Content uploaded by Karen Wyche
Author content
All content in this area was uploaded by Karen Wyche
Content may be subject to copyright.
ORIGINAL PAPER
Community Resilience as a Metaphor, Theory, Set of Capacities,
and Strategy for Disaster Readiness
Fran H. Norris ÆSusan P. Stevens Æ
Betty Pfefferbaum ÆKaren F. Wyche Æ
Rose L. Pfefferbaum
Published online: 22 December 2007
Springer Science+Business Media, LLC 2007
Abstract Communities have the potential to function
effectively and adapt successfully in the aftermath of
disasters. Drawing upon literatures in several disciplines,
we present a theory of resilience that encompasses con-
temporary understandings of stress, adaptation, wellness,
and resource dynamics. Community resilience is a process
linking a network of adaptive capacities (resources with
dynamic attributes) to adaptation after a disturbance or
adversity. Community adaptation is manifest in population
wellness, defined as high and non-disparate levels of
mental and behavioral health, functioning, and quality of
life. Community resilience emerges from four primary sets
of adaptive capacities—Economic Development, Social
Capital, Information and Communication, and Community
Competence—that together provide a strategy for disaster
readiness. To build collective resilience, communities must
reduce risk and resource inequities, engage local people in
mitigation, create organizational linkages, boost and pro-
tect social supports, and plan for not having a plan, which
requires flexibility, decision-making skills, and trusted
sources of information that function in the face of
unknowns.
Keywords Resilience Community resilience
Disaster Preparedness
Resilience as a Metaphor
When applied to people and their environments, ‘‘resil-
ience’’ is fundamentally a metaphor. With roots in the
sciences of physics and mathematics, the term originally
was used to describe the capacity of a material or system to
return to equilibrium after a displacement. A resilient
material, for example, bends and bounces back, rather than
breaks, when stressed (Bodin and Wiman 2004; Gordon
1978). In physics, resilience is not a matter of how large the
initial displacement is or even how severe the oscillations
are but is more precisely the speed with which homeostasis
is achieved. The image is a compelling one, capable of
sparking human imagination, as it clearly did for Holling
(1973) in his original and influential thesis about ‘‘eco-
logical resilience.’’ The concept of resilience has since
been applied to describe the adaptive capacities of indi-
viduals (e.g., Bonanno 2004; Butler et al. 2007; Rutter
1993; Werner and Smith 1982), human communities (e.g.,
Brown and Kulig 1996/97; Sonn and Fisher 1998), and
larger societies (e.g., Adger 2000; Godschalk 2003). As
references to resilience have continued to increase, so too
F. H. Norris S. P. Stevens
Dartmouth Medical School, Hanover, NH, USA
F. H. Norris (&)S. P. Stevens
Department of Veterans Affairs National Center for PTSD
(NCPTSD), VA Medical Center (116D), 215 North Main Street,
White River Junction, VT 05009, USA
e-mail: fran.norris@dartmouth.edu
F. H. Norris S. P. Stevens B. Pfefferbaum
K. F. Wyche R. L. Pfefferbaum
National Consortium for the Study of Terrorism and Responses
to Terrorism, College Park, MD, USA
B. Pfefferbaum K. F. Wyche
University of Oklahoma Health Sciences Center, Oklahoma
City, USA
B. Pfefferbaum K. F. Wyche R. L. Pfefferbaum
Terrorism and Disaster Center, National Child Traumatic Stress
Network, Oklahoma City, USA
R. L. Pfefferbaum
Phoenix College, Phoenix, USA
123
Am J Community Psychol (2008) 41:127–150
DOI 10.1007/s10464-007-9156-6
have criticisms that the concept may be inappropriate,
imprecise, or ‘‘glittery’’ (e.g., Bodin and Winman 2004;
Carpenter et al. 2001; Cowen 2001; Klein et al. 2003).
Looking back, one wonders if perhaps the social and psy-
chological sciences should have created their own
language, free from inherited meanings, but the term is
probably here to stay. Its metaphorical origins notwith-
standing, human resilience, we believe, must now be
studied on its own terms without undue concern with how
those meanings correspond to known physical properties or
laws. We return to the question of the meaning and defi-
nitions of resilience and related concepts in the next section
of this paper.
The concept of ‘‘community resilience’’ raises the same
concerns as the concept of resilience per se, but is further
complicated by variation in the meaning of community.
Not always, but typically, a community is an entity that has
geographic boundaries and shared fate. Communities are
composed of built, natural, social, and economic environ-
ments that influence one another in complex ways. Past
writings on community resilience have described every-
thing from grass-roots groups and neighborhoods to
complex amalgams of formal institutions and sectors in
larger geo-political units. This is not inappropriate, as
resilience can be understood and addressed at different
levels of analysis. Discussions of community resilience
often note that the ‘‘whole is more than the sum of its
parts,’’ meaning that a collection of resilient individuals
does not guarantee a resilient community (e.g., Pfeffer-
baum et al. 2005; Rose 2004). As Brown and Kulig (1996/
97, p. 43) observed, ‘‘People in communities are resilient
together, not merely in similar ways.’’
The primary focus of this paper is community resilience
as it applies to disasters. We adopted McFarlane and
Norris’ (2006, p. 4) definition of disaster as ‘‘a potentially
traumatic event that is collectively experienced, has an
acute onset, and is time delimited; disasters may be
attributed to natural, technological, or human causes.’’ This
definition includes acts of nature, such as hurricanes,
floods, and earthquakes; large industrial, transportation,
and nuclear accidents; and episodes of mass violence, such
as terrorist attacks and shooting sprees. It excludes chronic
environmental hazards, ongoing community and political
violence, war, and epidemics, not because they are less
important but because the dynamics of how such stressors
unfold over time are different enough to warrant bound-
aries of the potential applicability of theory and research.
Still, much of what is known or proposed here about
community resilience was gleaned from or applies equally
well to other types of collective stressors and adversities.
Several other papers on this topic have appeared in
recent years, and two pathways to the present state of
knowledge are discernible. One path has been relatively
more concerned with community resilience as it prevents
disaster-related health or mental health problems of com-
munity members. The other path has been relatively more
concerned with community resilience as it describes
effective organizational behavior and disaster management.
These two paths have sometimes, but not always, informed
each other. Although our work primarily follows the first
path, our goal was to integrate the various ways past writers
have conceptualized (and occasionally researched) resil-
ience and related terms. After all, postdisaster community
health depends in part on the effectiveness of organiza-
tional responses, and ultimately the purpose of disaster
management is to ensure the safety and well-being of the
public. In addition to these two literatures on community
resilience, we drew upon resilience literature in the fields
of psychology, sociology, geography, anthropology, public
health, ecology, technology, and communications
1
and
upon past stress theory and disaster research. In this paper,
we begin by discussing resilience as a theory. We present a
model that integrates various conceptualizations and define
each of the model’s components. We next discuss com-
munity resilience as a set of capacities and review themes
in past work that inform hypotheses about these capacities,
such as social capital. We then discuss community resil-
ience as a strategy for promoting effective disaster
readiness and response. We conclude with several caveats
or complexities that must be recognized and addressed in
future resilience theory, research, and application.
1
We used a wide array of search techniques in the literature review
including controlled vocabulary (for example, in the PsycINFO
database, using the term ‘‘resilience’’ which is included in the
PsycINFO thesaurus) and free text searching (for example, using the
term ‘‘community resilience,’’ which is not a term in the PsycINFO
thesaurus but would cue the database to identify papers with
community resilience in their titles or abstracts). Because one of
the goals of this project was to discover what had been written about
community resilience and how resilience was already conceptualized
in the literature, the initial search strategy was to cast as wide a net as
possible. In this way, free text searches using the terms ‘‘resilience’’
and ‘‘community resilience’’ were conducted in a variety of databases
(i.e., PsychINFO, PubMed, ERIC, PILOTS, Academic Search
Premier) to capture as many articles as possible that used the term.
As one might imagine, this broad search strategy yielded a diverse (an
oftentimes irrelevant) combination of articles from disciplines as
disparate as economics and food processing. The broad search
illustrated, if nothing else, the popularity of the term and the infinite
ways resilience is used to describe varying concepts across disci-
plines. Free text searches can also be helpful when the term is new
and perhaps has not yet been added to a thesaurus. Since community
resilience is a relatively new term, efforts were made to find articles
that might discuss the spirit of community resilience without using the
term community resilience. Examples of terms used to ferret out such
papers (used individually and in combination) included: cohesion,
adaptability, empowerment, mobilization, collective capacity, and
collective healing. In a more disciplined and focused search,
controlled vocabulary strategies were used to adhere to the search
terms used by specifics databases.
128 Am J Community Psychol (2008) 41:127–150
123
Resilience as a Theory
Resilience has been defined in a variety of ways (see
Table 1). Most definitions emphasize a capacity for suc-
cessful adaptation in the face of disturbance, stress, or
adversity. Although there are exceptions, most discussions,
if not the definitions themselves, distinguish resilience
from ‘‘resistance.’’ In mathematics and technology, resis-
tance refers to the force required to displace the system
from equilibrium, whereas resilience refers to the time
Table 1 Representative definitions of resilience
Citation first
author, year
Level of
analysis
Definition
Gordon, 1978 Physical The ability to store strain energy and deflect elastically under a load without breaking or being deformed
Bodin, 2004 Physical The speed with which a system returns to equilibrium after displacement, irrespective of how many
oscillations are required
Holling, 1973 Ecological
system
The persistence of relationships within a system; a measure of the ability of systems to absorb changes of
state variables, driving variables, and parameters, and still persist
Waller, 2001 Ecological
system
Positive adaptation in response to adversity; it is not the absence of vulnerability, not an inherent
characteristic, and not static
Klein, 2003 Ecological
system
The ability of a system that has undergone stress to recover and return to its original state; more precisely
(i) the amount of disturbance a system can absorb and still remain within the same state or domain of
attraction and (ii) the degree to which the system is capable of self-organization (see also Carpenter
et al. 2001)
Longstaff, 2005 Ecological
system
The ability by an individual, group, or organization to continue its existence (or remain more or less
stable) in the face of some sort of surprise….Resilience is found in systems that are highly adaptable
(not locked into specific strategies) and have diverse resources
Resilience Alliance,
2006
Ecological
system
The capacity of a system to absorb disturbance and reorganize while undergoing change so as to still
retain essentially the same function, structure and feedbacks—and therefore the same identity.
