and capacity build-ing (Joerin, Shaw, Takeuchi, & Krishnamurthy, 2014; Weich-selgartner & Kelman, 2015) or bouncing forward (Manyena,O’Brien, O’Keefe, & Rose, 2011).

Encompassing a variety of thematic disciplines, such com-munity capacities can be regarded as central to understandingdisaster responses. The loss or devaluation of settlement, liveli-hood, assets, social security, and ‘differences in accessibility’ asspatial inequalities or unequal vulnerability at the local levelthrough natural hazards can create new vulnerabilities orrisks associated with social discrimination, social exclusion,and violence (Oulahen et al., 2015; Parry et al., 2018, p. 127).

Drawing upon the proposition that disasters are ‘sociallyconstructed’ and reflect failures of a social system to supportcommunities in adapting to environmental hazards (Birkland,2006), social frameworks in disasters can be connected to‘social contours of risk’ (Kasperson & Kasperson, 2005), ‘social

roots of risk’ (Tierney, 2014), ‘double exposure in environ-mental change and globalization’ (O’Brien & Leichenko,2000), and ‘geographical differentiation of resilience’ (Pike,Dawley, & Tomaney, 2010). Communities affected by extremeweather events have insufficient resources with which to copewith livelihood disruptions (Kim & Marcouiller, 2016, 2018).These can include a lack of preparedness and planning, pooreconomic conditions prior to the event, and characteristicsthat reflect social injustice and social isolation. For instance,Tierney (2012, p. 347) argued that the Haiti earthquake in2010 is the ‘prototypical example of how physical hazardsand the social vulnerability that is characteristic of less-devel-oped nations combine to produce a catastrophe’.

Social and economic characteristics are critical componentsof community disaster resilience. Increasing the adaptivecapacity of counties, regions, communities, and social groupscan be thought to support wider societal goals of sustainabledevelopment (Berkes & Ross, 2013). Based on identifyingdifferential social vulnerabilities, Henly-Shepard et al. (2015,p. 360) highlighted ‘ … community-based coping and adaptivecapacity building and leadership development… ’ in order toenhance social-ecological resilience and planning. By adoptingcomponents of infrastructure, technology, and finance using acity-level index-based approach, Malakar and Mishra (2017)assessed socio-economic vulnerability to climate disasters inIndian cities and claimed that those approaches can be usefulin tracking vulnerability and enhancing disaster resilience.Similarly, Gerlitz et al. (2017) measured multidimensional live-lihood vulnerability indices based on the climatic, environ-mental, and socio-economic change to reflect Hindu KushMimalayas’s rapid social, economic, demographic, and politicalchanges. In their research, adaptive capacity, sensitivity, andexposure were integrated as categories of vulnerability. Byusing participatory group discussion and household surveysin Southern Laos, Jiao and Moinuddin (2016) examined vul-nerability to climate-induced change using the three dimen-sions of adaptive capacity, sensitivity, and exposure.

Our work reported here addresses the challenges and oppor-tunities faced by a community in response to natural hazards byexamining the role of community capacity in disaster riskreduction. It is based on the principles of social and economicvulnerability and resilience. Resilience reflects the ability ofpeople and systems to recover from significant disturbancesand is dependent on community capacity, democratic character-istics, social networks, and appropriate land use planning (Kim&Marcouiller, 2016). Variables for measuring resilience are typi-cally described by the competency of ecological, social, econ-omic, institutional, infrastructure, and community systems(Cutter et al., 2008, 2014; Kapucu, Hawkins, & Rivera, 2013; Pea-cock et al., 2012; Ross, 2014). In particular, ecological and insti-tutional systems are determined by factors like floodplain area,soil permeability, wetlands acreage and loss, erosion rates, imper-vious surfaces, precipitation, biodiversity, participation in hazardmitigation programs, hazard mitigation plans, emergency ser-vices, zoning and building standards, emergency responseplans, and continuity of operations plans (Brody & Gunn,2013; Cutter et al., 2008; Kim & Marcouiller, 2018).

Applying concepts of resilience to the nexus between climatechange, disaster, and development (Bahadur, Ibrahim, &

2 H. KIM AND D. W. MARCOUILLER

Tanner, 2013), indicators need to capture demographics, socialnetworks, social capital, community value-cohesion, faith-based organizations, employment, values of property, wealthgeneration, municipal finance, effective governance, diversity,and socio-economic equity (e.g. Aldrich & Meyer, 2015; Baha-dur et al., 2013; Cutter et al., 2008, 2014; Ersing & Kost, 2012;Peacock et al., 2012; Ross, 2014). In addition, infrastructureresilience factors in the context of environmental hazardsinclude lifelines and critical infrastructure, transportation net-works, age and quality of residential housing stock, commercialbusinesses, and manufacturing establishments (Cutter et al.,2008, 2014). These indicators address physical systems andthe interdependence of infrastructure.

Resilience shifts attention from purely growth and efficiencyto needed recovery and flexibility (Walker & Salt, 2006). Asnoted in the works of Walker et al. (2002), Walker, Holling,Carpenter, and Kinzig (2004) on ecological resilience derivedfrom Holling (1973), resilience can be defined as the capacityof a system to absorb disturbances and reorganize while under-going changes to retain function, structure and feedback. Thus,in this work, we integrate social and ecological resilience withclimate change and its impacts with particular focus on therole of planning and policy effort to attain sustainable develop-ment. In the pursuit of community resilience that involveshealth, wellness, and quality of life, we explore interconnectednotions of community ‘vulnerability, resilience, and adaptivecapacity to existing and potentially unprecedented environ-mental conditions and threats is an important step’ to supportcommunity coping with environmental hazard (Newman et al.,2014, p. 21).

3. Spatio-temporal framework in vulnerability andresilience to climate risks

Reflecting on the definitions of vulnerability and resiliencenoted above and adopting the work of Kim and Marcouiller

(2015)’s integrative tourism planning perspectives on vulner-ability and resilience, we outline the spatial and temporal fra-meworks that can be applicable to the context of developedas well as developing countries. Spatial and temporal elementscan be used to represent disaster vulnerability, risk, and resili-ence as illustrated in Figure 1. Note from this Figure that ta,ta−n, and ta+m reflect temporal elements, where t denotesthe point in time when a disaster event occurs and subscripta represents a location or community affected by the event;and n and m indicate time order with ta−n reflecting a pre-dis-aster time period and ta+m reflecting post-disaster period. Inthis model, i is identified as the interval between disaster events,and is equal to n +m. The interval i can serve as a key elementinvolved in the effectiveness of disaster planning and policy, aslonger intervals can lead to complacency, while shorter inter-vals can be associated with increased urgency and haste in resi-lience planning and policy. In this spatial and temporal context,the extent of loss attributed to shock or exposure at ta dependson the intensity and duration of the disaster event. Further-more, the temporal nature of climate change can be empiricallytracked; n +m can reflect disaster intervals while loss during theevent ta can reflect outcomes related to change in the severity ofa disaster effect.