(Retrieved 10/16/2006 from http://www.resalliance.org/564.php)
Adger, 2000 Social The ability of communities to withstand external shocks to their social infrastructure
Bruneau, 2003 Social The ability of social units to mitigate hazards, contain the effects of disasters when they occur, and carry
out recovery activities in ways that minimize social disruption and mitigate the effects of future
earthquakes
Godschalk, 2003 City A sustainable network of physical systems and human communities, capable of managing extreme
events; during disaster, both must be able to survive and function under extreme stress
Brown, 1996 Community The ability to recover from or adjust easily to misfortune or sustained life stress
Sonn, 1998 Community The process through which mediating structures (schools, peer groups, family) and activity settings
moderate the impact of oppressive systems
Paton, 2000 Community The capability to bounce back and to use physical and economic resources effectively to aid recovery
following exposure to hazards
Ganor, 2003 Community The ability of individuals and communities to deal with a state of continuous, long term stress; the ability
to find unknown inner strengths and resources in order to cope effectively; the measure of adaptation
and flexibility
Ahmed, 2004 Community The development of material, physical, socio-political, socio-cultural, and psychological resources that
promote safety of residents and buffer adversity
Kimhi, 2004 Community Individuals’ sense of the ability of their own community to deal successfully with the ongoing political
violence
Coles, 2004 Community A community’s capacities, skills, and knowledge that allow it to participate fully in recovery from
disasters
Pfefferbaum, 2005 Community The ability of community members to take meaningful, deliberate, collective action to remedy the impact
of a problem, including the ability to interpret the environment, intervene, and move on
Masten, 1990 Individual The process of, capacity for, or outcome of successful adaptation despite challenging or threatening
circumstances
Egeland, 1993 Individual The capacity for successful adaptation, positive functioning, or competence…despite high-risk status,
chronic stress, or following prolonged or severe trauma
Butler, 2007 Individual Good adaptation under extenuating circumstances; a recovery trajectory that returns to baseline
functioning following a challenge
Note. Because of our focus, definitions of community resilience are presented here in disproportionate frequency. Definitions describing larger
(ecological) and smaller (individual) levels of analysis were representative of others in the literature
Am J Community Psychol (2008) 41:127–150 129
123
required for the system to return to equilibrium once dis-
placed (Bodin and Wiman 2004). Across these definitions,
there is general consensus on two important points: first,
resilience is better conceptualized as an ability or process
than as an outcome (Brown and Kulig 1996/97; Pfeffer-
baum et al. 2005); and second, resilience is better
conceptualized as adaptability than as stability (e.g.,
Handmer and Dovers 1996; Waller 2001). In fact, in some
circumstances, stability (or failure to change) could point
to lack of resilience. The resilience of systems, for exam-
ple, depends upon one component of the system being able
to change or adapt in response to changes in other com-
ponents; and thus the system would fail to function if that
component remained stable (Adger 2000; Klein et al.
2003). Adaptability also takes different forms. ‘‘Engineer-
ing resilience’’ makes a system return to one pre-designed
state or function after a disturbance, whereas ‘‘ecological
resilience’’ allows for many possible desirable states that
match the environment (Gunderson 2000). The second type
of resilience is probably the relevant one for human com-
munities, organizations, and societies.
With these past definitions in mind, we define resilience
as: a process linking a set of adaptive capacities to a
positive trajectory of functioning and adaptation after a
disturbance. Adaptive capacities are resources with
dynamic attributes, a point to which we will return
momentarily. The specific elements of the definition would
be operationalized differently depending upon the level of
analysis, be it an individual, grass-roots group, larger
human community, or entire ecology, but our basic defi-
nition aims to apply across levels. In our definition, we
carefully did not equate resilience with the outcome, but
rather with the process linking resources (adaptive capac-
ities) to outcomes (adaptation).
Figure 1introduces a variety of concepts that we elab-
orate upon in the discussion to follow. Table 2provides a
summary of definitions of key terms in the model.
Although the notion of positive trajectories occurs in recent
writings (Butler et al. 2007), our model of resilience owes
its fundamental structure to Dohrenwend’s (1978) model of
psychosocial stress, presented as part of her presidential
address to Division 27 (Community Psychology) of the
American Psychological Association. Our model has been
updated, however, to encompass contemporary theories of
stress, to use the nomenclature common in discussions of
resilience, and to apply across multiple levels of analysis
(e.g., persistent dysfunction has replaced psychopathology
as the adverse outcome).
Fig. 1 Model of stress resistance and resilience over time: Resistance
occurs when resources are sufficiently robust, redundant, or rapid to
buffer or counteract the immediate effects of the stressor such that no
dysfunction occurs. Total resistance is hypothesized to be rare in the
case of severe, enduring, or highly surprising events, making transient
situational dysfunction the more likely and normative result in the
immediate aftermath of disasters. Resilience occurs when resources
are sufficiently robust, redundant, or rapid to buffer or counteract the
effects of the stressor such that a return to functioning, adapted to the
altered environment, occurs. For human individuals and communities,
this adaptation is manifest in wellness. Vulnerability occurs when
resources were not sufficiently robust, redundant, or rapid to create
resistance or resilience, resulting in persistent dysfunction. The more
severe, enduring, and surprising the stressor, the stronger the
resources must be to create resistance or resilience
130 Am J Community Psychol (2008) 41:127–150
123
Stress and Stress Resistance
Our model begins when a stressor, such as a disaster,
occurs. Stressors are aversive circumstances that threaten
the well-being or functioning of the individual, organiza-
tion, neighborhood, community, or society. Models of the
stress process typically consider (a) characteristics of the
stressor, (b) appraisals of the stressor, (c) the response to or
effects of the stressor, and (d) various conditions that
influence the relations between the stressor, stress apprai-
sal, and stress response. In stress-diathesis theory, level of
exposure is proposed to interact with pre-existing vulner-
abilities to influence the stress response. The influence of
this theory is evident in some of the earliest published
frameworks for disaster research (see Benight et al. 2006,
for further discussion of these points).
Stressors vary on a number of dimensions. Important
dimensions for characterizing disasters include severity,
duration, and surprise. Severity of exposure is a consistent
risk factor for adverse psychosocial consequences in the
aftermath of disasters (Norris et al. 2002a,b). Specific
stressors that have been found to affect postdisaster mental
health include bereavement, injury to self or family
member, life threat, property damage, financial loss, com-
munity destruction, and displacement. Disasters are
stressful not only for individuals experiencing personal loss
but also for the community-at-large (e.g., Norris et al.
1994). Communities share damages and disruptions to the
various environments of which they are composed.
Destruction of a ‘‘keystone’’ neighborhood, such as a
central business district, may challenge an entire city or
region (Fullilove and Saul 2006). And, in extreme cases,
such as the September 11th terrorist attacks, the stress may
be felt by an the entire nation (Silver et al. 2002).
Although disasters are rarely anticipated for more than a
few days in advance, if at all, they vary in duration of
environmental disruption. For example, after Hurricane
Katrina in August 2005, the most badly affected neigh-
borhoods in New Orleans remained uninhabitable for years
after the levee breech. The stressors were of long duration
even though the floodwaters receded within weeks. Certain
technological disasters may initiate long periods of threat
related to contamination; they essentially evolve into
chronic hazards, such as happened in the case of the
Chernobyl nuclear disaster (Bromet et al. 2000).
Longstaff (2005) provided a particularly thought-pro-
voking treatment of the nature of the stressors confronted
by systems and communities. She heavily emphasized the
unpredictable nature of the dangers confronting contem-
porary planners. Our world is increasingly interconnected
and complex, with systems in constant flux in reaction to
changes in other systems, making surprise more common
than predictability. Some dangers are familiar but unpre-
dictable as to where or when they will happen (known
Table 2 Key terms in the theoretical model
Concept Definition
Resilience A process linking a set of adaptive capacities to a positive trajectory of functioning and adaptation after a disturbance
Community
resilience
A process linking a set of networked adaptive capacities to a positive trajectory of functioning and adaptation in constituent
populations after a disturbance
Resilient (adj.) Having shown, currently showing, or eventually showing a positive trajectory of functioning and adaptation after an initial
disturbance as a result of adequate adaptive capacities
Adaptive capacities Resources with dynamic attributes, i.e., resources that are robust, redundant, or rapidly accessible
Resources Objects, conditions, characteristics, and energies that people value
Robustness One of three dynamic attributes of resources; resource strength, in combination with a low probability of resource
deterioration
Redundancy One of three dynamic attributes of resources; the extent to which elements are substitutable in the event of disruption or
degradation
Rapidity One of three dynamic attributes of resources; how quickly the resource can be accessed and used (mobilized)
Resilience-
resources
Synonymous with adaptive capacities
Resilience-
outcomes
The end result of resilience, characterized here as wellness for individuals and populations
Psychological
wellness
An individual-level outcome indicative of successful adaptation defined according to four criteria: (1) absence of stress-
related psychological disorders; (2) healthy patterns of behavior; (3) adequate role functioning at home, school, and/or
work; and (4) high quality of life
Population
wellness
A community-level outcome indicative of successful adaptation, defined as high and non-disparate levels of mental and
behavioral health, role functioning, and quality of life in constituent populations
Quality of life One component of wellness that captures how people generally feel about their lives as a whole and in domains of work or
school, family, health, leisure, and neighborhood.