Vulnerability to disaster or climate risk refers to the likeli-hood of being exposed to disaster damage (Adger, 2006; Turneret al., 2003). Vulnerabilities (ta-1) during the first stage before adisaster period (ta−n) include numerous characteristics thatdetermine the ability of local institutions to absorb the shockof natural disasters Social and economic conditions of a com-munity might make them more likely to suffer losses relativeto adjacent and similarly affected communities. Conditionsthat lead to increased exposure can include a multitude ofitems, such as a lack of disaster preparedness planning and pol-icy, high unemployment, low-income levels, high poverty inci-dence, social isolation, weak housing infrastructure, high levelsof impervious surface, and low levels of preventive

Figure 1. Integrative spatial and temporal disaster planning and policy model.Note: arrow indicates causal relationships, dotted arrow indicates feedback loop.

CLIMATE AND DEVELOPMENT 3

infrastructure. Based on these diverse factors, we can assess thedegree of community vulnerability to disaster events.

In addition, community planning and policy prior to a dis-aster event (ta-n) serve as an important vulnerability criterion.These attributes involve the extent to which the community haslocal emergency management plans, land use planning, devel-opment regulations (e.g. zoning ordinances), building stan-dards (e.g. building codes), property acquisition (e.g. buildingrelocation), critical public facilities, and taxation information(Beatley, 2009; Berke, Lyles, & Smith, 2014; Kapucu et al.,2013). In this process, the geographical and political character-istics of each community must be considered relative to vulner-ability factors that vary in accordance with differingcommunity attributes. Hazard, vulnerability and exposure,risk can be described as a triangle that reflects ‘ … the prob-ability of a loss’ (Crichton, 1999, p. 102). Within this risk tri-angle, exposure to a disaster involves the population, localinfrastructure, and the built environment (Crichton, 1999).

Disaster losses (ta+1) in the first post-disaster stage (ta+m)result from interactions among social and economic systemsand their many subsystems (population, culture, technology,social class, economics, politics, and infrastructure). Disasterlosses change a communities’ social and economic conditionand can affect vulnerability to subsequent disaster events.Such a vulnerability framework addresses an integrated socialand economic sphere (Davidson, 2010) within community-level planning and the policy context.

The conditions in the second stage after a disaster (ta+2) havebeen addressed in the literature (e.g. Cutter et al., 2008; Hawkins,2014; Kapucu et al., 2013; Ross, 2014) as including social andeconomic systems (e.g. demographic, employment, income,social networks), physical and environmental systems (e.g. hous-ing structures, floodplain areas, wetlands, impervious surfaces),and community planning and policies (e.g. disaster plans, devel-opment regulations, building standards, property acquisition,public facilities, taxation, and formal plan reviews). Furtheralong in the post-disaster phase, elements of adaptive resiliencebecome more evident. In the third stage post-disaster (ta+3),new community conditions affected by natural disasters willshow a considerable number of disaster responses such as natu-ral resource losses, household conditions, and community plan-ning effort. In the process of disaster response (ta+m),communities with better adaptive capacity can mitigate orreduce disaster losses and shorten the period of disaster recovery.

4. Research design and methods

4.1 Analytical framework and data collection

An analytical framework was developed for a broad disaster-prone case study region – the US Mississippi River Basin ana-lysed at the county-level. This framework includes two empiri-cal phases that include (1) measuring vulnerability and riskindices and using them to identify spatial clusters and (2) devel-oping spatial regression models to evaluate the role of commu-nity capacity and effort placed in planning and policydevelopment to address disaster risk reduction.

With respect to the first phase, socio-economic vulnerabilitycharacteristics were selected that encompassed various

environmental and geographic, human and social capital, econ-omic and housing, and planning and policy effort factors. Thisphase was based on available data collected from several pub-licly accessible research sources including the US CensusBureau (USCB), US County Business Pattern (USCBP), DaveLeip’s Atlas of US Presidential Elections (DLAP), the FederalEmergency Management Agency (FEMA), the National LeveeDatabase (NLD) and the National Inventory of Dams (NID)from US Army Corps of Engineers (USACE), the PRISM Cli-mate Group (PRISM), NASA’s Moderate Resolution ImagingSpectroradiometer classification (MODIS), the EconomicResearch Service (ERS), and the Spatial Hazard Events andLosses Database for the United States (SHELDUS) at theHazard Research Lab at the University of South Carolina.Descriptive statistics for these are summarized in Table 1.

Our case study region encompassed 1,266 counties withinthe Mississippi River basin from 22 different states affectedby flooding during the last 20 years (1990–2009) as depictedin Figure 2. We calculated socio-economic vulnerability indicesusing normalization by alternative scaling methods includingmin–max transformation, Z score, and maximum value trans-formation. These indices were then spatially mapped. For therisk index measure, flood duration and flood damage severitycharacteristics (disaster exposure attributes) were added intothe estimated composite vulnerability index in accordancewith the equation, Risk = Vulnerability + Exposure. The floodduration variable indicates days affected by flooding duringthe study period and flood damage severity was categorizedinto eleven levels (from 1 to 11) by each county’s flood damagevalue divided by total flood damage. To investigate the tem-poral differentiation or transition in vulnerability and risk,the composite index for each decade (1990s and 2000s) werecompared. Based on the temporal comparison of vulnerabilityand risk, spatial clusters of vulnerability and risk wereidentified.

In the second phase, social and economic factors contribut-ing to community resilience were determined in line with com-munity capacity characteristics that included human and socialcapital, economic and housing capacity, environmental andgeographic capacity, and planning and policy effort capacity.We used a log-linear spatial lag model to estimate spatialeffects.

We divided the approach into two models (i.e. Models 1 and2) to investigate temporal differentiation or transition in com-munity capacity between the 1990s and 2000s, respectively asdescribed in Table 1. As a dependent variable, the U.S. dollarvalue of per capita disaster losses (adjusted for inflation in2010) aggregated to the county level was log-transformed tobetter approximate a normal distribution. As controlling andindependent variables, the human and social capital and econ-omic and housing characteristic variables include educationalattainment (per cent of population with a bachelor’s degree,Bachelor), race (per cent of white population, White), age(per cent of population over 65, Age), per cent of female-headedhouseholder (Female), health access (total physicians per10,000 population, Health access), housing characteristics (percent of housing units built in 1989 or earlier, Housing age;the per cent of mobile housing units, Mobile home; per centof homeownership, Homeowner; housing value, Housing

4 H. KIM AND D. W. MARCOUILLER

Table 1. Concept measurement, descriptive statistics, and temporal comparison.