Am J Community Psychol (2008) 41:127–150 131
123
unknowns), but some dangers are new: ‘‘We don’t know
about and won’t know about these until they happen’’
(unknown unknowns). Longstaff used the concept of
‘‘surprise’’ to capture this discrepancy between what is
expected and what is experienced. These surprises are
nearly impossible to predict or prepare for, and thus call for
broad resilience strategies (Allenby and Fink 2005).
The ‘‘crisis’’ in Fig. 1symbolizes a hypothetical bal-
ancing act between stressors and resources. Do demands
outweigh resources? The implied interaction has a long
history in resilience theory (Rutter 1987) and is extremely
important in disaster theory. In fact, Quarantelli (1986)
defined disaster as a consensus-type crisis occasion where
demands exceed capabilities. Quarantelli sought to remind
us that the consequences of disasters follow not only from
the needs of the community but from the community’s
capacity to meet those needs. Longstaff (2005, p. 16)
described a crisis as ‘‘what happens when a surprise reveals
a failure of the rules, norms, behavior or infrastructure used
to handle that type of surprise.’’ It has become almost a
cliche
´to note that crises bring opportunities for growth as
well as risk for harm, as this idea has been with us for a
long time (Caplan 1964).
The ideal outcome after the crisis is resistance, meaning
that the resources have effectively blocked the stressor and,
accordingly, there is virtually no dysfunction, no mater
how temporary. Individuals benefit from resistance strate-
gies on a daily basis, in that the human immune system is
one of the most effective resistance strategies known to
exist. Resistance strategies are appropriate for dangers that
are likely to happen with some frequency and can be
planned for, such as fires, but they are not likely to be fully
effective against surprises (Longstaff 2005). Nonetheless,
there are critical systems that aspire for resistance in the
aftermath of disaster. Earthquake- and fire-resistant build-
ings, redundant power sources, and terrorist surveillance
and detection systems are but a few examples of actions
that help communities to prevent or resist disaster.
Resilience and Adaptation
Resistance is an unlikely course for individuals and com-
munities (and actually even for emergency systems) in the
aftermath of disasters, which brings us to the notion of
transient dysfunction, similar to Dohrenwend’s (1978)
concept of the transient stress reaction (see Fig. 1).
Disaster research indicates that the most typical pattern is
for distress to be nearly universal in the first weeks or
months postdisaster, even though only a minority of par-
ticipants experience criterion-level psychopathology
(Norris et al. 2002a,b). Most longitudinal disaster studies
find that an event’s adverse effects dissipate over time,
leaving only a minority of communities and a minority of
individuals within those communities chronically impaired.
Most of the time—especially if the severity of the stressor
has lessened and resources have been replenished—tran-
sient dysfunction is followed by a return to predisaster
levels of functioning.
The process that produces adapted outcomes is resil-
ience; the more rapid the return to pre-event functioning,
the greater the resilience. In his influential paper on psy-
chological resilience, Bonanno (2004) differentiated
between recovery and resilience trajectories. The former he
characterized as involving a period of dysfunction lasting
several months or more, followed by a gradual return to
pre-event functioning. A trajectory of resilience, on the
other hand, may involve transient perturbations, lasting as
long as several weeks, but it generally involves a stable
trajectory of healthy functioning. Most research does not
clearly distinguish between these two trajectories.
Our primary point here is that resilience does not pre-
clude dysfunction or distress. It is now commonly accepted
in the disaster field that some distress is a normal reaction
to an abnormal event (e.g., Flynn 1994). However, the
dysfunction or distress is transient, followed by a return to
functioning. Postevent functioning may not be the quali-
tative equivalent of pre-event functioning when there is a
need to adapt to an altered environment. The phrase ‘‘new
normal’’ was heard often in the aftermath of the September
11th terrorist attacks, as people speculated on America’s
capacity to adapt to the ongoing threat of terrorism and the
demands and inconveniences of heightened security
(Redlener and Morse 2006). We do not characterize this
adaptive trajectory as ‘‘growth,’’ although others might
legitimately dispute this choice (Brown and Kulig 1996/97;
Linley and Joseph 2005). Adapted functioning is not nec-
essarily superior in level or character or effectiveness to
pre-event functioning; it is simply different.
The alternative outcome to renewed and adapted func-
tioning is persistent dysfunction. It results from
‘‘vulnerability,’’ defined roughly as the antonym of stress
resistance or resilience. We say little about vulnerability in
this paper, but it has been and should remain an important
concept in hazard management (e.g., Cutter et al. 2003,
2006).
Wellness as the Manifestation of Adaptation
What is the end result of this process? Adaptation is the
theoretical result, but how is it manifest? This has been
debated. In the mental health arena, for example,
researchers have proposed that the appropriate individual
outcome is minimal impairment of functioning despite
distress, rapid recovery from distress, no distress at all, and
132 Am J Community Psychol (2008) 41:127–150
123
adversarial growth (Bonanno 2005; Linley and Joseph
2005; Litz 2005). In practice, resilience often has been
operationalized as the absence of psychopathology, which
is highly inadequate (Rutter 1993). A strict focus on
prevalence of disorders may cause us to erroneously or
prematurely conclude that a community has recovered
from the event when there is still considerable distress,
dysfunction, or dissatisfaction present, and it dismisses the
adverse consequences of suffering that does not qualify for
strict definitions of pathology. Here, we first discuss indi-
vidual resilience-outcomes and then turn our attention to
community resilience-outcomes.
Psychological Wellness
We suggest that ‘‘wellness’’ is a viable indicator that
adaptation has occurred on the individual level. We base
this conclusion largely on the writings of Cowen (1983,
1994,2000), who contended that psychological wellness
provides a broad and integrative frame for psychological
questions and activities, including those related to primary
prevention, empowerment, and resilience. By arguing that
‘‘wellness must be a matter of prime concern at all times,
not just when it fails,’’ Cowen (2000, p. 80) aimed to move
psychology beyond a focus on repairing dysfunction
(including efforts to increase the reach of services) to a
prime focus on promoting wellness. According to Cowen,
one advantage of this emphasis on wellness, as opposed to
some competing frames, was that it retained continuity
with the traditions and past developments in the field of
mental health. This is cogent advice for applying our work
to disaster readiness and recovery: a connection to wellness
places community resilience solidly within the domains of
concern for local, state, and federal policy-makers charged
with protecting health and behavioral health in the after-
math of major disasters and terrorist attacks.
As a manifestation of adaptation, we define psycholog-
ical wellness according to four criteria: (1) absence of
psychopathology; (2) healthy patterns of behavior; (3)
adequate role functioning at home, school, and/or work; and
(4) high quality of life. Quality of life captures how people
generally feel about their lives as a whole and in domains of
work, family life, health, leisure, and neighborhood (e.g.,
Zautra and Bachrach 2000). Our definition of wellness is
not identical to Cowen’s (2000, p. 83) but consistent with it,
as well as with areas of concern for public health and
behavioral health policies and programs. A criterion of
wellness serves to remind us that we must attend to disaster
victims’ abundant problems in living that may interfere with
their quality of life (Norris et al. 2002a,b).
Importantly, this definition retains the criterion of being
free of psychological disorder even though wellness is
more than this. Stress-related disorders remain a relevant
area for postdisaster research and intervention. It would be
unfortunate if a focus on resilience and wellness caused us
to neglect those persons who fare most poorly in disasters.
A growing body of research shows that posttraumatic stress
disorder, major depressive disorder, and other mental ill-
nesses strongly and adversely influence functioning and
quality of life (Kessler 2000; Thorp and Stein 2005).
Mental health, healthy behavior, adequate functioning, and
quality of life are interrelated but conceptually distinct.
Wellness might actually be a ‘‘higher bar’’ than has been
used in resilience research, but it is an appropriate standard
for concluding that adaptation to an altered environment
has occurred. As Cowen (2000) also noted, wellness is a
continuum. Individuals (and communities) show varying
degrees of wellness before as well as after disasters, and
this context must be taken into account in assessing post-
disaster adaptation.
Population Wellness
But what about community resilience? By what standard
do we conclude that a community has adapted? Although
we recognize that a community is not merely the sum total
(or average) of its members, we recommend that commu-
nity-level adaptation be understood as ‘‘population
wellness,’’ a high prevalence of wellness in the community,
defined as high and non-disparate levels of mental and
behavioral health, role functioning, and quality of life in
constituent populations. There are three reasons for this
choice: first, although emergency management systems are
not directly responsible for health, if they function effec-
tively to protect lives, reduce injuries, minimize damage to
public utilities, and connect community members to nec-
essary services, it is reasonable to expect the population to
remain well. Wellness is one criterion that should apply in
most circumstances. Second, wellness levels in the com-
munity can be monitored in postdisaster needs assessment
and surveillance initiatives to guide resource allocation
(Galea and Norris 2006). Third, this definition keeps the
outcome conceptually distinct from the community
resources that promote resilience (and wellness, in turn).