Variable name Definition /measurement Data source Model 1 (n = 1266) Model 2 (n = 1266)

t-valueMean SD Mean SD

Disaster loss variablesProperty Average property damage (1,000,000US$) a SHELDUS 5.65 17.2 14.32 240.1

Per capita property damage (US$) 331.13 1012.6 505.45 11,054.8Crop Average crop damage (1,000,000US$)a 3.63 14.7 1.41 10.07

Per capita property damage (US$) 177.54 670.58 56.71 397.09Human and social capital capacity variablesCivic Civic organizations per 10,000 population County Business Pattern 14.93 5.79 16.83 30.64 −2.24*Telephone Telephone-service (%) US census bureau 91.71 5.67 97.21 1.62 −38.33**Voter Voter turnout (%) Dave Leip’s Atlas of presidential elections 42.15 6.34 44.00 6.47 −21.96**Moving in Year householder moved into unit (%) US Census Bureau 55.32 6.86 56.97 6.80 −13.66**Language Language other than English (%) 4.05 4.33 4.66 4.05 −9.50**Bachelor Bachelor degree and over (%) 17.25 6.52 14.98 6.37 32.87**White White (%) 91.58 13.68 89.32 14.60 31.28**Age 65-year old and over (%) 15.76 4.03 15.36 3.81 10.91**Female Female householder (%) 11.90 5.23 9.85 3.97 32.31**Health access Total physicians per 10,000 population 11.39 14.31 12.00 14.65 −7.48**Economic and housing capacity variablesPoverty Poverty rate (%) US Census Bureau 16.75 8.16 14.01 6.97 35.06**Homeowner Owner-occupied housing (%) 73.85 6.38 75.02 6.30 −22.87**Income Per capita income (1,000 US$) 18.29 3.32 28.39 5.18 −1.42e + 02**Housing age Year housing structure built (%) 16.77 7.46 12.24 6.24 28.72*Employ Employment rate (%) 42.50 5.72 57.94 7.61 −1.6e + 02**Housing value Median housing value (1,000 US$) 42.46 14.56 105.60 37.51 −87.01**Mobile home Mobile home (%) 12.00 6.59 11.86 8.12 2.82*Economic diversity Farming, fishing, forestry industry (%) 0.96 1.95 0.76 1.48 6.88**Resilient industry Disaster-resilient industry (%) County Business Pattern 11.98 3.91 13.74 3.64 −18.38**Business diversity Small business establishments (%) 96.07 1.59 96.06 1.59 0.80Environmental and geographical capacity variablesResidential Residential area (%) NASA’s MODIS classification 15.48 17.44 17.47 18.39 −27.67**Precipitation Number of times precipitation exceeded the 75 percentile PRISM Climate Group 7.95 3.32 8.39 3.78 −13.56**Urban influence Population covered by urban characteristics per 1,000 population Economic Research Service, USDA 0.13 0.31 0.55 0.48 −30.42**Flooding duration The length of a flood (days) SHELDUS 36.77 54.09 19.40 24.01Flooding severity Categorized severity based on flooding damage costs 1.43 1.43 1.20 1.03Planning and policy effort capacity variablesRegulation Building regulation (1 = minimalist, 2 = enabling, 3 = mandatory, 4 = energetic) May (2013)’s State regulation provisions 2.12 1.14 2.12 1.14CRS class Community Rating System Class (1-10, 11: no class) FEMA 10.54 1.24 10.54 1.24Mitigation plan Population covered by multi-hazard approved mitigation plan (1,000 person) 40.25 109.95 43.24 114.53 −9.08**Storm-ready Population in Storm-ready counties (1,000 person) 25.34 99.43 27.72 104.19 −6.98**Levee Levee length (mile) National Levee Database, USACE 14.68 425.60 14.68 425.60Dam Storage of Dams (1,000Acre-feet) National Inventory of Dams, USACE 8.13 58.58 8.13 58.58

Note: Model 1 is for 1990s, Model 2 is for 2000s, a: adjusted in 2010, *: statistical significance at 5%, **: statistical significance at 1%, bold characters indicate variables used in measuring vulnerability index.

CLIM

ATE

ANDDEV

ELOPM

ENT

5

value), and language (the per cent of population using alanguage other than English, Language).

Characteristics reflective of economic capacity weremeasured using five variables that included poverty (per centof poverty), per capita income, employment rate, economicdiversity, and resilience industry. Following Ganapati, Cheng,and Ganapati (2013), we measured economic diversity usinga percentage of primary export-base sectors (agriculture, for-estry, fishing, and hunting industry as identified in NAICS).The resilience industry variable attempts to incorporate thosesectors expected to be more disaster-resilient (faster recoveryrate) and was proxied using per cent of public administration,education, and health care services (56, 61, and 62 in NAICS).Also, responding to the suggestion by Edwards (2013, p. 37)that ‘ … small local businesses are a key to economic recoveryin a community hit by disaster… ’, we selected the percentageof small establishments (firms employing less than 500 employ-ees) as a proxy for business diversity characteristics thought tobe crucial drivers in reducing disaster losses.

Responding to the positive role of social capital in enhancingcommunity resilience in the face of climate risks, we selectedproxies for social capital characteristic variables (as both con-trols and explanatory variables) that included voter partici-pation (per cent of voter turnout in the presidential generalelection in 1992, 1996, 2000, 2004, and 2008) and civic organ-ization (number of civic organizations consisting of religious,grantmaking, civic professional, and similar organizations,813 coded industries classified by the NAICS) collected fromUSPE and USCBP.

Environmental and geographically influenced character-istics involved the extent of urban characteristics which werecollected from the ERS. Geographically specific urban influenceattributes are also associated with economic characteristics asnoted before. The planning and policy effort characteristic vari-ables encompass nonstructural mitigation and structural miti-gation measures to reduce natural hazard losses. Asnonstructural mitigation measures, building regulation status(categorized into minimalist, enabling, mandatory, and ener-getic, as suggested by May (2013)’s state regulation provisions),community rating system class (classified from 1 to 11 byFEMA), population covered by multi-hazard approved mitiga-tion plan, and population covered by storm-ready countiesdesignated by FEMA were selected, In addition, levee lengthand storage dams were selected as variables to represent struc-tural mitigation measures. Data on damage or losses after aflood occurred, such as property damage, and crop damage atthe county level, were obtained from the SHELDUS.

4.2. Study area and potential application to the contextof developing countries

Increasing frequency and severity of flooding within the Missis-sippi River basin continue to put a large number of people andresources at risk and thus serves as an ideal case study regionfor examining the role of community capacity in disaster riskreduction over the last 20 years. This flood-prone basinincludes about 1,600 counties in 24 U.S. states. Among thecounties, as depicted in Figure 3, we focused on 1,266 counties(22 states) designated in FEMA presidential disaster declara-tions based on flood losses from 1990 to 2009. Within thisstudy areas, physical damage costs including property andcrop losses resulting from floods totalled roughly US$43 billionduring the study period, with the estimation that the damagecosts in the 2000s would increase compared to the 1990s(SHELDUS). In this study, we focused on the MississippiRiver basin region; much of which is rural and less-developed.

4.3. Methods

We measured the vulnerability index using equal weighted andnormalized scaling methods that include min–max transform-ation, Z score, and maximum value transformation, based onthe various factors such as environmental and geographical,human and social capital, economic and housing, and planningand policy effort. As suggested in the work of Cutter et al.(2008) and Kim and Marcouiller (2018), we measured the vul-nerability index using equally weighted and normalized scalingmethods with the selected 21 variables as described in Table 1.To represent capacity characteristics over time, we selected 21variables among a total of 31 variables to measure vulnerabilityindex (described in bold characters in Table 1). For human andsocial capital characteristics, we selected 9 variables thatincluded Civic, Voter, Moving in, Language, Bachelor, Whi,Age, Female, and Health access. For economic and housingand environmental and geographical characteristic variables,we selected Poverty, Homeowner, Income, Housing age, Employ,Mobile home, Economic diversity, Resilient industry, Businessdiversity, and Residential. For planning and policy effort

Figure 2. Analytical framework.

6 H. KIM AND D. W. MARCOUILLER

characteristics, we selected Mitigation plan and Storm-readyvariables.

Alternative normalization methods included min–maxtransformation (Xi − Xmin/Xmax − Xmin), Z score (Xi− �X/s),and maximum value transformation (Xi/Xmax), where Xi indi-cates each value of the variable in county i and X is the totalvalue of each variables among counties. s is a standard deviationand X max and Xmin are maximum value and minimum value ofX, respectively. Even though the values of each method weredifferent, the county ranking of vulnerability and risk indicesmeasured by each method were generally analogous.

In a similar approach, to measure the risk index we assumedthat risk was equal to vulnerability plus exposure. For exposure,we reclassified severity by the level of flood damage and calcu-lated flood duration. Based on these indices, we mapped thespatial distribution of vulnerability and risk index. The specificvulnerability index shown was based on a min–max transform-ation normalized method (values range from 0 to 1). Based onthe estimated flood vulnerability index discussed earlier, weexamined whether there were spatial associations for each vul-nerability index and risk index by using Local Moran’s I shownin Figure 4. In line with the result of the spatial distribution ofvulnerability and risk index and spatial clustering of the indexover time, we used a spatial analysis model to investigate thespatial relationships between community capacity character-istics and disaster losses in the Mississippi River basin areasduring the last 20 years. Prior to estimating the spatial associ-ations among the characteristics, a log-linear regressionmodel was conducted with an emphasis on the existence andmagnitude of potential multicollinearity among independentvariables. Within analogous social science empirical research,a VIF greater than 5 reflects serious multicollinearity results.The VIF (of 4.56) was below this threshold of 5.

As discussed earlier, in an effort to examine the determi-nants of community resilience in the study areas, a log-linearregression model was employed with the dependent variable(the per capita dollar value of flood losses), FL. Along with var-ious human and social capital capacity (HSC), economic andhousing capacity (EHC), environmental and geographicalcapacity (EGC), and planning and policy effort capacity(PEC), and each flood damaged county i, the empirical modelcan be addressed by the following equation:

ln (FLi = f (HSCi, EHCi, EGCi, PECi) (1)

Drawing upon the results of the log-linear regressionmodel, we tested spatial dependence to check model validityin accordance with the work of Anselin (1988). The diagnos-tic results suggested that there was statistical significance inboth the LM-lag (for example, in Model 1, LM-lag =35,112, p < .001) and LM-error (in Model 1, LM-error =18,935, p < .001; In Model 2, LM-lag = 36,337 p < .001; andLM-error = 19,119 p < 0.001) for spatial lag dependenceand spatial error dependence. In this case, since a robustLM-lag (in Model 1, robust LM-lag = 89,229, p < .001) wasstatistically significant compared to the LM-error (inModel 1, robust LM-error = 88,489, not significant; InModel 2, robust LM-lag = 91,526 (p < 0.001) and robustLM-error = 90,234, not significant), we selected the spatiallag model. Results of this model suggested that the depen-dent variable relied on the dependent variable observed inneighbouring units and on a set of observed local character-istics (LeSage & Pace, 2010). Based on Equation (1), thespatial lag model on the effect of community capacity charac-teristics (HSC, EHC, EGC, and PEC) on flood losses (FL) andeach flood damaged county i can be addressed by the

Figure 3. Study area.Note: PDD indicates Presidential Disaster Declaration.

CLIMATE AND DEVELOPMENT 7

following equation (2):

ln (FLi) = dW ln (FLi)+ a0(HSCi)+ a1(EHCi)

+ a2(EGCi)+ a3(PECi)+ 1i (2)

where α indicates parameters and ɛi is the error term. δ denotesthe spatial autoregressive coefficient and W indicates a spatialweights matrix with elements (i.e.W = 1 indicating two differentspatial units are neighbours and W = 0 otherwise).

5. Findings

On average across the study area, most of the communitycapacity variables in the 2000s showed increases compared to

the 1990s (except for educational attainment, percentage ofbachelor degrees). This was also true for human and socialcapital capacity characteristic variables which serve as proxiesfor social capital. The social justice characteristic of healthaccess (indicating total physicians per 10,000 population), onaverage within the study area, also increased by 1.3%. Overall,most of the variables associated with economic and housingcapacity suggested economic growth and physical improve-ment in housing that can be helpful in dealing with unexpectedevents and the development of resilient communities. Further-more, with regard to local economic structure, three proxy vari-ables in the 2000s used for economic diversity, resilientindustry, and business diversity characteristics suggestedsmall increases compared to the 1990s. Responding to the

Figure 4. Spatial distribution of vulnerability index (upper side) and risk index (lower side) between 2000 and 2009 (colour online).