Although much of the research on disasters has focused
on individual-level outcomes (Norris et al. 2002a,b),
effects can be observed at the community or population
level (Galea and Resnick 2005). In a prospective study of a
flood in eastern Kentucky (Appalachia), Norris et al. (1994)
differentiated between primary victims (those with personal
losses), secondary victims (others living in the flooded
counties), and non-victims (persons living in neighboring,
non-flooded counties). They documented community-wide
tendencies for residents to feel less positive about their
Am J Community Psychol (2008) 41:127–150 133
123
social networks and surroundings, less enthusiastic and
energetic, and less able to enjoy life after the flood. No one
would suggest that such consequences constitute psycho-
pathology, but they do indicate that disasters sometimes
impair the quality of life in the community for quite some
time. It is important to note that population wellness is not
captured fully by average behavioral health or quality of
life. Abnormal variability might be of greater concern,
suggesting that disparities in mental and behavioral health
within the population have been created or exacerbated by
the disaster or stressor (Galea et al. 2005).
Determinants of population wellness cannot be identi-
fied in studies of individuals within a single population.
This may sound blatantly obvious, but many examples to
the contrary exist in the literature, compelling us to
underscore this point. Within a population, the distribution
of illness is determined by individual susceptibility. Rose
(2001) shared several excellent examples from public
health research. One example concerned blood pressure in
Kenya and London. In both populations, blood pressure
varies. Through traditional risk factor research, we could
determine why some Kenyans have higher blood pressure
than others and why some Londoners have higher blood
pressure than others, but we would fail to observe that the
London setting shifted the entire distribution upward. Thus
to find the determinants of prevalence and incidence rates
we need to study characteristics of populations rather than
the characteristics of individuals.
These two approaches to etiology (individual and pop-
ulation) have their counterparts in prevention. The
population strategy attempts to control mean level of risk
factors, to shift the whole distribution in a favorable
direction. This shift may bring much benefit to the popu-
lation by reducing the incidence or prevalence of a
problem, but it offers only a small benefit to each indi-
vidual. This is the ‘‘prevention paradox’’ (Rose 1981).
Effects that seem small in analyses of individuals may be
quite large when extrapolated to entire populations.
Dynamic Attributes of Resilience-Resources
What determines which process (and consequent outcome)
occurs? We believe this depends on a host of objects,
conditions, characteristics, and energies that people
value—that is, resources (Hobfoll 1988,2006). Later, we
review the specific resources that have been shown or
theorized to influence community resilience but here our
primary goal is to identify the attributes of resources that
make them available and accessible for buffering or
counteracting stress.
We found the discussion of Bruneau et al. (2003)
especially enlightening for thinking about the dynamic
attributes that resources must have to engender resilience.
Bruneau et al. theorized that resilient systems have four
key properties. With unusual specificity for discussions of
resilience, Bruneau and colleagues provided numerous
examples of how their four properties could be operation-
alized for monitoring the performance of the technical,
organizational, social, and economic systems that compose
a community.
‘‘Robustness’’ is the ability to withstand stress without
suffering degradation. A resistance strategy is robust if it
keeps out or counteracts a wide variety of dangers, but it is
fragile if it works only under a small number of possible
scenarios (Longstaff 2005). Because of the importance of
resource loss in stress theory (more discussion to follow),
we characterize robustness as resource strength, in com-
bination with a low probability of resource deterioration.
‘‘Redundancy’’ is the extent to which elements are
substitutable in the event of disruption or degradation.
Many technological systems (airplanes, power grids) have
extensive redundancy built into them. People build in
redundancy too by having larger social networks or by
having more than one way to solve a problem or even by
having more than one lung or kidney. A condition related
to redundancy is ‘‘resource diversity.’’ Communities that
are dependent on a narrow range of resources are less able
to cope with change that involves the depletion of that
resource, a state that is sometimes referred to as ‘‘resource
dependency’’ (Adger 2000). For example, some ecologists
(Adger 2000; Klein et al. 2003) have argued that coastal
economies are more resilient than inland economies
because they have many different functions, linkages, and
niches. Resource dependency is somewhat the opposite of
redundancy.
The remaining two properties in Bruneau et al.’s
scheme were ‘‘rapidity,’’ the capacity to achieve goals in a
timely manner to contain losses and avoid disruption, and
‘‘resourcefulness,’’ the capacity to identify problems and
mobilize resources when conditions threaten the system.
The ‘‘4Rs’’ almost work for our purposes (i.e., as attributes
of resources rather than as properties of systems) but not
quite. Resourcefulness as a property of resources is
semantically problematic, and we view critical thinking as
a specific resource, rather than as a dynamic attribute that
resources share. Thus, we revised this framework slightly
by incorporating mobilization into rapidity, redefined to
reflect how quickly the resource can be accessed and used.
Thus, in summary, we propose that resilience-resources
have three dynamic properties: robustness, redundancy,
and rapidity. We speculate that a resource need have only
one of these attributes to engender resilience, but this is an
empirical question. In a case study of the Emergency
Operations Center (EOC) after the 2001 World Trade
Center Disaster, Kendra and Wachtendork (2003) showed
134 Am J Community Psychol (2008) 41:127–150
123
how the EOC’s ability to rapidly access other resources
substituted for lack of redundancy of personnel, equipment,
and space when the EOC was destroyed in the attack.
Resource Mobilization and Deterioration
Resilience can fail when resilience-resources are them-
selves damaged or disrupted by the stressor. Kimhi and
Shamai (2004) examined perceptions of community resil-
ience across four communities that varied in their
proximity to the Israel–Lebanon border and thus varied
with respect to threat of political violence. Individuals
living in the community that had experienced the highest
level of threat for the longest period of time perceived their
community as less resilient than did individuals in the other
communities. If these perceptions are accurate, this
example suggests that the community’s abilities to cope
with the threat were themselves harmed by the threat and
thus illustrates the importance of robustness as an essential
quality of resilience-resources, as discussed above. Disas-
ters and other stressors cause both resource mobilization
and resource deterioration. Ideally, the emergent resources
are sufficient to protect or replenish the vulnerable
resources, but this not always the case.
The concept of resource loss has become central in
stress theory, primarily because of the influence of Hob-
foll’s (1988,1998,2006) theory of ‘‘conservation of
resources’’ (COR). The basic tenet of COR theory is that
‘‘individuals strive to obtain, retain, protect, and foster
those things that they value,’’ which are termed resources
(Hobfoll 2006, p. 217). In Hobfoll’s theory, stress occurs
when resources are threatened, when resources are lost, or
when individuals fail to gain resources following a sig-
nificant investment of other resources. In COR theory,
people must invest resources in order to protect against
resource loss, recover from losses, and gain resources; this
serves to make those with greater resources less vulnerable
to resource loss and more capable of resource gain (the rich
get richer, so to speak).
COR theory has become highly influential in disaster
research because disasters and terrorism threaten a host of
object resources (housing), personal resources (e.g., opti-
mism, safety), social resources (companionship), and
energies (money, free time). This phenomenon wherein
resources are themselves harmed by the stressors they are
presumed to buffer severely limits the protection resources
can afford. Scored simply as a count of losses tallied from
an inventory, resource loss has correlated highly with
symptom severity in several disaster studies (Norris et al.
2002b).
Recognition that the loss of resources is shared by
members of a highly traumatized community was most
eloquently described by Erikson (1976), in his ethnography
of the Buffalo Creek, West Virginia dam collapse that
caused heavy loss of life and massive displacement in a
small mining town. Loss of social connections was severe,
leaving survivors feeling isolated and alone (see also Ab-
ramowitz 2005). Some years later, Kaniasty and Norris
(1993) proposed and validated a more specific ‘‘social
support deterioration model,’’ which posited that the
impact of disasters on mental health is both direct and
indirect, through disruptions of social networks and
declines in perceptions of support availability. Loss of
social supports may be especially severe and long-lasting
in the context of displacement, especially for women
(Norris et al. 2005a). Many things can lead to postdisaster
declines in social support, including not only severe
stressors, such as displacement and death of significant
others, but also loss of routine opportunities for compan-
ionship and leisure, community conflicts about the causes
of the disaster and appropriate responses, and violated
expectations of aid (Edelstein and Wandersman 1987;
Kaniasty and Norris 2004; Tobin and Whiteford 2002).
Fortunately we can say that postdisaster deterioration of
social resources is not inevitable. Using longitudinal data
collected after Hurricanes Hugo (1989) and Andrew
(1992), Norris and Kaniasty (1996) proposed and validated
a model of ‘‘social support deterioration deterrence’’ in
which the positive relation between severity of exposure
and received support (resource mobilization) to a greater or
lesser extent offset or counteracted the negative relation
between severity of exposure and perceived support
(resource deterioration). Victims who received high levels
of help following the disasters were thus protected against
salient erosion in their perceptions of belonging and
expectations of support.
Resilience as a Set of Capacities
In our framework, resilience emerges from a set of adaptive
capacities—community resilience from a set of networked
adaptive capacities. This is an important point: resilience
rests on both the resources themselves and the dynamic
attributes of those resources (robustness, redundancy,
rapidity); we use the term ‘‘adaptive capacities’’ to capture
this combination. This view is generally consistent with
Gunderson’s (2000) definition of adaptive capacity as the
property of the ecosystem that describes changes in stability
landscapes and resilience. Its transformational characteris-
tics are what distinguish ‘‘community resilience’’ from
other ways of characterizing community strengths, such as
‘‘community competence’’ or ‘‘social capital’’ (Brown and
Kulig 1996/97; Pfefferbaum et al. 2005), both of which are
viewed as resources, more properly as sets of resources,
Am J Community Psychol (2008) 41:127–150 135
123
here. However, our definition comes very close to the
meaning of ‘‘community capacity,’’ a term that has been
used for some years in public health, usually as it relates to
the implementation of a public health program or practice.
Goodman et al. (1998, p. 259) defined community capacity
in two ways: ‘‘(1) the characteristics of communities that
affect their ability to identify, mobilize, and address social
and public health problems and (2) the cultivation and use
of transferable knowledge, skills, systems, and resources
that affect community- and individual-level changes con-
sistent with public health-related goals and objectives.’’ We
might say that capacities become adaptive capacities when
they are robust, redundant, or rapidly accessible and thus
able to offset a new stressor, danger, or surprise. Goodman
et al. identified several dimensions of community capacity.
We include most of their dimensions within the sets
described subsequently.
On the basis of our review of the literature, we identified
four primary sets of networked resources: Economic
Development, Social Capital, Information and Communi-
cation, and Community Competence (see Fig. 2). These
sets are far from orthogonal, but they are also far from
synonymous. Much has been written about these concepts
that are introduced here only briefly. Our purpose was not
to provide a comprehensive, nuanced summary of what is
known about each capacity but rather to draw conclusions
about the set of capacities that should become the focus of
our attention in community resilience theory, research, and
application.
Economic Development
Resource Volume and Diversity
Communities are subject to larger sociological and eco-
nomic forces. Adger (2000) developed a set of key
parameters for observing ‘‘social resilience’’ (see Table 1).
The first parameter encompassed economic growth, sta-
bility of livelihoods, and equitable distribution of income
and assets within populations. Land and raw materials,
physical capital, accessible housing, health services,
schools, and employment opportunities create the essential
resource base of a resilient community (Godschalk 2003;
Pfefferbaum et al. 2005). Because of extensive interde-
pendencies at the macroeconomic level, economic
resilience depends not only on the capacities of individual
businesses but on the capacities of all the entities that
depend on them and on which they depend (Rose 2004,
2005).
Fig. 2 Community resilience as a set of networked adaptive capacities
136 Am J Community Psychol (2008) 41:127–150
123
The case for a link between economic resources and
postdisaster wellness is most evident in research on social
class as a buffer of disaster stress. Past disaster research has
shown that participants of lower socio-economic status
(SES) often experience more adverse psychological con-
sequences than do participants of higher SES (Norris et al.
2002a,b). Recent evidence (Ahern and Galea 2006) sug-
gests that the adverse psychosocial effects of low income
(an individual-level variable) after disasters are especially
strong in the context of income inequities (a neighborhood-
level variable). By taking a global perspective, we can
make this case all the more strongly because previous
research suggests that natural disasters are especially likely
to engender severe psychological distress when they occur
in the developing world. In fact, disaster location (United
States, other developed country, developing country) was a
stronger predictor of sample-level effects than either
disaster type (mass violence, natural, technological) or
sample type (child, adult, rescue/recovery) in Norris et al.’s
(2002a,b) empirical review.
Community resilience depends not only on the volume
of economic resources but also on their diversity. Using
mangrove agriculture in Southeast Asia as an example,
Adger (2000) showed how dependency on a narrow range
of natural resources can increase variance in income and
decrease social resilience. Extreme events, such as
droughts, floods, or infestations, increase the risk of being
dependent on particular resources and therefore decrease
resilience. Cutter et al. (2006) described one community
that was especially devastated by Hurricane Katrina in
August 2005 because residents were almost totally reliant
on the shrimping industry, on which the storm’s impact
was tremendous.
Resource Equity and Social Vulnerability
Societies do not allocate environmental risk equally, often
making the poorest communities the weakest links in hazard
mitigation (Cutter et al. 2003; Godschalk 2003; Tobin and
Whiteford 2002). Differential risk is all the more striking
from a global perspective because disasters are dispropor-
tionately likely to strike economically developing or poor
countries (De Girolamo and McFarlane 1996). Wisner
(2001) argued that mitigation plans in developing countries
often fail to address the ‘‘root causes of disaster vulnerability,
namely, the economic and political marginality of much of
the population and environmental degradation (pp. 251).’’
Quarantelli (1994) argued compellingly that increasing
industrialization, expanding urbanization, and decaying
infrastructures will lead to an escalation in the numbers of
disasters worldwide and that developing countries will bear
the brunt of this trend. In developing countries, housing
quality is poor relative to that found in developed countries,
so houses are less capable of withstanding severe force.
Lacking means for obtaining other property, families often
invade flood plains and other undesirable locations. Loss of
life often numbers in the thousands or tens of thousands when
major hurricanes or earthquakes strike poor countries. The
southeast Asian tsunami of December 26, 2004 caused an
unfathomable 276,000 deaths, according to a U.S. Geologi-
cal Survey (retrieved October 21, 2006 from http://
www.msnbc.msn.com/id/6948775).
Poor communities not only are at greater risk for death
and severe damage, but they often are less successful in
mobilizing support after disasters. In fact, disaster-stricken
communities are not ruled in the most egalitarian way
(Bolin and Bolton 1986; Beggs and Haines 1996; Kaniasty
and Norris 1995; Tobin and Whiteford 2002). Ideally, the
distribution or mobilization of support follows the ‘‘rule of
relative needs.’’ Simply put, the most support goes to those
who need it the most. Often, however, the distribution of
support follows the ‘‘rule of relative advantage’’ because
one’s embeddedness in the community, political connec-
tions, and social class determine the availability and
accessibility of resources (Kaniasty and Norris 1995,
2004). While these rules typically describe the distribution
of postdisaster support within communities, they also
describe the distribution of support across communities
(Norris et al. 2001). The capacity to distribute postdisaster
resources to those who most need them seems vitally
important for community resilience.
Social Capital
For developing a theory of community resilience, a highly
relevant theme is social capital. Like resilience, social
capital is a concept transferred from one discipline (in this
case, economics) to another (Bourdieu 1985). Also like
resilience, a variety of definitions and criticisms of those
definitions have appeared, which Kadushin (2004) nicely
summarized in his review of the concept. The basic idea of
social capital is that individuals invest, access, and use
resources embedded in social networks to gain returns (Lin
2001). It can also be defined as the aggregate of the actual
or potential resources that are linked to possession of a
durable network of relationships (Bourdieu 1985). Social
capital theorists debate the roles of self-interest and status
attainment and whether social capital should be conceived
as an individual, collective, or multi-level asset (Wellman
and Frank 2001). Theorists have also debated the extent to
which people actively aim to increase their social capital
(through investment) or whether, conversely, it arises from
structural positions, families, and friendships (Kadushin
2004).
Am J Community Psychol (2008) 41:127–150 137
123
Network Structures and Linkages
Quite a bit has been written about network structures,
which we group here under the umbrella of social capital.
One dimension of community capacity, according to
Goodman et al. (1998), is the presence of inter-organiza-
tional networks that are characterized by reciprocal links,
frequent supportive interactions, overlap with other net-
works, the ability to form new associations, and
cooperative decision-making processes. Comfort (2005)
noted that uncertainty often leads to efforts to broaden the
‘‘scope of actors, agents, and knowledge that can be mar-
shaled’’ (p. 347). More generally, this trend necessitates
networked as opposed to hierarchical systems for disaster
response. Longstaff (2005) highlighted the importance of
‘‘keystones’’ or ‘‘hubs,’’—‘‘super-connected’’ network
members who link one network to another (see also
Fullilove and Saul 2006). Despite the many values of dense
networks, they are also more complex and therefore more
uncertain. The efficiency of hubs may actually decrease
resilience because if the hub is compromised, the entire
system fails (Allenby and Fink 2005). Frequent flyers can
probably appreciate this point. Systems will be highly
vulnerable if there is little redundancy for these connective
functions.
Longstaff also noted the tendency to want individuals,
groups, and organizations to come together tightly to resist
danger. ‘‘Tight coupling’’ occurs when change in one
component engenders a response from other components.
This is not always bad, but tight coupling can also increase
danger in some circumstances and can lead to premature
convergence on solutions. ‘‘Loosely coupled’’ systems may
be better at responding to local changes since any change
they make does not require the whole system to respond
(Longstaff 2005).
The happy medium may be loosely coupled organiza-
tions (to better respond to local needs) that are able to
coordinate or collaborate (to facilitate access to their
resources). Gillespie and Murty (1994) noted that the
failure of relief organizations to work together results in
‘‘cracks’’ in the postdisaster service delivery network,
whereas an effective service delivery system provides a
complete set of services and linkages in which such cracks
do not appear. Gillespie and Murty distinguished between
two types of linkage cracks: ‘‘Isolates’’ are organizations
with essential services that do not interact with the rest of
the network. They impair service delivery because their
services are inaccessible to potential consumers who enter
the network elsewhere. ‘‘Peripherals’’ are organizations
that are connected to the network only by means of a long
and indirect series of links. They create inefficiency as a
great deal of time and effort may be required to access the
services they provide, and they are at risk of becoming
isolates if any one link is impaired. Gillespie and Murty
studied a disaster response network in a Midwestern setting
and identified both isolates and peripherals among the
organizations that had capacities and experience that were
essential for the functioning of the network as a whole.
Social Support
Social support refers to social interactions that provide
individuals with actual assistance and embed them into a
web of social relationships perceived to be loving, caring,
and readily available in times of need (Barrera 1986).
Research on social support (see Kaniasty and Norris 2004)
has long distinguished between ‘‘received support’’ (actual
receipt of help) and ‘‘perceived support’’ (the belief that
help would be available if needed). Social support, whether
received (enacted) or perceived (expected), varies on two
critical dimensions (Kaniasty and Norris 2000). The first
dimension, source, is reflected in the overall pattern of help
utilization. This pattern resembles a pyramid with its broad
foundation being the family, followed by other primary
support groups, such as friends, neighbors, and co-workers,
followed by formal agencies and other persons outside of
the victim’s immediate circle. The second dimension, type,
differentiates between emotional, informational, and tan-
gible support.