8 H. KIM AND D. W. MARCOUILLER

temporal effects on community capacity characteristics with acomparison between the 1990s and the 2000s using a t-test,most variables represented a statistically significant temporaldifference, except for the business diversity variable.

The spatial distribution of vulnerability and the risk indexbetween 2000 and 2009 is mapped in Figure 4 (darker coloursreflect higher indexed levels of vulnerability). Within the studyarea, many counties adjacent to the Mississippi River reflect asomewhat higher vulnerability index than other counties.Using our indices, the top five disaster vulnerable countiesincluded Franklin County, Missouri (calculated vulnerabilityindex is 1.000), Greene County, Missouri (0.936), VernonCounty, Missouri (0.8998), Lawrence County, Kentucky(0.7918), and Renville County, Nebraska (0.7720). By contrast,the five least vulnerable counties to flood events during the dec-ade of the 2000s included McLean County, Kentucky (0.000),

Adair County, Iowa (0.0027), Knox County, Ohio (0.0080),Scott County, Kentucky (0.0086), and Rock Island County, Illi-nois (0.0095).

The spatial distribution of the risk index between 2000 and2009 is shown in Figure 4; again darker shading indicatinghigher risk levels. Using our risk index, the top five countiesfor flood risk included Hancock County, Illinois (663.04), LaCrosse County, Wisconsin (616.03), Randolph County, Illinois(594.06), Brown County, Illinois (583.07), and Jersey County,Illinois (570.09). Those counties shaded in red indicate positiveautocorrelations with clustering of high values in the vulner-ability index while that shaded blue shows positive autocorrela-tion with clustering of low values in the index.

In terms of spatial econometric modelling and a Moran scat-ter plot, the former represents High-High (within the upper-right quadrant) and the latter are termed Low-Low (within the

Figure 5. Spatial clustering of vulnerability index (upper side) and risk index (lower side) between 2000 and 2009 (colour online).

CLIMATE AND DEVELOPMENT 9

lower-left quadrant) among the four different types of spatialassociations. As hotspots with flood vulnerability, several coun-ties that included Dubuque and Jackson County in Iowa, Jersey,St. Clair, and Monroe County in Illinois, Henry, Camden, Hick-ory, Taney, Greene, and Polk County in Missouri were associ-ated with spatial clustering of vulnerability. Similarly, based onthe calculated flood risk index, we identified whether therewere spatial associations for the risk index by using Local Mor-an’s I. As illustrated in the lower side of Figure 5, if we focus onthe red colour and the blue colour (High-High and Low-Low),there were several counties representing a spatial clustering ofthe flood risk index along the Mississippi River.

In view of the correlation between the vulnerability index andrisk index, as expected positive correlation (r = 0.6849, p < .001)were obtained. As illustrated in Table 2, the result of the spatiallag model was similar to that of the log-linear regression model.Several factors associated with community resilience appear toplay a pivotal role in modifying the disaster losses caused byflood events at the county level for temporal variations. As pre-dicted, an increase in the social service asset characteristic vari-able significantly decreases property and crop damages byfloods regardless of temporal change. Responding to the tem-poral comparison or effects on community capacity in floodrisk reduction, by and large, there are little differences in therole of various community capacity characteristics between the1990s and 2000s in Models 1 and 2 shown in Table 2.

With respect to economic and housing capacity character-istics, counties with a lower poverty rate and higher housingage had no discernible influence on reducing the flooddamages. On the contrary, counties with higher economicand housing capacities such as homeownership, income level,and business diversity (number of small businesses) were nega-tively associated with flood losses regardless of the decade con-sidered. Among various environmental and geographiccapacity characteristics, urban physical characteristics such asimpervious surfaces and level of urbanization were positivelyassociated with disaster losses. Consistent with the results ofBrody, Blessing, Sebastian, and Bedient (2014), both nonstruc-tural hazard mitigation strategies such as a community ratingsystem and building regulations and structural hazard mitiga-tion measures such as levee length were inversely correlatedwith disaster losses.

6. Conclusions and discussion

Natural disasters exacerbated by climate change can potentiallylead to an incremental rise in the vulnerability of economic,social, and environmental systems that affect such humanneeds as food or water availability, shelter, transportation,health, and ecosystem function. In this study, we examinedthe association between community capacity and resiliencewith specific reference to social, economic, environmental attri-butes and the effort placed on planning and policy relatedcharacteristics at the county-level within the U.S. MississippiRiver basin from 1990 to 2009.

Using the basic premise that natural disaster effects aresocial frameworks that require pro-active planning, we devel-oped a spatio-temporal conceptual model for disaster risk

Table 2. Spatial lag model and temporal effect results on the effect of communitycapacity on disaster losses.

Variable name Model 1 Model 2

Human and social capital capacity characteristics (HSCi)Civic −0.025***

[0.005]−0.003*[0.001]

Voter −0.008[0.006]

−0.008*[0.007]

Moving in −0.039***[0.008]

−0.021**[0.009]

Language 0.0007[0.005]

0.020**[0.008]

Bachelor −0.027***[0.007]

−0.013*[0.009]

White −0.0004[0.003]

−0.007*[0.005]

Age 0.035***[0.010]

0.003*[0.014]

Female 0.002[0.012]

0.019[0.023]

Health access −0.015**[0.005]

−0.006**[0.002]

Economic and housing capacity characteristics (EHCi)Poverty −0.017**

[0.0001]−0.005*[0.010]

Homeowner −0.038**[0.008]

−0.003*[0.009]

Income −0.0004**[0.0001]

−2.85E-06*[0.0001]

Housing age 0.053***[0.007]

0.006*[0.008]

Employ 0.003[0.008]

0.008[0.008]

Mobile home 0.007[0.005]

0.005[0.006]

Resilient industry −0.007[0.006]

−0.007[0.010]

Business diversity −0.018*[0.018]

−0.062**[0.028]

Environmental and geographical capacity characteristics (EGCi)Residential 0.007**

[0.002]0.003*[0.002]

Precipitation 0.109***[0.010]

0.025**[0.012]

Urban influence 0.0002*[0.0001]

0.0002***[0.00006]

Planning and policy effort capacity characteristics (PECi)Regulationa −0.086**

[0.026]−0.045*[0.034]

CRS classa 0.011**[0.025]

0.058**[0.027]

Mitigation plana 1.08E-06[4.62E-07]

2.78E-07[5.22E-07]

Storm-readya −8.91E-07**[4.00E-07]

−7.09E-07*[5.32E-07]

Leveeb −0.007**[0.002]

−0.011**[0.003]

Damb −3.01E-07[3.33E-07]

−2.96E-07[4.15E-07]

Constant 15.05***[2.09]

11.61***[3.51]

Number of observation 1,266 1,266Measures of fitLog likelihood −204,824 −204,525AIC 385,234 399,117SC 355,112 407,336Likelihood ratio 9899** 10,005**Spatial dependence testMoran’s I (error) 0.438*** 0.496***LM (lag) 35,765*** 37,110***LM (error) 85,626 87,117

Note: Dependent variable: log per capita flood losses, Model 1: 1990–1999, Model2: 2000–2009, *: statistical significance at 10%, **: statistical significance at 5%,***: statistical significance at 1%, standard errors in Brackets. a: nonstructuralmitigation measures, b: structural mitigation measures, AIC is Akaike Info Cri-terion, SC is Schwarz Criterion.

10 H. KIM AND D. W. MARCOUILLER

reduction and sustainable development that involved local vul-nerability, risk, and resilience processes. Prior research on natu-ral disaster effects attempted to reflect the spatial and temporalcharacteristics of socio-ecological situations or social systems.However, the results of previous studies generally lacked spa-tio-temporal changes and spatial dependencies with otherplaces even though the impacts of non-routine events havechanged over time at the local level. In this research, we filledthis void by exploring the socio-ecological components of com-munity resilience that can contribute to lower vulnerability andlead to enhanced resilience to climate impacts. This approachcan be promoted as a conceptual framework to achieve sustain-able development.

Drawing upon socio-ecological resilience and communitycapacity characteristics, we temporally and spatially analysedcommunity vulnerability using an integrative index and spatialclustering. In addition, we estimated the role of communitycapacity and planning and policy effort in disaster riskreduction using spatial lag models. Further, such an integratedand holistic analytical framework can be valuable in designingpolicies to achieve more robust sustainable development andcontribute to the accomplishment of 2030 agenda of the UnitedNations for sustainable development and the improvement ofhealth and education, reduction of inequality, and serious com-mitments to ameliorating climate change. Results suggest thatdisaster losses had inverse associations with community socialand economic structures, and that engaged social capital andsocial justice characteristics combined with local proactiveplanning and policy in place before a disaster resulted inlower disaster losses and enhances to community resilienceand environmental sustainability.

Our results suggest that planning and policy effort prior tothe occurrence of natural disasters lead to reduced disasterlosses; as did improve community capacity. To mitigate the vul-nerability, communities should set aside resources to make resi-dents safer. For this, policymakers or planners at the local levelneed to work with residents planning for disaster mitigation orresilience activities. Moreover, effective and proactive local pol-icy-making and planning can help minimize disaster loss whilehelping to make disaster-prone communities more resilient tofuture events.

Lessons learned relate to crucial disaster resilience planningpractice. Collaborative and comprehensive natural hazard miti-gation planning that addresses social and ecological com-ponents can improve local responses to natural disasters inthe future and serve to promote community resilience. Com-munities that develop and implement integrated resilience sol-utions that combine structural and nonstructural measures andincorporate specific tools and techniques can enhance andstrengthen responses to natural disasters and move commu-nities forward to sustainable development.

In this study, we stressed the application of this approach tothe context of less-developed regions and developing countriesby adopting a conceptual model for disaster vulnerability andresilience and an empirical framework to address spatial andtemporal aspects. However, our empirical results are geo-graphically limited to a mostly rural region encompassing themiddle region of the US. Secondary data availability will be asignificant challenge for future research that addresses cross-

national comparisons with mixed methods. Integration of thebest scientific evidence-based policies to address social dimen-sions of climate change, climate justice, the water-food-energynexus as a form of environmental security, and multiscale pro-poor climate policy is a necessary next step (Kim et al., 2018;Mearns & Norton, 2010; Waters & Adger, 2017).

Notes

1. Less-developed regional contexts can include much of the develop-ing world and rural or inner-city regions of the developed worldreflective of a general lack of agglomeration economies, amenityassets, declining social programs, and increased rates of povertyand income inequality.

2. Many of these studies are generally in line with the 2015 agendathat calls for ‘concerted efforts towards building an inclusive, sus-tainable and resilient future for people and planet’ (United Nations,2015).

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Hyun Kim is an assistant professor in the College of Social Sciences atChungnam National University, South Korea. His research driven byspatial econometrics, longitudinal data analysis, and content analysisencompasses environmental policy and planning, multilevel climate gov-ernance and climate adaptation policy, climate justice and vulnerability-readiness nexus, food-energy-water security nexus, planning and policyfor coastal community resilience, the intersections of urban (re) develop-ment, social capital, and social justice, urban gentrification and urban pol-icy, and science-policy integrated community participatory research.

David W. Marcouiller is a professor of regional development economics atthe University of Wisconsin–Madison. A resource economist by training,his work focuses on the linkages between natural resources and commu-nity economic development with a particular interest in land use compat-ibility, multifunctional rural landscapes, amenity-driven migration, andcommunity disaster resiliency. He has published more than 200 manu-scripts in a variety of outlets that span regional science, planning, resourceeconomics, and rural development.

References

Adger, W. N. (2003). Social capital, collective action and adaptation to cli-mate change. Economic Geography, 79(4), 387–404.

Adger, W. N. (2006). Vulnerability. Global Environmental Change, 16,268–281.

Aldrich, D. P., & Meyer, M. A. (2015). Social capital and community resi-lience. American Behavioral Scientist, 59(2), 254–269.

Anselin, L. (1988). Lagrange multiplier test diagnostics for spatial depen-dence and spatial heterogeneity. Geographical Analysis, 20, 1–17.

Bahadur, A. V., Ibrahim, M., & Tanner, T. (2013). Characterising resili-ence: Unpacking the concept for tackling climate change and develop-ment. Climate and Development, 5(1), 55–65.

Beatley, T. (2009). Planning for coastal resilience: Best practices for calami-tous times. Washington, DC: Island Press.

Berke, P. R., Lyles, W., & Smith, G. (2014). Impacts of federal and statehazard mitigation policies on local land use policy. Journal ofPlanning Education and Research, 34, 60–76.

Berkes, F., & Ross, H. (2013). Community resilience: Toward an integratedapproach. Society and Natural Resources, 26, 5–20.

Birkland, T. A. (2006). Lessons of disaster: Policy change after catastrophicevents. Washington, DC: Georgetown University Press.