Received support typically shows a mobilization pattern
by increasing in the aftermath of disasters and correlating
positively with severity of exposure. It protects against
erosion of perceived support, which in turn is a powerful
protective factor for mental health (Norris and Kaniasty
1996). Received support is most helpful when it is reci-
procal, meaning that there is a balance between receiving
and providing support (Hogan et al. 2002; Maton 1988;
Rook 1992). Being constantly on the receiving end of
support exchanges can threaten self-esteem (Dunkell-
Schetter and Bennett 1990), whereas being constantly on
the providing end creates stress and burden (Solomon et al.
1993).
Another critical function of social support is social
influence. In emergencies, people look to similar others to
help them make decisions about appropriate behaviors.
This idea, often characterized as ‘‘emergent norms,’’ is
among the oldest to be found in the sociology of disasters
(Fritz and Williams 1957). For example, the greater one’s
social ties, the more likely one is to receive information
about recommendations to evacuate. Evacuation is often
the only available strategy to save lives and reduce per-
sonal injuries. In an analysis of evacuation decisions before
Hurricanes Hugo and Andrew, Riad et al. (1999) found that
residents with stronger social support were twice as likely
to evacuate as were residents with weaker social support.
138 Am J Community Psychol (2008) 41:127–150
123
The important dimension was perceived support (e.g.,
ability to borrow money, get a ride, have a place to stay),
not merely the number of ties.
Previous discussions have likewise emphasized the
importance of social support for community resilience. One
of Pfefferbaum et al.’s (2005) factors of community resil-
ience was Support and Nurturance, referring to the
community’s care about the needs and well-being of
members. Tse and Liew (2004) and Ganor and Ben-Lavy
(2003) emphasized self-help and mutual support (‘‘cohe-
sion’’) as components. Goodman et al. (1998) similarly
proposed that social relationships, including both the fre-
quency and intensity of interactions (social embeddedness)
and the benefits members receive from their social ties
(received and perceived support), were dimensions of
community capacity.
Community Bonds, Roots, and Commitments
For the most part, social support captures helping behaviors
within family and friendship networks, but social capital
also encompasses relationships between individuals and
their larger neighborhoods and communities (Perkins et al.
2002; Perkins and Long 2002; Saegert and Winkel 2004).
Three key social psychological dimensions of social capital
are thus sense of community, place attachment, and citizen
participation.
‘‘Sense of community’’ is an attitude of bonding (trust
and belonging) with other members of one’s group or
locale (Perkins et al. 2002, p. 37), including mutual con-
cerns and shared values. Sense of community,
characterized by high concern for community issues,
respect for and service to others, sense of connection, and
needs fulfillment, is assumed to be a dimension of com-
munity capacity (Goodman et al. 1998, p. 261). It is also
believed to be an attribute of resilient communities (Ahmed
et al. 2004; Landau and Saul 2004; Pfefferbaum et al.
2005; Tse and Liew 2004). As noted earlier, disasters often
disrupt sense of community (e.g., Abramowitz 2005;
Erikson 1976; Kaniasty and Norris 2004). However, in
some circumstances, environmental threats can enhance
survivors’ sense of similarly and interdependence, leading
to increased sense of community (Edelstein 1988).
‘‘Place attachment’’ is closely related to one’s sense of
community (Tartaglia 2006). It implies an emotional con-
nection to one’s neighborhood or city, somewhat apart
from connections to the specific people who live there
(Altman and Low 1992; Manzo and Perkins 2006). Brown
and Perkins (1992) argued that place attachments are
integral to self-definitions; these attachments are holistic
and multi-faceted and provide stability. Place attachment
often underlies citizens’ efforts to revitalize a community
(Perkins et al. 2002) and thus may be essential for com-
munity resilience.
Place attachment may be of special note for disaster
recovery, as these events have spatial parameters and harm
built and natural environments. In the worst of cases,
people are displaced from neighborhoods and communities
in which they are deeply rooted. As Brown and Perkins
discussed (1992), disruptions in place attachments threaten
both individual and communal aspects of self-definitions,
and stronger attachments make such disruptions more
devastating. The impact of displacement after disasters has
often been profoundly adverse (e.g., Erikson 1976; Norris
et al. 2005b; Oliver-Smith 1986), and similar findings are
highly likely to emerge in research on Hurricane Katrina
(e.g., VanLandingham et al. 2007). These findings raise the
possibility that place attachment could, in some circum-
stances, impair rather than facilitate resilience. However, it
is simultaneously plausible that place attachments promote
healing (Cox and Holmes 2000) and increase the likelihood
that the community as a whole has the will to rebuild
(Manzo and Perkins 2006).
‘‘Citizen participation’’ is the engagement of community
members in formal organizations, including religious
congregations, school and resident associations, neighbor-
hood watches, and self-help groups (Perkins et al. 2002;
Wandersman 2000). Empowering community settings are
characterized by inspired, committed leadership and by
opportunities for members to play meaningful roles (Maton
and Salem 1995). Wandsmerman and Florin (2000) sum-
marized three primary areas of research on citizen
participation—who participates and why, how organiza-
tional characteristics influence participation, and the effects
of participation on community conditions and participants’
own feelings of efficacy—that provide a sound framework
for examining grass-roots participation in disaster readiness
and recovery efforts. Quarantelli (1989) summarized the
results of a series of observational studies of citizen groups
that emerge with respect to hazardous waste sites, noting
that they tend to have a small active core, a larger sup-
porting circle that can be mobilized for specific tasks, and a
greater number of nominal members. Most groups exist in
conflict. Their goals are often vague and lofty at the outset
but become more specific and achievable over time. These
findings relate to the broader aim of understanding the ‘‘life
cycle’’ of community organizations that form to address
environmental threats (Edelstein and Wandersman 1987).
Edelstein (1988) observed that leaders of citizen groups are
usually those who have strongest attachments to place.
Citizen participation is widely believed to be a funda-
mental element for community resilience. Two of the
factors hypothesized by Pfefferbaum et al. (2005) fall into
this set of capacities: Participation, referring to member
involvement and engagement and opportunities for such
Am J Community Psychol (2008) 41:127–150 139
123
that are sensitive to the diversity, ability, and interests of
members; and Structures, Roles, and Responsibilities,
referring to leadership, teamwork, clear organizational
structures, well-defined roles, and management of rela-
tionships with other communities. Similarly, Ganor and
Ben-Lavy (2003) argued that community resilience
requires authentic, grass-roots leadership, which they
labeled ‘‘credibility’’ in their scheme. Participation and
Leadership were also two of the dimensions of community
capacity described by Goodman et al. (1998).
Information and Communication
Information may be the primary resource in technical and
organizational systems that enables adaptive performance
(Comfort 2005). By communication, we refer to the crea-
tion of common meanings and understandings and the
provision of opportunities for members to articulate needs,
views, and attitudes. Pfefferbaum et al. (2005), Goodman
et al. (1998), and Ganor and Ben-Lavy (2003) have all
argued that good communication is essential for commu-
nity resilience or capacity.
Systems and Infrastructure for Informing the Public
Information and communication become vital in emergen-
cies. People need accurate information about the danger and
behavioral options, and they need it quickly. Public adher-
ence to recommendations cannot be taken for granted,
particularly when there is marked uncertainty about expo-
sure, consequences of exposure, or the risks involved with
following the recommendations (Reissman et al. 2005).
After an anthrax attack (to use Reissman et al.’s example),
people are neither certain they have been exposed nor certain
that prophylactic use of antibiotics can help them. How to
communicate risks and recommendations most effectively to
the public has been the focus of much past research, but ‘‘the
link between information provision and preparedness
remains tenuous’’ (Paton and Johnston 2001, p. 271).
On the basis of her review, Longstaff (2005, p. 55)
argued that information increases survival only if is ‘‘cor-
rect and correctly transmitted.’’ In emergencies, when there
is little time to check information, it is also important that
the sender of the information be trusted. Closer, local
sources of information are more likely to be relied upon
than unfamiliar, distant sources. In fact, Longstaff con-
cluded, ‘‘A trusted source of information is the most
important resilience asset that any individual or group can
have’’ (emphasis in original, p. 62). Similarly, the Working
Group on Governance Dilemmas (2004) concluded that
trusted communication treats the public as a capable ally,
invests in public outreach, and reflects the values and pri-
orities of local populations.
Communication infrastructure is also a valuable
resource. On the basis of their experiences in New York
City after the September 11th terrorist attacks, Draper et al.
(2006) maintained that it is advantageous for a life-line (or
hotline) system to be in place beforehand. These commu-
nication systems can be ramped up after the disaster to
coordinate and deploy volunteers, and later they provide a
central means for the pubic to learn about and access ser-
vices (see also Norris et al. 2006). Media also can be
engaged to publicize available services and educate the
public about typical reactions to disaster (e.g., Gist and
Stolz 1982; Norris et al. 2006).
Communication and Narrative
The remaining element in this set, less structural than the
others, is the presence of communal narratives that give the
experience shared meaning and purpose. As noted by Sonn
and Fisher (1998), who examined community resilience to
oppression, narratives provide insight into how communi-
ties see themselves and others (see also Harvey et al. 1995;
Rappaport 1995; Waller 2001). Members’ shared under-
standings of reality contribute to a sense of place and
connectedness, that in turn affect resilience (Alkon 2004).
Couto (1989) described how ‘‘group formulations’’ (nar-
ratives and symbols) became a mechanism for
empowerment in Aberfan, South Wales, after a horrific
environmental disaster took the lives of 104 school-chil-
dren and 20 adults. Writing about their own experiences in
the aftermath of the September 11th terrorist attacks in
lower Manhattan, Landau and Saul (2004) concluded that
community recovery depends partly on collectively telling
the story of the community’s experience and response.
The media shape how a disaster is framed in ways that
influence survivors’ and others’ understanding of the event,
including emergency managers. In a fascinating analysis of
disaster metaphors and myths, Tierney et al. (2006) traced
the impact of exaggerated and extreme portrayals of loot-
ing and lawlessness to critical policy decisions. Three days
after the New Orleans levee breach, leadership’s decision
to re-direct police officers to attend to lawbreakers rather
than to life-saving activities lessened the survival chances
of stranded residents. The authors also noted that stories
like these lead many policy-makers to believe that the
military should play a larger role in disaster response
despite 50 years of sociological research showing that the
emergency behavior of most disaster victims is orderly and
prosocial.
Abramowitz (2005) conducted an extraordinary anthro-
pological study of six Guinean communities attacked by
140 Am J Community Psychol (2008) 41:127–150
123
Sierra Leonean and Liberian forces in 2000–2001. The
author collected narratives from study participants as well
as measures of psychological distress. Symptoms of post-
traumatic stress were nearly universal but much higher in
three of the communities than in the others. Narratives
revealed that in those three communities, respondents
shared the feeling that government and non-governmental
organizations had neglected them. They mourned the loss
of their homes and markets. Social rituals and practices,
including reciprocity and charity, were abandoned. There
were widespread beliefs that some community members
had prospered at the expense of others. Distress was
present in the other communities, but less so. They had
managed to maintain many social rituals, and residents
shared a belief that customs and social practices would
return to normal as soon as economic conditions improved.
Most importantly, they had created a collective story that
emphasized their resistance to the violence.
Community Competence
Endangered communities must be able to learn about their
risks and options and work together flexibly and creatively
to solve problems. Longstaff (2005) argued that the
capacity to acquire trusted and accurate information, to
reflect on that information critically, and to solve emerging
problems is far more important for community resilience
than is a detailed security plan that rarely foresees all
contingencies (see also Comfort 2005; Handmer and
Dovers 1996). This point brings us to Community Com-
petence, which we view as the networked equivalent of
human agency, a term that is more broadly applicable
across levels of analysis. Community competence has to do
with collective action and decision-making, capacities that
may stem from collective efficacy and empowerment.
Collective Action and Decision-making
Cottrell (1976, p. 197) described a competent community
as one in which ‘‘the various component parts of the
community: (1) are able to collaborate effectively in
identifying the problems and needs of the community; (2)
can achieve a working consensus on goals and priorities;
(3) can agree on ways and means to implement the agreed
upon goals; and (4) can collaborate effectively in the
required actions.’’ Cottrell proposed that these competen-
cies arose from commitment to the community as a
relationship worthy of substantial effort, articulateness,
communication, participation, and means for debate, dis-
cussion, and decision-making. Many of these conditions
were previously considered as aspects of Social Capital or
Information and Communication, so it might be said that
social capital and communication are prerequisites for
community competence (see Fig. 2).
Drawing heavily upon earlier writings about community
competence, Brown and Kulig (1996/97, p. 30) argued
effectively that ‘‘the concept of resiliency in the context of
communities needs to be grounded in a notion of human
agency, understood in the sense of the capacity for mean-
ingful, intentional action. Individuals and collectives are
resilient in the first sense insofar as they act in such a way
as to recover from what they define as negative physical or
social events. Individuals and collectives are resilient in a
second sense insofar as they act to transform their physical
and social environments to mitigate against such events in
the future.’’ Brown and Kulig were making the point that
resilience is not simply a passive ‘‘bouncing back’’ (as
implied by the original metaphor) because people can
imagine how things might be and do things to bring those
conditions about. Perez-Sales et al. (2005) made similar
points from having observed the capacity of shelter resi-
dents to control their own lives after a severe earthquake in
El Salvador in 2001.
The skills identified by Goodman et al. (1998) included
several that are essential elements of community compe-
tence: abilities to engage constructively in group process,
resolve conflicts, collect and analyze data, and resist
opposing or undesirable influences (p. 261). There appears
to be high consensus that critical reflection and problem
solving are fundamental capacities for community com-
petence and resilience (Bruneau et al. 2003; Goodman
et al. 1998; Klein et al. 2003; Pfefferbaum et al. 2005).
Ganor and Ben-Lavy (2003) emphasized the importance of
the community’s ability to take action, which they called
‘‘coping.’’
Collective action is complex and challenging in the face
of environmental threats. Such threats, and concomitant
disappointment in political leaders, often lead to creation of
grass roots groups that represent affected families (Edel-
stein 1988; Oliver-Smith 1986). Sometimes, these groups
are effective in building consensus and mobilizing political
action and legislation, but other times, action is impeded by
mistrust, conflict, or ‘‘dissensus’’ (Edelstein and Wan-
dersman 1987; Kaniasty and Norris 2004). Which outcome
occurs is not necessarily predictable but may depend on the
extent of collective efficacy, the process of empowerment
(vs. disempowerment), and the nature of political interac-
tion (adversarial versus collaborative).
Collective Efficacy and Empowerment
‘‘Collective efficacy’’ reflects trust in the effectiveness of
organized community action (Perkins and Long 2002).
Am J Community Psychol (2008) 41:127–150 141
123
Sampson et al. (1997) defined collective efficacy as a
composite of mutual trust and shared willingness to work
for the common good of a neighborhood. In our model of
community resilience, collective efficacy bridges the
domains of Social Capital and Community Competence,
but is placed under the latter because of its fundamental
role in facilitating community action. A related concept is
communal mastery, defined as the sense that individuals
can overcome life challenges and obstacles through and
because of their being interwoven in a close social network
(Hobfoll et al. 2002). As one of their ‘‘six Cs’’ of com-
munity resilience, Ganor and Ben-Lavy (2003, p. 106)
included ‘‘credo,’’ ‘‘the vision of a community, one that
depicts a better future, a horizon of hope.’’ This idea is
similar to that of community hope (Ahmed et al. 2004).
Paton and Johnston (2001) proposed that an initial focus on
promoting collective efficacy would increase the likelihood
of achieving success in working with a community to adopt
mitigation strategies.
Benight (2004) offered a slightly different definition of
collective efficacy as the shared belief that a group can
effectively meet environmental demands and improve their
lives through concerted effort. In this study, collective
efficacy (as perceived by the individual) interacted with
resource loss after a flood to predict recovery from symp-
toms of posttraumatic stress. Persons with high collective
efficacy were less adversely affected by their losses than
were persons with low collective efficacy. In discussing his
findings, Benight noted that the people he studied had
responded successfully to a variety of problems after the
floods by creating an organized crisis committee to speak,
decide, and act on behalf of their small rural community.
Collective efficacy is highly related to empowerment
(Perkins et al. 2002), a process through which people
lacking an equal share of valued resources gain greater
access to and control over those resources (Rappaport
1995). A particular relevant discussion of empowerment
for our purposes was presented by Rich et al. (1995), who
examined the dynamics of community empowerment after
discovery of environmental hazards. Rich et al. reasoned
that meaningful participation in environmental action
groups can be empowering and, conversely, that lack of
opportunity for such participation can be disempowering.
The effectiveness of a community’s response to a hazard is
shaped by a combination of resources (such as sufficient
education to understand the technical issues, money to hire
lawyers or scientific advisors), a culture that permits
challenges to authority, institutions that provide a basis for
coordinating a response, and political mechanisms that
involve citizens in decision-making. In their model,
empowerment progresses through a sequence of formal
empowerment (mechanisms for citizen input), intraper-
sonal empowerment (feelings of personal competence and
confidence), instrumental empowerment (ability to partic-
ipate in and influence decision, as determined by
knowledge, material resources, and persuasiveness), and
substantive empowerment (ability to reach decisions that
solve problems). The last of these stages is virtually syn-
onymous with our broader meaning of Community
Competence.
Rich et al. (1995) furthermore noted that some political
approaches to citizen involvement in decisions are more
empowering than others. Whereas adversarial models grant
persons the right to express concerns and advocate for their
interests, they generally put citizens in the position of
reacting to proposals and may require legal action that
community groups can rarely afford. The adversarial model
may work against efforts to create consensus. The sug-
gested alternative is a partnership approach that ‘‘invites
proactive, not just reactive, thought that may produce
creative alternatives…and allows communities to move
toward sustainable environmental decisions (pp. 671).’’
The Network of Adaptive Capacities
Figure 2summarizes this section by showing the four
primary sets of networked resources that generate com-
munity resilience. There are innumerable possible linkages
between elements in these sets (and between these ele-
ments and population wellness) that could be researched
empirically across communities, a point to which we return
in the final section of this paper. The utility of the frame-
work could also be examined by means of qualitative case
studies of a particular community and particular response.
To recap, these resources qualify as resilience-resources
(adaptive capacities) to the extent they are robust, redun-
dant, or rapidly accessible.
Resilience as a Strategy for Disaster Readiness
Fueled by incomprehensible disasters, such as the Sep-
tember 11, 2001 terrorist attacks, the southeast Asian
tsunami of December 26, 2004, and Hurricane Katrina on
August 29, 2005, efforts to improve disaster preparedness
and response have increased dramatically since the turn of
the century. However, the trend toward greater prepared-
ness in the United States was already in play. The Disaster
Mitigation Act of 2000 gave new emphasis to proactive
mitigation and local all-hazards plans. Discussions of the
challenges and complexities of disaster management fre-
quently allude to community resilience as an important
ingredient. Some authors have speculated about the role of
community resilience in local preparedness; others have
advised how emergency management infrastructure can be
142 Am J Community Psychol (2008) 41:127–150
123
made more resilient; and others have illustrated how an
understanding and appreciation of community resilience
can guide the design of postdisaster psychosocial inter-
ventions. Community resilience may have particular value
in countering the impact of terrorism (Reissman et al.
2005).
The adaptive capacities illustrated in Fig. 2provide a
roadmap for enhancing community resilience to disasters.