Brinkman, E., Seekamp, E., Davenport, M. A., & Brehm, J. M. (2012).Community capacity for watershed conservation: A quantitative

CLIMATE AND DEVELOPMENT 11

assessment of indicators and core dimensions. EnvironmentalManagement, 50, 736–749.

Brody, S. D., Blessing, R., Sebastian, A., & Bedient, P. (2014). Examiningthe impact of land use/land cover characteristics on flood losses.Journal of Environmental Planning and Management, 57(8), 1252–1265.

Brody, S. D., & Gunn, J. R. (2013). Examining environmental factors con-tributing to community resilience along the Gulf of Mexico Coast. In N.Kapucu, C. V. Hawkins, & F. I. Rivera (Eds.), Disaster resiliency:Interdisciplinary perspectives (pp. 160–177). New York, NY: Routledge.

Buggy, L., & McNamara, K. E. (2016). The need to reinterpret “commu-nity” for climate change adaptation: A case study of Pele Island.Vanuatu. Climate and Development, 8(3), 270–280.

Chaskin, R. J., Brown, P., Venkatesh, S., & Vidal, A. (2001). Building com-munity capacity. New York, NY: Walter de Gruyter.

Crichton, D. (1999). The risk triangle. In J. Ingleton (Ed.), Natural disastermanagement (pp. 102–103). London: Tudor Rose.

Cutter, S. L., Ash, K. D., & Emrich, C. T. (2014). The geographies of com-munity disaster resilience. Global Environmental Change, 29, 65–77.

Cutter, S. L., Barnes, L., Berry, M., Burton, C., Evans, E., Tate, E., & Webb,J. (2008). A place-based model for understanding community resilienceto natural disasters. Global Environmental Change, 18(4), 598–606.

Cutter, S. L., & Derakhshan, S. (2018). Temporal and spatial change in dis-aster resilience in US counties, 2010-2015. Environmental Hazards,doi:10.1080/17477891.2018.1511405

Cutter, S. L., & Finch, C. (2008). Temporal and spatial changes in socialvulnerability to natural hazards. Proceedings of the National Academyof Sciences of the United States of America, 105(7), 2301–2306.

Davidson, D. J. (2010). The applicability of the concept of resilience tosocial systems: Some sources of optimism and nagging doubts. Societyand Natural Resources, 23, 1135–1149.

Donoghue, E. M., & Sturtevant, V. E. (2007). Social science constructs inecosystem assessments: Revisiting community capacity and communityresiliency. Society and Natural Resources, 20, 899–912.

Edwards, F. L. (2013). All hazards, whole community: Creating resiliency.In N. Kapucu, C. V. Hawkins, & F. I. Rivera (Eds.), Disaster resiliency:Interdisciplinary perspectives (pp. 21–47). New York, NY: Routledge.

Ensor, J., & Berger, R. (2009). Understanding climate change adaptation.Rugby: Practical Action.

Ersing, R. L., & Kost, K. A. (2012). Surviving disaster: The role of social net-works. Chicago, IL: Lyceum Books, Inc.

Frazier, T. G., Thompson, C. M., Dezzani, R. J., & Butsick, D. (2013).Spatial and temporal quantification of resilience at the communityscale. Applied Geography, 42, 95–107.

Ganapati, N. E., Cheng, S., & Ganapati, S. (2013). Resilient rural commu-nities: Housing recovery patterns following Hurricane Katrina. In N.Kapucu, C. V. Hawkins, & F. I. Rivera (Eds.), Disaster resiliency:Interdisciplinary perspectives (pp. 99–119). New York, NY: Routledge.

Gerlitz, J.-Y., Macchi, M., Brooks, N., Pandey, R., Banerjee, S., & Jha, S. K.(2017). The multidimensional livelihood vulnerability index: An instru-ment to measure livelihood vulnerability to change in The Hindu KushHimalayas. Climate and Development, 9(2), 124–140.

Hawkins, C. V. (2014). Planning and competing interests: Testing the mediat-ing influence of planning capacity on smart growth policy adoption.Journal of Environmental Planning and Management, 57(11), 1683–1703.

Henly-Shepard, S., Anderson, C., Burnett, K., Cox, L. J., Kittinger, J. N., &Ka’aumoana, M. (2015). Quantifying household social resilience: Aplace-based approach in a rapidly transforming community. NaturalHazards, 75, 343–363.

Holling, C. S. (1973). Resilience and stability of ecological systems. AnnualReview of Ecology, Evolution, and Systematics, 4, 1–23.

Jiao, X., & Moinuddin, H. (2016). Operationalizing analysis of micro-levelclimate change vulnerability and adaptive capacity. Climate andDevelopment, 8(1), 45–57.

Joerin, J., Shaw, R., Takeuchi, Y., & Krishnamurthy, R. (2014). The adop-tion of a climate disaster resilience index in Chennai, India.Disasters, 38(3), 540–561.

Jones, L., Ludi, E., Jeans, H., & Barihaihi, M. (2019). Revisiting the localadaptive capacity framework: Learning from the implementation of a

research and programming framework in Africa. Climate andDevelopment, 11(1), 3–13.

Kapucu, N., Hawkins, C. V., & Rivera, F. (2013). Disaster resiliency:Interdisciplinary perspectives. In N. Kapucu, C. V. Hawkins, & F. I.Rivera (Eds.), Disaster resiliency: Interdisciplinary perspectives (pp. 1–14). New York, London: Routledge.

Kasperson, R. E., & Kasperson, J. K. (2005). Climate change, vulnerabilityand social justice. In J. X. Kasperson, & R. E. Kasperson (Eds.), Thesocial contours of risk: Publics, risk communication and the socialamplification of risk (pp. 301–321). Sterling, VA: Earthscan.

Kim, H., & Marcouiller, D. W. (2015). Urban vulnerability and resiliencyto natural disasters: An integrative tourism planning perspective. In P.Filion, G. Sands, & M. Skidmore (Eds.), Cities at risk: Planning for andrecovering from natural disasters (pp. 158–187). Surry, UK: AshgatePublishing.

Kim, H., & Marcouiller, D. W. (2016). Natural disaster response, commu-nity resilience, and economic capacity: A case study of coastal Florida.Society and Natural Resources, 29(8), 981–997.

Kim, H., & Marcouiller, D. W. (2018). Mitigating flood risk and enhancingcommunity resilience to natural disasters: Plan quality matters.Environmental Hazards, 17(5), 397–417.

Kim, H., Marcouiller, D. W., & Woosnam, K. M. (2018). Rescaling socialdynamics in climate change: The implications of cumulative exposure,climate justice, and community resilience. Geoforum, 96, 129–140.

LeSage, J. P., & Pace, R. K. (2010). Spatial econometric models. In M. M.Fischer & A. Getis (Eds.), Handbook of applied spatial analysis:Software tools, methods and applications (pp. 355–376). London:Springer.

Ludi, E., Wiggins, S., Jones, L., Lofthouse, J., & Levine, S. (2014). Adaptingdevelopment: How wider development interventions can support adap-tive capacity at the community level. In E. L. F. Schipper, J. Ayers, H.Reid, S. Huq, & A. Rahman (Eds.), Community-based adaptation to cli-mate change: Scaling it up (pp. 36–51). London: Routledge.

McNamara, K. E., & Buggy, L. (2017). Community-based climate changeadaptation: A review of academic literature. Local Environment, 22(4), 443–460.

Magis, K. (2010). Community resilience: An indicator of social sustainabil-ity. Society and Natural Resources, 23, 401–416.

Malakar, K., & Mishra, T. (2017). Assessing socio-economic vulnerabilityto climate change: A city-level index-based approach. Climate andDevelopment, 9(4), 348–363.

Manyena, S. B., O’Brien, G., O’Keefe, P., & Rose, J. (2011). Disaster resili-ence: A bounce back or bounce forward ability? Local Environment, 16(5), 417–424.

Masterson, J. H., Peacock, W. G., Van Zandt, S. S., Grover, H., Schwarz, L.F., & Cooper, J. T. (2014). Planning for community resilience: A hand-book for reducing vulnerability to disasters. . London: Island Press.

May, P. J. (2013). Public risks and disaster resilience: Rethinking publicand private sector roles. In N. Kapucu, C. V. Hawkins, & F. I. Rivera(Eds.), Disaster resiliency: Interdisciplinary perspectives (pp. 126–145).New York, NY: Routledge.

Mearns, R., & Norton, A. (2010). Social dimensions of climate change:Equity and vulnerability in a warming world. Washington, DC: TheWorld Bank.

Mohan, D., & Sinha, S. (2016). Assessing vulnerability to climate change inthe Ganges Basin using a combined macro-and micro-level approach.Climate and Development, 8(4), 312–323.

Newman, S., Carroll, M., Jakes, P., & Higgins, L. (2014). Hurricanes andwildfires: Generic characteristics of community adaptive capacity.Environmental Hazards, 13(1), 21–37.

O’Brien, K., & Leichenko, R. (2000). Double exposure: Assessing theimpacts of climate change within the context of economic globalization.Global Environmental Change, 10, 221–232.

Oulahen, G., Mortsch, L., Tang, K., & Harford, D. (2015). Unequal vulner-ability to flood hazards: “ground truthing” a social vulnerability index offive municipalities in metro Vancouver, Canada. Annals of theAssociation of American Geographers, 105, 473–495.

Parry, L., Davies, G., Almeida, O., Frausin, G., de Moraés, A., Rivero, S.,…Torres, P. (2018). Social vulnerability to climate shocks is shaped by

12 H. KIM AND D. W. MARCOUILLER

urban accessibility. Annals of the American Association of Geographers,108, 125–143.

Peacock, W. G., Zandt, S. V., Henry, D., Grover, H., & Highfield, W.(2012). Using social vulnerability mapping to enhance coastal commu-nity resiliency in Texas. In P. B. Bedient (Ed.), Lessons from HurricaneIke (pp. 66–81). College Station, TX: Texas A&M University.

Pike, A., Dawley, S., & Tomaney, J. (2010). Resilience, adaptation and adap-tability. Cambridge Journal of Regions, Economy and Society, 3, 59–70.

Reid, H., & Schipper, E. L. F. (2014). Upscaling community-based adap-tation: An introduction to the edited volume. In E. L. F. Schipper, J.Ayers, H. Reid, S. Huq, & A. Rahman (Eds.), Community-based adap-tation to climate change: Scaling it up (pp. 3–21). Earthscan fromRoutledge, London: Routledge.

Ross, A. D. (2014). Local disaster resilience: Administrative and politicalperspectives. New York, NY: Routledge.

Ross, H., & Berkes, F. (2014). Research approaches for understanding,enhancing, and monitoring community resilience. Society andNatural Resources, 27, 787–804.

Schipper, E. L. F., Ayers, J., Reid, H., Huq, S., & Rahman, A. (2014).Community-based adaptation to climate change: Scaling it up.Earthscan from Routledge, London: Routledge.

Sherrieb, K., Norris, F. H., & Galea, S. (2010). Measuring capacities forcommunity resilience. Social Indicators Research, 99, 227–247.

Smit, B., & Wandel, J. (2006). Adaptation, adaptive capacity and vulner-ability. Global Environmental Change, 16(3), 282–292.

Tierney, K. (2012). Disaster governance: Social, political, and economicdimensions. Annual Review of Environment and Resources, 37, 341–363.

Tierney, K. (2014). The social roots of risk: Producing disasters, promotingresilience. Stanford, CA: Stanford Business Books.

Turner, B. L., Kasperson, R. E., Matson, P. A., McCarthy, J. J., Corell, R. W.,Christensen, L., & Schiller, A. (2003). A framework forvulnerability analysis in sustainability science. Proceedings of theNational Academy of Sciences of the United States of America, 100(14), 8074–8079.

United Nations. (2015). Transforming our world: The 2030 agenda for sus-tainable development. New York, NY: United Nations.

Walker, B., Carpenter, S., Anderies, J., Abel, N., Cumming, G., Janssen, M.,… Pritchard, R. (2002). Resilience management in social-ecological sys-tems: A working hypothesis for a participatory approach. ConservationEcology, 6(1), 14.

Walker, B., Holling, C. S., Carpenter, S. R., & Kinzig, A. (2004). Resilience,adaptability and transformability in social-ecological systems. Ecologyand Society, 9(2), 5.

Walker, B., & Salt, D. (2006). Resilience thinking: Sustaining ecosys-tems and people in a changing world. Washington, DC: IslandPress.

Waters, J., & Adger, W. N. (2017). Spatial, network and temporal dimen-sions of the determinants of adaptive capacity in poor urban areas.Global Environmental Change, 46, 42–49.

Weichselgartner, J., & Kelman, I. (2015). Geographies of resilience:Challenges and opportunities of a descriptive concept. Progress inHuman Geography, 39(3), 249–267.

Wilson, G. A. (2012). Community resilience and environmental transitions.London: Routledge.

CLIMATE AND DEVELOPMENT 13