This is perhaps more like a rotary than a highway, as one
can enter and exit anywhere. Nonetheless, we describe five
stops along this road that are likely to be necessary for most
travelers, although other stops undoubtedly could be made
as well.
First, to increase their resilience to disaster, communi-
ties must develop economic resources, reduce risk and
resource inequities, and attend to their areas of greatest
social vulnerability. Hazard risk is neither randomly nor
evenly distributed (Cutter et al. 2003). To mitigate social
vulnerability to urban hazards, Godschalk (2003) recom-
mended that cities set aside resources to make poor
neighborhoods safer. To accomplish this, city staff need to
work together with each neighborhood to identify its needs
and an appropriate mitigation approach. Godschalk (p. 140)
also recommended that urban hazard mitigation activities
be integrated with activities related to economic develop-
ment and social justice, thereby ‘‘achieving the multiple
objectives needed for a resilient system.’’ Efforts to create
economic diversity increase the probability that the com-
munity can withstand adversity or surprise. After a disaster,
residents and grass-roots leaders should be vigilant to the
equity of resource distribution. Those who need support the
most may have the least access to it, and there is extensive
evidence that many persons are excluded from emergent
‘‘altruistic communities’’ (Kaniasty and Norris 1995,2004).
Competent communities may influence these dynamics
through creative problem-solving and political action.
Second, to access social capital, one of the primary
resources of any community, local people must be engaged
meaningfully in every step of the mitigation process.
Enabled by professional practitioners, as necessary, com-
munity members must assess and address their own
vulnerabilities to hazards, identify and invest in their own
networks of assistance and information, and enhance their
own capacities to solve problems created by known or
unknown unknowns (e.g., Brown and Kulig 1996/97; Coles
and Buckle 2004; Longstaff 2005; Pfefferbaum et al.
2005). Non-indigenous practitioners best foster recovery
by providing settings and resources that allow the com-
munity to take charge of the direction of change (Fullilove
and Saul 2006; Landau and Saul 2004; Perez-Sales et al.
2005; van den Eynde and Veno 1999). Because trauma
emanates from profound powerlessness, interventions
should emphasize empowerment, meaning they need to
emphasize strengths, mobilize the community’s capabili-
ties, and help the community to become self-sufficient
(Harvey 1996). It is critical to make use of existing
resources that can be affirmed and integrated into the
response plan.
Third, pre-existing organizational networks and rela-
tionships are the key to rapidly mobilizing emergency and
ongoing support services for disaster survivors. Loosely
coupled but cooperative systems appear to provide the best
combination of linkages and flexibility (Gillespie and
Murty 1994; Longstaff 2005). A series of case studies of
mental health system responses to major disasters,
including the Oklahoma City bombing (Norris et al.
2005b), the September 11th terrorist attacks (Norris et al.
2006), and many others (Elrod et al. 2006) repeatedly
revealed that developing organizational networks, coali-
tions, and cooperative agreements ahead of time is crucial,
and that organizational plans should indicate how key
constituencies will be involved. Program directors relied
upon pre-existing relationships perhaps more than any
other single resource to implement programs quickly. To
work together after a disaster, systems must understand and
trust each other, which is challenging if they have not
worked together before. For example, organizations
mounting community outreach programs almost inevitably
experience difficulty in entering school systems if they
have not conducted programs with schools previously. This
barrier may greatly interfere with programs’ abilities to
reach children in the community, who appear to be a group
in need of additional support in the aftermath of major
disasters (Norris et al. 2002a,b).
Fourth, interventions are needed that boost and protect
naturally-occurring social supports in the aftermath of
disasters. Fostering natural supports helps to ensure that
communities and families retain the capacity to exchange
emotional and instrumental support (Landau and Saul
2004). Furthermore, well-functioning social networks keep
members informed about one another’s relative needs and
may improve equity of resource distribution. Although
helping behavior and cohesion are abundant initially, they
do not last, and certain social resources are not robust to the
impact of disasters (Kaniasty and Norris 2004). Social
support interventions are most effective when they build
social skills and mutual support (Hogan et al. 2002). Ide-
ally, postdisaster support interventions furnish participants
with knowledge, attitudes, and skills that can be used to
recruit their own supports (Gottlieb 1996; Layne et al.
2001).
Fifth, communities must plan, but they must also plan
for not having a plan; this means that communities must
exercise flexibility and focus on building effective and
trusted information and communication resources that
function in the face of unknowns. Uncertainty is almost
Am J Community Psychol (2008) 41:127–150 143
123
certain to exist after disasters. The most adaptive disaster
management strategy is one that acknowledges complexity
and uncertainty and relies on timely and trusted sources of
information for rapid decision-making as opposed to rigid
plans and command-and-control strategies (Longstaff
2005). In contrast to command and control styles, problem-
solving approaches allow for innovation and localized
variations in response. Similarly, Godschalk (2003) envi-
sioned that ‘‘the public and private organization of a
resilient city would both plan ahead and act spontane-
ously...They would eschew simple command and control
leadership, preferring to develop networks of leadership
and initiative. They would set goals and objectives, but be
prepared to adapt these in light of new information and
learning’’ (p. 139). In our age of advanced technology,
flexibility remains a simple but nonetheless essential
resource for managing a disaster response (Coles and
Buckle 2004; Handmer and Dovers 1996; Klein et al.
2003; Norris et al. 2005b).
Summary, Conclusions, Caveats, and One More
Metaphor
We have argued, as have others, that resilience is a process
that leads to adaptation, not an outcome, not stability. We
emphasized the likelihood that stress and crisis induce
transient periods of dysfunction, but adopted wellness as
the eventual outcome of interest, the manifestation of
adaptation to an altered environment. This choice ties the
concept of resilience to the traditional concerns of the
public health and mental health fields. Wellness, as defined
for individuals, incorporates the criterion of being free of
psychological disorder but is explicitly more than this. Far
too often, psychology papers begin by saying that we know
a lot about psychopathology but little about resilience and
then operationalize resilience as the lack of psychopa-
thology. To say, ‘‘social support promotes resilience’’ is not
very different from saying ‘‘social support prevents psy-
chopathology’’ if resilience is merely the lack of
psychopathology. An individual’s successful adaptation,
we believe, must be reflected also in healthy patterns of
behavior, adequate role functioning, and satisfactory
quality of life. Well communities show not only high
average levels of these conditions but limited disparities in
mental and behavioral health between rich and poor, young
and old, White and not, men and women, and so forth.
We furthermore proposed that resilience and wellness
emerge from a variety of adaptive capacities, which we
defined as resources with dynamic attributes, specifically
robustness, redundancy, and rapidity. By posing this defi-
nition, we aimed to integrate resilience perspectives with
evidence showing that resources are not static—they
evolve, strengthen, weaken, and rebound—and these tra-
jectories are of interest in their own right. Reaching a better
understanding of the impact of disasters on community
resources may be the most critical and complex challenge
for future research. It should be readily apparent from
Fig. 2that the network of adaptive capacities that yields
community resilience is not a singular condition that can be
measured or monitored simply. It is a set of sets with many
dynamic attributes and transactional linkages and rela-
tionships, far more than are shown in our diagram.
Moreover, political, economic, and natural forces operating
at larger ecological levels undoubtedly influence these
capacities that operate at the community level. That said,
the capacities identified here are all potentially observable
by using a mixture of quantitative and qualitative meth-
odologies, and recent work is advancing our ability to
assess various factors that influence or reflect community
resilience (Bruneau et al. 2003; Centre for Community
Enterprise 2000; Gibbon et al. 2002; Pfefferbaum et al.
2005; Rose 2004).
Our primary hope is to foster creative thinking about how
various pathways between Economic Development, Social
Capital, Information and Communication, and Community
Competence shape disaster readiness and recovery: Which
of these resources are most likely to be robust, meaning they
are strong and able to withstand the impact of a major
disaster? Which may be substitutable for others, thereby
building redundancy? Which resources can be mobilized
rapidly in the aftermath of disaster if they are not present or
strong in advance? Does diversity of economic resources
make equity of resource distribution more likely or maybe
less necessary? When social capital via informal ties is
strong, is social capital via formal ties less important? Is
there any resource that, taken alone, is necessary (likely) or
sufficient (unlikely) for resilience? Can disasters create new
roles and leaders and galvanize community competence?
Can community competence then foster equity of resource
distribution that in turn boosts social support? Can com-
munities work with trusted sources of information and
media to control discourse, counteract myths, and create
hopeful narratives about the event? How do these psycho-
social dimensions relate to social vulnerability, as it has
been measured objectively by geographers and social sci-
entists, or to demographic change in disaster-stricken areas?
Of course, there is then the question of how strongly the
various adaptive capacities in this network contribute to the
wellness of constituent populations. To date research on the
outcomes of community resilience is meager. Some studies
have examined how individual-level perceptions of com-
munity resilience (Kimhi and Shamai 2004; Pooley et al.
2006), sense of community (Paton et al. 2001), or collec-
tive efficacy (Benight 2004) correlate with individual-level
outcomes, but no study, to our knowledge, has truly
144 Am J Community Psychol (2008) 41:127–150
123
examined how independently assessed community resour-
ces influence the postdisaster wellness of constituent
populations. This is not a simple matter, but methods exist
that can facilitate multi-level epidemiologic research
(Kawachi and Subramanian 2006; Perkins and Taylor
1996).
The ‘‘prevention paradox’’ (Rose 1981,2001) is extre-
mely important for future judgments regarding the relative
influence (and significance for policy) of individual and
community resilience-resources. As we noted earlier,
effects that seem small in analyses of individuals may be
quite large when extrapolated to populations. Traditional
risk (or protective) factor research is almost assured to find
that individual-level resilience-resources (e.g., self-effi-
cacy) have stronger effects within a study populat